RS 17 Human Adaptation in the Grand Marais - University of Arkansas
RS 17 Human Adaptation in the Grand Marais - University of Arkansas
RS 17 Human Adaptation in the Grand Marais - University of Arkansas
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
ARKANSAS ARCHEOLOGICAL SURVEY RESEARCH SERIES <strong>17</strong><br />
W. Fredrick Limp, Series Editor<br />
Fayetteville, <strong>Arkansas</strong><br />
1982
Library <strong>of</strong> Congress #82-620028<br />
Pr<strong>in</strong>ted <strong>in</strong> <strong>the</strong> U.S.A. at Fayetteville, <strong>Arkansas</strong><br />
Production Editor: Mary Lynn Kennedy<br />
Graphic Artist: Jane Kellett<br />
Photographer: Pamela Ashford<br />
Typesetter: Louise Mull<strong>in</strong>s<br />
2006 Digital Repr<strong>in</strong>t: L<strong>in</strong>di J. Holmes<br />
Report submitted to<br />
U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers<br />
Vicksburg District
AbstrAct<br />
The Felsenthal Project is a multistage <strong>in</strong>terdiscipl<strong>in</strong>ary research project carried out<br />
by <strong>the</strong> <strong>Arkansas</strong> Archeological Survey for <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg<br />
District. The designated project area is <strong>the</strong> Felsenthal Navigation Pool, a portion <strong>of</strong> <strong>the</strong><br />
Ouachita and lower Sal<strong>in</strong>e River floodpla<strong>in</strong>s below 65 feet elevation. A broad based research<br />
design was developed prior to fieldwork and emphasized four research doma<strong>in</strong>s:<br />
settlement characteristics, locational characteristics, culture history, and site transformations.<br />
Eight hypo<strong>the</strong>ses relevant to regional prehistory and history are subsumed by<br />
<strong>the</strong>se doma<strong>in</strong>s.<br />
The field portion <strong>of</strong> <strong>the</strong> project <strong>in</strong>cluded overlapp<strong>in</strong>g <strong>in</strong>tensive site survey and test<strong>in</strong>g<br />
stages <strong>in</strong> late 1979; <strong>the</strong>se stages were followed by <strong>in</strong>tensive analysis at <strong>the</strong> Survey’s<br />
Contract Laboratory and by various specialists’ studies <strong>in</strong> 1980.<br />
The project area falls with<strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland, which is closely related<br />
geohydrologically and ecologically to <strong>the</strong> Lower Mississippi Alluvial Valley. The floodpla<strong>in</strong><br />
here is an overflow bottomland consist<strong>in</strong>g <strong>of</strong> an <strong>in</strong>tricate mosaic <strong>of</strong> swamp and<br />
woodland forest, river channels, open lakes, bayous, and backswamp dra<strong>in</strong>ages. The<br />
Ouachita River is <strong>the</strong> lowest elevation <strong>in</strong> <strong>the</strong> state <strong>of</strong> <strong>Arkansas</strong> at <strong>the</strong> Louisiana l<strong>in</strong>e (6<br />
km south <strong>of</strong> <strong>the</strong> project area), and <strong>the</strong> prehistoric and historic sites located and studied<br />
<strong>in</strong> 1979 are <strong>the</strong> lowest known sites presently recorded <strong>in</strong> <strong>Arkansas</strong>.<br />
Our environmental and cultural data <strong>in</strong>dicate that <strong>the</strong> floodpla<strong>in</strong> conta<strong>in</strong>ed abundant<br />
and diverse aquatic and riparian resources <strong>of</strong> fundamental importance to prehistoric<br />
populations aggregated on adjacent upland terraces. A dist<strong>in</strong>ctive characteristic <strong>of</strong><br />
<strong>the</strong> floodpla<strong>in</strong> ecosystem is <strong>the</strong> annual prolonged hydroperiod, or extensive deep submergence<br />
<strong>of</strong> lowlands dur<strong>in</strong>g w<strong>in</strong>ter and spr<strong>in</strong>g. This hydroperiod “pulse” <strong>in</strong>troduces<br />
and cycles nutrients to floodpla<strong>in</strong> habitats, and imposes a strong seasonal dimension on<br />
biological (<strong>in</strong>clud<strong>in</strong>g human) activity. Dur<strong>in</strong>g <strong>the</strong> historic period, Euramericans never<br />
permanently settled or farmed <strong>the</strong> overflow bottomland, and it is today encompassed by<br />
<strong>the</strong> Felsenthal National Wildlife Refuge.<br />
Site survey activity, designed for wetland and forested conditions, consisted pr<strong>in</strong>cipally<br />
<strong>of</strong> floodpla<strong>in</strong> transects and river or tributary bankl<strong>in</strong>e surveys. A total <strong>of</strong> 144 sites,<br />
<strong>in</strong>clud<strong>in</strong>g 132 prehistoric and 15 historic sites was recorded (<strong>the</strong>se totals reflect <strong>the</strong> occurrence<br />
<strong>of</strong> both prehistoric and historic components on three sites). Prehistoric<br />
i
floodpla<strong>in</strong> sites are predom<strong>in</strong>antly small extractive camps with relatively sparse artifact<br />
and debris content. The nature <strong>of</strong> such extractive camps and <strong>the</strong>ir relationship to regional<br />
subsistence and settlement systems are a major focus <strong>of</strong> this report. About one-third<br />
<strong>of</strong> all floodpla<strong>in</strong> sites provided depth data for occupational levels or artifacts <strong>in</strong> place,<br />
and a prelim<strong>in</strong>ary estimate <strong>of</strong> local rates <strong>of</strong> recent alluviation has been developed. About<br />
two-thirds <strong>of</strong> all floodpla<strong>in</strong> sites furnished typological, technological, or stratigraphic<br />
data sufficient to identify cultural components. The cultural sequence extends from Late<br />
Archaic <strong>in</strong>to <strong>the</strong> protohistoric and historic periods, or about 3000 years. Earlier sites are<br />
deeply buried or drowned by river level and high water table. The best available site<br />
sample comprises extractive camps <strong>of</strong> <strong>the</strong> Mississippi period (A.D. 1100-<strong>17</strong>00), <strong>in</strong>clud<strong>in</strong>g<br />
a dramatic concentration <strong>of</strong> such camps on <strong>the</strong> lower Sal<strong>in</strong>e River. One major conclusion<br />
<strong>of</strong> this report is <strong>the</strong> importance <strong>of</strong> small Mississippi period extractive sites, which are not<br />
farmsteads, as components <strong>of</strong> a regional subsistence-settlement system. Such sites have<br />
been little studied <strong>in</strong> eastern North America.<br />
Historic sites recorded dur<strong>in</strong>g survey <strong>in</strong>clude steamboat land<strong>in</strong>gs, earth mounds<br />
built as raised platforms for build<strong>in</strong>gs or lock construction facilities, an underwater<br />
steamboat wreck, and o<strong>the</strong>r sites related to river<strong>in</strong>e and extractive <strong>in</strong>dustry dur<strong>in</strong>g <strong>the</strong><br />
n<strong>in</strong>eteenth and early twentieth centuries.<br />
The test<strong>in</strong>g program <strong>in</strong>volved eight prehistoric floodpla<strong>in</strong> sites (and would have<br />
expanded to o<strong>the</strong>r prehistoric and historic sites, if not for <strong>the</strong> rise <strong>of</strong> river level above<br />
project contour <strong>in</strong> late November). Sites tested <strong>in</strong>clude a relatively deep stratified site<br />
on <strong>the</strong> Ouachita River with Poverty Po<strong>in</strong>t, Tchula period, and later occupation levels,<br />
two Mississippi period middens with evidence <strong>of</strong> bulk fish process<strong>in</strong>g, two Mississippi<br />
period extractive sites with multiple activity loci, a small Baytown-Coles Creek period<br />
site, a protohistoric site with a Quapaw phase assemblage previously unknown <strong>in</strong> <strong>the</strong><br />
region, and ano<strong>the</strong>r possible Poverty Po<strong>in</strong>t period site. The last seven sites are all located<br />
on <strong>the</strong> lower Sal<strong>in</strong>e River.<br />
Tactics for site test<strong>in</strong>g <strong>in</strong>cluded controlled surface collection where riverbank sites<br />
were well exposed, subsurface test<strong>in</strong>g <strong>of</strong> about 5% <strong>of</strong> known site area, and power auger<br />
or shovel test<strong>in</strong>g to determ<strong>in</strong>e extent <strong>of</strong> buried occupation levels. Screen<strong>in</strong>g and<br />
wet screen<strong>in</strong>g recovery techniques were employed, and soil and organic samples were<br />
obta<strong>in</strong>ed for all sites tested. The results <strong>of</strong> <strong>the</strong> test<strong>in</strong>g provided considerable data for <strong>the</strong><br />
nature <strong>of</strong> floodpla<strong>in</strong> extractive sites and variation among <strong>the</strong>m, as well as an empirical<br />
basis for assess<strong>in</strong>g significance.<br />
ii
Analyses <strong>of</strong> about 13,100 artifacts and items <strong>of</strong> debris from site survey and test<strong>in</strong>g<br />
provided a variety <strong>of</strong> useful regional data, some prelim<strong>in</strong>ary <strong>in</strong> nature, but suggest<strong>in</strong>g<br />
fruitful directions for future work. The more significant results <strong>of</strong> <strong>the</strong>se analyses would<br />
<strong>in</strong>clude <strong>the</strong> follow<strong>in</strong>g.<br />
1. partial temporal sequences for both arrow po<strong>in</strong>t and dart po<strong>in</strong>t styles<br />
2. functional studies <strong>of</strong> lithic reduction and lithic resources<br />
3. detailed analysis and comparison <strong>of</strong> Gran <strong>Marais</strong> phase (A.D. 1100-1400) ceramic<br />
assemblages from two sites, and<br />
4. ceramic seriation data for Mississippi period sites, based on temper differences.<br />
The limited program <strong>of</strong> soil and sediment analysis <strong>in</strong>dicates that organic materials<br />
are deficient <strong>in</strong> most floodpla<strong>in</strong> sites largely because <strong>of</strong> extreme acidity. High levels<br />
<strong>of</strong> organic carbon and phosphorus can occur <strong>in</strong> anthropogenically enriched soils or<br />
middens. Radiocarbon dat<strong>in</strong>g <strong>of</strong> floodpla<strong>in</strong> sites has been markedly unsuccessful. Five<br />
samples from two sites gave ages from 500 to 3000 years earlier than expected. The dimensions<br />
<strong>of</strong> this special geochronological problem are discussed <strong>in</strong> detail.<br />
In <strong>the</strong> conclud<strong>in</strong>g chapter two topical areas <strong>of</strong> particular research <strong>in</strong>terest, <strong>in</strong> <strong>the</strong><br />
context <strong>of</strong> environment and culture <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland, are expanded. The<br />
first area <strong>in</strong>cludes discussion <strong>of</strong> <strong>the</strong> archeological context and dimensions <strong>of</strong> seasonal or<br />
transient extractive camps and <strong>the</strong> material correlates <strong>of</strong> extractive site activity. These<br />
sites are dist<strong>in</strong>guished by high research potential <strong>in</strong> many areas <strong>of</strong> current archeological<br />
and scientific <strong>in</strong>terest. The second area discusses a seasonal model for extractive activity<br />
and resource use <strong>in</strong> an overflow bottomland environment. The variety <strong>of</strong> extractive<br />
activities, especially fish<strong>in</strong>g, cont<strong>in</strong>ues throughout <strong>the</strong> year, even dur<strong>in</strong>g prolonged deep<br />
<strong>in</strong>undation <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. This model draws upon <strong>the</strong> best available project data<br />
from Mississippi period extractive camps on <strong>the</strong> Sal<strong>in</strong>e River and from ethnohistoric<br />
data for <strong>the</strong> Atchafalaya Bas<strong>in</strong> <strong>in</strong> south-central Louisiana, an ecologically similar overflow<br />
bottomland environment.<br />
iii
tAble <strong>of</strong> contents<br />
Abstract ............................................................................................................................................i<br />
List <strong>of</strong> Tables .................................................................................................................................ix<br />
List <strong>of</strong> Figures ............................................................................................................................ xiii<br />
Management Summary .............................................................................................................xix<br />
Acknowledgments .................................................................................................................. xxiii<br />
Chapter 1. Introduction to <strong>the</strong> Felsenthal Project .................................................................. 1<br />
The Felsenthal Project and Its Research Potential .............................................. 1<br />
Elevation Constra<strong>in</strong>t ......................................................................................... 2<br />
Def<strong>in</strong>ition <strong>of</strong> Study Area .................................................................................. 3<br />
Development <strong>of</strong> Research Design ......................................................................... 6<br />
Schedule <strong>of</strong> Investigations ............................................................................... 6<br />
Schedule <strong>of</strong> Hypo<strong>the</strong>ses ................................................................................... 8<br />
Felsenthal Project Personnel ................................................................................ 10<br />
Chapter 2. <strong>Grand</strong> <strong>Marais</strong> Lowland Environment ................................................................ 13<br />
Sou<strong>the</strong>astern Floodpla<strong>in</strong> Ecosystems ................................................................. 15<br />
Riparian Ecosystems ....................................................................................... 15<br />
Aquatic Ecosystems ........................................................................................ <strong>17</strong><br />
Pleistocene Terrace Sequence .............................................................................. 19<br />
Recent Floodpla<strong>in</strong> Geohydrology ....................................................................... 24<br />
Flood Characteristics ...................................................................................... 24<br />
Floodpla<strong>in</strong> Features ........................................................................................ 25<br />
Topstratum/Substratum Deposits ................................................................ 28<br />
Floodpla<strong>in</strong> Soils ............................................................................................... 29<br />
Channel Geometry and Channel Deposits .................................................. 31<br />
Riverbanks and Bank Erosion ....................................................................... 34<br />
Recent Floodpla<strong>in</strong> Ecology .................................................................................. 36<br />
Hardwood and Swamp Forest ...................................................................... 36<br />
Vertebrates ....................................................................................................... 41<br />
Invertebrates .................................................................................................... 45<br />
Potential Paleoecological Studies ........................................................................ 46<br />
Potential Lowland Resources .............................................................................. 47<br />
Chapter 3. Prehistory <strong>of</strong> <strong>the</strong> Felsenthal Region, by John H. House .................................. 51<br />
Introduction ........................................................................................................... 51<br />
Previous Archeological Investigation ................................................................ 52<br />
Explorations ..................................................................................................... 52<br />
Recent Investigations ...................................................................................... 54<br />
v
The Sequence <strong>of</strong> Aborig<strong>in</strong>al Occupation ........................................................... 57<br />
Paleo-Indian Period, before 8000 B.C. .......................................................... 58<br />
Early Archaic Period, 8000-6000 B.C. ........................................................... 60<br />
Middle Archaic Period, 6000-4000 B.C. ........................................................ 61<br />
Late Archaic Period, 4000-1200 B.C. ............................................................. 62<br />
Poverty Po<strong>in</strong>t Period, 1200-500 B.C. ............................................................. 63<br />
Tchula Period, 500 B.C.-A.D. 1 ...................................................................... 64<br />
Marksville Period, A.D. 1-300 ....................................................................... 65<br />
Baytown Period, A.D. 300-700 ....................................................................... 66<br />
Coles Creek Period, A.D. 700-1100 ............................................................... 67<br />
Mississippi Period, A.D. 1100-<strong>17</strong>00 .............................................................. 68<br />
Discussion<br />
Chapter 4. History <strong>of</strong> <strong>the</strong> Felsenthal Region, by Beverly Watk<strong>in</strong>s .................................... 79<br />
Introduction ........................................................................................................... 79<br />
European Exploration and Settlement ............................................................... 79<br />
American Exploration and Settlement ............................................................... 80<br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g ..................................................................................... 83<br />
Disposition <strong>of</strong> Swamplands ................................................................................. 85<br />
River Commerce and Travel ................................................................................ 85<br />
Effects <strong>of</strong> <strong>the</strong> Civil War ......................................................................................... 87<br />
Improvements <strong>in</strong> Navigation .............................................................................. 87<br />
The Sternwheeler Lotawanna ......................................................................... 90<br />
Lumber<strong>in</strong>g and Railroads .................................................................................... 94<br />
The Felsenthal Community ................................................................................. 96<br />
Conclusions ............................................................................................................ 97<br />
Chapter 5. Variability Among Floodpla<strong>in</strong> Sites ................................................................... 99<br />
Survey Strategy and Tactics ............................................................................... 101<br />
Random and Judgmental Transects ........................................................... 105<br />
River Bankl<strong>in</strong>e Surveys ................................................................................ 112<br />
Tributary Bankl<strong>in</strong>e and Backswamp Surveys ........................................... 116<br />
Site Specific Surveys ..................................................................................... 1<strong>17</strong><br />
Site and Component Identification ............................................................ 1<strong>17</strong><br />
Survey Results ..................................................................................................... 121<br />
Unit 1: Ouachita River Above Mile 254 .................................................... 121<br />
Unit 2: Lower Sal<strong>in</strong>e River .......................................................................... 126<br />
Unit 3: Lower Eagle Creek .......................................................................... 130<br />
Unit 4: Ouachita River Below Mile 254 ..................................................... 130<br />
Unit 5: Lower Lapile Creek ........................................................................ 132<br />
Unit 6: Oxbow Lakes and Backswamp ..................................................... 133<br />
Unit 7: Pleistocene Terraces and Islands ................................................... 136<br />
Unit 8: Historic Sites .................................................................................... 138<br />
Evaluation <strong>of</strong> Survey Strategy and Tactics ...................................................... 146<br />
Site Locational and Historical Hypo<strong>the</strong>ses ..................................................... 150<br />
vi
Chapter 6. Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites ......................................................................... 155<br />
Selection <strong>of</strong> Sites for Test<strong>in</strong>g .............................................................................. 155<br />
Test Excavation Strategy and Tactics ................................................................ 157<br />
Marie Sal<strong>in</strong>e (3AS329) .................................................................................. 159<br />
False Indigo (3AS285) ................................................................................... 168<br />
Buttonbush (3BR58) ...................................................................................... <strong>17</strong>5<br />
One Cypress Po<strong>in</strong>t (3AS286) ....................................................................... <strong>17</strong>8<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76) .............................................................................. 185<br />
Jug Po<strong>in</strong>t 1 (3AS306) ..................................................................................... 187<br />
Jug Po<strong>in</strong>t 2 (3AS307) ..................................................................................... 191<br />
Mouth <strong>of</strong> Eagle Creek (3BR78) .................................................................... 194<br />
Evaluation <strong>of</strong> Test Excavation Strategy and Tactics ....................................... 194<br />
Summary Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites ....................................................... 197<br />
Site Functional Hypo<strong>the</strong>ses ............................................................................... 197<br />
Variation Among Extractive Sites ............................................................... 201<br />
Chapter 7. Soils, Sediments, and Chronology .................................................................... 205<br />
Selected Properties <strong>of</strong> Floodpla<strong>in</strong> Soils ............................................................ 205<br />
Sediment Textures and Floodpla<strong>in</strong> Alluviation .............................................. 208<br />
Radiocarbon Dates for Floodpla<strong>in</strong> Sites........................................................... 212<br />
Site Transformational Hypo<strong>the</strong>ses .................................................................... 216<br />
Chapter 8. Lithic and Ceramic Analysis .............................................................................. 219<br />
Lithic Source Study ............................................................................................. 219<br />
Methods <strong>of</strong> Collection and Analysis .......................................................... 221<br />
Results <strong>of</strong> Analysis ........................................................................................ 222<br />
Lithic Analysis and Results ................................................................................ 224<br />
Arrow Po<strong>in</strong>ts .................................................................................................. 224<br />
Dart Po<strong>in</strong>ts ..................................................................................................... 230<br />
Bifacial Preforms and O<strong>the</strong>r Bifaces ........................................................... 234<br />
O<strong>the</strong>r Flaked Stone Tools ............................................................................. 237<br />
Cores and Debitage ....................................................................................... 240<br />
Cobble and Heavy Duty Tools .................................................................... 244<br />
Ground Stone Tools ...................................................................................... 245<br />
M<strong>in</strong>eral Pigments .......................................................................................... 250<br />
Inorganic Debris ............................................................................................ 250<br />
Ceramic Analysis and Results ........................................................................... 251<br />
Tchefuncte Complex ..................................................................................... 252<br />
Baytown Complex ........................................................................................ 254<br />
Gran <strong>Marais</strong> Complex .................................................................................. 255<br />
Shell-tempered Complex ............................................................................. 263<br />
Caddoan Imports .......................................................................................... 266<br />
Culture Historical Hypo<strong>the</strong>ses .......................................................................... 266<br />
vii
Chapter 9. <strong>Human</strong> <strong>Adaptation</strong> <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland ....................................... 269<br />
On <strong>the</strong> Nature <strong>of</strong> Extractive Sites ...................................................................... 269<br />
Site Size ........................................................................................................... 271<br />
Site Configuration ......................................................................................... 272<br />
Site Assemblage ............................................................................................. 272<br />
Site Residue .................................................................................................... 273<br />
Site Location .................................................................................................. 273<br />
A Model for Resource Use <strong>in</strong> an Overflow Bottomland ................................ 275<br />
References Cited ........................................................................................................................ 279<br />
Appendix A. Site Descriptive and Analytical Data .........................................................A-1<br />
Prehistoric Sites .............................................................................................A-1<br />
Historic Sites ..................................................................................................A-2<br />
Appendix B. Faunal and Floral Analysis .......................................................................... B-1<br />
Vertebrate Fauna ........................................................................................... B-1<br />
Molluscan Fauna ........................................................................................... B-4<br />
Carbonized Plant Rema<strong>in</strong>s .......................................................................... B-6<br />
Subfossil Wood .............................................................................................. B-6<br />
Appendix C. Recommended Program for Mitigation <strong>of</strong> Impacts on Archeological<br />
Resources <strong>in</strong> <strong>the</strong> Felsenthal Navigation Pool ............................................C-1<br />
Introduction ...................................................................................................C-1<br />
Significance <strong>of</strong> Floodpla<strong>in</strong> Sites ..................................................................C-4<br />
National Register Evaluation ....................................................................C-10<br />
Assessment <strong>of</strong> Site Impacts .......................................................................C-14<br />
Recommended Program <strong>of</strong> Mitigation ....................................................C-<strong>17</strong><br />
General Recommendations .......................................................................C-29<br />
viii
list <strong>of</strong> tAbles<br />
Table 1. Resource Potential <strong>of</strong> Floodpla<strong>in</strong> Trees, Shrubs, and V<strong>in</strong>es ............................... 40<br />
Table 2. Floodpla<strong>in</strong> Mammals <strong>in</strong> Order <strong>of</strong> Body Weight .................................................. 42<br />
Table 3. Selected Food Fishes and Habitat Preferences, Lower Ouachita River<br />
System <strong>in</strong> <strong>Arkansas</strong> .................................................................................................. 44<br />
Table 4. Comparison <strong>of</strong> Archeological Sequences <strong>in</strong> <strong>the</strong> Felsenthal, Bartholomew-<br />
Macon, Mid-Ouachita, and Tensas Regions ......................................................... 59<br />
Table 5. Power Equipment and Accessories Employed for Wetland Archeological<br />
Survey <strong>in</strong> <strong>the</strong> Felsenthal Project Area .................................................................. 102<br />
Table 6. Characteristics <strong>of</strong> N<strong>in</strong>e Judgmental Transects <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> and<br />
Ouachita-Sal<strong>in</strong>e Channels Subareas .................................................................... 111<br />
Table 7. Channel Geometry and Prehistoric Site Distribution for Three Site Units<br />
on Ouachita and Sal<strong>in</strong>e Rivers ............................................................................. 123<br />
Table 8. Density <strong>of</strong> Prehistoric Riverbank Components <strong>in</strong> Three Site Units on<br />
Ouachita and Sal<strong>in</strong>e Rivers ................................................................................... 151<br />
Table 9. Summary <strong>of</strong> Test Excavation Tactics for Floodpla<strong>in</strong> Sites ................................ 158<br />
Table 10. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area,<br />
Marie Sal<strong>in</strong>e Site (3AS329) .................................................................................... 167<br />
Table 11. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area,<br />
False Indigo (3AS285) ............................................................................................ <strong>17</strong>5<br />
Table 12. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area,<br />
One Cypress Po<strong>in</strong>t (3AS286) ................................................................................. 183<br />
ix
Table 13. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area,<br />
Jug Po<strong>in</strong>t 1 (3AS306) ............................................................................................... 191<br />
Table 14. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area,<br />
Jug Po<strong>in</strong>t 2 (3AS307) ............................................................................................... 193<br />
Table 15. Summary Assessment <strong>of</strong> Eight Floodpla<strong>in</strong> Sites ............................................... 198<br />
Table 16. Proportions <strong>of</strong> Five Functional Categories <strong>of</strong> Imperishable Artifacts and<br />
Debris <strong>in</strong> N<strong>in</strong>e Prehistoric Components (Six Floodpla<strong>in</strong> Sites) ....................... 203<br />
Table <strong>17</strong>. Selected Properties <strong>of</strong> Soil Samples from Seven Floodpla<strong>in</strong> Sites ................... 206<br />
Table 18. Particle Size Frequencies by Depth <strong>in</strong> Sediment Samples from Test Pits 1<br />
and 6, Area C, Marie Sal<strong>in</strong>e Site (3AS329) .......................................................... 209<br />
Table 19. Characteristics and Results <strong>of</strong> Age Determ<strong>in</strong>ation for Radiocarbon Samples<br />
from Marie Sal<strong>in</strong>e (3AS329) and Jug Po<strong>in</strong>t 2 (3AS307) ..................................... 213<br />
Table 20. Proposed Sequence <strong>of</strong> Arrow Po<strong>in</strong>t Types <strong>in</strong> <strong>the</strong> Felsenthal Region .............. 226<br />
Table 21. Metric and Nonmetric Data for Dart Po<strong>in</strong>ts and Proposed General Sequence<br />
<strong>of</strong> Dart Po<strong>in</strong>t Types <strong>in</strong> <strong>the</strong> Felsenthal Region ..................................................... 231<br />
Table 22. Comparison <strong>of</strong> Debitage by Reduction Category and Raw Material<br />
Frequency among Three Floodpla<strong>in</strong> Site Assemblages .................................... 244<br />
Table 23. Descriptive Classification <strong>of</strong> Pla<strong>in</strong>ware from Two Sites <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong><br />
Phase, Mississippi Period ...................................................................................... 260<br />
Table 24. Descriptive Classification <strong>of</strong> Decorated Sherds from Two Sites <strong>of</strong> <strong>the</strong> Gran<br />
<strong>Marais</strong> Phase, Mississippi Period ..................................................................261-262<br />
Table 25. Proportions <strong>of</strong> Shell and Clay Temper<strong>in</strong>g <strong>in</strong> Ceramic Samples from Four<br />
Floodpla<strong>in</strong> Sites <strong>of</strong> <strong>the</strong> Mississippi Period .......................................................... 266<br />
Table 26. Three Classifications <strong>of</strong> Settlement Type which Emphasize or Dist<strong>in</strong>guish<br />
Small Extractive Sites ............................................................................................. 270<br />
x
Appendix A<br />
Table A-1. Prehistoric Sites (Units 1-7) ................................................................................A-4<br />
Table A-2. Historic Sites (Unit 8) ..........................................................................................A-9<br />
Table A-3. Occurrence <strong>of</strong> Temporally or Culturally Diagnostic Artifacts <strong>in</strong><br />
Prehistoric Floodpla<strong>in</strong> Sites .............................................................................A-10<br />
Appendix b<br />
Table B-1. Vertebrate Rema<strong>in</strong>s from Prehistoric Floodpla<strong>in</strong> Sites .................................. B-2<br />
Table B-2. Molluscs Identified <strong>in</strong> Prehistoric and Historic Site Samples, and<br />
<strong>in</strong> a Modern Sample from Felsenthal National Wildlife Refuge .................. B-5<br />
Table B-3. Floral Rema<strong>in</strong>s from Prehistoric and Historic Floodpla<strong>in</strong> Sites ................... B-7<br />
Table B-4. Wood and Subfossil Wood Specimens from Environmental Localities ...... B-8<br />
Appendix c<br />
Table C-1. Six Felsenthal Project Floodpla<strong>in</strong> Sites Recommended for Extensive<br />
and Intensive Investigation................................................................................C-7<br />
Table C-2. Twenty-n<strong>in</strong>e Felsenthal Project Floodpla<strong>in</strong> Sites Recommended for<br />
Two-stage Mitigation ..........................................................................................C-8<br />
Table C-3. Summary <strong>of</strong> Site Characteristics Relevant to Research Potential and<br />
Scientific or Historic Significance for 1<strong>17</strong> Floodpla<strong>in</strong> Sites .........................C-11<br />
Table C-4. Mitigation Budget Estimate (<strong>in</strong> field man days):<br />
PHASE 2: Mitigation <strong>of</strong> Six Prehistoric Floodpla<strong>in</strong> Sites ...........................C-19<br />
Table C-5. Mitigation Budget Estimate (<strong>in</strong> field man days):<br />
PHASE 3: Two-Stage Mitigation, 23 Prehistoric Floodpla<strong>in</strong> Sites..............C-20<br />
Table C-6. Mitigation Budget Estimate (<strong>in</strong> field man days):<br />
PHASE 3: Two-Stage Mitigation, Six Historic Sites .....................................C-21<br />
xi
xii
list <strong>of</strong> figures<br />
Figure 1. Location and features <strong>of</strong> Felsenthal National Wildlife Refuge <strong>in</strong><br />
sou<strong>the</strong>rn <strong>Arkansas</strong> .................................................................................................... 5<br />
Figure 2. Flow <strong>of</strong> <strong>in</strong>formation between research units from research design to proposed<br />
mitigation program for <strong>the</strong> Felsenthal project .......................................... 7<br />
Figure 3. Quaternary geology <strong>of</strong> <strong>the</strong> Ouachita River Valley <strong>in</strong> <strong>Arkansas</strong> ....................... 20<br />
Figure 4. Aerial view <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River-Eagle Creek floodpla<strong>in</strong> and Prairie Island ... 22<br />
Figure 5. Possible lowland prairie remnant on <strong>the</strong> Deweyville 2 Terrace near<br />
Redeye Lake ............................................................................................................. 23<br />
Figure 6. Hydrological features <strong>in</strong> <strong>the</strong> Felsenthal project area ......................................... 27<br />
Figure 7. Topstratum deposits exposed by <strong>the</strong> Ouachita River bankl<strong>in</strong>e ....................... 30<br />
Figure 8. Meander<strong>in</strong>g channel characteristics and term<strong>in</strong>ology ...................................... 32<br />
Figure 9. Typical riverbank pr<strong>of</strong>iles <strong>in</strong> <strong>the</strong> Felsenthal Project area .................................. 35<br />
Figure 10. Bottomland hardwood forest, predom<strong>in</strong>antly mature red oaks, <strong>in</strong> <strong>the</strong><br />
Ouachita River floodpla<strong>in</strong> ..................................................................................... 37<br />
Figure 11. Hardwood and swamp forest <strong>in</strong> <strong>the</strong> Felsenthal Project area ........................... 38<br />
Figure 12. Cross-valley section <strong>in</strong> <strong>the</strong> Felsenthal Project area show<strong>in</strong>g terraces,<br />
floodpla<strong>in</strong>, forest types, and o<strong>the</strong>r features ........................................................ 48<br />
Figure 13. A portion <strong>of</strong> <strong>the</strong> Ouachita River Valley designated <strong>the</strong> Felsenthal<br />
region and locations <strong>of</strong> some important prehistoric sites ................................. 53<br />
xiii
Figure 14. The Mississippi Period mound and midden complex at Watts Field<br />
(3UN18) .................................................................................................................... 70<br />
Figure 15. Historic sites, settlements, and placenames <strong>in</strong> <strong>the</strong> Felsenthal Region ............ 82<br />
Figure 16. Hydrographic survey chart <strong>of</strong> Ouachita River <strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area, show<strong>in</strong>g locations <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g (3AS299), <strong>Grand</strong> Mary<br />
(or <strong>Grand</strong> <strong>Marais</strong>) Land<strong>in</strong>g (3UN122), and o<strong>the</strong>r features ............................... 84<br />
Figure <strong>17</strong>. The sternwheeler Lotawanna fully loaded with cotton bales ............................ 89<br />
Figure 18. Portions <strong>of</strong> certificates <strong>of</strong> registry for <strong>the</strong> Lotawanna ......................................... 91<br />
Figure 19. Advertisement from unknown newspaper source (about 1868)<br />
announc<strong>in</strong>g Lotawanna’s Camden-New Orleans trade ..................................... 93<br />
Figure 20. The Felsenthal Project area, subareas, and 13 floodpla<strong>in</strong> transect locations .. 106<br />
Figure 21. Site survey tactics and equipment ...................................................................... 108<br />
Figure 22. Mounted air photo mosaics used for location f<strong>in</strong>d<strong>in</strong>g and site plott<strong>in</strong>g<br />
by boat survey and floodpla<strong>in</strong> transect survey teams ..................................... 109<br />
Figure 23. Routes and extent <strong>of</strong> bankl<strong>in</strong>e survey <strong>in</strong> river channels, tributaries,<br />
open lakes, and backswamp areas accessible by boat ..................................... 113<br />
Figure 24. Site Unit 1, Ouachita River above mile 254, with prehistoric sites (dots)<br />
located <strong>in</strong> relation to channel geometry. Some prehistoric sites <strong>in</strong> Unit 6<br />
and historic sites <strong>in</strong> Unit 8 (rectangles) are also shown on this map. ........... 122<br />
Figure 25. Site 3UN162, consist<strong>in</strong>g <strong>of</strong> fire-cracked rock exposed at 3 m depth <strong>in</strong><br />
<strong>the</strong> Ouachita River bankl<strong>in</strong>e ................................................................................ 125<br />
xiv
Figure 26. Site Unit 2, Lower Sal<strong>in</strong>e River, and Site Unit 3, Lower Eagle Creek,<br />
with prehistoric sites (dots) located <strong>in</strong> relation to channel geometry. Two<br />
historic sites with prehistoric components (rectangles enclos<strong>in</strong>g dots)<br />
<strong>in</strong> Unit 8 are also shown on this map. ............................................................... 127<br />
Figure 27. Mississippi period extractive sites on lower Sal<strong>in</strong>e River ............................... 129<br />
Figure 28. Site Unit 4, Ouachita River below mile 254, with prehistoric sites (dots)<br />
located <strong>in</strong> relation to channel geometry. Historic sites (rectangles) or<br />
historic sites with prehistoric components (rectangles enclos<strong>in</strong>g dots) <strong>in</strong><br />
Unit 8 are also shown on this map. .................................................................... 131<br />
Figure 29. Riverbank sites on Ouachita River ..................................................................... 133<br />
Figure 30. Site Unit 5, Lower Lapile Creek, with prehistoric sites (dots) located <strong>in</strong><br />
relation to channel geometry. Two mound sites on <strong>the</strong> adjacent terrace,<br />
Locust Ridge (3UN8) and Shallow Lake (3UN9/52) <strong>in</strong> Unit 7 are also<br />
shown on this map. .............................................................................................. 135<br />
Figure 31. Site Unit 6, oxbow lakes and backswamp, with prehistoric sites located<br />
<strong>in</strong> relation to floodpla<strong>in</strong> and terrace features ................................................... 137<br />
Figure 32. Historic artifacts from floodpla<strong>in</strong> sites .............................................................. 140<br />
Figure 33. Hydrographic survey chart <strong>of</strong> Ouachita River <strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area, show<strong>in</strong>g location <strong>of</strong> “Wreck <strong>of</strong> Lotti Warner” (Lotawanna) ................... 147<br />
Figure 34. The spatial array <strong>of</strong> Mississippi period sites <strong>in</strong> or near <strong>the</strong> Felsenthal<br />
project area ............................................................................................................. 152<br />
Figure 35. Aerial view <strong>of</strong> <strong>the</strong> Ouachita River-<strong>Marais</strong> Sal<strong>in</strong>e Lake locale and <strong>the</strong><br />
location <strong>of</strong> Marie Sal<strong>in</strong>e site ................................................................................ 160<br />
Figure 36. Marie Sal<strong>in</strong>e (3AS329), test excavations October 1979 .................................... 161<br />
xv
Figure 37. Marie Sal<strong>in</strong>e (3AS329), deep test pit pr<strong>of</strong>iles .................................................... 163<br />
Figure 38. Marie Sal<strong>in</strong>e (3AS329), pr<strong>of</strong>ile records for Test Pits 1 and 6 (south walls) ... 164<br />
Figure 39. Cultural stratigraphy <strong>in</strong> Area C, Marie Sal<strong>in</strong>e (3AS329) ................................. 166<br />
Figure 40. False Indigo (3AS285), controlled surface collection and test excavations,<br />
November 1979 ..................................................................................................... 169<br />
Figure 41. False Indigo (3AS285), controlled surface collection <strong>in</strong> Area A with<br />
scatter map <strong>of</strong> sherds and o<strong>the</strong>r artifacts .......................................................... <strong>17</strong>0<br />
Figure 42. False Indigo (3AS285), pr<strong>of</strong>ile with midden lens and o<strong>the</strong>r features <strong>in</strong><br />
Test Pit 15 ............................................................................................................... <strong>17</strong>2<br />
Figure 43. False Indigo (3AS285), Area A show<strong>in</strong>g 18 m 2 area <strong>of</strong> dense artifact<br />
concentration after surface collection and test excavation ............................. <strong>17</strong>3<br />
Figure 44. Buttonbush (3BR58), test excavations, November 1979 .................................. <strong>17</strong>6<br />
Figure 45. Buttonbush (3BR58), rema<strong>in</strong>s <strong>of</strong> two Baytown Pla<strong>in</strong> vessels <strong>in</strong> Test Pit 2 .... <strong>17</strong>7<br />
Figure 46. Aerial view <strong>of</strong> <strong>the</strong> lower Sal<strong>in</strong>e River floodpla<strong>in</strong> and <strong>the</strong> location <strong>of</strong><br />
One Cypress Po<strong>in</strong>t (3AS286) ............................................................................... <strong>17</strong>9<br />
Figure 47. One Cypress Po<strong>in</strong>t (3AS286), controlled surface collections and test<br />
excavations, November 1979 ............................................................................... 180<br />
Figure 48. Flaked stone tools and cores from One Cypress Po<strong>in</strong>t (3AS286) ................... 181<br />
Figure 49. One Cypress Po<strong>in</strong>t (3AS286), graphic presentation <strong>of</strong> lithic assemblage ..... 184<br />
Figure 50. Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76), test excavations, November 1979 .......................... 186<br />
Figure 51. Jug Po<strong>in</strong>t 1 (3AS306), test excavations, November 1979 ................................. 188<br />
Figure 52. Jug Po<strong>in</strong>t 1 (3AS306), test pit pr<strong>of</strong>iles ................................................................. 189<br />
xvi
Figure 53. Jug Po<strong>in</strong>t 2 (3AS307), test excavations, November 1979 ................................. 192<br />
Figure 54. Mouth <strong>of</strong> Eagle Creek (3BR78), test excavations, November 1979 ................ 195<br />
Figure 55. Dendrogram generated by cluster analysis <strong>of</strong> artifacts and debris <strong>in</strong><br />
n<strong>in</strong>e prehistoric components (six floodpla<strong>in</strong> sites) .......................................... 202<br />
Figure 56. Estimated rate <strong>of</strong> alluviation for <strong>the</strong> natural levees <strong>of</strong> Ouachita and<br />
lower Sal<strong>in</strong>e River <strong>in</strong> <strong>the</strong> Felsenthal Project area, based on selected<br />
riverbank sites ....................................................................................................... 210<br />
Figure 57. Histograms <strong>of</strong> size frequency <strong>in</strong> two samples <strong>of</strong> pebbles and small<br />
cobble from Ouachita Valley localities .............................................................. 223<br />
Figure 58. Arrow po<strong>in</strong>ts from <strong>the</strong> Felsenthal Project area <strong>in</strong> proposed chronological<br />
sequence ........................................................................................................... 227<br />
Figure 59. Dart po<strong>in</strong>ts from <strong>the</strong> Felsenthal Project area <strong>in</strong> proposed chronological<br />
sequence ................................................................................................................. 232<br />
Figure 60. Bifacial tools, preforms, and cores ..................................................................... 236<br />
Figure 61. Miscellaneous flaked stone tools ........................................................................ 239<br />
Figure 62. Schematic l<strong>in</strong>ear model for production <strong>of</strong> a bifacial tool and discard<br />
materials ................................................................................................................. 243<br />
Figure 63. Cobble tools ........................................................................................................... 246<br />
Figure 64. Groundstone tools and pigment ......................................................................... 247<br />
Figure 65. Tchefuncte complex ceramics ............................................................................. 253<br />
Figure 66. Baytown complex pla<strong>in</strong>ware ............................................................................... 256<br />
Figure 67. Miscellaneous sherds from floodpla<strong>in</strong> sites ...................................................... 257<br />
Figure 68. Decorated rim sherds, Gran <strong>Marais</strong> ceramic complex, False Indigo<br />
(3AS285) ................................................................................................................. 264<br />
xvii
Figure 69. Decorated body sherds, Gran <strong>Marais</strong> ceramic complex, False Indigo<br />
(3AS285) ................................................................................................................. 265<br />
Figure 70. A seasonal model for subsistence activity and resource use <strong>in</strong> <strong>the</strong><br />
<strong>Grand</strong> <strong>Marais</strong> Lowland ........................................................................................ 276<br />
xviii
MAnAgeMent suMMAry<br />
The <strong>Arkansas</strong> Archeological Survey conducted a multistage cultural resources<br />
survey and site test<strong>in</strong>g project <strong>in</strong> <strong>the</strong> Felsenthal Navigation Pool on <strong>the</strong> Ouachita and<br />
lower Sal<strong>in</strong>e rivers for <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg District, under<br />
terms <strong>of</strong> contract DACW 38-79-C-0002. Fieldwork was preceded by development <strong>of</strong> a<br />
broad-based, <strong>in</strong>terdiscipl<strong>in</strong>ary research design and by detailed archeological, historical,<br />
ecological, and geohydrological background research. Exist<strong>in</strong>g or proposed National<br />
Register properties were not previously recorded <strong>in</strong> <strong>the</strong> Felsenthal Project area, which is<br />
an overflow bottomland never permanently settled or farmed <strong>in</strong> historic times and now<br />
encompassed by <strong>the</strong> Felsenthal National Wildlife Refuge.<br />
Multiple field phases began <strong>in</strong> mid-August 1979, shortly after w<strong>in</strong>ter-spr<strong>in</strong>g floodwaters<br />
receded below <strong>the</strong> upper project contour (65 feet MSL), and cont<strong>in</strong>ued through<br />
late November 1979 when floodwaters aga<strong>in</strong> rose above this contour. Dur<strong>in</strong>g this threeand-one-half<br />
month field season, three floodpla<strong>in</strong> survey teams operated cont<strong>in</strong>uously<br />
through a seven week field phase; test excavation <strong>the</strong>n began with two test<strong>in</strong>g crews<br />
operat<strong>in</strong>g cont<strong>in</strong>uously through much <strong>of</strong> <strong>the</strong> next seven week period. Some site survey<br />
activity cont<strong>in</strong>ued on a reduced basis dur<strong>in</strong>g <strong>the</strong> test<strong>in</strong>g phase, so that site survey and<br />
test<strong>in</strong>g phases actually overlapped. A field laboratory also operated cont<strong>in</strong>uously at <strong>the</strong><br />
project headquarters throughout <strong>the</strong> field season.<br />
Site survey teams recorded 126 previously unknown archeological sites and revisited<br />
18 sites known from prior work <strong>in</strong> <strong>the</strong> area. These 144 sites are briefly described <strong>in</strong><br />
Appendix A. Essentially all floodpla<strong>in</strong> sites <strong>in</strong> <strong>the</strong> project area are buried, but many are<br />
exposed by riverbank erosion. Site survey tactics consisted <strong>of</strong> 13 random and judgmental<br />
transects travers<strong>in</strong>g about 62 km, and 10 river or tributary bankl<strong>in</strong>e surveys total<strong>in</strong>g<br />
about <strong>17</strong>7 km. Various site specific surveys were also carried out, as well as collection<br />
<strong>of</strong> environmental data <strong>in</strong> conjunction with survey activity. Subsurface test<strong>in</strong>g for buried<br />
sites <strong>in</strong>cluded use <strong>of</strong> power auger, shovel test<strong>in</strong>g, and close <strong>in</strong>spection <strong>of</strong> eroded banks<br />
or o<strong>the</strong>r exposed areas. All survey work was designed to contend with difficult wetland<br />
conditions and with extensive floodpla<strong>in</strong> forest cover.<br />
xix
Site test<strong>in</strong>g focused on eight prehistoric sites: a large, deep, and stratified site on<br />
<strong>the</strong> Ouachita River and seven shallow buried sites <strong>of</strong> relatively small size on <strong>the</strong> lower<br />
Sal<strong>in</strong>e River. Additional test<strong>in</strong>g <strong>of</strong> selected prehistoric and historic sites was prohibited<br />
by ris<strong>in</strong>g floodwaters. Site test<strong>in</strong>g tactics <strong>in</strong>cluded random, judgmental, and systematic<br />
test pits which sampled about 5% <strong>of</strong> known site areas, and also numerous power auger<br />
or shovel holes to del<strong>in</strong>eate site extent. Controlled surface collection was employed<br />
where riverbank erosion exposed discrete artifact scatters and arrays <strong>of</strong> debris. Each<br />
site was mapped by <strong>in</strong>strument and recorded by standard archeological methods. Site<br />
records filed at <strong>the</strong> <strong>Arkansas</strong> Archeological Survey run over 1000 pages.<br />
Extensive laboratory analysis, deal<strong>in</strong>g with about 13,100 artifacts and items <strong>of</strong><br />
debris and numerous o<strong>the</strong>r samples and residues, cont<strong>in</strong>ued through 1980 at <strong>the</strong> Contract<br />
Laboratory, <strong>Arkansas</strong> Archeological Survey, or <strong>in</strong> various specialists’ laboratories.<br />
Results <strong>of</strong> <strong>the</strong>se studies are reported <strong>in</strong> Chapters 7 and 8, and <strong>in</strong> Appendix B.<br />
All floodpla<strong>in</strong> sites recorded by site survey or tested dur<strong>in</strong>g 1979 were evaluated<br />
with regard to research potential <strong>in</strong> <strong>the</strong> context <strong>of</strong> regional or more broadly relevant<br />
problems <strong>in</strong> prehistory and history. This evaluation identified 37 floodpla<strong>in</strong> sites as significant<br />
resources. A detailed assessment <strong>of</strong> navigation pool <strong>in</strong>undation and related impacts<br />
<strong>in</strong>dicates that all significant floodpla<strong>in</strong> sites will be adversely affected. One <strong>of</strong> <strong>the</strong><br />
37 sites has been adequately mitigated by test<strong>in</strong>g and ano<strong>the</strong>r has recently been tested<br />
by Historic Preservation Associates.<br />
In compliance with <strong>the</strong> contract Scope <strong>of</strong> Services, <strong>the</strong> <strong>Arkansas</strong> Archeological<br />
Survey has prepared National Register nom<strong>in</strong>ations for 16 significant sites which have<br />
archeological deposits at or below <strong>the</strong> 65 foot contour <strong>in</strong>terval. However, based on <strong>the</strong><br />
evidence presented <strong>in</strong> this report, we have determ<strong>in</strong>ed that a total <strong>of</strong> 35 floodpla<strong>in</strong> sites<br />
appear eligible for nom<strong>in</strong>ation to <strong>the</strong> National Register <strong>of</strong> Historic Places, 19 <strong>of</strong> which<br />
are above 65 feet. These 35 floodpla<strong>in</strong> sites, for effective management purposes, could be<br />
nom<strong>in</strong>ated as a Multiple Resource Area; a recommendation to this effect is made <strong>in</strong> Appendix<br />
C.<br />
Among mitigation alternatives considered, only mitigation by data recovery has<br />
been considered feasible. We have <strong>the</strong>refore recommended a multistage <strong>in</strong>terdiscipl<strong>in</strong>ary<br />
program <strong>of</strong> mitigation. This program features extensive excavation and data recovery<br />
<strong>in</strong> six floodpla<strong>in</strong> sites dur<strong>in</strong>g a first field season, us<strong>in</strong>g concurrently operat<strong>in</strong>g field<br />
crews. A second full season <strong>of</strong> fieldwork would comprise two-stage mitigation <strong>of</strong> 29<br />
o<strong>the</strong>r significant floodpla<strong>in</strong> sites, <strong>in</strong>clud<strong>in</strong>g both prehistoric and historic sites. These 29<br />
sites are divided <strong>in</strong>to seven groups by cultural period; <strong>the</strong> flexible two-stage mitigation<br />
approach requires limited test<strong>in</strong>g <strong>of</strong> all or most sites with<strong>in</strong> each group, followed<br />
xx
y extensive excavation and data recovery at one or more selected sites with<strong>in</strong> each<br />
group. Aga<strong>in</strong>, multiple field crews would operate concurrently <strong>in</strong> order to meet objectives<br />
<strong>of</strong> <strong>the</strong> mitigation program (Appendix C). It is recognized that <strong>the</strong> Corps ma<strong>in</strong>ta<strong>in</strong>s<br />
adm<strong>in</strong>istrative responsibility only for <strong>the</strong>se sites below 65 feet; none<strong>the</strong>less it is <strong>the</strong><br />
Survey’s recommendation that <strong>the</strong> resources would be best served by a coord<strong>in</strong>ated<br />
research approach to <strong>the</strong> entire Felsenthal project.<br />
xxi
xxii
AcknowledgMents<br />
Many <strong>in</strong>dividuals and organizations contributed significantly to <strong>the</strong> performance<br />
<strong>of</strong> <strong>the</strong> Felsenthal Archeological Project and to preparation <strong>of</strong> this report. It is a pleasure<br />
to acknowledge this assistance which <strong>in</strong>cludes adm<strong>in</strong>istrative and scientific expertise,<br />
technical efforts or skill, and south <strong>Arkansas</strong> hospitality.<br />
First, we are most appreciative <strong>of</strong> <strong>the</strong> support and cooperation <strong>of</strong> <strong>the</strong> Vicksburg<br />
District, U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, under <strong>the</strong> direction <strong>of</strong> Col. Samuel P. Coll<strong>in</strong>s, Jr.,<br />
District Eng<strong>in</strong>eer. Mr. St. Clair Thompson, Contract<strong>in</strong>g Officer’s Representative and Ms.<br />
Shelia Lewis, Archeologist <strong>in</strong> <strong>the</strong> Environmental Branch, worked closely with us dur<strong>in</strong>g<br />
<strong>the</strong> research design and field phases. Ms. Kate Yarbrough, Archeologist <strong>in</strong> <strong>the</strong> Environmental<br />
Branch, worked with us dur<strong>in</strong>g later stages <strong>of</strong> <strong>the</strong> project. Among o<strong>the</strong>r contributions,<br />
<strong>the</strong> Vicksburg District provided us with excellent, recent air photo series and with<br />
hydrographic survey charts <strong>of</strong> <strong>the</strong> 1800s and early 1900s which proved to be <strong>in</strong>valuable.<br />
A number <strong>of</strong> colleagues at <strong>Arkansas</strong> Archeological Survey worked directly <strong>in</strong> various<br />
phases <strong>of</strong> <strong>the</strong> project and added immeasurably to its results. Dr. Charles R. McGimsey<br />
III, Director and Pr<strong>in</strong>cipal Investigator, assisted our work <strong>in</strong> numerous ways and at<br />
many times. Mr. Frank Rackerby, Contract Adm<strong>in</strong>istrator, was responsible for plann<strong>in</strong>g,<br />
logistical, and personnel matters, and also worked closely with <strong>the</strong> pr<strong>in</strong>cipal author on<br />
recommendations for mitigation (Appendix C). Ms. Hester Davis, State Archeologist,<br />
visited field headquarters dur<strong>in</strong>g site survey work and conferred with us and U.S. Army<br />
Eng<strong>in</strong>eers personnel on particular problems encountered and resources studied <strong>in</strong> <strong>the</strong><br />
Felsenthal project area. Ms. Sylvia Medl<strong>in</strong>, Accountant, skillfully handled many crucial<br />
arrangements for hous<strong>in</strong>g and services.<br />
Dr. Sandra C. Parker, Systems Analyst, and Ms. Susan Houston, Research Assistant,<br />
performed a statistical analysis which appears <strong>in</strong> Chapter 6. Ms. Cathy Moore-Jansen,<br />
Registrar, and Ms. Carolyn Cox, Assistant Registrar, worked conscientiously with<br />
more than 1000 pages <strong>of</strong> site records.<br />
A special acknowledgement is due three colleagues who reviewed <strong>the</strong> draft report<br />
<strong>in</strong> detail and made expert recommendations for its improvement: Dr. Fred W. Limp, Assistant<br />
Director, Dr. Frank Schambach, station archeologist at Sou<strong>the</strong>rn <strong>Arkansas</strong> <strong>University</strong>,<br />
and Dr. Marv<strong>in</strong> Jeter, station archeologist at <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>, Monticello.<br />
Mr. James E. Duncan skillfully supervised f<strong>in</strong>al process<strong>in</strong>g and various artifact analyses<br />
at <strong>the</strong> Contract Laboratory <strong>in</strong> Fayetteville. The production staff is gratefully acknowledged<br />
for its skill and hard work dur<strong>in</strong>g report preparation: Ms. Mary Lynn Kennedy,<br />
editor; Ms. Gwen Hamilton, typist; Ms. Louise Mull<strong>in</strong>s, typist; Ms. Christ<strong>in</strong>a Spees, typist;<br />
Ms. Jane Kellett, illustrator; and Ms. Pam Ashford, photographer.<br />
xiii
The Felsenthal Project field crew endured very difficult wetland conditions and<br />
ma<strong>in</strong>ta<strong>in</strong>ed high standards <strong>of</strong> work. Staff members <strong>in</strong>clude Jeyne Bennett, Ka<strong>the</strong>r<strong>in</strong>e<br />
D<strong>in</strong>nel, Andrew Hayes, and Clark Rogers; crew members <strong>in</strong>clude Paul Baumann, Diane<br />
Carpenter, John Paul Lane, Sheri Irv<strong>in</strong> Lane, Douglas Marsh, Jan Morrill, and William<br />
Radisch. Crew members employed for a portion <strong>of</strong> <strong>the</strong> project, or who rendered volunteer<br />
assistance dur<strong>in</strong>g site test<strong>in</strong>g, <strong>in</strong>clude Michael Hambacher, Douglas McKay, Henry<br />
McKelway, and Redus Montgomery. All are now veterans <strong>of</strong> sou<strong>the</strong>rn floodpla<strong>in</strong> forest<br />
and backswamp.<br />
A number <strong>of</strong> local residents were helpful to particular aspects <strong>of</strong> <strong>the</strong> Felsenthal<br />
Project or were gracious hosts, as <strong>in</strong>dicated below:<br />
Mr. Chris Avery, Soil Scientist, Soil Conservation Service, El Dorado: project<br />
area soils <strong>in</strong>formation<br />
Mrs. Garland Brantley, Felsenthal: headquarters and laboratory space<br />
Mr. James Howe, Forester, Felsenthal National Wildlife Refuge: project area<br />
floral <strong>in</strong>formation<br />
Mr. Roland Johnson, Felsenthal: local historical <strong>in</strong>formation, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong><br />
tale <strong>of</strong> steamboat Lotawanna<br />
Mr. Bill Julian, Manager, Felsenthal National Wildlife Refuge: project area<br />
<strong>in</strong>formation<br />
Mr. Jack Lee, El Dorado: historical references and regional historical <strong>in</strong>formation<br />
Mr. Bud Poole, El Dorado: boat dock<strong>in</strong>g facilities and project area <strong>in</strong>formation<br />
Mr. John Robertson, Crossett: repair and servic<strong>in</strong>g <strong>of</strong> all-terra<strong>in</strong> vehicles and<br />
outboard motors<br />
Mr. W. C. Saunders, Strong: crew hous<strong>in</strong>g<br />
Mr. Ken Short, Felsenthal: local archeological site <strong>in</strong>formation<br />
Mr. James Waterston, Felsenthal: crew hous<strong>in</strong>g<br />
Mr. and Mrs. Ed White, Hamburg: regional archeological site <strong>in</strong>formation<br />
Mr. Jack C. Womble, Assistant Manager, Felsenthal National Wildlife Refuge:<br />
project area <strong>in</strong>formation<br />
xxiv
In addition to <strong>the</strong>se <strong>in</strong>dividuals, we are grateful to <strong>the</strong> group <strong>of</strong> pr<strong>of</strong>essionals and<br />
scientists who provided analyses or materially assisted certa<strong>in</strong> research efforts, <strong>in</strong>clud<strong>in</strong>g<br />
<strong>the</strong> follow<strong>in</strong>g:<br />
Mr. Mark E. Gordon, Department <strong>of</strong> Zoology, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>: mollusc<br />
identification<br />
Dr. Nancy G. McCartney, Curator <strong>of</strong> Botany, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong> Museum:<br />
wetland vegetation and ecology<br />
Mr. Charles Reeves, Federal Archives and Records Center: archival research<br />
Dr. E. Moye Rutledge, Department <strong>of</strong> Agronomy, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>:<br />
sediment particle size analysis<br />
Mr. Allen R. Saltus Jr., Archeological Research and Survey: magnetometry<br />
data for steamboat Lotawanna<br />
Dr. Wayne E. Sabbe, Soil Test<strong>in</strong>g and Diagnostic Laboratory, <strong>University</strong> <strong>of</strong><br />
<strong>Arkansas</strong>: soil test data<br />
Dr. Frank Schambach, <strong>Arkansas</strong> Archeological Survey station archeologist at<br />
Sou<strong>the</strong>rn <strong>Arkansas</strong> <strong>University</strong>: regional prehistory and ceramics<br />
Mr. David W. Stahle, Research Assistant, <strong>Arkansas</strong> Archeological Survey:<br />
wood and charcoal identification<br />
Dr. Jerry J. Stipp, Beta-Analytic, Inc.: radiocarbon assays<br />
Dr. Murray Tamers, Beta-Analytic, Inc.: radiocarbon assays<br />
F<strong>in</strong>ally, <strong>the</strong> pr<strong>in</strong>cipal author expresses his thanks to <strong>the</strong> contributors to this volume,<br />
Mr. John H. House and Ms. Beverly Watk<strong>in</strong>s, Research Assistants at <strong>Arkansas</strong><br />
Archeological Survey, for <strong>the</strong>ir comprehensive reviews <strong>of</strong> prehistory and history <strong>in</strong><br />
<strong>the</strong> Felsenthal region (Chapters 3 and 4). Their <strong>in</strong>sights are a welcome addition to <strong>the</strong><br />
emerg<strong>in</strong>g perspective <strong>of</strong> human adaptation <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland environment,<br />
as presented <strong>in</strong> <strong>the</strong> follow<strong>in</strong>g pages.<br />
xxv
chapter 1<br />
<strong>in</strong>troduction to <strong>the</strong> felsenthAl project<br />
<strong>the</strong> felsenthAl contrAct And its reseArch potentiAl<br />
This report describes diverse results <strong>of</strong> a large scale <strong>in</strong>terdiscipl<strong>in</strong>ary, archeological<br />
research and cultural resource management project <strong>in</strong> sou<strong>the</strong>rn <strong>Arkansas</strong>. There is little<br />
question that we are tell<strong>in</strong>g <strong>the</strong> middle <strong>of</strong> a story, s<strong>in</strong>ce our efforts were preceded by sporadic,<br />
but important, prior <strong>in</strong>vestigations (mostly <strong>in</strong> <strong>the</strong> last decade), and s<strong>in</strong>ce <strong>the</strong> most fruitful<br />
archeological research and management efforts <strong>in</strong> this region lie ahead. One purpose <strong>of</strong> this<br />
<strong>in</strong>troductory chapter is to alert <strong>the</strong> reader to important data and conclusions that will follow.<br />
Certa<strong>in</strong>ly <strong>the</strong> boldest conclusion is this—prehistoric and historic floodpla<strong>in</strong> sites <strong>in</strong> <strong>the</strong><br />
Felsenthal Navigation Pool lie with<strong>in</strong>, and comprise a portion <strong>of</strong>, one <strong>of</strong> <strong>the</strong> richest and least<br />
disturbed archeological zones <strong>in</strong> North America.<br />
The contract for cultural resources survey and evaluation (DACW 38-79-C-0002) <strong>in</strong> <strong>the</strong><br />
Felsenthal Navigation Pool was awarded by <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg<br />
District, to <strong>the</strong> <strong>Arkansas</strong> Archeological Survey <strong>in</strong> October 1978. By authority <strong>of</strong> <strong>the</strong> River and<br />
Harbor Act <strong>of</strong> July 1960, <strong>the</strong> Corps is modify<strong>in</strong>g <strong>the</strong> Ouachita River to provide a 9-foot (2.7<br />
m) deep navigation channel below Camden, <strong>Arkansas</strong>. Among four new 84 x 600 foot (25.6<br />
x 182.9 m) locks and dams to be constructed on <strong>the</strong> river, Felsenthal and Calion locks and<br />
dams <strong>in</strong> <strong>Arkansas</strong> are scheduled for completion <strong>in</strong> 1982 and 1983 respectively. (Our contract<br />
was concerned with both projects, but only Felsenthal Navigation Pool’s cultural resources<br />
and related studies are reported here; see Kelley 1981.) The Felsenthal Project Scope <strong>of</strong><br />
Services (Appendix D) requires all <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g: (1) a research design, (2) background<br />
research, (3) a check for exist<strong>in</strong>g or proposed National Register properties, (4) <strong>in</strong>tensive field<br />
survey, (5) test<strong>in</strong>g <strong>of</strong> selected sites for extent and significance, (6) evaluation <strong>of</strong> National Register<br />
eligibility for all properties, (7) proposed mitigation program, (8) recommendations for<br />
<strong>in</strong>terpretive displays and, <strong>of</strong> course, (9) a comprehensive report <strong>of</strong> <strong>the</strong>se <strong>in</strong>vestigations. All<br />
<strong>of</strong> <strong>the</strong>se requirements are addressed <strong>in</strong> this report <strong>in</strong> subsequent chapters and appendixes<br />
to <strong>the</strong> best <strong>of</strong> our ability. The table <strong>of</strong> contents is very detailed with respect to where <strong>the</strong>se<br />
k<strong>in</strong>ds <strong>of</strong> <strong>in</strong>formation are presented and discussed.<br />
Fieldwork was <strong>in</strong>itiated by <strong>the</strong> <strong>Arkansas</strong> Archeological Survey <strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area <strong>in</strong> mid-August and term<strong>in</strong>ated <strong>in</strong> early December 1979. Our field season had to be<br />
delayed considerably until floodwaters subsided and was <strong>the</strong>n curtailed by ris<strong>in</strong>g flood-
2 Hemm<strong>in</strong>gs<br />
waters <strong>in</strong> late November. Overlapp<strong>in</strong>g site survey and test<strong>in</strong>g stages are described <strong>in</strong> Chapters<br />
5 and 6. One aspect <strong>of</strong> our work is repeatedly emphasized <strong>in</strong> this report; <strong>the</strong> project area<br />
lies with<strong>in</strong> an overflow bottomland and, as <strong>in</strong>dicated above, is subject to annual, prolonged<br />
<strong>in</strong>undation. This environmental dimension has affected <strong>the</strong> project at every stage—from<br />
research design through performance <strong>of</strong> fieldwork, analysis <strong>of</strong> site data, and proposal for<br />
mitigation measures. The 1<strong>17</strong> floodpla<strong>in</strong> sites we <strong>in</strong>vestigated (among a total <strong>of</strong> 144 sites<br />
recorded) are <strong>the</strong> most lowly<strong>in</strong>g archeological sites yet recorded <strong>in</strong> <strong>the</strong> state <strong>of</strong> <strong>Arkansas</strong><br />
(see discussion <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland <strong>in</strong> Chapter 2). Because <strong>of</strong> flood conditions, <strong>the</strong><br />
floodpla<strong>in</strong> here has never been developed for agriculture and no floodpla<strong>in</strong> site has ever<br />
been plowed; also, to <strong>the</strong> best <strong>of</strong> our knowledge, <strong>the</strong>se sites, because <strong>the</strong>y are so unobtrusive,<br />
have never been collected or dug <strong>in</strong>to by unauthorized <strong>in</strong>dividuals. Therefore our site<br />
survey and test<strong>in</strong>g methods, and especially our logistical arrangements, were designed for<br />
unusual preservation and wetlands conditions.<br />
Moreover, <strong>the</strong> k<strong>in</strong>ds <strong>of</strong> cultural resources we deal with most fully <strong>in</strong> this report are<br />
prehistoric, seasonal, specialized extractive camps that go under several meters <strong>of</strong> water<br />
every year (Chapter 9). If <strong>the</strong>se floodpla<strong>in</strong> sites are not unique, <strong>the</strong>y are certa<strong>in</strong>ly members<br />
<strong>of</strong> a class <strong>of</strong> human settlements vastly underrepresented <strong>in</strong> current archeological studies.<br />
Based on site survey and test<strong>in</strong>g data from <strong>the</strong> Felsenthal Project area, we believe <strong>the</strong>se<br />
small extractive camps have high potential for regional research and broad anthropological<br />
problems (see also Research Design Outl<strong>in</strong>e <strong>in</strong> Appendix C).<br />
In Chapter 2 we emphasize ecological and geohydrological characteristics <strong>of</strong> this<br />
region and <strong>in</strong>terrelationships with past human subsistence and settlement. The seasonal<br />
<strong>in</strong>undation (hydroperiod) <strong>in</strong> this lowland and its <strong>in</strong>tricate mosaic <strong>of</strong> aquatic and riparian<br />
communities (which disperse or concentrate certa<strong>in</strong> food resources) are keys to regional<br />
subsistence-settlement systems. However we identify a press<strong>in</strong>g need for conjunctive,<br />
environmental, archeological study <strong>in</strong> upland terra<strong>in</strong> <strong>of</strong> <strong>the</strong> region. Also <strong>the</strong> floodpla<strong>in</strong> and<br />
river<strong>in</strong>e features enumerated <strong>in</strong> Chapter 2 are <strong>in</strong>dicators <strong>of</strong> a dynamic geomorphic history<br />
which has barely been exam<strong>in</strong>ed by qualified scientists. The po<strong>in</strong>t emphasized here is<br />
that great research potential exists for regional cultural-environmental study <strong>in</strong> <strong>the</strong> <strong>Grand</strong><br />
<strong>Marais</strong> Lowland and <strong>the</strong> Felsenthal archeological region.<br />
elevation constra<strong>in</strong>t<br />
The Felsenthal Project Scope <strong>of</strong> Services def<strong>in</strong>ed <strong>the</strong> navigation pool and <strong>the</strong> study<br />
area as beg<strong>in</strong>n<strong>in</strong>g “at <strong>the</strong> Felsenthal Lock and Dam [and cont<strong>in</strong>u<strong>in</strong>g] upstream to cover all<br />
areas below <strong>the</strong> 65-foot contour l<strong>in</strong>e” (Appendix D). The <strong>Arkansas</strong> Archeological Survey’s<br />
proposal (1978) <strong>in</strong>terpreted this to mean that <strong>the</strong> Scope <strong>of</strong> Services “specifically excludes any<br />
land adjacent to or with<strong>in</strong> <strong>the</strong> navigation pool that is at or above <strong>the</strong> 65-foot contour <strong>in</strong>terval.”<br />
In po<strong>in</strong>t <strong>of</strong> fact, this constra<strong>in</strong>t <strong>in</strong>itially recognized by both parties to <strong>the</strong> contract, has<br />
caused numerous problems dur<strong>in</strong>g all phases <strong>of</strong> <strong>the</strong> project. Although we have discussed<br />
<strong>the</strong>se problems with <strong>the</strong> Corps <strong>of</strong> Eng<strong>in</strong>eers on several occasions and reached <strong>in</strong>formal
Introduction 3<br />
agreement (especially dur<strong>in</strong>g a field conference October 15, 1979), we would like to summarize<br />
<strong>the</strong>m here as an <strong>in</strong>troduction to many scientific and management considerations <strong>of</strong><br />
floodpla<strong>in</strong> elevation that appear throughout this report.<br />
1. Site survey transects crossed and recrossed terra<strong>in</strong> above 65 feet (19.8 m)<br />
because it is <strong>in</strong>tricately <strong>in</strong>terspersed with lower terra<strong>in</strong> and water courses;<br />
accord<strong>in</strong>g to accepted archeological research standards, we recorded sites<br />
above 65 feet when encountered, as well as those below 65 feet (Chapter 5).<br />
2. Site survey emphasized riverbank exposures where frequently artifacts and<br />
debris were discovered from <strong>the</strong> upper portion <strong>of</strong> <strong>the</strong> bankl<strong>in</strong>e (above 65<br />
feet) to river level at 62 feet; accord<strong>in</strong>g to accepted archeological research<br />
standards, we attempted to locate and record <strong>in</strong> situ cultural levels, whe<strong>the</strong>r<br />
or not above 65 feet (Chapter 5).<br />
3. Site survey and site test<strong>in</strong>g activities del<strong>in</strong>eated sites <strong>in</strong> which multiple<br />
occupation levels occurred above and below 65 feet elevation; <strong>in</strong> <strong>the</strong> most<br />
significant stratified site (3AS329) elevation necessarily began above and<br />
term<strong>in</strong>ated below 65 feet; <strong>in</strong> o<strong>the</strong>r cases an occupation level was recorded<br />
above 65 feet, but a contemporary feature extended below 65 feet (Chapter<br />
6).<br />
4. Because <strong>of</strong> natural downvalley slope <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>, a site at 70 feet elevation<br />
<strong>in</strong> <strong>the</strong> nor<strong>the</strong>rn part <strong>of</strong> our project area could be ecologically and geohydrologically<br />
identical to a site at 65 foot <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn part <strong>of</strong> our project<br />
area; however, research design and field methodology determ<strong>in</strong>ed that <strong>the</strong><br />
70-foot site would not likely be discovered unless well exposed <strong>in</strong> a riverbank<br />
(Chapter 2).<br />
5. The impacts <strong>of</strong> <strong>the</strong> Felsenthal Lock and Dam project and <strong>of</strong> <strong>the</strong> navigation<br />
and seasonally raised wildlife pools do not stop at <strong>the</strong> 65-foot contour, and<br />
we have attempted to assess a wide range <strong>of</strong> impacts <strong>in</strong> Appendix C.<br />
These detailed comments lead us to propose that <strong>the</strong> floodpla<strong>in</strong> itself (terra<strong>in</strong> <strong>in</strong>undated<br />
annually) is <strong>the</strong> m<strong>in</strong>imal, mean<strong>in</strong>gful, ecological, geomorphic, and cultural unit <strong>of</strong> study<br />
<strong>in</strong> this area, and not some portion <strong>of</strong> it arbitrarily def<strong>in</strong>ed by a horizontal plane (<strong>the</strong> navigation<br />
pool). While we respect this eng<strong>in</strong>eer<strong>in</strong>g feature and consequent contract specification,<br />
we have also approached our floodpla<strong>in</strong> work with <strong>the</strong> logic <strong>of</strong> science.<br />
Def<strong>in</strong>ition <strong>of</strong> Study Area<br />
Four geographical units <strong>of</strong> study are referred to frequently <strong>in</strong> this report, and <strong>the</strong>ir<br />
extent and relationship should be made clear at <strong>the</strong> outset.
4 Hemm<strong>in</strong>gs<br />
1. Felsenthal Project area (or simply “project area”) <strong>in</strong>cludes all terra<strong>in</strong> below<br />
65 feet elevation <strong>in</strong> <strong>the</strong> Ouachita and Sal<strong>in</strong>e River floodpla<strong>in</strong>s above Felsenthal<br />
Lock and Dam. For practical purposes this terra<strong>in</strong> lies with<strong>in</strong> <strong>the</strong> new<br />
navigation pool and surrounded by Felsenthal National Wildlife Refuge,<br />
and has been divided <strong>in</strong>to two subareas by <strong>the</strong> U.S. Highway 82 cross<strong>in</strong>g.<br />
Figure 20 <strong>in</strong> Chapter 5 best portrays <strong>the</strong> Felsenthal Project area.<br />
2. Ouachita and Sal<strong>in</strong>e River floodpla<strong>in</strong> (or simply “floodpla<strong>in</strong>”) <strong>in</strong>cludes all<br />
terra<strong>in</strong> and water courses with<strong>in</strong> <strong>the</strong> project area def<strong>in</strong>ed above which are<br />
annually overflowed. This <strong>in</strong>cludes bottomland terra<strong>in</strong> at about 62-75 feet<br />
elevation; we occasionally refer to such terra<strong>in</strong> as “overflow bottom.” It consists<br />
primarily <strong>of</strong> Holocene floodpla<strong>in</strong> as mapped by Fleetwood (1969), but<br />
<strong>in</strong>cludes also m<strong>in</strong>or extensions <strong>of</strong> Deweyville 3 Terrace (Figure 3 <strong>in</strong> Chapter<br />
2). The floodpla<strong>in</strong> so def<strong>in</strong>ed appears to be a useful cultural-environmental<br />
study unit for <strong>the</strong> past 3,000 years at least. Throughout this report we refer<br />
to floodpla<strong>in</strong> environments and floodpla<strong>in</strong> sites.<br />
3. Felsenthal National Wildlife Refuge (or “study area” as <strong>in</strong>dicated <strong>in</strong> this report)<br />
is an irregular area <strong>of</strong> 26,305 ha (65,000 acres) established and operated<br />
by <strong>the</strong> U.S. Fish and Wildlife Service <strong>in</strong> sou<strong>the</strong>rn <strong>Arkansas</strong>. Figure 1 <strong>in</strong>dicates<br />
<strong>the</strong> boundaries and some <strong>of</strong> <strong>the</strong> features with<strong>in</strong> <strong>the</strong> refuge, and also its<br />
proximity to <strong>the</strong> Louisiana border (see discussion <strong>of</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland<br />
<strong>in</strong> Chapter 2). Because <strong>the</strong> refuge encompasses <strong>the</strong> range <strong>of</strong> environments<br />
and <strong>of</strong> upland and lowland cultural resources <strong>in</strong> <strong>the</strong> region, we have made<br />
<strong>in</strong>tensive use <strong>of</strong> data perta<strong>in</strong><strong>in</strong>g to all refuge areas, although current field<br />
operations focused on <strong>the</strong> project area and floodpla<strong>in</strong>. The refuge as a study<br />
area is a research concept used <strong>in</strong> this report, and is not <strong>the</strong> area <strong>of</strong> field<br />
study def<strong>in</strong>ed <strong>in</strong> <strong>the</strong> contract.<br />
4. Felsenthal archeological region (or simply <strong>the</strong> “region”) is def<strong>in</strong>ed by<br />
Schambach (1979; Rol<strong>in</strong>gson and Schambach 1981:103-105) as follows:<br />
it lies with<strong>in</strong> a zone <strong>of</strong> <strong>the</strong> Lower Mississippi Valley archeological area.... The<br />
nor<strong>the</strong>rn [boundary] on <strong>the</strong> Ouachita River...I put at Camden, <strong>Arkansas</strong>...<br />
on good archeological and ecological evidence....The sou<strong>the</strong>rn boundary...<br />
would be well down <strong>the</strong> Ouachita Valley <strong>in</strong> Louisiana, most probably between<br />
Ouachita City and Monroe....All major and m<strong>in</strong>or tributaries <strong>of</strong> <strong>the</strong><br />
Ouachita River between Ouachita City and Camden, except Bayou Bartholomew,<br />
were controlled all <strong>the</strong> way to <strong>the</strong>ir headwaters by Felsenthal region<br />
peoples.<br />
Figure 13 <strong>in</strong> Chapter 3 best portrays <strong>the</strong> extent <strong>of</strong> this region and <strong>the</strong> distribution<br />
<strong>of</strong> some important prehistoric sites. Schambach is <strong>of</strong> course follow-
Figure 1. Location and features <strong>of</strong> Felsenthal National Wildlife Refuge <strong>in</strong> sou<strong>the</strong>rn <strong>Arkansas</strong>.<br />
5
6 Hemm<strong>in</strong>gs<br />
<strong>in</strong>g <strong>the</strong> concept <strong>of</strong> an archeological region as a spatial division, where <strong>the</strong><br />
“relationship between culture and a highly characteristic environment has<br />
been particularly close throughout <strong>the</strong> span <strong>of</strong> <strong>the</strong> archaeological period”<br />
(Willey and Phillips 1958:19).<br />
deVelopMent <strong>of</strong> reseArch design<br />
S<strong>in</strong>ce <strong>the</strong> research design is <strong>in</strong>cluded <strong>in</strong> Appendix D, only its salient characteristics will<br />
be noted here. First, it should be acknowledged that our research design takes advantage <strong>of</strong><br />
an earlier stimulat<strong>in</strong>g proposal for archeological research <strong>in</strong> <strong>the</strong> Felsenthal region by Raab<br />
(1976a), who <strong>in</strong> turn drew upon nearly a decade <strong>of</strong> varied research activity by <strong>the</strong> <strong>Arkansas</strong><br />
Archeological Survey. However, where Raab assumed that a major segment <strong>of</strong> <strong>the</strong> Ouachita<br />
dra<strong>in</strong>age bas<strong>in</strong> (i.e., <strong>the</strong> Felsenthal region) would constitute <strong>the</strong> research area, we have necessarily<br />
adapted our research design to floodpla<strong>in</strong> sites and environments (and to elevation<br />
constra<strong>in</strong>ts noted earlier).<br />
Second, <strong>the</strong> strategies for site survey and test<strong>in</strong>g were conservatively devised for <strong>the</strong><br />
formidable conditions <strong>of</strong> bottomland hardwood and swamp forest, <strong>in</strong>terspersed with river<br />
channels, bayous, open lakes, and backswamps (Chapters 5, 6). Our research design also<br />
anticipated burial <strong>of</strong> sites <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> topstratum and marked “<strong>in</strong>visibility” <strong>of</strong> sites<br />
buried deeply and those below water table or river stage. At several po<strong>in</strong>ts <strong>in</strong> this report we<br />
po<strong>in</strong>t out that all prehistoric floodpla<strong>in</strong> sites, and probably many historic sites, are buried<br />
through all or most <strong>of</strong> <strong>the</strong>ir extent <strong>in</strong> our project area. Under <strong>the</strong>se conditions, we cannot<br />
claim to have located all cultural resources, although we elicit strong evidence for where<br />
buried sites occur frequently and where <strong>the</strong>y do not (Chapter 5).<br />
Schedule <strong>of</strong> Investigations<br />
The Felsenthal Project was organized as a series <strong>of</strong> overlapp<strong>in</strong>g tasks with numerous<br />
pathways for <strong>in</strong>formation flow between research units (Figure 2). These tasks may be summarized<br />
as follows:<br />
Task 1—Background Research<br />
a. local and regional prehistory (Chapter 3)<br />
b. local and regional history (Chapter 4)<br />
c. regional ecology and geohydrology (Chapter 2)<br />
d. relevant <strong>the</strong>oretical and methodological perspectives (Appendix D)<br />
e. preparation <strong>of</strong> project research design (Appendix D)<br />
Task 2—Implementation <strong>of</strong> Site Survey<br />
a. transects (Chapter 5)<br />
b. bankl<strong>in</strong>e surveys (Chapter 5)<br />
c. <strong>in</strong>itiation <strong>of</strong> field laboratory operations (Chapter 5)<br />
d. <strong>in</strong>itial analysis <strong>of</strong> site locational data (Chapter 5)<br />
Task 3—Implementation <strong>of</strong> Site Test<strong>in</strong>g<br />
a. ongo<strong>in</strong>g site survey (Chapter 5)
Introduction 7<br />
Figure 2. Flow <strong>of</strong> <strong>in</strong>formation between research units from research design to proposed<br />
mitigation program for <strong>the</strong> Felsenthal Project (after <strong>Arkansas</strong> Archeological Survey<br />
1978:Figure 2).
8 Hemm<strong>in</strong>gs<br />
b. <strong>in</strong>itial assessment <strong>of</strong> potential significance and selection <strong>of</strong> sites for test<strong>in</strong>g<br />
(Chapter 6)<br />
c. <strong>in</strong>itiation <strong>of</strong> test<strong>in</strong>g to establish site significance (Chapter 6)<br />
d. ongo<strong>in</strong>g field laboratory operation (Chapter 5)<br />
e. <strong>in</strong>itial analysis <strong>of</strong> site contextual data and research potential (Chapter 6)<br />
Task 4—Detailed Analysis <strong>of</strong> Project Research Data<br />
a. lithic and ceramic samples (Chapter 8)<br />
b. soil, sediment, and chronometric samples (Chapter 7)<br />
c. faunal and floral samples (Appendix B)<br />
d. o<strong>the</strong>r laboratory analyses (Chapter 8)<br />
e. <strong>in</strong>tegration <strong>of</strong> background and analytical data (Chapters 5-9)<br />
Task 5—Evaluation <strong>of</strong> Site Significance<br />
a. local and regional criteria (Appendix C)<br />
b. National Register criteria (Appendix C)<br />
c. preparation <strong>of</strong> recommended program <strong>of</strong> mitigation (Appendix C)<br />
Task 6—Preparation <strong>of</strong> Draft Report<br />
Schedule <strong>of</strong> Hypo<strong>the</strong>ses<br />
The research design (Appendix D) provides a broad, flexible orientation by identify<strong>in</strong>g<br />
four problem doma<strong>in</strong>s and present<strong>in</strong>g a series <strong>of</strong> eight hypo<strong>the</strong>ses (and several subhypo<strong>the</strong>ses)<br />
with<strong>in</strong> <strong>the</strong>se problem doma<strong>in</strong>s. These hypo<strong>the</strong>ses are conservatively framed <strong>in</strong><br />
regard to <strong>the</strong> k<strong>in</strong>ds <strong>of</strong> data recovery expected and exigencies <strong>of</strong> site survey and test<strong>in</strong>g <strong>in</strong> <strong>the</strong><br />
overflow bottomland environment. In <strong>the</strong> research design we emphasized <strong>the</strong> near absence<br />
<strong>of</strong> current, systematic, environmental, or archeological studies <strong>in</strong> <strong>the</strong> project area and region.<br />
The problem doma<strong>in</strong>s and hypo<strong>the</strong>ses are stated <strong>in</strong> full below, with a notation <strong>of</strong> where test<br />
data and discussion <strong>of</strong> hypo<strong>the</strong>ses appear <strong>in</strong> <strong>the</strong> text <strong>of</strong> this report. (No data were obta<strong>in</strong>ed<br />
for Hypo<strong>the</strong>sis 5 below.)<br />
Problem Doma<strong>in</strong> 1: Settlement Characteristics (Chapter 6)<br />
H 1 If subsistence strategies employed wild plant and animal resources, impermanent<br />
specialized extractive sites would implement <strong>the</strong>se strategies <strong>in</strong> true floodpla<strong>in</strong><br />
environments.<br />
Test Implications: Recognition <strong>of</strong> specialized extractive sites as settlement types, dist<strong>in</strong>ct<br />
from base settlements.<br />
Data Requirements: Floodpla<strong>in</strong> site assemblages with high ratio <strong>of</strong> extractive/ma<strong>in</strong>tenance<br />
tool kits; faunal or floral <strong>in</strong>dicators <strong>of</strong> seasonality and procurement strategy;<br />
absence <strong>of</strong> permanent structures for habitation, storage, or burial.<br />
Problem Doma<strong>in</strong> 2: Locational Characteristics (Chapter 5)<br />
H 2 If specialized extractive sites were established for resource procurement, <strong>the</strong>se<br />
sites would be located for most efficient acquisition and delivery to base settlements.
Introduction 9<br />
Subhypo<strong>the</strong>ses: A. Such sites will correlate with specific resource zones.<br />
B. Such sites will correlate with access routes to base settlements.<br />
C. Such sites will correlate with both factors.<br />
D. None <strong>of</strong> <strong>the</strong> above, but o<strong>the</strong>r unknown factors.<br />
Test Implications: Significant correlations obta<strong>in</strong>ed for location factors; recognition <strong>of</strong><br />
contemporary base settlements, access routes, and resource zones.<br />
Data Requirements: Map distributions; measures <strong>of</strong> spatial relationships; crossdat<strong>in</strong>g.<br />
Problem Doma<strong>in</strong> 3: Culture History (Chapters 5, 8)<br />
H 3 If upland mound complexes represent a dom<strong>in</strong>ant regional culture, numerous<br />
sites <strong>of</strong> this culture should also be identified <strong>in</strong> <strong>the</strong> lowland sett<strong>in</strong>g.<br />
Test Implications: Discovery <strong>of</strong> culturally related sites <strong>in</strong> study area.<br />
Data Requirements: Common artifact styles, especially ceramics.<br />
H 4 If aborig<strong>in</strong>al Felsenthal populations occupied a frontier region, local styles <strong>of</strong> artifacts<br />
and settlement rema<strong>in</strong>s will occur mixed with extraregional styles.<br />
Test Implications: Recognition <strong>of</strong> local and alien styles.<br />
Data Requirements: Typological and comparative analyses disclos<strong>in</strong>g <strong>in</strong>troduced<br />
styles, especially <strong>in</strong> ceramics.<br />
H 5 If early historical sites are present <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>, <strong>the</strong>y were occupied briefly for<br />
hunt<strong>in</strong>g and trad<strong>in</strong>g activity (before <strong>17</strong>80).<br />
Test Implications: Discovery <strong>of</strong> sites with simple technological assemblages.<br />
Data Requirements: Tools or discards primarily associated with hunt<strong>in</strong>g; trade items;<br />
datable objects; absence <strong>of</strong> permanent structures.<br />
H 6 If later historical sites are present <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>, <strong>the</strong>y were occupied <strong>in</strong> association<br />
with river<strong>in</strong>e activity and extractive <strong>in</strong>dustry (after <strong>17</strong>80).<br />
Test Implications: Discovery <strong>of</strong> sites with complex technological assemblages.<br />
Data Requirements: Tools or components associated with eng<strong>in</strong>es.
10 Hemm<strong>in</strong>gs<br />
Problem Doma<strong>in</strong> 4: Site Transformations (Chapter 7)<br />
H 7 If geologically Recent topstratum deposits have aggraded 20 feet or more <strong>in</strong> <strong>the</strong><br />
Ouachita Valley, methods <strong>of</strong> <strong>in</strong>tensive survey <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> will fail to sample<br />
adequately early occupation sites.<br />
Test Implications: Recognition <strong>of</strong> <strong>the</strong> age <strong>of</strong> sediments drowned and exposed; discovery<br />
<strong>of</strong> buried sites <strong>in</strong> datable context.<br />
Data Requirements: Significant disparity <strong>in</strong> observed and expected occurrence <strong>of</strong><br />
early sites; geochronological controls.<br />
H 8 If differential preservation <strong>of</strong> organic matter occurs with<strong>in</strong> or between sites, soil<br />
acidity has adversely affected <strong>the</strong> archeological context.<br />
Test Implications: Recognition <strong>of</strong> preservation gradients by sampl<strong>in</strong>g soil and refuse;<br />
demonstrable loss <strong>of</strong> organic matter.<br />
Data Requirements: Translation <strong>of</strong> preservation gradient to a numerical scale; significant<br />
positive correlation with pH (low soil pH—low degree <strong>of</strong> preservation).<br />
felsenthAl project personnel<br />
The follow<strong>in</strong>g list describes key <strong>in</strong>dividuals <strong>in</strong>volved <strong>in</strong> <strong>the</strong> Felsenthal Project from<br />
plann<strong>in</strong>g <strong>in</strong> 1978 to draft report preparation <strong>in</strong> 1981 (see also Acknowledgments for many<br />
o<strong>the</strong>r contributors). This list is presented here as an additional perspective on Felsenthal<br />
Project organization and expertise.<br />
Charles R. McGimsey III (Ph.D., Harvard, 1958), Director <strong>of</strong> <strong>Arkansas</strong> Archeological<br />
Survey, was Pr<strong>in</strong>cipal Investigator for <strong>the</strong> Felsenthal Project. He was similarly <strong>in</strong>volved <strong>in</strong> all<br />
earlier Survey research activities <strong>in</strong> <strong>the</strong> Felsenthal region. He bears overall responsibility for<br />
all adm<strong>in</strong>istrative and scientific aspects <strong>of</strong> <strong>the</strong> project. Dr. McGimsey is <strong>the</strong> author <strong>of</strong> several<br />
books and numerous articles <strong>in</strong> American archeology and cultural resource management.<br />
W. Fredrick Limp (Ph.D., Indiana, 1980), Assistant Director <strong>of</strong> <strong>Arkansas</strong> Archeological<br />
Survey, was <strong>in</strong>volved with <strong>the</strong> project as research coord<strong>in</strong>ator and pr<strong>in</strong>cipal reviewer <strong>of</strong> <strong>the</strong><br />
draft report. In <strong>the</strong>se capacities he contributed to many aspects <strong>of</strong> <strong>the</strong> research, especially<br />
to <strong>the</strong> proposed mitigation program <strong>in</strong> Appendix C. His research <strong>in</strong>terests and publications<br />
<strong>in</strong>clude development <strong>of</strong> data recovery techniques such as flotation, research design, and<br />
quantitative analyses <strong>in</strong> archeology.
Introduction 11<br />
Frank Rackerby (ABD, Northwestern, 1968) is Contract Adm<strong>in</strong>istrator at <strong>the</strong> <strong>Arkansas</strong><br />
Archeological Survey, and served as <strong>the</strong> critical l<strong>in</strong>k between <strong>the</strong> Survey and <strong>the</strong> project<br />
sponsor. He was deeply <strong>in</strong>volved <strong>in</strong> preparation <strong>of</strong> <strong>the</strong> proposal (Appendix D), complex<br />
schedul<strong>in</strong>g and logistical arrangements for survey and test<strong>in</strong>g, and personnel and fiscal<br />
matters. He has published various monographs and articles <strong>in</strong> Midwestern archeology and<br />
cultural resource management.<br />
E. Thomas Hemm<strong>in</strong>gs (Ph.D., Arizona, 1970), Associate Archeologist at <strong>the</strong> <strong>Arkansas</strong><br />
Archeological Survey, was Felsenthal Project Archeologist, developed <strong>the</strong> research design<br />
(Appendix D), supervised all field and field laboratory operations, supervised or conducted<br />
analyses <strong>of</strong> project data <strong>in</strong> <strong>the</strong> Contract Laboratory, and prepared this draft report. He also<br />
provided <strong>the</strong> <strong>in</strong>terdiscipl<strong>in</strong>ary dimension <strong>of</strong> <strong>the</strong> project at this stage by virtue <strong>of</strong> tra<strong>in</strong><strong>in</strong>g<br />
and experience <strong>in</strong> both environmental archeology and geology/geochronology. Hemm<strong>in</strong>gs<br />
directly supervised test<strong>in</strong>g <strong>of</strong> a deep stratified site on <strong>the</strong> Ouachita River, Marie Sal<strong>in</strong>e<br />
(3AS329), described <strong>in</strong> Chapter 6. Dr. Hemm<strong>in</strong>gs is <strong>the</strong> author <strong>of</strong> numerous monographs<br />
and articles <strong>in</strong> Sou<strong>the</strong>astern and Ohio Valley archeology, and is actively <strong>in</strong>volved <strong>in</strong> <strong>in</strong>terdiscipl<strong>in</strong>ary<br />
research <strong>in</strong> <strong>Arkansas</strong>.<br />
Beverly Watk<strong>in</strong>s (ABD, Auburn, 1980), Project Historian, performed historical background<br />
research which was essential to project plann<strong>in</strong>g and historic site survey work, and<br />
also prepared <strong>the</strong> regional historical summary <strong>in</strong> Chapter 4. Watk<strong>in</strong>s was notably successful<br />
<strong>in</strong> obta<strong>in</strong><strong>in</strong>g archival data and oral history perta<strong>in</strong><strong>in</strong>g to <strong>the</strong> steamboat Lotawanna (3UN153),<br />
which burned and sank <strong>in</strong> 1874 on <strong>the</strong> Ouachita River with<strong>in</strong> <strong>the</strong> project area. She is actively<br />
engaged <strong>in</strong> historical research on <strong>the</strong> n<strong>in</strong>eteenth century period <strong>in</strong> <strong>Arkansas</strong>.<br />
John H. House (B.A., <strong>Arkansas</strong>, 1972), Research Assistant at <strong>Arkansas</strong> Archeological<br />
Survey, has extensive research experience <strong>in</strong> <strong>the</strong> Sou<strong>the</strong>ast United States <strong>in</strong>clud<strong>in</strong>g <strong>Arkansas</strong><br />
and Louisiana. He recently prepared a syn<strong>the</strong>ses <strong>of</strong> certa<strong>in</strong> prehistoric periods <strong>in</strong> eastcentral<br />
and sou<strong>the</strong>ast <strong>Arkansas</strong> for <strong>the</strong> <strong>Arkansas</strong> State Plan (Davis 1981), and on this basis<br />
was <strong>in</strong>vited to contribute a review <strong>of</strong> prehistory <strong>in</strong> <strong>the</strong> Felsenthal region, which appears as<br />
Chapter 3 <strong>of</strong> this report.<br />
Marv<strong>in</strong> Jeter (Ph.D., Arizona State, 1977), <strong>Arkansas</strong> Archeological Survey station<br />
archeologist at <strong>the</strong> <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong> at Monticello, is an expert regional archeologist,<br />
and served as a project consultant. He visited ongo<strong>in</strong>g field operations on several occasions<br />
and participated <strong>in</strong> test<strong>in</strong>g <strong>of</strong> a significant Mississippi period extractive camp, False Indigo<br />
(3AS285), described <strong>in</strong> Chapter 6. Dr. Jeter has made many suggestions for improvement <strong>of</strong><br />
this report. He has published articles and monographs <strong>in</strong> Sou<strong>the</strong>astern and Southwestern<br />
archeology, and is currently active <strong>in</strong> prehistoric studies <strong>in</strong> sou<strong>the</strong>ast <strong>Arkansas</strong>.<br />
Clark Rogers (B.A., SUNY Geneseo, 1971) served as senior crew chief and was particularly<br />
concerned with all-terra<strong>in</strong>-vehicle and pedestrian survey transects, and with test<strong>in</strong>g on<br />
sites 3AS285, 3AS306, and 3BR78 (Chapter 6).
12 Hemm<strong>in</strong>gs<br />
Ka<strong>the</strong>r<strong>in</strong>e D<strong>in</strong>nel (M.A., Florida State, 1977) served as site survey crew chief and was<br />
particularly concerned with boat surveys on both <strong>the</strong> Ouachita and Sal<strong>in</strong>e rivers, and with<br />
test<strong>in</strong>g on sites 3AS286 and 3AS307 (Chapter 6). D<strong>in</strong>nel also held an appo<strong>in</strong>ted position as<br />
laboratory and research analyst follow<strong>in</strong>g <strong>the</strong> fieldwork stage, and contributed significantly<br />
to this report <strong>in</strong> all <strong>the</strong>se respects (Chapter 8).<br />
Andrew Hayes (B.A., Appalachian State, 1978) served as site survey crew chief and<br />
was particularly <strong>in</strong>volved <strong>in</strong> all-terra<strong>in</strong>-vehicle and pedestrian survey transects, and with<br />
test<strong>in</strong>g on sites 3BR58 and 3BR76 (Chapter 6).<br />
Jeyne Bennett (M.A., <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>, 1980) operated <strong>the</strong> field laboratory at<br />
Felsenthal and was concerned with process<strong>in</strong>g and <strong>in</strong>itial analysis <strong>of</strong> collections. She was<br />
additionally <strong>in</strong>volved with personnel and logistical record keep<strong>in</strong>g, and frequently served<br />
as communication l<strong>in</strong>k between field personnel and o<strong>the</strong>r <strong>in</strong>dividuals or agencies concerned<br />
with <strong>the</strong> Felsenthal Project.
chapter 2<br />
grAnd MArAis lowlAnd enVironMent<br />
In <strong>the</strong> West Gulf Coastal Pla<strong>in</strong> <strong>of</strong> sou<strong>the</strong>rn <strong>Arkansas</strong>, one <strong>of</strong> <strong>the</strong> most dist<strong>in</strong>ctive geographical<br />
features is <strong>the</strong> low trough <strong>of</strong> <strong>the</strong> Ouachita River Valley. At <strong>the</strong> <strong>Arkansas</strong>-Louisiana<br />
state l<strong>in</strong>e (about 33° N latitude) <strong>the</strong> floodpla<strong>in</strong> is 40 feet (12.2 m) lower than <strong>the</strong> alluvial<br />
ridge <strong>of</strong> <strong>the</strong> Mississippi River at <strong>the</strong> same latitude. This floodpla<strong>in</strong> corridor is <strong>the</strong> lowest<br />
terra<strong>in</strong> <strong>in</strong> <strong>the</strong> state <strong>of</strong> <strong>Arkansas</strong> and, not surpris<strong>in</strong>gly, shares many environmental characteristics<br />
with <strong>the</strong> Lower Mississippi Alluvial Valley and with large tributary rivers and floodbas<strong>in</strong>s<br />
<strong>in</strong> Louisiana.<br />
The sou<strong>the</strong>rn character or aff<strong>in</strong>ity <strong>of</strong> <strong>the</strong> Felsenthal Project area is exemplified by a<br />
small brake <strong>of</strong> bald cypress draped with Spanish moss just northwest <strong>of</strong> <strong>the</strong> U.S. Highway<br />
82 bridge across <strong>the</strong> Ouachita River (locally called Mouton Brake). This occurrence is essentially<br />
<strong>the</strong> nor<strong>the</strong>rn limit <strong>of</strong> Spanish moss <strong>in</strong> <strong>Arkansas</strong> (Tucker 1974). The journals and maps<br />
<strong>of</strong> Dunbar and Hunter who explored <strong>the</strong> Ouachita River <strong>in</strong> 1804-1805 make several references<br />
to <strong>the</strong> disappearance <strong>of</strong> “long moss/Tillandsia/above Latitude 33°” (Rowland 1930:309).<br />
Explorers, travelers, and especially geologists have commented on <strong>the</strong> low-ly<strong>in</strong>g floodpla<strong>in</strong><br />
and overflow conditions which prevail <strong>in</strong> <strong>the</strong> region <strong>of</strong> Felsenthal. We <strong>in</strong>troduce some<br />
<strong>of</strong> those comments here as environmental and historical background to later sections <strong>of</strong> this<br />
report.<br />
The face <strong>of</strong> <strong>the</strong> Country beg<strong>in</strong>s to change [above present-day Alabama<br />
Land<strong>in</strong>g, Louisiana]; <strong>the</strong> banks are low and steep, and <strong>the</strong> river generally deeper,<br />
and much contracted, be<strong>in</strong>g from 30 to 50 yards wide; this low Country is...<br />
liable to overflow 12 or 15 feet above <strong>the</strong> level <strong>of</strong> <strong>the</strong> land....It may be supposed<br />
that <strong>the</strong>re existed a great Lake with<strong>in</strong> <strong>the</strong> space now occupied by this alluvial<br />
tract....18 or 20 feet <strong>of</strong> soil perpendicular is yet want<strong>in</strong>g to render it a fit habitation<br />
for man; it appears never <strong>the</strong> less to be well peopled by <strong>the</strong> beasts <strong>of</strong> <strong>the</strong> forest....We<br />
now beg<strong>in</strong> to see quantities <strong>of</strong> water fowl (Wm. Dunbar, November 14,<br />
1804, <strong>in</strong> Rowland 1930:240).<br />
A slight movement along <strong>the</strong>...fault l<strong>in</strong>e near Alabama Land<strong>in</strong>g La., [caused]<br />
extreme swamp<strong>in</strong>g <strong>of</strong> <strong>the</strong> bottoms from that po<strong>in</strong>t to above <strong>the</strong> mouth <strong>of</strong> Bayou<br />
Moro <strong>in</strong> <strong>Arkansas</strong>....[Fault<strong>in</strong>g] has amounted to 25 feet, and <strong>the</strong> movement is so
14 Hemm<strong>in</strong>gs<br />
recent that, although occurr<strong>in</strong>g <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> <strong>of</strong> a large river, it has not been<br />
perceptibly obliterated by sedimentation....The persistence <strong>of</strong> this break must be<br />
regarded as an extreme illustration <strong>of</strong> <strong>the</strong> small amount <strong>of</strong> sediment carried by<br />
Ouachita River (Veatch 1906:59, 65).<br />
Much <strong>of</strong> <strong>the</strong> territory border<strong>in</strong>g <strong>the</strong> Ouachita is low-ly<strong>in</strong>g and subject to <strong>in</strong>undation,<br />
and probably <strong>in</strong> aborig<strong>in</strong>al times, as at present, was not occupied for permanent<br />
habitation....From just above Ouachita City for about seventy-five miles up<br />
by water, is an almost un<strong>in</strong>habited region (Moore 1909:16).<br />
Between <strong>the</strong> Sal<strong>in</strong>e River and Alabama Land<strong>in</strong>g, <strong>the</strong> floodpla<strong>in</strong> widens to about<br />
4 miles where it transects <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> lowland....The topographically<br />
depressed <strong>Grand</strong> <strong>Marais</strong> area is not structurally controlled, but ra<strong>the</strong>r is due to<br />
pond<strong>in</strong>g <strong>of</strong> <strong>the</strong> lower portion <strong>of</strong> <strong>the</strong> river brought about dur<strong>in</strong>g <strong>the</strong> formation <strong>of</strong><br />
<strong>the</strong> relatively higher <strong>Arkansas</strong> River meander belt near Ouachita City (Fleetwood<br />
1969:Figure 1, fac<strong>in</strong>g page).<br />
Alluvial drown<strong>in</strong>g...is manifested by <strong>the</strong> large, frequently <strong>in</strong>undated, swampy<br />
lowland or flood bas<strong>in</strong> called <strong>Grand</strong> <strong>Marais</strong>...between Felsenthal and Alabama<br />
Land<strong>in</strong>g, Louisiana. With<strong>in</strong> <strong>the</strong> flood bas<strong>in</strong> <strong>the</strong> Ouachita River natural levees are<br />
poorly developed, <strong>the</strong> floodpla<strong>in</strong> relief is m<strong>in</strong>imal, and <strong>the</strong> river has abandoned<br />
meander belts more frequently than anywhere else....Fault<strong>in</strong>g was postulated as<br />
<strong>the</strong> cause <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong>; however nei<strong>the</strong>r physiographic nor subsurface<br />
evidence has ever been discovered to substantiate this (Saucier and Fleetwood<br />
1970:878).<br />
Fleetwood and Saucier have thus <strong>in</strong>troduced <strong>the</strong> concept <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland as a<br />
geomorphic division <strong>of</strong> <strong>the</strong> Ouachita Bas<strong>in</strong> and Lower Mississippi Alluvial Valley. The use<br />
<strong>of</strong> <strong>the</strong> historical placename “<strong>Grand</strong> <strong>Marais</strong>” is highly appropriate for this lowland, s<strong>in</strong>ce <strong>in</strong><br />
Mississippi Valley French a “marais” is more likely to refer to an open lake than a marsh or<br />
swamp (McDermott 1941:98). A pr<strong>of</strong>usion <strong>of</strong> oxbow lakes and abandoned river courses, reoccupied<br />
by bayous, is a salient characteristic <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> <strong>in</strong> <strong>the</strong> Felsenthal Project area.<br />
Our concept <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> is both geohydrological and ecological. The dist<strong>in</strong>ctive<br />
lowland flora and fauna have <strong>in</strong>teracted with and pr<strong>of</strong>oundly <strong>in</strong>fluenced resident human<br />
populations over <strong>the</strong> past few thousand years. In advance <strong>of</strong> <strong>the</strong> summary <strong>of</strong> regional<br />
prehistory <strong>in</strong> Chapter 3 and site data presented later <strong>in</strong> this report, we note that a dense<br />
zone <strong>of</strong> prehistoric occupation sites co<strong>in</strong>cides with <strong>the</strong> elevated outer edge (terrace) adjacent<br />
to <strong>the</strong> floodpla<strong>in</strong>, and that an array <strong>of</strong> small impermanent camps co<strong>in</strong>cides with channel<br />
marg<strong>in</strong>s with<strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>. Many sites <strong>in</strong> both zones are firmly attributed to <strong>the</strong> Mississippi<br />
period (A.D. 1100-<strong>17</strong>00), but several millenia <strong>of</strong> earlier human occupation sites are also<br />
well represented. It is clear that <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland is <strong>the</strong> axis and core environment
Environment 15<br />
<strong>of</strong> <strong>the</strong> Felsenthal region, and that procurement <strong>of</strong> lowland resources was a major regional<br />
strategy <strong>of</strong> prehistoric human populations.<br />
sou<strong>the</strong>Astern floodplA<strong>in</strong> ecosysteMs<br />
A limitation <strong>of</strong> our study is <strong>the</strong> lack <strong>of</strong> specific, detailed, ecological analyses <strong>in</strong> <strong>the</strong><br />
Ouachita River Bas<strong>in</strong> and <strong>in</strong> <strong>the</strong> Felsenthal Project area. Most available data are <strong>in</strong> <strong>the</strong> form<br />
<strong>of</strong> checklists and <strong>in</strong>ventories (e.g., <strong>Arkansas</strong> Department <strong>of</strong> Plann<strong>in</strong>g 1974; U.S. Department<br />
<strong>of</strong> Agriculture 1978). Certa<strong>in</strong> general statements can be made from <strong>the</strong> results <strong>of</strong> ecological<br />
studies <strong>in</strong> o<strong>the</strong>r regions <strong>of</strong> <strong>the</strong> sou<strong>the</strong>astern United States.<br />
In broad terms <strong>the</strong> Felsenthal region can be subdivided as upland and lowland ecological<br />
zones; <strong>the</strong> latter zone, our focus <strong>of</strong> <strong>in</strong>terest, is a dist<strong>in</strong>ctive juxtaposition <strong>of</strong> riparian and<br />
aquatic ecosystems.<br />
Riparian Ecosystems<br />
These wetlands are def<strong>in</strong>ed <strong>in</strong> part as follows:<br />
Riparian ecosystems usually occur as an ecotone between aquatic and<br />
upland ecosystems, but have dist<strong>in</strong>ct vegetation and soil characteristics....These<br />
ecosystems are most commonly recognized as bottomland hardwood and floodpla<strong>in</strong><br />
forests <strong>in</strong> <strong>the</strong> eastern and central U.S....Riparian ecosystems are uniquely<br />
characterized by <strong>the</strong> comb<strong>in</strong>ation <strong>of</strong> high species diversity, high species densities,<br />
and high productivity. Cont<strong>in</strong>uous <strong>in</strong>teractions occur betwen riparian, aquatic,<br />
and upland terrestrial ecosystems through exchanges <strong>of</strong> energy, nutrients, and<br />
species (Brown et al. 1979:18).<br />
Two salient characteristics that dist<strong>in</strong>guish riparian ecosystems among o<strong>the</strong>r wetlands<br />
are <strong>the</strong> hydroperiod pulse (prolonged annually <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong>) and <strong>the</strong> high degree<br />
<strong>of</strong> connectedness with o<strong>the</strong>r ecosystems (Ewel 1979). The composition and structure <strong>of</strong> <strong>the</strong><br />
bottomland hardwood forest is highly sensitive to <strong>the</strong> tim<strong>in</strong>g and duration <strong>of</strong> flood<strong>in</strong>g; relatively<br />
fewer tree species, e.g., bald cypress, tolerate long hydroperiods and soil saturation<br />
extend<strong>in</strong>g <strong>in</strong>to <strong>the</strong> grow<strong>in</strong>g season (Bed<strong>in</strong>ger 1971; Huffman 1976). The stand<strong>in</strong>g biomass <strong>of</strong><br />
bottomland hardwood and swamp forest is enormous, and contributes to high productivity<br />
and high carry<strong>in</strong>g capacity for a variety <strong>of</strong> consumer organisms (Conner and Day 1976; Robertson<br />
and Weaver 1979). Among Sou<strong>the</strong>astern ecosystems, floodpla<strong>in</strong> forests rank among<br />
<strong>the</strong> highest <strong>in</strong> primary productivity with litterfall (leaf and twig debris) values exceed<strong>in</strong>g<br />
those reported for upland forests (Wharton and Br<strong>in</strong>son 1979). This detritus is cycled laterally<br />
or downstream dur<strong>in</strong>g flood stages and grazed by bacteria and <strong>in</strong>vertebrates, which<br />
are <strong>in</strong> turn consumed by spawn<strong>in</strong>g or forag<strong>in</strong>g fish and crayfish, seasonal <strong>in</strong>habitants <strong>of</strong> <strong>the</strong><br />
submerged bottomland forest.
16 Hemm<strong>in</strong>gs<br />
The riparian ecosystem is connected or l<strong>in</strong>ked to its upland watershed, to wetlands<br />
upstream and downstream, and to distant ecosystems by a variety <strong>of</strong> processes. Important<br />
among <strong>the</strong>se are long distance migrations <strong>of</strong> fish and eels, seasonally resident birds (swimm<strong>in</strong>g,<br />
wad<strong>in</strong>g, and perch<strong>in</strong>g birds <strong>of</strong> great diversity), and more restricted seasonal or periodic<br />
movements <strong>of</strong> terrestrial vertebrates (e.g., white-tailed deer, turkey). In <strong>the</strong> <strong>Grand</strong><br />
<strong>Marais</strong> Lowland numerous terrestrial and semi-aquatic vertebrates <strong>in</strong>habit or <strong>in</strong>vade <strong>the</strong><br />
bottomland forest and floodpla<strong>in</strong> water courses on a seasonal schedule, withdraw<strong>in</strong>g to upland<br />
marg<strong>in</strong>s dur<strong>in</strong>g <strong>the</strong> w<strong>in</strong>ter-spr<strong>in</strong>g hydroperiod. Among <strong>the</strong>se seasonal foragers is man<br />
himself.<br />
Smith (1978) skillfully relates <strong>the</strong> ecology <strong>of</strong> major river valleys to Mississippi period<br />
settlement and subsistence activity <strong>in</strong> terms broadly applicable to <strong>the</strong> eastern United States.<br />
In his view meander belt zones with closely spaced levees, oxbow lakes, and backswamps<br />
comprised <strong>the</strong> adaptive niche <strong>of</strong> Mississippian populations (Smith 1978:484). These populations<br />
subsisted by cultivat<strong>in</strong>g <strong>the</strong> aborig<strong>in</strong>al tr<strong>in</strong>ity <strong>of</strong> corn, beans, and squash (as well as<br />
secondary crops) on fertile, arable, but localized levee soils, and by exploit<strong>in</strong>g five wild species<br />
groups, primarily (1) backwater fish species, (2) migratory waterfowl, (3) <strong>the</strong> “terrestrial<br />
tr<strong>in</strong>ity” <strong>of</strong> deer, raccoon, and turkey, (4) nuts, fruits, and berries; and (5) pioneer seed-bear<strong>in</strong>g<br />
plants.<br />
Spr<strong>in</strong>g flood<strong>in</strong>g provided an essential external energy subsidy to <strong>the</strong> meander belt<br />
zone, <strong>in</strong>troduc<strong>in</strong>g and dispers<strong>in</strong>g nutrients. Diverse meander belt features were bounded by<br />
maximum edge areas and susta<strong>in</strong>ed a high biomass <strong>of</strong> plants and animals. Regional “energy<br />
capture” variables <strong>of</strong> significance to Mississippian populations were <strong>the</strong> extent <strong>of</strong> arable<br />
land emergent for spr<strong>in</strong>g plant<strong>in</strong>g, and <strong>the</strong> extent <strong>of</strong> permanent or seasonally flooded lakes.<br />
Smith suggests that a dispersed pattern <strong>of</strong> small settlements was <strong>the</strong> optimum strategy for<br />
exploit<strong>in</strong>g meander belt resources, but that aggregated patterns, adapted to ma<strong>in</strong>tenance<br />
and defense <strong>of</strong> <strong>the</strong> resource catchment was <strong>the</strong> most common Mississippian strategy.<br />
Lewis (1974) employed a similar ecological approach <strong>in</strong> a detailed study <strong>of</strong> Mississippian<br />
settlements <strong>in</strong> sou<strong>the</strong>ast Missouri. He emphasizes <strong>the</strong> constra<strong>in</strong>t <strong>of</strong> flood<strong>in</strong>g on permanent<br />
settlement <strong>in</strong> <strong>the</strong> Lower Mississippi Alluvial Valley, which may be resolved by one or<br />
more <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g strategies:<br />
1. locat<strong>in</strong>g settlements on meander belt sites least affected by <strong>in</strong>undation,<br />
2. <strong>in</strong>corporat<strong>in</strong>g “flood-pro<strong>of</strong>” or elevated structures with<strong>in</strong> flood-prone communities,<br />
and<br />
3. reduc<strong>in</strong>g flood risk by construct<strong>in</strong>g levee and dra<strong>in</strong>age systems.<br />
Lewis identifies <strong>the</strong> first strategy as that adopted by Mississippian groups <strong>in</strong> his study area.
Environment <strong>17</strong><br />
Although many <strong>of</strong> <strong>the</strong> ecological concepts and settlement data advanced by Smith and<br />
Lewis are relevant to our analysis <strong>of</strong> <strong>the</strong> Felsenthal Project area, we see here a significant regional<br />
variation <strong>of</strong> <strong>the</strong> Mississippian meander belt strategy. In our study area floodpla<strong>in</strong> resources<br />
were heavily exploited by Mississippi period populations aggregated <strong>in</strong> settlements<br />
on elevated terraces adjacent to <strong>the</strong> floodpla<strong>in</strong> (Chapters 5, 6). Floodpla<strong>in</strong> farm<strong>in</strong>g was<br />
prohibited by <strong>in</strong>fertile, poorly dra<strong>in</strong>ed soils and by <strong>the</strong> prolonged annual period <strong>of</strong> overflow,<br />
but was presumably practiced on terrace lands adjacent to permanent settlements. These<br />
“flood-pro<strong>of</strong>” settlements were “poised” at <strong>the</strong> outer edge <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> to take advantage<br />
<strong>of</strong> slope, dra<strong>in</strong>age, and access to lowland resources by means <strong>of</strong> creeks and bayous. We<br />
doubt whe<strong>the</strong>r this disposition <strong>of</strong> Mississippian settlements is unique to <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong><br />
and <strong>the</strong> Felsenthal region, but here <strong>the</strong> emerg<strong>in</strong>g pattern is dist<strong>in</strong>ctive (Chapter 5:Figure 34).<br />
With regard to settlement and flood risk, we may also make some observations for<br />
European and American <strong>in</strong>habitants <strong>of</strong> <strong>the</strong> region dur<strong>in</strong>g <strong>the</strong> historic period. Clearly, overflow<br />
conditions were a well known, practical matter to early hunters or traders and to later<br />
settlers, as <strong>in</strong>dicated by <strong>the</strong> Dunbar-Hunter journals and o<strong>the</strong>r regional historical sources<br />
(e.g., U.S. Congress 1874). In <strong>the</strong> aborig<strong>in</strong>al fashion, permanent settlement and farm<strong>in</strong>g were<br />
relegated to upland terraces, and riverbank encampments for hunt<strong>in</strong>g, fish<strong>in</strong>g, logg<strong>in</strong>g <strong>of</strong><br />
hardwood, cutt<strong>in</strong>g staves, etc., were made dur<strong>in</strong>g periods <strong>of</strong> normal flow. River land<strong>in</strong>gs<br />
and woodyards dur<strong>in</strong>g <strong>the</strong> n<strong>in</strong>eteenth century seem to have employed moored float<strong>in</strong>g<br />
rafts, usable year-round, serv<strong>in</strong>g especially <strong>the</strong> high water steamboat traffic. In <strong>the</strong> twentieth<br />
century substantial fish<strong>in</strong>g and hunt<strong>in</strong>g camps cont<strong>in</strong>ue to employ float<strong>in</strong>g rafts, pilesupported<br />
platforms, and, <strong>in</strong> a few cases, raised earth mounds. Low earth mounds were also<br />
constructed as refuges for pigs turned out to forage <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> forest. In <strong>the</strong> Congaree<br />
River swamp <strong>of</strong> South Carol<strong>in</strong>a, an environment markedly similar to <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong>,<br />
Michie (1980:102-111) describes comparable, but larger, earth platforms called “cattle<br />
mounts.” In <strong>the</strong>se lowland environments historic <strong>in</strong>habitants have necessarily adapted and<br />
limited <strong>the</strong>ir activities <strong>in</strong> response to seasonal <strong>in</strong>undation.<br />
Aquatic Ecosystems<br />
Sou<strong>the</strong>astern river and floodpla<strong>in</strong> systems <strong>in</strong>clude a variety <strong>of</strong> aquatic habitats, differ<strong>in</strong>g<br />
for example <strong>in</strong> hydrodynamics, waterborne <strong>in</strong>organic and organic matter, water hardness,<br />
acidity/alkal<strong>in</strong>ity, penetration <strong>of</strong> sunlight, and, <strong>of</strong> course, abundance and diversity <strong>of</strong><br />
organisms (Gossel<strong>in</strong>k and Turner 1978; Wharton and Br<strong>in</strong>son 1979). These diverse habitats<br />
may be characterized by an energy gradient, with high energy (lentic) flow<strong>in</strong>g water systems<br />
at one extreme and low energy (lotic) stand<strong>in</strong>g water at <strong>the</strong> o<strong>the</strong>r. Heterogeneity or<br />
niche differentiation and species richness <strong>in</strong>crease with hydrologic energy. In <strong>the</strong> context<br />
<strong>of</strong> a meander<strong>in</strong>g river system <strong>in</strong> an alluvial valley, heterogeneity is created by hydrological<br />
variation between reaches <strong>of</strong> a river (e.g., alternat<strong>in</strong>g deep pools and shallow riffles) and<br />
between <strong>the</strong> river, its tributary streams, oxbow lakes, and backswamps. These habitats may<br />
be <strong>in</strong>terconnected as a broad pool dur<strong>in</strong>g flood stage (an energy pulse) and isolated dur<strong>in</strong>g
18 Hemm<strong>in</strong>gs<br />
low flow. Among o<strong>the</strong>r aspects <strong>of</strong> <strong>the</strong>se aquatic ecosystems, we are particularly concerned<br />
with river fisheries ecology (Welcomme 1979). How have human populations <strong>in</strong> <strong>the</strong> <strong>Grand</strong><br />
<strong>Marais</strong> Lowland adapted to <strong>the</strong> spatial heterogeneity, marked seasonality, and diversity <strong>of</strong><br />
fisheries resources?<br />
Both Smith (1978) and Limp and Reidhead (1979) emphasize <strong>the</strong> importance <strong>of</strong> fish as<br />
an aborig<strong>in</strong>al prote<strong>in</strong> source, <strong>the</strong> seasonal concentration <strong>of</strong> fish biomass, and <strong>the</strong> efficiency<br />
<strong>of</strong> mass-capture techniques <strong>in</strong> <strong>the</strong> eastern United States. Limp and Reidhead experimentally<br />
“harvested” an oxbow lake remnant or slough dur<strong>in</strong>g summer low water stage; gizzard<br />
shad, high <strong>in</strong> oil or fat and food energy value, was <strong>the</strong> predom<strong>in</strong>ant species (21.8 kg <strong>of</strong> <strong>the</strong><br />
45.5 kg total catch <strong>in</strong> one experiment). They conclude that floodpla<strong>in</strong> backwaters had high<br />
potential for low <strong>in</strong>put/high output aborig<strong>in</strong>al fish<strong>in</strong>g. In advance <strong>of</strong> our site survey and<br />
test<strong>in</strong>g results, we note that significant evidence for <strong>in</strong>tensive fish<strong>in</strong>g was obta<strong>in</strong>ed from two<br />
Mississippi period sites <strong>in</strong> <strong>the</strong> Felsenthal Project area (see 3AS285 and 3BR76 <strong>in</strong> Chapter 6).<br />
However, <strong>the</strong>se sites and a high proportion <strong>of</strong> all o<strong>the</strong>r floodpla<strong>in</strong> sites are located on river<br />
channels ra<strong>the</strong>r than oxbow lakes or backswamp dra<strong>in</strong>ages. The local fisheries potential and<br />
relevant archeological data are discussed <strong>in</strong> more detail later <strong>in</strong> this chapter.<br />
A variety <strong>of</strong> potentially useful aquatic and semiaquatic animals, o<strong>the</strong>r than fish, <strong>in</strong>habit<br />
<strong>the</strong> diverse niches <strong>of</strong> Sou<strong>the</strong>astern aquatic ecosystems. Among <strong>the</strong>se are furbear<strong>in</strong>g mammals,<br />
alligators, turtles, frogs, crawfish, and molluscs. For most <strong>of</strong> <strong>the</strong>se species we have<br />
little or no regional archeological or historical data, molluscs be<strong>in</strong>g <strong>the</strong> pr<strong>in</strong>cipal exception<br />
(see below and Appendix B). An analysis <strong>of</strong> commonly occurr<strong>in</strong>g freshwater mussels by<br />
Parmalee and Klippel (1974) <strong>in</strong>dicates that <strong>the</strong>se animals represent low food energy, supplementary<br />
food sources, even though <strong>the</strong>y may be obtrusive constituents <strong>of</strong> an archeological<br />
deposit. None <strong>of</strong> <strong>the</strong> animals listed above are <strong>in</strong>cluded <strong>in</strong> Smith’s (1978) five species groups<br />
exploited as primary food resources by Mississippian populations. But consider<strong>in</strong>g <strong>the</strong><br />
ethno-historic use <strong>of</strong> some <strong>of</strong> <strong>the</strong>se (e.g., crawfish and frogs) <strong>in</strong> <strong>the</strong> Atchafalaya Bas<strong>in</strong> and<br />
o<strong>the</strong>r lowland regions <strong>of</strong> <strong>the</strong> sou<strong>the</strong>rn Mississippi Alluvial Valley, one might predict a sixth<br />
group <strong>of</strong> important aquatic or semiaquatic species (Comeaux 1972).<br />
We noted earlier <strong>the</strong> limitation <strong>of</strong> specific ecological data for <strong>the</strong> Felsenthal Project<br />
area. Fortunately, it is not our task to del<strong>in</strong>eate <strong>in</strong> all particulars an obviously complex<br />
<strong>Grand</strong> <strong>Marais</strong> Lowland ecosystem or to appraise and reconstruct all animal and plant populations<br />
<strong>in</strong> this region. Flannery (1968) and o<strong>the</strong>rs note that human groups are not adapted to<br />
environments, but to particular animal and plant species <strong>in</strong> a systemic relationship. Subsistence<br />
strategies implemented by past human groups <strong>in</strong>volved schedul<strong>in</strong>g <strong>of</strong> seasonal procurement<br />
tasks and also technological choices for maximum efficiency. The follow<strong>in</strong>g review<br />
<strong>of</strong> geohydrological and ecological data for <strong>the</strong> Felsenthal Project area leads to considerations<br />
<strong>of</strong> potential lowland resources and procurement strategies that may have been used by aborig<strong>in</strong>al<br />
and early historic populations.
pleistocene terrAce seQuence<br />
Environment 19<br />
Above <strong>the</strong> lowly<strong>in</strong>g floodpla<strong>in</strong>, <strong>the</strong> valley walls and uplands consist <strong>of</strong> a steplike series<br />
<strong>of</strong> Pleistocene terraces (Figure 3). These terraces have been mapped and <strong>in</strong>terpreted by Fleetwood<br />
(1969), Saucier and Fleetwood (1970), and Saucier (1974). Correlation with cont<strong>in</strong>ental<br />
Pleistocene events and with Mississippi Alluvial Valley landforms and deposits is based on<br />
stratigraphic and o<strong>the</strong>r relative age data ra<strong>the</strong>r than chronometric data. (So far as we know,<br />
no radiocarbon dates for geological contexts have been obta<strong>in</strong>ed <strong>in</strong> <strong>the</strong> Ouachita Valley <strong>of</strong><br />
<strong>Arkansas</strong>; archeological dates from this region are reported <strong>in</strong> Chapters 3 and 7.) Westward<br />
<strong>of</strong> <strong>the</strong> river about 5 km, Eocene Claiborne Group deposits are exposed at <strong>the</strong> surface, while<br />
eastward <strong>of</strong> <strong>the</strong> river <strong>the</strong>se older sediments are almost completely buried by a huge expanse<br />
<strong>of</strong> <strong>the</strong> Prairie Terrace. Proceed<strong>in</strong>g upstream on <strong>the</strong> Ouachita and Sal<strong>in</strong>e rivers, well beyond<br />
our project area, bluffs and hills <strong>of</strong> Claiborne sediments occasionally rise immediately above<br />
<strong>the</strong> modern channels (Haley 1976).<br />
The Prairie Terrace, a relict alluvial pla<strong>in</strong> <strong>of</strong> <strong>the</strong> ancestral <strong>Arkansas</strong> and Ouachita rivers,<br />
has an average elevation <strong>of</strong> about 33.5 m (110 feet) above <strong>the</strong> natural low water stage<br />
<strong>of</strong> <strong>the</strong> Ouachita River (Saucier and Fleetwood 1970:873). Aggradation <strong>of</strong> <strong>the</strong> valley to this<br />
high level occurred dur<strong>in</strong>g Sangamon Interglacial time or about 100,000-80,000 years B.P.<br />
(Saucier 1974:Figure 3). One <strong>of</strong> <strong>the</strong> characteristic features <strong>of</strong> this terrace surface is <strong>the</strong> occurrence<br />
<strong>of</strong> numerous low circular “prairie mounds” <strong>of</strong> presumed eolian/biologic orig<strong>in</strong> several<br />
thousand years B.P. (Ca<strong>in</strong> 1974; Saucier 1978). These mounds are well represented east and<br />
sou<strong>the</strong>ast <strong>of</strong> <strong>the</strong> Felsenthal Project area, not only on <strong>the</strong> Prairie Terrace, but also on younger<br />
surfaces. As <strong>the</strong> name implies, prairie grasslands <strong>in</strong>terrupted <strong>the</strong> presettlement p<strong>in</strong>e-oak forests<br />
<strong>of</strong> <strong>the</strong> Prairie Terrace uplands. These prairies were limited <strong>in</strong> extent, occurr<strong>in</strong>g south <strong>of</strong><br />
<strong>the</strong> <strong>Arkansas</strong> River and north and east <strong>of</strong> <strong>the</strong> Felsenthal Project area. Remnants exist today<br />
<strong>in</strong> Ashley and Bradley counties. In Chapter 8 we describe a sample <strong>of</strong> chert gravel collected<br />
from Prairie Terrace alluvium near Sulphur Spr<strong>in</strong>gs, <strong>Arkansas</strong>, for <strong>the</strong> purpose <strong>of</strong> study<strong>in</strong>g<br />
local toolmak<strong>in</strong>g raw materials.<br />
The Deweyville 1 Terrace, restricted <strong>in</strong> its extent, occurs east and north <strong>of</strong> <strong>the</strong> Felsenthal<br />
Project area at an average elevation <strong>of</strong> 16 m (52 feet) above natural low water stage<br />
(Saucier and Fleetwood 1970: 873). This terrace formed partly as an aggrad<strong>in</strong>g floodpla<strong>in</strong><br />
and partly as backbarrier flats <strong>of</strong> a large “f<strong>in</strong>ger lake” impounded by <strong>Arkansas</strong> River alluvium<br />
(Macon Ridge) at Monroe, Louisiana. Lake Monroe, mid-Wiscons<strong>in</strong> <strong>in</strong> age or about<br />
30,000-28,000 years B.P., covered an area <strong>of</strong> 500-700 square miles, <strong>in</strong>clud<strong>in</strong>g much <strong>of</strong> <strong>the</strong><br />
Ouachita and lower Sal<strong>in</strong>e River valleys and tributary dra<strong>in</strong>ages with<strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area. The detailed evidence for this ext<strong>in</strong>ct lake is beyond <strong>the</strong> scope <strong>of</strong> this summary. The<br />
Deweyville 1 Terrace surface today supports poorly dra<strong>in</strong>ed p<strong>in</strong>e flatwoods <strong>in</strong> some areas.<br />
The Deweyville 2 Terrace flanks <strong>the</strong> Felsenthal Project area as an almost cont<strong>in</strong>uous,<br />
irregular band <strong>of</strong> terra<strong>in</strong> characterized by little surface relief. This terrace surface is <strong>in</strong>ter-
20 Hemm<strong>in</strong>gs<br />
Figure 3. Quaternary geology <strong>of</strong> <strong>the</strong> Ouachita River Valley <strong>in</strong> <strong>Arkansas</strong> (after Saucier and Fleetwood 1970:Figure 1).
Environment 21<br />
preted as a lacustr<strong>in</strong>e pla<strong>in</strong> that developed dur<strong>in</strong>g dra<strong>in</strong><strong>in</strong>g <strong>of</strong> Lake Monroe about 28,000-<br />
25,000 years B.P. (Saucier and Fleetwood 1970:883). The average elevation <strong>of</strong> <strong>the</strong> Deweyville<br />
2 Terrace surface above natural low water stage is about 12.2 m (40 feet). This terrace has<br />
special significance for understand<strong>in</strong>g <strong>the</strong> record <strong>of</strong> past human use <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong><br />
Lowland <strong>in</strong> several respects:<br />
1. On <strong>the</strong> west side <strong>of</strong> <strong>the</strong> Felsenthal Project area and to a lesser extent elsewhere<br />
<strong>in</strong> <strong>the</strong> valley, <strong>the</strong> <strong>in</strong>ner edge <strong>of</strong> <strong>the</strong> Deweyville 2 Terrace is a prom<strong>in</strong>ent<br />
scarp ris<strong>in</strong>g 6-7 m above overflow bottomlands; <strong>the</strong> crest <strong>of</strong> this scarp<br />
(i.e., <strong>the</strong> <strong>in</strong>ner edge <strong>of</strong> <strong>the</strong> terrace) was <strong>the</strong> preferred location for substantial<br />
permanent prehistoric settlements, and probably also was farmed <strong>in</strong> <strong>the</strong><br />
vic<strong>in</strong>ity <strong>of</strong> <strong>the</strong>se settlements (see Rol<strong>in</strong>gson 1972; Rol<strong>in</strong>gson and Schambach<br />
1981; and Appendix A, Unit 7, for data perta<strong>in</strong><strong>in</strong>g to some major sites).<br />
2. Degradation <strong>of</strong> <strong>the</strong> terrace resulted <strong>in</strong> isolated remnants or islands above<br />
overflow bottomlands <strong>in</strong> some areas; examples <strong>in</strong> <strong>the</strong> Felsenthal Project area<br />
are Prairie Island (Figure 4) and Goulett Island on <strong>the</strong> Sal<strong>in</strong>e River. These<br />
outliers, commonly called “p<strong>in</strong>e islands,” are also known to have been<br />
favored locations for prehistoric settlement and for later historic activity<br />
(Rol<strong>in</strong>gson 1972; Appendix A, Unit 7).<br />
3. The terrace scarp or <strong>in</strong>ner edge <strong>in</strong> <strong>the</strong> Felsenthal Project area marks <strong>the</strong><br />
lower limit <strong>of</strong> loblolly p<strong>in</strong>e forest and upper limit <strong>of</strong> bottomland hardwood<br />
forest; this edge reflects <strong>the</strong> control on seedl<strong>in</strong>g reproduction and o<strong>the</strong>r<br />
stress factors exerted by overflow conditions. P<strong>in</strong>e islands are bounded by<br />
<strong>the</strong> same edge between forest types. The elevation <strong>of</strong> this ecological and<br />
hydrological edge is about 75 feet MSL <strong>in</strong> our study area.<br />
4. The Deweyville 2 Terrace surface is ecologically diverse <strong>in</strong> o<strong>the</strong>r respects<br />
as well. As <strong>in</strong> <strong>the</strong> case <strong>of</strong> higher surfaces, presettlement p<strong>in</strong>e-oak forest was<br />
<strong>in</strong>terrupted by prairie open<strong>in</strong>gs <strong>of</strong> restricted extent. Prairie Island is an exist<strong>in</strong>g<br />
example. Ano<strong>the</strong>r possible prairie remnant near Redeye Lake at 74 feet<br />
MSL is subject to frequent overflow (Figure 5). This oblong area was clear <strong>of</strong><br />
trees by 1934 and rema<strong>in</strong>s unchanged to <strong>the</strong> present, based on topographic<br />
maps and air photos. Soil scientists mapped <strong>the</strong> clear<strong>in</strong>g as Crevasse loamy<br />
f<strong>in</strong>e sand, and referred to it as an “idle field,” but we believe it may be a<br />
natural prairie remnant (Gill et al. 1979:38).<br />
The Deweyville 3 Terrace is a relict floodpla<strong>in</strong> formed at approximately 20,000 years<br />
B.P., and is dist<strong>in</strong>guished by channel scars <strong>of</strong> an ancestral Ouachita River with much greater<br />
discharge than <strong>the</strong> present river (Saucier and Fleetwood 1970:876). Us<strong>in</strong>g relict channel dimensions<br />
and geometry, Saucier and Fleetwood calculated paleodischarge five or more
22<br />
Figure 4. Aerial view <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River-Eagle Creek floodpla<strong>in</strong> and Prairie Island (air photo December 1975, courtesy<br />
<strong>of</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers) (AAS neg. 814148).
Environment 23<br />
Figure 5. Possible lowland prairie remnant on <strong>the</strong> Deweyville 2 Terrace near Redeye<br />
Lake (AAS neg. 797152).
24 Hemm<strong>in</strong>gs<br />
times greater than <strong>the</strong> present river, and attribute this flow to Late Wiscons<strong>in</strong> pluvial conditions.<br />
The average elevation <strong>of</strong> <strong>the</strong> terrace above low water is 10 m (33 feet), but throughout<br />
its extent <strong>the</strong> Deweyville 3 surface is barely elevated above <strong>the</strong> Recent floodpla<strong>in</strong> or is<br />
buried by uppermost Recent sediments. For <strong>the</strong> purposes <strong>of</strong> our study <strong>the</strong> Deweyville 3<br />
Terrace is ecologically and hydrologically cont<strong>in</strong>uous with Recent floodpla<strong>in</strong> and overflow<br />
bottomland, although Fleetwood (1969) has mapped with<strong>in</strong> this unit some higher terra<strong>in</strong>, for<br />
example <strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> Eagle Lake Mounds (3BR4). Certa<strong>in</strong>ly, <strong>the</strong> terrace surface was diverse<br />
through its range <strong>in</strong> our study area, for lowly<strong>in</strong>g prairies exist or formerly existed at several<br />
locations (e.g., P<strong>in</strong>e Prairie and C<strong>of</strong>fee Prairie).<br />
recent floodplA<strong>in</strong> geohydrology<br />
Post-Pleistocene sea level rise was accompanied by a complex sequence <strong>of</strong> events <strong>in</strong><br />
<strong>the</strong> Lower Mississippi Alluvial Valley. A reduced sediment load, net aggradation <strong>of</strong> floodpla<strong>in</strong>,<br />
and shifts <strong>of</strong> meander belt characterized <strong>the</strong> Mississippi River and its tributaries <strong>in</strong><br />
various degrees. The sequence <strong>of</strong> events <strong>in</strong> lower courses <strong>of</strong> <strong>the</strong> <strong>Arkansas</strong> River had pr<strong>of</strong>ound<br />
effects upstream on its major tributary, <strong>the</strong> Ouachita River. After about 10,000 years<br />
B.P., <strong>the</strong> <strong>Arkansas</strong> River formed and abandoned a series <strong>of</strong> meander belts now occupied<br />
by Bayou Macon, Boeuf River, and Bayou Bartholomew (Saucier and Fleetwood 1970:884;<br />
Saucier 1974:Figure 1). Above Monroe, Louisiana aggradation <strong>of</strong> <strong>the</strong> Bartholomew course<br />
from about 3000-1000 years B.P. brought about “extreme swamp<strong>in</strong>g” or alluvial drown<strong>in</strong>g <strong>of</strong><br />
<strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland upstream. The f<strong>in</strong>al <strong>Arkansas</strong> River abandonment occurred here<br />
at about 1500 to 1000 years ago. These events provide a temporal and geological framework<br />
for assess<strong>in</strong>g human use <strong>of</strong> <strong>the</strong> Recent floodpla<strong>in</strong> <strong>of</strong> Ouachita River. Later <strong>in</strong> this report we<br />
<strong>in</strong>troduce a record <strong>of</strong> buried archeological components and alluviation <strong>in</strong> <strong>the</strong> Felsenthal<br />
Project area which is reasonably complete for <strong>the</strong> past 3000 years. Earlier components <strong>in</strong> <strong>the</strong><br />
Recent floodpla<strong>in</strong> are deeply buried and “drowned” by <strong>the</strong> modern river (Chapter 7).<br />
Flood Characteristics<br />
Patterson (1971:183) provides a 56-year record <strong>of</strong> annual peak stages for <strong>the</strong> Ouachita<br />
River gaug<strong>in</strong>g station at Lock and Dam No. 6. These peak stages consistently occur from<br />
February to May, reflect<strong>in</strong>g storms mov<strong>in</strong>g nor<strong>the</strong>astward from <strong>the</strong> western Gulf Coast to<br />
<strong>the</strong> Ouachita Mounta<strong>in</strong> region. For <strong>the</strong> period 1912-1968 bankful stage was exceeded <strong>in</strong> all<br />
but two years; <strong>in</strong> all o<strong>the</strong>r years peak floodwaters <strong>in</strong>undated <strong>the</strong> floodpla<strong>in</strong> to at least 65 feet<br />
MSL locally, <strong>the</strong> upper project contour for our site survey work (see Chapters 1 and 5). Very<br />
high peak stages were recorded <strong>in</strong> 1927, 1932, 1945, and 1958, when floodwaters reached<br />
<strong>the</strong> 85-foot contour locally, barely overtopp<strong>in</strong>g <strong>the</strong> Deweyville 2 Terrace. These peak stages<br />
<strong>in</strong>undated <strong>the</strong> lower portions <strong>of</strong> major Mississippi period mound groups <strong>in</strong> <strong>the</strong> study area,<br />
and <strong>in</strong>undated a few mound groups throughout <strong>the</strong>ir extent (e.g., Big Mound Ridge, 3AS6).<br />
It is probable that <strong>the</strong>se and o<strong>the</strong>r large permanent prehistoric settlements were subjected to<br />
comparable, but not greater, flood risk <strong>in</strong> <strong>the</strong> past. An assumption <strong>of</strong> similar flood<strong>in</strong>g conditions<br />
<strong>in</strong> <strong>the</strong> recent past is based on Saucier and Fleetwood’s study and on our own observa-
Environment 25<br />
tions with regard to floodpla<strong>in</strong> topstratum deposits (Chapter 7). This assumption is made<br />
with some confidence for <strong>the</strong> past few centuries and may also be valid for <strong>the</strong> past 1500-3000<br />
years, but requires fur<strong>the</strong>r work.<br />
Duration <strong>of</strong> flood<strong>in</strong>g, not reported by Patterson, is <strong>of</strong> equal or greater importance for<br />
<strong>in</strong>terpret<strong>in</strong>g seasonal occupation or use <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. Accord<strong>in</strong>g to Bed<strong>in</strong>ger (1971),<br />
<strong>the</strong> Ouachita River <strong>in</strong> our study area floods 35-40% <strong>of</strong> <strong>the</strong> year (dur<strong>in</strong>g w<strong>in</strong>ter and spr<strong>in</strong>g)<br />
and greatly resembles <strong>the</strong> lower White River overflow bottomland <strong>in</strong> distribution <strong>of</strong> flood<br />
tolerant hardwoods. In 1979 we kept an <strong>in</strong>termittent record <strong>of</strong> river levels <strong>in</strong> order to plan<br />
and carry out our floodpla<strong>in</strong> survey and test<strong>in</strong>g work; we show bankful stage (about 65 feet<br />
MSL) exceeded for <strong>the</strong> first 5 1/2 months and last 1 1/2 months <strong>of</strong> 1979 (about 58% duration<br />
on a calendar year basis). Floodwaters peaked at about 83 feet MSL <strong>in</strong> early May. The<br />
Ouachita River is, however, regulated by dams <strong>in</strong> its mounta<strong>in</strong>ous headwaters which must<br />
affect peak stage and duration to some extent. Never<strong>the</strong>less, all available modern data lead<br />
one to predict that <strong>the</strong> floodpla<strong>in</strong> <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland will be overflowed <strong>in</strong> w<strong>in</strong>ter-spr<strong>in</strong>g<br />
and emergent <strong>in</strong> summer-fall portions <strong>of</strong> <strong>the</strong> year.<br />
We have not located many historical (pre-1912) records or specific accounts <strong>of</strong> flood<strong>in</strong>g<br />
for <strong>the</strong> Ouachita or Sal<strong>in</strong>e rivers <strong>in</strong> sou<strong>the</strong>rn <strong>Arkansas</strong>. Dunbar’s journal makes an oblique<br />
reference to unusual flood<strong>in</strong>g here <strong>in</strong> <strong>17</strong>99 (Rowland 1930:241). One authoritative report <strong>of</strong><br />
Ouachita River conditions below Camden stated that “a large portion <strong>of</strong> <strong>the</strong> adjacent country<br />
is annually overflowed....Both banks are submerged at even moderate stages and vast<br />
tracts <strong>of</strong> bottomland, covered with forest and canebrake, are underwater five or six months<br />
<strong>of</strong> <strong>the</strong> year” (U.S. Congress 1874:12).<br />
Floodpla<strong>in</strong> Features<br />
A variety <strong>of</strong> features associated with present or past regimes <strong>of</strong> <strong>the</strong> Ouachita River and<br />
its tributaries can be dist<strong>in</strong>guished <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland and <strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area. Modern meander belt features occupy a zone 1-2 km wide def<strong>in</strong>ed by <strong>the</strong> s<strong>in</strong>uous<br />
channels <strong>of</strong> Ouachita and Sal<strong>in</strong>e rivers. The alluvial ridges or natural levees flank<strong>in</strong>g <strong>the</strong>se<br />
channels are more or less cont<strong>in</strong>uous downvalley, and subdued <strong>in</strong> relief, though <strong>in</strong> fact <strong>the</strong>y<br />
are <strong>the</strong> highest terra<strong>in</strong> <strong>in</strong> <strong>the</strong> Recent floodpla<strong>in</strong>. Typically, <strong>the</strong> greatest elevation occurs at <strong>the</strong><br />
riverbank and <strong>the</strong> levee surface drops rapidly 1 or 2 m to lowly<strong>in</strong>g backswamp. Just south <strong>of</strong><br />
U.S. Highway 82 at an historic river land<strong>in</strong>g location (Marie Sal<strong>in</strong>e Land<strong>in</strong>g, 3AS299) <strong>the</strong> follow<strong>in</strong>g<br />
elevations across <strong>the</strong> Ouachita River levee are noted (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers<br />
1979:Plate 1-1):<br />
Distance Eastward<br />
Elevation (MSL) from Left Bank<br />
70.7 feet (21.6 m) 10 feet (3.0 m)<br />
70.2 feet (21.4 m) 280 feet (85.4 m)<br />
67.8 feet (20.7 m) 730 feet (222.6 m)<br />
66.6 feet (20.3 m) 1<strong>17</strong>0 feet (356.7 m)
26 Hemm<strong>in</strong>gs<br />
This levee pr<strong>of</strong>ile is typical <strong>of</strong> many site locations recorded dur<strong>in</strong>g our survey work (Chapter<br />
5). Prehistoric sites frequently occur as small buried lenses <strong>in</strong> <strong>the</strong> levee, truncated and<br />
exposed at <strong>the</strong> riverbank. Historic sites <strong>of</strong>ten occur at or near <strong>the</strong> surface and adjacent to<br />
<strong>the</strong> bank. Both k<strong>in</strong>ds <strong>of</strong> sites took advantage <strong>of</strong> <strong>the</strong> elevation and dra<strong>in</strong>age <strong>of</strong> levee crests.<br />
In Chapter 7 we discuss <strong>the</strong> rate <strong>of</strong> alluviation <strong>in</strong> <strong>the</strong> Recent floodpla<strong>in</strong> <strong>of</strong> <strong>the</strong> Ouachita and<br />
Sal<strong>in</strong>e rivers based on <strong>the</strong> occurrence <strong>of</strong> buried prehistoric sites <strong>in</strong> natural levees.<br />
On <strong>the</strong> <strong>in</strong>side <strong>of</strong> many concave bends a series <strong>of</strong> po<strong>in</strong>t bar ridges, separated by swales,<br />
has resulted from lateral accretion. These ridges may be best developed on <strong>the</strong> downstream<br />
arm or slip<strong>of</strong>f slope <strong>of</strong> a bend, but are uniformly low features <strong>of</strong> <strong>the</strong> modern meander belt<br />
on both rivers. In general our survey work does not extend to po<strong>in</strong>t bar ridges (which lie<br />
mostly above <strong>the</strong> 65-foot upper project contour), and we have recorded few sites <strong>in</strong> such a<br />
location. At least two prehistoric sites on <strong>the</strong> Sal<strong>in</strong>e River occupy a low ridge adjacent to <strong>the</strong><br />
chute or high water passage across <strong>the</strong> neck <strong>of</strong> a bend (Jug Po<strong>in</strong>t Cut<strong>of</strong>f, 3BR76, and Persimmon<br />
3, 3AS315).<br />
Between narrow natural levees and valley walls is a broad, lowly<strong>in</strong>g floodbas<strong>in</strong> or<br />
backswamp zone, usually 3-5 km wide on both sides <strong>of</strong> <strong>the</strong> Ouachita River and somewhat<br />
less on <strong>the</strong> Sal<strong>in</strong>e River. This floodbas<strong>in</strong> zone is underla<strong>in</strong> partly by Recent sediments and<br />
partly by Pleistocene Deweyville 3 Terrace sediments, but <strong>the</strong> whole constitutes an expanse<br />
<strong>of</strong> terra<strong>in</strong> with imperceptible relief. A few irregular ridges (Gum Ridge, Pea Ridge) and p<strong>in</strong>e<br />
islands rise above <strong>the</strong> backswamp, but <strong>the</strong> dom<strong>in</strong>ant features are hydrological <strong>in</strong> orig<strong>in</strong>,<br />
ei<strong>the</strong>r relict or modern. These hydrological features <strong>in</strong> <strong>the</strong> Felsenthal Project area are illustrated<br />
<strong>in</strong> Figure 6 and are listed below by name, with orig<strong>in</strong> noted where ascerta<strong>in</strong>able<br />
(Fleetwood 1969:Figure 3 and Felsenthal map):<br />
Open oxbow or cut<strong>of</strong>f lakes:<br />
Eagle Lake<br />
Pereogee<strong>the</strong> Lake<br />
St. Marys Lake<br />
Open abandoned river courses:<br />
Jones Lake<br />
Fishtrap Lake<br />
Redeye Lake<br />
Horseshoe Slough<br />
Wildcat Lake<br />
Irregular open lakes and partly filled depressions:<br />
<strong>Marais</strong> Sal<strong>in</strong>e<br />
Boeuf Brake<br />
Mouton Brake<br />
Open Lake<br />
<strong>Grand</strong> <strong>Marais</strong>
Figure 6. Hydrological features <strong>in</strong> <strong>the</strong> Felsenthal Project area.<br />
Environment 27
28 Hemm<strong>in</strong>gs<br />
Bayous partly reoccupy<strong>in</strong>g abandoned river courses:<br />
Eagle Creek<br />
Caney Bayou<br />
Lapoile Creek<br />
Lapile Bayou<br />
Sloughs, dra<strong>in</strong>s, crevasses, and backswamp distributaries:<br />
Carroll Slough<br />
Deep Slough<br />
Brushy Creek<br />
Many smaller hydrological features have only local names or lack names altoge<strong>the</strong>r.<br />
Those names which appear above commonly persist from eighteenth century French or<br />
have been anglicized; we f<strong>in</strong>d that some are correctly translated and o<strong>the</strong>rs are consistently<br />
mistranslated by historical writers. Placenames are, <strong>of</strong> course, <strong>the</strong>mselves a source <strong>of</strong> cultural<br />
and environmental <strong>in</strong>formation, as <strong>in</strong> <strong>the</strong> follow<strong>in</strong>g few examples (see Branner 1899;<br />
McDermott 1941):<br />
Pereogee<strong>the</strong> Lake—certa<strong>in</strong>ly from pirogue or periogue, “dugout canoe.”<br />
Boeuf Brake—shortened from boeuf sauvage, “buffalo.”<br />
Lapile Bayou—Hunter’s journal renders this as Bayu de la pelle, most correctly<br />
translated as “shovelnose sturgeon creek.”<br />
These names thus appear to recall an early type <strong>of</strong> watercraft and several animals no longer<br />
present <strong>in</strong> <strong>the</strong> region (see Buchanan 1974:78 on <strong>the</strong> shovelnose sturgeon).<br />
Our list <strong>of</strong> hydrological features emphasizes <strong>the</strong> prom<strong>in</strong>ence, diversity, complex distribution,<br />
and seasonality <strong>of</strong> aquatic habitats <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>. All such habitats are <strong>in</strong>terconnected<br />
dur<strong>in</strong>g peak flood stage; many sluggish meander<strong>in</strong>g sloughs or bayous <strong>in</strong>crease <strong>the</strong>ir<br />
velocities, and may even reverse direction <strong>of</strong> flow, dur<strong>in</strong>g ris<strong>in</strong>g and fall<strong>in</strong>g stages. Even at<br />
low water stage, circuitous dra<strong>in</strong>ages connect most backswamp areas with river channels<br />
and upland terrace edges.<br />
Topstratum/Substratum Deposits<br />
The Recent floodpla<strong>in</strong> is underla<strong>in</strong> by a 35-m thick fill <strong>of</strong> clay, silt, sand, and gravel<br />
(Fleetwood 1969). As <strong>in</strong> <strong>the</strong> case <strong>of</strong> earlier terrace formations, Recent alluvium becomes<br />
f<strong>in</strong>er-gra<strong>in</strong>ed upward, and can be subdivided as substratum and topstratum deposits. The<br />
coarser substratum consists predom<strong>in</strong>antly <strong>of</strong> channel or bedload materials while f<strong>in</strong>er<br />
suspended load materials, deposited overbank, overlie <strong>the</strong> substratum (Allen 1965:127). The<br />
topstratum clays and silts <strong>in</strong> <strong>the</strong> Felsenthal Project area do not exceed about 8 m <strong>in</strong> thickness,<br />
and <strong>of</strong> <strong>the</strong>se usually only 1 or 2 m, but not more than 5 m, are exposed above present
Environment 29<br />
river level (lock stage at 62.2 feet MSL). Substratum deposits are exposed locally only <strong>in</strong> <strong>the</strong><br />
deep excavation for <strong>the</strong> new Felsenthal Lock and Dam, which we have visited but not exam<strong>in</strong>ed<br />
<strong>in</strong> detail. All <strong>of</strong> our site survey and test<strong>in</strong>g work and all <strong>of</strong> <strong>the</strong> known geoarcheological<br />
record <strong>in</strong> <strong>the</strong> Felsenthal Project area perta<strong>in</strong> to <strong>the</strong> f<strong>in</strong>e gra<strong>in</strong>ed topstratum.<br />
Observations made dur<strong>in</strong>g riverbank site survey and test<strong>in</strong>g allow us to characterize<br />
<strong>the</strong> floodpla<strong>in</strong> topstratum. First, <strong>the</strong>se sediments are primarily clays and silts with m<strong>in</strong>or<br />
occurrence <strong>of</strong> sands. Gravels, or dispersed gravel-sized particles, are exceed<strong>in</strong>gly rare <strong>in</strong><br />
<strong>the</strong> upper few meters <strong>of</strong> topstratum. At one location on <strong>the</strong> levee <strong>of</strong> Ouachita River we<br />
collected and analyzed 1.5 and 1.1 m column samples which consisted <strong>of</strong> clayey, silty, and<br />
sandy loams (Chapter 7). Second, <strong>the</strong> sediments ord<strong>in</strong>arily appear to be horizontally bedded<br />
without obtrusive breaks between units (Figure 7). Third, <strong>the</strong>re is little variation <strong>in</strong> <strong>the</strong> drab<br />
grayish brown color <strong>of</strong> sediments. And last, organic components such as wood, bone, and<br />
shell are rarely preserved under <strong>the</strong> prevail<strong>in</strong>g, strongly acid soil conditions (Chapter 7).<br />
(However, we know from published borehole data, from dredged spoil at a pipel<strong>in</strong>e cross<strong>in</strong>g,<br />
and from lock construction backdirt that subfossil wood, especially oak, is frequently<br />
encountered <strong>in</strong> water-logged levels <strong>of</strong> <strong>the</strong> topstratum; see specimens described <strong>in</strong> Appendix<br />
B.) We emphasize <strong>the</strong>se characteristics <strong>of</strong> topstratum deposits because <strong>the</strong>y are relevant to<br />
<strong>the</strong> Recent geological, archeological, and paleoecological record <strong>of</strong> <strong>the</strong> Felsenthal Project<br />
area, and suggest both limits and potential for future work.<br />
Floodpla<strong>in</strong> Soils<br />
Published soil survey reports <strong>in</strong> <strong>the</strong> Felsenthal Project area are available for Ashley and<br />
Bradley counties (Larance 1961; Gill et al. 1979), but only a general soil map is available for<br />
Union County (U.S. Department <strong>of</strong> Agriculture 1968). The overflow bottomlands are uniformly<br />
mapped as Chasta<strong>in</strong> series <strong>in</strong> Bradley and Union counties and as Guyton series <strong>in</strong><br />
Ashley County. The latter series represents a revision <strong>of</strong> classification and term<strong>in</strong>ology still<br />
<strong>in</strong>complete for <strong>the</strong> region. We thus have <strong>the</strong> obvious <strong>in</strong>consistency <strong>of</strong> soil series which co<strong>in</strong>cide<br />
with political boundaries. For example, where a “panhandle” <strong>of</strong> Bradley County extends<br />
across <strong>the</strong> Sal<strong>in</strong>e River at Goulett Island one f<strong>in</strong>ds Bibb silt loam abutt<strong>in</strong>g Guyton soils<br />
along section l<strong>in</strong>es. We are faced, however, with o<strong>the</strong>r sources <strong>of</strong> confusion. For example, a<br />
typical pedon <strong>of</strong> Guyton silt loam is described from a location mapped <strong>in</strong> <strong>the</strong> same report as<br />
<strong>the</strong> Pheba soil series (Gill et al. 1979:40, Sheet 41). This makes chemical and physical analyses<br />
reported for samples from this “typical pedon” doubtful.<br />
Never<strong>the</strong>less, we can briefly describe some properties <strong>of</strong> Guyton soils as an additional<br />
environmental characteristic <strong>of</strong> <strong>the</strong> Recent floodpla<strong>in</strong>. These soils are strongly acid, siliceous,<br />
silt loams over f<strong>in</strong>er textured subsoils, low <strong>in</strong> natural fertility and high <strong>in</strong> available water<br />
capacity. They are, <strong>of</strong> course, frequently flooded <strong>in</strong> w<strong>in</strong>ter and spr<strong>in</strong>g. A humic horizon or<br />
topsoil is weakly developed due to frequent prolonged flood<strong>in</strong>g and repeated fluvial <strong>in</strong>crements<br />
<strong>of</strong> silt and clay. Forest floor organisms which cont<strong>in</strong>uously degrade leaf litter and<br />
o<strong>the</strong>r debris <strong>in</strong> upland forests are limited by <strong>the</strong> flood<strong>in</strong>g regimen. These lowly<strong>in</strong>g floodpla<strong>in</strong><br />
soils, even on slightly elevated levees and ridges, very likely had no agricultural po-
30<br />
Figure 7. Topstratum deposits exposed by <strong>the</strong> Ouachita River bankl<strong>in</strong>e. a. partially<br />
obscured bankl<strong>in</strong>e with old cypress butts exposed; b. actively erod<strong>in</strong>g bank<br />
section with bedd<strong>in</strong>g emphasized by a permeable sandy horizon (AAS neg.<br />
796513, 796544).<br />
A<br />
b
Environment 31<br />
tential whatever for aborig<strong>in</strong>al human groups <strong>in</strong> <strong>the</strong> region. They have supported, however,<br />
a dense productive hardwood forest with many o<strong>the</strong>r useful wetland species <strong>of</strong> wild plants<br />
and animals.<br />
Channel Geometry and Channel Deposits<br />
Figure 8 clarifies <strong>the</strong> various terms which may be used to describe <strong>the</strong> channel and<br />
flow characteristics <strong>of</strong> a meander<strong>in</strong>g river system. In <strong>the</strong> upper part <strong>of</strong> this figure are units<br />
<strong>of</strong> measurement which show, for example, how channel length (<strong>in</strong> river km or river miles)<br />
is related to downvalley distance. The s<strong>in</strong>uosity <strong>of</strong> a river is <strong>the</strong> ratio <strong>of</strong> channel length to<br />
downvalley distance (Leopold et al. 1964:281). In <strong>the</strong> lower part <strong>of</strong> this figure are l<strong>in</strong>es <strong>of</strong><br />
maximum flow and loci <strong>of</strong> maximum erosion and deposition with relation to <strong>the</strong> axis <strong>of</strong><br />
a meander. The thalweg <strong>of</strong> a river is its low water channel; <strong>in</strong> a typical meander<strong>in</strong>g channel<br />
<strong>the</strong> thalweg connects deep scour pools near <strong>the</strong> concave banks. In such a channel scour<br />
pools alternate with riffles or shoals at <strong>the</strong> crossover po<strong>in</strong>t between bends. In this brief<br />
discussion we shall employ such terms and concepts without delv<strong>in</strong>g more deeply <strong>in</strong>to hydraulic<br />
processes; several terms expla<strong>in</strong>ed by Figure 8 appear frequently <strong>in</strong> this report.<br />
The modern channels <strong>of</strong> <strong>the</strong> Ouachita and Sal<strong>in</strong>e rivers are s<strong>in</strong>uous patterns <strong>of</strong> meanders,<br />
usually asymmetrical <strong>in</strong> form, with rare straight reaches (Figure 6). An <strong>in</strong>terest<strong>in</strong>g<br />
characteristic <strong>of</strong> <strong>the</strong>se channels is <strong>the</strong>ir relative stability. We were able to make a detailed<br />
comparison between General Land Office maps (1827) with section l<strong>in</strong>es and modern 7.5<br />
m<strong>in</strong>ute topographic maps (1978), and this comparison shows virtually no measurable<br />
river channel changes <strong>in</strong> <strong>the</strong> Felsenthal Project area. (S<strong>in</strong>uosity and stability are noted for<br />
particular reaches <strong>of</strong> <strong>the</strong>se rivers <strong>in</strong> Chapter 5, Site Units 1, 2, and 4). While bank erosion<br />
has certa<strong>in</strong>ly proceeded, especially <strong>in</strong> <strong>the</strong> concave bank <strong>of</strong> meanders, this erosion has been<br />
markedly slow over <strong>the</strong> 151-year span <strong>of</strong> our maps. Lateral or down-valley migration <strong>of</strong><br />
bends can hardly be detected, and no natural cut<strong>of</strong>fs can be identified for this period; bend<br />
for bend, <strong>the</strong> river channels <strong>of</strong> 1827 and 1978 are approximately congruent. Moreover, <strong>the</strong><br />
various prom<strong>in</strong>ent oxbow lakes and abandoned courses <strong>in</strong> <strong>the</strong> project area were all abandoned<br />
before 1827. Certa<strong>in</strong>ly, this stability is attributable to a complex <strong>of</strong> factors, <strong>in</strong>clud<strong>in</strong>g<br />
magnitude <strong>of</strong> discharge, suspended load and bed load <strong>in</strong> transport, and cohesiveness <strong>of</strong><br />
bed and bank materials (topstratum deposits; see Fisk 1947). It is clear that <strong>the</strong> major annual<br />
flood episodes on <strong>the</strong> Ouachita and lower Sal<strong>in</strong>e rivers are not also major erosional episodes;<br />
deep scour<strong>in</strong>g and cutt<strong>in</strong>g <strong>of</strong> new channels is not <strong>in</strong>dicated for <strong>the</strong> historic period. We<br />
are not so much concerned with <strong>the</strong>se fluvial processes as with <strong>the</strong> fact <strong>of</strong> channel stability,<br />
for it furnishes two general approaches to evaluat<strong>in</strong>g riverbank archeological sites:<br />
1. site selection factors—we know, for example, that deep scour pools at <strong>the</strong> axis<br />
<strong>of</strong> bends and gravel shoals at <strong>the</strong> crossover po<strong>in</strong>t between bends have persisted<br />
for a century and a half; <strong>the</strong> spatial association <strong>of</strong> historic sites or late<br />
prehistoric sites with <strong>the</strong>se channel environments may thus be related to site<br />
functions; Chapter 5 presents some site distributional data relevant to <strong>the</strong>se<br />
locational questions.
32 Hemm<strong>in</strong>gs<br />
Figure 8. Meander<strong>in</strong>g channel characteristics and term<strong>in</strong>ology (after Leopold<br />
et al. 1964:Figure 7-40). a. measurable channel features; b.<br />
flow characteristics and loci <strong>of</strong> erosion or deposition.
Environment 33<br />
2. site preservation/destruction factors—we now have a scale for assess<strong>in</strong>g<br />
lateral erosion (pr<strong>in</strong>cipally on concave banks) and deposition (pr<strong>in</strong>cipally on<br />
convex banks); <strong>in</strong> fact we perceive a surpris<strong>in</strong>gly slow rate at which riverbank<br />
sites will become <strong>in</strong>visible or will be removed altoge<strong>the</strong>r by <strong>the</strong>se processes;<br />
such an assessment is complicated by recent and prospective artificial controls<br />
on river regimen; bankl<strong>in</strong>e erosion is discussed <strong>in</strong> more detail later <strong>in</strong> this<br />
chapter.<br />
The bed load <strong>of</strong> a river consists <strong>of</strong> more or less coarse particles which move along or<br />
very near <strong>the</strong> bottom. These materials are transported pr<strong>in</strong>cipally dur<strong>in</strong>g episodes <strong>of</strong> <strong>in</strong>creased<br />
discharge and may o<strong>the</strong>rwise be “stored” as gravel bars or shoals, <strong>in</strong> contrast to <strong>the</strong><br />
more cont<strong>in</strong>uous transit <strong>of</strong> f<strong>in</strong>e gra<strong>in</strong>ed, suspended load particles. One <strong>of</strong> <strong>the</strong> pr<strong>in</strong>cipal river<br />
channel locations <strong>of</strong> storage has been noted, <strong>the</strong> crossover po<strong>in</strong>t between bends. The shallow<br />
riffles or shoals at such a location may be extensive, and may be broadly exposed at low water<br />
stage. The location and nature <strong>of</strong> shoals are <strong>of</strong> <strong>in</strong>terest because <strong>the</strong>y represent dist<strong>in</strong>ctive,<br />
shallow, and sometimes turbulent habitats preferred by certa<strong>in</strong> fish and molluscs, among<br />
o<strong>the</strong>r species. The potential for human exploitation <strong>of</strong> <strong>the</strong>se conditions is great, as for example<br />
by <strong>the</strong> use <strong>of</strong> fishweirs dur<strong>in</strong>g seasonal spawn<strong>in</strong>g runs. We are also <strong>in</strong>terested <strong>in</strong> gravel<br />
shoals as <strong>the</strong> probable source <strong>of</strong> stone for toolmak<strong>in</strong>g and for various modes <strong>of</strong> cook<strong>in</strong>g <strong>in</strong><br />
an o<strong>the</strong>rwise nearly stoneless alluvial pla<strong>in</strong>. In Chapter 8 we describe a sample <strong>of</strong> bed load<br />
material from <strong>the</strong> Ouachita River collected for <strong>the</strong> purpose <strong>of</strong> test<strong>in</strong>g <strong>the</strong>se latter propositions.<br />
This sample was necessarily derived from dredged material s<strong>in</strong>ce shoals are no longer<br />
exposed above river level (i.e., lock stage).<br />
The distribution <strong>of</strong> shoals and bars on <strong>the</strong> Ouachita River is well known from historical<br />
sources <strong>of</strong> <strong>the</strong> late n<strong>in</strong>eteenth century when river improvements and navigational<br />
hazards were matters <strong>of</strong> great concern. Unfortunately, we lack <strong>the</strong> same data for <strong>the</strong> Sal<strong>in</strong>e<br />
River. From an authoritative report on Ouachita River conditions, we quote <strong>the</strong> follow<strong>in</strong>g<br />
entries for <strong>the</strong> Felsenthal Project area (U.S. Congress 1874:“Table <strong>of</strong> Shoals from Camden,<br />
Ark. to <strong>the</strong> State L<strong>in</strong>e”):<br />
Name Length Remarks<br />
Parrigee<strong>the</strong> 2,500 feet (762 m) Gravel and sand...<br />
Eutaw Rapids 2,000 feet (609 m) Sand<br />
Caney Mary 8,000 feet (2439 m) Sand<br />
No name 16,500 feet (5030 m) Series <strong>of</strong> sand bars<br />
The U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers hydrographic survey for 1873 locates <strong>the</strong>se and o<strong>the</strong>r<br />
shoals by name, ord<strong>in</strong>arily also <strong>in</strong>dicat<strong>in</strong>g <strong>the</strong> presence <strong>of</strong> gravel and <strong>the</strong> expected occurrence<br />
<strong>of</strong> shoals at <strong>the</strong> crossovers between bends. All <strong>of</strong> <strong>the</strong> shoals listed or mapped by <strong>the</strong>se<br />
sources are found on <strong>the</strong> upper part <strong>of</strong> <strong>the</strong> Ouachita River (above <strong>the</strong> mouth <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e<br />
River), <strong>in</strong> contrast to <strong>the</strong> deeper channel characteristic <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland
34 Hemm<strong>in</strong>gs<br />
section downstream. One might expect, <strong>the</strong>refore, differences <strong>in</strong> archeological site density or<br />
distribution between <strong>the</strong>se reaches <strong>of</strong> river (see Chapter 5, Site Units 1 and 4).<br />
Riverbanks and Bank Erosion<br />
Dur<strong>in</strong>g site survey work (described <strong>in</strong> Chapter 5), we exam<strong>in</strong>ed riverbank exposures<br />
total<strong>in</strong>g over 100 km <strong>in</strong> length. One result <strong>of</strong> this work is an appraisal <strong>of</strong> ongo<strong>in</strong>g riverbank<br />
erosion and variation <strong>in</strong> erosional processes related to channel pattern. Bankl<strong>in</strong>e erosion<br />
affects <strong>the</strong> <strong>in</strong>itial exposure, dissection or truncation, and eventual destruction <strong>of</strong> numerous<br />
buried floodpla<strong>in</strong> sites discussed <strong>in</strong> Chapters 5 and 6. Our observations are a prelim<strong>in</strong>ary<br />
attempt to understand present and future impacts on <strong>the</strong>se cultural resources.<br />
Figure 9 represents two typical bank pr<strong>of</strong>iles which may be representative <strong>of</strong> many<br />
reaches on both Ouachita and Sal<strong>in</strong>e rivers and which may also represent erosional extremes.<br />
Numerous archeological sites were discovered on erod<strong>in</strong>g bankl<strong>in</strong>es <strong>of</strong> each type,<br />
<strong>in</strong>dependent, we believe, <strong>of</strong> <strong>the</strong> <strong>in</strong>tensity <strong>of</strong> erosion. Although we detected little map evidence<br />
<strong>of</strong> lateral erosion, <strong>the</strong>re are many field <strong>in</strong>dications that banks are retreat<strong>in</strong>g differentially.<br />
On <strong>the</strong> o<strong>the</strong>r hand, we noted few <strong>in</strong>dications <strong>of</strong> lateral accretion except for obviously<br />
fresh sand bars at <strong>the</strong> axis <strong>of</strong> a bend or slightly downstream on <strong>the</strong> convex bank. Such<br />
prograd<strong>in</strong>g bankl<strong>in</strong>es are not subsumed by <strong>the</strong> pr<strong>of</strong>iles <strong>in</strong> Figure 9. The dist<strong>in</strong>ction between<br />
high <strong>in</strong>tensity and low <strong>in</strong>tensity pr<strong>of</strong>iles is borrowed and re<strong>in</strong>terpreted from a study <strong>of</strong> <strong>the</strong><br />
Port Hudson Bluff, Louisiana (Brunsden and Kesel 1973).<br />
Our high <strong>in</strong>tensity pr<strong>of</strong>ile is steep, concave to near vertical, 1 to 5 m <strong>in</strong> height where<br />
observed. This pr<strong>of</strong>ile is associated with <strong>the</strong> concave bank <strong>of</strong> meanders and is believed to<br />
be retreat<strong>in</strong>g at a relatively rapid rate for <strong>the</strong> Felsenthal Project area (but at a dist<strong>in</strong>ctly slow<br />
rate by comparison to o<strong>the</strong>r river channels like <strong>the</strong> Red River). The topstratum deposits are<br />
so uniformly f<strong>in</strong>e gra<strong>in</strong>ed and cohesive, that sapp<strong>in</strong>g, cav<strong>in</strong>g, slump<strong>in</strong>g, and talus accumulation<br />
at <strong>the</strong> foot <strong>of</strong> <strong>the</strong> slope are rare (Fisk 1947:70). The shr<strong>in</strong>k-swell tendency <strong>of</strong> <strong>the</strong>se deposits<br />
causes blocky peds to dislodge periodically and fall from <strong>the</strong> upper slope, while a plume<br />
<strong>of</strong> muddy sediment washes away cont<strong>in</strong>uously at river level. Wett<strong>in</strong>g and dry<strong>in</strong>g by ra<strong>in</strong> or<br />
fluctuat<strong>in</strong>g water level (<strong>in</strong>clud<strong>in</strong>g wave wash) must accelerate this removal <strong>of</strong> particles. One<br />
<strong>of</strong> <strong>the</strong> most important aspects <strong>of</strong> <strong>the</strong>se steep bank pr<strong>of</strong>iles is <strong>the</strong> potential exposure <strong>of</strong> deeply<br />
buried archeological materials at any elevation above river level, unobscured by talus (see<br />
3UN162 <strong>in</strong> Figure 25). A second important aspect is that sites so exposed are relatively more<br />
endangered by ongo<strong>in</strong>g erosion.<br />
Our low <strong>in</strong>tensity pr<strong>of</strong>ile has low slope, slight convexity upward, and a height <strong>of</strong> 1 to<br />
3 m where observed. This pr<strong>of</strong>ile is associated with convex to nearly straight reaches, and is<br />
believed to represent a slow rate <strong>of</strong> retreat, if not a steady state. The same uniform top-stratum<br />
deposits are exposed, and blocky peds are mov<strong>in</strong>g downslope to be removed at river<br />
level. However, sheet erosion is <strong>in</strong>dicated by <strong>the</strong> frequent shallow exposure <strong>of</strong> tree roots<br />
(Figure 9b). Rills or gullies are also developed which must result from heavy ra<strong>in</strong>fall and
Environment 35<br />
Figure 9. Typical riverbank pr<strong>of</strong>iles <strong>in</strong> <strong>the</strong> Felsenthal Project area. a. high<br />
<strong>in</strong>tensity, relatively rapidly erod<strong>in</strong>g bank pr<strong>of</strong>ile; b. low <strong>in</strong>tensity,<br />
slowly degrad<strong>in</strong>g or nearly stable bank pr<strong>of</strong>ile.
36 Hemm<strong>in</strong>gs<br />
from <strong>the</strong> fall<strong>in</strong>g stages <strong>of</strong> flood<strong>in</strong>g. Although little talus accumulates at <strong>the</strong> foot <strong>of</strong> <strong>the</strong>se<br />
slopes, deeply buried archeological materials are not so well exposed. On <strong>the</strong> o<strong>the</strong>r hand<br />
we found that discovery and exposure <strong>of</strong> shallow buried sites was excellent near <strong>the</strong> upper<br />
slope and zone <strong>of</strong> exposed tree roots. In such sites <strong>the</strong> f<strong>in</strong>e gra<strong>in</strong>ed sediments may be w<strong>in</strong>nowed<br />
away, leav<strong>in</strong>g an array <strong>of</strong> artifacts and debris appropriate for controlled collection<br />
and precise plott<strong>in</strong>g techniques (see False Indigo, 3AS285, and One Cypress Po<strong>in</strong>t, 3AS286,<br />
<strong>in</strong> Chapter 6). We believe <strong>the</strong>se sites are threatened far less by bankl<strong>in</strong>e retreat than by sheet<br />
erosion and broad areal exposure <strong>of</strong> shallow rema<strong>in</strong>s.<br />
recent floodplA<strong>in</strong> ecology<br />
Our pr<strong>in</strong>cipal concern <strong>in</strong> this section is to identify and appraise potential human food<br />
resources <strong>of</strong> <strong>the</strong> overflow bottomland <strong>in</strong> <strong>the</strong> Felsenthal Project area and <strong>Grand</strong> <strong>Marais</strong> Lowland.<br />
Therefore, we exclude some animal and plant species found on terraces or floodpla<strong>in</strong>s<br />
where less rigorous hydrological conditions prevail. In do<strong>in</strong>g so, we make an assumption<br />
that “extreme swamp<strong>in</strong>g” <strong>in</strong> this lowland, attested <strong>in</strong> n<strong>in</strong>eteenth century and modern sources,<br />
also characterized <strong>the</strong> preced<strong>in</strong>g several centuries when numerous prehistoric extractive<br />
camps were established <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> (Chapters 5, 6). Regional archeological data, which<br />
do not now exist, are needed to syn<strong>the</strong>size and verify a model for late prehistoric human<br />
adaptation <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland (Chapter 9).<br />
In order to present summaries <strong>of</strong> important animal and plant species we drew upon<br />
published sources <strong>of</strong> vary<strong>in</strong>g scope <strong>in</strong>clud<strong>in</strong>g: (1) broad areas <strong>of</strong> <strong>the</strong> sou<strong>the</strong>ast United States<br />
or Gulf Coastal Pla<strong>in</strong>, (2) <strong>the</strong> Lower Mississippi Valley region, (3) <strong>the</strong> Ouachita Valley <strong>in</strong><br />
sou<strong>the</strong>rn <strong>Arkansas</strong>, and (4) localities near, but not <strong>in</strong>, <strong>the</strong> Felsenthal Project area. To remedy<br />
<strong>the</strong> lack <strong>of</strong> specific published sources, we also made limited field records <strong>of</strong> species occurrence<br />
<strong>in</strong> <strong>the</strong> overflow bottomland environment dur<strong>in</strong>g site survey and test<strong>in</strong>g work, and<br />
consulted with expert <strong>in</strong>formants at Felsenthal National Wildlife Refuge and <strong>the</strong> <strong>University</strong><br />
<strong>of</strong> <strong>Arkansas</strong>, Fayetteville.<br />
Hardwood and Swamp Forest<br />
Dense hardwood and swamp forest occupies more than 85% <strong>of</strong> <strong>the</strong> area <strong>of</strong> Recent<br />
floodpla<strong>in</strong>, <strong>the</strong> rema<strong>in</strong>der be<strong>in</strong>g open water. Species <strong>of</strong> <strong>the</strong> red oak group predom<strong>in</strong>ate <strong>in</strong><br />
hardwood stands which are flooded approximately 40% <strong>of</strong> <strong>the</strong> year (Figure 10), but variation<br />
occurs among stands <strong>in</strong> response to hydrological and o<strong>the</strong>r environmental factors (Bed<strong>in</strong>ger<br />
1971). Swamp forest refers to brakes and galleries <strong>of</strong> bald cypress and water tupelo,<br />
<strong>in</strong>terspersed with hardwood forest, where <strong>in</strong>undation approaches 100% (Figure 11). The<br />
Ouachita and Sal<strong>in</strong>e River bottomlands <strong>in</strong> and near <strong>the</strong> Felsenthal Project area support some<br />
<strong>of</strong> <strong>the</strong> largest trees on record <strong>in</strong> <strong>the</strong> state, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> follow<strong>in</strong>g (Moore 1960:22, 41, 55, 66,<br />
108):
Figure 10. Bottomland hardwood forest, predom<strong>in</strong>antly mature red oaks, <strong>in</strong> <strong>the</strong> Ouachita River floodpla<strong>in</strong> (U.S.<br />
Forest Service photo, July 1937, from Reynolds 1980:Figure 16) (AAS neg. 813844).<br />
37
38<br />
Figure 11. Hardwood and swamp forest <strong>in</strong> <strong>the</strong> Felsenthal Project area. a. bald cypress<br />
stand <strong>in</strong> an open lake; b. swamp privet, o<strong>the</strong>r shrubs, and sapl<strong>in</strong>gs <strong>in</strong> <strong>the</strong><br />
understory (AAS neg. 796506, 797155).<br />
A<br />
b
Environment 39<br />
Bald cypress, Taxodium distichum, Union County, height 120 feet, dbh 9 feet 3 <strong>in</strong>ches<br />
Water Tupelo, Nyssa aquatics, Ashley County, height 85 feet, dbh 5 feet 3.5 <strong>in</strong>ches<br />
P<strong>in</strong> Oak, Quercus palustris, Union County, height ?, dbh 4 feet 1.8 <strong>in</strong>ches<br />
River Birch, Betula nigra, Ashley County, height 50 feet, dbh ?<br />
Water Elm, Planera aquatics, Union County, height 40 feet, dbh 15 <strong>in</strong>ches<br />
Dur<strong>in</strong>g floodpla<strong>in</strong> survey work, marked differences <strong>in</strong> composition <strong>of</strong> species were<br />
observed for various sites, e.g., river edges and bars, natural levees, oxbow lakes, sloughs,<br />
and backswamps. The follow<strong>in</strong>g classification <strong>of</strong> forest types (American Society <strong>of</strong> Foresters<br />
1954) seems to encompass stands <strong>of</strong> similar composition <strong>in</strong> <strong>the</strong> Felsenthal Project area (only<br />
<strong>the</strong> dom<strong>in</strong>ant overstory species are listed):<br />
Hardwood forest types<br />
(92) Sweetgum-Nuttall Oak-Willow Oak<br />
(93) Sugarberry-American Elm-Green Ash<br />
(95) Black Willow<br />
(96) Overcup Oak-Water Hickory<br />
Swamp forest types<br />
(101-103) Bald Cypress-Water Tupelo<br />
In Table 1 we have compiled a list <strong>of</strong> overstory and understory species which occur<br />
<strong>in</strong> <strong>the</strong> project area; field identifications were made by reference to Maisenhelder (1958) and<br />
Moore (1960). Although this list is unlikely to be complete, we believe it is representative <strong>of</strong><br />
<strong>the</strong> most common woody species which tolerate prolonged overflow <strong>in</strong> this floodpla<strong>in</strong> area.<br />
More importantly, we annotated potential food resources, o<strong>the</strong>r technological resources,<br />
and certa<strong>in</strong> wildlife uses by reference to Halls (1977) and o<strong>the</strong>r sources cited above. This<br />
summary suggests that bottomland forests <strong>of</strong> this type do not yield abundant, concentrated,<br />
wild foods for human consumption as do <strong>the</strong> loblolly p<strong>in</strong>e-oak-hickory forests on elevated<br />
terraces. However, a detailed comparison <strong>of</strong> upland forest and bottomland forest resources<br />
has not been attempted for this region. A variety <strong>of</strong> technologically useful products are present<br />
(Table 1): wood, fuelwood, fiber, bark, and gum. Mast, fruits, and browse for wildlife<br />
consumption are evidently excellent dur<strong>in</strong>g <strong>the</strong> summer-fall stage <strong>of</strong> low water.<br />
One understory species not listed <strong>in</strong> Table 1 is cane (Arund<strong>in</strong>aria gigantea). Although<br />
cane is an important element <strong>of</strong> floodpla<strong>in</strong> vegetation <strong>in</strong> <strong>the</strong> Lower Mississippi Valley<br />
region, it does not tolerate prolonged overflow (Marsh 1977). Canebrakes or stands are not<br />
present today <strong>in</strong> <strong>the</strong> lowly<strong>in</strong>g Felsenthal Project area, a fact also noted <strong>in</strong> 1804 by Dunbar<br />
and Hunter (Rowland 1930:245). The importance <strong>of</strong> canebrakes to wild turkeys and o<strong>the</strong>r<br />
floodpla<strong>in</strong> animals <strong>in</strong> lower <strong>Arkansas</strong> River forests is noted by Meanley (1972:99).
40 Hemm<strong>in</strong>gs<br />
Table 1. Resource Potential <strong>of</strong> Floodpla<strong>in</strong> Trees, Shrubs, and V<strong>in</strong>es.<br />
Common Name Scientific Name Pr<strong>in</strong>cipal Uses and Seasonality<br />
OverstorySpecies:<br />
American Elm Ulmus americana bark can be peeled <strong>in</strong> summer; human use<br />
Baldcypress Taxodium distichum moisture- and decay-resistant wood; human use<br />
bitter pecan Carya aquatica nut crop <strong>in</strong> fall; m<strong>in</strong>or wildlife use<br />
Black Willow Salix nigra fiber for cordage; human use<br />
carol<strong>in</strong>a Ash Frax<strong>in</strong>us carol<strong>in</strong>iana wood for implements; m<strong>in</strong>or human use<br />
Cedar Elm Ulmus crassifolia<br />
Cherrybark Oak Quercus falcata var. pagodaefolia mast crop <strong>in</strong> fall; wildlife use<br />
Honey Locust Gleditsia triacanthos legume seeds and pulp <strong>in</strong> fall; human and wildlife use<br />
Nuttall Oak Quercus nuttalli excellent mast crop <strong>in</strong> fall; wildlife use<br />
Overcup Oak Quercus lyrata mast crop <strong>in</strong> fall; large sweet acorns; wildlife and human use?<br />
P<strong>in</strong> Oak Quercus palustris mast crop <strong>in</strong> fall; wildlife use<br />
Persimmon Diospyros virg<strong>in</strong>iana large fruits <strong>in</strong> fall; wildlife and human use<br />
Pumpk<strong>in</strong> Ash Frax<strong>in</strong>us tomentosa wood used for tool handles; human use<br />
Red Maple Acer rubrum sugary sap <strong>in</strong> spr<strong>in</strong>g; human use<br />
silver Maple Acer sacchar<strong>in</strong>um sugary sap <strong>in</strong> spr<strong>in</strong>g; human use<br />
Sugarberry Celtis laevigata small fruits <strong>in</strong> late summer; wildlife use<br />
Swamp Cottonwood Populus heterophylla<br />
Sweetgum Liquidambar styraciflua mast crop <strong>in</strong> late summer-fall; res<strong>in</strong>ous gum used by humans<br />
Sycamore Platanus occidentalis<br />
Swamp Tupelo Nyssa sylvatica var. biflora small fruits <strong>in</strong> fall; wildlife use<br />
Water Oak Quercus nigra excellent mast crop <strong>in</strong> fall; wildlife use<br />
Water Locust Gleditsia aquatica legume seeds <strong>in</strong> fall; wildlife use?<br />
Water Tupelo Nyssa aquatica small fruits <strong>in</strong> fall; wildlife use<br />
Willow Oak Quercus phellos mast crop <strong>in</strong> fall; wildlife use<br />
Understory Species:<br />
American Hornbeam Carp<strong>in</strong>us carol<strong>in</strong>iana small nuts <strong>in</strong> late summer-early fall; wildlife use<br />
Boxelder Acer negundo sap is edible and available <strong>in</strong> spr<strong>in</strong>g; human use<br />
Buttonbush Cephalanthus occidentalis bark and twigs; used by beaver<br />
False Indigo Amorpha fruticosa legume seeds <strong>in</strong> fall; wildlife use<br />
Mayhaw Crataegus opaca small spr<strong>in</strong>g pomes <strong>in</strong> clusters; wildlife and human use<br />
Possumhaw Ilex decidua small fruits <strong>in</strong> early fall; wildlife use<br />
river birch Betula nigra wood used for implements, basketry; m<strong>in</strong>or human use<br />
Swamp Privet Forestiera acum<strong>in</strong>ata small fruits <strong>in</strong> summer; wildlife use?<br />
Water Elm Planera aquatica fuelwood; m<strong>in</strong>or human use<br />
V<strong>in</strong>es:<br />
Wild Grape Vitis sp. clusters <strong>of</strong> fruit <strong>in</strong> late summer-fall; wildlife and human use<br />
Note: Many species <strong>of</strong> hardwoods listed above may be important as fuelwoods, e.g., oaks and maples
Vertebrates<br />
Environment 41<br />
Lowlands <strong>of</strong> <strong>the</strong> Ouachita Valley support an abundant, diverse, vertebrate fauna<br />
<strong>in</strong>clud<strong>in</strong>g, for <strong>the</strong> most part, species common to forests and wetlands <strong>of</strong> <strong>the</strong> Lower Mississippi<br />
Valley and sou<strong>the</strong>ast United States (U.S. Department <strong>of</strong> Agriculture 1978:Exhibit 12).<br />
We will attempt to make a selective presentation <strong>of</strong> mammals, birds, reptiles, amphibians,<br />
and fishes which <strong>in</strong>habit or regularly <strong>in</strong>trude riparian and aquatic habitats <strong>in</strong> <strong>the</strong> Felsenthal<br />
Project area, and which also appear to represent potentially important food resources for<br />
aborig<strong>in</strong>al human populations <strong>in</strong> <strong>the</strong> region. Many wildlife elements are mentioned <strong>in</strong> <strong>the</strong><br />
Dunbar-Hunter journals <strong>of</strong> 1804-1805. These historical notes <strong>in</strong>clude some species rare or ext<strong>in</strong>ct<br />
<strong>in</strong> <strong>the</strong> region today, e.g., a large trumpeter (?) swan and alligators noted <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity<br />
<strong>of</strong> Caney <strong>Marais</strong> (McDermott 1963:93) and buffalo on <strong>the</strong> upper Ouachita River (Rowland<br />
1930:264). Archeological records <strong>of</strong> vertebrates are few for <strong>the</strong> Felsenthal region as a whole,<br />
but a pert<strong>in</strong>ent sample from <strong>the</strong> Shallow Lake site (3UN9/52) is described by Rol<strong>in</strong>gson<br />
and Schambach (1981:Chapter 6). The ext<strong>in</strong>ct passenger pigeon is rare, but present, <strong>in</strong> this<br />
sample.<br />
Table 2 lists mammals <strong>of</strong> potential importance <strong>in</strong> order <strong>of</strong> body weight, omitt<strong>in</strong>g<br />
small rodents, shrews, and bats (all generally less than 50 g). Data <strong>in</strong>cluded are taken from<br />
Sealander (1979) and o<strong>the</strong>r sources noted. This tabulation represents most <strong>of</strong> <strong>the</strong> large and<br />
medium sized mammals <strong>of</strong> <strong>the</strong> Ouachita Bas<strong>in</strong> <strong>in</strong> <strong>Arkansas</strong> (55 species appear <strong>in</strong> a m<strong>in</strong>imal<br />
modern checklist; U.S. Department <strong>of</strong> Agriculture 1978). The order <strong>of</strong> presentation by body<br />
weight is, <strong>of</strong> course, not necessarily <strong>the</strong> order <strong>of</strong> importance <strong>in</strong> an aborig<strong>in</strong>al subsistence<br />
system; many <strong>of</strong> <strong>the</strong>se animals are solitary, nocturnal, wary, or relatively less palatable, and<br />
some are pr<strong>in</strong>cipally important as furbearers. Archeological data from o<strong>the</strong>r regions <strong>of</strong> <strong>the</strong><br />
eastern United States suggest that white-tailed deer would far exceed all o<strong>the</strong>r mammals<br />
<strong>in</strong> importance as a food animal, with raccoons a distant second <strong>in</strong> importance (Smith 1975).<br />
From <strong>the</strong> Shallow Lake site (3UN9/52), immediately adjacent to <strong>the</strong> floodpla<strong>in</strong> <strong>in</strong> <strong>the</strong> Felsenthal<br />
Project area, <strong>the</strong> follow<strong>in</strong>g mammals are reported (Rol<strong>in</strong>gson and Schambach 1981:Table<br />
4): black bear, white-tailed deer, bobcat, raccoon, gray fox, opossum, skunk, cottontail<br />
rabbit, fox squirrel, gray squirrel, m<strong>in</strong>k, and small rodents. Some animals from both floodpla<strong>in</strong><br />
and upland habitats are certa<strong>in</strong>ly miss<strong>in</strong>g here because <strong>of</strong> <strong>the</strong> size and nature <strong>of</strong> <strong>the</strong><br />
archeological sample. We conclude from modern and limited archeological data that deer<br />
and o<strong>the</strong>r mammals taken on <strong>the</strong> emergent floodpla<strong>in</strong> dur<strong>in</strong>g summer-fall, and occasionally<br />
also under flood conditions <strong>in</strong> w<strong>in</strong>ter-spr<strong>in</strong>g, would contribute a significant <strong>in</strong>crement <strong>of</strong><br />
prote<strong>in</strong> and fat <strong>in</strong> <strong>the</strong> aborig<strong>in</strong>al diet.<br />
A modern checklist <strong>of</strong> birds for <strong>the</strong> Ouachita River Bas<strong>in</strong> <strong>in</strong> <strong>Arkansas</strong> <strong>in</strong>cludes 62<br />
resident species, 59 summer and 56 w<strong>in</strong>ter migrants, and 97 transient species (U.S. Department<br />
<strong>of</strong> Agriculture 1978). A number <strong>of</strong> species formerly present <strong>in</strong> <strong>the</strong> region are omitted<br />
from this list, <strong>in</strong>clud<strong>in</strong>g passenger pigeon, Carol<strong>in</strong>a parakeet, and ivory-billed woodpecker<br />
(James 1974). Numerous o<strong>the</strong>r species are endangered, <strong>in</strong>clud<strong>in</strong>g large wad<strong>in</strong>g birds, raptors,<br />
and some waterfowl characteristic <strong>of</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland habitats. Aga<strong>in</strong> referr<strong>in</strong>g
42 Hemm<strong>in</strong>gs<br />
Table 2. Floodpla<strong>in</strong> Mammals <strong>in</strong> Order <strong>of</strong> Body Weight.<br />
Common Name Scientific Name Body Weight Potential Uses/Comments<br />
Buffalo or Bison Bison bison 1000-600 kg, flesh, hides; probably not present <strong>in</strong> numbers before<br />
rarely 1200 kg <strong>17</strong>00s; last large herd killed <strong>in</strong> Sal<strong>in</strong>e River bottoms<br />
1808 (Sealander 1979:16); ext<strong>in</strong>ct <strong>in</strong> region<br />
Black Bear Ursus americanus 227-100 kg, flesh, fur, tallow or grease; prefers bottomland forrarely<br />
275 kg est and canebrakes; heavily exploited by French<br />
and Indian hunters <strong>in</strong> early historic period; probably<br />
taken commonly by aborig<strong>in</strong>al hunters <strong>in</strong> region<br />
White-Tailed Deer Odocoileus virg<strong>in</strong>ianus 136-40 kg flesh, hides, s<strong>in</strong>ew; prefers bottomlands for mast,<br />
browse fleshy fruits (Stransky 1969); species <strong>of</strong><br />
major importance to aborig<strong>in</strong>al hunters <strong>in</strong> region<br />
Mounta<strong>in</strong> Lion or Felis concolor 103-36 kg flesh, fur; prefers heavily forested habitat with<br />
Cougar dense deer population; an endangered species still<br />
present <strong>in</strong> Ouachita and Sal<strong>in</strong>e River bottomlands<br />
Red Wolf Canis rufus 41-16 kg flesh, fur; wide-rang<strong>in</strong>g species <strong>in</strong> Gulf Coastal<br />
Pla<strong>in</strong>, <strong>in</strong>clud<strong>in</strong>g bottomland forest; ext<strong>in</strong>ct <strong>in</strong> region<br />
(Sealander and Gipson 1974:124)<br />
beaver Castor canadensis 35-11 kg flesh, fur; feeds on sweetgum, buttonbush, o<strong>the</strong>r<br />
trees and shrubs <strong>of</strong> bottomland forest<br />
Bobcat Lynx rufus 20-6 kg flesh, fur; probably not a heavily exploited species,<br />
but known to have been taken by aborig<strong>in</strong>al hunters<br />
<strong>in</strong> region<br />
river otter Lutra canadensis 14-45 kg flesh, fur; a possibly endangered species still present<br />
<strong>in</strong> Ouachita and Sal<strong>in</strong>e River bottomlands<br />
raccoon Procyon lotor 14-4 kg flesh, fur; prefers bottomland hardwood forest and<br />
stream or lake marg<strong>in</strong>s; species <strong>of</strong> major importance<br />
to aborig<strong>in</strong>al hunters <strong>in</strong> region<br />
Gray Fox Urocyon c<strong>in</strong>ereoargenteus 6.4-2.5 kg flesh, fur; preferred habitat <strong>in</strong>cludes bottomland<br />
forest<br />
Opossum Didelphis virg<strong>in</strong>iana 6-3 kg usually taken pr<strong>in</strong>cipally for flesh; prefers bottomland<br />
forest<br />
Swamp Rabbit Sylvilagus aquaticus 2.9-1.2 kg flesh, fur; prefers bottomland hardwood forest and<br />
canebrakes<br />
Muskrat Ondatra zibethicus 1.8-0.5 kg flesh, fur; prefers river and stream banks, as well as<br />
lakes and ponds<br />
M<strong>in</strong>k Mustela vison 1.6-0.8 kg flesh, fur; prefers river and stream banks<br />
Gray Squirrel Sciurus carol<strong>in</strong>ensis 710-3<strong>17</strong> g flesh, fur; preferred habitat <strong>in</strong>cludes bottomland<br />
hardwood forest<br />
Long-tailed Weasel Mustela frenata 250-85 g flesh, fur; preferred habitat <strong>in</strong>cludes bottomland<br />
hardwood forest<br />
Sou<strong>the</strong>rn Fly<strong>in</strong>g Glaucomys volans 98-40 g usually only taken for fur; preferred habitat <strong>in</strong>cludes<br />
Squirrel bottomland hardwood forest
Environment 43<br />
to <strong>the</strong> archeological record from Shallow Lake (3UN9/52), we f<strong>in</strong>d a substantial representation<br />
<strong>of</strong> w<strong>in</strong>ter waterfowl such as Canada geese and ducks, turkey, and a s<strong>in</strong>gle occurrence<br />
<strong>of</strong> passenger pigeon (Rol<strong>in</strong>gson and Schambach 1981:Table 4). O<strong>the</strong>r bird rema<strong>in</strong>s were<br />
unidentified, and sampl<strong>in</strong>g error here is likely to be present. Wild turkey, for example, is<br />
exceeded by waterfowl <strong>in</strong> this sample. In fact we would expect heavy use <strong>of</strong> turkey by <strong>the</strong><br />
aborig<strong>in</strong>al residents <strong>of</strong> this region, as elsewhere <strong>in</strong> eastern United States (Smith 1975).<br />
The same modern checklist for <strong>the</strong> Ouachita River Bas<strong>in</strong> <strong>in</strong>cludes 61 reptiles and 32<br />
amphibians, although many <strong>of</strong> <strong>the</strong>se occur <strong>in</strong> upland habitats not under consideration here.<br />
Those animals occurr<strong>in</strong>g <strong>in</strong> riparian or aquatic habitats <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> which can feasibly<br />
be exploited for human food <strong>in</strong>clude alligator, snapp<strong>in</strong>g turtles, sliders, s<strong>of</strong>tshells, map<br />
turtles, mud and musk turtles, water snakes, western cottonmouth, salamanders, siren,<br />
amphiuma, Red River waterdog, and <strong>the</strong> bullfrog. Alligators occasionally reached over 5<br />
m length prior to historic settlement and commercial exploitation <strong>in</strong> <strong>the</strong> region. S<strong>in</strong>ce that<br />
time <strong>the</strong> species has been elim<strong>in</strong>ated, restocked, and rema<strong>in</strong>s endangered (Reagan 1974).<br />
Aborig<strong>in</strong>al uses for this animal <strong>in</strong>cluded flesh, hides, oil, and ivorylike teeth. One very large<br />
alligator tooth has been recovered from <strong>the</strong> Gordon site (3AS152) near Sal<strong>in</strong>e River, just<br />
north <strong>of</strong> <strong>the</strong> Felsenthal Project area (<strong>Arkansas</strong> Archeological Survey site files). Many o<strong>the</strong>r<br />
species noted above have been recorded occasionally <strong>in</strong> archeological deposits. At Shallow<br />
Lake (3UN9/52) only snapp<strong>in</strong>g turtle, slider, and <strong>the</strong> upland box turtle have been identified<br />
(Rol<strong>in</strong>gson and Schambach 1981: Table 4). We believe that <strong>the</strong> use <strong>of</strong> various aquatic turtles,<br />
alligator, and frogs could be more important than this limited record <strong>in</strong>dicates.<br />
The f<strong>in</strong>al group <strong>of</strong> vertebrates considered here are fishes which may have constituted<br />
a highly important food source for <strong>the</strong> aborig<strong>in</strong>al <strong>in</strong>habitants <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland.<br />
In general, fishes here are abundant, diverse, high <strong>in</strong> food value, and susceptible to capture<br />
by simple techniques. Regional <strong>in</strong>ventories and studies <strong>in</strong>clude those <strong>of</strong> Reynolds (1971),<br />
Buchanan (1974), and Robison (1975). Some records are from with<strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area. Table 3 summarizes <strong>the</strong> most important species or groups by body weight and habitat<br />
preference. Many <strong>of</strong> <strong>the</strong> smaller fishes such as darters and sh<strong>in</strong>ers are omitted. However,<br />
body weight is far less likely to determ<strong>in</strong>e aborig<strong>in</strong>al use than behavior <strong>of</strong> a species. Certa<strong>in</strong><br />
fishes are more easily exploited by mass capture techniques. Large differences <strong>in</strong> usable<br />
meat, palatability, and food value also exist among species. A historical account notes, for example,<br />
that buffal<strong>of</strong>ish “floated” <strong>in</strong> large numbers dur<strong>in</strong>g August and could be easily gigged<br />
(W. T. Mart<strong>in</strong> <strong>in</strong> E<strong>the</strong>ridge 1959:xviii). Buffal<strong>of</strong>ish also can be taken <strong>in</strong> quantity dur<strong>in</strong>g spr<strong>in</strong>g<br />
flood<strong>in</strong>g when spawn<strong>in</strong>g <strong>in</strong> shallow areas <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> forest (Comeaux 1972). Mass<br />
capture <strong>of</strong> gizzard shad <strong>in</strong> a dw<strong>in</strong>dl<strong>in</strong>g floodpla<strong>in</strong> lake was described earlier <strong>in</strong> this chapter<br />
(Limp and Reidhead 1979). Shad and eel, among species listed <strong>in</strong> Table 3, are particularly<br />
high <strong>in</strong> fat content. Migratory and spawn<strong>in</strong>g species which ascend or descend <strong>the</strong> rivers
44 Hemm<strong>in</strong>gs<br />
Table 3. Selected Food Fishes and Habitat Preferences, Lower Ouachita River System <strong>in</strong><br />
<strong>Arkansas</strong>.<br />
Approximate Habitat Preferences<br />
Maximum River Sloughs/ Oxbows/<br />
Common Name Scientific Name Weight (kg) Channels Bayous Open Lakes Comments<br />
Lake Sturgeon Acipenser fulvescens 135 x x Endangered species<br />
Paddlefish Polyodon spathula 68 x x x Endangered species<br />
Bowf<strong>in</strong> Amia calva - x x x<br />
Bigmouth Buffalo Ictiobus cypr<strong>in</strong>ellus 30 x x<br />
Fla<strong>the</strong>ad Catfish Pylodictis olivaris 23<br />
Freshwater Drum Aplod<strong>in</strong>otus grunniens 20 x x<br />
Channel Catfish Ictalurus punctatus 9 x<br />
Catfish Ictalurus sp. - x<br />
Smallmouth Buffalo Ictiobus bubalus 9 x x<br />
gars Lepisosteus sp. - Alligator gar is depleted<br />
Walleye Stizostedion vitreum 5 x x<br />
Largemouth Bass Micropterus salmoides 4 x x<br />
American Eel Anguilla rostrata - x Catadromous species<br />
Shovelnose Sturgeon Scaphirynchus platorynchus 3 x Endangered species<br />
Bullheads Ictalurus sp. 1 x x<br />
crappie Pomoxis sp. 1 x x<br />
Pickerel Esox sp.
Environment 45<br />
were susceptible to capture at shoals by gigg<strong>in</strong>g, spear<strong>in</strong>g, or trapp<strong>in</strong>g <strong>in</strong> weirs. Species frequent<strong>in</strong>g<br />
pools or ponds were sometimes poisoned or taken with nets, techniques that leave<br />
no direct evidence.<br />
M<strong>in</strong>imal data from <strong>the</strong> Felsenthal Project area are available for fish utilization dur<strong>in</strong>g<br />
<strong>the</strong> prehistoric period. At Shallow Lake (3UN9/52) fishbone was abundant <strong>in</strong> refuse pits and<br />
midden; only catfish, gar, and drum are identified (Stacy 1976:61; Rol<strong>in</strong>gson and Schambach<br />
1981:Table 4). In Chapter 6 we present test excavation data from two floodpla<strong>in</strong> extractive<br />
sites <strong>of</strong> <strong>the</strong> early Mississippi period (A.D. 1100-1400) which have small, but dense, midden<br />
deposits with abundant fishbone (False Indigo, 3AS285, and Jug Po<strong>in</strong>t Cut<strong>of</strong>f, 3BR76). These<br />
small camps’ <strong>in</strong>habitants were evidently <strong>in</strong>tensively engaged <strong>in</strong> cook<strong>in</strong>g or render<strong>in</strong>g fish<br />
and fish oil.<br />
Invertebrates<br />
Two groups <strong>of</strong> <strong>in</strong>vertebrates are likely to have been exploited regularly by aborig<strong>in</strong>al<br />
<strong>in</strong>habitants <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland. The first group, freshwater mussels, was most<br />
accessible <strong>in</strong> river channels dur<strong>in</strong>g summer-fall low water stage, and mussel shell rema<strong>in</strong>s<br />
have been preserved <strong>in</strong> numerous sites <strong>in</strong> <strong>the</strong> region. The second group, crawfish, were collected<br />
or trapped pr<strong>in</strong>cipally dur<strong>in</strong>g w<strong>in</strong>ter-spr<strong>in</strong>g flood<strong>in</strong>g on bottomland.<br />
In Appendix B we summarize a modern collection <strong>of</strong> molluscs from <strong>the</strong> Felsenthal<br />
Project area and limited archeological collections from both prehistoric and historic sites<br />
(5 species <strong>in</strong> each context). A slightly longer species list has been compiled for <strong>the</strong> Shallow<br />
Lake site (3UN9/52), but is equally unlikely to be a comprehensive sample; buckhorn, pigtoe,<br />
and three-ridge mussels predom<strong>in</strong>ated <strong>in</strong> this collection. Mussels now known from prehistoric<br />
sites represent a variety <strong>of</strong> habitats, <strong>in</strong>clud<strong>in</strong>g shoals, sloughs, and oxbow or open<br />
lakes. As noted earlier <strong>in</strong> this chapter, mussels would have been supplementary ra<strong>the</strong>r than<br />
staple food resources. A recently published checklist <strong>of</strong> unionid molluscs for <strong>the</strong> Ouachita<br />
River system <strong>in</strong> <strong>Arkansas</strong> <strong>in</strong>dicates that <strong>the</strong> lower reaches rema<strong>in</strong> poorly known (Gordon et<br />
al. 1979).<br />
Crawfish <strong>of</strong> several species occur frequently <strong>in</strong> riparian and aquatic habitats <strong>of</strong> <strong>the</strong><br />
<strong>Grand</strong> <strong>Marais</strong> Lowland. The species <strong>of</strong> potential importance are Procambarus clarkii (swamp<br />
crawfish) and P. bland<strong>in</strong>gi (river crawfish), both taken commercially <strong>in</strong> <strong>the</strong> Atchafalaya Bas<strong>in</strong><br />
and o<strong>the</strong>r regions <strong>of</strong> Louisiana (Comeaux 1972:Chapter 5). In this area young crawfish are<br />
baited with a stick or l<strong>in</strong>e or trapped dur<strong>in</strong>g w<strong>in</strong>ter-spr<strong>in</strong>g flood<strong>in</strong>g <strong>of</strong> bottomlands. The<br />
exoskeleton <strong>of</strong> this animal does not persist <strong>in</strong> archeological deposits, and its record is correspond<strong>in</strong>gly<br />
rare or lack<strong>in</strong>g. Almost certa<strong>in</strong>ly, crawfish were frequently used <strong>in</strong> <strong>the</strong> Felsenthal<br />
Project area dur<strong>in</strong>g <strong>the</strong> prehistoric period.<br />
The last potential foodstuff considered here is that most desirable product <strong>of</strong> <strong>the</strong> honeybee.<br />
Hives and honey were (and still are) located <strong>in</strong> <strong>the</strong> hollows <strong>of</strong> many large trees above
46 Hemm<strong>in</strong>gs<br />
<strong>the</strong> level <strong>of</strong> annual flood<strong>in</strong>g. In historic times woodsmen were able to track honeybees to<br />
<strong>the</strong> bee tree and cut or burn it down, <strong>of</strong>ten tak<strong>in</strong>g a large quantity <strong>of</strong> honey and beeswax<br />
(E<strong>the</strong>ridge 1959:80). This low water, summer-fall activity became an important early historic<br />
<strong>in</strong>dustry <strong>in</strong> <strong>the</strong> Atchafalaya Bas<strong>in</strong> and o<strong>the</strong>r areas <strong>of</strong> extensive bottomland forest (Comeaux<br />
1972:92). We would expect considerable use <strong>of</strong> this foodstuff by aborig<strong>in</strong>al <strong>in</strong>habitants <strong>of</strong> <strong>the</strong><br />
<strong>Grand</strong> <strong>Marais</strong> Lowland. Unfortunately, <strong>the</strong>re is little possibility <strong>of</strong> an archeological record <strong>of</strong><br />
its use.<br />
potentiAl pAleoecologicAl studies<br />
This chapter presented a summary <strong>of</strong> environmental data for <strong>the</strong> Felsenthal Project<br />
area and lower Ouachita Valley <strong>in</strong> <strong>Arkansas</strong>. In this section we acknowledge that paleoenvironmental<br />
data, which could support our view <strong>of</strong> regional prehistoric adaptations, are truly<br />
deficient. There are no palynological, paleontological, or geochronological studies with<strong>in</strong> or<br />
near <strong>the</strong> project area. But <strong>the</strong> work <strong>of</strong> geologists cited earlier (chiefly mapp<strong>in</strong>g and map <strong>in</strong>terpretation)<br />
<strong>in</strong>dicated a dynamic geomorphic history for <strong>the</strong> last 100,000 years or so. Relict<br />
terraces, lacustr<strong>in</strong>e features, abandoned river courses, cut<strong>of</strong>f lakes, and a sequence <strong>of</strong> alluvial<br />
deposits attest to this dynamic history. We are conv<strong>in</strong>ced that a wide range <strong>of</strong> paleoecological<br />
data are recoverable, and that <strong>in</strong>terdiscipl<strong>in</strong>ary paleoenvironmental study is badly<br />
needed <strong>in</strong> <strong>the</strong> Felsenthal region, especially where modern land or river uses are modify<strong>in</strong>g<br />
or effac<strong>in</strong>g <strong>the</strong> paleoenvironmental record.<br />
One locality <strong>in</strong> <strong>the</strong> Felsenthal Project area, Goulett Island, may serve here as an example<br />
<strong>of</strong> potentially significant study sites. Goulett Island is a “p<strong>in</strong>e island,” an erosional<br />
remnant <strong>of</strong> <strong>the</strong> Deweyville 2 Terrace, truncated by <strong>the</strong> modern channel <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River.<br />
We refer to archeological and historical rema<strong>in</strong>s on Goulett Island at various po<strong>in</strong>ts <strong>in</strong> this<br />
report (Chapter 5, Appendixes A, C). Here we are concerned with Goulett Island’s natural<br />
values and scientific potential. Goulett Island was <strong>in</strong> fact first studied by a geologist, David<br />
Dale Owen (1860), who described a riverbottom lignite ledge or outcrop m<strong>in</strong>ed for fuel by<br />
local <strong>in</strong>habitants (see historic site 3BR8 <strong>in</strong> Chapter 5). This outcrop is no longer visible, even<br />
at low water stage. Goulett Island was also visited by <strong>the</strong> early archeological entrepreneur,<br />
C. B. Moore (1913), who made no useful observations beyond his perfunctory mound <strong>in</strong>vestigations<br />
(he does mention a “great flood <strong>of</strong> 1912” on <strong>the</strong> Sal<strong>in</strong>e River). We doubt whe<strong>the</strong>r<br />
geologists have studied <strong>the</strong> Goulett Island locality or published any observations perta<strong>in</strong><strong>in</strong>g<br />
to it s<strong>in</strong>ce <strong>the</strong> time <strong>of</strong> Owen.<br />
Dur<strong>in</strong>g our own riverbank survey work, we exam<strong>in</strong>ed <strong>the</strong> exposure <strong>of</strong> Deweyville 2<br />
Terrace sediments at Goulett Island. This exposure is a section <strong>of</strong> riverbank about 4 m high<br />
and 1000 m long. The topstratum deposits <strong>of</strong> <strong>the</strong> Deweyville 2 Terrace formation here consist<br />
<strong>of</strong> gray lam<strong>in</strong>ated silty clays and clays with charcoal and/or carbonized wood exposed at<br />
certa<strong>in</strong> levels. No artifacts were observed <strong>in</strong> place, although <strong>the</strong>re are abundant pre-
Environment 47<br />
historic and historic artifacts erod<strong>in</strong>g downslope or discarded overbank from <strong>the</strong> surface<br />
<strong>of</strong> Goulett Island. If artifacts were discovered <strong>in</strong> place <strong>the</strong>y would presumably be mid-Wiscons<strong>in</strong><br />
<strong>in</strong> age or about 30,000-25,000 years B.P., accord<strong>in</strong>g to Saucier and Fleetwood’s (1970)<br />
<strong>in</strong>terpretation <strong>of</strong> Ouachita Valley chronology. Moreover, <strong>the</strong> sediments exposed at Goulett<br />
Island may represent lake bottom or nearshore clays and silts deposited by Pleistocene Lake<br />
Monroe, at or near its upstream (nor<strong>the</strong>rnmost) marg<strong>in</strong> as proposed and mapped by Saucier<br />
and Fleetwood (1970:Figure 5).<br />
Goulett Island is probably <strong>the</strong> most significant outcrop <strong>of</strong> Lake Monroe sediments<br />
<strong>in</strong> <strong>the</strong> Felsenthal region, and could furnish samples to verify a lacustr<strong>in</strong>e paleoenvironment<br />
and especially to corroborate (by radiocarbon assay) <strong>the</strong> age <strong>of</strong> Lake Monroe and <strong>the</strong><br />
Deweyville 2 Terrace. Although we recommend a broad range <strong>of</strong> <strong>in</strong>terdiscipl<strong>in</strong>ary study <strong>in</strong><br />
conjunction with mitigation <strong>of</strong> archeological sites <strong>in</strong> <strong>the</strong> Felsenthal Project area (Appendix<br />
C), we specifically recommend detailed sedimentological and geochronological study <strong>of</strong> <strong>the</strong><br />
Goulett Island locality.<br />
potentiAl lowlAnd resources<br />
In <strong>the</strong> preced<strong>in</strong>g summary <strong>of</strong> biophysical characteristics <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland<br />
and Felsenthal Project area we relied upon historical sources, published ecological and<br />
geohydrological data, regional archeological data, and our own field observations. Figure<br />
12 <strong>in</strong>tegrates some <strong>of</strong> <strong>the</strong>se data <strong>in</strong> <strong>the</strong> form <strong>of</strong> a schematic cross section extend<strong>in</strong>g across<br />
<strong>the</strong> Felsenthal Project area and Ouachita Valley at about <strong>the</strong> latitude <strong>of</strong> Shallow Lake site<br />
(3UN9/52). Variation between upland p<strong>in</strong>e-oak forest, locus <strong>of</strong> prehistoric permanent settlements,<br />
and lowland hardwood and swamp forest with numerous water courses, locus <strong>of</strong><br />
impermanent extractive camps, is emphasized <strong>in</strong> this draw<strong>in</strong>g.<br />
We have attempted to del<strong>in</strong>eate potential floodpla<strong>in</strong> food and o<strong>the</strong>r useful resources<br />
through limited modern and archeological <strong>in</strong>formation from <strong>the</strong> region (Tables 1-3). Many<br />
fundamental questions about local resource potential rema<strong>in</strong> to be answered. Direct evidence,<br />
<strong>in</strong> <strong>the</strong> form <strong>of</strong> well preserved floral and faunal rema<strong>in</strong>s from floodpla<strong>in</strong> sites <strong>in</strong> this<br />
region, is likely to be rare (and has <strong>in</strong> fact proved to be so; see Chapters 6 and 7). The follow<strong>in</strong>g<br />
questions deal with <strong>the</strong> Felsenthal Project area specifically, but are broadly relevant to<br />
study<strong>in</strong>g prehistoric subsistence systems <strong>in</strong> <strong>the</strong> region:<br />
1. What was <strong>the</strong> relative importance <strong>of</strong> floodpla<strong>in</strong> floral and faunal resources?<br />
We are <strong>in</strong>cl<strong>in</strong>ed to believe that fish was an exceed<strong>in</strong>gly important prote<strong>in</strong> source dur<strong>in</strong>g <strong>the</strong><br />
Mississippi period (A.D. 1100-<strong>17</strong>00), and probably also dur<strong>in</strong>g earlier periods. Certa<strong>in</strong> fish<br />
species were seasonally abundant or concentrated <strong>in</strong> floodpla<strong>in</strong> aquatic habitats and were
48<br />
Figure 12. Cross-valley section <strong>in</strong> <strong>the</strong> Felsenthal Project area show<strong>in</strong>g terraces, floodpla<strong>in</strong>, forest types, and o<strong>the</strong>r features (geology<br />
adapted from Fleetwood 1969 and Saucier and Fleetwood 1970: Figure 8f; no horizontal or vertical scale).
Environment 49<br />
quite likely harvested <strong>in</strong> bulk quantities. Floodpla<strong>in</strong> mammals (especially white-tailed deer),<br />
birds, reptiles, amphibians, molluscs, and o<strong>the</strong>r <strong>in</strong>vertebrates were certa<strong>in</strong>ly also supplementary<br />
prote<strong>in</strong> sources, and many <strong>of</strong> <strong>the</strong>se species were migratory and seasonally abundant,<br />
or were seasonally concentrated on p<strong>in</strong>e islands by ris<strong>in</strong>g floodwaters.<br />
Among floodpla<strong>in</strong> floral resources, nut crops were <strong>the</strong> major food supply available <strong>in</strong><br />
both quantity and variety. The various red oaks, such as willow oak (Quercus phellos), occur<br />
<strong>in</strong> dense stands on <strong>the</strong> natural levees and produce substantial mast crops <strong>in</strong> fall, prior to <strong>in</strong>undation.<br />
Hilliard (1981) shows that red oak acorns were heavily used by prehistoric people<br />
<strong>of</strong> <strong>the</strong> Ozarks, could be collected, stored, and processed efficiently, and are higher than<br />
white oak acorns <strong>in</strong> caloric value. One very abundant nut tree <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> forest, bitter<br />
pecan (Carya aquatica), does not appear to have any importance as a human food source, <strong>in</strong><br />
spite <strong>of</strong> <strong>the</strong> productivity and usefulness <strong>of</strong> o<strong>the</strong>r hickories and pecans (Smith 1953).<br />
On <strong>the</strong> whole, we identify more “first l<strong>in</strong>e” animal foods than plant foods <strong>in</strong> <strong>the</strong> overflow<br />
bottomland environment, and we suggest that prehistoric subsistence strategies <strong>in</strong> <strong>the</strong><br />
region focused on <strong>the</strong>se lowland faunal resources.<br />
2. What was <strong>the</strong> relative importance <strong>of</strong> uplands (terra<strong>in</strong> above prolonged overflow)<br />
and lowlands (overflow bottomlands) as resource zones?<br />
This question is at least partly beyond <strong>the</strong> scope <strong>of</strong> our study, s<strong>in</strong>ce we would have to control<br />
environmental and paleoenvironmental data from uplands <strong>in</strong> <strong>the</strong> region. In particular<br />
we would require knowledge <strong>of</strong> prehistoric agricultural systems <strong>in</strong> <strong>the</strong> region which is<br />
not now available. Schambach proposes some local agricultural adaptations, but <strong>the</strong>re are<br />
presently no archeological data to support <strong>the</strong>se propositions (Rol<strong>in</strong>gson and Schambach<br />
1981:105):<br />
In late periods, if <strong>the</strong>re was any horticulture, it must have been slash and burn horticulture<br />
<strong>in</strong> <strong>the</strong> uplands by people liv<strong>in</strong>g <strong>in</strong> small hamlets or s<strong>in</strong>gle farmsteads scattered<br />
along <strong>the</strong> tributary streams....Only extensive exploitation <strong>of</strong> <strong>the</strong> uplands...could have<br />
supported <strong>the</strong> populations [represented] <strong>in</strong> <strong>the</strong> Felsenthal region.<br />
Smith’s (1978) identification <strong>of</strong> major Mississippi period food resource groups, mentioned<br />
earlier <strong>in</strong> this chapter, might be more specifically adapted to <strong>the</strong> Felsenthal region as<br />
follows:
50 Hemm<strong>in</strong>gs<br />
Major Resource Groups Primary Environments<br />
1 corn, beans, squash upland terraces<br />
2 channel, lake, and backwater fish species floodpla<strong>in</strong><br />
3 migratory waterfowl floodpla<strong>in</strong><br />
4 deer, raccoon, turkey terraces and floodpla<strong>in</strong><br />
5 nut, fruits, berries terraces and floodpla<strong>in</strong><br />
6 pioneer seed plants floodpla<strong>in</strong><br />
7 alligators, crawfish, frogs, turtles floodpla<strong>in</strong><br />
This simple tabulation suggests that overflow bottomlands were <strong>of</strong> enormous importance<br />
to <strong>the</strong> local Mississippi period population at any given period which had “two str<strong>in</strong>gs<br />
to its bow—wild and agricultural resource....time devoted to hunt<strong>in</strong>g and ga<strong>the</strong>r<strong>in</strong>g [here<br />
primarily <strong>in</strong> floodpla<strong>in</strong> environments]... was almost equal to that devoted to agriculture” [<strong>in</strong><br />
uplands] (Peebles 1978:392).<br />
3. What are some “second l<strong>in</strong>e” food resources and nonfood materials specifically<br />
available or abundant <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>?<br />
We cannot claim to have made a systematic appraisal <strong>of</strong> <strong>the</strong>se resources, but <strong>the</strong> follow<strong>in</strong>g<br />
items are suggested:<br />
Second L<strong>in</strong>e Foods O<strong>the</strong>r Raw Materials<br />
snakes, small turtles alluvial clays<br />
mussels, snails novaculite and o<strong>the</strong>r gravels<br />
beaver, otter, muskrat cypress and o<strong>the</strong>r implement woods<br />
honey, maple sap hardwood fuels<br />
persimmon, riverbank grapes, small fruits bark, fiber, gum<br />
seeds, shoots, greens furs, sk<strong>in</strong>s, fea<strong>the</strong>rs<br />
tallow, grease
chapter 3<br />
prehistory <strong>of</strong> <strong>the</strong> felsenthAl region<br />
by John H. House<br />
<strong>in</strong>troduction<br />
The Felsenthal archeological region, recognized <strong>in</strong> <strong>the</strong> State Plan for <strong>the</strong> Conservation<br />
<strong>of</strong> Archeological Resources <strong>in</strong> <strong>Arkansas</strong> (Davis 1981) is a subdivision <strong>of</strong> <strong>the</strong> Lower Mississippi<br />
Valley archeological subarea <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast. The Felsenthal region was given formal def<strong>in</strong>ition<br />
by Schambach (Rol<strong>in</strong>gson and Schambach 1981:103-106) and assigned to <strong>the</strong> Lower<br />
Valley on <strong>the</strong> basis <strong>of</strong> physiographic and biogeographic cont<strong>in</strong>uities with <strong>the</strong> Lower Valley,<br />
as well as archeological evidence. Throughout much <strong>of</strong> <strong>the</strong> prehistoric and protohistoric<br />
aborig<strong>in</strong>al sequence, components <strong>in</strong> <strong>the</strong> Felsenthal region exhibit greater formal similarities<br />
to contemporary archeological culture units <strong>in</strong> <strong>the</strong> Lower Valley to <strong>the</strong> east than to culture<br />
units <strong>in</strong> <strong>the</strong> Caddo subarea to <strong>the</strong> west.<br />
The Felsenthal archeological region encompasses <strong>the</strong> Felsenthal Navigation Pool<br />
project area and <strong>the</strong> Felsenthal National Wildlife Refuge (see Chapter 1). The region centers<br />
on <strong>the</strong> Ouachita River between Camden, <strong>Arkansas</strong>, and Ouachita City, Louisiana, where<br />
<strong>the</strong> Ouachita River enters <strong>the</strong> Mississippi Alluvial Pla<strong>in</strong>. It also extends up <strong>the</strong> Sal<strong>in</strong>e River<br />
to <strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> Rison, <strong>Arkansas</strong> and <strong>in</strong>cludes contiguous West Gulf Coastal Pla<strong>in</strong> upland<br />
areas with<strong>in</strong> <strong>the</strong> Ouachita River and Sal<strong>in</strong>e River watersheds. In <strong>Arkansas</strong>, adjo<strong>in</strong><strong>in</strong>g archeological<br />
regions recognized <strong>in</strong> <strong>the</strong> State Plan (Davis 1981) are: <strong>the</strong> Great Bend <strong>of</strong> <strong>the</strong> Red River<br />
and <strong>the</strong> Little Missouri regions to <strong>the</strong> west, <strong>the</strong> Middle Ouachita (cf. Hodges and Hodges<br />
1945) and Middle Sal<strong>in</strong>e regions on <strong>the</strong> north, and <strong>the</strong> Bartholomew-Macon region on <strong>the</strong><br />
east.<br />
The assignment <strong>of</strong> <strong>the</strong> Felsenthal region to <strong>the</strong> Lower Mississippi Valley subarea reverses<br />
<strong>the</strong> previous assignment <strong>of</strong> this portion <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast to <strong>the</strong> Caddoan subarea (Davis<br />
1961; Phillips 1970:861; cf. Rol<strong>in</strong>gson 1976a:100). Schambach (1979:21) has emphasized,<br />
however, that <strong>the</strong> Felsenthal region is on a frontier or border area between <strong>the</strong>se two major<br />
archeological subareas <strong>of</strong> North America and that <strong>the</strong> archeological record <strong>in</strong> <strong>the</strong> region<br />
should reflect this frontier situation.<br />
The purpose <strong>of</strong> <strong>the</strong> present chapter is to review <strong>the</strong> history <strong>of</strong> archeological <strong>in</strong>vestigation<br />
<strong>in</strong> <strong>the</strong> Felsenthal region and briefly summarize our current knowledge <strong>of</strong> <strong>the</strong> cultural<br />
sequence and <strong>the</strong> nature <strong>of</strong> cultural development <strong>in</strong> <strong>the</strong> region dur<strong>in</strong>g prehistoric and protohistoric<br />
times. The history <strong>of</strong> <strong>the</strong> region will be discussed by Watk<strong>in</strong>s <strong>in</strong> Chapter 4 <strong>of</strong> this
52 House<br />
report. Locations <strong>of</strong> <strong>the</strong> pr<strong>in</strong>cipal archeological sites mentioned <strong>in</strong> this chapter are <strong>in</strong>dicated<br />
on <strong>the</strong> map <strong>of</strong> <strong>the</strong> Felsenthal region, Figure 13.<br />
preVious ArcheologicAl <strong>in</strong>VestigAti0n<br />
exploration<br />
Although sporadic references to aborig<strong>in</strong>al mounds are made <strong>in</strong> <strong>the</strong> writ<strong>in</strong>gs <strong>of</strong> early<br />
explorers <strong>in</strong> <strong>the</strong> region, notably Dunbar and Hunter (Rowland 1930; McDermott 1963; see<br />
also Chapter 4 <strong>of</strong> this report), <strong>the</strong> first scientific archeological <strong>in</strong>vestigation <strong>in</strong> or near <strong>the</strong><br />
Felsenthal region appears to have been made by Cyrus Thomas (1894:250) at Pargoud Land<strong>in</strong>g<br />
(16OU1) on <strong>the</strong> Ouachita River near Monroe, Louisiana. The first systematic reconnaissance<br />
<strong>in</strong> <strong>the</strong> region, however, was by Clarence B. Moore <strong>of</strong> <strong>the</strong> Academy <strong>of</strong> Natural Sciences<br />
<strong>of</strong> Philadelphia <strong>in</strong> <strong>the</strong> early decades <strong>of</strong> this century. In 1908-1909, Moore (1909) ascended<br />
<strong>the</strong> Ouachita River as far as Camden, <strong>Arkansas</strong>. On <strong>the</strong> sou<strong>the</strong>rn marg<strong>in</strong> <strong>of</strong> <strong>the</strong> region<br />
Moore briefly <strong>in</strong>vestigated Pargoud Land<strong>in</strong>g and dug extensively at <strong>the</strong> Glendora site on<br />
<strong>the</strong> Ouachita and <strong>the</strong> nearby Keno site on <strong>the</strong> lower reaches <strong>of</strong> Bayou Bartholomew. In <strong>the</strong><br />
Felsenthal region <strong>in</strong> <strong>Arkansas</strong>, Moore <strong>in</strong>vestigated <strong>the</strong> follow<strong>in</strong>g sites <strong>in</strong> 1908-1909:<br />
Mound near Lock Number 6 (3AS6), Ashley County<br />
Mounds near Green Lake (3BR1), Bradley County<br />
Cemetery near Caryville Land<strong>in</strong>g (3UN11), Union County<br />
Mounds near Pigeon Hill (3UN12), Union County<br />
Cemetery <strong>in</strong> Boytts Field (3UN13), Union County<br />
Mounds near Purdue Wood-Camp (no assigned state site number), Union County<br />
Cemetery at Bell G<strong>in</strong> Land<strong>in</strong>g (3UN99), Union County<br />
Mounds at Bell Land<strong>in</strong>g (3UN51), Union County<br />
Mounds at Boone Place (3CA9), Calhoun County<br />
Mounds at Keller Place (3CA13), Calhoun County<br />
Mounds and sites near Pyles Land<strong>in</strong>g (3CA26), Calhoun County<br />
Mound and cemetery at Kent (3OU6), Ouachita County<br />
Return<strong>in</strong>g to <strong>the</strong> region <strong>in</strong> 1912-1913, Moore (1913) ascended <strong>the</strong> Sal<strong>in</strong>e River as far as<br />
<strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> Warren, <strong>Arkansas</strong>. Sites <strong>in</strong>vestigated by Moore on <strong>the</strong> Sal<strong>in</strong>e River <strong>in</strong> <strong>Arkansas</strong><br />
are:<br />
Site on Goulett Island (3BR8), Bradley County<br />
Site at Hampton Land<strong>in</strong>g (no assigned state site number), Bradley County<br />
Site near Gee Land<strong>in</strong>g (3DR<strong>17</strong>), Drew County<br />
Site near Lowrie Land<strong>in</strong>g (3BR5), Bradley County<br />
Mound near Wherry Land<strong>in</strong>g (no assigned state site number), Bradley County
Figure 13. A portion <strong>of</strong> <strong>the</strong> Ouachita River Valley designated <strong>the</strong> Felsenthal region and<br />
locations <strong>of</strong> some important prehistoric sites.<br />
53
54 House<br />
Site near Wire Fence Land<strong>in</strong>g (no assigned state site number), Bradley County<br />
Site near Sutton Ferry (3BR6), Bradley County<br />
Site near Brooks Land<strong>in</strong>g (no assigned state site number), Bradley County<br />
The excavations <strong>in</strong> aborig<strong>in</strong>al cemeteries at Glendora (Moore 1909: 27-80) and Keno<br />
(Moore 1909:120-151) yielded polished and engraved ceramic bowls and bottles occurr<strong>in</strong>g<br />
with early European trade goods. Recovery <strong>of</strong> <strong>the</strong>se vessels by Moore led to <strong>the</strong> subsequent<br />
long stand<strong>in</strong>g assignment <strong>of</strong> this portion <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast to <strong>the</strong> Caddoan subarea (Davis<br />
1961). Observations on an extensive human skeletal series recovered by Moore at Boytts<br />
Field (3UN13) <strong>in</strong> Union County, <strong>Arkansas</strong>, were reported by Hrdlicka (1909).<br />
In <strong>the</strong> decades follow<strong>in</strong>g Moore’s work, archeological <strong>in</strong>vestigations <strong>in</strong> <strong>the</strong> Ouachita<br />
River bas<strong>in</strong> were largely conf<strong>in</strong>ed to portions <strong>of</strong> <strong>the</strong> bas<strong>in</strong> above <strong>the</strong> Felsenthal region<br />
(e.g., Harr<strong>in</strong>gton 1920; Hodges and Hodges 1945; Wood 1963; Schambach 1970) or below<br />
<strong>the</strong> Felsenthal region (e.g., Ford 1935, 1936; Walker 1936; Ford and Willey 1940). Price and<br />
Heartfield (1977:4) refer to unpublished <strong>in</strong>vestigations by James A. Ford at Pargoud Land<strong>in</strong>g<br />
(16OU1) <strong>in</strong> 1935 result<strong>in</strong>g <strong>in</strong> recovery <strong>of</strong> a flexed burial.<br />
Dur<strong>in</strong>g <strong>the</strong> 1940s and 1950s, amateur archeologist Frank J. Soday and associates <strong>of</strong><br />
El Dorado, <strong>Arkansas</strong>, excavated at a number <strong>of</strong> late prehistoric sites with<strong>in</strong> <strong>the</strong> Felsenthal<br />
region primarily at Locust Ridge (3UN8), Watts Field (3UN18), and Jones Lake (3UN81).<br />
No report on <strong>the</strong>se excavations has ever been made available but copies <strong>of</strong> site records were<br />
given by Soday to <strong>the</strong> <strong>Arkansas</strong> Archeological Survey. Frank Schambach and J. Cynthia<br />
Weber <strong>of</strong> <strong>the</strong> <strong>Arkansas</strong> Archeological Survey subsequently made photographic records <strong>of</strong><br />
ceramic vessels with site and grave lot provenience <strong>in</strong> <strong>the</strong> Soday and Franz collections from<br />
<strong>the</strong> Felsenthal region.<br />
Recent Investigations<br />
Archeological reconnaissance and small scale excavations <strong>in</strong> <strong>the</strong> Felsenthal region by<br />
amateur members <strong>of</strong> <strong>the</strong> <strong>Arkansas</strong> Archeological Society began <strong>in</strong> <strong>the</strong> 1960s. After 1968 <strong>the</strong>se<br />
efforts were augmented by reconnaissance and test<strong>in</strong>g by <strong>Arkansas</strong> Archeological Survey<br />
station archeologists from Magnolia and Monticello. These activities have contributed to a<br />
grow<strong>in</strong>g regional data base <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong> portion <strong>of</strong> <strong>the</strong> Felsenthal region. Excavations<br />
carried out at <strong>the</strong> Paw Paw site (3OU22) near Camden as a part <strong>of</strong> <strong>the</strong> <strong>Arkansas</strong> Archeological<br />
Society tra<strong>in</strong><strong>in</strong>g session <strong>in</strong> 1971 yielded data on Fourche Mal<strong>in</strong>e and Archaic occupations<br />
at <strong>the</strong> nor<strong>the</strong>rn marg<strong>in</strong> <strong>of</strong> <strong>the</strong> Felsenthal region (Weber, excavation and analysis notes, 1973,<br />
<strong>Arkansas</strong> Archeological Survey).
Prehistory 55<br />
The history <strong>of</strong> recent archeological <strong>in</strong>vestigations along <strong>the</strong> Ouachita River <strong>in</strong> <strong>the</strong> Monroe,<br />
Louisiana, locality is recounted by Price and Heartfield (1977:6-7) and by Price (U.S.<br />
Army Corps <strong>of</strong> Eng<strong>in</strong>eers 1980). In <strong>the</strong> 1960s <strong>the</strong> Nor<strong>the</strong>ast Louisiana Archeological Society<br />
conducted excavations at Myatt’s Land<strong>in</strong>g, one <strong>of</strong> <strong>the</strong> sites <strong>in</strong>vestigated by Moore, but no<br />
report on this work has been published. Excavations by <strong>the</strong> Nor<strong>the</strong>ast Louisiana Archeological<br />
Society and Nor<strong>the</strong>astern Louisiana <strong>University</strong> at <strong>the</strong> suspected site <strong>of</strong> Colonial Spanish<br />
Fort Miro (16OU3) <strong>in</strong> downtown Monroe are reported by Greene et al. (1975). Follow-up excavations<br />
at this site were funded by <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg District,<br />
supported by <strong>the</strong> city <strong>of</strong> Monroe and directed by archeologists from Nor<strong>the</strong>ast Louisiana<br />
<strong>University</strong> (Price et al. 1975).<br />
Subsequent <strong>in</strong>vestigations <strong>in</strong> <strong>the</strong> Monroe locality <strong>in</strong>clude small-scale cultural resource<br />
management surveys conducted by archeologists from Nor<strong>the</strong>astern Louisiana <strong>University</strong>,<br />
as well as ongo<strong>in</strong>g surveys and excavations sponsored by <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers,<br />
Vicksburg District <strong>in</strong> conjunction with navigation channel works on <strong>the</strong> Ouachita River (U.S.<br />
Army Corps <strong>of</strong> Eng<strong>in</strong>eers 1980). The latter efforts <strong>in</strong>clude salvage excavations at <strong>the</strong> T. E.<br />
Salsbury site (16OU15) where numerous Mississippi period burials accompanied by ceramic<br />
vessels, largely assignable to <strong>the</strong> Plaquem<strong>in</strong>e Culture, were recovered (Price and Heartfield<br />
1977).<br />
Glen S. Green <strong>of</strong> Nor<strong>the</strong>astern Louisiana <strong>University</strong> conducted fur<strong>the</strong>r excavations at<br />
Pargoud Land<strong>in</strong>g (16OU1) <strong>in</strong> 1975, 1976, 1977, and 1979. No report on <strong>the</strong>se <strong>in</strong>vestigations<br />
is yet available but prelim<strong>in</strong>ary results are briefly summarized <strong>in</strong> a recent cultural resources<br />
survey report (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers 1980, Chapter 5).<br />
Investigations by <strong>the</strong> <strong>Arkansas</strong> Archeological Survey <strong>in</strong> <strong>the</strong> Felsenthal National Wildlife<br />
Refuge began with an <strong>in</strong>itial survey or reconnaissance conducted by Rol<strong>in</strong>gson and<br />
Schambach <strong>in</strong> 1971. This reconnaissance <strong>in</strong>volved visits to previously recorded sites and<br />
record<strong>in</strong>g <strong>of</strong> new sites (Rol<strong>in</strong>gson 1972). The 65 sites on record at <strong>the</strong> conclusion <strong>of</strong> <strong>the</strong><br />
project <strong>in</strong>cluded a number <strong>of</strong> late prehistoric mound centers and habitation sites on terraces<br />
overlook<strong>in</strong>g <strong>the</strong> floodpla<strong>in</strong> and several small, low density prehistoric artifact scatters <strong>in</strong> <strong>the</strong><br />
floodpla<strong>in</strong>. Artifacts recovered suggested closer ties to <strong>the</strong> Coles Creek and Plaquem<strong>in</strong>e<br />
cultures <strong>of</strong> <strong>the</strong> Lower Mississippi Valley than had been anticipated. Evidence <strong>of</strong> occupation<br />
dat<strong>in</strong>g to <strong>the</strong> Archaic, Poverty Po<strong>in</strong>t, Tchula, and Marksville periods was also recognized.<br />
The probability that many sites may be buried under recent alluvium <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> was<br />
also noted (Rol<strong>in</strong>gson 1972).
56 House<br />
The follow<strong>in</strong>g year, test excavations were conducted at six sites <strong>in</strong> <strong>the</strong> Felsenthal<br />
National Wildlife Refuge by <strong>the</strong> <strong>Arkansas</strong> Archeological Survey under a cooperative agreement<br />
with <strong>the</strong> National Park Service. This project, directed by Lischka, <strong>in</strong>cluded test<strong>in</strong>g at<br />
<strong>the</strong> follow<strong>in</strong>g sites: Shallow Lake (3UN9/52), Bent Tree (3UN75), Locust Ridge (3UN8), Watts<br />
Field (3UN18), Disturbed Cemetery (3UN84), and Oven (3UN53). The major components<br />
represented at <strong>the</strong>se sites were attributed by Lischka (1973) to Coles Creek and Plaquem<strong>in</strong>e<br />
culture occupation. Evidence <strong>of</strong> Marksville and Archaic period occupations was also encountered.<br />
The program <strong>of</strong> test<strong>in</strong>g <strong>in</strong>itiated by Lischka was cont<strong>in</strong>ued by <strong>in</strong>com<strong>in</strong>g Survey archeologist<br />
V. K. Pheriba Stacy <strong>in</strong> 1975. Her work, jo<strong>in</strong>tly sponsored by <strong>the</strong> Interagency Archeological<br />
Services, <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg District, and <strong>the</strong> <strong>Arkansas</strong><br />
Archeological Survey, consisted <strong>of</strong> excavation <strong>of</strong> two test pits at Pereogee<strong>the</strong> Lake 1 (3BR52)<br />
and <strong>in</strong>tensive excavation <strong>in</strong> Mound C at Shallow Lake (3UN9/52) which had been subject to<br />
vandalism. Stacy’s notes on <strong>the</strong>se excavations also drew toge<strong>the</strong>r data from Lischka’s work<br />
at Shallow Lake and o<strong>the</strong>r sites. A report on <strong>the</strong> Shallow Lake site was written by Martha<br />
Ann Rol<strong>in</strong>gson and Frank F. Schambach (1981). Schambach had done considerable work <strong>in</strong><br />
<strong>the</strong> Felsenthal region, and used <strong>the</strong> data as a vehicle for a regional summary and <strong>in</strong>troduction<br />
<strong>of</strong> a new ceramic typology (Rol<strong>in</strong>gson and Schambach 1981: 101-<strong>17</strong>6). His ceramic classification<br />
is a response to difficulties previously encountered <strong>in</strong> describ<strong>in</strong>g and classify<strong>in</strong>g<br />
Felsenthal region ceramics by ei<strong>the</strong>r <strong>the</strong> current Lower Mississippi Valley (Phillips 1970) or<br />
Caddoan areas (Suhm and Jelks 1964; cf. Webb 1981) typologies. This analytical classification<br />
is discussed and applied to a large ceramic sample from a newly recorded floodpla<strong>in</strong> site by<br />
Hemm<strong>in</strong>gs <strong>in</strong> Chapter 8 <strong>of</strong> this report.<br />
A cultural resources assessment <strong>of</strong> <strong>the</strong> projected Corps <strong>of</strong> Eng<strong>in</strong>eers Calion Lock and<br />
Dam project on <strong>the</strong> Ouachita River near Calion, <strong>Arkansas</strong> was carried out by David Kelley<br />
<strong>of</strong> <strong>the</strong> <strong>Arkansas</strong> Archeological Survey <strong>in</strong> 1978-1979. The fieldwork <strong>in</strong>volved systematic subsurface<br />
test<strong>in</strong>g <strong>in</strong> <strong>the</strong> entire 244.4 ha impact zone located <strong>in</strong> <strong>the</strong> Ouachita River floodpla<strong>in</strong>.<br />
The study provided valuable new <strong>in</strong>formation on <strong>the</strong> Bell G<strong>in</strong> Land<strong>in</strong>g site (3UN99) which<br />
had been <strong>in</strong>vestigated by Moore <strong>in</strong> 1908-1909 and revealed one previously unrecorded site<br />
<strong>in</strong> <strong>the</strong> impact zone, an isolated f<strong>in</strong>d <strong>of</strong> a late Archaic po<strong>in</strong>t (Kelley 1979).<br />
A cultural resources reconnaissance <strong>of</strong> <strong>the</strong> Felsenthal National Wildlife Refuge was<br />
conducted by Heartfield, Price and Greene, Inc. <strong>of</strong> Monroe, Louisiana, <strong>in</strong> fall 1979 concurrent<br />
with present <strong>in</strong>vestigations. This study identified five additional archeological sites,<br />
four isolated f<strong>in</strong>ds, and n<strong>in</strong>e stand<strong>in</strong>g structures <strong>in</strong> addition to <strong>the</strong> 63 archeological sites<br />
already on record with<strong>in</strong> <strong>the</strong> refuge (Heartfield, Price and Greene, Inc. 1980).
Prehistory 57<br />
Recent and ongo<strong>in</strong>g efforts to identify and evaluate archeological resources <strong>in</strong> proposed<br />
lignite strip m<strong>in</strong><strong>in</strong>g areas <strong>in</strong> <strong>the</strong> West Gulf Coastal Pla<strong>in</strong> uplands near Hampton <strong>in</strong><br />
Calhoun County, <strong>Arkansas</strong>, have focused attention on <strong>the</strong> archeological record <strong>of</strong> prehistoric<br />
and historic occupation <strong>in</strong> <strong>the</strong> h<strong>in</strong>terlands <strong>of</strong> <strong>the</strong> Felsenthal region. An overview study<br />
conducted by Timothy C. Kl<strong>in</strong>ger (1979) <strong>in</strong>volved reconnaissance and <strong>in</strong>formant <strong>in</strong>terview<br />
survey <strong>in</strong> <strong>the</strong> 16,194 ha Hampton Prospect. This resulted <strong>in</strong> <strong>the</strong> identification and record<strong>in</strong>g<br />
<strong>of</strong> a number <strong>of</strong> prehistoric sites <strong>in</strong> spite <strong>of</strong> <strong>the</strong> difficult survey conditions encountered <strong>in</strong> <strong>the</strong><br />
heavily wooded project area.<br />
Phase II assessment at Hampton (Lafferty et al. 1981) <strong>in</strong>volved an <strong>in</strong>tensive survey <strong>of</strong><br />
10% <strong>of</strong> <strong>the</strong> redef<strong>in</strong>ed (and renamed) 14,<strong>17</strong>5 ha Sparta M<strong>in</strong>e area. As a cumulative result <strong>of</strong><br />
<strong>the</strong>se efforts, 27 prehistoric sites have been identified <strong>in</strong> <strong>the</strong> project area. Although most <strong>of</strong><br />
<strong>the</strong>se sites were discovered by systematic shovel test<strong>in</strong>g <strong>in</strong> wooded areas and yielded only<br />
very small samples <strong>of</strong> cultural material, <strong>the</strong>y represent Archaic, Poverty Po<strong>in</strong>t, Tchula or<br />
Marksville, and Mississippi period occupations. Fur<strong>the</strong>r analysis <strong>of</strong> <strong>the</strong> Sparta M<strong>in</strong>e data<br />
<strong>in</strong>volved exam<strong>in</strong>ation <strong>of</strong> <strong>the</strong> relationship between prehistoric site location and selected biophysical<br />
environmental variables <strong>in</strong> <strong>the</strong> project area and formulation <strong>of</strong> a predictive model<br />
<strong>of</strong> site location (to be tested <strong>in</strong> future work <strong>in</strong> <strong>the</strong> Sparta M<strong>in</strong>e area).<br />
Recent <strong>in</strong>vestigation <strong>in</strong> <strong>the</strong> Sal<strong>in</strong>e River watershed portion <strong>of</strong> <strong>the</strong> Felsenthal region<br />
has been limited to reconnaissance, test<strong>in</strong>g, and small scale salvage excavation by members<br />
<strong>of</strong> <strong>the</strong> <strong>Arkansas</strong> Archeological Society and <strong>the</strong> Survey research station at Monticello and<br />
several small cultural resource management surveys. Published accounts <strong>in</strong>clude White’s<br />
(1970) report on salvage excavations at a very late prehistoric cemetery at Gee’s Land<strong>in</strong>g<br />
(3DR<strong>17</strong>), one <strong>of</strong> <strong>the</strong> sites <strong>in</strong>vestigated by Moore <strong>in</strong> 1912-1913, and Reaves and McClurkan’s<br />
(1970) brief note on salvage work at <strong>the</strong> Dansbury Creek site (3CV24) near Rison, which also<br />
represents very late prehistoric or protohistoric occupation.<br />
<strong>the</strong> seQuence <strong>of</strong> Aborig<strong>in</strong>Al occupAtion<br />
Knowledge <strong>of</strong> prehistoric and protohistoric cultural development <strong>in</strong> <strong>the</strong> Felsenthal archeological<br />
region is <strong>the</strong> result <strong>of</strong> slow accumulation <strong>of</strong> data from early explorations, reconnaissance,<br />
and small scale excavation by pr<strong>of</strong>essional and amateur archeologists, and recent<br />
cultural resource management efforts. The follow<strong>in</strong>g chronological survey <strong>of</strong> this topic relies<br />
heavily on a summary by Schambach <strong>in</strong> <strong>the</strong> Hampton Prospect (Sparta M<strong>in</strong>e) overview<br />
(Kl<strong>in</strong>ger 1979), and discussions by Schambach and Early <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong> State Plan (Davis<br />
1981).
58 House<br />
Table 4 presents an outl<strong>in</strong>e <strong>of</strong> <strong>the</strong> Felsenthal archeological sequence, as it is currently<br />
known, and aligns it with <strong>the</strong> sequence <strong>of</strong> adjo<strong>in</strong><strong>in</strong>g regions. The column on <strong>the</strong> left consists<br />
<strong>of</strong> archeological periods recognized <strong>in</strong> <strong>the</strong> Lower Mississippi Valley. For <strong>the</strong> latter part <strong>of</strong><br />
<strong>the</strong> sequence, <strong>the</strong>se periods follow Phillips (1970:Figure 2). The archeological culture units<br />
under <strong>the</strong> various regional head<strong>in</strong>gs, not <strong>in</strong> paren<strong>the</strong>ses, are phases (Willey and Phillips<br />
1957:22). Entities <strong>in</strong> paren<strong>the</strong>ses are artifact complexes, archeological cultures or, <strong>in</strong> <strong>the</strong> case<br />
<strong>of</strong> <strong>the</strong> Great Bend region, periods recognized <strong>in</strong> <strong>the</strong> Caddoan subarea. Intervals <strong>in</strong> regional<br />
sequences for which occupation is yet poorly def<strong>in</strong>ed archeologically are designated “Unnamed.”<br />
Correspond<strong>in</strong>g <strong>in</strong>tervals for which few or no published data exist are left blank.<br />
The Middle Ouachita region sequence follows Schambach (1970:Table 27). The Great Bend<br />
sequence follows Davis (1970) and subsequent discussions by Schambach <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong><br />
State Plan (Davis 1981). The Bartholomew sequence follows Rol<strong>in</strong>gson (1973, 1976a). The<br />
Upper Tensas sequence follows Hally (1967) and Belmont (1979) and <strong>the</strong> discussion <strong>of</strong> phase<br />
distributions <strong>in</strong> Phillips (1970:861-976).<br />
S<strong>in</strong>ce archeological cultures and archeological periods <strong>in</strong> <strong>the</strong> Lower Mississippi Valley<br />
may have <strong>the</strong> same name, it will be important <strong>in</strong> <strong>the</strong> follow<strong>in</strong>g discussion to rigorously<br />
dist<strong>in</strong>guish between <strong>the</strong> two. Assignment <strong>of</strong> a given component to a certa<strong>in</strong> archeological<br />
culture (e.g., Marksville culture) will imply a high degree <strong>of</strong> formal similarity on <strong>the</strong> total assemblage<br />
level to <strong>the</strong> named archeological culture. Assignment <strong>of</strong> a component to a certa<strong>in</strong><br />
period (e.g., Marksville period), on <strong>the</strong> o<strong>the</strong>r hand, will imply only that <strong>the</strong> component dates<br />
to <strong>the</strong> specified time <strong>in</strong>terval; similarities on <strong>the</strong> total assemblage level may be slight.<br />
Paleo-Indian Period, before 8000 B.C.<br />
The archeological components <strong>of</strong> <strong>the</strong> Paleo-Indian period found throughout North<br />
America are regarded as represent<strong>in</strong>g hunter-ga<strong>the</strong>rer populations liv<strong>in</strong>g <strong>in</strong> term<strong>in</strong>al<br />
Pleistocene or transitional Pleistocene-Holocene environments. The fluted projectile po<strong>in</strong>t<br />
horizon represented by <strong>in</strong>frequent sites, cont<strong>in</strong>entwide, is dated to ca 9,000 B.C. while <strong>the</strong><br />
Dalton culture, largely conf<strong>in</strong>ed to <strong>the</strong> Sou<strong>the</strong>ast, appears to have flourished immediately after<br />
<strong>the</strong> Pleistocene-Holocene transition, ca 8500 B.C. (Goodyear et al. 1979:96-99). Dalton culture<br />
exhibits both cont<strong>in</strong>uities and discont<strong>in</strong>uities with earlier fluted po<strong>in</strong>t complexes; some<br />
<strong>in</strong>vestigators, emphasiz<strong>in</strong>g <strong>the</strong> latter, assign Dalton to <strong>the</strong> Early Archaic period (e.g., Goodyear<br />
et al. 1979:96). Scottsbluff and related projectile po<strong>in</strong>t forms may be roughly coeval with<br />
Dalton but have contrast<strong>in</strong>g biogeographic distribution. The extent <strong>of</strong> Paleo-Indian period<br />
occupation <strong>in</strong> many regions is particularly difficult to assess because <strong>of</strong> alluvial burial <strong>of</strong><br />
sites <strong>in</strong> river<strong>in</strong>e zones.
Prehistory 59<br />
Table 4. Comparison <strong>of</strong> Archeological Sequences <strong>in</strong> <strong>the</strong> Felsenthal and Adjacent Archeological<br />
Regions.
60 House<br />
Schambach (1979:24) notes that fluted po<strong>in</strong>ts resembl<strong>in</strong>g <strong>the</strong> Clovis type have been<br />
recorded <strong>in</strong> small numbers from several counties <strong>in</strong> sou<strong>the</strong>ast <strong>Arkansas</strong>. The one specimen<br />
on record from <strong>the</strong> Felsenthal region is <strong>in</strong> an amateur’s collection from <strong>the</strong> Shallow Lake site<br />
(3UN9/52) on <strong>the</strong> Pleistocene Deweyville Terrace surface.<br />
Scottsbluff and Dalton po<strong>in</strong>ts are comparatively frequent <strong>in</strong> southwest <strong>Arkansas</strong>.<br />
Schambach (1979:24) reports a Scottsbluff po<strong>in</strong>t from site 3OU148 on <strong>the</strong> Ouachita River<br />
across from <strong>the</strong> mouth <strong>of</strong> Champagnolle Creek. With<strong>in</strong> <strong>the</strong> Felsenthal National Wildlife Refuge<br />
area, a buried Dalton culture component appears present at Coon Island (3BR10) on <strong>the</strong><br />
Ouachita River floodpla<strong>in</strong> (Schambach 1979:26). A private collection from <strong>the</strong> Shallow Lake<br />
site <strong>in</strong>cludes six Dalton po<strong>in</strong>ts (Rol<strong>in</strong>gson and Schambach 1981:92). Rol<strong>in</strong>gson and Schambach<br />
(1981:<strong>17</strong>8) note that Dalton po<strong>in</strong>ts <strong>in</strong> <strong>the</strong> Felsenthal region have been found as much as<br />
a meter deep <strong>in</strong> Pleistocene terrace deposits.<br />
The Snow Hill complex is <strong>the</strong> name given by Schambach (1979:24) to a reported—but<br />
unverified—discovery by a collector <strong>of</strong> apparent Paleo-Indian period artifacts at five sites<br />
near Smackover <strong>in</strong> Union County, <strong>Arkansas</strong>. The artifacts consist <strong>of</strong> some 50 Angostura-like<br />
po<strong>in</strong>ts, termed “Snow Hill po<strong>in</strong>ts” by Schambach, plus choppers, uniface scrapers, gravers,<br />
denticulates and polished magnetite “bola stones.” Although <strong>the</strong> specimens do appear to be<br />
au<strong>the</strong>ntic artifacts, <strong>the</strong>y were excavated and collected <strong>in</strong> secrecy. No similar assemblage has<br />
ever been identified at any o<strong>the</strong>r site <strong>in</strong> southwest <strong>Arkansas</strong>.<br />
Early Archaic Period, 8000-6000 B.C.<br />
The Early Archaic period, toge<strong>the</strong>r with <strong>the</strong> earlier Dalton culture horizon, is regarded<br />
as <strong>the</strong> period <strong>of</strong> <strong>in</strong>itial adaptation to <strong>the</strong> varied postglacial environments <strong>of</strong> eastern North<br />
America. Dalton and Early Archaic sites are far more frequent than earlier fluted po<strong>in</strong>t sites<br />
and settlement pattern shifts are <strong>in</strong>dicated; this period <strong>of</strong>ten represents <strong>the</strong> earliest occupation<br />
<strong>in</strong> many caves and alluvial sites with subsequent long prehistoric sequences. Technological<br />
discont<strong>in</strong>uities with <strong>the</strong> earlier fluted po<strong>in</strong>t complexes appear to become more<br />
pronounced through <strong>the</strong> time span <strong>of</strong> <strong>the</strong> Early Archaic period (Griff<strong>in</strong> 1967:<strong>17</strong>8; Morse<br />
1973; Goodyear et al. 1979:97-99).<br />
The little known San Patrice culture (Webb et al. 1971) is localized <strong>in</strong> north Louisiana,<br />
south <strong>Arkansas</strong>, and east Texas. Its relationship to <strong>the</strong> Dalton culture is yet undeterm<strong>in</strong>ed<br />
but Schambach (1979:26) suggests that it may locally represent <strong>the</strong> beg<strong>in</strong>n<strong>in</strong>gs <strong>of</strong> “regional<br />
differentiation” <strong>of</strong> Sou<strong>the</strong>astern Archaic cultures <strong>in</strong> post-Dalton times. San Patrice projectile<br />
po<strong>in</strong>ts (Suhm and Jelks 1962:242), side notched po<strong>in</strong>ts, and “Albany scrapers,” all attributes<br />
<strong>of</strong> <strong>the</strong> San Patrice culture, have been found <strong>in</strong> <strong>the</strong> Felsenthal National Wildlife Refuge at<br />
Watts Field (3UN18) and Shallow Lake (3UN9/52).
Middle Archaic Period, 6000-4000 B.C.<br />
Prehistory 61<br />
The Middle Archaic period <strong>in</strong> eastern North America is marked by some technological<br />
<strong>in</strong>novations such as stone atlatl weights and <strong>the</strong> earliest widespread use <strong>of</strong> ground and polished<br />
stone tools and by evidence <strong>of</strong> more <strong>in</strong>tensive and prolonged occupation <strong>of</strong> <strong>in</strong>dividual<br />
sites with possible <strong>in</strong>creased reliance on aquatic resources (Griff<strong>in</strong> 1967:<strong>17</strong>8). This <strong>in</strong>terval<br />
correlates generally with <strong>the</strong> Atlantic climatic episode (Bryson et al. 1970). Paleoenvironmental<br />
data suggest dramatic shifts toward drier and warmer climate <strong>in</strong> nearby portions <strong>of</strong><br />
<strong>the</strong> East (Wood and McMillan 1976; K<strong>in</strong>g and Allen 1977). Response by human populations<br />
to this climatic change <strong>in</strong> some regions may have <strong>in</strong>cluded dramatic demographic shifts<br />
(Morse 1975:191).<br />
Investigation <strong>in</strong> southwest <strong>Arkansas</strong> has resulted <strong>in</strong> recognition <strong>of</strong> Tom’s Brook culture.<br />
Diagnostic attributes <strong>of</strong> this culture, first def<strong>in</strong>ed as a phase by Schambach (1970:384-<br />
385) at <strong>the</strong> Cooper site <strong>in</strong> <strong>the</strong> Middle Ouachita region, <strong>in</strong>clude Johnson po<strong>in</strong>ts (Scholtz 1967;<br />
Bartlett 1963), stemmed scrapers, and side-notched pebble “net s<strong>in</strong>kers.” Subsequent excavations<br />
at <strong>the</strong> Paw Paw site (3OU22) on <strong>the</strong> nor<strong>the</strong>rn fr<strong>in</strong>ge <strong>of</strong> <strong>the</strong> Felsenthal region revealed a<br />
1 m thick, deeply buried Tom’s Brook culture midden and <strong>the</strong> earliest human burials identified<br />
to date <strong>in</strong> south <strong>Arkansas</strong>. These excavations also yielded a radiocarbon date <strong>of</strong> 4690 ±<br />
70 B.C. (TX-1550) for <strong>the</strong> Tom’s Brook culture (Weber excavation and analysis notes, 1973,<br />
<strong>Arkansas</strong> Archeological Survey).<br />
Tom’s Brook culture components <strong>in</strong> <strong>the</strong> Felsenthal region have been grouped <strong>in</strong>to<br />
<strong>the</strong> Spoon Bend phase (Rol<strong>in</strong>gson and Schambach 1981:<strong>17</strong>5). A deeply buried Tom’s Brook<br />
culture component has been identified at <strong>the</strong> Short Brake site (3BR20) <strong>in</strong> <strong>the</strong> Ouachita River<br />
floodpla<strong>in</strong> (Schambach 1979:26).<br />
The Tom’s Brook culture is considered by Schambach (1979:26-27) to be <strong>the</strong> earliest <strong>in</strong><br />
this portion <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast to show a strong river<strong>in</strong>e orientation. An upland component<br />
<strong>in</strong> <strong>the</strong> Felsenthal region has been identified at 3UN36. This upland site assemblage lacks net<br />
s<strong>in</strong>kers but <strong>in</strong>cludes gr<strong>in</strong>d<strong>in</strong>g equipment and uniface scrapers. Components <strong>of</strong> <strong>the</strong> Tom’s<br />
Brook culture are widespread, occurr<strong>in</strong>g throughout <strong>the</strong> Ouachita River watershed <strong>in</strong> both<br />
<strong>the</strong> West Gulf Coastal Pla<strong>in</strong> and Ouachita Mounta<strong>in</strong>s, <strong>the</strong> <strong>Arkansas</strong> Valley and <strong>the</strong> Ozarks<br />
(Bartlett 1963) and perhaps <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong> River Lowland <strong>in</strong> <strong>the</strong> Lower Mississippi Alluvial<br />
Valley as well (House 1980).<br />
Side-notched Big Sandy projectile po<strong>in</strong>ts (Lewis and Lewis 1961:34, 37), diagnostic <strong>of</strong><br />
Schambach’s (1970:385-386) Crystal Mounta<strong>in</strong> phase, have been identified <strong>in</strong> private collections<br />
from <strong>the</strong> Felsenthal region but no sites yield<strong>in</strong>g <strong>the</strong>se po<strong>in</strong>ts <strong>in</strong> abundance or <strong>in</strong> good<br />
stratigraphic context have yet been discovered.
62 House<br />
In <strong>the</strong> Sou<strong>the</strong>ast <strong>the</strong>re is considerable confusion about <strong>the</strong> correct temporal placement<br />
<strong>of</strong> <strong>the</strong> side-notched po<strong>in</strong>t horizon on horizons (cf. Fowler 1959:Figure 14; Dejarnette et al.<br />
1962; Bartlett 1963:Table 2; Schambach 1970:387; Goodyear et al. 1979:100-105) but a general<br />
Middle Archaic position seems likely. Schambach (1970:386) tentatively associates a bone atlatl<br />
hook and a full-grooved axe with <strong>the</strong> Crystal Mounta<strong>in</strong> phase component at <strong>the</strong> Cooper<br />
site <strong>in</strong> <strong>the</strong> Middle Ouachita region.<br />
Late Archaic Period, 4000-1500 B.C.<br />
In much <strong>of</strong> <strong>the</strong> East it has proved useful to divide <strong>the</strong> 4000-500 B.C. <strong>in</strong>terval, generally<br />
known as <strong>the</strong> Late Archaic, <strong>in</strong>to an earlier and later segment. In <strong>the</strong> Lower Mississippi Valley<br />
<strong>the</strong> earlier segment, 4000-1200 B.C., is <strong>the</strong> Late Archaic proper while <strong>the</strong> latter segment<br />
is <strong>the</strong> Poverty Po<strong>in</strong>t period. The Late Archaic period throughout eastern North America is<br />
seen as a time <strong>of</strong> considerable human population growth, dist<strong>in</strong>ct regional adaptations, and<br />
<strong>in</strong>terregional exchange <strong>of</strong> raw materials (Griff<strong>in</strong> 1967:<strong>17</strong>8). Rema<strong>in</strong>s <strong>of</strong> both tropical and native<br />
North American cultigens have been recovered from some eastern North American sites<br />
(Chomko and Crawford 1978) as early as 2300 B.C. (Marquardt and Watson 1976). Recent<br />
data from Mississippi (Connaway et al. 1977) demonstrate that <strong>the</strong> beg<strong>in</strong>n<strong>in</strong>gs <strong>of</strong> an elaborate<br />
lapidary <strong>in</strong>dustry <strong>in</strong> <strong>the</strong> Lower Valley dat<strong>in</strong>g to <strong>the</strong> 3000 B.C. time level or earlier.<br />
The White Oak phase <strong>in</strong> <strong>the</strong> Middle Ouachita region was def<strong>in</strong>ed by Schambach<br />
(1970:388) on <strong>the</strong> basis <strong>of</strong> <strong>the</strong> compact horizontal distribution <strong>of</strong> corner-notched Williams<br />
po<strong>in</strong>ts (Suhm and Jelks 1962:259) at <strong>the</strong> Cooper site. Stratigraphic data support a ca 2000<br />
B.C. placement for this component (Schambach 1970:Table 27). This horizon is poorly def<strong>in</strong>ed<br />
but present throughout southwest <strong>Arkansas</strong>, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> Felsenthal region. Schambach<br />
(1970:388) suggests that Williams projectile po<strong>in</strong>ts may be largely conf<strong>in</strong>ed to upland<br />
areas <strong>in</strong> this part <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast. Williams-like Big Creek po<strong>in</strong>ts (Morse 1970), however, are<br />
markers <strong>of</strong> <strong>the</strong> Frierson phase <strong>in</strong> nor<strong>the</strong>ast <strong>Arkansas</strong> which can probably be assigned to <strong>the</strong><br />
3000-1500 B.C. <strong>in</strong>terval (Morse 1975:192).<br />
Schambach (<strong>in</strong> Davis 1981:57-59) def<strong>in</strong>es <strong>the</strong> archeological Big Creek culture which<br />
subsumes <strong>the</strong> previously def<strong>in</strong>ed Dorcheat phase (Schambach 1970:389-390) <strong>of</strong> <strong>the</strong> Middle<br />
Ouachita region. Markers <strong>of</strong> <strong>the</strong> Big Creek culture are Bulverde (Suhm and Jelks 1962:404)<br />
and Evans (Ford and Webb 1956:64-65) projectile po<strong>in</strong>ts. Schambach (1970:390) attributes a<br />
series <strong>of</strong> polished red slate items to <strong>the</strong> Dorcheat phase component at Cooper. The type site<br />
<strong>of</strong> <strong>the</strong> Big Creek culture is <strong>the</strong> Big Creek site near Rison, <strong>Arkansas</strong> on <strong>the</strong> nor<strong>the</strong>rn fr<strong>in</strong>ge<br />
<strong>of</strong> <strong>the</strong> Felsenthal region. A s<strong>in</strong>gle amateur collection from this site conta<strong>in</strong>s over 500 Evans<br />
po<strong>in</strong>ts. Rol<strong>in</strong>gson and Schambach (1981:<strong>17</strong>9) note that <strong>the</strong> Big Creek culture is well represented<br />
<strong>in</strong> <strong>the</strong> Felsenthal region.
Poverty Po<strong>in</strong>t Period, 1200-500 B.C.<br />
Prehistory 63<br />
The <strong>in</strong>terval 1200-500 B.C. brackets <strong>the</strong> major portion <strong>of</strong> <strong>the</strong> development <strong>of</strong> <strong>the</strong><br />
Poverty Po<strong>in</strong>t culture (Webb 1977:Table 1). This archeological culture, recognized at sites<br />
throughout <strong>the</strong> Lower Mississippi Valley below <strong>the</strong> latitude <strong>of</strong> Memphis plus adjo<strong>in</strong><strong>in</strong>g<br />
upland areas and Gulf Coast areas, is characterized by diagnostic forms <strong>of</strong> baked clay balls<br />
or “Poverty Po<strong>in</strong>t objects,” elaborate lapidary items, a microlith <strong>in</strong>dustry, far-flung <strong>in</strong>terregional<br />
exchange <strong>in</strong> lithic raw materials, and—at <strong>the</strong> Poverty Po<strong>in</strong>t type site and some o<strong>the</strong>r<br />
sites—construction <strong>of</strong> mounds and earthworks (Webb 1968, 1977). This archeological culture<br />
is generally conceded to represent <strong>the</strong> most complex cultural development <strong>in</strong> North America<br />
on this time level. Its economic base is a topic <strong>of</strong> considerable <strong>in</strong>terest and debate. The<br />
few data available at this time suggest that a cont<strong>in</strong>uation <strong>of</strong> <strong>the</strong> Archaic pattern <strong>of</strong> harvest<strong>in</strong>g<br />
nondomesticated faunal and floral resources formed <strong>the</strong> basis <strong>of</strong> Poverty Po<strong>in</strong>t Culture<br />
subsistence (Webb 1977:3-4; Byrd and Neuman 1978).<br />
Several sites <strong>in</strong> and near <strong>the</strong> Felsenthal region have yielded such Poverty Po<strong>in</strong>t culture<br />
diagnostic artifacts as baked clay balls <strong>in</strong> several forms, hematite and magnetite plummets,<br />
steatite vessel fragments, microliths, and greenstone celts. These sites <strong>in</strong>clude Woodward<br />
Lake (3OU118), Calion (3UN51), Green Island (3BR1), Str<strong>in</strong>gfellow (3CA15, Hoover Levee<br />
(3UN90), Bangs Slough (3CA3), and Coon Island (3BR10) <strong>in</strong> <strong>Arkansas</strong> and Monroe and<br />
Pargoud (16OU1) <strong>in</strong> Louisiana (Webb 1977:Figure 4; Schambach <strong>in</strong> Davis 1981). Schambach<br />
fur<strong>the</strong>r notes that <strong>the</strong>se diagnostic artifacts occur <strong>in</strong> sufficient number and variety that <strong>the</strong>se<br />
components may readily be assigned to Poverty Po<strong>in</strong>t culture as well as to <strong>the</strong> Poverty Po<strong>in</strong>t<br />
period. This set <strong>of</strong> Felsenthal components has been assigned to <strong>the</strong> Calion phase by Rol<strong>in</strong>gson<br />
and Schambach (1981:Table 19).<br />
With<strong>in</strong> <strong>the</strong> Felsenthal region Schambach (1979:27) notes that Poverty Po<strong>in</strong>t culture sites<br />
recorded to date tend to cluster <strong>in</strong> <strong>the</strong> Calion-Champagnolle-Moro Creek locality with only<br />
scattered occurrences <strong>of</strong> diagnostic artifacts <strong>in</strong> <strong>the</strong> Felsenthal National Wildlife Refuge area<br />
to <strong>the</strong> south; this record persists <strong>in</strong> spite <strong>of</strong> extensive survey work <strong>in</strong> <strong>the</strong> latter area (but see<br />
Chapters 5 and 6 <strong>of</strong> this report). Many <strong>of</strong> <strong>the</strong> sites <strong>in</strong> <strong>the</strong> Calion locality have yielded large<br />
quantities <strong>of</strong> unf<strong>in</strong>ished magnetite plummets, worked magnetite debris and small unaltered<br />
pieces <strong>of</strong> magnetite. Schambach (1979:27-28) proposes that a yet-undiscovered natural occurrence<br />
<strong>of</strong> magnetite <strong>in</strong> gravel form <strong>in</strong> <strong>the</strong> uplands <strong>of</strong> <strong>the</strong> Champagnolle and Moro Creek<br />
watersheds may have been <strong>the</strong> resource that attracted Poverty Po<strong>in</strong>t culture people to <strong>the</strong><br />
region.<br />
Large contract<strong>in</strong>g stemmed po<strong>in</strong>ts correspond<strong>in</strong>g to Schambach’s (1970:188-191) Gary,<br />
variety Gary, may be <strong>the</strong> predom<strong>in</strong>ant po<strong>in</strong>t form associated with Poverty Po<strong>in</strong>t period components<br />
<strong>in</strong> <strong>the</strong> Felsenthal region. Webb (1977:37-38) notes that this form predom<strong>in</strong>ates at <strong>the</strong>
64 House<br />
Poverty Po<strong>in</strong>t site. In <strong>the</strong> <strong>Arkansas</strong> River Lowland to <strong>the</strong> east, use <strong>of</strong> <strong>Arkansas</strong> novaculite as<br />
a raw material for flaked stone tools appears to peak at this time (House 1980; cf. Giard<strong>in</strong>o<br />
1979).<br />
Dur<strong>in</strong>g <strong>the</strong> course <strong>of</strong> <strong>the</strong> present <strong>in</strong>vestigation, test<strong>in</strong>g at Marie Sal<strong>in</strong>e (3AS329) revealed<br />
an apparent Poverty Po<strong>in</strong>t culture component at a depth <strong>of</strong> about 1.1 m below surface.<br />
Recovered specimens <strong>in</strong>cluded steatite sherds, muscovite schist, novaculite flakes,<br />
and fire-cracked rock (Chapter 6). The survey results also <strong>in</strong>dicated a decided emphasis on<br />
novaculite versus chert on this time level <strong>in</strong> <strong>the</strong> project area (see Chapter 8).<br />
Tchula Period, 500 B.C.-A.D. 1<br />
Manufacture <strong>of</strong> ceramics became widespread <strong>in</strong> North America between 500 B.C.<br />
and A.D. 1. In <strong>the</strong> Lower Mississippi Valley this <strong>in</strong>terval, known as <strong>the</strong> Tchula period, is<br />
characterized by <strong>the</strong> development <strong>of</strong> <strong>the</strong> Tchefuncte archeological culture <strong>in</strong> <strong>the</strong> south and<br />
<strong>the</strong> Lake Cormorant and possibly o<strong>the</strong>r cultures <strong>in</strong> <strong>the</strong> north (Phillips 1970:16). Data from<br />
throughout eastern North America establishes <strong>the</strong> early cultivation <strong>of</strong> certa<strong>in</strong> tropical and<br />
native North American cultigens at this time. Of particular <strong>in</strong>terest is <strong>the</strong> recovery <strong>of</strong> rema<strong>in</strong>s<br />
<strong>of</strong> squash and bottle gourd <strong>in</strong> Tchefuncte levels at Morton Shell Mound <strong>in</strong> Coastal<br />
Louisiana (Byrd and Neuman 1978:11).<br />
Attributes <strong>of</strong> <strong>the</strong> Tchefuncte Culture (Ford and Quimby 1945) <strong>in</strong>clude grog-tempered<br />
or “untempered” pottery (types Tchefuncte Pla<strong>in</strong>, Tchefuncte Stamped, Tchefuncte Incised,<br />
Lake Borgne Incised), tetrapodal bases, and possibly construction <strong>of</strong> conical burial mounds.<br />
Sand-tempered Alexander series ceramics also occur <strong>in</strong> some Tchefuncte culture components<br />
(Gibson 1966; Phillips 1970:881). The Lake Cormorant culture <strong>in</strong> <strong>the</strong> nor<strong>the</strong>rn Lower<br />
Mississippi Valley is equivalent to Phillips, Ford and Griff<strong>in</strong>’s (1951:432-433) Nor<strong>the</strong>rn Tchula<br />
period complex which exhibits strong formal similarities to Early Woodland complexes<br />
<strong>in</strong> <strong>the</strong> Midwest. The nature <strong>of</strong> Tchula period occupation <strong>in</strong> <strong>the</strong> nor<strong>the</strong>rn Lower Mississippi<br />
Valley west <strong>of</strong> <strong>the</strong> Mississippi is yet poorly understood. The few Tchula period components<br />
recognized <strong>in</strong> <strong>the</strong> Bartholomew-Macon region appear to exhibit closer similarities to <strong>the</strong><br />
Tchefuncte culture to <strong>the</strong> south than to Lake Cormorant culture to <strong>the</strong> north (cf. Rol<strong>in</strong>gson<br />
1981; Jeter 1981).<br />
One <strong>of</strong> <strong>the</strong> surprises <strong>of</strong> <strong>the</strong> recent survey and test<strong>in</strong>g work <strong>in</strong> <strong>the</strong> Felsenthal National<br />
Wildlife Refuge has been <strong>the</strong> recognition <strong>of</strong> an apparent Tchefuncte culture component at<br />
Coon Island (3BR10). Artifacts <strong>in</strong>clude dist<strong>in</strong>ctive ceramic types Tchefuncte Incised, Lake<br />
Borgne Incised, Tchefuncte Stamped, and Tchefuncte Pla<strong>in</strong>, and characteristic tetrapodal<br />
bases. Four o<strong>the</strong>r sites 3UN9/52, 3UN63, 3UN75, and 3UN86 have also yielded small numbers<br />
<strong>of</strong> Tchefuncte sherds (Rol<strong>in</strong>gson 1972:3; Schambach 1979:29). These components have<br />
been grouped by Schambach (1979) <strong>in</strong>to <strong>the</strong> Coon Island phase.
Prehistory 65<br />
No data perta<strong>in</strong><strong>in</strong>g to Tchula period occupation along <strong>the</strong> Ouachita River <strong>in</strong> <strong>the</strong> Monroe,<br />
Louisiana locality have been published. Moore (1909:Figure 4) illustrates a basal fragment<br />
<strong>of</strong> an apparent Tchefuncte vessel from Booth Land<strong>in</strong>g on <strong>the</strong> lower reaches <strong>of</strong> <strong>the</strong><br />
Ouachita <strong>in</strong> Catahoula Parish.<br />
Dur<strong>in</strong>g <strong>the</strong> course <strong>of</strong> <strong>the</strong> present survey, Tchula period components were identified at<br />
Marie Sal<strong>in</strong>e (3AS329) and False Indigo (3AS285) and m<strong>in</strong>or components were identified at<br />
three additional sites. Ceramic variability among <strong>the</strong> components suggests that <strong>the</strong> Coon<br />
Island phase <strong>in</strong> <strong>the</strong> Felsenthal region may eventually be divided <strong>in</strong>to earlier (500-250 B.C.)<br />
and later (250 B.C.-A.D. 1) subphases (see Chapter 8).<br />
Marksville Period A.D. 1-300<br />
The concept <strong>of</strong> Marksville culture and Marksville period have been important <strong>in</strong><br />
Lower Mississippi Valley archeology s<strong>in</strong>ce Setzler (1933, cited <strong>in</strong> Toth 1974:21) recognized<br />
<strong>the</strong> strong connection to Midwestern Hopewell <strong>in</strong> ceramic vessels recovered from <strong>the</strong><br />
Marksville site <strong>in</strong> Avoyelles Parish, Louisiana. Marksville culture attributes <strong>in</strong>clude dist<strong>in</strong>ctive<br />
<strong>in</strong>cised and stamped ceramics, mound and earthwork construction, and, <strong>in</strong> some cases,<br />
mortuary ceremonialism <strong>in</strong>volv<strong>in</strong>g <strong>in</strong>terment <strong>of</strong> large numbers <strong>of</strong> <strong>in</strong>dividuals on ear<strong>the</strong>n<br />
platforms with<strong>in</strong> mounds (Fowke 1928:420-421, cited <strong>in</strong> Toth 1974:18; Ford and Willey 1940).<br />
A chronological subdivision <strong>of</strong> <strong>the</strong> Marksville period <strong>in</strong>to early or “classic” and late subperiods<br />
has been accomplished <strong>in</strong> some regions (Greengo 1964; Phillips 1970:10-11; Toth<br />
1974:101-131). Data from this <strong>in</strong>terval <strong>in</strong> <strong>the</strong> Midwest <strong>in</strong>dicate an economy based on <strong>in</strong>tensive<br />
harvest<strong>in</strong>g <strong>of</strong> nondomesticated resources, particularly fish, nuts, and starchy seeds <strong>of</strong><br />
annual weeds, plus cultivation <strong>of</strong> native North American cultigens and limited cultivation <strong>of</strong><br />
tropical cultigens (Asch et al. 1979). In <strong>the</strong> Lower Mississippi Valley, however, little progress<br />
has been achieved toward reconstruction <strong>of</strong> subsistence, settlement systems, or population<br />
biology dur<strong>in</strong>g <strong>the</strong> Marksville period.<br />
In a 1939 article, Dick<strong>in</strong>son and Lemley recognized ceramics <strong>of</strong> <strong>the</strong> “Marksville complex”<br />
(cf. Ford 1936) at <strong>the</strong> Kirkham site (3CL29) <strong>in</strong> <strong>the</strong> Middle Ouachita region <strong>of</strong> <strong>Arkansas</strong>.<br />
Schambach (1970:396-399) notes <strong>the</strong> occurrence <strong>of</strong> Marksville types <strong>in</strong> <strong>the</strong> Lost Bayou Phase<br />
<strong>of</strong> this region. H<strong>of</strong>fman (1970:151-153) reported additional data on Marksville period occupation<br />
<strong>in</strong> southwest <strong>Arkansas</strong> def<strong>in</strong><strong>in</strong>g <strong>the</strong> Bellvue phase <strong>of</strong> Marksville culture. Schambach<br />
(<strong>in</strong> Davis 1981:88-89), subsequently has taken issue with H<strong>of</strong>fman’s identification <strong>of</strong><br />
Marksville culture, emphasiz<strong>in</strong>g <strong>the</strong> environmental contrasts between <strong>the</strong> Lower Mississippi<br />
Valley and Trans-Mississippi South and <strong>the</strong> fact that, <strong>in</strong> most regions <strong>in</strong> <strong>the</strong> latter area,<br />
Marksville attributes occur <strong>in</strong> only very low frequencies <strong>in</strong> assemblages that reflect Fourche<br />
Mal<strong>in</strong>e culture cont<strong>in</strong>uities. Accord<strong>in</strong>g to Schambach, this lack <strong>of</strong> comparability on <strong>the</strong> total<br />
assemblage level implies major systemic differences <strong>in</strong> culture and adaptation.
66 House<br />
The question <strong>of</strong> <strong>the</strong> presence <strong>of</strong> Marksville culture <strong>in</strong> <strong>the</strong> Felsenthal region, however,<br />
rema<strong>in</strong>s open. Marksville sherds have been found at sites throughout <strong>the</strong> Felsenthal region<br />
but no pure Marksville period components have been isolated (Schambach 1979:29). It is<br />
<strong>in</strong>terest<strong>in</strong>g to note that both ceramic and radiometric dat<strong>in</strong>g suggest placement for this<br />
occupation on a late Marksville period time level. Stratum 6 at Paw Paw (3OU22) yielded<br />
sherds tentatively classified as Marksville Incised, var. Yokena, and Charupa Punctated (Phillips<br />
1970), and uncorrected radiocarbon dates <strong>of</strong> A.D. 460 ± 40 (TX 1547) and A.D. 660 ± 60<br />
(TX-1552) (Weber Excavation and analysis notes, <strong>Arkansas</strong> Archeological Survey 1973). The<br />
Shallow Lake (3UN9/52) excavations likewise yielded sherds resembl<strong>in</strong>g Marksville Incised,<br />
var. Yokena and Goose Lake (Rol<strong>in</strong>gson and Schambach 1981:181).<br />
Baytown Period, A.D. 300-700<br />
Phillips (1970:18) def<strong>in</strong>es <strong>the</strong> Baytown culture period as <strong>the</strong> <strong>in</strong>terval between <strong>the</strong> end<br />
<strong>of</strong> Hopewellian and emergence <strong>of</strong> Coles Creek culture. In <strong>the</strong> Lower Mississippi Valley, this<br />
<strong>in</strong>terval is occupied by <strong>the</strong> archeological Baytown culture, a concept which—along with<br />
<strong>the</strong> Baytown period—is def<strong>in</strong>ed somewhat negatively and has proven difficult to characterize.<br />
Baytown period phases <strong>in</strong> portions <strong>of</strong> <strong>the</strong> Lower Mississippi Valley near <strong>the</strong> Felsenthal<br />
region (e.g., Dry Bayou [Rol<strong>in</strong>gson 1973], Marsden [Phillips 1970:908] and Troyville [Phillips<br />
1970:908-910; cf. Walker 1936; Ford 1936, 1951]) are characterized by a predom<strong>in</strong>ance <strong>of</strong> Baytown<br />
Pla<strong>in</strong> ceramics with consistent m<strong>in</strong>orities <strong>of</strong> Mulberry Creek Cordmarked and pa<strong>in</strong>ted,<br />
<strong>in</strong>cised, and brushed ceramic types. Belmont and Williams (1981) suggest that horizon styles<br />
<strong>in</strong> grog-tempered pa<strong>in</strong>ted ceramics may provide improved chronological controls with<strong>in</strong><br />
<strong>the</strong> Baytown period. Absolute dates for Baytown culture phases and comparable units <strong>in</strong><br />
<strong>the</strong> Lower Valley cluster <strong>in</strong> <strong>the</strong> A.D. 600s (Phillips 1970:Table 18; Morse et al. 1980). The few<br />
available subsistence data for <strong>the</strong> Baytown period <strong>in</strong> <strong>the</strong> Lower Mississippi Valley <strong>in</strong>dicate<br />
cont<strong>in</strong>ued reliance on nondomesticated food resources with perhaps supplemental cultivation<br />
<strong>of</strong> native North American and tropical cultigens (Byrd and Neuman 1978).<br />
Contemporary components <strong>in</strong> <strong>the</strong> Middle Ouachita region are grouped <strong>in</strong>to <strong>the</strong> Oak<br />
Grove phase <strong>of</strong> Fourche Mal<strong>in</strong>e culture (Schambach 1970:399-400). Like Baytown assemblages,<br />
Fourche Mal<strong>in</strong>e culture ceramic assemblages at this time level are dom<strong>in</strong>ated by grogtempered<br />
pla<strong>in</strong>ware (Williams Pla<strong>in</strong>).<br />
The extent <strong>of</strong> Baytown period occupation <strong>in</strong> <strong>the</strong> Felsenthal region has proved difficult<br />
to determ<strong>in</strong>e—for <strong>the</strong> same reason that Baytown culture has proved a difficult concept to<br />
apply <strong>in</strong> <strong>the</strong> Lower Mississippi Valley. Schambach (<strong>in</strong> Davis 1981:91) notes that <strong>the</strong>se components<br />
with pla<strong>in</strong>ware <strong>in</strong> <strong>the</strong> Felsenthal region can present a problem <strong>of</strong> dist<strong>in</strong>guish<strong>in</strong>g<br />
between Baytown and Fourche Mal<strong>in</strong>e cultures. He outl<strong>in</strong>es two dist<strong>in</strong>ctions between <strong>the</strong><br />
Baytown and Fourche Mal<strong>in</strong>e archeological cultures which should prove useful: (1) Fourche<br />
Mal<strong>in</strong>e ceramic assemblages are dom<strong>in</strong>ated by flat-based jar forms, whereas Baytown
Prehistory 67<br />
assemblages are characterized by high proportions <strong>of</strong> round-based bowls, and (2) Gary<br />
po<strong>in</strong>ts (small varieties) are very abundant <strong>in</strong> Fourche Mal<strong>in</strong>e assemblages but <strong>in</strong>frequent <strong>in</strong><br />
Baytown assemblages. In this report <strong>the</strong> concept <strong>of</strong> Baytown culture has proved most useful<br />
(Chapters 5, 6, 8).<br />
Coles Creek Period, A.D. 700-1100<br />
As Jeter (1981) observes, <strong>the</strong> Coles Creek period cont<strong>in</strong>ues to be a “poorly def<strong>in</strong>ed concept<br />
applied to a number <strong>of</strong> poorly dated sites” (cf. Greengo 1964:14-15; Phillips 1970: 958;<br />
Schambach <strong>in</strong> Davis 1981:92). The Coles Creek period is characterized by Phillips (1970:20)<br />
as <strong>the</strong> <strong>in</strong>terval <strong>of</strong> Coles Creek culture <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Mississippi Valley.<br />
The Coles Creek archeological culture, first recognized as a ceramic complex by Ford<br />
(1935, 1936) has been widely used as a culture-level concept but is currently undergo<strong>in</strong>g a<br />
somewhat pa<strong>in</strong>ful process <strong>of</strong> redef<strong>in</strong>ition (e.g., Rol<strong>in</strong>gson 1979; Belmont 1979) and is only<br />
now beg<strong>in</strong>n<strong>in</strong>g to be dated adequately (Phillips 1970:Table 13; Rol<strong>in</strong>gson <strong>in</strong> Davis 1981:21;<br />
Schambach <strong>in</strong> Davis 1981:76; House and Jeter 1981).<br />
Evidence <strong>of</strong> Coles Creek period occupation is practically ubiquitous throughout <strong>the</strong><br />
Lower Valley and adjacent south <strong>Arkansas</strong>, possibly <strong>in</strong>dicative <strong>of</strong> <strong>in</strong>creased populations<br />
over earlier periods. In <strong>the</strong> Lower Mississippi Valley, this <strong>in</strong>terval is known as <strong>the</strong> Early<br />
Mississippi period (e.g., Morse 1980a) and is characterized not only by <strong>the</strong> appearance <strong>of</strong><br />
shell-tempered ceramics but also by <strong>the</strong> earliest substantial occurrence <strong>of</strong> maize <strong>in</strong> archeological<br />
sites (Morse 1980b). In <strong>the</strong> sou<strong>the</strong>rn Lower Mississippi Valley, ceramics cont<strong>in</strong>ue<br />
to be overwhelm<strong>in</strong>gly grog-tempered with characteristic types <strong>of</strong> <strong>in</strong>cised and punctated<br />
decoration on rims. Corner-notched and stemmed arrow po<strong>in</strong>ts are associated with Coles<br />
Creek period components <strong>in</strong> many regions (Rol<strong>in</strong>gson 1979). Although maize agriculture is<br />
thought to have been important <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Valley by this time (Haag 1971:26),<br />
no direct evidence has been reported. The few available data <strong>in</strong>dicate cont<strong>in</strong>u<strong>in</strong>g importance<br />
<strong>of</strong> nondomesticated plant foods (Byrd and Neuman 1978). Quite as significant, however, is<br />
evidence <strong>of</strong> construction <strong>of</strong> platform mounds at a few sites <strong>in</strong> each region (Belmont 1967;<br />
Rol<strong>in</strong>gson 1979; House and Jeter 1981), imply<strong>in</strong>g <strong>the</strong> development <strong>of</strong> more hierarchically<br />
organized settlement systems than characterize <strong>the</strong> Baytown period.<br />
In southwest <strong>Arkansas</strong>, Coles Creek ceramics were first recognized by Dick<strong>in</strong>son<br />
and Lemley (1939) at <strong>the</strong> Kirkham Place (3CL29) <strong>in</strong> <strong>the</strong> Middle Ouachita region. Coles Creek<br />
ceramics were subsequently recognized by Wood (1963) at Crenshaw (3MI6), by H<strong>of</strong>fman<br />
(1970b) <strong>in</strong> <strong>the</strong> Millwood Reservoir bas<strong>in</strong>, and by Schambach (1970:401-406) <strong>in</strong> <strong>the</strong> Dutchman’s<br />
Garden phase <strong>in</strong> <strong>the</strong> Middle Ouachita region. H<strong>of</strong>fman’s (1970a:153-155, 157-158)<br />
assignment <strong>of</strong> Coles Creek period components <strong>in</strong> <strong>the</strong> Little River region and at Crenshaw<br />
(3MI6) to Coles Creek culture has been questioned by Schambach (<strong>in</strong> Davis 1981:92-93) who<br />
cites major differences on <strong>the</strong> total assemblage level between Coles Creek period compo-
68 House<br />
nents <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Mississippi Valley and those <strong>in</strong> <strong>the</strong> Trans-Mississippi South,<br />
aga<strong>in</strong> with <strong>the</strong> implication <strong>of</strong> major systemic differences <strong>in</strong> culture and adaptation.<br />
Ongo<strong>in</strong>g <strong>in</strong>vestigation has revealed abundant evidence <strong>of</strong> Coles Creek period occupation<br />
<strong>in</strong> <strong>the</strong> Felsenthal Refuge area (Rol<strong>in</strong>gson 1972; Lischka 1973) and elsewhere <strong>in</strong> <strong>the</strong><br />
Felsenthal region (White 1970:14; Jeter 1981). Schambach (<strong>in</strong> Davis 1981:95) cites <strong>the</strong> near<br />
ubiquity <strong>of</strong> Coles Creek Incised ceramics <strong>in</strong> <strong>the</strong> region and suggests heavy occupation dur<strong>in</strong>g<br />
<strong>the</strong> Coles Creek period. Schambach (1979:30) also po<strong>in</strong>ts out that Coles Creek period<br />
middens are present at all <strong>of</strong> <strong>the</strong> major mound groups <strong>in</strong> <strong>the</strong> region and that sherd collections<br />
from mounds damaged by vandals and o<strong>the</strong>r destructive agents <strong>in</strong>dicate that some<br />
mounds <strong>in</strong> each group were built or begun dur<strong>in</strong>g <strong>the</strong> Coles Creek period.<br />
The taxonomic status <strong>of</strong> Coles Creek period components <strong>in</strong> <strong>the</strong> Felsenthal region rema<strong>in</strong>s<br />
to be addressed by isolation and description <strong>of</strong> total assemblages. Prelim<strong>in</strong>ary <strong>in</strong>dications<br />
(Schambach <strong>in</strong> Davis 1981:92-94) are that <strong>the</strong> ceramics exhibit greater formal similarity<br />
to <strong>the</strong> Coles Creek culture <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Mississippi Valley than to contemporary<br />
components <strong>of</strong> ei<strong>the</strong>r <strong>the</strong> Plum Bayou culture <strong>of</strong> east and central <strong>Arkansas</strong> (Rol<strong>in</strong>gson 1979)<br />
or to preCaddoan culture at Crenshaw and o<strong>the</strong>r sites <strong>in</strong> southwest <strong>Arkansas</strong>. The precise<br />
temporal placement <strong>of</strong> Coles Creek period components <strong>in</strong> <strong>the</strong> Felsenthal region is yet poorly<br />
known. Schambach (1979:30) states that <strong>the</strong> full temporal range <strong>of</strong> Coles Creek period occupation,<br />
from early to late, appears to be present. The Coles Creek period component at<br />
Shallow Lake (Rol<strong>in</strong>gson and Schambach 1981:182-189) appears to <strong>in</strong>clude burials beneath<br />
Mound C. Ceramics from <strong>the</strong> overly<strong>in</strong>g midden appear to represent Early Coles Creek period<br />
occupation.<br />
Mississippi Period, A.D.1100-<strong>17</strong>00<br />
The Mississippi period <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Valley is generally divided <strong>in</strong>to Early,<br />
ca A.D. 1100-1400, and Late, ca A.D. 1400-<strong>17</strong>00, subperiods. The two subperiods are characterized<br />
by <strong>the</strong> successive dom<strong>in</strong>ation <strong>of</strong> <strong>the</strong> Plaquem<strong>in</strong>e and Mississippian archeological<br />
cultures, respectively (Phillips 1970:20). S<strong>in</strong>ce <strong>the</strong> divid<strong>in</strong>g l<strong>in</strong>e between <strong>the</strong> nor<strong>the</strong>rn and<br />
sou<strong>the</strong>rn portions <strong>of</strong> <strong>the</strong> Lower Valley is usually placed at about 33° north latitude (i.e., <strong>the</strong><br />
<strong>Arkansas</strong>-Louisiana state l<strong>in</strong>e) <strong>the</strong> Felsenthal region dur<strong>in</strong>g <strong>the</strong> Mississippi period may be<br />
seen to have been on <strong>the</strong> fluctuat<strong>in</strong>g frontiers <strong>of</strong> three archeological cultures, Mississippian,<br />
Plaquem<strong>in</strong>e, and Caddoan (cf. Jeter 1981).<br />
Like <strong>the</strong> preced<strong>in</strong>g Coles Creek culture, <strong>the</strong> Plaquem<strong>in</strong>e culture, def<strong>in</strong>ed by Quimby<br />
(1951), persists <strong>in</strong> be<strong>in</strong>g a ra<strong>the</strong>r loose group<strong>in</strong>g <strong>of</strong> attributes. The culture is considered as<br />
“Mississippianized” Coles Creek by Bra<strong>in</strong> (1978:345), who cites <strong>the</strong> development <strong>in</strong> some<br />
regions <strong>of</strong> impressive site plans characterized by multiple mounds dom<strong>in</strong>ated by a s<strong>in</strong>gle<br />
mound as much as 20 m <strong>in</strong> height with dist<strong>in</strong>ct plaza areas (cf. Phillips 1970:560). In contrast<br />
to <strong>the</strong> ceramics <strong>of</strong> <strong>the</strong> nor<strong>the</strong>rn Lower Valley dur<strong>in</strong>g this <strong>in</strong>terval, those <strong>of</strong> <strong>the</strong> sou<strong>the</strong>rn<br />
Lower Valley Plaquem<strong>in</strong>e culture persist <strong>in</strong> be<strong>in</strong>g predom<strong>in</strong>antly grog-tempered (though
Prehistory 69<br />
perhaps with some technical improvements <strong>in</strong> paste over preced<strong>in</strong>g Baytown-Coles Creek<br />
ceramics). Brushed, <strong>in</strong>cised, and punctated decoration is frequently applied to body as well<br />
as rim portions <strong>of</strong> vessels. A new major vessel form, <strong>the</strong> bottle, appears as do additional vessel<br />
form modes <strong>in</strong> bowls and jars. Stemmed arrow po<strong>in</strong>ts, called Ashley po<strong>in</strong>ts by Rol<strong>in</strong>gson<br />
(1971), occur <strong>in</strong> Early Mississippi period components <strong>in</strong> <strong>the</strong> Bartholomew-Macon region.<br />
Intensive agriculture <strong>in</strong>volv<strong>in</strong>g both maize and beans as well as squash appears to<br />
have been widespread <strong>in</strong> eastern North America by this time (Griff<strong>in</strong> 1967:189) but no direct<br />
subsistence data from Plaquem<strong>in</strong>e culture components are yet reported. Absolute dates for<br />
<strong>the</strong> Plaquem<strong>in</strong>e culture are yet relatively few (Phillips 1970:Table 18; House and Jeter 1981)<br />
but progress has been made <strong>in</strong> some regions <strong>in</strong> chronologically subdivid<strong>in</strong>g <strong>the</strong> Plaquem<strong>in</strong>e<br />
culture on <strong>the</strong> basis <strong>of</strong> artifactual attributes (Hally 1967; Phillips 1970:558, 560; cf. Jeter 1981).<br />
Early Mississippian period Plaquem<strong>in</strong>e culture phases <strong>in</strong> <strong>the</strong> Lower Valley east <strong>of</strong><br />
<strong>the</strong> Felsenthal region <strong>in</strong>clude <strong>the</strong> Routh phase <strong>of</strong> <strong>the</strong> Upper Tensas region (Hally 1967) and<br />
<strong>the</strong> Bartholomew phase <strong>of</strong> <strong>the</strong> Bartholomew-Macon region (Rol<strong>in</strong>gson 1976a). To <strong>the</strong> west,<br />
<strong>the</strong> Caddoan culture Haley phase (Caddo II period) (Schambach <strong>in</strong> Davis 1980:107-108) <strong>in</strong><br />
<strong>the</strong> Great Bend region is roughly coeval with <strong>the</strong> Plaquem<strong>in</strong>e culture while <strong>in</strong> <strong>the</strong> Middle<br />
Ouachita region much <strong>of</strong> <strong>the</strong> Caddoan culture Mid-Ouachita phase (Hodges and Hodges<br />
1945) appears to date to this <strong>in</strong>terval (Early <strong>in</strong> Davis 1981:109-110).<br />
The first recent <strong>in</strong>vestigations <strong>in</strong> <strong>the</strong> Felsenthal region (Rol<strong>in</strong>gson 1972; Lischka 1973)<br />
identified Plaquem<strong>in</strong>e culture occupation at a number <strong>of</strong> sites. S<strong>in</strong>ce <strong>the</strong> region had previously<br />
been considered part <strong>of</strong> <strong>the</strong> Caddoan archeological culture area, <strong>the</strong> low frequency <strong>of</strong><br />
Caddoan attributes observed <strong>in</strong> <strong>the</strong> recovered ceramic assemblages was somewhat surpris<strong>in</strong>g.<br />
Analysis by Rol<strong>in</strong>gson and Schambach (1981) <strong>of</strong> material from Stacy’s 1975 excavation<br />
<strong>of</strong> a circular structure beneath Mound C at Shallow Lake (3UN9/52) have resulted <strong>in</strong> def<strong>in</strong>ition<br />
<strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> phase <strong>of</strong> <strong>the</strong> Early Mississippi period, culture unknown. Ceramics<br />
<strong>in</strong>clude Baytown Pla<strong>in</strong>, var. Shallow Lake, Harrison Bayou Incised, Kiam Incised, var. Redeye<br />
Lake and Pargoud Incised, var. Monroe. Both a radiocarbon date and an archeomagnetic<br />
date place this occupation <strong>in</strong> <strong>the</strong> A.D. 1250-1300 range. This date is supported by an archeomagnetic<br />
date <strong>of</strong> A.D. 1350 for Watts Field (3UN18) Mound D. Analysis <strong>of</strong> faunal and<br />
floral specimens from <strong>the</strong> Coles Creek period and Gran <strong>Marais</strong> phase midden at Shallow<br />
Lake Mound C <strong>in</strong>dicated exploitation <strong>of</strong> both terrestrial and aquatic habitats <strong>in</strong> <strong>the</strong> locality<br />
and probable multiseasonal occupation. Surpris<strong>in</strong>gly, <strong>in</strong> light <strong>of</strong> <strong>the</strong> date <strong>of</strong> <strong>the</strong> context, no<br />
rema<strong>in</strong>s <strong>of</strong> cultigens were recovered (Rol<strong>in</strong>gson 1981).<br />
Gran <strong>Marais</strong> phase components appear to be present at every one <strong>of</strong> <strong>the</strong> large mound<br />
groups <strong>in</strong> <strong>the</strong> Felsenthal region. Figure 14 presents <strong>the</strong> plan <strong>of</strong> one <strong>of</strong> <strong>the</strong>se mound groups,
70 House<br />
Figure 14. The Mississippi period mound and midden complex at Watts Field<br />
(3UN18) (after Lischka 1973 and <strong>Arkansas</strong> Archeological Survey notes on<br />
file).
Prehistory 71<br />
Watts Field (3UN18). These components, however, appear primarily to represent ceremonial<br />
and mortuary behavior ra<strong>the</strong>r than <strong>in</strong>tensive habitation, lead<strong>in</strong>g Schambach (1979:30)<br />
to propose that <strong>the</strong> Early Mississippi period settlement pattern <strong>in</strong> <strong>the</strong> Felsenthal region is<br />
characterized by sites <strong>of</strong> highly dispersed hamlets and farmsteads.<br />
Dur<strong>in</strong>g <strong>the</strong> course <strong>of</strong> <strong>the</strong> present <strong>in</strong>vestigation, Gran <strong>Marais</strong> phase components were<br />
identified at False Indigo (3AS285) and Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76). Application <strong>of</strong> Schambach’s<br />
analytical classification system (Rol<strong>in</strong>gson and Schambach 1981:101-<strong>17</strong>6) to <strong>the</strong> large<br />
ceramic sample from False Indigo Area A revealed <strong>the</strong> sample to be strik<strong>in</strong>gly similar to <strong>the</strong><br />
Gran <strong>Marais</strong> phase assemblage from Shallow Lake (3UN9/52) Mound C. The contrast<strong>in</strong>g<br />
higher percentage <strong>of</strong> shell temper<strong>in</strong>g <strong>in</strong> <strong>the</strong> ceramic sample from Jug Po<strong>in</strong>t Cut<strong>of</strong>f suggests<br />
placement <strong>of</strong> this component later <strong>in</strong> <strong>the</strong> span <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> phase (Chapter 8). In<br />
Chapter 9 <strong>of</strong> this report, Hemm<strong>in</strong>gs presents a model <strong>of</strong> human adaptation to <strong>the</strong> Ouachita<br />
River floodpla<strong>in</strong> by Gran <strong>Marais</strong> phase people, <strong>in</strong>corporat<strong>in</strong>g <strong>in</strong>formation on <strong>the</strong> structure<br />
<strong>of</strong> <strong>the</strong> environment and data from <strong>the</strong> present and previous <strong>in</strong>vestigations <strong>in</strong> <strong>the</strong> Felsenthal<br />
region.<br />
On <strong>the</strong> sou<strong>the</strong>rn fr<strong>in</strong>ges <strong>of</strong> <strong>the</strong> Felsenthal region, recent excavations <strong>in</strong> Area I <strong>of</strong> <strong>the</strong><br />
badly-damaged T. E. Salsbury site (16OU15) <strong>in</strong> Ouachita Parish, Louisiana (Price and Heartfield<br />
1977) yielded ceramics resembl<strong>in</strong>g those <strong>of</strong> <strong>the</strong> (subsequently-def<strong>in</strong>ed) Gran <strong>Marais</strong><br />
phase <strong>in</strong> association with aborig<strong>in</strong>al burials and a village midden. A small amount <strong>of</strong> Caddoan<br />
ceramics were also recovered <strong>in</strong> <strong>the</strong> cemetery and midden area. Subsequent radiocarbon<br />
dates (TX-2839 through TX-2846) <strong>in</strong>dicate this occupation to date between A.D. 1200<br />
and 1500 (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers 1980:5:7). An important Plaquem<strong>in</strong>e time-level<br />
component also appears to be present at <strong>the</strong> Pargoud Land<strong>in</strong>g site (16OU1) near Monroe,<br />
though no report on recent work at that site is yet available (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers<br />
1980:5:6).<br />
The Late Mississippi subperiod (A.D. 1400-<strong>17</strong>00) is characterized by <strong>the</strong> dom<strong>in</strong>ation <strong>of</strong><br />
Mississippian culture throughout much <strong>of</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Valley as well as <strong>in</strong> <strong>the</strong> north.<br />
In this sense, Mississippian culture may (with some simplification) be equated with shelltempered<br />
pottery (cf. Holmes 1903). Alternative def<strong>in</strong>itions <strong>of</strong> “Mississippian” as a type <strong>of</strong><br />
adaptation (Griff<strong>in</strong> 1967:189; Smith 1978:486), however, have also been proposed.<br />
Frequencies <strong>of</strong> shell-tempered pottery appear to <strong>in</strong>crease gradually through time <strong>in</strong><br />
some regions (Hally 1967) and more abruptly <strong>in</strong> o<strong>the</strong>rs (Phillips 1970:501-567). In much <strong>of</strong><br />
<strong>the</strong> Lower Valley east <strong>of</strong> <strong>the</strong> Felsenthal region, <strong>the</strong> late shell-tempered complexes equate<br />
with <strong>the</strong> “Tunican” complex identified by Ford (1936:98-106) <strong>in</strong> <strong>the</strong> Lower Yazoo bas<strong>in</strong> (cf.<br />
Lemley and Dick<strong>in</strong>son 1936). Important decorated types on this horizon <strong>in</strong>clude W<strong>in</strong>terville<br />
Incised, Leland Incised, and Park<strong>in</strong> Punctated. In many Lower Valley sites on this time level,<br />
Caddoan ceramics—or at least polished and engraved types whose distribution extends <strong>in</strong>to<br />
<strong>the</strong> Caddoan area—are part <strong>of</strong> predom<strong>in</strong>antly shell-tempered “Mississippian” assemblages.
72 House<br />
Willow Leaf or Nodena and triangular arrowpo<strong>in</strong>ts tend to replace stemmed forms <strong>in</strong> many<br />
regions by this time (Phillips et al. 1951:450).<br />
The Late Mississippi period lifeway <strong>in</strong> <strong>the</strong> Lower Valley appears to have focused on<br />
<strong>in</strong>tensive agriculture <strong>in</strong>volv<strong>in</strong>g maize, beans, and squash but considerable utilization <strong>of</strong> nondomesticated<br />
floral and faunal resources persisted (Byrd and Neuman 1978). Dramatic settlement<br />
pattern and demographic shifts are perceived <strong>in</strong> some regions after A.D. 1400 (Bra<strong>in</strong><br />
1978; Morse 1981; House 1981) but <strong>the</strong>ir nature is not yet understood. The Late Mississippi<br />
period <strong>in</strong> <strong>the</strong> Lower Valley extends through <strong>the</strong> 1541-1543 De Soto expedition datel<strong>in</strong>e to <strong>the</strong><br />
time <strong>of</strong> Early French contact, ca A.D. <strong>17</strong>00. This contact had drastic and wide-rang<strong>in</strong>g effects<br />
on aborig<strong>in</strong>al populations and cultural development <strong>in</strong> this portion <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast but<br />
documentation derived from <strong>the</strong>se episodes (Swanton 1911, 1942, 1946) provides ethnohistoric<br />
data pert<strong>in</strong>ent to many phenomena observed archeologically <strong>in</strong> prehistory. There are<br />
also many sites <strong>of</strong> aborig<strong>in</strong>al occupation, particularly <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Valley, at which<br />
contact is attested to by <strong>the</strong> occurrence <strong>of</strong> objects <strong>of</strong> European manufacture <strong>in</strong> metal and<br />
glass.<br />
Prehistoric Late Mississippi period phases <strong>in</strong> <strong>the</strong> Lower Mississippi Valley east <strong>of</strong> <strong>the</strong><br />
Felsenthal region <strong>in</strong>clude <strong>the</strong> Wilmot (Rol<strong>in</strong>gson 1973) and Bellaire (Phillips 1970:944; Jeter<br />
1981) phases <strong>in</strong> <strong>the</strong> Bartholomew-Macon region <strong>of</strong> <strong>Arkansas</strong> and <strong>the</strong> subsequent Fitzhugh<br />
and Transylvania phases <strong>of</strong> <strong>the</strong> Upper Tensas Bas<strong>in</strong> <strong>in</strong> Louisiana (Rally 1967; cf. Phillips<br />
1970:945). In <strong>the</strong> latter region, Hally emphasizes <strong>the</strong> essential cont<strong>in</strong>uity <strong>of</strong> Plaquem<strong>in</strong>e culture<br />
despite <strong>the</strong> <strong>in</strong>creas<strong>in</strong>g frequency <strong>of</strong> shell-tempered pottery. To <strong>the</strong> west, <strong>the</strong> Texarkana<br />
(Krieger 1946) and Belcher (Webb 1959) phases <strong>of</strong> <strong>the</strong> Caddo III period on <strong>the</strong> Red River are<br />
roughly contemporary units.<br />
Def<strong>in</strong>ed phases which <strong>in</strong>clude European trade goods and clearly date to <strong>the</strong> Contact<br />
Historic horizon <strong>in</strong>clude <strong>the</strong> Quapaw (Phillips 1970:443-444; H<strong>of</strong>fman 1977; House<br />
and McKelway 1980) and Tensas (Phillips 1970: 945) phases. On <strong>the</strong> sou<strong>the</strong>rn fr<strong>in</strong>ge <strong>of</strong> <strong>the</strong><br />
Felsenthal region, <strong>in</strong> Ouachita Parish, Louisiana, Moore (1909:27-80, 120-151) recovered<br />
abundant European trade goods <strong>in</strong>clud<strong>in</strong>g glass beads and discs and cones <strong>of</strong> sheet brass<br />
from aborig<strong>in</strong>al burials <strong>in</strong> association with elaborate pa<strong>in</strong>ted, polished, engraved, and <strong>in</strong>cised<br />
ceramics at Glendora and Keno. Only a s<strong>in</strong>gle Caddoan site <strong>in</strong> <strong>the</strong> Great Bend region,<br />
Rosenbrough Lake <strong>in</strong> Bowie County, Texas, has produced early European trade goods but<br />
contemporary occupations are seen at a number <strong>of</strong> o<strong>the</strong>r sites <strong>in</strong> <strong>the</strong> region (Schambach <strong>in</strong><br />
Davis 1981:124).<br />
On <strong>the</strong> basis <strong>of</strong> <strong>the</strong> f<strong>in</strong>d<strong>in</strong>gs <strong>of</strong> two amateurs’ 1965 salvage excavation <strong>of</strong> an aborig<strong>in</strong>al<br />
cemetery at Shallow Lake (3UN9/52), Schambach and Rol<strong>in</strong>gson (1981:193-198) have def<strong>in</strong>ed<br />
<strong>the</strong> Caney Bayou phase <strong>of</strong> <strong>the</strong> Late Mississippi period, culture unknown. The cemetery<br />
yielded <strong>in</strong>complete bundle burials accompanied by both shell-tempered and grog-tempered<br />
vessels <strong>in</strong> <strong>the</strong> types Baytown Pla<strong>in</strong>, var. Shallow Lake, Coleman Incised, Cowhide Stamped,
Prehistory 73<br />
Glassell Engraved, Park<strong>in</strong> Punctated, and W<strong>in</strong>terville Incised. Additional grave goods<br />
<strong>in</strong>clude discoidals and a cannel coal pendant. Three ceramic elbow pipes were found <strong>in</strong> <strong>the</strong><br />
cemetery area but not <strong>in</strong> direct grave association. No European trade goods were recovered.<br />
This occupation is placed <strong>in</strong> <strong>the</strong> A.D. 1500-<strong>17</strong>00 range.<br />
Schambach (1979:30) notes that o<strong>the</strong>r late prehistoric pottery vessels have reportedly<br />
been dug by collectors from shallow <strong>in</strong>trusive graves <strong>in</strong> <strong>the</strong> tops <strong>of</strong> mounds <strong>in</strong> <strong>the</strong> Felsenthal<br />
region. Schambach and Rol<strong>in</strong>gson (1981:201) propose that <strong>the</strong> apparent absence <strong>of</strong> large sites<br />
dat<strong>in</strong>g to this time level is due to a post-De Soto depopulation or to a highly dispersed settlement<br />
pattern. They suggest that <strong>the</strong> Caney Bayou phase population may have visited <strong>the</strong><br />
preexist<strong>in</strong>g mound centers not to build or ma<strong>in</strong>ta<strong>in</strong> mounds but to bury <strong>the</strong>ir dead <strong>in</strong> small<br />
cemeteries, some <strong>of</strong> which were located on top <strong>of</strong> mounds. The present survey has identified<br />
probable Caney Bayou phase components, possibly represent<strong>in</strong>g river<strong>in</strong>e extraction camps,<br />
at <strong>the</strong> Jug Po<strong>in</strong>t 2 (3AS307) site (see below, Chapter 8).<br />
Ceramics which can be roughly dated to <strong>the</strong> Late Mississippi period were recovered<br />
by Clarence B. Moore at a number <strong>of</strong> sites <strong>in</strong> <strong>the</strong> Felsenthal region. These <strong>in</strong>clude Boytt’s<br />
Field (3UN13), Keller (3CA13), and Kent (3OU6) (Moore 1909:82-89, 91-96, 96-100) and Wire<br />
Fence Land<strong>in</strong>g on <strong>the</strong> Sal<strong>in</strong>e (Moore 1913:91-92).<br />
Ceramic vessels recovered from an aborig<strong>in</strong>al cemetery at Gee’s Land<strong>in</strong>g (3DR<strong>17</strong>) on<br />
<strong>the</strong> Sal<strong>in</strong>e River <strong>in</strong> 1967-68 <strong>in</strong>clude shell-tempered and grog-tempered vessels with <strong>in</strong>cised,<br />
engraved, and ridge-p<strong>in</strong>ched decorations. Collectively, this series <strong>of</strong> vessels resembles<br />
those recovered by Moore at Glendora and Keno and series from contact horizon sites <strong>in</strong><br />
<strong>the</strong> Lower Mississippi Valley. No European trade goods, however, were recovered at Gee’s<br />
Land<strong>in</strong>g (3DR<strong>17</strong>) (White 1970). Ano<strong>the</strong>r component on <strong>the</strong> nor<strong>the</strong>rn fr<strong>in</strong>ges <strong>of</strong> <strong>the</strong> Felsenthal<br />
region <strong>in</strong> <strong>the</strong> Sal<strong>in</strong>e River watershed which dates to this time level was recognized at Dansbury<br />
Creek (3CV24) <strong>in</strong> Cleveland County, <strong>Arkansas</strong> (Reaves and McClurkan 1970).<br />
In <strong>the</strong> Felsenthal Navigation Pool area <strong>in</strong>vestigated by <strong>the</strong> present survey, an additional<br />
very Late Mississippi period complex, called “term<strong>in</strong>al Mississippi” <strong>in</strong> this report, has<br />
been identified at One Cypress Po<strong>in</strong>t (3AS286) on <strong>the</strong> Sal<strong>in</strong>e River. The lithic assemblage recovered<br />
at this site (see Chapter 6, below) <strong>in</strong>cludes Nodena or Willow Leaf po<strong>in</strong>ts and small<br />
hafted end scrapers, both attributes <strong>of</strong> <strong>the</strong> Quapaw phase on <strong>the</strong> Lower <strong>Arkansas</strong> (Ford<br />
1961:156-158; House and McKelway 1980; cf. Phillips 1970:945; Williams 1980).<br />
No ethnohistoric sources from <strong>the</strong> sixteenth and seventeenth centuries appear to refer<br />
directly to <strong>the</strong> Felsenthal region. Swanton (1946:55) suggests that <strong>the</strong> aborig<strong>in</strong>al town <strong>of</strong><br />
Utiangue, where <strong>the</strong> De Soto expedition w<strong>in</strong>tered <strong>in</strong> 1541-42, was on <strong>the</strong> Ouachita River<br />
near Camden or Calion but it has generally proved impossible to positively identify any<br />
regions west <strong>of</strong> <strong>the</strong> Mississippi mentioned <strong>in</strong> <strong>the</strong> De Soto expedition chronicles. In <strong>17</strong>00, <strong>the</strong>
74 House<br />
Washita, identified by Swanton (1946:204) as a Caddoan group, were encountered on <strong>the</strong><br />
lower Ouachita near present-day Columbia, Louisiana by French explorer Bienville. Bienville<br />
reported <strong>the</strong> Washita town at this location to consist <strong>of</strong> five cab<strong>in</strong>s with about 70<br />
men. Somewhat confus<strong>in</strong>g late sixteenth century documentary data reviewed by Swanton<br />
(1911:329-330) place part <strong>of</strong> <strong>the</strong> Tunican-speak<strong>in</strong>g Koroa on <strong>the</strong> west side <strong>of</strong> <strong>the</strong> Mississippi<br />
River, perhaps near lower Bayou Bartholomew or <strong>the</strong> Ouachita River (Swanton 1911:Plate<br />
I). The possible presence <strong>of</strong> Tunican-speakers near <strong>the</strong> Felsenthal region <strong>in</strong> <strong>the</strong> sixteenth and<br />
seventeenth centuries (cf. Rol<strong>in</strong>gson and Schambach 1981:106) is especially <strong>in</strong>terest<strong>in</strong>g <strong>in</strong><br />
light <strong>of</strong> <strong>the</strong> marked similarities between Felsenthal region ceramics on this time level and<br />
contemporary Late Mississippi period ceramics <strong>in</strong> <strong>the</strong> Yazoo Bas<strong>in</strong> east <strong>of</strong> <strong>the</strong> Mississippi<br />
River.<br />
discussion<br />
The Felsenthal archeological region is <strong>the</strong> same territory which Schambach <strong>in</strong> 1971<br />
described <strong>in</strong> <strong>the</strong>se terms:<br />
Few areas <strong>of</strong> <strong>Arkansas</strong> are as wild, as rich archeologically and as little studied<br />
as this portion <strong>of</strong> <strong>the</strong> Ouachita Valley. When Clarence B. Moore (1909) traveled<br />
up <strong>the</strong> river to <strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> Camden <strong>in</strong> 1908, he found a few ‘plantations’<br />
and some wood camps for <strong>the</strong> steamboat traffic carved out <strong>of</strong> <strong>the</strong> dense forest<br />
along <strong>the</strong> banks. S<strong>in</strong>ce <strong>the</strong>n most <strong>of</strong> <strong>the</strong>se clear<strong>in</strong>gs have reverted to timber and<br />
for <strong>the</strong> most part <strong>the</strong> Ouachita bottoms south <strong>of</strong> Camden are wilder than <strong>the</strong>y<br />
were <strong>in</strong> Moore’s time. (Schambach 1971)<br />
S<strong>in</strong>ce 1971 <strong>the</strong> Felsenthal region has undergone considerable archeological <strong>in</strong>vestigation,<br />
much <strong>of</strong> it <strong>in</strong> conjunction with <strong>the</strong> development <strong>of</strong> <strong>the</strong> Felsenthal National Wildlife<br />
Refuge and U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers navigation channel improvements on <strong>the</strong> Ouachita<br />
River. This <strong>in</strong>vestigation has revealed a long prehistoric and contact historic aborig<strong>in</strong>al<br />
sequence extend<strong>in</strong>g from Paleo-Indian through <strong>the</strong> era <strong>of</strong> earliest French exploration. One<br />
<strong>of</strong> <strong>the</strong> <strong>in</strong>trigu<strong>in</strong>g results <strong>of</strong> this <strong>in</strong>vestigation are <strong>in</strong>dications that, throughout much <strong>of</strong> this<br />
aborig<strong>in</strong>al sequence, components <strong>in</strong> <strong>the</strong> Felsenthal region exhibit much greater similarity to<br />
cultures <strong>in</strong> <strong>the</strong> Lower Mississippi Valley to <strong>the</strong> east than to cultures <strong>in</strong> <strong>the</strong> Caddoan area to<br />
<strong>the</strong> west. Indeed, <strong>the</strong>se Felsenthal region occupations appear to exhibit greater similarities<br />
to cultures <strong>of</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Mississippi Valley (e.g., Tchefuncte, “classic” Coles Creek,<br />
Plaquem<strong>in</strong>e) than to geographically close cultures <strong>of</strong> <strong>the</strong> nor<strong>the</strong>rn Lower Mississippi Valley.<br />
In spite <strong>of</strong> <strong>the</strong> impressive advances <strong>of</strong> <strong>the</strong> last decade, however, <strong>the</strong> archeological record<br />
<strong>of</strong> <strong>the</strong> Felsenthal region is only beg<strong>in</strong>n<strong>in</strong>g to be explored on <strong>the</strong> regional level. Because<br />
<strong>of</strong> present-day land use patterns, <strong>the</strong> archeological resource base <strong>in</strong> <strong>the</strong> region appears to be<br />
especially well-preserved compared with <strong>the</strong> resource base <strong>in</strong> <strong>the</strong> Lower Mississippi Valley.
Prehistory 75<br />
Investigation <strong>in</strong> <strong>the</strong> Ouachita River floodpla<strong>in</strong> particularly, reveals <strong>the</strong> presence <strong>of</strong> alluvially<br />
buried and stratified sites. Such contexts are <strong>of</strong> limited occurrence <strong>in</strong> <strong>the</strong> Lower Mississippi<br />
Valley proper. The present survey <strong>of</strong> <strong>the</strong> Felsenthal Navigation Pool project area has been<br />
especially productive <strong>of</strong> <strong>in</strong>formation on buried and stratified floodpla<strong>in</strong> sites and associated<br />
geomorphological processes (Chapter 7).<br />
Ongo<strong>in</strong>g prehistoric archeology <strong>in</strong> <strong>the</strong> Felsenthal region has been heavily <strong>in</strong>volved<br />
with taxonomy, on both <strong>the</strong> artifact and culture unit levels. By explicitly address<strong>in</strong>g measurement<br />
<strong>of</strong> prehistoric social <strong>in</strong>teraction and del<strong>in</strong>eation <strong>of</strong> prehistoric sociocultural<br />
boundaries (e.g., Schambach <strong>in</strong> Davis 1981:88-98), this taxonomic emphasis may be seen<br />
to transcend traditional Sou<strong>the</strong>astern preoccupations with chronology and culture history.<br />
Particularly, Schambach’s design-mode based classification <strong>of</strong> aborig<strong>in</strong>al ceramics from <strong>the</strong><br />
region (Rol<strong>in</strong>gson and Schambach 1981: 106-<strong>17</strong>6) holds <strong>the</strong> promise <strong>of</strong> allow<strong>in</strong>g quantification<br />
<strong>of</strong> formal similarity/dissimilarity between components, at least on <strong>the</strong> <strong>in</strong>traregional<br />
level.<br />
The development <strong>of</strong> <strong>in</strong>dependent chronological controls <strong>in</strong> <strong>the</strong> region has only begun<br />
<strong>in</strong> recent years but dated components <strong>in</strong> <strong>the</strong> region range from Middle Archaic through<br />
Mississippi <strong>in</strong> time and from <strong>the</strong> Camden locality to <strong>the</strong> Monroe locality <strong>in</strong> space. Contexts<br />
suitable for archeomagnetic dat<strong>in</strong>g are widespread <strong>in</strong> <strong>the</strong> region. Archeomagnetic dat<strong>in</strong>g<br />
<strong>of</strong> samples from Shallow Lake (3UN9/52) and Watts Field (3UN18) have already provided<br />
valuable corroboration <strong>of</strong> radiocarbon dates for <strong>the</strong> Gran <strong>Marais</strong> phase <strong>of</strong> <strong>the</strong> Early Mississippi<br />
period (Rol<strong>in</strong>gson and Schambach 1981:52-53). Dur<strong>in</strong>g <strong>the</strong> present <strong>in</strong>vestigation, attempts<br />
to radiocarbon-date samples recovered from low-ly<strong>in</strong>g floodpla<strong>in</strong> sites encountered<br />
unanticipated technical problems and yielded disappo<strong>in</strong>t<strong>in</strong>g results (see Chapter 7, below).<br />
Biocultural anthropology <strong>in</strong> <strong>the</strong> Felsenthal region can be said to have begun with<br />
Hrdlicka’s (1909) analysis <strong>of</strong> human skeletal series from Boytt’s Field (3UN13) <strong>in</strong> Union<br />
County, <strong>Arkansas</strong>. S<strong>in</strong>ce this time, aborig<strong>in</strong>al human rema<strong>in</strong>s have been unear<strong>the</strong>d at several<br />
sites <strong>in</strong> <strong>the</strong> region but only <strong>the</strong> probable Coles Creek period series from beneath Shallow<br />
Lake Mound C has been analyzed and reported (Rol<strong>in</strong>gson and Schambach 1981: 48, 51-52,<br />
89). Though a biological picture <strong>of</strong> past aborig<strong>in</strong>al populations <strong>in</strong> <strong>the</strong> Felsenthal region is yet<br />
to emerge, <strong>the</strong>se studies demonstrate that a great potential exists.<br />
The present-day environment <strong>in</strong> <strong>the</strong> Felsenthal archeological region, <strong>in</strong> both <strong>the</strong> uplands<br />
and lowlands, presents formidable challenges to <strong>in</strong>tensive and systematic site survey.<br />
The Sparta M<strong>in</strong>e (Hampton) project <strong>in</strong> <strong>the</strong> West Gulf Coastal Pla<strong>in</strong> uplands (Lafferty et al.<br />
1981) and <strong>the</strong> present study <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland (Chapter 5) both represent <strong>in</strong>novative<br />
attempts to meet this challenge and both have yielded encourag<strong>in</strong>g results.<br />
Our current site sample <strong>in</strong> <strong>the</strong> Felsenthal region, a cumulative result <strong>of</strong> both <strong>in</strong>tensive<br />
and systematic site survey <strong>in</strong> recent years and traditional reconnaissance and collector leads
76 House<br />
spann<strong>in</strong>g decades, encompasses a wide range <strong>of</strong> habitat types and significant functional<br />
variability. The number and magnitude <strong>of</strong> prehistoric mound centers <strong>in</strong> this hi<strong>the</strong>rto obscure<br />
region <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>ast is particularly impressive. These mound centers appear <strong>in</strong> every<br />
case to have complex histories and perhaps chang<strong>in</strong>g functions through time. Though data<br />
on specific community patterns are yet very limited, <strong>the</strong>re appears to be a shift from more<br />
nucleated settlements <strong>in</strong> <strong>the</strong> Coles Creek period to more dispersed settlement <strong>in</strong> subsequent<br />
times. The proposed seasonal occupation and extractive function <strong>of</strong> late prehistoric sites on<br />
<strong>the</strong> floodpla<strong>in</strong> (Rol<strong>in</strong>gson 1972) receives considerable support from <strong>the</strong> results <strong>of</strong> <strong>the</strong> present<br />
<strong>in</strong>vestigation (see Chapter 6 below).<br />
Recovery and analysis <strong>of</strong> large assemblages <strong>of</strong> ec<strong>of</strong>actual rema<strong>in</strong>s is thus far conf<strong>in</strong>ed<br />
to Shallow Lake Mound C (Rol<strong>in</strong>gson and Schambach 1981: 55-68) though analysis <strong>of</strong> a<br />
small sample <strong>of</strong> faunal rema<strong>in</strong>s from badly-disturbed contexts at <strong>the</strong> T. E. Salsbury site<br />
(16OU15) is presented by Price and Heartfield (1977:92-94). Small samples <strong>of</strong> floral and<br />
faunal rema<strong>in</strong>s recovered by test<strong>in</strong>g at floodpla<strong>in</strong> sites dur<strong>in</strong>g <strong>the</strong> present <strong>in</strong>vestigation are<br />
described <strong>in</strong> Appendix B <strong>of</strong> this report. Until extensive series <strong>of</strong> faunal and floral samples<br />
from sites <strong>in</strong> a variety <strong>of</strong> habitats have been analyzed and reported, <strong>the</strong> most basic questions<br />
about paleoenvironments and aborig<strong>in</strong>al economies will rema<strong>in</strong> unanswered. The floral and<br />
faunal specimens recovered by <strong>the</strong> present survey are particularly valuable <strong>in</strong> demonstrat<strong>in</strong>g<br />
that even floodpla<strong>in</strong> extractive sites may have a potential to yield direct subsistence data.<br />
On <strong>the</strong> macroregional level, Schambach (Rol<strong>in</strong>gson and Schambach 1931:103-106) has<br />
addressed <strong>the</strong> Lower Mississippi Valley aff<strong>in</strong>ities <strong>of</strong> aborig<strong>in</strong>al components <strong>in</strong> <strong>the</strong> Felsenthal<br />
region <strong>in</strong> terms <strong>of</strong> hydrological, topographic, and biogeographic similarities between<br />
<strong>the</strong> Felsenthal lowland environment and that <strong>in</strong> <strong>the</strong> Lower Mississippi Valley. To br<strong>in</strong>g this<br />
problem <strong>in</strong>to sharper focus, it is necessary to explicitly acknowledge that <strong>the</strong> Lower Mississippi<br />
Valley is not a homogeneous set <strong>of</strong> regions and to beg<strong>in</strong> compar<strong>in</strong>g <strong>the</strong> Felsenthal<br />
region to specific Lower Mississippi Valley regions <strong>in</strong> terms <strong>of</strong> resources and <strong>the</strong>ir seasonality<br />
and abundance. In Chapter 2 <strong>of</strong> this report, Hemm<strong>in</strong>gs provides a detailed discussion <strong>of</strong><br />
<strong>the</strong> Gran <strong>Marais</strong> Lowland environment <strong>in</strong> <strong>the</strong>se terms.<br />
It may ultimately prove especially important that <strong>the</strong> aborig<strong>in</strong>al components <strong>in</strong> <strong>the</strong><br />
Felsenthal region exhibit strong similarities to units <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn ra<strong>the</strong>r than <strong>the</strong> nor<strong>the</strong>rn<br />
Lower Mississippi Valley. The nor<strong>the</strong>rn/sou<strong>the</strong>rn dichotomy <strong>in</strong> <strong>the</strong> Mississippi Alluvial Valley<br />
has long been recognized. In a recent review <strong>of</strong> Lower Mississippi Valley faunal studies,<br />
Spr<strong>in</strong>ger (1980) noted that <strong>in</strong> <strong>the</strong> sou<strong>the</strong>rn Lower Mississippi Valley fish tend to exhibit a<br />
markedly greater importance relative to terrestrial species <strong>in</strong> comparison to <strong>the</strong> nor<strong>the</strong>rn<br />
Lower Mississippi Valley. Fur<strong>the</strong>rmore, this geographic dichotomy <strong>in</strong> archeologically-recovered<br />
faunal assemblages appears to override localized habitat variability. Thus, suggests<br />
Spr<strong>in</strong>ger, <strong>the</strong> observed north-south cultural dichotomy <strong>in</strong> <strong>the</strong> Lower Mississippi Valley may
Prehistory 77<br />
have important underly<strong>in</strong>g ecological dimensions. It will be <strong>in</strong>terest<strong>in</strong>g to see what faunal<br />
resource exploitation patterns ultimately emerge from <strong>in</strong>vestigation <strong>of</strong> sites <strong>in</strong> <strong>the</strong> Felsenthal<br />
region. This question is obviously <strong>of</strong> direct relevance to socioculture, anthropology’s current<br />
concern with <strong>the</strong> ecological basis <strong>of</strong> ethnicity, and <strong>the</strong> adaptive significance <strong>of</strong> cultural variability<br />
(cf. Barth 1969; Smith 1978; Hardesty 1980:162-163).
Chapter 4<br />
history <strong>of</strong> <strong>the</strong> felsenthAl region<br />
by Beverly Watk<strong>in</strong>s<br />
<strong>in</strong>troduction<br />
This chapter exam<strong>in</strong>es <strong>the</strong> history <strong>of</strong> south-central <strong>Arkansas</strong> along <strong>the</strong> Ouachita River.<br />
Focus<strong>in</strong>g on <strong>the</strong> importance <strong>of</strong> <strong>the</strong> river itself to <strong>the</strong> history <strong>of</strong> <strong>the</strong> region, this study follows<br />
<strong>the</strong> chang<strong>in</strong>g economy <strong>of</strong> <strong>the</strong> area from hunt<strong>in</strong>g, to cotton, to lumber<strong>in</strong>g, and traces fluctuations<br />
<strong>in</strong> river commerce as a result <strong>of</strong> <strong>the</strong>se changes.<br />
Two archeological sites that are closely tied to <strong>the</strong>se <strong>the</strong>mes are discussed <strong>in</strong> some detail.<br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g was established as a cotton shipp<strong>in</strong>g po<strong>in</strong>t and died when lumber<br />
replaced cotton as <strong>the</strong> area’s most important product. The Lotawanna, which burned and<br />
sank <strong>in</strong> <strong>the</strong> Felsenthal Project area, was <strong>the</strong> largest sternwheel steamboat on <strong>the</strong> Ouachita<br />
River <strong>in</strong> <strong>the</strong> 1870s, and carried both passengers and freight for <strong>the</strong> Blanks Steamboat L<strong>in</strong>e.<br />
These two sites are representative <strong>of</strong> <strong>the</strong> area <strong>in</strong> <strong>the</strong> second half <strong>of</strong> <strong>the</strong> n<strong>in</strong>eteenth century at<br />
<strong>the</strong> peak <strong>of</strong> its cotton economy.<br />
europeAn explorAtion And settleMent<br />
The Ouachita River Valley was visited by Europeans on <strong>the</strong>ir earliest explorations west<br />
<strong>of</strong> <strong>the</strong> Mississippi River. Traversed by <strong>the</strong> De Soto expedition <strong>in</strong> 1541-1542, <strong>the</strong> next recorded<br />
visits were those <strong>of</strong> Henri de Tonti <strong>in</strong> 1689 and <strong>the</strong> Sieur De Bienville <strong>in</strong> <strong>17</strong>00. The first European<br />
settlements on <strong>the</strong> Ouachita River were started about <strong>17</strong>18 near Monroe, Louisiana, but<br />
<strong>the</strong> river valley <strong>in</strong> <strong>Arkansas</strong> was known only to hunters and trappers (Mitchell and Calhoun<br />
1937:291-293).<br />
The European population grew, and a census was ordered soon after Spa<strong>in</strong> took possession<br />
<strong>of</strong> Louisiana under <strong>the</strong> Treaty <strong>of</strong> <strong>17</strong>62, which ended <strong>the</strong> French and Indian War. The<br />
count was made <strong>in</strong> <strong>17</strong>69. At that time <strong>the</strong>re were 110 white persons liv<strong>in</strong>g <strong>in</strong> <strong>the</strong> Washita<br />
district <strong>of</strong> south <strong>Arkansas</strong> and nor<strong>the</strong>ast Louisiana (Mitchell and Calhoun 1937: 294). This<br />
figure may have <strong>in</strong>cluded <strong>the</strong> families that founded Longview on <strong>the</strong> Sal<strong>in</strong>e River at about<br />
this same date (Goodspeed 1890:876). The area came under <strong>the</strong> supervision <strong>of</strong> <strong>the</strong> Commandant<br />
<strong>of</strong> <strong>the</strong> Natchitoches District, Athanase de Mezieres, who was ordered to clear <strong>the</strong><br />
Ouachita River <strong>of</strong> <strong>the</strong> vagabonds (wanderers) liv<strong>in</strong>g <strong>the</strong>re. He accomplished this <strong>in</strong> <strong>17</strong>74<br />
with <strong>the</strong> help <strong>of</strong> <strong>the</strong> Caddo Indians, and reported <strong>the</strong> removal <strong>of</strong> at least 14 people, <strong>in</strong>clud<strong>in</strong>g<br />
Pedro Champagnolle and <strong>the</strong> Le Boeuf family (Bolton 1914:114).
80 Watk<strong>in</strong>s<br />
Throughout <strong>the</strong> <strong>17</strong>70s Spa<strong>in</strong> was concerned by <strong>the</strong> growth <strong>of</strong> American settlements<br />
along <strong>the</strong> Ohio River, and <strong>in</strong> Kentucky and Tennessee, <strong>in</strong> defiance <strong>of</strong> <strong>the</strong> Proclamation <strong>of</strong><br />
<strong>17</strong>63 restrict<strong>in</strong>g English settlements to <strong>the</strong> lands east <strong>of</strong> <strong>the</strong> Appalachian Mounta<strong>in</strong>s. When<br />
<strong>the</strong> United States defeated England <strong>in</strong> <strong>the</strong> War <strong>of</strong> <strong>the</strong> American Revolution and was given<br />
title to <strong>the</strong> lands east <strong>of</strong> <strong>the</strong> Mississippi River, Esteban Miro, <strong>the</strong> new governor <strong>of</strong> Louisiana,<br />
decided that <strong>the</strong> time had come to establish a post and settlements along <strong>the</strong> Ouachita River<br />
as a foil to any fur<strong>the</strong>r expansion <strong>the</strong> Americans might have <strong>in</strong> m<strong>in</strong>d. He commissioned Jean<br />
Baptiste Filhoil as <strong>the</strong> first commandant <strong>of</strong> <strong>the</strong> Ouachita District, and sent him up <strong>the</strong> river<br />
to ga<strong>the</strong>r <strong>the</strong> European population <strong>in</strong>to settlements. Filhoil established his first post at Ecore<br />
a Fabre (Camden, <strong>Arkansas</strong>), but moved two years later to Prairie des Canots (now Monroe,<br />
Louisiana) where he was granted a large tract <strong>of</strong> land (Mitchell and Calhoun 1937:295;<br />
Sprague 1974:194-213). The population <strong>of</strong> <strong>the</strong> area grew so that by <strong>17</strong>90 <strong>the</strong>re were 242 persons<br />
(<strong>in</strong>clud<strong>in</strong>g 19 slaves) <strong>in</strong> <strong>the</strong> Ouachita District, more than double <strong>the</strong> <strong>17</strong>70 count (Greene<br />
et al. 1975:7-13).<br />
Louisiana was returned to France <strong>in</strong> 1800, but through an amicable arrangement <strong>the</strong><br />
Spanish <strong>of</strong>ficials cont<strong>in</strong>ued to adm<strong>in</strong>ister <strong>the</strong> affairs <strong>of</strong> <strong>the</strong> prov<strong>in</strong>ce. By early 1803 <strong>the</strong> French<br />
colonial adm<strong>in</strong>istration <strong>in</strong> New Orleans had become <strong>in</strong>terested <strong>in</strong> <strong>the</strong> Ouachita District.<br />
Charles F. Adrien le Paul<strong>in</strong>ier, <strong>the</strong> Chevalier D’Anemours, former Consul General <strong>of</strong> France<br />
to <strong>the</strong> United States, who had settled at Ouachita Post <strong>in</strong> <strong>17</strong>96, submitted a lengthy report on<br />
<strong>the</strong> area. He described <strong>the</strong> rivers, land, trees, and m<strong>in</strong>erals <strong>of</strong> <strong>the</strong> District. In list<strong>in</strong>g agricultural<br />
products he stated that <strong>the</strong> residents grew corn, sweet potatoes, Irish potatoes, peas,<br />
and o<strong>the</strong>r vegetables for <strong>the</strong>ir own use, and tobacco for trade. Hunters brought <strong>in</strong> <strong>the</strong> sk<strong>in</strong>s<br />
<strong>of</strong> deer, bear, beaver, and otter, and bear grease and tallow. D’Anemours did not say how<br />
many farmers <strong>the</strong>re were, but that <strong>the</strong>re were about 60 white hunters and 600 to 800 Indian<br />
hunters liv<strong>in</strong>g <strong>in</strong> <strong>the</strong> Ouachita District (D’Anemours 1803:26, 34).<br />
AMericAn explorAtion And settleMent<br />
After <strong>the</strong> United States bought Louisiana from France <strong>in</strong> 1803, a number <strong>of</strong> expeditions<br />
were organized to explore <strong>the</strong> new territory. One <strong>of</strong> <strong>the</strong>se, led by William Dunbar and<br />
George Hunter, <strong>in</strong>vestigated <strong>the</strong> Ouachita River as far north as Hot Spr<strong>in</strong>gs to learn someth<strong>in</strong>g<br />
<strong>of</strong> <strong>the</strong> <strong>in</strong>terior <strong>of</strong> <strong>the</strong> country between <strong>the</strong> Mississippi River and <strong>the</strong> Red River. Dunbar,<br />
a well known naturalist from Natchez, and Hunter, a chemist and astronomer, were well<br />
tra<strong>in</strong>ed to make observations on <strong>the</strong> plants, animals, and geology <strong>of</strong> <strong>the</strong> area as well as to analyze<br />
spr<strong>in</strong>gs and salt licks and to record locations so that a map <strong>of</strong> <strong>the</strong> new territory could<br />
be drawn.<br />
The expedition left Fort Miro (Monroe) late on <strong>the</strong> afternoon <strong>of</strong> November 11, 1804,<br />
and entered <strong>the</strong> Felsenthal Project area four days later. Their guide was a man who had<br />
lived <strong>in</strong> <strong>the</strong> area for over ten years and who had traveled to <strong>the</strong> hot spr<strong>in</strong>gs several times. On
History 81<br />
November 15 <strong>the</strong>y passed <strong>the</strong> entrance to <strong>Grand</strong> <strong>Marais</strong> and camped just below Lapile<br />
Creek. The next day <strong>the</strong>y passed <strong>Marais</strong> Sal<strong>in</strong>e and <strong>the</strong> Sal<strong>in</strong>e River and camped near St.<br />
Mary’s Lake (Figure 15). Dur<strong>in</strong>g <strong>the</strong> day <strong>the</strong>y passed several hunters’ cab<strong>in</strong>s, but all were<br />
empty. On November <strong>17</strong> <strong>the</strong>y passed Caney <strong>Marais</strong> and made <strong>the</strong>ir camp outside <strong>the</strong> project<br />
area. The follow<strong>in</strong>g day, November 18, just below Moro Bay, Dunbar and Hunter met two<br />
French hunters, part <strong>of</strong> a larger party us<strong>in</strong>g dogs to hunt bear. On <strong>the</strong> n<strong>in</strong>eteenth Hunter<br />
noted now low <strong>the</strong> surround<strong>in</strong>g country was as <strong>the</strong>y passed through <strong>the</strong> area <strong>of</strong> <strong>the</strong> new<br />
Calion lock and dam; after pass<strong>in</strong>g <strong>the</strong> hills <strong>of</strong> Champagnolle, <strong>the</strong>y made camp just above<br />
Champagnolle Creek.<br />
As <strong>the</strong> expedition returned through <strong>the</strong> area on its way downriver, Hunter recorded<br />
on January 12, 1805, that <strong>the</strong>y passed Champagnolle Creek, <strong>the</strong>n John Sk<strong>in</strong>ner’s camp, before<br />
camp<strong>in</strong>g two leagues below <strong>the</strong> Champagnolle hills. (A league equals about 3 mar<strong>in</strong>e<br />
miles or less than 5 km.) The next day after <strong>the</strong>y passed Moro Bay and <strong>the</strong> place used by<br />
Tulipe for stor<strong>in</strong>g his hides, <strong>the</strong>y came to a clear<strong>in</strong>g called <strong>the</strong> Eagle (L’Aigle) and to a place<br />
known as la pirague d’Auguste (now corrupted to Perrogee<strong>the</strong>e and o<strong>the</strong>r forms). They<br />
camped 1.5 leagues far<strong>the</strong>r down <strong>the</strong> river. On <strong>the</strong> fourteenth Hunter noted both Lapoile<br />
and Lapile Creeks before <strong>the</strong>y left <strong>the</strong> project area and returned to Louisiana (McDermott<br />
1963:91-94, 111-112; Rowland 1930:242-246, 310-312).<br />
The camps Dunbar and Hunter saw, and <strong>the</strong> names <strong>of</strong> <strong>the</strong> places <strong>the</strong>y visited emphasize<br />
<strong>the</strong> importance <strong>of</strong> hunters to <strong>the</strong> area. Hunter described a typical camp that <strong>the</strong>y passed<br />
<strong>in</strong> Louisiana as hav<strong>in</strong>g a bark cab<strong>in</strong> one-story high, about 15 feet square with no w<strong>in</strong>dows<br />
and a grass and mud chimney. The furnish<strong>in</strong>gs for <strong>the</strong> family <strong>of</strong> six were a bed, three<br />
blocks <strong>of</strong> wood for stools, a trough for pound<strong>in</strong>g corn, a rifle, and a shot pouch (McDermott<br />
1963:90).<br />
Settlement <strong>of</strong> <strong>the</strong> lands near <strong>the</strong> project area proceeded slowly. The basic pattern <strong>of</strong><br />
growth can be seen <strong>in</strong> <strong>the</strong> creation <strong>of</strong> new counties. Union County, which covered all <strong>of</strong><br />
south-central <strong>Arkansas</strong>, was formed <strong>in</strong> 1829. Bradley County was created from parts <strong>of</strong><br />
Union and Chicot counties <strong>in</strong> 1840. Ouachita County was formed from <strong>the</strong> nor<strong>the</strong>rn part <strong>of</strong><br />
Union County <strong>in</strong> 1842. Ashley County was added <strong>in</strong> 1848, and Calhoun County was created<br />
<strong>in</strong> 1852.<br />
As <strong>the</strong> population grew, land<strong>in</strong>gs on <strong>the</strong> Ouachita River became trade centers for <strong>the</strong><br />
<strong>in</strong>terior. Champagnolle, which was established before 1830 by Lawrence, John, and Silas<br />
Scarborough, became a prosperous town and <strong>the</strong> most important shipp<strong>in</strong>g po<strong>in</strong>t on <strong>the</strong> river<br />
between Camden and <strong>the</strong> Louisiana l<strong>in</strong>e. The county seat was at Champagnolle from 1838 to<br />
1844, and a government land <strong>of</strong>fice was opened <strong>the</strong>re <strong>in</strong> 1845 shortly after <strong>the</strong> surveys were<br />
f<strong>in</strong>ished. The town had several stores, a tavern, and a jail, and several <strong>of</strong> <strong>the</strong> houses had<br />
plastered walls and formal gardens (Cordell 1951:37-42).
82 Watk<strong>in</strong>s<br />
Figure 15. Historic sites, settlements, and place-names <strong>in</strong> <strong>the</strong> Felsenthal region.
History 83<br />
The land<strong>in</strong>gs at Pigeon Hill, Wilm<strong>in</strong>gton, and Caryville (also called Burk’s Land<strong>in</strong>g)<br />
ga<strong>in</strong>ed importance <strong>in</strong> <strong>the</strong> late 1840s. Even though <strong>the</strong>y were closer than Champagnolle, <strong>the</strong>y<br />
still were not conve nient to <strong>the</strong> people who had settled <strong>in</strong> sou<strong>the</strong>ast Union County <strong>in</strong> what<br />
was called <strong>the</strong> “Dark Corner” until Lapile Township was created <strong>in</strong> 1846. William Rowland,<br />
who had settled <strong>in</strong> 1843 near present-day Huttig met <strong>the</strong> need for a land<strong>in</strong>g by build<strong>in</strong>g a<br />
float<strong>in</strong>g dock at <strong>the</strong> nearest po<strong>in</strong>t on <strong>the</strong> river. Known as Rowland’s Raft, this land<strong>in</strong>g helped<br />
<strong>the</strong> Lapile community grow. Wesley Rowland and his bro<strong>the</strong>r William Lorenzo started <strong>the</strong><br />
first store <strong>the</strong>re about 1850, and John V. McHenry built <strong>the</strong> first g<strong>in</strong> about <strong>the</strong> same time. The<br />
community had Methodist and Primitive Baptist churches, and a school named <strong>the</strong> Lapile<br />
Academy (Cordell 1947:275-281).<br />
O<strong>the</strong>r improvements were also needed as <strong>the</strong> population grew away from <strong>the</strong> river.<br />
The Black family who settled sou<strong>the</strong>ast <strong>of</strong> Burk’s Land<strong>in</strong>g <strong>in</strong> <strong>the</strong> 1830s had to cut a road<br />
from Farmerville, Louisiana to <strong>the</strong>ir new land. This road was later extended to <strong>the</strong> river at<br />
Burk’s Land<strong>in</strong>g and was eventually made a part <strong>of</strong> a road from Camden to Monroe (Spencer<br />
1953:228, 233-234). O<strong>the</strong>r roads ran from Pigeon Hill to Black’s, from Wallace’s (east <strong>of</strong> New<br />
London) to Lapile Creek, and from Wallace’s to Watt’s Land<strong>in</strong>g on Lapoile Creek (Union<br />
County Court Record B:98). Similar improvements were be<strong>in</strong>g made on <strong>the</strong> east side <strong>of</strong> <strong>the</strong><br />
river (General Land Office 1845).<br />
Ferries were established to make river cross<strong>in</strong>gs easier. The first license on this part<br />
<strong>of</strong> <strong>the</strong> river was issued to William Burk <strong>in</strong> 1834 for his ferry at <strong>the</strong> mouth <strong>of</strong> Moro Creek<br />
(Union County Court Record A:27). He later established <strong>the</strong> land<strong>in</strong>g and ferry that became<br />
Caryville. O<strong>the</strong>r ferries were established on <strong>the</strong> Sal<strong>in</strong>e River (Bradley County Court Record<br />
A:25, 56-58; Union County Court Record A:55).<br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g<br />
The community with <strong>the</strong> most direct impact on <strong>the</strong> project area was Marie Sal<strong>in</strong>e<br />
Land<strong>in</strong>g (see Site 3AS299 <strong>in</strong> Chapter 5) on <strong>the</strong> Ouachita River (Figure 16). The land<strong>in</strong>g<br />
and ferry at Marie Sal<strong>in</strong>e were established <strong>in</strong> 1852 by Nathanial M. Mulholland and Joseph<br />
D. Christian on <strong>the</strong> Ouachita River at <strong>the</strong> cross<strong>in</strong>g for <strong>the</strong> Hamburg-El Dorado Road<br />
(State Land Office Tract Book:18S 10W). Mulholland soon sold his <strong>in</strong>terest <strong>in</strong> <strong>the</strong> land<strong>in</strong>g<br />
to James A. Mason, and Mason and Christian received a license to run a ferry at <strong>the</strong> land<strong>in</strong>g<br />
(E<strong>the</strong>ridge 1959:31, 136). The land<strong>in</strong>g served a community that stretched along <strong>the</strong> road<br />
from Hamburg to <strong>the</strong> river, and along <strong>the</strong> road from Founta<strong>in</strong> Hill <strong>in</strong> <strong>the</strong> nor<strong>the</strong>rn part <strong>of</strong><br />
Ashley County to Extra Land<strong>in</strong>g on <strong>the</strong> river below Lock 6. Warehouses were built to store<br />
cotton and merchandise, and a canal was dug so that boats could reach <strong>the</strong> warehouses<br />
dur<strong>in</strong>g low water. Dur<strong>in</strong>g high water, flatboats were used to transport goods from <strong>the</strong><br />
warehouses to wagons wait<strong>in</strong>g on higher ground (E<strong>the</strong>ridge 1959:136). By 1856 <strong>the</strong> land<strong>in</strong>g<br />
served enough people that a post <strong>of</strong>fice was established with Joseph Christian as postmaster<br />
(U.S. Post Office n.d.). By 1860 <strong>the</strong> extended community <strong>in</strong>cluded 56 families, <strong>in</strong>clud<strong>in</strong>g doctors,<br />
surveyors, clerks, carpenters, mechanics, and a teacher (E<strong>the</strong>ridge 1959:135-136).
84 Watk<strong>in</strong>s<br />
Figure 16. Hydrographic survey chart <strong>of</strong> Ouachita<br />
River <strong>in</strong> <strong>the</strong> Felsenthal Project area, show<strong>in</strong>g<br />
locations <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g (3AS299),<br />
<strong>Grand</strong> Mary (or <strong>Grand</strong> <strong>Marais</strong>) Land<strong>in</strong>g<br />
(3UN122), and o<strong>the</strong>r features (U.S. Army<br />
Corps <strong>of</strong> Eng<strong>in</strong>eers 1871:Sheet 6).
History 85<br />
Trade on <strong>the</strong> river almost ceased dur<strong>in</strong>g <strong>the</strong> Civil War, but by 1870 Marie Sal<strong>in</strong>e Land<strong>in</strong>g<br />
was aga<strong>in</strong> a major shipp<strong>in</strong>g po<strong>in</strong>t for cotton. In 1871, 5,000 bales were shipped from<br />
<strong>the</strong>re, mak<strong>in</strong>g it <strong>the</strong> third largest shipper on <strong>the</strong> Ouachita River <strong>in</strong> <strong>Arkansas</strong>, after Camden<br />
and Caryville (U.S. Congress 1872:5). A map made <strong>the</strong> same year (Figure 16) shows three<br />
build<strong>in</strong>gs just north <strong>of</strong> <strong>the</strong> canal (Raynolds 1871:Sheet 6). The land<strong>in</strong>g changed owners several<br />
times <strong>in</strong> <strong>the</strong> late 1800s, go<strong>in</strong>g first to Fred Erickson and <strong>the</strong>n to Dock Bird, who also ran<br />
Extra Land<strong>in</strong>g. Marie Sal<strong>in</strong>e Land<strong>in</strong>g went out <strong>of</strong> bus<strong>in</strong>ess soon after <strong>the</strong> railroad reached<br />
<strong>the</strong> river, and <strong>the</strong> economy shifted from cotton to timber (E<strong>the</strong>ridge 1959:137).<br />
disposition <strong>of</strong> swAMplAnds<br />
In contrast to <strong>the</strong> growth <strong>of</strong> <strong>the</strong> surround<strong>in</strong>g areas, few people settled <strong>in</strong> lowly<strong>in</strong>g<br />
lands <strong>of</strong> <strong>the</strong> Felsenthal Project area, primarily because <strong>the</strong>re was still plenty <strong>of</strong> good land<br />
available that was not subject to flood<strong>in</strong>g. In 1850 Congress gave all unclaimed land that was<br />
swamp or subject to overflow to <strong>the</strong> states. Called swamplands, <strong>the</strong> proceeds from <strong>the</strong> sale<br />
<strong>of</strong> <strong>the</strong>se lands was to be used to reclaim more land by build<strong>in</strong>g levees and dra<strong>in</strong>s. <strong>Arkansas</strong><br />
eventually got 8,600,000 acres <strong>of</strong> swampland <strong>in</strong>clud<strong>in</strong>g most <strong>of</strong> <strong>the</strong> land <strong>in</strong> <strong>the</strong> project<br />
area (Harrison and Kollmorgen 1947:369-371). <strong>Arkansas</strong> found it very difficult to sell its<br />
swamplands, so <strong>in</strong> <strong>the</strong> late 1850s discussion began on donat<strong>in</strong>g swamplands to a number <strong>of</strong><br />
railroads <strong>in</strong> exchange for stock, <strong>the</strong> justification be<strong>in</strong>g that railroad beds and trestles acted<br />
as levees and dra<strong>in</strong>s. The <strong>in</strong>terest from <strong>the</strong> stock was to be used to reclaim lands as called<br />
for <strong>in</strong> <strong>the</strong> orig<strong>in</strong>al act. The lands <strong>in</strong> <strong>the</strong> project area, along with all <strong>of</strong> <strong>the</strong> swamplands <strong>in</strong> <strong>the</strong><br />
Champagnolle Land District, were given to <strong>the</strong> Mississippi, Ouachita, and Red River Railroad.<br />
This clouded title to <strong>the</strong> lands and prevented private ownership <strong>of</strong> <strong>the</strong>se lands until<br />
problems with this grant were resolved <strong>in</strong> 1877 (Watk<strong>in</strong>s 1981:343-44). Consequently, by<br />
1860 only 1200 acres with<strong>in</strong> <strong>the</strong> area now designated Felsenthal National Wildlife Refuge<br />
had been claimed, mostly <strong>the</strong> higher ground at <strong>the</strong> edges <strong>of</strong> <strong>the</strong> area (State Land Office Tract<br />
Books).<br />
riVer coMMerce And trAVel<br />
Central to <strong>the</strong> study <strong>of</strong> <strong>the</strong> history <strong>of</strong> <strong>the</strong> Felsenthal Project area is <strong>the</strong> Ouachita River<br />
itself. The river was used for commerce as soon as <strong>the</strong>re were settlers to serve, first by flatboats<br />
and keelboats, <strong>the</strong>n <strong>in</strong> <strong>the</strong> 1820s by steamboats. The first steamboats on <strong>the</strong> Ouachita<br />
River <strong>in</strong> <strong>Arkansas</strong> were <strong>the</strong> Dime owned by Jacob Barkman, and <strong>the</strong> Natchitoches and Enterprise<br />
owned by John Nunn (WPA Author n.d.). The first to navigate <strong>the</strong> lower Sal<strong>in</strong>e River<br />
was <strong>the</strong> Howard piloted by Capta<strong>in</strong> Bob Wi<strong>the</strong>rs (Mart<strong>in</strong> 1936). Steamboat<strong>in</strong>g on <strong>the</strong>se rivers<br />
was seasonal, be<strong>in</strong>g dependent on high water, and <strong>the</strong> boats were usually shallow draft<br />
sidewheelers, which could be easily maneuvered through narrow channels and around tight<br />
bends. In spite <strong>of</strong> <strong>the</strong> difficulties, as <strong>the</strong> population grew and <strong>the</strong> production <strong>of</strong> cotton <strong>in</strong>-
86 Watk<strong>in</strong>s<br />
creased, river commerce prospered. By 1855 <strong>the</strong>re were n<strong>in</strong>e boats mak<strong>in</strong>g regular trips to<br />
Camden (Camden <strong>in</strong> 1855:10-11).<br />
The trade centered around cotton, corn, and merchandise. The higher lands along <strong>the</strong><br />
Ouachita River were prime agricultural lands, well suited to a plantation economy <strong>of</strong> slaves<br />
and cotton. As early as 1840, Union County had one <strong>of</strong> <strong>the</strong> ten largest slave populations <strong>in</strong><br />
<strong>the</strong> state. By 1850 Union County had <strong>the</strong> largest slave population <strong>in</strong> <strong>the</strong> state (4,767), with<br />
Ouachita County <strong>in</strong> fourth place (3,304). By 1860, Union County had dropped to fourth<br />
place (6,331), Ouachita County had dropped to seventh (4,478), and Ashley County had<br />
moved <strong>in</strong>to tenth place (3,761) (Taylor 1958:51-52). Obviously most <strong>of</strong> <strong>the</strong>se slaves did not<br />
live and work immediately along <strong>the</strong> river, but <strong>the</strong> products <strong>of</strong> <strong>the</strong>ir labor were shipped to<br />
New Orleans by boat.<br />
At this same time, Camden became an important commercial center, second only to<br />
Little Rock <strong>in</strong> population. By 1860, <strong>the</strong> town had 56 merchants, n<strong>in</strong>e <strong>of</strong> <strong>the</strong>m do<strong>in</strong>g bus<strong>in</strong>ess<br />
<strong>in</strong> excess <strong>of</strong> $30,000, as well as three newspapers, three hotels, a woolen mill, a brick factory,<br />
a foundry, a sash factory, and a tannery. In addition <strong>the</strong> town boasted 21 lawyers, 11 doctors,<br />
six churches, and three schools (Kellam 1970:<strong>17</strong>-18). In 1861 roughly 20,000 bales <strong>of</strong> cotton<br />
(at 400 pounds per bale) were shipped down <strong>the</strong> river from Camden (Kellam 1971:18); <strong>the</strong><br />
return freights were merchandise and corn. The o<strong>the</strong>r communities along <strong>the</strong> river contributed<br />
to <strong>the</strong> trade <strong>in</strong> proportion to <strong>the</strong>ir size, and all <strong>of</strong> this trade passed through <strong>the</strong> Felsenthal<br />
Project area.<br />
By <strong>the</strong> 1840s most boats were also carry<strong>in</strong>g passengers, and a few had begun to specialize<br />
<strong>in</strong> this bus<strong>in</strong>ess. The accommodations varied. In 1854 <strong>the</strong> Louisa, owned by Frank<br />
Keel<strong>in</strong>g, was advertised as “elegantly fitted with modern improvements...hav<strong>in</strong>g new furniture<br />
and carpets” (Griff<strong>in</strong> 1932:65). The J. F. Pargoud was described <strong>in</strong> 1860, shortly after it<br />
was built, as hav<strong>in</strong>g rosewood furniture, lace curta<strong>in</strong>s, chandeliers, silverware, bathrooms,<br />
and “all o<strong>the</strong>r conveniences” (Thatcher 1970:6). It became a standard way <strong>of</strong> life to travel by<br />
boat for both bus<strong>in</strong>ess and pleasure (Williamson and Williamson 1939:208-212).<br />
River travel was not without its hazards, however. Boats hit snags, float<strong>in</strong>g debris, or<br />
sandbars, and sank. Boilers exploded, boats burned, or were trapped by fall<strong>in</strong>g river levels.<br />
The first loss on <strong>the</strong> Ouachita River <strong>in</strong> <strong>Arkansas</strong> was <strong>the</strong> Fair Star which sank at Pereogee<strong>the</strong>e<br />
Shoals <strong>in</strong> 1827. She was followed by <strong>the</strong> Chiefta<strong>in</strong> which hit a snag at Walnut Hills 20<br />
miles (32 km) below Camden <strong>in</strong> 1835, <strong>the</strong> Hadie sank at Ragged Island (now Haidee Shoals)<br />
<strong>in</strong> 1850, <strong>the</strong> Romeo which went down above Horsehead Shoals <strong>in</strong> 1851, and <strong>the</strong> John Ray<br />
which burned at P<strong>in</strong>e Prairie <strong>in</strong> 1859 (Williamson and Williamson 1939:218; <strong>Arkansas</strong> Gazette<br />
April 28, 1835).
effects <strong>of</strong> <strong>the</strong> ciVil wAr<br />
History 87<br />
The com<strong>in</strong>g <strong>of</strong> <strong>the</strong> Civil War brought trade on <strong>the</strong> Ouachita River almost to a standstill.<br />
The blockade <strong>of</strong> New Orleans and its capture by <strong>the</strong> Union Navy <strong>in</strong> 1862 left no market<br />
for <strong>the</strong> sale <strong>of</strong> cotton, nor any place to buy o<strong>the</strong>r merchandise for trad<strong>in</strong>g. Although some<br />
planters cont<strong>in</strong>ued to grow cotton, more attention was given to corn, wheat, and vegetables.<br />
Some trade was ma<strong>in</strong>ta<strong>in</strong>ed by <strong>the</strong> smaller boats; <strong>the</strong> larger boats served <strong>the</strong> Confederacy<br />
as transports and cotton-clad gunships. Some were sunk near Vicksburg as obstacles to <strong>the</strong><br />
Union gunboats (Owens 1979:63-103). On <strong>the</strong> Ouachita River, a few boats were beached or<br />
abandoned to keep <strong>the</strong>m out <strong>of</strong> Union hands. One <strong>of</strong> <strong>the</strong>se, <strong>the</strong> Beauregard, escaped from<br />
New Orleans with a load <strong>of</strong> sugar and molasses, and was left at Pereogee<strong>the</strong>e Shoals. After<br />
she was recovered, she served <strong>the</strong> Confederacy by haul<strong>in</strong>g soldiers and corn on Bayou Macon<br />
(McMahan [1913]).<br />
There were no battles or skirmishes <strong>in</strong> <strong>the</strong> Felsenthal Project area. The Union Army<br />
came no fur<strong>the</strong>r south than Camden, although <strong>the</strong>y did capture <strong>the</strong> steamer Homer (Green<br />
1864). The Union Navy came only as far north as Ouachita City, stopp<strong>in</strong>g at Monroe to burn<br />
<strong>the</strong> courthouse and <strong>the</strong> railroad bridge and depot (Foster 1864).<br />
iMproVeMents <strong>in</strong> nAVigAtion<br />
Follow<strong>in</strong>g <strong>the</strong> Civil War, <strong>the</strong> Ouachita River was once aga<strong>in</strong> <strong>the</strong> focus <strong>of</strong> <strong>the</strong> commercial<br />
life <strong>of</strong> south-central <strong>Arkansas</strong>. Planters who had stored cotton dur<strong>in</strong>g <strong>the</strong> war were<br />
anxious to ship it to New Orleans quickly to get <strong>the</strong> best price. In 1865 <strong>the</strong>re were 45,000<br />
bales <strong>of</strong> cotton at Camden, and as many as 11 steamboats at <strong>the</strong> wharf at one time (McMahan<br />
[1913]). Cotton prices fell because <strong>of</strong> <strong>the</strong> surplus available, but even so it rema<strong>in</strong>ed <strong>the</strong><br />
pr<strong>in</strong>cipal agricultural product as tenant farm<strong>in</strong>g and sharecropp<strong>in</strong>g agreements forced farmers<br />
to concentrate on a cash crop <strong>in</strong> order to pay <strong>the</strong>ir debts.<br />
Although <strong>the</strong> railroad reached Arkadelphia <strong>in</strong> 1873, for farmers far<strong>the</strong>r south it was<br />
easier to ship <strong>the</strong>ir cotton on <strong>the</strong> river. Approximately 68,000 bales were shipped each year<br />
from Camden and o<strong>the</strong>r land<strong>in</strong>gs on <strong>the</strong> Ouachita River <strong>in</strong> <strong>Arkansas</strong>, <strong>in</strong>clud<strong>in</strong>g 5,000 bales<br />
from Marie Sal<strong>in</strong>e (U.S. Congress 1872:5). Dur<strong>in</strong>g <strong>the</strong> 1872 season 19 steamboats made 67<br />
trips up <strong>the</strong> river to Camden (U.S. Congress 1874:7). In 1876, eight steamboats made 43 trips<br />
to Camden (U.S. Congress 1877:479). And <strong>in</strong> <strong>the</strong> 1899 season, just before <strong>the</strong> railroads were<br />
extended <strong>in</strong>to <strong>the</strong> extreme sou<strong>the</strong>rn part <strong>of</strong> <strong>the</strong> state, eight boats made 24 trips to Camden,<br />
plus 74 trips to o<strong>the</strong>r land<strong>in</strong>gs <strong>in</strong> <strong>Arkansas</strong> (U.S. Congress 1900:19-20). Altoge<strong>the</strong>r <strong>in</strong> <strong>the</strong><br />
30 years follow<strong>in</strong>g <strong>the</strong> Civil War, 105 steamboats operated on <strong>the</strong> Ouachita River (Sifford<br />
n.d.:Vol. 4).
88 Watk<strong>in</strong>s<br />
As river trade prospered, sternwheel boats replaced sidewheel boats because <strong>the</strong>y<br />
had a larger cargo area and more power. Two bro<strong>the</strong>rs from Monroe, Fred and Jack Blanks,<br />
established a steamboat l<strong>in</strong>e to provide regular service from New Orleans to Monroe and as<br />
far north as Arkadelphia with connections with smaller boats up various tributaries <strong>of</strong> <strong>the</strong><br />
Ouachita River (Ouachita Telegraph May 1, 1874). Their largest boat was <strong>the</strong> Ouachita Belle<br />
which called at land<strong>in</strong>gs <strong>in</strong> <strong>the</strong> project area (see Site 3UN155 <strong>in</strong> Chapter 5) dur<strong>in</strong>g <strong>the</strong> 1870s.<br />
In <strong>the</strong> 1880s <strong>the</strong> Blanks L<strong>in</strong>e was jo<strong>in</strong>ed by Capta<strong>in</strong> L. V. Cooley, whose boats had dom<strong>in</strong>ated<br />
<strong>the</strong> Tensas River trade, and <strong>the</strong> new company was named <strong>the</strong> Ouachita Consolidated Company<br />
(Williamson and Williamson 1939:221-223).<br />
Steamboat<strong>in</strong>g on <strong>the</strong> Ouachita River cont<strong>in</strong>ued to be a hazardous occupation. Among<br />
<strong>the</strong> boats lost were <strong>the</strong> Willie sunk above Camden <strong>in</strong> 1868; <strong>the</strong> Native sunk at Camden <strong>in</strong><br />
1872; <strong>the</strong> McCullough burned at Rowland’s Raft <strong>in</strong> 1873; and <strong>the</strong> Lotawanna (Figure <strong>17</strong>)<br />
burned below <strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g <strong>in</strong> 1874 (see below and Chapter 5) (Williamson and<br />
Williamson 1939:218-219). The greatest hazards were still snags and shoals.<br />
As early as 1840 citizens <strong>of</strong> Clark County had asked Congress for funds to improve<br />
navigation on <strong>the</strong> Ouachita River (<strong>Arkansas</strong> Gazette April 15, 1840) but noth<strong>in</strong>g came <strong>of</strong> <strong>the</strong><br />
request. Follow<strong>in</strong>g <strong>the</strong> Civil War, local bus<strong>in</strong>essmen felt that it would be possible to build<br />
a series <strong>of</strong> locks and dams that would make <strong>the</strong> river navigable all year, <strong>the</strong>reby <strong>in</strong>creas<strong>in</strong>g<br />
trade and br<strong>in</strong>g<strong>in</strong>g a return <strong>of</strong> prosperity to <strong>the</strong> area. The first appropriation to study this<br />
idea was made <strong>in</strong> 1870, for a survey from Arkadelphia to <strong>the</strong> Louisiana l<strong>in</strong>e. The work was<br />
done by Just<strong>in</strong> Strazser <strong>of</strong> <strong>the</strong> Corps <strong>of</strong> Eng<strong>in</strong>eers, and resulted <strong>in</strong> <strong>the</strong> recommendation that<br />
$56,000 be spent to remove snags and lean<strong>in</strong>g trees (Scott 1940:7). In 1871 a civil eng<strong>in</strong>eer,<br />
Clement Smith, was hired to survey <strong>the</strong> river from Tr<strong>in</strong>ity, Louisiana, to Camden. He recommended<br />
a series <strong>of</strong> five locks and dams to provide a 4-foot deep channel, and <strong>of</strong>fered cost estimates<br />
from $1.1 to $2.1 million depend<strong>in</strong>g on whe<strong>the</strong>r <strong>the</strong> locks were <strong>of</strong> timber or masonry<br />
(U.S. Congress 1872:6-7). The project was approved by Congress and a contract was signed,<br />
but doubts soon arose about <strong>the</strong> accuracy <strong>of</strong> <strong>the</strong> survey and <strong>the</strong> wisdom <strong>of</strong> <strong>the</strong> recommendations.<br />
The Chief <strong>of</strong> Eng<strong>in</strong>eers <strong>the</strong>n authorized a new survey <strong>of</strong> <strong>the</strong> river. This work was done<br />
<strong>in</strong> 1872 by C. W. Durham under <strong>the</strong> direction <strong>of</strong> W. H. H. Benyaurd, who agreed that a system<br />
<strong>of</strong> locks and dams was <strong>the</strong> only way to make <strong>the</strong> river navigable year-round. He found,<br />
however, that <strong>the</strong> locks recommended by Smith were poorly placed and were too small for<br />
<strong>the</strong> largest boats. Benyaurd outl<strong>in</strong>ed several alternative plans us<strong>in</strong>g from seven to ten locks<br />
and dams, with <strong>the</strong> cost rang<strong>in</strong>g from $3 to $6 million, depend<strong>in</strong>g on <strong>the</strong> option chosen and<br />
<strong>the</strong> type <strong>of</strong> construction used. However, he felt that <strong>the</strong> amount <strong>of</strong> trade did not justify this<br />
much expense and recommended a special snagboat be built <strong>in</strong>stead (U.S. Congress 1874:3-<br />
9; Mills 1978:43-44).<br />
Snagg<strong>in</strong>g operations had begun <strong>in</strong> 1871 on <strong>the</strong> Ouachita River us<strong>in</strong>g two flatboats fitted<br />
with cranes, before Strazser submitted his report. These were hampered by a lack <strong>of</strong>
Figure <strong>17</strong>. The sternwheeler Lotawanna fully loaded with cotton bales (photo<br />
courtesy <strong>of</strong> Harry P. Fischer Collection, Marietta College Library)<br />
(AAS neg. 803785).<br />
History 89
90 Watk<strong>in</strong>s<br />
power and by <strong>the</strong> depth <strong>of</strong> water required. In 1874, us<strong>in</strong>g funds from <strong>the</strong> cancelled lock and<br />
dam project, Benyaurd had an iron-hulled, light draft snagboat, <strong>the</strong> O. G. Wagner, built and<br />
snagg<strong>in</strong>g began <strong>in</strong> earnest <strong>in</strong> 1875 (Mills 1978:44-49). Benyaurd also tried to improve navigation<br />
by <strong>the</strong> construction <strong>of</strong> w<strong>in</strong>g dams at <strong>the</strong> worst shoals. These concentrated <strong>the</strong> flow <strong>of</strong><br />
water <strong>in</strong>to one channel, which would <strong>the</strong>n be deeper at low water levels. Two <strong>of</strong> <strong>the</strong>se dams<br />
were built <strong>in</strong> <strong>Arkansas</strong> at Spoon Camp and at Buffalo Flats (Mills 1978:49; U.S. Congress<br />
1877:478).<br />
Dur<strong>in</strong>g <strong>the</strong> 1880s work on <strong>the</strong> river was limited to snagg<strong>in</strong>g operations to keep <strong>the</strong><br />
natural channels open. New surveys were made <strong>in</strong> 1882 and 1887, but each one concluded<br />
that <strong>the</strong> needed system <strong>of</strong> locks and dams would cost too much. In 1889 a new snagboat, <strong>the</strong><br />
Hooker, came to south <strong>Arkansas</strong>, but her wooden hull only lasted a few years (Scott 1940:10).<br />
At this same time, trade on <strong>the</strong> river was <strong>in</strong>creas<strong>in</strong>g and bus<strong>in</strong>essmen <strong>in</strong> Camden and<br />
Arkadelphia were anxious to provide for year-round navigation. In May 1893 <strong>the</strong> Ouachita<br />
Valley Improvement Association was formed at Camden. Led by Colonel Asa S. Morgan<br />
and Adolph Felsenthal, <strong>the</strong> group pressured Congress to authorize a slack water navigation<br />
system as far up <strong>the</strong> Ouachita River as possible. Congress responded with funds for a new,<br />
<strong>in</strong>tensive, and scientific survey (Thatcher 1970:18; Sifford n.d.; Scott 1940:10).<br />
The new survey was done <strong>in</strong> 1895 and 1896 by M. H. Marshall whose report, issued <strong>in</strong><br />
1902, recommended a series <strong>of</strong> n<strong>in</strong>e locks and dams, <strong>in</strong>clud<strong>in</strong>g Lock 6 at Rowland’s Raft and<br />
Lock 7 at Pereogee<strong>the</strong>e Shoals, to provide a 6.5-foot slack water channel. These plans were<br />
enthusiastically supported by <strong>the</strong> Corps <strong>of</strong> Eng<strong>in</strong>eers, and Congress authorized <strong>the</strong> project<br />
<strong>in</strong> June 1902. Work <strong>the</strong>n began on Lock 6 <strong>in</strong> 1904 and on Lock 7 <strong>in</strong> 1910. However, modifications<br />
to Lock 6 elim<strong>in</strong>ated <strong>the</strong> need for Lock 7, and work on that <strong>in</strong>stallation ended <strong>in</strong> 1912<br />
(see Site 3BR32 <strong>in</strong> Chapter 5). Although Lock 6 was f<strong>in</strong>ished <strong>in</strong> 1915, <strong>the</strong> entire project was<br />
not completed until 1925 (Thatcher 1970:13-14; Mills 1978:92).<br />
The Sternwheeler Lotawanna<br />
The Lotawanna was built <strong>in</strong> 1867 at Marietta, Ohio for Henry J. Br<strong>in</strong>ker <strong>of</strong> New Orleans.<br />
A temporary license, called an enrollment certificate, issued on October 29, 1867 at Wheel<strong>in</strong>g,<br />
West Virg<strong>in</strong>ia, described her as a steamboat with three decks, no mast, a wheel stern,<br />
and a cab<strong>in</strong> on deck (Figure 18). The <strong>of</strong>ficial measurements given were 155 feet long, 35<br />
feet wide, and 4 feet deep (draw<strong>in</strong>g four feet <strong>of</strong> water). She was rated as be<strong>in</strong>g able to carry<br />
479.22 tons <strong>of</strong> cargo (U.S. Bureau <strong>of</strong> Mar<strong>in</strong>e Inspection and Navigation). Although tonnage<br />
rat<strong>in</strong>gs were never very accurate, this would give her a capacity <strong>of</strong> over 2,000 bales <strong>of</strong> cotton<br />
at 400 pounds per bale.
A b<br />
History 91<br />
Figure 18. Portions <strong>of</strong> certificates <strong>of</strong> registry for <strong>the</strong> Lotawanna. a. temporary<br />
certificate issued at Wheel<strong>in</strong>g, West Virg<strong>in</strong>ia, October 27, 1867; b.<br />
permanent certificate issued at New Orleans, Louisiana, November<br />
27, 1867 (U.S. Bureau <strong>of</strong> Mar<strong>in</strong>e Inspection and Navigation) (AAS<br />
neg. 803787, 803789).
92 Watk<strong>in</strong>s<br />
Br<strong>in</strong>ker <strong>the</strong>n took her to New Orleans where her Permanent Enrollment Certificate<br />
was issued on November 27, 1867, and she began trad<strong>in</strong>g on <strong>the</strong> Ouachita River (U.S. Bureau<br />
<strong>of</strong> Mar<strong>in</strong>e Inspection and Navigation). In an advertisement <strong>the</strong> follow<strong>in</strong>g year Br<strong>in</strong>ker<br />
described <strong>the</strong> Lotawanna as a “splendid new and light draught Packet, built expressly for <strong>the</strong><br />
(Camden-New Orleans) trade,” and promised punctuality and comfortable accommodations<br />
(Sifford n.d.:Vol. 1:50; Figure 19).<br />
The Lotawanna had several owners. In 1869 Br<strong>in</strong>ker sold her to Phillip A. Work, his<br />
clerk, and Joseph Morgan. They ran her for just over a year before sell<strong>in</strong>g her to H. G. McComas<br />
<strong>in</strong> January 1871. In May 1871 McComas mortgaged <strong>the</strong> boat for $13,613.53, which he apparently<br />
was not able to pay, because <strong>in</strong> August she became <strong>the</strong> property <strong>of</strong> William O’Hern,<br />
Michael Shelly, and Arthur Po<strong>in</strong>cy (with E. R. Hart jo<strong>in</strong><strong>in</strong>g <strong>the</strong> partnership <strong>in</strong> September). In<br />
June 1872 Po<strong>in</strong>cy, O’Hern, and Shelly bought out Hart’s <strong>in</strong>terest before sell<strong>in</strong>g <strong>the</strong> Lotawanna<br />
to F. A. Blanks, <strong>of</strong> <strong>the</strong> Blanks Steamboat L<strong>in</strong>e, <strong>in</strong> December (U.S. Bureau <strong>of</strong> Mar<strong>in</strong>e Inspection<br />
and Navigation).<br />
She was <strong>the</strong> largest sternwheeler on <strong>the</strong> river <strong>in</strong> <strong>the</strong> early 1870s. Dur<strong>in</strong>g <strong>the</strong> 1872-1873<br />
season she made seven trips to Camden from New Orleans, exceeded only by <strong>the</strong> Mayflower<br />
and Sab<strong>in</strong>e (eight trips each) and <strong>the</strong> Bertha Bruner (n<strong>in</strong>e trips) (U.S. Congress 1874:7). She<br />
also did some trad<strong>in</strong>g on <strong>the</strong> Yazoo River (Owens 1979:169). In 1874, as one <strong>of</strong> <strong>the</strong> stars <strong>of</strong><br />
<strong>the</strong> Blanks L<strong>in</strong>e, <strong>the</strong> Ouachita Belle be<strong>in</strong>g <strong>the</strong> o<strong>the</strong>r, <strong>the</strong> Lotawanna was advertised as a passenger<br />
packet leav<strong>in</strong>g New Orleans once a week and connect<strong>in</strong>g with <strong>the</strong> tributary steamers<br />
Bertha Bruner to Farmerville, Ora on Bayou Bartholomew, Stella Block on <strong>the</strong> Sal<strong>in</strong>e River, and<br />
<strong>the</strong> Willie from Monroe to Camden <strong>in</strong> low water (Ouachita Telegraph May 1, 1874).<br />
In May 1874 <strong>the</strong> Lotawanna was headed up to Camden with 25-30 passengers. Because<br />
<strong>the</strong> river was too hazardous to navigate after dark, she tied up for <strong>the</strong> night just below<br />
<strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g. Dur<strong>in</strong>g <strong>the</strong> night <strong>the</strong> boat caught fire, and was destroyed, strand<strong>in</strong>g<br />
<strong>the</strong> people on a small bar where <strong>the</strong>y had tied up. There was a young man <strong>in</strong> <strong>the</strong> crew<br />
who was learn<strong>in</strong>g to be a pilot, and <strong>the</strong> next morn<strong>in</strong>g he volunteered to swim ashore and get<br />
help. He did so, f<strong>in</strong>d<strong>in</strong>g wagon tracks and follow<strong>in</strong>g <strong>the</strong>m to <strong>the</strong> home <strong>of</strong> Arnold Nash (near<br />
present-day Dollar Junction). Nash and a neighbor, Jim Dollar, rode through <strong>the</strong> neighborhood<br />
ga<strong>the</strong>r<strong>in</strong>g bedd<strong>in</strong>g and o<strong>the</strong>r camp<strong>in</strong>g equipment which <strong>the</strong>y took out to <strong>the</strong> people <strong>in</strong><br />
a flatboat. The passengers were rescued when a steamboat came down from Camden to pick<br />
<strong>the</strong>m up (Johnson 1979).<br />
The wreck <strong>of</strong> <strong>the</strong> Lotawanna was a hazard to navigation, especially dur<strong>in</strong>g low water,<br />
so Capta<strong>in</strong> Just<strong>in</strong> Straszer <strong>of</strong> <strong>the</strong> U.S. Snagboat O. G. Wagner was ordered to remove it. He<br />
visited <strong>the</strong> wreck <strong>the</strong> week <strong>of</strong> June 14, 1875 but found <strong>the</strong> water too high to make work practical.<br />
The follow<strong>in</strong>g week he had better luck and was able to salvage <strong>the</strong> follow<strong>in</strong>g articles:
Figure 19. Advertisement from unknown newspaper source<br />
(about 1868) announc<strong>in</strong>g Lotawanna’s Camden-<br />
New Orleans trade (after Sifford n.d.; Vol. 1, p. 50)<br />
(AAS neg. 803780).<br />
History 93
94 Watk<strong>in</strong>s<br />
1. one shaft and cranks—<strong>the</strong> axle which supported <strong>the</strong> sternwheel and <strong>the</strong><br />
mechanisms at each end for attach<strong>in</strong>g <strong>the</strong> shaft to <strong>the</strong> pitmans. The shaft was<br />
wrought iron and had recently been replaced at <strong>the</strong> Leeds foundry <strong>in</strong> New<br />
Orleans for $1600.<br />
2. two pitmans—<strong>the</strong> long rods connect<strong>in</strong>g <strong>the</strong> eng<strong>in</strong>e to <strong>the</strong> sternwheel and, with<br />
<strong>the</strong> cranks, chang<strong>in</strong>g <strong>the</strong> l<strong>in</strong>ear motion <strong>of</strong> <strong>the</strong> cyl<strong>in</strong>der <strong>in</strong>to circular motion;<br />
usually made <strong>of</strong> wood with metal ends.<br />
3. one eng<strong>in</strong>e—<strong>the</strong> cyl<strong>in</strong>der which turned steam pressure <strong>in</strong>to l<strong>in</strong>ear motion, a<br />
sternwheeler would have two eng<strong>in</strong>es, one connected to each end <strong>of</strong> <strong>the</strong> wheel<br />
by a pitman.<br />
4. doctor—a small auxiliary eng<strong>in</strong>e run by steam for operat<strong>in</strong>g w<strong>in</strong>ches and<br />
hoists.<br />
5. capstan gear<strong>in</strong>g—a w<strong>in</strong>ch.<br />
6. gaspipes—pipes used to run steam from <strong>the</strong> boilers to <strong>the</strong> eng<strong>in</strong>es.<br />
7. hog cha<strong>in</strong>s—truss rods with turnbuckles runn<strong>in</strong>g from bow to stern up over<br />
<strong>the</strong> top <strong>of</strong> <strong>the</strong> boat to keep <strong>the</strong> hull from break<strong>in</strong>g from <strong>the</strong> weight <strong>of</strong> <strong>the</strong> boilers<br />
<strong>in</strong> <strong>the</strong> bow and <strong>the</strong> eng<strong>in</strong>es and wheel at <strong>the</strong> stern.<br />
8. smaller articles<br />
Straszer took <strong>the</strong> salvaged materials to Monroe to be sold at auction. He estimated <strong>the</strong> sale<br />
would br<strong>in</strong>g $600 to $700 if all articles sold for scrap (at $.02/pound for wrought iron and<br />
$.01/pound for cast iron) (Straszer to Benyaurd, June 20, June 25, August 4; U.S. Army Corps<br />
<strong>of</strong> Eng<strong>in</strong>eers 1875).<br />
This salvage operation did not clear <strong>the</strong> Lotawanna from <strong>the</strong> channel because Straszer<br />
was only able to work on <strong>the</strong> parts <strong>of</strong> <strong>the</strong> wreck that were out <strong>of</strong> <strong>the</strong> water or just below<br />
<strong>the</strong> surface. He returned to <strong>the</strong> wreck later <strong>in</strong> <strong>the</strong> year and blew up <strong>the</strong> forward part <strong>of</strong> <strong>the</strong><br />
hull with gunpowder (U.S. Congress 1876:592). However, <strong>the</strong> wreck rema<strong>in</strong>ed enough <strong>of</strong><br />
an obstruction that it was marked on river maps as late as 1896, and <strong>the</strong> deck<strong>in</strong>g <strong>of</strong> <strong>the</strong> hull<br />
could still be seen dur<strong>in</strong>g low water as late as about 1904 (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers<br />
1896:Sheet 14; Johnson 1979).<br />
luMber<strong>in</strong>g And rAilroAds<br />
In <strong>the</strong> 55 years that passed between <strong>the</strong> first survey and <strong>the</strong> completion <strong>of</strong> <strong>the</strong> lock and<br />
dam system, great changes had taken place <strong>in</strong> <strong>the</strong> economy <strong>of</strong> <strong>the</strong> Ouachita River Valley <strong>in</strong>
History 95<br />
south <strong>Arkansas</strong> <strong>in</strong> general, and <strong>in</strong> <strong>the</strong> project area <strong>in</strong> particular. Even before <strong>the</strong> Civil War,<br />
attempts had been made to develop <strong>the</strong> area’s timber and m<strong>in</strong>eral resources. Timber was<br />
rafted down <strong>the</strong> river as early as 1826; lignite was m<strong>in</strong>ed from <strong>the</strong> bed <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River<br />
<strong>in</strong> <strong>the</strong> 1850s; and axle grease was made from an oil seep at Wilm<strong>in</strong>gton (E<strong>the</strong>ridge 1959:25;<br />
Owen 1860: 139; Spencer 1957). However, <strong>the</strong> lack <strong>of</strong> reliable transportation, <strong>the</strong> distance to<br />
markets, and <strong>the</strong> scarcity <strong>of</strong> labor caused <strong>the</strong>se enterprises to fail.<br />
In <strong>the</strong> 1870s <strong>the</strong> economy was dependent on cotton, while timber<strong>in</strong>g and sawmill<strong>in</strong>g<br />
were local activities by farmers who owned boilers for g<strong>in</strong>s or grist mills and wanted<br />
to make a little extra money by keep<strong>in</strong>g <strong>the</strong>ir mills runn<strong>in</strong>g all year long (Curry 1960:115).<br />
The first group to ship f<strong>in</strong>ished wood products from <strong>the</strong> Felsenthal area were <strong>the</strong> “Slavonians.”<br />
Orig<strong>in</strong>ally from <strong>the</strong> mounta<strong>in</strong>s <strong>of</strong> Yugoslavia, <strong>the</strong>se men came to <strong>the</strong> United States <strong>in</strong><br />
<strong>the</strong> early 1880s to work <strong>in</strong> <strong>the</strong> steel mills. Every w<strong>in</strong>ter a group <strong>of</strong> 8 to 25 men would come<br />
to south <strong>Arkansas</strong> to cut pipe staves and claret staves. The men lived <strong>in</strong> shelters <strong>of</strong> red oak<br />
(which was free) and cut <strong>the</strong> staves from white oak (which <strong>the</strong>y bought). The rough staves<br />
were cut with a broadaxe and trimmed with a draw knife. Dur<strong>in</strong>g one season <strong>the</strong> men<br />
could cut 90,000 staves, which <strong>the</strong>y floated to New Orleans on barges <strong>in</strong> <strong>the</strong> spr<strong>in</strong>g (Curry<br />
1960:1<strong>17</strong>-118). Their venture was successful enough that o<strong>the</strong>rs followed <strong>the</strong>ir example. In<br />
1894 <strong>the</strong> H. M. Townsend left Camden with 50,000 staves and expected to pick up a barge<br />
<strong>of</strong> staves at <strong>the</strong> Sal<strong>in</strong>e River (Sifford n.d. Vol. 1). Stave cutt<strong>in</strong>g cont<strong>in</strong>ued <strong>in</strong>to <strong>the</strong> 1920s, at<br />
which time <strong>the</strong> standard price was $.03 per stave (Johnson 1979).<br />
Just as <strong>the</strong> new Ouachita River navigation project was gett<strong>in</strong>g underway, <strong>the</strong> economic<br />
focus <strong>of</strong> <strong>the</strong> area shifted away from <strong>the</strong> river. The St. Louis, Iron Mounta<strong>in</strong>, and Sou<strong>the</strong>rn<br />
Railroad (now <strong>the</strong> Missouri and Pacific) reached El Dorado <strong>in</strong> 1891, and extended its tracks<br />
to Bastrop, Louisiana, <strong>in</strong> 1904, cross<strong>in</strong>g <strong>the</strong> Ouachita River near <strong>the</strong> new town <strong>of</strong> Felsenthal.<br />
Its com<strong>in</strong>g made practical <strong>the</strong> build<strong>in</strong>g <strong>of</strong> logg<strong>in</strong>g railroads which opened <strong>the</strong> entire area to<br />
extensive timber<strong>in</strong>g. Logg<strong>in</strong>g began at <strong>the</strong> railroad and advanced to new areas by build<strong>in</strong>g<br />
tramways (Curry 1960:118; see also Watk<strong>in</strong>s 1981:54-56).<br />
Between 1900 and 1904, <strong>17</strong> sawmills and lumber companies were chartered <strong>in</strong> south<br />
<strong>Arkansas</strong> (Crockett 1904:359-399). The Crossett Lumber Company was <strong>the</strong> largest on <strong>the</strong><br />
east side <strong>of</strong> <strong>the</strong> river. It established Crossett as its company town <strong>in</strong> 1902 and shipped out<br />
its products over <strong>the</strong> Rock Island Railroad (Buckner 1979). On <strong>the</strong> west side <strong>of</strong> <strong>the</strong> river <strong>the</strong><br />
largest company was <strong>the</strong> Union Sawmill which established Huttig <strong>in</strong> 1902. This company<br />
rafted logs down <strong>the</strong> river <strong>in</strong> w<strong>in</strong>ter (7 million logs <strong>in</strong> 1904), but relied on <strong>the</strong> railroad for<br />
br<strong>in</strong>g<strong>in</strong>g <strong>the</strong> logs <strong>in</strong>to <strong>the</strong> mill and for shipp<strong>in</strong>g out f<strong>in</strong>ished products (Felsenthal Press March<br />
4, 1905).<br />
Dur<strong>in</strong>g this early twentieth century period timber<strong>in</strong>g <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> concentrated on<br />
bald cypress, oaks, and o<strong>the</strong>r hardwoods, while both hardwoods and p<strong>in</strong>e were cut <strong>in</strong> up-
96 Watk<strong>in</strong>s<br />
land areas. Extensive upland clear<strong>in</strong>g and p<strong>in</strong>e plant<strong>in</strong>g began <strong>in</strong> <strong>the</strong> 1920s and 1930s (Buckner<br />
1979; Reynolds 1980). Trees were cut by two-man teams with large bucksaws and skidded<br />
by mules or oxen teams to tramways or to <strong>the</strong> riverbank for transportation to mills at<br />
Crossett and Huttig.<br />
The shift <strong>in</strong> <strong>the</strong> economy from cotton to timber and from river to railroad was reflected<br />
<strong>in</strong> <strong>the</strong> chang<strong>in</strong>g status <strong>of</strong> <strong>the</strong> towns. Communities like Marie Sal<strong>in</strong>e Land<strong>in</strong>g and Champagnolle<br />
which were dependent on river trade died from lack <strong>of</strong> bus<strong>in</strong>ess when <strong>the</strong> railroad<br />
passed <strong>the</strong>m by (E<strong>the</strong>ridge 1959:137; Cordell 1951:43-44). New communities (Bold<strong>in</strong>g,<br />
Strong, Dollar Junction) grew up along <strong>the</strong> railroad to replace <strong>the</strong>m (Cordell 1947:284).<br />
<strong>the</strong> felsenthAl coMMunity<br />
The town <strong>of</strong> Felsenthal is <strong>the</strong> best example <strong>of</strong> changes that took place <strong>in</strong> and around<br />
<strong>the</strong> project area. There had been a community associated with Lake Land<strong>in</strong>g, where <strong>Grand</strong><br />
<strong>Marais</strong> Lake entered <strong>the</strong> Ouachita River, at least as early as 1899 (U.S. Congress 1900:19).<br />
Adolph Felsenthal and his bro<strong>the</strong>r Ike bought property and laid out <strong>the</strong> town <strong>of</strong> Felsenthal<br />
as a manufactur<strong>in</strong>g and commercial center for nearby Huttig, a company town. The first<br />
public sale <strong>of</strong> lots was held on April 27, 1904. At that time <strong>the</strong> town already had <strong>the</strong> Lake<br />
Lumber Company sawmill, a stave factory, and a brick factory. The day <strong>of</strong> <strong>the</strong> sale was a big<br />
celebration with excursion tra<strong>in</strong>s com<strong>in</strong>g from Gurdon, Camden, and Monroe. Two brass<br />
bands and a speech by Governor Jeff Davis, as well as races, contests, and “o<strong>the</strong>r unique<br />
amusements” were provided for enterta<strong>in</strong>ment; and an auctioneer was brought <strong>in</strong> from<br />
Atlanta, Georgia, to make sure <strong>the</strong> sale was properly conducted. The big sell<strong>in</strong>g po<strong>in</strong>ts for<br />
<strong>the</strong> new “Magic City <strong>of</strong> South <strong>Arkansas</strong>” were that it was at <strong>the</strong> junction <strong>of</strong> four railroads<br />
(<strong>the</strong> Little Rock and Monroe Railroad and three branches <strong>of</strong> <strong>the</strong> St. Louis, Iron Mounta<strong>in</strong>,<br />
and Sou<strong>the</strong>rn) and that <strong>the</strong> government had appropriated funds to build a lock on <strong>the</strong> river<br />
nearby (<strong>Arkansas</strong> Gazette April 24, 1904).<br />
The sale apparently went well, because Felsenthal prospered. In its prime it had a<br />
bank, a newspaper, several hotels, grocery stores, and drug stores. The brick factory was<br />
jo<strong>in</strong>ed by a bottl<strong>in</strong>g plant and a gar factory that ground up rough fish for dog food (Johnson<br />
1979). The town reached its height <strong>in</strong> <strong>the</strong> 1920s while <strong>the</strong> timber<strong>in</strong>g <strong>in</strong>dustry was at its peak<br />
and when <strong>the</strong> completion <strong>of</strong> <strong>the</strong> navigation project brought <strong>in</strong>creased river trade.<br />
In <strong>the</strong> meantime events just west <strong>of</strong> El Dorado signaled ano<strong>the</strong>r shift <strong>in</strong> economic emphasis.<br />
On January 10, 1921 an oil well owned by Dr. Samuel T. Busey, came <strong>in</strong> as a gusher<br />
and started <strong>the</strong> south <strong>Arkansas</strong> oil boom (Green 1954:51-53). As <strong>the</strong> hopeful and <strong>the</strong> curious<br />
flocked to El Dorado (which grew from 4,000 to 32,000), areas at <strong>the</strong> eastern end <strong>of</strong> <strong>the</strong><br />
county, where most <strong>of</strong> <strong>the</strong> timber had already been logged out, lost population.
History 97<br />
Trade on <strong>the</strong> Ouachita River cont<strong>in</strong>ued to grow. By <strong>the</strong> 1930s sternwheel steamboats<br />
had been replaced by gasol<strong>in</strong>e or diesel powered towboats capable <strong>of</strong> push<strong>in</strong>g several<br />
large barges. With <strong>the</strong> start <strong>of</strong> World War II, however, <strong>the</strong> river trade came to a temporary<br />
halt—partly because <strong>the</strong> barges were needed elsewhere, partly because <strong>the</strong> channel was not<br />
ma<strong>in</strong>ta<strong>in</strong>ed (Thatcher 1970:16). After <strong>the</strong> war bus<strong>in</strong>essmen all along <strong>the</strong> river were <strong>in</strong>terested<br />
<strong>in</strong> reestablish<strong>in</strong>g this trade, and this <strong>in</strong>terest, through <strong>the</strong> work <strong>of</strong> <strong>the</strong> Ouachita River Valley<br />
Association, has brought about <strong>the</strong> current 9-Foot Navigation Project (Mills 1978:165-169).<br />
conclusions<br />
The history <strong>of</strong> <strong>the</strong> Felsenthal Project area is closely tied to <strong>the</strong> history <strong>of</strong> trade and<br />
travel on <strong>the</strong> Ouachita River. Beg<strong>in</strong>n<strong>in</strong>g as a highway for hunters, trappers, and traders,<br />
<strong>the</strong> river was <strong>the</strong> only practical means <strong>of</strong> shipp<strong>in</strong>g agricultural products and merchandise<br />
throughout <strong>the</strong> n<strong>in</strong>eteenth century. The importance <strong>of</strong> this trade was reflected <strong>in</strong> <strong>the</strong> number<br />
<strong>of</strong> communities and land<strong>in</strong>gs on <strong>the</strong> river which served as shipp<strong>in</strong>g po<strong>in</strong>ts for <strong>the</strong> <strong>in</strong>land<br />
farms and plantations.<br />
The build<strong>in</strong>g <strong>of</strong> railroads <strong>in</strong>to <strong>the</strong> area early <strong>in</strong> <strong>the</strong> twentieth century brought a change<br />
<strong>of</strong> economic focus. Shipp<strong>in</strong>g and commerce shifted <strong>in</strong>land to <strong>the</strong> railroad, and new towns<br />
took <strong>the</strong> place <strong>of</strong> land<strong>in</strong>gs which went out <strong>of</strong> bus<strong>in</strong>ess as <strong>the</strong> volume <strong>of</strong> trade on <strong>the</strong> river<br />
decl<strong>in</strong>ed. At <strong>the</strong> same time, <strong>the</strong> railroads opened <strong>the</strong> area to lumber<strong>in</strong>g, especially <strong>in</strong> bottomlands<br />
which had never been cleared because <strong>the</strong>y were not suitable for farm<strong>in</strong>g.<br />
Permanent improvements <strong>in</strong> navigation kept <strong>the</strong> river open for commerce, but <strong>the</strong> end<br />
<strong>of</strong> <strong>in</strong>tensive lumber<strong>in</strong>g and <strong>the</strong> oil boom near El Dorado caused <strong>the</strong> area to lose population.<br />
Today towboats pass through <strong>the</strong> project area on <strong>the</strong>ir way to Camden, but do not stop, and<br />
<strong>the</strong> bottomlands are once aga<strong>in</strong> <strong>the</strong> preserve <strong>of</strong> hunters and fishermen.
Chapter 5<br />
VAriAbility AMong floodplA<strong>in</strong> sites<br />
River levels <strong>in</strong> <strong>the</strong> Felsenthal Project area rema<strong>in</strong>ed above <strong>the</strong> upper project contour<br />
(65 feet elevation) for <strong>the</strong> first half <strong>of</strong> 1979, dropp<strong>in</strong>g to lock stage (62.2 feet) f<strong>in</strong>ally <strong>in</strong> midsummer.<br />
In August we assembled our field personnel and a large quantity <strong>of</strong> equipment and<br />
supplies at <strong>the</strong> project headquarters <strong>in</strong> Felsenthal, conducted a three-day tra<strong>in</strong><strong>in</strong>g session,<br />
and sent out survey teams <strong>in</strong>to <strong>the</strong> field on August 20. The <strong>in</strong>tensive survey stage <strong>of</strong> <strong>the</strong><br />
Felsenthal Project, which deployed three survey teams simultaneously, <strong>the</strong>n proceeded essentially<br />
without <strong>in</strong>terruption for seven weeks until <strong>the</strong> test<strong>in</strong>g stage was <strong>in</strong>itiated on October<br />
6. After this date, a s<strong>in</strong>gle small survey team cont<strong>in</strong>ued to operate on an occasional basis,<br />
concurrent with test<strong>in</strong>g <strong>of</strong> selected sites (Chapter 6), until ris<strong>in</strong>g river levels term<strong>in</strong>ated all<br />
field activity <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> on November 24. A few site-specific surveys on higher terra<strong>in</strong><br />
were conducted <strong>in</strong> <strong>the</strong> first week <strong>of</strong> December, but all aspects <strong>of</strong> fieldwork had ended<br />
by December 7. The term<strong>in</strong>ation <strong>of</strong> work <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> <strong>in</strong> late November was about three<br />
weeks earlier than <strong>in</strong>itially projected, affect<strong>in</strong>g <strong>the</strong> objectives <strong>of</strong> survey only <strong>in</strong> m<strong>in</strong>or respects,<br />
but curtail<strong>in</strong>g test<strong>in</strong>g more markedly.<br />
In this chapter we present both <strong>the</strong> methods and results <strong>of</strong> <strong>in</strong>tensive survey, focus<strong>in</strong>g<br />
on variability <strong>in</strong> <strong>the</strong> physical context (and to a lesser extent cultural content) <strong>of</strong> floodpla<strong>in</strong><br />
sites deduced from site survey data and subsequent analyses <strong>of</strong> <strong>the</strong>se data. Dur<strong>in</strong>g <strong>the</strong><br />
1979 Felsenthal Project, 126 previously unknown archeological sites were recorded and 18<br />
sites known from prior work were revisited. The <strong>Arkansas</strong> Archeological Survey site forms<br />
generated by this survey activity are volum<strong>in</strong>ous, averag<strong>in</strong>g seven pages for each <strong>of</strong> <strong>the</strong> 144<br />
sites, or more than 1,000 pages total. A limited selection <strong>of</strong> <strong>the</strong>se site descriptive and analytical<br />
data are presented as two tables (Prehistoric and Historic Sites) <strong>in</strong> Appendix A. For<br />
<strong>the</strong> purposes <strong>of</strong> data presentation, comparison, and generalization we have organized <strong>the</strong><br />
survey results <strong>in</strong> this chapter and <strong>in</strong> Appendix A as a series <strong>of</strong> eight site units as follows:<br />
Unit 1: Ouachita River above Mile 254 (18 sites)<br />
Unit 2: Lower Sal<strong>in</strong>e River (40 sites)<br />
Unit 3: Lower Eagle Creek (8 sites)<br />
Unit 4: Ouachita River below Mile 254 (20 sites)<br />
Unit 5: Lower Lapile Creek (8 sites)<br />
Unit 6: Oxbow Lakes and Backswamp (11 sites)<br />
Unit 7: Pleistocene Terraces and Islands (27 sites)
100 Hemm<strong>in</strong>gs<br />
Unit 8: Historic Sites (15 sites, 3 <strong>of</strong> which have prehistoric components<br />
listed <strong>in</strong> o<strong>the</strong>r site units above)<br />
These units are not survey subareas, although <strong>the</strong>y reflect <strong>the</strong> differences among various<br />
methods employed (see Survey Strategy and Tactics below), but were devised after all<br />
analyses <strong>of</strong> site data were completed. They do represent geomorphic units which vary by<br />
dra<strong>in</strong>age area with<strong>in</strong> <strong>the</strong> Recent floodpla<strong>in</strong> (Units 1-6) or occur on older landforms above<br />
<strong>the</strong> floodpla<strong>in</strong> (Unit 7). The division <strong>of</strong> <strong>the</strong> Ouachita River at mile 254 is not an arbitrary one,<br />
but represents <strong>the</strong> approximate present location <strong>of</strong> <strong>the</strong> mouth <strong>of</strong> Sal<strong>in</strong>e River. Historic sites<br />
have been arbitrarily grouped as Unit 8, but all occur on or below <strong>the</strong> Recent floodpla<strong>in</strong>. As<br />
an additional organiz<strong>in</strong>g device, sites and site data are presented <strong>in</strong> an order from upstream<br />
to downstream location with<strong>in</strong> each <strong>of</strong> <strong>the</strong> eight site units segregated for discussion.<br />
The environmental and adm<strong>in</strong>istrative constra<strong>in</strong>ts on performance <strong>of</strong> archeological<br />
survey <strong>in</strong> <strong>the</strong> Felsenthal Project area were discussed <strong>in</strong> detail <strong>in</strong> Chapter 1. In this chapter<br />
we def<strong>in</strong>e two major subareas divided by <strong>the</strong> Ouachita River bridge at U.S. Highway 82:<br />
<strong>the</strong> Ouachita-Sal<strong>in</strong>e Channels subarea on <strong>the</strong> north and <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland subarea<br />
on <strong>the</strong> south (Figure 20). In <strong>the</strong> nor<strong>the</strong>rnmost subarea <strong>the</strong> upper project contour (65 feet) is<br />
for <strong>the</strong> most part restricted to <strong>the</strong> bankl<strong>in</strong>es <strong>of</strong> <strong>the</strong> rivers and <strong>the</strong>ir tributaries, and even <strong>the</strong><br />
lowest floodpla<strong>in</strong> surfaces (<strong>in</strong> backswamp) lie above this elevation. Our survey efforts necessarily<br />
employed boat transportation <strong>in</strong> this subarea. The sou<strong>the</strong>rn-most subarea, by contrast,<br />
<strong>in</strong>cludes 9330 acres (3776 ha) <strong>of</strong> floodpla<strong>in</strong> terra<strong>in</strong> below 65 feet and above lock stage at 62.2<br />
feet, most <strong>of</strong> this terra<strong>in</strong> represent<strong>in</strong>g backswamp. In this <strong>Grand</strong> <strong>Marais</strong> Lowland subarea<br />
we have employed boat, vehicle, and pedestrian tactics <strong>in</strong> a variety <strong>of</strong> ways. For both subareas<br />
<strong>the</strong> environmental constra<strong>in</strong>ts on mobility <strong>in</strong>clude <strong>the</strong> follow<strong>in</strong>g:<br />
1. There are no permanent or improved roads or boat land<strong>in</strong>gs <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong><br />
except at <strong>the</strong> U.S. Highway 82 bridge and at <strong>the</strong> new lock construction site.<br />
2. The floodpla<strong>in</strong> surface is hardwood and swamp forest complexly <strong>in</strong>terspersed<br />
with river channels, creeks, sloughs, and open lakes.<br />
3. There are no agricultural clear<strong>in</strong>gs, fencel<strong>in</strong>es, or artificial levees <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>.<br />
4. The modern cultural features that do occupy <strong>the</strong> floodpla<strong>in</strong> are widely scattered<br />
and ord<strong>in</strong>arily occur on terra<strong>in</strong> above 65 feet elevation (4-wheel drive<br />
tracks, fishcamps and deercamps, tramways, a pipel<strong>in</strong>e right-<strong>of</strong>-way, an abandoned<br />
railroad cross<strong>in</strong>g, and <strong>the</strong> Highway 82 cross<strong>in</strong>g).
Variability Among Floodpla<strong>in</strong> Sites 101<br />
The Felsenthal Project area as a whole, <strong>the</strong>refore, is an extensive, forested overflow bottom<br />
where access to floodpla<strong>in</strong> terra<strong>in</strong> is prohibited, or at least sharply restricted, by high water<br />
and wet wea<strong>the</strong>r, and where <strong>the</strong> logistical basis for archeological survey must be specially<br />
adapted to wetland conditions.<br />
Table 5 summarizes <strong>the</strong> pr<strong>in</strong>cipal items <strong>of</strong> power equipment and accessories employed<br />
by our survey teams <strong>in</strong> various modes <strong>of</strong> survey described below. Dur<strong>in</strong>g <strong>the</strong> four months <strong>of</strong><br />
fieldwork some items <strong>of</strong> power equipment were operated more or less cont<strong>in</strong>uously under<br />
adverse conditions. For <strong>the</strong> purpose <strong>of</strong> rout<strong>in</strong>e ma<strong>in</strong>tenance and repair we had technical<br />
manuals and an assortment <strong>of</strong> tools and spare parts. In <strong>the</strong> event <strong>of</strong> breakdowns we were<br />
usually able to substitute a backup item and keep <strong>the</strong> survey teams <strong>in</strong> <strong>the</strong> field. Fortunately,<br />
also, arrangements were made for almost overnight mechanical repair service for trucks,<br />
all-terra<strong>in</strong>-vehicles (ATVs), and outboard motors, and <strong>the</strong>se arrangements were called upon<br />
frequently. Almost all <strong>of</strong> <strong>the</strong> power equipment and some <strong>of</strong> <strong>the</strong> accessory items listed <strong>in</strong><br />
Table 5 had failures <strong>of</strong> some sort, but could be put back <strong>in</strong>to service <strong>in</strong> a day or two.<br />
The diverse k<strong>in</strong>ds <strong>of</strong> hand operated test<strong>in</strong>g equipment, record<strong>in</strong>g equipment, miscellaneous<br />
items and supplies used by <strong>the</strong> Felsenthal Project are largely those that would be<br />
employed also <strong>in</strong> large scale upland survey and need not be tabulated here. We should,<br />
however, emphasize <strong>the</strong> necessity for special protective or safety equipment and for waterpro<strong>of</strong><br />
cloth<strong>in</strong>g, footwear, and conta<strong>in</strong>ers or cover<strong>in</strong>gs <strong>of</strong> various k<strong>in</strong>ds for <strong>the</strong> purpose<br />
<strong>of</strong> survey <strong>in</strong> a remote wetland. Our survey teams, for example, were provided with snake<br />
bite and first aid kits, snake legg<strong>in</strong>gs, hard hats, blaze orange and flotation vests, machetes,<br />
compasses, canteens, daypacks, and voyager bags, and <strong>the</strong>se items were used extensively<br />
throughout <strong>the</strong> floodpla<strong>in</strong> and river channel survey effort.<br />
The three-day orientation and tra<strong>in</strong><strong>in</strong>g session at <strong>the</strong> outset <strong>of</strong> <strong>the</strong> survey stage was<br />
an important mechanism for <strong>in</strong>troduc<strong>in</strong>g project <strong>in</strong>formation and tactics, outl<strong>in</strong><strong>in</strong>g safety<br />
responsibilities and measures, and demonstrat<strong>in</strong>g correct use <strong>of</strong> power equipment. Approximately<br />
one half <strong>of</strong> <strong>the</strong> field crew had never before entered a Sou<strong>the</strong>rn floodpla<strong>in</strong> forest. Survey<br />
teams had numerous close encounters with poisonous snakes and a few m<strong>in</strong>or accidents<br />
which did not lead to significant difficulties or loss <strong>of</strong> time.<br />
surVey strAtegy And tActics<br />
The strategy for cultural resources survey <strong>in</strong> <strong>the</strong> Felsenthal Navigation Pool was<br />
outl<strong>in</strong>ed <strong>in</strong> <strong>Arkansas</strong> Archeological Survey’s proposal (1978) and a subsequent research<br />
design (Chapter 1, Appendix D). Our general objective was to perform an “<strong>in</strong>tensive, on-<strong>the</strong>ground<br />
survey and test<strong>in</strong>g <strong>of</strong> [<strong>the</strong>] area sufficient to determ<strong>in</strong>e <strong>the</strong> number and extent <strong>of</strong> <strong>the</strong><br />
resources present, <strong>the</strong>ir cultural and scientific importance, and to estimate <strong>the</strong> time and cost
102 Hemm<strong>in</strong>gs<br />
Table 5. Power Equipment and Accessories Employed for Wetland Archeological Survey<br />
<strong>in</strong> <strong>the</strong> Felsenthal Project Area.<br />
Power Equipment No. <strong>of</strong> Items Accessory Equipment Pr<strong>in</strong>cipal Applications<br />
4WD Dodge trucks 2 w<strong>in</strong>ches, camper shells, mov<strong>in</strong>g personnel and<br />
tow cables, flatbed equipment to floodpla<strong>in</strong><br />
trailer, tool kits, first- stag<strong>in</strong>g areas<br />
aid kits<br />
All-Terra<strong>in</strong>-Vehicles 2 come-alongs, jerry cans, haul<strong>in</strong>g personnel and<br />
(Hustler and Coot) tool kits, first-aid kits equipment on floodpla<strong>in</strong><br />
transects<br />
Jon boats 2 flotation and safety gear, haul<strong>in</strong>g personnel and<br />
(16-foot and 12-foot) boat trailer, anchors, equipment on river channel<br />
moor<strong>in</strong>g l<strong>in</strong>es, and and tributary surveys<br />
paddles<br />
Outboard motors 4 auxiliary tanks, spare as above<br />
(25 HP and 7.5 HP) parts, tool kits<br />
Power augers 3 bits, extensions, jerry subsurface test<strong>in</strong>g on floodcans,<br />
ear protectors pla<strong>in</strong> transects<br />
Cha<strong>in</strong> saw 1 gas can, safety goggles, brush clear<strong>in</strong>g as needed<br />
ear protectors
Variability Among Floodpla<strong>in</strong> Sites 103<br />
for preserv<strong>in</strong>g, recover<strong>in</strong>g, or o<strong>the</strong>rwise mitigat<strong>in</strong>g adverse effects on <strong>the</strong>m” (33CFR Part<br />
305). The strategy adopted for this <strong>in</strong>tensive, on-<strong>the</strong>-ground survey is a conservative one,<br />
based strongly on <strong>the</strong> results <strong>of</strong> nearly a decade <strong>of</strong> difficult fieldwork and critical evaluation<br />
by <strong>Arkansas</strong> Archeological Survey <strong>in</strong> this portion <strong>of</strong> <strong>the</strong> Felsenthal region (Rol<strong>in</strong>gson<br />
1972; Lishka 1973; Raab 1976b; Stacy 1976; Schambach 1976, 1979; Rol<strong>in</strong>gson and Schambach<br />
1981). While much <strong>of</strong> this earlier work was restricted to elevated Pleistocene terraces, a brief<br />
floodpla<strong>in</strong> reconnaissance <strong>in</strong> 1976 and <strong>the</strong> sum total <strong>of</strong> work <strong>in</strong> <strong>the</strong> area <strong>in</strong>dicated that archeological<br />
survey <strong>in</strong> <strong>the</strong> overflow bottoms would contend with formidable environmental<br />
conditions. These conditions were expected to reduce both <strong>the</strong> mobility <strong>of</strong> survey teams and<br />
success <strong>of</strong> site discovery <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>, which is largely backswamp, but not on navigable<br />
river channels and tributaries. Our strategy adjusted to this expectation by employ<strong>in</strong>g<br />
successive stages <strong>of</strong> survey, which <strong>in</strong>tegrated newly obta<strong>in</strong>ed results, and by employ<strong>in</strong>g<br />
vary<strong>in</strong>g methods and <strong>in</strong>tensities <strong>of</strong> survey <strong>in</strong> each <strong>of</strong> <strong>the</strong> two project subareas.<br />
The <strong>Grand</strong> <strong>Marais</strong> Lowland subarea is an almost enclosed bas<strong>in</strong>, 5.8 x 6.4 km <strong>in</strong> maximum<br />
dimensions (Figure 20). Us<strong>in</strong>g <strong>the</strong> best available maps (USGS Felsenthal NW, NE,<br />
SW and SE 7.5 m<strong>in</strong>ute advance workpr<strong>in</strong>ts 1978) and an acreage grid overlay, <strong>the</strong> follow<strong>in</strong>g<br />
proportions <strong>of</strong> terra<strong>in</strong> and water were calculated:<br />
Area <strong>of</strong> all water surface at 62.2 feet MSL 976.9 ha (2414 acres or 15.1%)<br />
(lock stage)<br />
Area <strong>of</strong> terra<strong>in</strong> at or below 65 feet MSL 3775.8 ha (9330 acres or 58.5%)<br />
(survey area)<br />
Area <strong>of</strong> terra<strong>in</strong> above 65 feet MSL <strong>17</strong>05.0 ha (4213 acres or 26.4%)<br />
(above survey area)<br />
Total Area <strong>of</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland subarea 6457.7 ha (15,957 acres or 100.0%)<br />
The potential survey area <strong>of</strong> 3775.8 ha is, however, widely and <strong>in</strong>tricately <strong>in</strong>terspersed with<br />
bayous, lakes, and sloughs. Except where <strong>in</strong>terrupted by open water, this floodpla<strong>in</strong> was<br />
expected to be (and <strong>in</strong>deed was) 100% hardwood or swamp forest with essentially 100%<br />
ground cover <strong>of</strong> leaf litter. Our maps showed a maximum relief <strong>of</strong> about 8 feet (62-70 feet<br />
MSL) <strong>in</strong> this expanse <strong>of</strong> floodpla<strong>in</strong>, with few clearly def<strong>in</strong>able elevated features except<strong>in</strong>g<br />
an irregular, narrow, low natural levee zone adjo<strong>in</strong><strong>in</strong>g <strong>the</strong> modern river channel. We were<br />
concerned that habitable rises, ridges, or hummocks <strong>of</strong> little extent, not <strong>in</strong>dicated on maps<br />
with a 5-foot contour <strong>in</strong>terval, could be present, but <strong>in</strong> fact <strong>the</strong>re appear to be no such small<br />
scale features <strong>in</strong> this lowland subarea.<br />
Previously recorded archeological sites <strong>in</strong> this subarea <strong>in</strong>cluded three small prehistoric<br />
artifact scatters (3AS201 which was not relocated, 3UN123, and 3UN121) and one historic
104 Hemm<strong>in</strong>gs<br />
site (3UN122), all on <strong>the</strong> Ouachita River bankl<strong>in</strong>e (see Units 4 and 8 below). Very little was<br />
known about <strong>the</strong> size, depth, content, or age <strong>of</strong> <strong>the</strong>se sites. The survey tactics employed by<br />
us <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland subarea thus assumed all <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g conditions:<br />
1. no floodpla<strong>in</strong> zone or feature can be ruled out as a potential site location,<br />
2. most sites must be discovered by subsurface test<strong>in</strong>g, except on exposed<br />
bankl<strong>in</strong>es, and<br />
3. mobility and location f<strong>in</strong>d<strong>in</strong>g by survey teams will be very difficult.<br />
The last condition, <strong>in</strong> particular, required <strong>the</strong> choice <strong>of</strong> transect, as opposed to quadrat,<br />
methods <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>, and <strong>the</strong>se transects were comb<strong>in</strong>ed with extensive bankl<strong>in</strong>e surveys.<br />
The Ouachita-Sal<strong>in</strong>e Channels subarea contrasts with <strong>the</strong> previously described subarea<br />
<strong>in</strong> many respects, most importantly <strong>in</strong> <strong>the</strong> disposition <strong>of</strong> potential survey area. In this subarea,<br />
extend<strong>in</strong>g upstream from <strong>the</strong> U.S. Highway 82 bridge, terra<strong>in</strong> below <strong>the</strong> 65-foot contour<br />
is restricted to <strong>the</strong> bankl<strong>in</strong>es <strong>of</strong> <strong>the</strong> rivers and <strong>the</strong>ir tributaries, and to <strong>the</strong> marg<strong>in</strong>s <strong>of</strong> lakes,<br />
swamps, and sloughs. The extent <strong>of</strong> potential survey area is more realistically represented<br />
by l<strong>in</strong>eal ra<strong>the</strong>r than areal calculations (see River Bankl<strong>in</strong>e Surveys, and Tributary Bankl<strong>in</strong>e<br />
and Backswamp Surveys below). For this subarea we had prior knowledge <strong>of</strong> <strong>the</strong> relative<br />
stability <strong>of</strong> <strong>the</strong> river channels s<strong>in</strong>ce <strong>the</strong> early 1800s (Chapter 2), and a record <strong>of</strong> three small<br />
prehistoric artifact scatters (3BR33, 3UN87, 3UN93, <strong>the</strong> last two not relocated) and three<br />
historic sites (3BR32, 3UN88, 3UN92), aga<strong>in</strong> all located on <strong>the</strong> Ouachita River bankl<strong>in</strong>e (see<br />
Units 1, 4, and 8 below). For <strong>the</strong> prehistoric sites little was known about size, depth, content,<br />
or age. The survey tactics devised for <strong>the</strong> Ouachita-Sal<strong>in</strong>e Channels subarea thus assumed<br />
all <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g conditions:<br />
1. no reach <strong>of</strong> a river or tributary can be ruled out as a potential site location,<br />
2. most sites can be discovered by careful <strong>in</strong>spection <strong>of</strong> exposed bankl<strong>in</strong>es, and<br />
3. mobility and location f<strong>in</strong>d<strong>in</strong>g by boat survey teams will present no extraord<strong>in</strong>ary<br />
problems <strong>in</strong> <strong>the</strong> accessible waterways.<br />
Extensive bankl<strong>in</strong>e surveys, comb<strong>in</strong>ed with two judgmental transects <strong>in</strong> backswamp areas,<br />
were <strong>the</strong> methods chosen for use <strong>in</strong> this subarea. Bankl<strong>in</strong>e surveys are <strong>the</strong>mselves a variation<br />
<strong>of</strong> <strong>the</strong> transect method where <strong>the</strong> <strong>in</strong>tensity <strong>of</strong> survey depends pr<strong>in</strong>cipally on <strong>the</strong> degree<br />
and pattern <strong>of</strong> subsurface exposure.<br />
One <strong>of</strong> <strong>the</strong> fundamental f<strong>in</strong>d<strong>in</strong>gs <strong>of</strong> <strong>the</strong> 1979 Felsenthal Project was derived from <strong>the</strong><br />
nature and relationships <strong>of</strong> buried archeological sites. Our research design anticipated <strong>the</strong>
Variability Among Floodpla<strong>in</strong> Sites 105<br />
aggrad<strong>in</strong>g regime <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> and <strong>the</strong> need for subsurface test<strong>in</strong>g methods and evaluation<br />
<strong>of</strong> <strong>the</strong>se methods. Based on all available site data from <strong>the</strong> project area and analyses <strong>of</strong><br />
<strong>the</strong>se data, we can say with some confidence that all floodpla<strong>in</strong> archeological sites greater<br />
than 50 years <strong>in</strong> age are buried. The only exception, and it is a m<strong>in</strong>or one, is <strong>the</strong> occurrence<br />
<strong>of</strong> substantial above-ground structures <strong>in</strong> historic sites (e.g., structural earth mounds at<br />
3BR32 and 3UN92). The alluvial stratigraphic context <strong>of</strong> floodpla<strong>in</strong> sites is exam<strong>in</strong>ed fur<strong>the</strong>r<br />
<strong>in</strong> Chapters 6 and 7. In this chapter we evaluate <strong>the</strong> use <strong>of</strong> power auger and shovel test<br />
methods, and more importantly <strong>the</strong> method <strong>of</strong> bankl<strong>in</strong>e survey. Regrettably, we have not<br />
had <strong>the</strong> opportunity here to evaluate <strong>the</strong> use <strong>of</strong> backhoe trench<strong>in</strong>g, core drill<strong>in</strong>g, grad<strong>in</strong>g,<br />
plow<strong>in</strong>g, or o<strong>the</strong>r mechanical techniques <strong>of</strong> site location and exam<strong>in</strong>ation. These techniques<br />
are presently regarded as too logistically complex, too costly, or too destructive for employment<br />
<strong>in</strong> an overflow bottom environment (which is also a wildlife refuge). Moreover, <strong>in</strong><br />
<strong>the</strong> context <strong>of</strong> <strong>the</strong> Felsenthal Project area, only a th<strong>in</strong> slice <strong>of</strong> terra<strong>in</strong> is accessible to test<strong>in</strong>g,<br />
below 65 feet and above 62.2 feet MSL. Excavations <strong>of</strong> any significant depth would require<br />
pump<strong>in</strong>g if <strong>the</strong>y are to be entered and exam<strong>in</strong>ed <strong>in</strong> detail. We have <strong>the</strong>refore implemented<br />
relatively commonplace subsurface test<strong>in</strong>g methods, auger<strong>in</strong>g, shovel test<strong>in</strong>g, and bank<br />
<strong>in</strong>spection, which are dist<strong>in</strong>guished chiefly by mobility. We have no great or detailed knowledge<br />
<strong>of</strong> deep floodpla<strong>in</strong> and underwater sites, while <strong>the</strong>re is virtual certa<strong>in</strong>ty that <strong>the</strong>se are<br />
present (see Units 1 and 8 below).<br />
Random and Judgmental Transects<br />
In <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland subarea <strong>the</strong> <strong>in</strong>itial survey stage consisted <strong>of</strong> four, randomly<br />
placed, east-west transects (Figure 20). This subarea was stratified on <strong>the</strong> basis <strong>of</strong><br />
prelim<strong>in</strong>ary map observations which <strong>in</strong>dicated that <strong>the</strong> floodpla<strong>in</strong> varies more or less<br />
cont<strong>in</strong>uously <strong>in</strong> terra<strong>in</strong> and dra<strong>in</strong>age features along <strong>the</strong> downvalley direction (N-S) and<br />
also perpendicular to this direction (E-W). Essentially all <strong>of</strong> <strong>the</strong> potential survey area is<br />
backswamp environment with<strong>in</strong> <strong>the</strong> Recent floodpla<strong>in</strong>, characterized by a s<strong>in</strong>gle soil series<br />
and by bottomland hardwood forest, undifferentiated on available maps (U.S. Department<br />
<strong>of</strong> Agriculture 1968; Fleetwood 1969; Gill et al. 1979). Open lakes and sloughs comprise a<br />
greater proportion <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> area on <strong>the</strong> south (<strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> itself and Wildcat<br />
Lake). This hydrographic variation probably existed before impoundment by Felsenthal<br />
Lock and Dam <strong>in</strong> 1915, as <strong>in</strong>dicated by historical sources.<br />
For <strong>the</strong> purpose <strong>of</strong> stratification, <strong>the</strong> subarea was divided east-to-west by <strong>the</strong> Ouachita<br />
River and north-to-south by an imag<strong>in</strong>ary l<strong>in</strong>e across <strong>the</strong> valley at its midpo<strong>in</strong>t. Each <strong>of</strong> <strong>the</strong><br />
four strata def<strong>in</strong>ed <strong>in</strong> this manner was subdivided <strong>in</strong>to 250 potential east-west transects <strong>of</strong><br />
20-m width and variable length, and one transect was selected from each by use <strong>of</strong> a table <strong>of</strong><br />
random numbers. Transects 1 to 4 (Figure 20) are 4.9, 4.5, 2.8, and 3.0 km <strong>in</strong> length respectively,<br />
and appear to traverse all known variation <strong>in</strong> floodpla<strong>in</strong> terra<strong>in</strong>.
106 Hemm<strong>in</strong>gs<br />
Figure 20. The Felsenthal Project area, subareas, and 13 floodpla<strong>in</strong> transect locations.
Variability Among Floodpla<strong>in</strong> Sites 107<br />
The designated tactic for exploration <strong>of</strong> <strong>the</strong>se random transects was to send <strong>the</strong> survey<br />
teams from a po<strong>in</strong>t on higher terra<strong>in</strong> <strong>in</strong>to <strong>the</strong> floodpla<strong>in</strong> on a due-east or due-west compass<br />
course, flagg<strong>in</strong>g and test<strong>in</strong>g at 50 m <strong>in</strong>tervals by power auger (10 cm diameter by 1 m depth)<br />
below <strong>the</strong> 65-foot contour, except where prohibited by water. Two random transect teams<br />
were deployed simultaneously, one proceed<strong>in</strong>g on an odd numbered transect and one on<br />
an even numbered transect (i.e., on opposite sides <strong>of</strong> <strong>the</strong> valley). Each team (three to four<br />
persons) was responsible for <strong>in</strong>spect<strong>in</strong>g all tree-throw disturbances, bankl<strong>in</strong>es, game trails,<br />
or o<strong>the</strong>r rare exposures <strong>of</strong> soil throughout <strong>the</strong> 20 m wide transect swath, and could test or<br />
<strong>in</strong>spect beyond <strong>the</strong>se limits at <strong>the</strong> discretion <strong>of</strong> <strong>the</strong> crew chief. An all-terra<strong>in</strong> vehicle (ATV)<br />
was completely essential for ferry<strong>in</strong>g personnel and equipment across sloughs, tributary<br />
channels, and even portions <strong>of</strong> lakes <strong>in</strong>tersected by <strong>the</strong> transect route. We planned to extend<br />
and complete each <strong>of</strong> <strong>the</strong> four <strong>in</strong>itial transects over a few days’ time, mov<strong>in</strong>g <strong>in</strong>to and withdraw<strong>in</strong>g<br />
from <strong>the</strong> floodpla<strong>in</strong> by <strong>the</strong> same route each day, until <strong>the</strong> Ouachita River term<strong>in</strong>ation<br />
was reached. A 1976 aerial photo mosaic at 1:10,000 scale was prepared and marked<br />
appropriately, mounted on masonite, covered by plastic film, and carried <strong>in</strong>to <strong>the</strong> field as <strong>the</strong><br />
pr<strong>in</strong>cipal locat<strong>in</strong>g and plott<strong>in</strong>g device. Figures 21 and 22 illustrate <strong>the</strong>se items <strong>of</strong> equipment<br />
as employed for random transect surveys.<br />
The results <strong>of</strong> <strong>the</strong>se random transects were largely negative; no sites were recorded <strong>in</strong><br />
<strong>the</strong> backswamp portion <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> and <strong>the</strong> method <strong>of</strong> proceed<strong>in</strong>g by ATV was necessarily<br />
modified. Two sites were located on or near <strong>the</strong> riverbank and several o<strong>the</strong>rs on higher<br />
terra<strong>in</strong> near <strong>the</strong> orig<strong>in</strong> <strong>of</strong> transects, all by <strong>in</strong>spection ra<strong>the</strong>r than auger<strong>in</strong>g. In <strong>the</strong> case <strong>of</strong><br />
all four transects we discovered that un<strong>in</strong>terrupted, end-to-end, ATV travel was impossible<br />
because <strong>of</strong> localized dense shrub growth and cypress knees, and that cross<strong>in</strong>gs <strong>of</strong> expanses<br />
<strong>of</strong> open water by ATV were too slow and cumbersome. Clear<strong>in</strong>g <strong>of</strong> woody growth was not<br />
considered feasible and attempts to make short circumventions <strong>of</strong> <strong>the</strong>se obstacles usually<br />
failed. We <strong>the</strong>refore attempted to complete <strong>the</strong> random transects <strong>in</strong> three segments each: an<br />
outer floodpla<strong>in</strong> segment reached by ATV from <strong>the</strong> elevated terrace, an <strong>in</strong>ner floodpla<strong>in</strong> segment<br />
reached by boat and pedestrian travel from <strong>the</strong> riverbank, and a midtransect segment<br />
reached by ATV on <strong>the</strong> best available, circuitous, floodpla<strong>in</strong> route. These attempts were<br />
successful to <strong>the</strong> extent that 50% to 80% <strong>of</strong> each transect route was completed, <strong>the</strong> rema<strong>in</strong>der<br />
represent<strong>in</strong>g water surface or saturated terra<strong>in</strong> areas.<br />
Random transect work exceeded <strong>the</strong> time tentatively scheduled for this <strong>in</strong>itial stage,<br />
and riverbank survey was briefly delayed by <strong>the</strong> compet<strong>in</strong>g need for boat transportation.<br />
The two teams emplaced a total <strong>of</strong> 390 auger holes to 1 m depth, proportionately more on<br />
<strong>the</strong> longer and drier transects (Tl and T2). These auger holes each required only a few m<strong>in</strong>utes<br />
to complete, but none recovered an artifact or trace <strong>of</strong> organic rema<strong>in</strong>s which could be<br />
attributed to cultural orig<strong>in</strong> (e.g., shell, bone, charcoal). One <strong>of</strong> <strong>the</strong> detrimental qualities <strong>of</strong><br />
auger test<strong>in</strong>g was <strong>the</strong> tendency to br<strong>in</strong>g up a cohesive mass or fillet <strong>of</strong> wet clay <strong>in</strong> water-
108 Hemm<strong>in</strong>gs<br />
Figure 21. Site survey tactics and equipment. a. <strong>of</strong>f-load<strong>in</strong>g <strong>the</strong> Hustler all terra<strong>in</strong><br />
vehicle (ATV) from a flatbed trailer at a po<strong>in</strong>t <strong>of</strong> access to <strong>the</strong> floodpla<strong>in</strong>;<br />
b. subsurface test<strong>in</strong>g by means <strong>of</strong> power auger<strong>in</strong>g <strong>in</strong> bottomland hardwood<br />
forest (AAS neg. 796415, 796537).<br />
A<br />
b
Variability Among Floodpla<strong>in</strong> Sites 109<br />
Figure 22. Mounted air photo mosaics used for location f<strong>in</strong>d<strong>in</strong>g and site plott<strong>in</strong>g<br />
by boat survey and floodpla<strong>in</strong> transect survey teams (AAS neg. 814156,<br />
814153, 813887, 813888, 814154).
110 Hemm<strong>in</strong>gs<br />
logged areas, <strong>in</strong> which small sparse artifacts or organic rema<strong>in</strong>s would be obscured. The<br />
negative results <strong>of</strong> <strong>the</strong>se four random transects may <strong>in</strong>dicate that no sites are present <strong>in</strong><br />
backswamp, but more likely that sites are too few, too small, scattered, or deeply buried to<br />
be located by this method. Great difficulty <strong>in</strong> complet<strong>in</strong>g <strong>the</strong>se transects conv<strong>in</strong>ced us to<br />
ensure greater mobility <strong>in</strong> <strong>the</strong> succeed<strong>in</strong>g survey stage.<br />
The series <strong>of</strong> six judgmental transects next employed <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland<br />
subarea was considerably different <strong>in</strong> concept and execution. Three o<strong>the</strong>r judgmental<br />
transects (T8, T9, and T10) <strong>in</strong> <strong>the</strong> Ouachita-Sal<strong>in</strong>e Channels subarea are similar <strong>in</strong> execution.<br />
These n<strong>in</strong>e transects (Figure 20), compris<strong>in</strong>g <strong>the</strong> second stage <strong>of</strong> floodpla<strong>in</strong> survey, were<br />
equally based on <strong>the</strong> need for greater mobility and <strong>the</strong> opportunity to explore a variety <strong>of</strong><br />
localities and geohydrological sett<strong>in</strong>gs with<strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>. To ensure mobility we planned<br />
transect routes to orig<strong>in</strong>ate on higher terra<strong>in</strong> and to follow closely 4-wheel drive tracks<br />
shown on air photos, <strong>the</strong> toe <strong>of</strong> elevated terraces, and exist<strong>in</strong>g borrow pits, all routes travers<strong>in</strong>g<br />
floodpla<strong>in</strong> but vary<strong>in</strong>g considerably <strong>in</strong> extent <strong>of</strong> terra<strong>in</strong> below <strong>the</strong> 65-foot contour. These<br />
routes m<strong>in</strong>imized cross<strong>in</strong>gs <strong>of</strong> open water, and were surveyed by pedestrian teams us<strong>in</strong>g <strong>the</strong><br />
ATV capability only when necessary. Characteristics <strong>of</strong> <strong>the</strong>se judgmental transects, <strong>in</strong>clud<strong>in</strong>g<br />
<strong>the</strong> diverse floodpla<strong>in</strong> areas and geohydrological sett<strong>in</strong>gs targeted for survey, are summarized<br />
<strong>in</strong> Table 6.<br />
The pr<strong>in</strong>cipal tactic for <strong>the</strong>se judgmental transects was to send <strong>the</strong> survey teams from<br />
a po<strong>in</strong>t on higher terra<strong>in</strong> <strong>in</strong>to <strong>the</strong> floodpla<strong>in</strong> on designated pedestrian routes, shovel test<strong>in</strong>g<br />
at 50 m <strong>in</strong>tervals (25 cm diameter by 50 cm depth) below <strong>the</strong> 65 foot contour, except where<br />
tree-throw disturbance or deeply rutted vehicle tracks provided comparable exposure or<br />
where prohibited by water. These shovel tests were closely <strong>in</strong>spected, but not screened.<br />
Each team (three to four persons) was aga<strong>in</strong> responsible for close <strong>in</strong>spection <strong>of</strong> a 20 m wide<br />
transect swath, and carried aerial photo mosaics at 1:10,000 scale with <strong>the</strong> route marked<br />
(Figure 22). Two survey teams were deployed simultaneously, <strong>in</strong> most cases operat<strong>in</strong>g on<br />
opposite sides <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. Judgmental transects were designed to proceed rapidly,<br />
and were <strong>in</strong> fact able to do so. Depend<strong>in</strong>g on length <strong>of</strong> transect, difficulty <strong>in</strong> water cross<strong>in</strong>g,<br />
and degree <strong>of</strong> subsurface exposure, <strong>the</strong>se transects required one-half to four days to complete.<br />
This relatively rapid method essentially sacrificed <strong>the</strong> capability <strong>of</strong> discover<strong>in</strong>g sites at<br />
greater than 50 cm depth, but <strong>in</strong>creased <strong>the</strong> extent <strong>of</strong> survey markedly for this second stage<br />
and presumably <strong>the</strong> ability to discover shallow buried sites.<br />
Transects 9 and 10 have special importance for achiev<strong>in</strong>g and evaluat<strong>in</strong>g results dur<strong>in</strong>g<br />
this second survey stage. In carry<strong>in</strong>g out <strong>the</strong>se transects we took advantage <strong>of</strong> exist<strong>in</strong>g,<br />
l<strong>in</strong>ear, water-filled, borrow pits which parallel U.S. Highway 82 and extend nearly across <strong>the</strong><br />
Recent floodpla<strong>in</strong>. These borrow pits have steep eroded walls which extend from <strong>the</strong> water<br />
surface at 62.2 feet to <strong>the</strong> modern floodpla<strong>in</strong> surface at 65-70-foot elevations. In effect <strong>the</strong>re is<br />
an excellent, almost cont<strong>in</strong>uous exposure <strong>of</strong> floodpla<strong>in</strong> sediments about 1 m high and 8 km
Variability Among Floodpla<strong>in</strong> Sites 111<br />
Table 6. Characteristics <strong>of</strong> N<strong>in</strong>e Judgmental Transects <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> and Ouachita-<br />
Sal<strong>in</strong>e Channels Subareas.<br />
Transect Length<br />
Designation (Kilometers) Survey Target Area* Tactics Employed<br />
T5 3.3 Ouachita River levee and backslope to shovel test and <strong>in</strong>spect<br />
Open Lake<br />
T6 7.7 backswamp at Redeye Lake and Horseshoe shovel test and <strong>in</strong>spect<br />
Lake; Ouachita River levee at Henderson<br />
Bend<br />
T7 2.8 Gum Ridge <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> backswamp shovel test and <strong>in</strong>spect<br />
T8 5.5 Brushy Creek-<strong>Marais</strong> Sal<strong>in</strong>e backswamp; shovel test and <strong>in</strong>spect<br />
Sal<strong>in</strong>e River levee at mile 1.0<br />
T9 6.0 Ouachita River levee at mile 252.1; Pea <strong>in</strong>spect subsurface exposure<br />
Ridge-Lapoile Creek backswamp <strong>in</strong> borrow pit<br />
T10 2.0 Ouachita River levee at mile 252.1; terrace- <strong>in</strong>spect subsurface exposure<br />
floodpla<strong>in</strong> transition <strong>in</strong> borrow pit<br />
T11 8.4 terrace-floodpla<strong>in</strong> transition; <strong>Grand</strong> <strong>Marais</strong>- shovel test and <strong>in</strong>spect<br />
Lapile Creek backswamp<br />
T12 7.3 terrace-floodpla<strong>in</strong> transition; backswamp at shovel test and <strong>in</strong>spect<br />
Wildcat Lake<br />
T13 3.7 Ouachita River levee and backslope to <strong>the</strong> shovel test and <strong>in</strong>spect<br />
<strong>Grand</strong> <strong>Marais</strong> backswamp<br />
* Place-names and river mileages from USGS Felsenthal, Ark.-La. 15 m<strong>in</strong>ute (1939); floodpla<strong>in</strong> geomorphic divisions<br />
from Fleetwood (1969).
112 Hemm<strong>in</strong>gs<br />
<strong>in</strong> length. Three small sites were recorded here (3UN<strong>17</strong>0, 3AS295, 3AS294; see Unit 6 <strong>in</strong><br />
Appendix A) which are, <strong>of</strong> course, shallow buried components disturbed by borrow pit<br />
excavation. It was <strong>of</strong> great <strong>in</strong>terest and importance to know at this stage that so few buried<br />
sites are present <strong>in</strong> an extensive transect <strong>of</strong> backswamp and <strong>in</strong> <strong>the</strong> upper meter <strong>of</strong> sediments<br />
exposed.<br />
The results <strong>of</strong> n<strong>in</strong>e judgmental transects <strong>in</strong> both subareas are on <strong>the</strong> whole more positive<br />
and more useful than those <strong>of</strong> <strong>the</strong> preced<strong>in</strong>g random four. All but one <strong>of</strong> <strong>the</strong> second<br />
stage transects recorded one or more sites; <strong>the</strong>se sites were most commonly located on<br />
terrace-floodpla<strong>in</strong> transition areas, but a few shallow buried sites <strong>in</strong> backswamp environments<br />
were also located. While we deplore <strong>the</strong> lack <strong>of</strong> an efficient and reliable method <strong>of</strong><br />
subsurface survey <strong>in</strong> <strong>the</strong> backswamp, we are more confident that <strong>the</strong> number and density <strong>of</strong><br />
shallow buried sites here are low. The backswamp surveys carried out by boat, to be described<br />
later, tend to corroborate this conclusion.<br />
River Bankl<strong>in</strong>e Surveys<br />
Background research and some limited prior site data <strong>in</strong>dicated that riverbank sites<br />
could be expected to occur <strong>in</strong> both subareas. Accord<strong>in</strong>gly, we devoted a major effort to<br />
bankl<strong>in</strong>e survey, employ<strong>in</strong>g a boat survey team (three to four persons) cont<strong>in</strong>uously for<br />
seven weeks and <strong>in</strong>termittently <strong>the</strong>reafter on both river courses. A secondary function <strong>of</strong> this<br />
boat survey team was to acquire floodpla<strong>in</strong> environmental data <strong>of</strong> several k<strong>in</strong>ds (Chapters<br />
2, 7). River bankl<strong>in</strong>e surveys, for example, provided a major source <strong>of</strong> <strong>in</strong>formation about <strong>the</strong><br />
alluvial stratigraphic context <strong>of</strong> buried sites <strong>in</strong> <strong>the</strong> project area. The assumption made about<br />
potential site location <strong>in</strong> any reach <strong>of</strong> river channel was noted earlier. This assumption and<br />
o<strong>the</strong>rs regard<strong>in</strong>g <strong>the</strong> relationship <strong>of</strong> sites and site discovery to bank forms and channel geometry<br />
could be thoroughly tested aga<strong>in</strong>st survey results (see Units 1, 2, and 4). The extent<br />
<strong>of</strong> river bankl<strong>in</strong>e survey completed is summarized below for both subareas (see Figure 23):<br />
Ouachita-Sal<strong>in</strong>e Channels subarea:<br />
Ouachita River above U.S. Highway 82 bridge = 37.8 river miles or 60.8 km<br />
(total for left and right banks)<br />
Lower Sal<strong>in</strong>e River = 22.0 river miles or 35.4 km (total for left and right banks)<br />
<strong>Grand</strong> <strong>Marais</strong> Lowland subarea:<br />
Ouachita River below U.S. Highway 82 bridge = 19.6 river miles or 31.6 km<br />
(total for left and right banks)<br />
These figures were calculated by doubl<strong>in</strong>g navigational distances <strong>in</strong>dicated on <strong>the</strong> USGS<br />
Felsenthal, Ark.-La. 15 m<strong>in</strong>ute quadrangle (1939). Our upstream and downstream limits <strong>of</strong><br />
bankl<strong>in</strong>e survey were miles 271.0 and 242.3 on <strong>the</strong> Ouachita River, and miles 11.0 and 0.0
Variability Among Floodpla<strong>in</strong> Sites 113<br />
Figure 23. Routes and extent <strong>of</strong> bankl<strong>in</strong>e survey <strong>in</strong> river channels, tributaries,<br />
open lakes, and backswamp areas accessible by boat.
114 Hemm<strong>in</strong>gs<br />
on <strong>the</strong> lower Sal<strong>in</strong>e River, for a total navigational distance <strong>of</strong> 39.7 river miles (63.9 km). The<br />
limits here were partly determ<strong>in</strong>ed by <strong>the</strong> practical daily range <strong>of</strong> boat travel and survey<br />
activity. We found it essential to operate boats from <strong>the</strong> improved launch<strong>in</strong>g area on <strong>the</strong><br />
right bank just above U.S. Highway 82. This arrangement provided us with a secure dock<strong>in</strong>g<br />
facility, centrally located with respect to <strong>the</strong> entire Felsenthal Project area.<br />
River bankl<strong>in</strong>e survey tactics were specifically devised to meet field conditions at <strong>the</strong><br />
outset <strong>of</strong> this work, and were consistently practiced throughout <strong>the</strong> seven weeks <strong>of</strong> cont<strong>in</strong>uous<br />
survey and <strong>the</strong>reafter. The major requirement was cont<strong>in</strong>uous close scrut<strong>in</strong>y <strong>of</strong> all accessible<br />
bankl<strong>in</strong>e by observers on foot, usually work<strong>in</strong>g as a pair or team. The only <strong>in</strong>accessible<br />
areas were low slop<strong>in</strong>g bankl<strong>in</strong>es with impenetrable undergrowth at <strong>the</strong> water’s edge,<br />
usually found on <strong>the</strong> slip<strong>of</strong>f slope <strong>of</strong> prom<strong>in</strong>ent bends and at <strong>the</strong> entrance to sloughs, and,<br />
more rarely, vertical cutbanks with no foot<strong>in</strong>g. We estimate that only 10 to 20% <strong>of</strong> Ouachita<br />
or Sal<strong>in</strong>e River bankl<strong>in</strong>es was so thoroughly <strong>in</strong>accessible. In <strong>the</strong> rema<strong>in</strong><strong>in</strong>g 80 to 90% survey<br />
teams <strong>in</strong>spected <strong>the</strong> beach or water’s edge, <strong>the</strong> face or slope <strong>of</strong> <strong>the</strong> bank, and <strong>the</strong> summit <strong>of</strong><br />
<strong>the</strong> bank, compris<strong>in</strong>g at least a 10 m wide swath extend<strong>in</strong>g back from <strong>the</strong> shorel<strong>in</strong>e. At <strong>the</strong><br />
discretion <strong>of</strong> <strong>the</strong> crew chief, boat survey personnel could, and frequently did, extend <strong>the</strong><br />
survey <strong>in</strong>to coves, sloughs, crevasses, or clear<strong>in</strong>gs adjacent to <strong>the</strong> channel (bankl<strong>in</strong>e survey<br />
<strong>in</strong> tributary dra<strong>in</strong>ages is described below). This degree <strong>of</strong> scrut<strong>in</strong>y is necessary because <strong>of</strong><br />
<strong>the</strong> nature <strong>of</strong> floodpla<strong>in</strong> sites <strong>in</strong> this region—under <strong>the</strong> best conditions <strong>of</strong> riverbank exposure<br />
<strong>the</strong>y have extremely low visibility.<br />
It should be emphasized here also that bankl<strong>in</strong>e survey tactics encompassed <strong>the</strong><br />
entire riverbank exposure and not just a portion below <strong>the</strong> 65-foot contour. Archeological<br />
rema<strong>in</strong>s occurred <strong>in</strong> situ, or eroded and transported, at any elevation or on any portion <strong>of</strong><br />
an exposed bankl<strong>in</strong>e, and no case can be made for ignorance <strong>of</strong> <strong>the</strong> sources or stratigraphic<br />
relationships <strong>of</strong> <strong>the</strong>se rema<strong>in</strong>s. This <strong>in</strong>troduces a second major requirement <strong>of</strong> bankl<strong>in</strong>e<br />
survey—exposed artifacts and debris which rema<strong>in</strong>ed <strong>in</strong> situ were so identified, and depth<br />
below modern surface was recorded by actual taped measurement as an <strong>in</strong>dicator <strong>of</strong> relative<br />
age and site significance (see Chapter 7 and Appendix C). Site elevations were recorded by<br />
survey teams referr<strong>in</strong>g to river stage (read on a river guage daily), and were checked aga<strong>in</strong>st<br />
U.S. Geological Survey advance work pr<strong>in</strong>ts (1978) for <strong>the</strong> project area.<br />
For both <strong>the</strong> Ouachita and Sal<strong>in</strong>e River bankl<strong>in</strong>e surveys, aerial photo mosaics identical<br />
to those employed <strong>in</strong> floodpla<strong>in</strong> transect work were <strong>the</strong> pr<strong>in</strong>cipal locat<strong>in</strong>g or site plott<strong>in</strong>g<br />
device (Figure 22). Site locations and extents were marked with grease pencil on <strong>the</strong> clear<br />
plastic overlay <strong>of</strong> <strong>the</strong>se air photos, and later transferred to project base maps. The upstreamdownstream<br />
length <strong>of</strong> artifact scatters was easily measured by tap<strong>in</strong>g or pac<strong>in</strong>g along <strong>the</strong><br />
riverbank (Appendix A, Table A-1, column 6). The perpendicular dimension or site width,<br />
however, proved to be elusive beyond <strong>the</strong> eroded bankl<strong>in</strong>e, for, as noted earlier, all floodpla<strong>in</strong><br />
sites are buried. Considerable experimentation with power auger<strong>in</strong>g and shovel test<strong>in</strong>g<br />
at 5 m and 2.5 m <strong>in</strong>tervals did not achieve <strong>the</strong> desired result where occupation levels
Variability Among Floodpla<strong>in</strong> Sites 115<br />
were sparse, or deep, or both <strong>of</strong> <strong>the</strong>se (and <strong>in</strong>deed most sites were markedly sparse). Therefore,<br />
<strong>in</strong> order to make timely and efficient progress, we abandoned <strong>the</strong> attempt to trace<br />
site width below <strong>the</strong> surface (thus also precise buried site area), and <strong>the</strong> l<strong>in</strong>ear configuration<br />
recorded for most floodpla<strong>in</strong> sites (as width x length <strong>in</strong> column 6, Table A-1) is partly<br />
a measure <strong>of</strong> exposure and exploration technique. However, we know from various sites<br />
with substantial exposure, and from o<strong>the</strong>rs tested extensively for site limits (Chapter 6), that<br />
floodpla<strong>in</strong> sites, regardless <strong>of</strong> overall size, consistently do exhibit a pronounced l<strong>in</strong>ear configuration,<br />
where <strong>the</strong> length along <strong>the</strong> shorel<strong>in</strong>e far exceeds <strong>the</strong> width. Test<strong>in</strong>g also demonstrated<br />
that l<strong>in</strong>ear configuration <strong>in</strong> <strong>the</strong>se sites was a function <strong>of</strong> cultural activity ra<strong>the</strong>r than<br />
scour<strong>in</strong>g and downstream transport <strong>of</strong> artifacts or debris by floodwaters (Chapter 6).<br />
S<strong>in</strong>ce riverbank exposures are a k<strong>in</strong>d <strong>of</strong> s<strong>in</strong>uous transect, and one that can be effectively<br />
surveyed (that is, sites <strong>in</strong>tersected by <strong>the</strong> river have a high probability <strong>of</strong> discovery), <strong>the</strong><br />
results <strong>of</strong> our bankl<strong>in</strong>e site survey provide some controlled distributional data. The exist<strong>in</strong>g<br />
navigation pool enters this problem by vary<strong>in</strong>g <strong>the</strong> amount <strong>of</strong> riverbank exposure accessible<br />
to survey. As <strong>the</strong> floodpla<strong>in</strong> slopes downstream, <strong>the</strong> navigation pool’s horizontal surface<br />
<strong>in</strong>creas<strong>in</strong>gly “drowns” <strong>the</strong> riverbank exposures. In our project area Ouachita River bankl<strong>in</strong>e<br />
heights decrease more or less cont<strong>in</strong>uously downstream from about 5 m to 1 m above lock<br />
stage, and Sal<strong>in</strong>e River bankl<strong>in</strong>e heights decrease from over 2 m to over 1 m. Thus, if archeological<br />
sites were randomly distributed <strong>in</strong> <strong>the</strong> irregular prism <strong>of</strong> floodpla<strong>in</strong> sediments, <strong>the</strong><br />
number <strong>of</strong> such sites <strong>in</strong>tersected (and recorded) would steadily decrease downstream. Shift<strong>in</strong>g<br />
<strong>of</strong> channel reaches and destruction <strong>of</strong> sites <strong>in</strong> <strong>the</strong>se reaches would not affect <strong>the</strong> overall<br />
steady decrease. In advance <strong>of</strong> <strong>the</strong> results presented later <strong>in</strong> this chapter (Units 1, 2, and 4),<br />
we can say that random distribution is all but excluded, and that o<strong>the</strong>r distributional factors<br />
are <strong>in</strong>dicated. The same drown<strong>in</strong>g <strong>of</strong> riverbank exposure downstream should also cause<br />
older, more deeply buried sites to decrease <strong>in</strong> that direction, and this relationship is corroborated<br />
by our data (compare Site Units 1 and 4 later <strong>in</strong> this chapter).<br />
Bankl<strong>in</strong>e survey on both <strong>the</strong> Ouachita and Sal<strong>in</strong>e rivers produced significant results <strong>in</strong><br />
terms <strong>of</strong> newly discovered sites, our understand<strong>in</strong>g <strong>of</strong> <strong>the</strong>ir physical context, and our application<br />
and evaluation <strong>of</strong> survey methods. Records <strong>of</strong> 68 buried prehistoric sites on <strong>the</strong><br />
narrow natural levees <strong>of</strong> <strong>the</strong>se rivers, 64 previously unknown, are essential to both scientific<br />
and resource management goals <strong>of</strong> <strong>the</strong> Felsenthal Project. All 40 floodpla<strong>in</strong> sites recorded<br />
on <strong>the</strong> lower Sal<strong>in</strong>e River, a dramatic and important concentration, were previously unknown.<br />
Most historic sites (Unit 8) also occurred on or near <strong>the</strong> riverbanks, and were newly<br />
recorded or re<strong>in</strong>vestigated by boat survey teams, sometimes us<strong>in</strong>g also archival maps or<br />
o<strong>the</strong>r historical <strong>in</strong>formation. One fundamental contribution <strong>of</strong> riverbank survey is an emerg<strong>in</strong>g<br />
body <strong>of</strong> data perta<strong>in</strong><strong>in</strong>g to floodpla<strong>in</strong> prehistoric extractive sites (Chapter 6) and to <strong>the</strong><br />
systematic relationship <strong>of</strong> site depth, age, and alluviation <strong>in</strong> <strong>the</strong> natural levee zone <strong>of</strong> this<br />
lowland region (Chapter 7).
116 Hemm<strong>in</strong>gs<br />
Tributary Bankl<strong>in</strong>e and Backswamp Surveys<br />
Boat or pedestrian survey <strong>of</strong> bankl<strong>in</strong>es <strong>in</strong> tributary waterways was conducted us<strong>in</strong>g<br />
<strong>the</strong> pr<strong>in</strong>ciples and most <strong>of</strong> <strong>the</strong> tactics described above. All <strong>of</strong> this work was scheduled late <strong>in</strong><br />
<strong>the</strong> seven-week period <strong>of</strong> <strong>in</strong>tensive survey, when floodpla<strong>in</strong> transect crews could be reassigned,<br />
and when substantial results from riverbank survey were available. These surveys<br />
thus represent a third stage <strong>of</strong> exploratory work, although <strong>the</strong>y cont<strong>in</strong>ued to emphasize<br />
bankl<strong>in</strong>e <strong>in</strong>spection with<strong>in</strong> <strong>the</strong> backswamp portion <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. A major determ<strong>in</strong>ant <strong>of</strong><br />
<strong>the</strong> extent <strong>of</strong> survey was aga<strong>in</strong> accessibility and <strong>the</strong> practical range <strong>of</strong> boat travel. For much<br />
<strong>of</strong> <strong>the</strong> work considered here a smaller (12-foot) boat with a two-person survey team was employed,<br />
although a larger pedestrian crew was used <strong>in</strong> some areas accessible overland. The<br />
extent <strong>of</strong> accessible tributary channels and backswamp waterways surveyed is characterized<br />
below by subarea (Figures 6, 23):<br />
Ouachita Sal<strong>in</strong>e Channels subarea:<br />
Eagle Creek (total for left and right banks) 5.6 km<br />
Jones Lake (south bank) 1.0 km<br />
St. Mary’s Lake (west bank) 2.4 km<br />
<strong>Grand</strong> <strong>Marais</strong> Lowland subarea:<br />
<strong>Marais</strong> Sal<strong>in</strong>e Lake (perimeter) 2.2 km<br />
Lapile Creek (total for left and right banks) 5.8 km<br />
Lapile Bayou, Fishtrap Lake, and backswamp<br />
distributaries (total for left and right banks) 21.6 km<br />
Wildcat Lake and distributaries (perimeter<br />
and irregular route) 4.5 km<br />
<strong>Grand</strong> <strong>Marais</strong> and distributaries (perimeter<br />
and irregular route) 6.5 km<br />
Bankl<strong>in</strong>e exposure <strong>in</strong> <strong>the</strong>se smaller channels and backswamp areas was generally poor<br />
with dense undergrowth <strong>of</strong>ten obscur<strong>in</strong>g low shorel<strong>in</strong>es. We estimate that less than 10% <strong>of</strong><br />
<strong>the</strong> routes summarized above were well enough exposed for effective bankl<strong>in</strong>e survey with<br />
one major exception. This higher degree <strong>of</strong> exposure was well developed on Lapile Bayou <strong>in</strong><br />
backswamp and greatly exceeded on <strong>the</strong> Lapile Creek segment which extends <strong>in</strong>to <strong>the</strong> Pleistocene<br />
terrace (Figure 23). Heavy vehicle and foot traffic here by fishermen and consequent<br />
bankl<strong>in</strong>e erosion have thoroughly exposed <strong>the</strong> marg<strong>in</strong>s <strong>of</strong> this tributary channel (see Unit 5<br />
below).<br />
Our survey efforts <strong>in</strong> tributary waterways proceeded fairly rapidly, chiefly because <strong>the</strong><br />
most accessible routes were selected and because good bankl<strong>in</strong>e exposure was rare. Relatively<br />
few sites were recorded <strong>in</strong> an actual backswamp environment, and <strong>the</strong>se were usually<br />
exceed<strong>in</strong>gly small and sparse (Unit 6), parallel<strong>in</strong>g results from floodpla<strong>in</strong> transect work. We<br />
can state with greater assurance that shallow buried sites are few and scattered <strong>in</strong> <strong>the</strong> back-
Variability Among Floodpla<strong>in</strong> Sites 1<strong>17</strong><br />
swamp zone <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. If additional time were available for <strong>in</strong>vestigation <strong>of</strong> less accessible<br />
areas, we would have extended <strong>the</strong> survey to <strong>the</strong> Caney Bayou-Lapoile Creek channels<br />
and to large oxbow lakes on <strong>the</strong> north. At Pereogee<strong>the</strong> Lake one previously recorded<br />
prehistoric site (3BR19) was tested by Lishka (1973) with modest, but <strong>in</strong>terest<strong>in</strong>g, results<br />
from our floodpla<strong>in</strong> perspective.<br />
Site Specific Surveys<br />
Site specific survey was employed as a means <strong>of</strong> acquir<strong>in</strong>g regional comparative data,<br />
both archeological and environmental, ra<strong>the</strong>r than as a technique for locat<strong>in</strong>g previously<br />
unknown sites. This work was conducted by <strong>the</strong> project archeologist, usually assisted by a<br />
survey team, but <strong>in</strong> some cases work<strong>in</strong>g with consult<strong>in</strong>g personnel or <strong>in</strong>formants. The follow<strong>in</strong>g<br />
site specific work contributed diverse k<strong>in</strong>ds <strong>of</strong> data and results:<br />
1. Observed new lock construction site, lock chamber excavation, and spoil areas.<br />
2. Collected samples for lithic source study at Recent floodpla<strong>in</strong> and Pleistocene<br />
terrace locations (Chapter 8).<br />
3. Visited and recorded historic sawmill site with <strong>in</strong>formant (3BR55).<br />
4. Visited and recorded newly disturbed prehistoric site on Pleistocene terrace<br />
(3AS328).<br />
5. Visited and recorded natural prairie area and associated archeological sites on<br />
Pleistocene terrace (3AS160, 3AS293).<br />
6. Revisited and recorded major mound groups on Pleistocene terraces and islands<br />
with <strong>Arkansas</strong> Archeological Survey station personnel (3BR4, 3BR8, 3UN18,<br />
3UN8, 3UN9/52, 3AS1, 3AS6).<br />
7. Located and recorded underwater steamboat wreck with mar<strong>in</strong>e survey archeologist<br />
(3UN153).<br />
Site and Component Identification<br />
Our record <strong>of</strong> floodpla<strong>in</strong> sites is a product not only <strong>of</strong> specific survey tactics, but also<br />
<strong>of</strong> operational def<strong>in</strong>itions and levels <strong>of</strong> analysis. We shall briefly <strong>in</strong>troduce certa<strong>in</strong> <strong>of</strong> those<br />
def<strong>in</strong>itions and levels here, and note that differences may exist with o<strong>the</strong>r large scale cultural<br />
resource surveys (e.g., House and Schiffer 1975).<br />
The pr<strong>in</strong>cipal unit <strong>of</strong> observation for Felsenthal Project survey teams was an artifact<br />
or artifact scatter. Debris <strong>of</strong> presumed cultural orig<strong>in</strong> or <strong>of</strong> uncerta<strong>in</strong> orig<strong>in</strong> was similarly<br />
recorded. In <strong>the</strong> stoneless floodpla<strong>in</strong> topstratum, any medium to large pebble or coarse rock<br />
fragment was regarded as a potential “manuport,” and recorded and collected, but was<br />
subsequently judged as to modification and context. Survey teams made 100% surface collections<br />
on prehistoric artifact scatters, and ei<strong>the</strong>r 100% or select representative collections<br />
on historic scatters. A series <strong>of</strong> consecutive project numbers was assigned for use by each<br />
survey team, and ultimately about 200 numbers were used, although some refer to environ-
118 Hemm<strong>in</strong>gs<br />
mental localities or samples. At this level <strong>of</strong> operation (survey team record<strong>in</strong>g) no decision<br />
had been made about “sites.”<br />
In record<strong>in</strong>g historic artifacts, scatters, or structures, we specified that “modern”<br />
rema<strong>in</strong>s be rejected, and those rema<strong>in</strong>s 50 years <strong>of</strong> age or greater be recorded and collected.<br />
Doubtful examples were recorded, collected, and judged subsequently. In practice this elim<strong>in</strong>ated<br />
all <strong>the</strong> “ongo<strong>in</strong>g” rema<strong>in</strong>s <strong>of</strong> fishcamps, hunters’ camps, vehicle tracks, and timber<br />
harvest<strong>in</strong>g activities. While historical rema<strong>in</strong>s less than 50 years <strong>of</strong> age can have regional<br />
and even scientific importance, record<strong>in</strong>g <strong>the</strong>se clearly becomes an endless task for specialists.<br />
A contemporary survey <strong>in</strong> <strong>the</strong> Felsenthal Project area differs on this po<strong>in</strong>t (Heartfield et<br />
al. 1980).<br />
Our second level <strong>of</strong> operation consisted <strong>of</strong> process<strong>in</strong>g, catalog<strong>in</strong>g, and limited study <strong>of</strong><br />
collections at <strong>the</strong> Felsenthal Project field laboratory. Individual objects, collections, or localities<br />
were occasionally elim<strong>in</strong>ated when no archeological or environmental importance could<br />
be established. All archeological collections were now entered <strong>in</strong> <strong>the</strong> permanent catalog<strong>in</strong>g<br />
system <strong>of</strong> <strong>Arkansas</strong> Archeological Survey, while unassociated environmental localities and<br />
samples reta<strong>in</strong>ed only project numbers. Aga<strong>in</strong> no decision about “sites” had been made,<br />
except by elim<strong>in</strong>at<strong>in</strong>g certa<strong>in</strong> localities or collections as <strong>in</strong>dicated.<br />
The third level <strong>of</strong> operation consisted <strong>of</strong> detailed analysis <strong>of</strong> both field data and collections,<br />
conducted upon our return to permanent laboratory and research facilities. At this<br />
stage <strong>the</strong> numbered floodpla<strong>in</strong> loci <strong>of</strong> artifacts or o<strong>the</strong>r cultural rema<strong>in</strong>s were designated as<br />
sites, or comb<strong>in</strong>ed as sites, based on all survey and analytical results. We now had to dist<strong>in</strong>guish<br />
s<strong>in</strong>gle f<strong>in</strong>ds (an apparently isolated artifact or item <strong>of</strong> debris firmly attributed to<br />
human activity) from sparse scatters <strong>of</strong> such items, and <strong>the</strong>se from more measurable, dense,<br />
discrete clusters, usually hav<strong>in</strong>g diameter <strong>of</strong> a few tens <strong>of</strong> meters. A “50 m rule” was developed<br />
at this po<strong>in</strong>t—scattered f<strong>in</strong>ds or discrete scatters less than 50 m apart could be comb<strong>in</strong>ed<br />
as sites when no o<strong>the</strong>r data mitigated aga<strong>in</strong>st this assignment. A s<strong>in</strong>gle fire-cracked<br />
rock fragment could be designated a site (3UN162 is just such a site) when careful <strong>in</strong>spection<br />
produced no o<strong>the</strong>r artifacts with<strong>in</strong> 50 m, and so also could two fl<strong>in</strong>t flakes 49 m apart. Given<br />
<strong>the</strong> widely scattered nature <strong>of</strong> our f<strong>in</strong>ds, this rule was rarely needed, but Unit 5 (discussed<br />
below) is a most notable exception. In Chapter 6 relatively dense concentrations <strong>of</strong> artifacts<br />
(Figure 41) and closely spaced activity loci (Figures 47, 51, 53) are illustrated. We do not recommend<br />
use <strong>of</strong> a 50 m rule <strong>in</strong> any o<strong>the</strong>r regional sett<strong>in</strong>g without careful consideration.<br />
For each <strong>of</strong> <strong>the</strong> 126 previously unknown sites dist<strong>in</strong>guished at this level, standard <strong>Arkansas</strong><br />
Archeological Survey site forms were completed, checked, and distributed to station<br />
archeologists for comments and assignment <strong>of</strong> tr<strong>in</strong>omial site designations. Site names were<br />
also assigned, except <strong>in</strong> <strong>the</strong> case <strong>of</strong> s<strong>in</strong>gle f<strong>in</strong>ds, by <strong>the</strong> project archeologist. These site names
Variability Among Floodpla<strong>in</strong> Sites 119<br />
were drawn from local place-names where available, and from prom<strong>in</strong>ent landscape features<br />
or common names <strong>of</strong> local woody species (Appendix A, Table A-1, column 10). For<br />
each <strong>of</strong> <strong>the</strong> 18 sites known from prior work, updated site forms were completed, and similarly<br />
checked and distributed to station archeologists.<br />
We now def<strong>in</strong>e a site as “any spatially discrete locus <strong>of</strong> rema<strong>in</strong>s <strong>of</strong> past human activity.”<br />
These rema<strong>in</strong>s may or may not be preserved <strong>in</strong> place on <strong>the</strong> locus <strong>of</strong> use and abandonment,<br />
and <strong>the</strong>y may or may not result from a s<strong>in</strong>gle episode <strong>of</strong> past human activity. These<br />
questions and o<strong>the</strong>rs must be answered by detailed study <strong>of</strong> association or context for any<br />
designated site, and by application <strong>of</strong> general pr<strong>in</strong>ciples <strong>of</strong> site formation and transformation<br />
process (Schiffer 1976; B<strong>in</strong>ford 1980). In <strong>the</strong> case <strong>of</strong> Felsenthal floodpla<strong>in</strong> sites, <strong>the</strong> loci <strong>of</strong><br />
rema<strong>in</strong>s are comparatively very small <strong>in</strong> size and <strong>the</strong> site assemblages are small <strong>in</strong> quantity<br />
and low <strong>in</strong> diversity. As a class, such small sites have frequently been attributed to brief,<br />
one-time use by a task unit or small social group (Talmage and Chesler 1977). These sites<br />
have high potential for behavioral analysis <strong>in</strong> general, and for settlement and subsistence<br />
studies <strong>in</strong> <strong>the</strong> Felsenthal region (Chapter 6).<br />
Essentially <strong>the</strong> last operation performed with site data was assignment <strong>of</strong> prehistoric<br />
components, where possible. A “component” is a contemporaneous assemblage <strong>of</strong> rema<strong>in</strong>s,<br />
sometimes a level, with<strong>in</strong> a specific site (Willey and Phillips 1958:21). Typological and<br />
stratigraphic correlation permits assignment <strong>of</strong> related components to “phases” <strong>in</strong> a locality<br />
or region. Two difficulties beset our assignments: (1) <strong>the</strong> Felsenthal regional sequence<br />
is still <strong>in</strong> an early stage <strong>of</strong> formulation, based pr<strong>in</strong>cipally on large upland sites (Chapter 3),<br />
and (2) floodpla<strong>in</strong> site artifact samples are usually small and restricted <strong>in</strong> variety (Appendix<br />
A). However, our floodpla<strong>in</strong> site data <strong>in</strong>clude numerous discrete clusters <strong>of</strong> artifacts and<br />
some buried “floors” where co-occurrence <strong>of</strong> artifacts and raw materials can be observed<br />
and verified aga<strong>in</strong> and aga<strong>in</strong>. Most artifact f<strong>in</strong>ds were made on river banks with more or less<br />
slope and which expose generally flat-ly<strong>in</strong>g sedimentary units. Artifacts or debris still firmly<br />
embedded <strong>in</strong> sediment, essentially flat-ly<strong>in</strong>g and arranged along a bedd<strong>in</strong>g plane or th<strong>in</strong><br />
sedimentary unit, were assumed to be <strong>in</strong> place. Artifacts repos<strong>in</strong>g at angles on <strong>the</strong> surface<br />
<strong>of</strong> <strong>the</strong> bank, scattered <strong>in</strong> location or elevation, were assumed to be transported (usually a<br />
short vertical distance downslope). The follow<strong>in</strong>g attributes <strong>of</strong> components <strong>in</strong> floodpla<strong>in</strong><br />
sites have been determ<strong>in</strong>ed, or at least are strongly suggested, by our site data (see site and<br />
artifact data <strong>in</strong> Chapters 3, 6, 7, 8, and Appendix A):<br />
Mississippi Period, Caney Bayou Phase<br />
1. buried levels at less than 0.2 m depth<br />
2. shell temper<strong>in</strong>g <strong>in</strong> 75-100% <strong>of</strong> ceramics<br />
3. Eagle Creek, Ashley, Bassett and o<strong>the</strong>r arrow po<strong>in</strong>t styles<br />
4. high proportion <strong>of</strong> brown chert among debitage
120 Hemm<strong>in</strong>gs<br />
Mississippi Period, Gran <strong>Marais</strong> Phase<br />
1. buried levels at about 0.2 m depth<br />
2. Gran <strong>Marais</strong> ceramic complex featur<strong>in</strong>g clay-tempered vessels with diverse<br />
rim and body decoration (<strong>in</strong>cis<strong>in</strong>g, punctat<strong>in</strong>g, and o<strong>the</strong>r); shell temper<strong>in</strong>g <strong>in</strong><br />
only 5-15% <strong>of</strong> ceramics<br />
3. Ashley arrow po<strong>in</strong>t style (a variable type cluster)<br />
4. high proportion <strong>of</strong> brown chert among debitage<br />
Baytown-Coles Creek Period, phases undifferentiated<br />
1. buried levels at about 0.2-0.6 m depth<br />
2. high proportion <strong>of</strong> clay-tempered pla<strong>in</strong> vessels with disc bases; occurrence <strong>of</strong><br />
<strong>in</strong>cised rims <strong>in</strong> Coles Creek style; no shell-tempered ceramics<br />
3. small variety <strong>of</strong> Gary dart po<strong>in</strong>t, arrow po<strong>in</strong>ts (?), and o<strong>the</strong>r dart po<strong>in</strong>t styles<br />
4. both novaculite and chert well represented <strong>in</strong> debitage<br />
Marksville Period (no floodpla<strong>in</strong> site data)<br />
Marksville sites may be few, but present, deeply buried, and sparse, as are Tchula sites.<br />
Such sites would probably lack dist<strong>in</strong>ctive decorated sherds (such as Marksville Stamped)<br />
and have only pla<strong>in</strong>ware, <strong>in</strong> contrast to Tchula sites which are ceramically dist<strong>in</strong>ctive. Thus,<br />
Marksville sites may have been missed due to sampl<strong>in</strong>g error or are misclassified as deeply<br />
buried Baytown sites.<br />
Tchula Period, Coon Island Phase<br />
1. buried levels at about 0.6-1.0 m depth<br />
2. <strong>in</strong>cised, stamped, or drag-and-jab decoration on clay-tempered or temperless(?)<br />
vessels<br />
3. Gary dart po<strong>in</strong>ts and perhaps o<strong>the</strong>r dart po<strong>in</strong>t styles<br />
4. high proportion <strong>of</strong> novaculite among debitage<br />
Poverty Po<strong>in</strong>t Period, Calion Phase<br />
1. buried levels at 1.0 m or greater depth<br />
2. no ceramics<br />
3. large Gary dart po<strong>in</strong>ts and o<strong>the</strong>r dart po<strong>in</strong>t styles<br />
4. high proportion <strong>of</strong> light colored novaculite among debitage<br />
5. exotic materials <strong>in</strong>clud<strong>in</strong>g steatite, schist<br />
6. high frequency <strong>of</strong> fire-cracked rock compared to ceramic sites<br />
Late Archaic Period, phases undifferentiated<br />
1. buried levels at about 1.0-2.0 m depth<br />
2. no ceramics<br />
3. large stemmed dart po<strong>in</strong>ts<br />
Middle and Early Archaic Periods (no floodpla<strong>in</strong> site data)
Paleo-Indian Period (no floodpla<strong>in</strong> site data)<br />
Variability Among Floodpla<strong>in</strong> Sites 121<br />
In Appendix A, Table A-1, components represented <strong>in</strong> prehistoric sites are identified<br />
where sufficient comparative data are available. All such identifications represent <strong>in</strong>ferences<br />
where confidence decreases with fewer typological, technological, or stratigraphic observations.<br />
Never<strong>the</strong>less, one or a few diagnostic artifacts were assumed to be good criteria <strong>of</strong><br />
identification <strong>in</strong> <strong>the</strong> context <strong>of</strong> small floodpla<strong>in</strong> sites, just as <strong>the</strong>y may be <strong>in</strong> small upland<br />
sites (Hemm<strong>in</strong>gs 1975). We make an attempt to identify cultural period, where phase cannot<br />
be dist<strong>in</strong>guished; or to <strong>in</strong>dicate an undifferentiated span (“Archaic”) <strong>in</strong> some cases. About<br />
one third <strong>of</strong> all prehistoric floodpla<strong>in</strong> components (Units 1-7) rema<strong>in</strong> <strong>in</strong> an “unknown” category.<br />
Large mound complexes on Pleistocene terraces (Unit 8) are known from exist<strong>in</strong>g site<br />
records to have a long sequence <strong>of</strong> site occupation <strong>in</strong> most cases, and sequences <strong>of</strong> cultural<br />
periods are <strong>in</strong>dicated <strong>in</strong> Table A-1 (“Archaic-Mississippi”).<br />
In Appendix A, Table A-2, historic components are assigned to decade spans with<strong>in</strong><br />
<strong>the</strong> n<strong>in</strong>eteenth and twentieth century, based on regional historical data (Chapter 4), or more<br />
rarely, diagnostic historic artifacts. Archival data for more precise dates is available <strong>in</strong> a few<br />
cases (Unit 8).<br />
surVey results<br />
Unit 1: Ouachita River above Mile 254<br />
Unit 1 <strong>in</strong>cludes 18 prehistoric sites on a s<strong>in</strong>uous <strong>17</strong> mile (27.4 km) stretch <strong>of</strong> Ouachita<br />
River (Figure 24). In this stretch <strong>the</strong> river proceeds only 8 miles (12.9 km) <strong>in</strong> <strong>the</strong> down-valley<br />
direction, and is marked by prom<strong>in</strong>ent bends and oxbow lakes, <strong>the</strong> latter cut <strong>of</strong>f prior to <strong>the</strong><br />
Dunbar-Hunter expedition <strong>of</strong> 1804-1805 (Rowland 1930; McDermott 1963). Dunbar’s map, as<br />
well as <strong>the</strong> General Land Office maps <strong>of</strong> 1827 and later, show that <strong>the</strong> river has ma<strong>in</strong>ta<strong>in</strong>ed<br />
its course, bend for bend, dur<strong>in</strong>g <strong>the</strong> last <strong>17</strong>5 years. The U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers’ hydrographic<br />
surveys <strong>of</strong> 1873 and later also <strong>in</strong>dicate that low water shoals <strong>of</strong> gravel and sand<br />
were extensive <strong>in</strong> this stretch <strong>of</strong> river, <strong>in</strong> contrast to <strong>the</strong> stretch immediately downstream<br />
(Unit 4). The floodpla<strong>in</strong> here also exhibits greater relief, with dist<strong>in</strong>ctive po<strong>in</strong>t bar ridges<br />
<strong>in</strong>side bends and <strong>in</strong>side oxbow lakes. Dunbar and Hunter, travel<strong>in</strong>g upstream, noted <strong>the</strong><br />
appearance <strong>of</strong> small canes and a few p<strong>in</strong>es, on or near <strong>the</strong> riverbank, attribut<strong>in</strong>g this growth<br />
to reduced extent and duration <strong>of</strong> flood<strong>in</strong>g. Thus, <strong>in</strong> certa<strong>in</strong> respects <strong>the</strong> river channel and<br />
its marg<strong>in</strong> <strong>in</strong> Unit 1 represent less extreme overflow conditions than are found <strong>in</strong> <strong>the</strong> <strong>Grand</strong><br />
<strong>Marais</strong> Lowland subarea downstream.<br />
Beg<strong>in</strong>n<strong>in</strong>g with this group <strong>of</strong> 18 prehistoric sites, we will exam<strong>in</strong>e <strong>the</strong> relationship <strong>of</strong><br />
site distribution to river channel geometry. The record <strong>of</strong> riverbank sites may be a product <strong>of</strong><br />
all <strong>the</strong> follow<strong>in</strong>g factors:
122 Hemm<strong>in</strong>gs<br />
Figure 24. Site Unit 1, Ouachita River above mile 254, with prehistoric sites (dots)<br />
located <strong>in</strong> relation to channel geometry. Some prehistoric sites <strong>in</strong> Unit 6<br />
and historic sites <strong>in</strong> Unit 8 (rectangles) are also shown on this map.
Variability Among Floodpla<strong>in</strong> Sites 123<br />
1. Preferential site location, e.g., adjacent to shoals <strong>in</strong> <strong>the</strong> crossover po<strong>in</strong>t between<br />
bends or on higher levee crests adjacent to <strong>the</strong> concave bank <strong>of</strong> major bends.<br />
2. Differential site destruction, e.g., on concave cutbanks <strong>of</strong> migrat<strong>in</strong>g bends.<br />
3. Differential site visibility, e.g., on erod<strong>in</strong>g cutbanks versus prograd<strong>in</strong>g slip<strong>of</strong>f<br />
slopes obscured by vegetation.<br />
4. Artificially imposed bias, e.g., when survey teams give unequal consideration<br />
to left or right banks or to various reaches.<br />
Us<strong>in</strong>g meander term<strong>in</strong>ology <strong>in</strong>troduced <strong>in</strong> Chapter 2 (Figure 8), we can classify all riverbank<br />
site locations <strong>in</strong> eight possible categories, rate <strong>the</strong>se categories by expected site density (from<br />
<strong>the</strong> first three criteria above), and compare <strong>the</strong> observed site distributions. Table 7 makes<br />
this comparison for Unit 1 and two o<strong>the</strong>r site units on <strong>the</strong> Ouachita and Sal<strong>in</strong>e rivers. In Unit<br />
1, riverbank sites are most frequently represented on <strong>the</strong> downstream arm (below <strong>the</strong> axis)<br />
Table 7. Channel Geometry and Prehistoric Site Distribution for Three Site Units on<br />
Ouachita and Sal<strong>in</strong>e Rivers.<br />
Expected Observed Site Density<br />
River Channel Location Site Density* Unit 1 Unit 2 Unit 4 Total<br />
Concave bank above bend axis high 1 16 7 24<br />
Concave bank at axis moderate 3 3 1 7<br />
Concave bank below axis high 9 6 5 20<br />
Crossover po<strong>in</strong>t between bends moderate 0 7 4 11<br />
Convex bank above bend axis moderate 5 2 2 9<br />
Convex bank at axis low 0 3 1 4<br />
Convex bank below axis low 0 1 0 1<br />
Straight reach low 0 2 0 2<br />
Site totals 18 40 20 78<br />
* Based on processes observed or <strong>in</strong>ferred for <strong>the</strong> relatively stable courses <strong>of</strong> <strong>the</strong>se rivers and not<br />
necessarily applicable to o<strong>the</strong>r meander<strong>in</strong>g river regimens.
124 Hemm<strong>in</strong>gs<br />
<strong>of</strong> <strong>the</strong> concave or convex bank. In certa<strong>in</strong> respects <strong>the</strong> Unit 1 site distribution does not agree<br />
with “expected site density” or with distributions observed <strong>in</strong> Units 2 and 4. One possible<br />
explanation <strong>of</strong> this discordance is <strong>the</strong> high frequency <strong>of</strong> Archaic sites <strong>in</strong> Unit 1, as discussed<br />
fur<strong>the</strong>r below. These older sites may well have been established on or near river courses<br />
which vary markedly from <strong>the</strong> “modern” channel serv<strong>in</strong>g as our geometric reference.<br />
It is <strong>of</strong> <strong>in</strong>terest to note that <strong>the</strong> 18 prehistoric sites recorded <strong>in</strong> Unit 1 are equally distributed<br />
on <strong>the</strong> left and right banks (Figure 24). We suspect that extensive shifts <strong>of</strong> channel<br />
dur<strong>in</strong>g <strong>the</strong> past several centuries or imbalance <strong>in</strong> survey <strong>in</strong>tensity would result <strong>in</strong> an asymmetrical<br />
distribution.<br />
The follow<strong>in</strong>g components are represented <strong>in</strong> Unit 1: Archaic 8, Baytown-Coles Creek<br />
3, Tchula/Coon Island 1, and unknown 6 (Appendix A, Table A-1). Three <strong>of</strong> 18 sites recorded<br />
are s<strong>in</strong>gle f<strong>in</strong>ds. The large proportion <strong>of</strong> Archaic sites is very likely a function <strong>of</strong> high banks<br />
and exposure <strong>of</strong> deep alluvium <strong>in</strong> Unit 1. Site 3UN162, consist<strong>in</strong>g <strong>of</strong> a s<strong>in</strong>gle fire-cracked<br />
cobble <strong>in</strong> place at 2.95 m below <strong>the</strong> levee surface, is <strong>the</strong> deepest artifact occurrence recorded<br />
by us <strong>in</strong> <strong>the</strong> Felsenthal Project area (Figure 25). Shovel scrap<strong>in</strong>g along <strong>the</strong> alluvial horizon<br />
<strong>in</strong>dicated by this artifact did not produce o<strong>the</strong>r cultural material. This site exemplifies <strong>the</strong><br />
problem <strong>of</strong> locat<strong>in</strong>g and assess<strong>in</strong>g significance <strong>of</strong> sparse buried sites <strong>in</strong> an alluvial floodpla<strong>in</strong>.<br />
We have no basis for decid<strong>in</strong>g whe<strong>the</strong>r it represents <strong>the</strong> vestige <strong>of</strong> an eroded occupation<br />
level, <strong>the</strong> <strong>in</strong>itial <strong>in</strong>dicator <strong>of</strong> an undisturbed level, or an isolated object transported from<br />
some o<strong>the</strong>r location.<br />
Prehistoric sites <strong>in</strong> Unit 1 are small, sparse, and relatively deep compared with o<strong>the</strong>r<br />
site units (Appendix A, Table A-1). Few survey collections exceed a dozen objects, and <strong>in</strong>organic<br />
debris (usually fire-cracked rock) is <strong>the</strong> major constituent <strong>of</strong> <strong>the</strong>se collections. Diagnostic<br />
ceramics and stone tools are characteristically rare. A large novaculite Macon dart po<strong>in</strong>t<br />
from Small Cane 2 (3BR73) is shown <strong>in</strong> Figure 59, and Tchefuncte Stamped sherds from<br />
Hunter’s Swan (3BR70) are shown <strong>in</strong> Figure 65f. Although Unit 1 sites are sparse, buried<br />
materials are known to occur <strong>in</strong> place <strong>in</strong> some <strong>of</strong> <strong>the</strong>se and may be expected to occur <strong>in</strong> o<strong>the</strong>rs.<br />
The probability <strong>of</strong> Middle Archaic occupations (before about 4000 B.C.) <strong>in</strong> deeply buried<br />
zones is good, and even older occupations could be present (above river stage) as well. Identifiable<br />
Late Archaic and early ceramic occupations are known, while late prehistoric sites<br />
(after A.D. 1100) are curiously absent, or at least presently unknown.<br />
Monitor<strong>in</strong>g <strong>of</strong> <strong>the</strong> reach <strong>of</strong> Ouachita River presented by Unit l is desirable <strong>in</strong> <strong>the</strong><br />
future. Three <strong>of</strong> <strong>the</strong> presently recorded sites have been recommended for mitigation (sites<br />
3BR72, 3BR73, 3UN166 <strong>in</strong> Appendix C). Ongo<strong>in</strong>g riverbank erosion could reveal sites to be<br />
more substantial than <strong>the</strong>y now appear, disclose unrecorded sites, and clearly could remove<br />
<strong>the</strong> ephemeral traces <strong>of</strong> some sites now known. The functional characteristics <strong>of</strong> <strong>the</strong>se sites
Variability Among Floodpla<strong>in</strong> Sites 125<br />
Figure 25. Site 3UN162, consist<strong>in</strong>g <strong>of</strong> fire-cracked rock exposed at 3 m depth <strong>in</strong> <strong>the</strong><br />
Ouachita River bankl<strong>in</strong>e (AAS neg. 796519).
126 Hemm<strong>in</strong>gs<br />
rema<strong>in</strong> to be established through test<strong>in</strong>g and analysis <strong>of</strong> site data, as we have attempted to<br />
do <strong>in</strong> Units 2 and 4 (Chapter 6). One <strong>of</strong> <strong>the</strong> <strong>in</strong>terest<strong>in</strong>g possibilities is that prehistoric fish<strong>in</strong>g<br />
stations here were established at riverbank locations adjacent to extensive shoals.<br />
Unit 2: Lower Sal<strong>in</strong>e River<br />
Unit 2 comprises <strong>the</strong> lower 11 miles (<strong>17</strong>.7 km) <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River where 40 riverbank<br />
sites, all previously unknown, were recorded (Figure 26). The Sal<strong>in</strong>e River’s course is<br />
slightly less s<strong>in</strong>uous than <strong>the</strong> Ouachita River <strong>in</strong> Unit 1, mak<strong>in</strong>g about 6.5 miles (20.1 km) <strong>in</strong><br />
<strong>the</strong> down-valley direction through this reach. Meander length, amplitude, and radius, and<br />
channel width, are all smaller than observed for <strong>the</strong> Ouachita River, reflect<strong>in</strong>g <strong>the</strong> Sal<strong>in</strong>e<br />
River’s lower discharge. Relatively many spits, l<strong>in</strong>ed by cypress trees and buttonbush, and<br />
elongate coves or bays give this lower Sal<strong>in</strong>e River course a drowned appearance (Figure<br />
46). Bank height decreases from greater than 2.0 m at river mile 11.0 to greater than 1.0 m at<br />
<strong>the</strong> mouth.<br />
The General Land Office maps <strong>of</strong> 1827 and later years <strong>in</strong>dicate that <strong>the</strong> lower Sal<strong>in</strong>e<br />
River channel has also rema<strong>in</strong>ed stable for at least a century and a half. Historic and modern<br />
maps <strong>of</strong> <strong>the</strong> channel can be correlated bend for bend, with a s<strong>in</strong>gle exception. This exception<br />
is <strong>the</strong> cut<strong>of</strong>f <strong>of</strong> a small bend at <strong>the</strong> mouth <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River about 1895, probably artificially<br />
assisted, as shown on a U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers hydrographic survey (Sheet 11)<br />
for that year. The riverbank sett<strong>in</strong>g and perhaps <strong>the</strong> <strong>in</strong>tegrity <strong>of</strong> two sites (3BR77 <strong>in</strong> Unit 1,<br />
3AS284 <strong>in</strong> Unit 2) are affected by this cut<strong>of</strong>f (Figures 24, 26).<br />
Table 8 shows considerable agreement between expected and observed site densities<br />
among <strong>the</strong> eight possible channel locations previously discussed. Sites were most frequently<br />
recorded along concave bankl<strong>in</strong>es above <strong>the</strong> axis <strong>of</strong> a bend. Relatively many sites were also<br />
recorded at crossover po<strong>in</strong>ts between bends. One or more sites occur <strong>in</strong> all eight categories.<br />
The distribution <strong>of</strong> sites by left and right bank is slightly asymmetrical (23/<strong>17</strong>), which could<br />
reflect survey team bias. The boat survey team ord<strong>in</strong>arily proceeded upstream, work<strong>in</strong>g<br />
along <strong>the</strong> left bank, <strong>in</strong> <strong>the</strong> morn<strong>in</strong>g, and downstream along <strong>the</strong> right bank <strong>in</strong> <strong>the</strong> afternoon.<br />
The most <strong>in</strong>terest<strong>in</strong>g aspect <strong>of</strong> site distribution <strong>in</strong> this Sal<strong>in</strong>e River unit is <strong>the</strong> <strong>in</strong>creas<strong>in</strong>g site<br />
density downstream, contrary to <strong>the</strong> probability <strong>of</strong> site discovery as bank height and exposure<br />
decrease (Figure 26). Twenty-eight <strong>of</strong> 40 sites recorded lie along <strong>the</strong> 5 mile (8 km) reach<br />
above <strong>the</strong> mouth. This is a dramatic concentration <strong>of</strong> floodpla<strong>in</strong> sites unequalled elsewhere<br />
<strong>in</strong> <strong>the</strong> project area. The downstream <strong>in</strong>crease <strong>in</strong> site density is believed to be significant and<br />
real, not an artifact <strong>of</strong> survey <strong>in</strong>tensity (see Site Locational Hypo<strong>the</strong>ses, this chapter).<br />
The follow<strong>in</strong>g 41 prehistoric components are represented <strong>in</strong> Unit 2 (Appendix A, Table<br />
A-1):
Variability Among Floodpla<strong>in</strong> Sites 127<br />
Figure 26. Site Unit 2, lower Sal<strong>in</strong>e River, and Site Unit 3, lower Eagle Creek, with<br />
prehistoric sites (dots) located <strong>in</strong> relation to channel geometry. Two<br />
historic sites with prehistoric components (rectangles enclos<strong>in</strong>g dots) <strong>in</strong><br />
Unit 8 are also shown on this map.
128 Hemm<strong>in</strong>gs<br />
Mississippi/Term<strong>in</strong>al 2<br />
Mississippi/Caney Bayou 4<br />
Mississippi/Gran <strong>Marais</strong> 6<br />
Mississippi 4<br />
Baytown-Coles Creek 4<br />
Tchula/Coon Island 1<br />
Poverty Po<strong>in</strong>t/Calion 2<br />
Archaic 4<br />
Unknown 14<br />
Only one multicomponent site has def<strong>in</strong>itely been identified; False Indigo (3AS285) has<br />
a Gran <strong>Marais</strong> phase component overly<strong>in</strong>g a Coon Island phase component, hence <strong>the</strong> total<br />
<strong>of</strong> 41 components reported above for 40 sites. This site and six o<strong>the</strong>rs <strong>in</strong> Unit 2 were systematically<br />
tested, and are reported <strong>in</strong> detail <strong>in</strong> Chapter 6. It is important to note that 40% <strong>of</strong><br />
components identified here are Mississippi period, while no component <strong>of</strong> this period was<br />
identified <strong>in</strong> Unit 1. The lower reach <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River thus appears to have been selectively<br />
utilized by Mississippi period groups, a po<strong>in</strong>t we return to <strong>in</strong> discuss<strong>in</strong>g o<strong>the</strong>r site units<br />
below.<br />
While <strong>the</strong> riverbank sites <strong>in</strong> Unit 2 are generally <strong>of</strong> small size, one-fourth <strong>of</strong> <strong>the</strong>se<br />
exceed 50 m <strong>in</strong> length. These latter sites frequently <strong>in</strong>clude multiple discrete artifact scatters.<br />
Jug Po<strong>in</strong>t 1 (3AS306) and Jug Po<strong>in</strong>t 2 (3AS307), for example, circumscribe five such scatters<br />
with<strong>in</strong> a 225 m stretch <strong>of</strong> Sal<strong>in</strong>e River bankl<strong>in</strong>e (Figures 51, 53). In some cases <strong>the</strong>se scatters<br />
appear to be coeval activity loci and <strong>in</strong> o<strong>the</strong>rs merely successive occupations <strong>of</strong> closely<br />
spaced loci (Chapter 6). The typical lower Sal<strong>in</strong>e River site is evidently a m<strong>in</strong>iscule 10-20 m<br />
<strong>in</strong> length, but relatively densely marked by artifacts and debris where shallow buried occupation<br />
levels are erod<strong>in</strong>g (Figure 27).<br />
Collections <strong>of</strong> artifacts and debris obta<strong>in</strong>ed <strong>in</strong> <strong>the</strong>se Sal<strong>in</strong>e River sites range from a<br />
s<strong>in</strong>gle item (five sites) to nearly 5,000 items (One Cypress Po<strong>in</strong>t, 3AS286). Mississippi period<br />
sites furnished <strong>the</strong> greatest proportion <strong>of</strong> this material, and our analyses <strong>of</strong> ceramics, arrow<br />
po<strong>in</strong>ts, o<strong>the</strong>r stone tools, and both organic and <strong>in</strong>organic debris have been <strong>the</strong> fundamental<br />
basis for understand<strong>in</strong>g Mississippian exploitative activities <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> (Chapters 6, 8,<br />
Appendix B).<br />
Unit 2 is a unique and significant group <strong>of</strong> 40 prehistoric sites, all buried and none<br />
plowed or disturbed to any extent except by processes <strong>of</strong> bankl<strong>in</strong>e erosion. No o<strong>the</strong>r site<br />
unit <strong>in</strong> <strong>the</strong> Felsenthal Project area has produced <strong>the</strong> quantity and quality <strong>of</strong> data represented<br />
by this group. Extensive fur<strong>the</strong>r <strong>in</strong>vestigation and measures <strong>of</strong> mitigation have been<br />
proposed for eight sites <strong>in</strong> Unit 2 (Appendix C). In <strong>the</strong> next chapter we use test excavation<br />
data from Sal<strong>in</strong>e River sites to <strong>in</strong>vestigate functional characteristics <strong>of</strong> floodpla<strong>in</strong> extractive<br />
camps and also to exam<strong>in</strong>e variation among extractive sites <strong>of</strong> <strong>the</strong> Mississippi period.
Variability Among Floodpla<strong>in</strong> Sites 129<br />
Figure 27. Mississippi period extractive sites on lower Sal<strong>in</strong>e River. a. Jug Po<strong>in</strong>t 1<br />
(3AS306); b. One Cypress Po<strong>in</strong>t (3AS286) (AAS neg. 797088, 797133).<br />
b<br />
A
130 Hemm<strong>in</strong>gs<br />
Unit 3: Lower Eagle Creek<br />
All eight sites recorded <strong>in</strong> this unit are clustered near <strong>the</strong> mouth <strong>of</strong> Eagle Creek, although<br />
boat survey was extended throughout <strong>the</strong> accessible 1.7 mile (2.8 km) lower reach<br />
(Figure 26). While Fleetwood (1969) shows Eagle Creek as an abandoned Ouachita River<br />
course, <strong>the</strong> lower channel is so atypically straight from NNW to SSE that some structural<br />
control may be suspected. No Quaternary fault<strong>in</strong>g is known to have occurred <strong>in</strong> this region<br />
(Saucier and Fleetwood 1970:870).<br />
Lower Eagle Creek sites are essentially with<strong>in</strong> <strong>the</strong> slightly elevated natural levee zone<br />
<strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River, and <strong>the</strong>se sites might easily be considered ecologically equivalent <strong>in</strong> location<br />
to Unit 2. The eight recorded sites occurred equally on concave and convex bank-l<strong>in</strong>es,<br />
but are asymmetrically distributed with regard to left and right banks (2/6); two sites on <strong>the</strong><br />
right bank are s<strong>in</strong>gle f<strong>in</strong>ds, however (Appendix A, Table A-1).<br />
The components identified <strong>in</strong> Unit 3 appear to extend <strong>the</strong> dense cluster <strong>of</strong> Mississippi<br />
period components on <strong>the</strong> adjacent reach <strong>of</strong> Sal<strong>in</strong>e River. Unit 3 components <strong>in</strong>clude <strong>the</strong><br />
follow<strong>in</strong>g (Appendix A, Table A-1): Mississippi/Caney Bayou-one, Mississippi/Gran <strong>Marais</strong>one,<br />
Mississippi-four, and unknown-two. These components are generally shallow, small<br />
<strong>in</strong> size, with moderate amounts <strong>of</strong> artifacts and debris erod<strong>in</strong>g out, <strong>in</strong>clud<strong>in</strong>g occasional<br />
organic rema<strong>in</strong>s. Three <strong>of</strong> <strong>the</strong>se small but substantial sites, 3BR65, 3BR66, and 3BR79, have<br />
been recommended for mitigation (Appendix C).<br />
As <strong>in</strong> <strong>the</strong> case <strong>of</strong> Unit 2, <strong>the</strong> lower Eagle Creek components are largely undisturbed<br />
and may have high potential for <strong>in</strong>vestigat<strong>in</strong>g <strong>the</strong> nature <strong>of</strong> floodpla<strong>in</strong> extractive sites.<br />
Unit 4: Ouachita River below Mile 254<br />
The 20 prehistoric sites <strong>in</strong> Unit 4 occur along <strong>the</strong> lower 11.7 mile (18.8 km) stretch <strong>of</strong><br />
Ouachita River below <strong>the</strong> mouth <strong>of</strong> Sal<strong>in</strong>e River (Figure 28). The river here is just slightly<br />
less s<strong>in</strong>uous than <strong>the</strong> reach immediately upstream (Unit 1), proceed<strong>in</strong>g 6.0 miles (9.7 km) <strong>in</strong><br />
<strong>the</strong> downvalley direction. The prom<strong>in</strong>ent bends <strong>of</strong> <strong>the</strong> present channel are found also on <strong>the</strong><br />
Government Land Office maps <strong>of</strong> 1827 and later years, and on Dunbar’s 1804-1805 map with<br />
<strong>the</strong> exception <strong>of</strong> one bend (at mile 251.0) on <strong>the</strong> latter. The omission <strong>of</strong> this unnamed bend<br />
on <strong>the</strong> Dunbar map appears merely to be a discrepancy ra<strong>the</strong>r than rapid development <strong>of</strong> a<br />
meander between 1805 and 1827. The floodpla<strong>in</strong> <strong>in</strong> Unit 4 is characterized by low, narrow<br />
natural levees adjo<strong>in</strong><strong>in</strong>g <strong>the</strong> stable channel, and extensive backswamp areas with bayous<br />
and open lakes.<br />
Table 8 shows a distribution <strong>of</strong> bankl<strong>in</strong>e sites <strong>in</strong> Unit 4 much like that <strong>of</strong> <strong>the</strong> lower Sal<strong>in</strong>e<br />
River, but less dense. Sites were recorded most frequently on concave bankl<strong>in</strong>es and on<br />
crossover po<strong>in</strong>ts between bends, as expected. There is ra<strong>the</strong>r marked asymmetry <strong>in</strong> <strong>the</strong>
Variability Among Floodpla<strong>in</strong> Sites 131<br />
Figure 28. Site Unit 4, Ouachita River below mile 254, with prehistoric sites<br />
(dots) located <strong>in</strong> relation to channel geometry. Historic sites (rectangles)<br />
or historic sites with prehistoric components (rectangles<br />
enclos<strong>in</strong>g dots) <strong>in</strong> Unit 8 are also shown on this map.
132 Hemm<strong>in</strong>gs<br />
occurrence <strong>of</strong> sites on left and right banks (14/6). We cannot presently attribute this asymmetry<br />
ei<strong>the</strong>r to survey <strong>in</strong>tensity or to a shift <strong>in</strong> river channel. The dearth <strong>of</strong> sites on <strong>the</strong> right<br />
bank above Lapile Creek was noted while boat survey proceeded, and was checked by a second<br />
boat survey team without positive results. Given <strong>the</strong> extreme low visibility <strong>of</strong> riverbank<br />
sites <strong>in</strong> this area, we would expect some sites to be recorded here by future efforts under<br />
ideal conditions <strong>of</strong> bankl<strong>in</strong>e survey.<br />
Riverbank sites <strong>in</strong> Unit 4 compare <strong>in</strong> small size and sparse content to our sample from<br />
Unit 1 particularly. Only one site, Marie Sal<strong>in</strong>e (3AS329), exceeds 50 m <strong>in</strong> length. This multicomponent<br />
stratified site is reported <strong>in</strong> detail <strong>in</strong> Chapter 6, based on text excavation results.<br />
Ceramics are relatively well represented <strong>in</strong> collections from Unit 4 sites, while all o<strong>the</strong>r categories<br />
<strong>of</strong> tools and debris are rare (Appendix A, Table A-1). Five <strong>of</strong> <strong>the</strong> 20 sites recorded are<br />
s<strong>in</strong>gle f<strong>in</strong>ds. In Appendix C five ceramic sites with relatively substantial artifact assemblages<br />
(3AS287, 3AS321, 3AS322, 3AS327, 3UN168) have been recommended for mitigation, <strong>in</strong> addition<br />
to extensive fur<strong>the</strong>r work at Marie Sal<strong>in</strong>e (3AS329).<br />
In contrast to results <strong>of</strong> survey <strong>in</strong> Unit 1, few Archaic and many later ceramic sites<br />
were recorded <strong>in</strong> Unit 4. This is certa<strong>in</strong>ly attributable to low bank height and limited exposure<br />
<strong>of</strong> deeper alluvium (Figure 29). Mississippi period components are also much less<br />
frequent than observed for Unit 2. The follow<strong>in</strong>g 23 components from 20 sites <strong>in</strong> Unit 4 are<br />
identified:<br />
Mississippi/Caney Bayou 1<br />
Mississippi/Gran <strong>Marais</strong> 2<br />
Mississippi/Caddo II 1<br />
Mississippi 1<br />
Baytown-Coles Creek 9<br />
Tchula/Coon Island 2<br />
Poverty Po<strong>in</strong>t/Calion 1<br />
Late Archaic 1<br />
Unknown 5<br />
The small buried riverbank sites <strong>of</strong> Unit 4 are a significant sample <strong>of</strong> floodpla<strong>in</strong> sites<br />
<strong>in</strong> <strong>the</strong> region. The depths <strong>of</strong> relatively many components are known (0.1 to 2.0 m below<br />
<strong>the</strong> modern levee surface), and correlate well with typological ages <strong>of</strong> <strong>the</strong>se components.<br />
In Chapter 6 we argue that stratified site data from Marie Sal<strong>in</strong>e (3AS329) are fundamental<br />
to understand<strong>in</strong>g prehistory <strong>in</strong> <strong>the</strong> Felsenthal region. Unit 4 sites are essentially <strong>the</strong> lowest<br />
known prehistoric sites <strong>in</strong> <strong>the</strong> state <strong>of</strong> <strong>Arkansas</strong>, by reason <strong>of</strong> <strong>the</strong>ir occurrence at <strong>the</strong> downstream<br />
end <strong>of</strong> <strong>the</strong> project area and near <strong>the</strong> Ouachita River cross<strong>in</strong>g <strong>of</strong> <strong>the</strong> state l<strong>in</strong>e.<br />
Unit 5: Lower Lapile Creek<br />
The eight prehistoric sites <strong>in</strong> this unit lie along <strong>the</strong> marg<strong>in</strong>s <strong>of</strong> Lapile Creek just upstream<br />
from <strong>the</strong> Recent floodpla<strong>in</strong> <strong>of</strong> Ouachita River (Figure 30). This location represents a
Variability Among Floodpla<strong>in</strong> Sites 133<br />
Figure 29. Riverbank sites on Ouachita River. a. Marie Sal<strong>in</strong>e (3AS329); b. mouth<br />
<strong>of</strong> Lapile Creek (3UN123) (AAS neg. 796403, 797124).<br />
A<br />
b
134 Hemm<strong>in</strong>gs<br />
1 km wide tongue <strong>of</strong> low floodpla<strong>in</strong>, bordered north and south by higher Pleistocene terraces<br />
(Fleetwood 1369). Although we regard this unit as an extension <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong><br />
Lowland subarea, it varies somewhat <strong>in</strong> environmental character be<strong>in</strong>g hemmed <strong>in</strong> by<br />
higher terraces and <strong>in</strong> site distribution from all o<strong>the</strong>r areas surveyed.<br />
Lapile Creek sites were resolved (by <strong>the</strong> 50 m rule) from a great many scattered f<strong>in</strong>ds,<br />
amount<strong>in</strong>g <strong>in</strong> some cases to almost cont<strong>in</strong>uous distribution <strong>of</strong> artifacts and debris along<br />
both bankl<strong>in</strong>es. The many project numbers, large site dimensions, and high totals for artifacts<br />
and debris listed <strong>in</strong> Appendix A (Table A-1, Unit 5) testify to this complex situation. At<br />
least six <strong>of</strong> <strong>the</strong>se sites <strong>in</strong>clude multiple activity loci which imp<strong>in</strong>ge on or overlap each o<strong>the</strong>r.<br />
More than one component is identified <strong>in</strong> two sites, based on survey collections available,<br />
but o<strong>the</strong>rs may rema<strong>in</strong> unidentified. The Lapile Creek bankl<strong>in</strong>e is today heavily used by<br />
fishermen, extensively denuded <strong>of</strong> ground cover, and may be subject to casual artifact collect<strong>in</strong>g<br />
activity. In addition to <strong>the</strong> complex problems <strong>of</strong> density and distribution <strong>of</strong> artifact<br />
scatters along lower Lapile Creek, we are not satisfied with <strong>the</strong> field records furnished by<br />
our survey team <strong>in</strong> two respects: precision <strong>of</strong> l<strong>in</strong>ear dimensions with<strong>in</strong> and between scatters,<br />
and absence <strong>of</strong> depth observations for buried material. Therefore, <strong>the</strong> record <strong>of</strong> prehistoric<br />
sites <strong>in</strong> Unit 5 is a prelim<strong>in</strong>ary one, not completely understood, and <strong>in</strong> need <strong>of</strong> additional<br />
carefully designed survey work. In particular, this survey work could be extended upstream<br />
from <strong>the</strong> known site concentration.<br />
The follow<strong>in</strong>g 10 components were identified from eight sites recorded <strong>in</strong> Unit 5:<br />
Mississippi 1<br />
Baytown-Coles Creek 5<br />
Archaic 2<br />
Unknown 2<br />
Two Unit 5 sites <strong>of</strong> relatively large size and substantial content have been recommended<br />
for mitigation (3UN151 and 3UN157, <strong>in</strong> Appendix C). Ceramics, stone tools, and <strong>in</strong>organic<br />
debris <strong>of</strong> several k<strong>in</strong>ds, <strong>in</strong>clud<strong>in</strong>g daub, are well represented <strong>in</strong> <strong>the</strong> collections from <strong>the</strong>se<br />
sites.<br />
Lower Lapile Creek sites recorded by us are located <strong>in</strong> close proximity to many substantial<br />
habitation sites on <strong>the</strong> elevated Pleistocene terraces (cf. Rol<strong>in</strong>gson 1972). One <strong>in</strong>terest<strong>in</strong>g<br />
aspect <strong>of</strong> <strong>the</strong>se floodpla<strong>in</strong> sites is <strong>the</strong>ir functional relationship to upland settlements.<br />
Intensive use <strong>of</strong> <strong>the</strong> Lapile Creek channel marg<strong>in</strong>, even though subject to frequent prolonged<br />
overflow, is <strong>in</strong>dicated by exist<strong>in</strong>g site data.<br />
Unit 6: Oxbow Lakes and Backswamp<br />
Unit 6 comprises 11 prehistoric sites <strong>in</strong> both subareas which occur <strong>in</strong> low backswamp<br />
areas, more or less remote from <strong>the</strong> modern channel and levee <strong>of</strong> Ouachita River (Figure 31).
Variability Among Floodpla<strong>in</strong> Sites 135<br />
Figure 30. Site Unit 5, lower Lapile Creek, with prehistoric sites (dots) located <strong>in</strong><br />
relation to channel geometry. Two mound sites on <strong>the</strong> adjacent terrace,<br />
Locust Ridge (3UN8) and Shallow Lake (3UN9/52) <strong>in</strong> Unit 7 are also<br />
shown on this map.
136 Hemm<strong>in</strong>gs<br />
The extent and difficulty <strong>of</strong> backswamp survey were noted earlier <strong>in</strong> this chapter. Except to<br />
note that exceed<strong>in</strong>gly small, sparse, artifact scatters were recorded, it is not possible to generalize<br />
from <strong>the</strong>se backswamp sites. One Mississippi period and two Baytown-Coles Creek<br />
period components were identified; eight o<strong>the</strong>r prehistoric components rema<strong>in</strong> unidentified<br />
(Appendix A, Table A-1).<br />
Although backswamp sites are few, difficult to locate, and usually nondescript <strong>in</strong> content,<br />
we do not assume that all such sites are <strong>in</strong>significant. Small fish<strong>in</strong>g stations, <strong>in</strong> particular,<br />
could be represented at favorable locations on oxbow lakes and bayous. Sites recorded<br />
on St. Mary’s Lake and Fishtrap Lake (<strong>the</strong> lower bayou <strong>of</strong> Lapile Creek) may represent such<br />
stations. Two sites on Fishtrap Lake (3UN<strong>17</strong>9 and 3UN180) have been recommended for<br />
mitigation (Appendix C).<br />
Unit 7: Pleistocene Terraces and Islands<br />
The 27 sites <strong>in</strong>cluded <strong>in</strong> this unit occur on elevated Pleistocene terraces or terrace<br />
remnants (“islands”) above <strong>the</strong> Recent floodpla<strong>in</strong>, and <strong>the</strong>ir discovery or re<strong>in</strong>vestigation was<br />
secondary to <strong>the</strong> survey objectives discussed earlier <strong>in</strong> this chapter (see Elevation Constra<strong>in</strong>t<br />
<strong>in</strong> Chapter 1). Project records for <strong>the</strong>se sites are summarized <strong>in</strong> Appendix A, Table A-1, and<br />
are discussed briefly here. Detailed comparison among <strong>the</strong>se sites or with floodpla<strong>in</strong> site<br />
data is beyond <strong>the</strong> scope <strong>of</strong> this report.<br />
Major sites on elevated terrace surfaces are rarely subject to overflow, and are not alluviated<br />
as are all floodpla<strong>in</strong> sites. Depth <strong>of</strong> material <strong>in</strong> <strong>the</strong>se sites is not well known, but early<br />
components may underlie later ones as <strong>the</strong> result <strong>of</strong> local colluvial and cultural deposition.<br />
Mound and midden complexes at seven terrace or island locations were noted <strong>in</strong> Chapter<br />
3 (3BR4, 3BR8, 3UN18, 3UN8, 3UN9/52, 3AS1, and 3AS6). Well dra<strong>in</strong>ed silt or sandy loams<br />
characterize <strong>the</strong> terrace edges, <strong>in</strong> contrast to poorly dra<strong>in</strong>ed clay loams <strong>of</strong> <strong>the</strong> Recent floodpla<strong>in</strong>.<br />
Soil conditions and overflow characteristics ensure <strong>the</strong> growth <strong>of</strong> oak-p<strong>in</strong>e forest on<br />
terraces and islands, as opposed to <strong>the</strong> bottomland hardwood forest (Chapter 2).<br />
From <strong>the</strong> report <strong>of</strong> Rol<strong>in</strong>gson (1972) and to a lesser extent from our own observations<br />
(Appendix A, Table A-1), it is clear that a wide variety <strong>of</strong> prehistoric sites occupy terrace<br />
locations <strong>in</strong> <strong>the</strong> Felsenthal region. Large multicomponent habitation sites are conspicuous,<br />
but smaller components, vary<strong>in</strong>g <strong>in</strong> size and <strong>in</strong>tensity <strong>of</strong> occupation, are much more numerous<br />
and widespread <strong>in</strong> distribution. Systematic survey or test<strong>in</strong>g <strong>in</strong> <strong>the</strong>se small upland sites<br />
is regrettably lack<strong>in</strong>g, and no good basis for assess<strong>in</strong>g significance has been established as<br />
it has for certa<strong>in</strong> large sites (Rol<strong>in</strong>gson and Schambach 1981). The impact <strong>of</strong> surface disturbance<br />
<strong>of</strong> various k<strong>in</strong>ds poses a much greater threat to small terrace surface sites than to<br />
buried floodpla<strong>in</strong> sites, so that additional <strong>in</strong>tensive survey <strong>of</strong> upland areas is badly needed<br />
<strong>in</strong> this region (Appendix C:General Recommendations).
Variability Among Floodpla<strong>in</strong> Sites 137<br />
Figure 31. Site Unit 6, oxbow lakes and backswamp, with prehistoric sites located<br />
<strong>in</strong> relation to floodpla<strong>in</strong> and terrace features.
138 Hemm<strong>in</strong>gs<br />
unit 8: historic sites<br />
A total <strong>of</strong> 15 historic sites was recorded for <strong>the</strong> entire project area by comb<strong>in</strong>ed use <strong>of</strong><br />
survey techniques, archival research, <strong>in</strong>formants, and prior site records (Chapter 4). This<br />
very low number for so large an area reflects <strong>the</strong> limited use made <strong>of</strong> overflow bottomlands<br />
dur<strong>in</strong>g <strong>the</strong> historic period and also <strong>the</strong> difficulty <strong>of</strong> locat<strong>in</strong>g small impermanent floodpla<strong>in</strong><br />
camps. In fact we were unable to record any early historic sites (before about <strong>17</strong>80) and can<br />
br<strong>in</strong>g no data to bear on Hypo<strong>the</strong>sis 5 regard<strong>in</strong>g such hunt<strong>in</strong>g or trad<strong>in</strong>g camps (Chapter 1).<br />
The sample <strong>of</strong> later historic sites (after <strong>17</strong>80) associated with river<strong>in</strong>e activity and extractive<br />
<strong>in</strong>dustry is discussed <strong>in</strong> relation to Hypo<strong>the</strong>sis 6 later <strong>in</strong> this chapter, although we have very<br />
little pert<strong>in</strong>ent artifact data.<br />
Each <strong>of</strong> <strong>the</strong> 15 historic sites is summarized <strong>in</strong> Appendix A (Table A-2), and is described<br />
briefly below <strong>in</strong> terms <strong>of</strong> observed content, relevant archival or documentary sources, and<br />
research potential.<br />
Ganer Mound (3BR32). This square, flat-topped earth mound (ca. 50 x 50 x 3.5 m) was<br />
built at Stormhole Bend (Pereogee<strong>the</strong> Shoals) about 1910-1912 as a stag<strong>in</strong>g area for construction<br />
<strong>of</strong> Lock No. 7, and presumably to raise <strong>the</strong> lockkeeper’s residence above w<strong>in</strong>ter-spr<strong>in</strong>g<br />
flood stage (Figure 24). Modification <strong>of</strong> Lock No. 6 at Felsenthal <strong>in</strong> 1912 elim<strong>in</strong>ated <strong>the</strong> need<br />
for Lock No. 7, and no fur<strong>the</strong>r improvements were made (Thatcher 1952:19). The government<br />
tract at Stormhole Bend is shown on <strong>the</strong> 1895 Ouachita River hydrographic survey<br />
(Sheet 10). The earth mound itself is shown only on <strong>the</strong> most recent U.S. Geological Survey<br />
topographic map at 1:24,000 (USGS Felsenthal NW, 1978 advance workpr<strong>in</strong>t).<br />
Ganer Mound was first recorded by <strong>Arkansas</strong> Archeological Survey <strong>in</strong> 1971, and was<br />
revisited by our boat survey team. A few oyster shells were collected on <strong>the</strong> mound flanks,<br />
and a massive cast iron mach<strong>in</strong>e component was noted. A modern corrugated metal cab<strong>in</strong><br />
and associated debris occupy <strong>the</strong> mound summit. O<strong>the</strong>r artifacts <strong>of</strong> <strong>the</strong> 1910-1912 construction<br />
period may be present at <strong>the</strong> water’s edge or adjacent to <strong>the</strong> mound and buried by silt<br />
and leaf litter. Ganer Mound perta<strong>in</strong>s to river improvement and eng<strong>in</strong>eer<strong>in</strong>g activity early<br />
<strong>in</strong> this century, and thus has some <strong>in</strong>terest for regional history <strong>of</strong> <strong>the</strong> Ouachita Valley. No<br />
mitigation recommendation has been made for this site, but destruction should be avoided.<br />
Caney Marie Land<strong>in</strong>g (3BR56). This river land<strong>in</strong>g <strong>of</strong> <strong>the</strong> late 1800s and early 1900s<br />
was located on a concave bank at <strong>the</strong> axis <strong>of</strong> a prom<strong>in</strong>ent bend (Figure 24). The place-name,<br />
spelled variously (“<strong>Marais</strong> de Cannes” was recorded by Hunter <strong>in</strong> 1804), once applied to a<br />
cane marsh, <strong>the</strong> riverbend, <strong>the</strong> land<strong>in</strong>g itself, and shoals below <strong>the</strong> land<strong>in</strong>g. We note here <strong>the</strong><br />
location <strong>of</strong> four o<strong>the</strong>r riverboat land<strong>in</strong>gs at <strong>the</strong> axis <strong>of</strong> bends (3BR81, 3UN155, 3UN88, and<br />
3UN122). Apparently this location conferred an advantage by virtue <strong>of</strong> <strong>the</strong> deep scour pool<br />
near <strong>the</strong> concave bank.
Variability Among Floodpla<strong>in</strong> Sites 139<br />
Caney Marie Land<strong>in</strong>g is shown on U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers river survey sheets<br />
and <strong>in</strong>dex maps for 1873, 1895, and 1897, usually as a s<strong>in</strong>gle small onshore rectangle (structure?).<br />
Caney Mary Shoal downstream was described as 8000 feet <strong>in</strong> length, 0.6 feet deep at<br />
low water (U.S. Congress 1874). This document also noted dangerous snags near <strong>the</strong> land<strong>in</strong>g.<br />
We have no map or documentary references for cont<strong>in</strong>ued use <strong>of</strong> <strong>the</strong> land<strong>in</strong>g early <strong>in</strong><br />
<strong>the</strong> twentieth century. However, our boat survey team recovered a 1905 Liberty Head dime<br />
and stoneware bottle sherd on <strong>the</strong> erod<strong>in</strong>g bank, precisely at <strong>the</strong> n<strong>in</strong>eteenth century map<br />
location (Figure 32a, b). A sandstone block erod<strong>in</strong>g out <strong>of</strong> <strong>the</strong> bank was also noted, but no<br />
o<strong>the</strong>r traces <strong>of</strong> <strong>the</strong> structure(s) which may have existed here. Mitigation by data recovery,<br />
<strong>in</strong>clud<strong>in</strong>g additional archival research, is needed for this site which reta<strong>in</strong>s some evidence <strong>of</strong><br />
early Ouachita River commerce (Appendix C).<br />
Goulett Island (3BR8). Goulett Island on <strong>the</strong> Sal<strong>in</strong>e River (Figure 26) has significant<br />
prehistoric rema<strong>in</strong>s known s<strong>in</strong>ce <strong>the</strong> time <strong>of</strong> C. B. Moore (1913), but also a variety <strong>of</strong> historic<br />
structures and rema<strong>in</strong>s which are essentially unstudied. <strong>Arkansas</strong> Archeological Survey revisited<br />
and recorded Moore’s site <strong>in</strong> 1961 and 1971, and our survey team made ano<strong>the</strong>r brief<br />
visit and supplementary site record <strong>in</strong> 1979. However, <strong>the</strong> bulk <strong>of</strong> Goulett Island lies above<br />
our project contour and we were not able to justify a comprehensive search or test<strong>in</strong>g <strong>of</strong> this<br />
important locality.<br />
Goulett Island is <strong>the</strong> first high terra<strong>in</strong>, immediately adjo<strong>in</strong><strong>in</strong>g <strong>the</strong> channel, which one<br />
encounters proceed<strong>in</strong>g upstream on Sal<strong>in</strong>e River. One suspects that historic settlement began<br />
early, or may even have been cont<strong>in</strong>uous, after <strong>the</strong> prehistoric occupation <strong>of</strong> this “p<strong>in</strong>e<br />
island.” However, Government Land Office maps <strong>of</strong> 1827 and 1830 show no settlement or<br />
agricultural fields until Godfrey Land<strong>in</strong>g is reached 3 miles (4.8 km) upstream. David Dale<br />
Owen, <strong>the</strong> geologist, made note <strong>in</strong> 1860 <strong>of</strong> a lignite outcrop <strong>in</strong> <strong>the</strong> bed <strong>of</strong> Sal<strong>in</strong>e River at<br />
Goulett Island which was quarried for fuel by local <strong>in</strong>habitants at low water. Land adjo<strong>in</strong><strong>in</strong>g<br />
<strong>the</strong> river, evidently <strong>in</strong>clud<strong>in</strong>g Goulett Island, was owned by <strong>Arkansas</strong> Governor Elias N.<br />
Conway (served two terms 1852-1860) at this time (Ferguson and Atk<strong>in</strong>son 1966). Sections<br />
23 and 24 <strong>in</strong> T<strong>17</strong>S, R10W, which encompass Goulett Island and adjacent floodpla<strong>in</strong> terra<strong>in</strong>,<br />
rema<strong>in</strong>ed an isolated “panhandle” <strong>of</strong> Bradley County on <strong>the</strong> left bank <strong>of</strong> Sal<strong>in</strong>e River when<br />
Ashley County was formed <strong>in</strong> 1846. Conway’s <strong>in</strong>fluence may have affected this boundary.<br />
Archival research for this early to middle n<strong>in</strong>eteenth century settlement and use <strong>of</strong> Goulett<br />
Island is badly needed.<br />
Stave cutt<strong>in</strong>g (white oak) by “Slavonians” was an important w<strong>in</strong>ter activity <strong>in</strong> <strong>the</strong><br />
lower Sal<strong>in</strong>e River area from about 1890 to 1920 (Chapter 4). Logg<strong>in</strong>g <strong>of</strong> cypress first, and<br />
<strong>the</strong>n p<strong>in</strong>e and hardwoods, began <strong>in</strong> <strong>the</strong> early 1800s; logg<strong>in</strong>g cont<strong>in</strong>ues today <strong>in</strong> various Sal<strong>in</strong>e<br />
River floodpla<strong>in</strong> and terrace areas (E<strong>the</strong>ridge 1959; Reynolds 1980). This activity is likely<br />
to have <strong>in</strong>volved Goulett Island as a logg<strong>in</strong>g camp and river land<strong>in</strong>g.
140 Hemm<strong>in</strong>gs<br />
Figure 32. Historic artifacts from floodpla<strong>in</strong> sites. a. 1905 Liberty Head dime,<br />
Caney Marie Land<strong>in</strong>g (3BR56); b. stoneware bottle sherd, Caney<br />
Marie Land<strong>in</strong>g (3BR56); c. cast iron cha<strong>in</strong> drive l<strong>in</strong>k, Keelboat Brake<br />
Sawmill (3BR55); d. whiteware plate fragment, Marie Sal<strong>in</strong>e Ferry<br />
(3AS331); e. stoneware jug handle, South <strong>of</strong> Highway 82 Bridge<br />
(3AS300) (AAS neg. 813882).
Variability Among Floodpla<strong>in</strong> Sites 141<br />
Our survey team recorded abundant, diverse historic refuse along <strong>the</strong> erod<strong>in</strong>g riverbank<br />
at Goulett Island, <strong>in</strong>clud<strong>in</strong>g domestic utensils and debris greater than 50 years <strong>in</strong> age.<br />
Only a few selected items were collected (Appendix A, Table A-2). We also observed and<br />
photographed a small p<strong>in</strong>e log barn with hand split boards on gable ends and shake ro<strong>of</strong>,<br />
situated on a low aborig<strong>in</strong>al (?) earth mound near <strong>the</strong> riverbank. U.S. Geological Survey<br />
maps <strong>of</strong> <strong>the</strong> 1930s <strong>in</strong>dicate one centrally located structure and agricultural (?) clear<strong>in</strong>g <strong>of</strong><br />
<strong>the</strong> island. We recommend that Goulett Island be <strong>in</strong>tensively surveyed and tested <strong>in</strong> order<br />
to assess <strong>the</strong> significance <strong>of</strong> <strong>the</strong> complex <strong>of</strong> prehistoric and historic rema<strong>in</strong>s that occur here<br />
(Appendix C). The historic refuse on <strong>the</strong> riverbank is considered significant and is recommended<br />
for mitigation (Appendix C). This locality is evidently slated for development <strong>of</strong><br />
boat access facilities by <strong>the</strong> Felsenthal National Wildlife Refuge (U.S. Fish and Wildlife Service<br />
1979). In Chapter 2 (Potential Paleoecological Studies) we noted also <strong>the</strong> importance <strong>of</strong><br />
<strong>the</strong> geological exposure at Goulett Island.<br />
Prairie Island Land<strong>in</strong>g (3BR81). This land<strong>in</strong>g downstream from Goulett Island on <strong>the</strong><br />
Sal<strong>in</strong>e River also conta<strong>in</strong>s both prehistoric and historic components (Figure 26). The historic<br />
component comprises a po<strong>in</strong>t at <strong>the</strong> axis <strong>of</strong> a bend where a track from Prairie Island<br />
<strong>in</strong>tersects <strong>the</strong> riverbank (Deeter et al. 1925), and where at least one structure is located on<br />
topographic maps <strong>of</strong> <strong>the</strong> 1930s. S<strong>in</strong>ce <strong>the</strong> riverbank lies above our project contour, no <strong>in</strong>tensive<br />
search was made for early historic rema<strong>in</strong>s and only modern fishermen’s debris was<br />
observed.<br />
Prairie Island Land<strong>in</strong>g does not appear on Government Land Office maps <strong>of</strong> <strong>the</strong> early<br />
1800s, and we have no o<strong>the</strong>r n<strong>in</strong>eteenth century Sal<strong>in</strong>e River maps available. Nor does it appear<br />
on C. B. Moore’s map <strong>of</strong> Sal<strong>in</strong>e River, which is, however, deficient <strong>of</strong> most detail (Moore<br />
1913:80). Prairie Island itself, a “p<strong>in</strong>e island” elevated well above <strong>the</strong> floodpla<strong>in</strong>, may be<br />
significant archeologically, historically, and even ecologically, and should be <strong>in</strong>tensively surveyed<br />
(Chapter 2). At <strong>the</strong> present time we have no knowledge <strong>of</strong> significant historic rema<strong>in</strong>s<br />
at <strong>the</strong> riverbank land<strong>in</strong>g, only <strong>in</strong>dications from early twentieth century maps.<br />
Keelboat Brake Sawmill (3BR55). The rema<strong>in</strong>s <strong>of</strong> cypress logg<strong>in</strong>g, a sawmill, and a<br />
tramway lead<strong>in</strong>g to <strong>the</strong> sawmill <strong>in</strong> a backswamp area were po<strong>in</strong>ted out to us by a local<br />
trapper, David Marshall (Figure 24). Cypress was be<strong>in</strong>g rafted out <strong>of</strong> this area, <strong>the</strong> “forks <strong>of</strong><br />
Sal<strong>in</strong>e and Ouachita,” <strong>in</strong> 1855 and probably a decade or two earlier (E<strong>the</strong>ridge 1959:25). Old<br />
cypress butts and tops are ly<strong>in</strong>g along <strong>the</strong> marg<strong>in</strong> <strong>of</strong> Keelboat Brake. This brake shows on<br />
Government Land Office maps <strong>of</strong> 1827 and later years. The sawmill rema<strong>in</strong>s, a burned slab<br />
pile 10 m <strong>in</strong> diameter and about 0.5 m high, were shovel tested and produced only charcoal.<br />
A cast iron, drive cha<strong>in</strong> l<strong>in</strong>k was collected on <strong>the</strong> surface nearby (Figure 32c), and o<strong>the</strong>r artifacts<br />
and debris may be silted over <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity. The low tramway road extend<strong>in</strong>g many<br />
miles north and nor<strong>the</strong>ast to <strong>the</strong> Chicago, Rock Island, and Pacific rail l<strong>in</strong>e shows up well on<br />
1951 air photos (Larance 1961:Sheet 39). This sawmill operation was evidently very small <strong>in</strong>
142 Hemm<strong>in</strong>gs<br />
scale and was used decades later than <strong>the</strong> period <strong>of</strong> cypress raft<strong>in</strong>g, perhaps <strong>in</strong> <strong>the</strong> early<br />
1900s. Most hardwood logs could have been skidded or hauled out <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> via <strong>the</strong><br />
tramway. The sawmill and associated rema<strong>in</strong>s at Keelboat Brake are potentially significant<br />
to <strong>the</strong> history <strong>of</strong> <strong>the</strong> development <strong>of</strong> forestry <strong>in</strong> this region (Reynolds 1980). We have recommended<br />
mitigation <strong>of</strong> <strong>the</strong> site, s<strong>in</strong>ce it perta<strong>in</strong>s to regional research questions concern<strong>in</strong>g<br />
extractive <strong>in</strong>dustry (Appendix C).<br />
Ouachita Belle Land<strong>in</strong>g (3UN155). This riverboat land<strong>in</strong>g is known presently from<br />
map data only. The U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers’ hydrographic survey for 1873 (Sheet 11)<br />
identifies <strong>the</strong> land<strong>in</strong>g on <strong>the</strong> concave bank <strong>of</strong> Five Mile Bend, <strong>in</strong>dicat<strong>in</strong>g possibly an <strong>of</strong>fshore<br />
land<strong>in</strong>g stage or raft. Choice <strong>of</strong> this land<strong>in</strong>g location probably reflects <strong>the</strong> occurrence<br />
<strong>of</strong> a deep scour pool along <strong>the</strong> concave bank and <strong>the</strong> presence <strong>of</strong> high terra<strong>in</strong> less than a<br />
mile southwestward (Figure 24). Our survey team found no surface <strong>in</strong>dication <strong>of</strong> <strong>the</strong> 1870s<br />
steamboat land<strong>in</strong>g.<br />
The Ouachita Belle <strong>of</strong> <strong>the</strong> Blanks L<strong>in</strong>e, largest steamboat on <strong>the</strong> river <strong>in</strong> <strong>the</strong> 1870s, was<br />
a sidewheeler 250 feet long, 67.5 feet at <strong>the</strong> beam, draw<strong>in</strong>g 8.5 feet <strong>of</strong> water (Thatcher 1952).<br />
She made <strong>the</strong> New Orleans-Camden run only dur<strong>in</strong>g high water, and was considered <strong>the</strong><br />
“pride <strong>of</strong> <strong>the</strong> Ouachita.” The land<strong>in</strong>g site should be monitored and protected from surface<br />
disturbance, s<strong>in</strong>ce undiscovered rema<strong>in</strong>s may be present or may be revealed by ongo<strong>in</strong>g<br />
erosion.<br />
Fletchers Land<strong>in</strong>g and Woodyard (3UN88). This mussel shell concentration on <strong>the</strong><br />
riverbank was recorded by <strong>Arkansas</strong> Archeological Survey <strong>in</strong> 1971 and revisited by our<br />
survey team. U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers’ hydrographic surveys for 1873 and 1895 <strong>in</strong>dicate<br />
<strong>the</strong> presence <strong>of</strong> a land<strong>in</strong>g and woodyard here with one <strong>of</strong>fshore land<strong>in</strong>g stage or raft (?).<br />
The location on <strong>the</strong> concave bank at <strong>the</strong> axis <strong>of</strong> <strong>the</strong> bend, and <strong>in</strong> proximity to higher terra<strong>in</strong><br />
southwest (Figure 24), is similar to that noted above for Ouachita Belle Land<strong>in</strong>g.<br />
The only rema<strong>in</strong>s known at 3UN88 are erod<strong>in</strong>g shallow lenses <strong>of</strong> mussel shell with<br />
rare pebbles and rocks, none attributable to prehistoric occupation. Four or more species <strong>of</strong><br />
mussel have been identified (Appendix B). Modern fishcamp debris and logg<strong>in</strong>g roads and<br />
clear<strong>in</strong>gs are also present. The shell lenses are discont<strong>in</strong>uous over an area <strong>of</strong> about 40 x 60<br />
m adjo<strong>in</strong><strong>in</strong>g <strong>the</strong> riverbank and a small tributary slough or crevasse. These shell patches are<br />
clearly not modern, but are not certa<strong>in</strong>ly associated with <strong>the</strong> late n<strong>in</strong>eteenth century river<br />
land<strong>in</strong>g ei<strong>the</strong>r. This site pr<strong>in</strong>cipally occupies <strong>the</strong> levee crest above our project contour, and<br />
was not extensively searched or tested. We recommend cont<strong>in</strong>ued monitor<strong>in</strong>g and protection<br />
from surface disturbance.<br />
Brown Camp (3UN92). The historic site known as Brown Camp or Hogan Property<br />
was recorded by <strong>Arkansas</strong> Archeological Survey <strong>in</strong> 1971 and revisited by our survey team
Variability Among Floodpla<strong>in</strong> Sites 143<br />
which also recorded prehistoric rema<strong>in</strong>s (Figure 28). Brown Camp consists <strong>of</strong> a large earth<br />
mound surmounted by a log house near <strong>the</strong> riverbank, a borrow pit, and a second smaller<br />
earth mound about 150 m southward. The owner <strong>of</strong> this property, <strong>in</strong>clud<strong>in</strong>g adjacent boat<br />
launch<strong>in</strong>g facilities and riverbank lots, is Mrs. Louise M. Hogan <strong>of</strong> Little Rock, <strong>Arkansas</strong>. Accord<strong>in</strong>g<br />
to our <strong>in</strong>formant, Mr. Bud Poole, who operates <strong>the</strong> boat land<strong>in</strong>g, <strong>the</strong>se mounds were<br />
built about 1930 when <strong>the</strong> approaches to <strong>the</strong> Highway 82 bridge were under construction,<br />
and <strong>the</strong> log house was built soon <strong>the</strong>reafter. Brown Camp lies above our project contour, and<br />
was not exam<strong>in</strong>ed closely by <strong>the</strong> survey team. We did note prehistoric artifacts on <strong>the</strong> slopes<br />
<strong>of</strong> <strong>the</strong> larger mound which are redeposited with fill from <strong>the</strong> borrow pit nearby. At <strong>the</strong> present<br />
time nei<strong>the</strong>r <strong>the</strong> historic structures nor <strong>the</strong> prehistoric artifacts are considered to have<br />
research potential, although one or more buried prehistoric components, partly disturbed,<br />
may be present. Additional construction on this site should be preceded by subsurface test<strong>in</strong>g.<br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g (3AS299) and associated historic sites. The most important historic<br />
site <strong>in</strong> <strong>the</strong> Felsenthal Project area <strong>in</strong> terms <strong>of</strong> n<strong>in</strong>eteenth century settlement and commerce<br />
was Marie Sal<strong>in</strong>e Land<strong>in</strong>g on <strong>the</strong> left bank <strong>of</strong> <strong>the</strong> Ouachita River where U.S. Highway<br />
82 now crosses (Figure 28). This land<strong>in</strong>g is listed <strong>in</strong> an early steamboat directory as hav<strong>in</strong>g<br />
a population <strong>of</strong> 300 (Lloyd 1856). Habitation and farm<strong>in</strong>g, however, would have been on<br />
higher terra<strong>in</strong> nearby. The history <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g was recounted briefly <strong>in</strong> Chapter<br />
4. Here we describe <strong>the</strong> limited historic rema<strong>in</strong>s which were recorded by our survey team at<br />
or near <strong>the</strong> land<strong>in</strong>g location shown on historic maps, and note <strong>the</strong> significance <strong>of</strong> <strong>the</strong> locale.<br />
Site 3AS299 is directly identified with <strong>the</strong> land<strong>in</strong>g, but represents only a small part <strong>of</strong><br />
<strong>the</strong> area once utilized on and near <strong>the</strong> riverbank. It consists <strong>of</strong> a canal, 4.5 m wide, extend<strong>in</strong>g<br />
200 m due east from <strong>the</strong> river, partly obscured by a mud bar at <strong>the</strong> mouth. Waterlogged<br />
tree trunks, limbs, and debris are submerged <strong>in</strong> its shallow depression. This canal seems<br />
to be <strong>in</strong>dicated on <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers hydrographic survey (1871, Sheet 6),<br />
reproduced as our Figure 16; one large rectangular structure and two smaller structures<br />
(?) were evidently situated near <strong>the</strong> canal’s north bank. E<strong>the</strong>ridge (1959:136) briefly noted<br />
this canal “lead<strong>in</strong>g from <strong>the</strong> river to each [warehouse] so that boats might reach [<strong>the</strong>m] for<br />
load<strong>in</strong>g and unload<strong>in</strong>g.” The exact date and details <strong>of</strong> canal construction are not presently<br />
known. There is a good possibility that submerged artifacts, <strong>in</strong>clud<strong>in</strong>g even wood or o<strong>the</strong>r<br />
organic materials, are preserved <strong>in</strong> <strong>the</strong> mud fill <strong>of</strong> <strong>the</strong> canal depression. The adjacent terra<strong>in</strong><br />
lies above our project contour and was not <strong>in</strong>vestigated extensively or tested. No stand<strong>in</strong>g<br />
rema<strong>in</strong>s <strong>of</strong> warehouses or o<strong>the</strong>r contemporary structures or surface refuse are known for<br />
<strong>the</strong> canal’s vic<strong>in</strong>ity, although buried rema<strong>in</strong>s could be present. Much <strong>of</strong> <strong>the</strong> surround<strong>in</strong>gs are<br />
forested, but dirt roads, clear<strong>in</strong>gs, and modern fishcamps are scattered through this area. As<br />
<strong>in</strong>dicated <strong>in</strong> Appendix A, Table A-2, we regard <strong>the</strong> canal site as a significant vestige <strong>of</strong> Marie<br />
Sal<strong>in</strong>e Land<strong>in</strong>g with excellent potential for regional archeological and historical study. The<br />
canal itself warrants fur<strong>the</strong>r <strong>in</strong>vestigation and mitigation (Appendix C).
144 Hemm<strong>in</strong>gs<br />
First Slough (3AS330) north <strong>of</strong> <strong>the</strong> Highway 82 Bridge is a small shorel<strong>in</strong>e mussel shell<br />
lens with rusted iron, ceramics, and glass among <strong>the</strong> shell. Three mussel species have been<br />
identified (Appendix B). The historic artifacts collected are few and fragmentary, but are<br />
likely to have been contemporary with <strong>the</strong> n<strong>in</strong>eteenth century occupation <strong>of</strong> Marie Sal<strong>in</strong>e<br />
Land<strong>in</strong>g. Several heavy iron spikes and large rusted objects suggest that <strong>the</strong> rema<strong>in</strong>s at First<br />
Slough are not merely domestic refuse, but perta<strong>in</strong> to riverboat or land<strong>in</strong>g construction <strong>of</strong><br />
some sort. This mussel shell lens with artifacts is potentially associated with Marie Sal<strong>in</strong>e<br />
Land<strong>in</strong>g, and should be protected from disturbance.<br />
Marie Sal<strong>in</strong>e Ferry (3AS331), not far downstream from First Slough on <strong>the</strong> same<br />
bankl<strong>in</strong>e (Figure 28), is <strong>the</strong> approximate location <strong>of</strong> a ferry cross<strong>in</strong>g shown on 1873 and 1895<br />
Ouachita River hydrographic survey maps. E<strong>the</strong>ridge (1959:31) <strong>in</strong>dicates that <strong>the</strong> owners <strong>of</strong><br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g were licensed to operate a ferry here <strong>in</strong> 1852. This ferry connected <strong>the</strong><br />
Hamburg and El Dorado roads, and must have been one <strong>of</strong> <strong>the</strong> busiest cross<strong>in</strong>gs <strong>in</strong> <strong>the</strong> region<br />
(Chapter 4). Our survey team observed about 10 sandstone blocks at <strong>the</strong> water’s edge,<br />
and collected one large whiteware sherd (Figure 32d) <strong>of</strong> probable early twentieth century<br />
manufacture on <strong>the</strong> erod<strong>in</strong>g bank. Much modern fishcamp debris is strewn along <strong>the</strong> shorel<strong>in</strong>e<br />
at this location, and surface disturbance is more extensive than elsewhere <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity.<br />
We cannot be certa<strong>in</strong> that <strong>the</strong> rema<strong>in</strong>s noted perta<strong>in</strong> directly to <strong>the</strong> operation <strong>of</strong> <strong>the</strong> ferry, or<br />
that o<strong>the</strong>r rema<strong>in</strong>s <strong>of</strong> ferry and land<strong>in</strong>g activity are not present. As with o<strong>the</strong>r areas <strong>of</strong> activity<br />
associated with Marie Sal<strong>in</strong>e Land<strong>in</strong>g, this site should be protected and may also warrant<br />
more <strong>in</strong>tensive survey.<br />
Site 3AS300, between <strong>the</strong> U.S. Highway 82 Bridge and <strong>the</strong> Marie Sal<strong>in</strong>e canal (3AS299),<br />
consists <strong>of</strong> a small mussel shell lens on <strong>the</strong> riverbank with an early twentieth century s<strong>of</strong>t<br />
dr<strong>in</strong>k bottle buried among <strong>the</strong> shell and also one stoneware jug handle found on <strong>the</strong> erod<strong>in</strong>g<br />
bank nearby (Figure 32e). There is no basis for attribut<strong>in</strong>g <strong>the</strong>se rema<strong>in</strong>s directly to activity<br />
at Marie Sal<strong>in</strong>e Land<strong>in</strong>g except that it falls with<strong>in</strong> <strong>the</strong> land<strong>in</strong>g area generally delimited by<br />
n<strong>in</strong>eteenth century hydrographic survey maps. Modern fishcamps and debris are extensive<br />
along this shorel<strong>in</strong>e. This small site is potentially significant if it can be l<strong>in</strong>ked with <strong>the</strong> history<br />
<strong>of</strong> <strong>the</strong> land<strong>in</strong>g, and on that basis warrants protection and additional study.<br />
Our study <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g rests on results <strong>of</strong> riverbank survey and prelim<strong>in</strong>ary<br />
search <strong>of</strong> historical sources. Except for <strong>the</strong> canal depression, all <strong>of</strong> <strong>the</strong> natural levee <strong>of</strong><br />
Ouachita River, where n<strong>in</strong>eteenth century land<strong>in</strong>g structures and roads were situated, lies<br />
several feet above our project contour. We were thus unable to perform <strong>in</strong>tensive survey or<br />
test<strong>in</strong>g on this levee. The same may be said for elevated Pleistocene terrace surface, knownlocally<br />
as Parker Ridge, just eastward <strong>of</strong> <strong>the</strong> river, where portions <strong>of</strong> <strong>the</strong> Marie Sal<strong>in</strong>e settle-
Variability Among Floodpla<strong>in</strong> Sites 145<br />
ment must have stood. Much <strong>of</strong> <strong>the</strong> locale <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g is today traversed by<br />
a highway, borrow pits, and dirt roads, but much more rema<strong>in</strong>s forested. The canal itself<br />
(3AS299) appears to be partly filled, but o<strong>the</strong>rwise undisturbed. Submerged rema<strong>in</strong>s may be<br />
expected with<strong>in</strong> its depression. We recommend that a program <strong>of</strong> <strong>in</strong>tensive survey and test<strong>in</strong>g,<br />
encompass<strong>in</strong>g <strong>the</strong> extent <strong>of</strong> sites and structures associated with Marie Sal<strong>in</strong>e Land<strong>in</strong>g,<br />
be implemented <strong>in</strong> <strong>the</strong> near future. Additional construction on <strong>the</strong> Highway 82 bridge and<br />
eastern approach, but especially a proposed “Crossett Harbor” project, pose serious threats<br />
to <strong>the</strong> rema<strong>in</strong>s <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers 1979). In addition to<br />
<strong>the</strong> historic rema<strong>in</strong>s, we have recorded two small prehistoric sites (3AS301, 3AS319) on <strong>the</strong><br />
riverbank with<strong>in</strong> this area (Figure 20). The four historic and two prehistoric riverbank sites<br />
<strong>in</strong> this reach <strong>of</strong> Ouachita River were all previously unknown.<br />
Artesian Well (3UN<strong>17</strong>4). An artesian well <strong>in</strong> <strong>the</strong> Ouachita River floodpla<strong>in</strong> near Open<br />
Lake (Figure 28) is <strong>in</strong>dicated on U.S. Geological survey maps <strong>of</strong> 1934 and later years. Our<br />
survey team visited <strong>the</strong> location <strong>in</strong> <strong>the</strong> course <strong>of</strong> complet<strong>in</strong>g Transect 1. The well site consists<br />
<strong>of</strong> a low, cyl<strong>in</strong>drical, brick and concrete tank (some <strong>in</strong>scribed bricks were recorded; see site<br />
form for 3UN<strong>17</strong>4). A metal shed and modern debris are also present near this brick structure.<br />
The date <strong>of</strong> development and <strong>in</strong>tended use <strong>of</strong> this well are unknown, although local<br />
drillers’ records might conta<strong>in</strong> such <strong>in</strong>formation. This well does not appear <strong>in</strong> a compilation<br />
<strong>of</strong> wells and spr<strong>in</strong>gs by Veatch (1906:Chapter 5). Ins<strong>of</strong>ar as it is now known, this well site<br />
does not appear to have historical significance.<br />
<strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g (3UN122). This late n<strong>in</strong>eteenth century riverboat land<strong>in</strong>g, at or<br />
just above <strong>the</strong> axis <strong>of</strong> Coon Glory Bend on <strong>the</strong> right concave bank (Figures 16, 28) was first<br />
recorded by <strong>Arkansas</strong> Archeological Survey <strong>in</strong> 1976, but its nature and significance were<br />
only established by our work with various historical sources. U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers’<br />
hydrographic survey maps for 1871, 1873, 1895, and 1897 show “<strong>Grand</strong> Mary Land<strong>in</strong>g,”<br />
usually with one or two riverbank structures just above <strong>the</strong> axis <strong>of</strong> <strong>the</strong> bend. The deep scour<br />
pool would have provided sufficient water for steamboats to tie up along <strong>the</strong> concave bank.<br />
Higher, permanently settled terra<strong>in</strong> (present-day Felsenthal), was less than 2 km southwestward,<br />
across <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> backswamp.<br />
An artificial depression, bricks, and mussel shell were reported dur<strong>in</strong>g <strong>the</strong> <strong>in</strong>itial site<br />
visit, but not relocated by our survey team. However, we did exam<strong>in</strong>e, record, and shovel<br />
test a th<strong>in</strong> shallow buried organic lens, 25 x 40 m, partially exposed <strong>in</strong> <strong>the</strong> cutbank. This<br />
lens at about 64.0 feet MSL lies immediately adjacent to <strong>the</strong> upper approach channel to <strong>the</strong><br />
new Felsenthal Lock and Dam, and may be affected by construction, as may o<strong>the</strong>r unknown<br />
buried rema<strong>in</strong>s. One split p<strong>in</strong>e board or stave fragment (Appendix B) was recovered from<br />
a shovel test, but no o<strong>the</strong>r artifacts were recovered here. No surface <strong>in</strong>dications <strong>of</strong> <strong>the</strong> land<strong>in</strong>g’s<br />
structures were noted <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> <strong>the</strong> bend.
146 Hemm<strong>in</strong>gs<br />
Systematic subsurface test<strong>in</strong>g <strong>of</strong> this area was scheduled by us for fall 1979, but was<br />
prohibited by ris<strong>in</strong>g river level (Chapter 6). We believe that o<strong>the</strong>r rema<strong>in</strong>s <strong>of</strong> <strong>Grand</strong> <strong>Marais</strong><br />
Land<strong>in</strong>g may be present below <strong>the</strong> levee surface, and that historical significance <strong>of</strong> <strong>the</strong> site is<br />
established by our test<strong>in</strong>g and archival research (Appendix C). A specific recommendation<br />
to this effect has been submitted to <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers (C. R. McGimsey letter<br />
to Shelia Lewis 29 August 1980).<br />
Wreck <strong>of</strong> <strong>the</strong> Lotawanna (3UN153). The 1874 loss <strong>of</strong> this steamboat <strong>in</strong> <strong>the</strong> Felsenthal<br />
Project area was discovered from historical sources prior to fieldwork, and has subsequently<br />
been confirmed and <strong>in</strong>vestigated by various methods. Only after much o<strong>the</strong>r field and archival<br />
work was undertaken did we acquire a hydrographic survey map which p<strong>in</strong>po<strong>in</strong>ts <strong>the</strong><br />
underwater wreck about 200 m below <strong>the</strong> axis <strong>of</strong> Coon Glory Bend (Figure 33). In September<br />
1979 we <strong>in</strong>terviewed an elderly Felsenthal man, Roland Johnson, who had considerable<br />
knowledge <strong>of</strong> local river land<strong>in</strong>gs and who had seen <strong>the</strong> submerged wreckage <strong>of</strong> Lotawanna<br />
as a child (transcript on file at <strong>Arkansas</strong> Archeological Survey; see also Chapter 4). In May<br />
1980 one <strong>of</strong> us (Hemm<strong>in</strong>gs) <strong>the</strong>n accompanied a mar<strong>in</strong>e survey archeologist who obta<strong>in</strong>ed<br />
magnetometer data for <strong>the</strong> wreck site (a marked anomaly match<strong>in</strong>g <strong>the</strong> map data, presumed<br />
to be <strong>the</strong> Lotawanna <strong>in</strong> about 24 feet <strong>of</strong> water; <strong>in</strong>formation from Allen R. Saltus, Jr., Archeological<br />
Research and Survey, Prairieville, Louisiana). The wreck has not yet been directly<br />
observed.<br />
The underwater Lotawanna wreck site is located nearly with<strong>in</strong> or just below <strong>the</strong> upstream<br />
approach to <strong>the</strong> new Felsenthal Lock and Dam, so far as we can tell now. Precise<br />
mapp<strong>in</strong>g <strong>of</strong> this wreck <strong>in</strong> relation to river channel and construction right-<strong>of</strong>-way has been<br />
undertaken by Saltus, us<strong>in</strong>g both magnetometer and bathymetric data. We also have exam<strong>in</strong>ed<br />
a 1976 construction plan with bathymetric contours which may <strong>in</strong>dicate <strong>the</strong> wreckage.<br />
It is expected that lock construction and deep <strong>in</strong>undation will have an adverse impact on <strong>the</strong><br />
rema<strong>in</strong>s <strong>of</strong> Lotawanna (Appendix C). On this basis we have recommended to <strong>the</strong> U.S. Army<br />
Corps <strong>of</strong> Eng<strong>in</strong>eers that a qualified diver-archeologist <strong>in</strong>spect <strong>the</strong> wreck site, establish its extent<br />
and <strong>in</strong>tegrity, evaluate impact, and recommend mitigation procedures (C. R. McGimsey<br />
letter to Shelia Lewis August 29, 1980). At <strong>the</strong> present time <strong>the</strong> Lotawanna is <strong>the</strong> only steamboat<br />
wreck, p<strong>in</strong>po<strong>in</strong>ted and archeologically recorded, on <strong>the</strong> Ouachita River <strong>in</strong> <strong>Arkansas</strong>.<br />
eVAluAtion <strong>of</strong> surVey strAtegy And tActics<br />
Intensive archeological survey was undertaken <strong>in</strong> <strong>the</strong> Felsenthal Project area <strong>in</strong> order<br />
to locate cultural resources present, assess <strong>the</strong>ir cultural and scientific importance, and<br />
identify means for mitigat<strong>in</strong>g adverse effects <strong>of</strong> a 65-foot navigation pool and o<strong>the</strong>r related<br />
impacts. A seven-week survey stage was first completed, <strong>in</strong>volv<strong>in</strong>g a successive series <strong>of</strong><br />
survey procedures and target areas <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>; survey activity <strong>the</strong>n cont<strong>in</strong>ued on a<br />
reduced basis dur<strong>in</strong>g a seven-week test excavation stage reported <strong>in</strong> Chapter 6. A systematic
Variability Among Floodpla<strong>in</strong> Sites 147<br />
Figure 33. Hydrographic survey chart <strong>of</strong> Ouachita River <strong>in</strong> <strong>the</strong> Felsenthal Project area, show<strong>in</strong>g location <strong>of</strong> “Wreck <strong>of</strong><br />
Lottie Warner” (Lotawanna) (U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers 1896:Sheet 14).
148 Hemm<strong>in</strong>gs<br />
body <strong>of</strong> survey data for 144 prehistoric and historic sites has been summarized <strong>in</strong> this chapter<br />
and <strong>in</strong> Appendix A. Much new substantive data and a variety <strong>of</strong> new <strong>in</strong>terpretations<br />
have resulted from this work. These results perta<strong>in</strong> to both scientific and cultural resource<br />
management objectives <strong>of</strong> <strong>the</strong> Felsenthal Project, and to many aspects <strong>of</strong> prehistory and<br />
history <strong>in</strong> <strong>the</strong> Felsenthal region. In this section we briefly evaluate our strategy and tactics,<br />
especially <strong>in</strong> regard to survey capability <strong>in</strong> a remote wetland environment where resources<br />
lie buried beneath <strong>the</strong> modern surface.<br />
1. The project area was stratified <strong>in</strong>to two subareas, and one <strong>of</strong> <strong>the</strong>se subdivided<br />
as four sampl<strong>in</strong>g units for <strong>the</strong> purpose <strong>of</strong> random transect survey. This stratification<br />
was based pr<strong>in</strong>cipally on <strong>the</strong> constra<strong>in</strong>t <strong>of</strong> project contour (terra<strong>in</strong> below<br />
65 feet MSL) and secondarily on geohydrological variation between floodpla<strong>in</strong><br />
areas. Ideally, a cont<strong>in</strong>uous block or segment <strong>of</strong> floodpla<strong>in</strong> terra<strong>in</strong> would serve<br />
as <strong>the</strong> project area, ra<strong>the</strong>r than <strong>the</strong> mosaic <strong>of</strong> patches or ribbons <strong>of</strong> terra<strong>in</strong> with<br />
which we have had to deal (see Chapter 1).<br />
2. Environmental variation with<strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> was poorly known at <strong>the</strong> outset<br />
<strong>of</strong> this study and was <strong>in</strong>tegrated <strong>in</strong>to survey strategy with difficulty. Detailed<br />
studies or maps <strong>of</strong> biotic and abiotic variables <strong>in</strong> <strong>the</strong> Felsenthal floodpla<strong>in</strong> do<br />
not exist. Guyton silty clay loam, bottomland hardwood or swamp forest, and<br />
m<strong>in</strong>imal relief characterize <strong>the</strong> entire project area. However, a variety <strong>of</strong> dra<strong>in</strong>age<br />
features <strong>of</strong> geologically Recent orig<strong>in</strong> are present, and are differentiated<br />
generally by Fleetwood (1969). We now perceive significant microenvironmental<br />
variation between <strong>the</strong> <strong>in</strong>ner valley natural levee zone, poorly developed as<br />
it is, and <strong>the</strong> outer valley backswamp zone, and also between <strong>the</strong> upvalley and<br />
downvalley sections <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. Site distributions are believed to reflect<br />
this variation (see Site Locational Hypo<strong>the</strong>ses below).<br />
3. Floodpla<strong>in</strong> transects <strong>in</strong>volved po<strong>in</strong>t sampl<strong>in</strong>g (auger<strong>in</strong>g or shovel test<strong>in</strong>g) to<br />
0.5 or 1.0 m depth and at 50 m <strong>in</strong>tervals. Four transects were randomly located<br />
and n<strong>in</strong>e o<strong>the</strong>rs judgmentally located. These transects largely <strong>in</strong>tersected backswamp<br />
portions <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> and proceeded with difficulty. From results<br />
<strong>of</strong> all floodpla<strong>in</strong> survey work we now know that eight <strong>of</strong> 10 sites are likely to<br />
extend less than 50 m <strong>in</strong> greatest dimension, and also that some sites are buried<br />
at depths greater than 0.5 m. The deficiency <strong>in</strong> survey <strong>in</strong>tensity is obvious. However,<br />
<strong>in</strong>creased <strong>in</strong>tensity and <strong>the</strong>refore cost <strong>of</strong> survey would not be justified on<br />
<strong>the</strong> basis <strong>of</strong> our present understand<strong>in</strong>g <strong>of</strong> backswamp sites and site distribution.<br />
Transects 9 and 10 which provided cont<strong>in</strong>uous exposure <strong>of</strong> sediments to at least<br />
1.0 m depth, 8 km across <strong>the</strong> floodpla<strong>in</strong>, <strong>in</strong>dicate that backswamp sites are few,<br />
small, and sparse <strong>in</strong> content.
Variability Among Floodpla<strong>in</strong> Sites 149<br />
4. Adequate procedure for rapid, extensive, subsurface test<strong>in</strong>g <strong>in</strong> this floodpla<strong>in</strong><br />
context has yet to be developed. We are conv<strong>in</strong>ced that power augers are <strong>in</strong>efficient<br />
devices for locat<strong>in</strong>g sparse rema<strong>in</strong>s to 1.0 m depth, although <strong>the</strong>y penetrated<br />
tough clayey sediments with ease. Shovel test<strong>in</strong>g proved effective to 0.5<br />
or 0.6 m depth, but decisions to employ hundreds <strong>of</strong> time-consum<strong>in</strong>g shovel<br />
tests should be carefully considered. Manually operated posthole diggers,<br />
said to be effective to 1.5 m or more (Wood 1976), are by no means well suited<br />
to tough clays. Lack <strong>of</strong> quick, effective, and nondestructive test<strong>in</strong>g procedure<br />
prevented us from obta<strong>in</strong><strong>in</strong>g precise areal extent <strong>of</strong> buried sites <strong>in</strong> most cases.<br />
Portability is a requirement for any test<strong>in</strong>g device used extensively <strong>in</strong> <strong>the</strong>se<br />
overflow bottoms.<br />
5. Riverbank survey tactics were carefully conceived and consistently carried<br />
out with good results. An understand<strong>in</strong>g <strong>of</strong> meander activity, prograd<strong>in</strong>g,<br />
degrad<strong>in</strong>g, or stable bankl<strong>in</strong>es, alluviation, and o<strong>the</strong>r local floodpla<strong>in</strong> or river<br />
channel processes is essential to design<strong>in</strong>g and evaluat<strong>in</strong>g riverbank surveys.<br />
The procedures we employed <strong>in</strong> <strong>the</strong> Felsenthal Project area would not be well<br />
suited to <strong>the</strong> dynamic floodpla<strong>in</strong> regime <strong>in</strong> <strong>the</strong> Great Bend region <strong>of</strong> <strong>the</strong> Red<br />
River (Hemm<strong>in</strong>gs 1980). We were not able to devise a precise workable method<br />
<strong>of</strong> evaluat<strong>in</strong>g and record<strong>in</strong>g degree <strong>of</strong> bankl<strong>in</strong>e exposure with<strong>in</strong> various channel<br />
reaches. The record <strong>of</strong> small ephemeral riverbank sites <strong>of</strong> low visibility is<br />
presumably strongly biased by exposure.<br />
6. Among o<strong>the</strong>r site survey data, emphasis was placed on record<strong>in</strong>g <strong>the</strong> depth <strong>of</strong><br />
buried artifacts or debris <strong>in</strong> place on prehistoric riverbank sites. About onethird<br />
<strong>of</strong> all floodpla<strong>in</strong> sites furnished depth observations, and about two-thirds<br />
<strong>of</strong> all floodpla<strong>in</strong> sites furnished typological, technological, or stratigraphic data<br />
sufficient to identify cultural components. We are able to advance a sequence <strong>of</strong><br />
floodpla<strong>in</strong> sites and components and general rates <strong>of</strong> floodpla<strong>in</strong> alluviation for<br />
<strong>the</strong> past 3000 years (Chapter 7). We believe this is a significant contribution to<br />
regional prehistory.<br />
7. We do not calculate a sample fraction from <strong>the</strong> extent <strong>of</strong> transect and river-bank<br />
survey and <strong>the</strong> total extent <strong>of</strong> project area. Given <strong>the</strong> mosaiclike distribution<br />
<strong>of</strong> terra<strong>in</strong> below 65 feet elevation, <strong>the</strong> uniform occurrence <strong>of</strong> buried sites, and<br />
<strong>the</strong> tactics <strong>of</strong> survey such a calculation would be <strong>in</strong>tricate and spurious. Never<strong>the</strong>less,<br />
<strong>the</strong> empirical data for distribution <strong>of</strong> archeological sites <strong>in</strong> <strong>the</strong> upper<br />
1.0 meters <strong>of</strong> floodpla<strong>in</strong> sediments, especially <strong>in</strong> <strong>the</strong> stable natural levee zone<br />
<strong>of</strong> Ouachita and Sal<strong>in</strong>e rivers, provides a systematic basis for evaluat<strong>in</strong>g <strong>the</strong>se<br />
resources.
150 Hemm<strong>in</strong>gs<br />
site locAtionAl And historicAl hypo<strong>the</strong>ses<br />
Site survey data, especially those data obta<strong>in</strong>ed by riverbank survey, are relevant to<br />
most <strong>of</strong> <strong>the</strong> hypo<strong>the</strong>ses proposed <strong>in</strong> <strong>the</strong> research design (Chapter 1). We shall here exam<strong>in</strong>e<br />
two hypo<strong>the</strong>ses regard<strong>in</strong>g <strong>the</strong> location <strong>of</strong> prehistoric sites and <strong>the</strong> function <strong>of</strong> historic sites <strong>in</strong><br />
<strong>the</strong> floodpla<strong>in</strong>. The problem doma<strong>in</strong>s and hypo<strong>the</strong>ses were stated as follows:<br />
Problem Doma<strong>in</strong> 2: Locational Characteristics<br />
H 2<br />
If specialized extractive sites were established for resource procurement, <strong>the</strong>se sites<br />
would be located for most efficient acquisition and delivery to base settlements.<br />
Subhypo<strong>the</strong>ses: A. Such sites will correlate with specific resource zones.<br />
B. Such sites will correlate with access routes to base settlements.<br />
C. Such sites will correlate with both factors.<br />
D. None <strong>of</strong> <strong>the</strong> above, but o<strong>the</strong>r unknown factors.<br />
Test Implications: Significant correlations obta<strong>in</strong>ed for location factors; recognition <strong>of</strong><br />
contemporary base settlements, access routes, and resource zones.<br />
Data Requirements: Map distributions; measures <strong>of</strong> spatial relationships; cross-dat<strong>in</strong>g.<br />
Problem Doma<strong>in</strong> 3: Culture History<br />
H 6<br />
If later historical sites are present <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>, <strong>the</strong>y were occupied <strong>in</strong> association<br />
with river<strong>in</strong>e activity and extractive <strong>in</strong>dustry (after <strong>17</strong>80).<br />
Test Implications: Discovery <strong>of</strong> sites with complex technological assemblages.<br />
Data Requirements: Tools or components associated with eng<strong>in</strong>es.<br />
Hypo<strong>the</strong>sis 2 seeks to expla<strong>in</strong> <strong>the</strong> distribution <strong>of</strong> floodpla<strong>in</strong> sites <strong>in</strong> terms <strong>of</strong> how a<br />
resident local prehistoric population might best acquire riparian and aquatic resources. Multiple<br />
variables are <strong>in</strong>volved <strong>in</strong> <strong>the</strong> several subhypo<strong>the</strong>ses and also <strong>in</strong> test data we can br<strong>in</strong>g<br />
to bear, so that elegant pro<strong>of</strong> <strong>of</strong> this hypo<strong>the</strong>sis is not presently possible. We believe that<br />
nonrandom distribution <strong>of</strong> riverbank sites can be established and that locational factors can<br />
at least be systematically exam<strong>in</strong>ed. The most marked departures from random distribution<br />
are relatively high frequencies <strong>of</strong> Archaic components <strong>in</strong> Unit 1 (at least partly due to deep<br />
exposure), <strong>of</strong> Tchula and Baytown-Coles Creek components <strong>in</strong> Unit 4, and <strong>of</strong> Mississippi<br />
period components <strong>in</strong> Unit 2. By far <strong>the</strong> best available site contextual data perta<strong>in</strong> to <strong>the</strong><br />
Mississippi period, where:
Variability Among Floodpla<strong>in</strong> Sites 151<br />
1. visibility <strong>of</strong> shallow buried components is good,<br />
2. more floodpla<strong>in</strong> components can be assigned to this period or its three phases,<br />
3. <strong>the</strong> nature <strong>of</strong> extractive sites has been <strong>in</strong>vestigated (Chapter 6),<br />
4. <strong>the</strong> presumption <strong>of</strong> stable channels and stable levee environment over <strong>the</strong> past<br />
several centuries is relatively strong, and<br />
5. some understand<strong>in</strong>g <strong>of</strong> base settlements on terraces and islands can be advanced.<br />
In order to emphasize <strong>the</strong> strik<strong>in</strong>g concentration <strong>of</strong> Mississippi period components on<br />
<strong>the</strong> lower Sal<strong>in</strong>e River we present some simple density calculations <strong>in</strong> Table 8. On <strong>the</strong> average,<br />
n<strong>in</strong>e such components occur here with<strong>in</strong> a 10 km reach <strong>of</strong> <strong>the</strong> river. We know also from<br />
o<strong>the</strong>r survey and test excavation data (Chapter 6) that multiple activity loci characterize<br />
some <strong>of</strong> <strong>the</strong>se Mississippi period components on <strong>the</strong> lower Sal<strong>in</strong>e River. The densities for<br />
Ouachita River components attributed to this period, however, are markedly lower where<br />
such components have been discovered at all. Our locational hypo<strong>the</strong>sis and any o<strong>the</strong>r that<br />
might be formulated must contend with this dense cluster with<strong>in</strong> a complex regional system<br />
<strong>of</strong> Mississippi period settlements.<br />
Table 8. Density <strong>of</strong> Prehistoric Riverbank Components <strong>in</strong> Three Site Units on Ouachita and<br />
Sal<strong>in</strong>e Rivers.<br />
All Prehistoric L<strong>in</strong>ear Density Mississippi Period L<strong>in</strong>ear Density<br />
Site Unit and Length <strong>of</strong> Unit Components (Components/ Components (Components/<br />
Location (river km) Recorded 10 river km) Identified river km)<br />
unit 1<br />
(Ouachita River<br />
above Mile 254) 27.4 18 6.6 0 0.0<br />
Unit 4<br />
(Ouachita River<br />
below Mile 254) 18.8 23 12.2 5 2.7<br />
unit 2<br />
(Lower Sal<strong>in</strong>e River) <strong>17</strong>.7 41 23.2 16 9.0<br />
Figure 34 illustrates <strong>the</strong> spatial array <strong>of</strong> Mississippi period components <strong>in</strong> or near <strong>the</strong><br />
Felsenthal Project area. In Chapter 3 we proposed an hierarchy <strong>of</strong> Mississippi period settlements<br />
where major and m<strong>in</strong>or mound centers on terraces and islands are reasonably well<br />
known, but where habitation sites (villages, hamlets, and farmsteads) on terraces are very<br />
poorly known. We <strong>the</strong>refore have <strong>in</strong>complete knowledge <strong>of</strong> <strong>the</strong> Mississippian settlement<br />
hierarchy and <strong>of</strong> <strong>the</strong> nature and distribution <strong>of</strong> base settlements, and this bias is reflected <strong>in</strong><br />
Figure 34. The seven mound centers <strong>in</strong>dicated around <strong>the</strong> periphery <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> are
152 Hemm<strong>in</strong>gs<br />
Figure 34. The spatial array <strong>of</strong> Mississippi period sites <strong>in</strong> or near <strong>the</strong> Felsenthal<br />
Project area.
Variability Among Floodpla<strong>in</strong> Sites 153<br />
not merely ceremonial loci for a dispersed population, but evidently were <strong>the</strong>mselves occupied<br />
by a sizable population segment as <strong>in</strong>dicated by extensive dense refuse and, <strong>in</strong> some<br />
cases, multiple cemeteries (Rol<strong>in</strong>gson and Schambach 1981; <strong>Arkansas</strong> Archeological Survey<br />
site records on file). The very large Mississippi mound and midden complex at Watts Field<br />
(3UN18) was illustrated <strong>in</strong> Figure 14.<br />
Inspection <strong>of</strong> <strong>the</strong> spatial array <strong>in</strong> Figure 34 reveals several very <strong>in</strong>terest<strong>in</strong>g characteristics.<br />
First, <strong>the</strong>re is a pattern <strong>of</strong> base settlements arranged around a cluster <strong>of</strong> extractive<br />
camps (our Unit 2), ra<strong>the</strong>r than extractive camps arranged about base settlements. This “centripetal”<br />
pattern is clear with respect to floodpla<strong>in</strong> site data, but lacks data for upland extractive<br />
sites. Second, <strong>the</strong>re may be regularity <strong>in</strong> spac<strong>in</strong>g among Mississippi period settlements;<br />
<strong>the</strong> seven mound centers average about 7 km distance between each pair, and <strong>the</strong> shortest<br />
distance between any mound center and <strong>the</strong> lower Sal<strong>in</strong>e River cluster <strong>of</strong> extractive sites<br />
also averages about 7 km. Variations from <strong>the</strong>se average distances are not especially great,<br />
and are apparent <strong>in</strong> Figure 34. We suggest that cultural factors primarily, not environmental<br />
differences, have operated to produce this spatial array.<br />
Some, and conceivably all, mound centers <strong>in</strong> <strong>the</strong> project area were occupied contemporaneously<br />
dur<strong>in</strong>g <strong>the</strong> Gran <strong>Marais</strong> phase (A.D. 1100-1400), and <strong>in</strong>teracted with<strong>in</strong> a regional<br />
sociopolitical and economic network. At least some <strong>of</strong> <strong>the</strong> larger mound centers seem to<br />
have been occupied also dur<strong>in</strong>g <strong>the</strong> Caney Bayou phase (A.D. 1400-1600). When <strong>the</strong>se base<br />
settlements (or o<strong>the</strong>rs along <strong>the</strong> terrace edges) undertook to exploit floodpla<strong>in</strong> resources<br />
with<strong>in</strong> a radius <strong>of</strong> a few kilometers, no temporary camp would be required; when <strong>the</strong> radius<br />
was extended to approximately 7 km (or more) a work party encamped <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong><br />
to maximize its extractive and process<strong>in</strong>g activity, and m<strong>in</strong>imize unproductive daily travel.<br />
That is to say, a pr<strong>in</strong>ciple <strong>of</strong> least effort for maximum return is reflected <strong>in</strong> <strong>the</strong> spac<strong>in</strong>g <strong>of</strong><br />
floodpla<strong>in</strong> extractive sites and base settlements on adjacent terrace edges. Cooperative procurement<br />
activities among base settlements might also have <strong>in</strong>fluenced this spac<strong>in</strong>g.<br />
Return<strong>in</strong>g now to Hypo<strong>the</strong>sis 2, we would argue that test implications are generally<br />
supported by <strong>the</strong> Mississippi period settlement data. Extractive sites are highly correlated<br />
with a resource zone, <strong>the</strong> river channel and its levee marg<strong>in</strong>s, where important faunal and<br />
floral resources are concentrated (Chapter 2). River channels and <strong>the</strong>ir tributaries provided<br />
easiest access, across <strong>in</strong>terven<strong>in</strong>g backswamp, to base settlements on elevated terraces.<br />
An “unknown factor” is also implied by <strong>the</strong> concentration <strong>of</strong> extractive sites on <strong>the</strong> lower<br />
Sal<strong>in</strong>e River, a relatively remote <strong>in</strong>terior floodpla<strong>in</strong> zone with respect to <strong>the</strong> peripheral r<strong>in</strong>g<br />
<strong>of</strong> base settlements. We tentatively attribute this “centripetal pattern” to resource procurement<br />
efficiency. Recognition <strong>of</strong> this pattern is one <strong>of</strong> <strong>the</strong> most important research results <strong>of</strong><br />
our floodpla<strong>in</strong> survey work. Biases and deficiencies <strong>in</strong> regional Mississippian settlement<br />
data are great, but a productive direction for site survey work and spatial analysis is clearly<br />
<strong>in</strong>dicated.
154 Hemm<strong>in</strong>gs<br />
In Hypo<strong>the</strong>sis 6 we proposed that later historical sites <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> would reflect<br />
river<strong>in</strong>e and extractive activity. The data requirement, tools or components associated with<br />
eng<strong>in</strong>es, is m<strong>in</strong>imally met by <strong>the</strong> metallic rema<strong>in</strong>s collected or observed at Ganer Mound<br />
(3BR32), Keelboat Brake Sawmill (3BR55), and First Slough (3AS330). However, <strong>the</strong> location<br />
and function <strong>of</strong> most o<strong>the</strong>r historic sites is documented by maps and o<strong>the</strong>r historical sources;<br />
steamboat travel, logg<strong>in</strong>g, or o<strong>the</strong>r river<strong>in</strong>e commerce is clearly associated with n<strong>in</strong>e<br />
<strong>of</strong> 15 historic sites recorded and not associated with two sites (Brown Camp [3UN92] and<br />
Artesian Well [3UN<strong>17</strong>4]). The hypo<strong>the</strong>sis is well supported by all available data, but not by<br />
<strong>the</strong> few tools or eng<strong>in</strong>e components we were able to locate and recover <strong>in</strong> floodpla<strong>in</strong> historic<br />
sites.<br />
One o<strong>the</strong>r result <strong>of</strong> historic site survey was <strong>the</strong> correlation <strong>of</strong> river land<strong>in</strong>gs with<br />
meander geometry. These land<strong>in</strong>gs were consistently located on <strong>the</strong> concave bank at <strong>the</strong><br />
axis <strong>of</strong> a bend, <strong>in</strong> order to take advantage <strong>of</strong> <strong>the</strong> deep scour pool adjo<strong>in</strong><strong>in</strong>g this bank. Bends<br />
were selected for <strong>the</strong>ir proximity to upland terra<strong>in</strong>, which m<strong>in</strong>imized <strong>the</strong> difficult cross<strong>in</strong>g<br />
<strong>of</strong> backswamp. From n<strong>in</strong>eteenth century hydrographic surveys and o<strong>the</strong>r maps we believe<br />
<strong>the</strong>se preferred land<strong>in</strong>g locations are repeated upstream on <strong>the</strong> Ouachita and Sal<strong>in</strong>e rivers.
Chapter 6<br />
AssessMent <strong>of</strong> floodplA<strong>in</strong> sites<br />
On October 6, 1979 <strong>the</strong> test<strong>in</strong>g stage <strong>of</strong> <strong>the</strong> Felsenthal Project was <strong>in</strong>itiated with a daylong<br />
tra<strong>in</strong><strong>in</strong>g session for all personnel on test excavation objectives, procedures, and equipment.<br />
Survey activity was to cont<strong>in</strong>ue on a reduced basis on <strong>the</strong> lower Sal<strong>in</strong>e River and <strong>the</strong><br />
Ouachita River below mile 254, consistent with <strong>the</strong> needs <strong>of</strong> site test<strong>in</strong>g. The test<strong>in</strong>g stage<br />
was projected to end <strong>in</strong> mid-December, but on November 24 <strong>the</strong> rivers rose above 65 feet<br />
and never aga<strong>in</strong> dropped below 64 feet dur<strong>in</strong>g 1979. This curtailed <strong>the</strong> test<strong>in</strong>g stage, prevented<br />
satisfactory completion <strong>of</strong> test<strong>in</strong>g <strong>the</strong>n underway on one site, and, <strong>of</strong> course, reduced<br />
<strong>the</strong> number <strong>of</strong> sites tested and data available for floodpla<strong>in</strong> site assessment.<br />
Test<strong>in</strong>g proceeded without <strong>in</strong>terruption for seven weeks, with one to three test excavation<br />
crews and one boat survey crew deployed, <strong>of</strong>ten simultaneously. Nearly all floodpla<strong>in</strong><br />
sites were reached most efficiently by boat, and are essentially o<strong>the</strong>rwise <strong>in</strong>accessible <strong>in</strong> wet<br />
wea<strong>the</strong>r. We <strong>the</strong>refore employed two boats on a busy schedule to deliver crews and equipment<br />
and to extend bankl<strong>in</strong>e surveys. We also necessarily scheduled simultaneous test<strong>in</strong>g<br />
operations on sites close to one ano<strong>the</strong>r, although no site was selected for test<strong>in</strong>g on this<br />
basis. Considerations <strong>of</strong> safety and logistics are <strong>in</strong>volved <strong>in</strong> mov<strong>in</strong>g <strong>the</strong> test<strong>in</strong>g operation<br />
between site clusters, that is, <strong>in</strong> centraliz<strong>in</strong>g effort at <strong>in</strong>tervals with<strong>in</strong> <strong>the</strong> test<strong>in</strong>g stage.<br />
In <strong>the</strong> follow<strong>in</strong>g section we are at pa<strong>in</strong>s to describe <strong>the</strong> process <strong>of</strong> selection <strong>of</strong> sites for<br />
test<strong>in</strong>g and limitations <strong>of</strong> this process. Clearly <strong>the</strong> results <strong>of</strong> test<strong>in</strong>g are critically important to<br />
assessment and management <strong>of</strong> floodpla<strong>in</strong> sites, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> direction that future archeological<br />
work may take <strong>in</strong> this project area and region.<br />
selection <strong>of</strong> sites for test<strong>in</strong>g<br />
Decisions about selection <strong>of</strong> sites for test<strong>in</strong>g were made when substantial survey results<br />
were available, <strong>in</strong>clud<strong>in</strong>g both field observations and field laboratory process<strong>in</strong>g <strong>of</strong> collections.<br />
In accordance with contract requirements and our project research design, test<strong>in</strong>g<br />
was conducted to assess <strong>the</strong> significance <strong>of</strong> selected archeological resources <strong>in</strong> <strong>the</strong> context <strong>of</strong><br />
certa<strong>in</strong> research questions and hypo<strong>the</strong>ses. Our criteria for selection were based on site survey<br />
data recovered and also on artificial constra<strong>in</strong>ts <strong>of</strong> location and accessibility. Thus any
156 Hemm<strong>in</strong>gs<br />
site selected for test<strong>in</strong>g has produced or is expected to produce at least three <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g:<br />
1. cultural rema<strong>in</strong>s below <strong>the</strong> 65-foot contour, whe<strong>the</strong>r or not this material is<br />
known to be <strong>in</strong> place<br />
2. one or more substantial samples <strong>of</strong> artifacts and debris<br />
3. temporally or functionally diagnostic artifacts<br />
4. superposed assemblages <strong>of</strong> artifacts<br />
5. one or more samples <strong>of</strong> organic rema<strong>in</strong>s<br />
6. subsurface or above-ground structural rema<strong>in</strong>s or features<br />
7. data representative <strong>of</strong> a larger functional, temporal, or areal site group<br />
8. data evidently unique to <strong>the</strong> locality or region<br />
9. data bear<strong>in</strong>g on specific research questions<br />
10. also any site selected must be reasonably and safely accessible to personnel<br />
and equipment.<br />
Consultation with <strong>the</strong> pr<strong>in</strong>cipal <strong>in</strong>vestigator, station archeologists, and o<strong>the</strong>r personnel<br />
closely <strong>in</strong>volved with <strong>the</strong> project contributed to <strong>the</strong> decision process. Although criteria 1 and<br />
10 were considered necessary, no o<strong>the</strong>r weight or order <strong>of</strong> importance was ascribed, so that<br />
judgment <strong>of</strong> <strong>the</strong> project archeologist and staff was ultimately exercised.<br />
At <strong>the</strong> outset <strong>of</strong> <strong>the</strong> test<strong>in</strong>g stage 16 floodpla<strong>in</strong> sites were selected and <strong>in</strong>formation<br />
perta<strong>in</strong><strong>in</strong>g to <strong>the</strong>se sites (location, elevation, size, observed content, proposed work) was<br />
submitted to <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers. The total number <strong>of</strong> sites <strong>in</strong>dicated for test<strong>in</strong>g<br />
was based on our <strong>in</strong>itial proposal, our experience with floodpla<strong>in</strong> mobility and logistics,<br />
and on <strong>the</strong> length <strong>of</strong> time scheduled for test<strong>in</strong>g (curtailed as noted above). It should be<br />
emphasized here that many o<strong>the</strong>r floodpla<strong>in</strong> sites meet some <strong>of</strong> <strong>the</strong> criteria, and a few sites<br />
have emerged as potentially significant resources dur<strong>in</strong>g <strong>the</strong> last survey efforts (e.g., Marie<br />
Sal<strong>in</strong>e Land<strong>in</strong>g, 3AS299) or as analysis proceeded. One such site, <strong>the</strong> Wreck <strong>of</strong> <strong>the</strong> Lotawanna<br />
(3UN153), described previously (Chapters 4, 5) has now also been recommended to <strong>the</strong> U.S.<br />
Army Corps <strong>of</strong> Eng<strong>in</strong>eers for assessment by special techniques. The stand<strong>in</strong>g list <strong>of</strong> <strong>17</strong> floodpla<strong>in</strong><br />
sites selected for test<strong>in</strong>g, <strong>in</strong>clud<strong>in</strong>g those eight sites (<strong>in</strong>dicated by asterisks) described<br />
<strong>in</strong> <strong>the</strong> rema<strong>in</strong>der <strong>of</strong> this chapter, is as follows:<br />
Unit 2: Lower Sal<strong>in</strong>e River<br />
*Mouth <strong>of</strong> Eagle Creek (3BR78)<br />
*Jug Po<strong>in</strong>t 1 (3AS307)<br />
*Jug Po<strong>in</strong>t 1 (3AS306)<br />
*Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76)<br />
*One Cypress Po<strong>in</strong>t (3AS286)<br />
*Buttonbush (3BR58)<br />
*False Indigo (3AS285)<br />
Unit 3: Lower Eagle Creek<br />
Eagle Creek 4 (3BR66)
Unit 4: Ouachita River below Mile 254<br />
*Marie Sal<strong>in</strong>e (3AS329)<br />
Water Elm 3 (3AS287)<br />
Coon Glory Bend (3UN168)<br />
Unit 5: Lower Lapile Creek<br />
Lapile Creek 5 (3UN157)<br />
Lapile Creek 7 (3UN151)<br />
Unit 6: Oxbow Lake and Backswamp<br />
St. Mary’s Lake North (3UN187)<br />
Unit 8: Historic Sites<br />
Keelboat Brake Sawmill (3BR55)<br />
<strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g (3UN122)<br />
Wreck <strong>of</strong> <strong>the</strong> Lotawanna (3UN153)<br />
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 157<br />
Under ideal field conditions, we would have expected to test m<strong>in</strong>imally six to eight<br />
more sites, not <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> Lotawanna, s<strong>in</strong>ce most <strong>of</strong> <strong>the</strong> rema<strong>in</strong><strong>in</strong>g ones (not <strong>in</strong>dicated by<br />
asterisks) are small <strong>in</strong> extent and relatively accessible. Also, it would have been desirable to<br />
accomplish test<strong>in</strong>g on <strong>the</strong> historic sawmill and land<strong>in</strong>g sites.<br />
test excAVAtion strAtegy And tActics<br />
A consistent set <strong>of</strong> test excavation procedures was briefly outl<strong>in</strong>ed <strong>in</strong> <strong>the</strong> research<br />
design, <strong>in</strong>troduced dur<strong>in</strong>g a field tra<strong>in</strong><strong>in</strong>g session, and implemented on <strong>the</strong> first site tested<br />
(3AS329) under <strong>the</strong> direction <strong>of</strong> <strong>the</strong> project archeologist. All succeed<strong>in</strong>g test excavations employed<br />
<strong>the</strong>se procedures under <strong>the</strong> direction <strong>of</strong> crew chiefs with m<strong>in</strong>or adjustments for field<br />
conditions. Test<strong>in</strong>g strategy required not only judicious selection <strong>of</strong> sites, but also efficient<br />
allocation <strong>of</strong> test<strong>in</strong>g effort with<strong>in</strong> sites and management <strong>of</strong> excavator’s bias. These requirements<br />
are relevant to all test<strong>in</strong>g work, but assume added importance where sites are buried<br />
through most <strong>of</strong> <strong>the</strong>ir extent, as are <strong>in</strong> fact all prehistoric floodpla<strong>in</strong> sites recorded by us<br />
(Chapter 5). Table 9 summarizes test excavation procedures as proposed <strong>in</strong> outl<strong>in</strong>e and <strong>the</strong>n<br />
implemented or adjusted <strong>in</strong> <strong>the</strong> field. (This summary does not address <strong>the</strong> cont<strong>in</strong>gencies <strong>of</strong><br />
test excavation <strong>in</strong> historic sites with structural or o<strong>the</strong>rwise complex rema<strong>in</strong>s.)<br />
We turn now to summaries <strong>of</strong> results <strong>of</strong> test excavation for eight floodpla<strong>in</strong> sites <strong>in</strong><br />
<strong>the</strong> order <strong>in</strong> which test<strong>in</strong>g proceeded. All <strong>of</strong> <strong>the</strong>se sites are newly discovered and recorded.<br />
These summaries emphasize data perta<strong>in</strong><strong>in</strong>g to <strong>the</strong> physical context, lithic and ceramic<br />
analyses, and o<strong>the</strong>r specialized analyses, as well as <strong>in</strong>ferences about site functions and age<br />
<strong>of</strong> components represented. Additional site data may be found <strong>in</strong> Appendixes A and B and<br />
<strong>in</strong> Chapters 7 and 8. Location maps for <strong>the</strong>se eight sites <strong>in</strong>clude Figures 26 and 28.
158 Hemm<strong>in</strong>gs<br />
Table 9. Summary <strong>of</strong> Test Excavation Tactics for Floodpla<strong>in</strong> Sites.<br />
Tasks Proposed Tactics Implemented<br />
1. Area tested to approximate 5% <strong>of</strong> site Area tested approximated 5% <strong>of</strong> maximum site extent<br />
extent known from survey<br />
2. Site boundaries to be established by Site boundaries and depths established approximately by<br />
auger<strong>in</strong>g shovel test<strong>in</strong>g (60 cm) at close<br />
regular <strong>in</strong>tervals; contents <strong>of</strong> all holes screened<br />
3. 1 m 2 test pits located randomly with<strong>in</strong> 1 m 2 test pits located systematically along a basel<strong>in</strong>e,<br />
site boundaries randomly with<strong>in</strong> site boundaries, and judgmentally where<br />
subsurface rema<strong>in</strong>s already exposed (on small sites only<br />
judgmental or systematic test pits employed)<br />
4. Test pits to be dug <strong>in</strong> 10 cm levels, Test pits dug to sterile soil or water table as proposed;<br />
screen<strong>in</strong>g all fill through ¼ <strong>in</strong>ch (6.4 mm) because <strong>of</strong> clay content, all fill screened thoroughly and<br />
hardware cloth coarse residue searched and dumped<br />
5. Collection <strong>of</strong> radiocarbon and o<strong>the</strong>r Opportunistic collection <strong>of</strong> charcoal for radiocarbon and<br />
datable samples wood identification; no archeomagnetic samples available<br />
from any site<br />
6. Collection <strong>of</strong> soil and constant volume Limited program <strong>of</strong> soil, sediment, and wet screen or sieve<br />
flotation samples samples from all sites<br />
7. Field laboratory process<strong>in</strong>g Field laboratory wash<strong>in</strong>g, sort<strong>in</strong>g, and catalog<strong>in</strong>g <strong>of</strong> all<br />
collections as received<br />
8. Instrument mapp<strong>in</strong>g Surface features, artifact concentrations, basel<strong>in</strong>e, and test<br />
excavation units mapped for each site with light portable<br />
transit and tape; elevations determ<strong>in</strong>ed approximately by<br />
reference to river level and gauge read<strong>in</strong>g; semipermanent<br />
benchmarks established<br />
9. Standard record<strong>in</strong>g Site form, field notes, level forms, pr<strong>of</strong>iles, site maps, photo<br />
record, lab record compiled for all sites
Marie Sal<strong>in</strong>e (3AS329)<br />
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 159<br />
The Marie Sal<strong>in</strong>e site (3AS329), located on <strong>the</strong> low natural levee <strong>of</strong> <strong>the</strong> Ouachita River<br />
(Figure 35) below <strong>the</strong> entrance to <strong>the</strong> <strong>Marais</strong> Sal<strong>in</strong>e Lake (named by <strong>the</strong> French before 1804)<br />
and its <strong>in</strong>tricate system <strong>of</strong> sloughs, is a relatively large, deep, multicomponent, and stratified<br />
site. It is <strong>in</strong> fact <strong>the</strong> largest and most complex floodpla<strong>in</strong> site now known <strong>in</strong> <strong>the</strong> project area.<br />
We suggest <strong>the</strong> possibility that <strong>Marais</strong> Sal<strong>in</strong>e Lake represents <strong>the</strong> former mouth <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e<br />
River, and that early occupation here between about 1200 B.C. and A.D. 1 was contemporary<br />
with this confluence <strong>of</strong> rivers. Although Fleetwood (1969) does not make this dist<strong>in</strong>ction, he<br />
does present geomorphic and subsurface data for former courses <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River at or<br />
near <strong>Marais</strong> Sal<strong>in</strong>e. Future work <strong>in</strong> this site should address this geohydrological problem<br />
(Appendix C).<br />
Dur<strong>in</strong>g <strong>in</strong>itial survey and later test<strong>in</strong>g, <strong>the</strong> Marie Sal<strong>in</strong>e site was recorded and collected<br />
as three adjacent and dist<strong>in</strong>ctive occupation areas (Figure 36). These artifact scatters extend<br />
over 200 m along <strong>the</strong> riverbank and constitute a very narrow elongate configuration <strong>in</strong>s<strong>of</strong>ar<br />
as now known. The bankl<strong>in</strong>e has a maximum height <strong>of</strong> 1.6 m (62.2-67.5 feet MSL). Artifacts<br />
or debris have been recovered <strong>in</strong> place throughout this vertical span, albeit sparsely, and<br />
from <strong>the</strong> modern levee surface, especially <strong>in</strong> Area C upstream. The follow<strong>in</strong>g archeological<br />
periods and phases appear to be represented by one or more components at <strong>the</strong> Marie Sal<strong>in</strong>e<br />
site (o<strong>the</strong>r components may be identified <strong>in</strong> future work):<br />
Mississippi period, Caney Bayou phase (A.D. 1400-1600)<br />
Baytown-Coles Creek period, phases unnamed (A.D. 300-1100)<br />
Tchula period, Coon Island phase (500 B.C.-A.D. 1)<br />
Poverty Po<strong>in</strong>t period, Calion phase (1200-500 B.C.)<br />
Area A. The presumed extent <strong>of</strong> this area was sampled by means <strong>of</strong> systematic (10,<br />
11) random (14, 15) and judgmental test pits (12) to sterile soil at about 0.6 m and six auger<br />
holes to 1.0 m depth (Figure 36). All <strong>of</strong> <strong>the</strong> test pits produced sparse artifacts and debris<br />
from <strong>the</strong> upper half meter <strong>of</strong> levee deposit which are best attributed to Baytown-Coles Creek<br />
occupations <strong>of</strong> low <strong>in</strong>tensity. A novaculite dart po<strong>in</strong>t stem, Baytown Pla<strong>in</strong> sherds, and chert,<br />
novaculite, and clear crystal quartz debitage were <strong>the</strong> pr<strong>in</strong>cipal materials recovered. No<br />
organic rema<strong>in</strong>s were noted or recovered except <strong>in</strong> Test Pit 14 which <strong>in</strong>tersected a modern<br />
disturbance. All auger holes were sterile, presumably because this technique is <strong>in</strong>effective<br />
under conditions <strong>of</strong> low artifact density. Based on test excavation and good riverbank exposure<br />
from levee surface to river stage, no cultural materials are known to be present below<br />
0.6 m depth.<br />
Area B. This small area produced two notable concentrations <strong>of</strong> chert debitage, one<br />
broadly scattered by erosion and disturbance, but ano<strong>the</strong>r dense cluster rema<strong>in</strong><strong>in</strong>g largely<br />
<strong>in</strong> place at about 0.5 to 0.6 m depth. A systematic test pit (9), a judgmental test pit (13) <strong>in</strong> <strong>the</strong><br />
rema<strong>in</strong><strong>in</strong>g cluster <strong>of</strong> chert debitage, and four auger holes were emplaced <strong>in</strong> Area B (Figure<br />
36). Surface collection from <strong>the</strong> eroded scatter (shown to <strong>the</strong> north <strong>of</strong> Test Pit 9) recovered a
160 Hemm<strong>in</strong>gs<br />
Figure 35. Aerial view <strong>of</strong> <strong>the</strong> Ouachita River-<strong>Marais</strong> Sal<strong>in</strong>e Lake locale and <strong>the</strong><br />
location <strong>of</strong> Marie Sal<strong>in</strong>e site (3AS329) (air photo December 1195,<br />
courtesy <strong>of</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers) (AAS neg. 814150).
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 161<br />
Figure 36. Marie Sal<strong>in</strong>e (3AS329), test excavations, October 1979.
162 Hemm<strong>in</strong>gs<br />
Late Archaic dart po<strong>in</strong>t and two Mississippi period arrow po<strong>in</strong>ts, chert debitage, and Baytown<br />
Pla<strong>in</strong> sherds (Figures 58e, 59q <strong>in</strong> Chapter 8). Test Pit 9 produced only a few novaculite<br />
flakes, while all auger holes were aga<strong>in</strong> sterile to 1.0 m depth. This heterogeneous assemblage<br />
suggests to us localized occupation and associated fl<strong>in</strong>tknapp<strong>in</strong>g dur<strong>in</strong>g Baytown-<br />
Coles Creek time, and subsequent low <strong>in</strong>tensity use late <strong>in</strong> <strong>the</strong> Mississippi period. No<br />
cultural rema<strong>in</strong>s were observed below 0.7 m depth, and <strong>the</strong> Archaic dart po<strong>in</strong>t is quite likely<br />
out <strong>of</strong> context. Area B is partly disturbed by vehicle and boat launch<strong>in</strong>g tracks as <strong>in</strong>dicated<br />
<strong>in</strong> Figure 36.<br />
Area C. Area C furnished key stratigraphic data for this site and for <strong>the</strong> project area<br />
as a whole. Dur<strong>in</strong>g <strong>in</strong>itial survey, fire-cracked rock and debitage were observed <strong>in</strong> <strong>the</strong> riverbank<br />
at or near river level, extend<strong>in</strong>g <strong>in</strong>termittently <strong>the</strong> length <strong>of</strong> Area C (130 m). This material<br />
was subsequently mapped and elevations obta<strong>in</strong>ed (Figure 36). Similar debris and two<br />
dart po<strong>in</strong>t fragments were found eroded out and concentrated at <strong>the</strong> water’s edge. Test<strong>in</strong>g<br />
by means <strong>of</strong> six systematic test pits (1-6), two random test pits (7, 8), and 15 auger holes was<br />
designed to <strong>in</strong>tersect a presumed Archaic horizon at 1.0 m or greater depth below <strong>the</strong> levee<br />
surface. Younger and stratigraphically higher ceramic components were expected on <strong>the</strong><br />
basis <strong>of</strong> a few Baytown Pla<strong>in</strong> sherds from <strong>the</strong> riverbank and two shell-tempered pla<strong>in</strong> sherds<br />
erod<strong>in</strong>g out at less than 0.2 m depth.<br />
All Area C test pits were excavated to depths <strong>of</strong> 1.2 to 1.5 m where ground water was<br />
encountered, and all produced artifacts or manuports <strong>in</strong> <strong>the</strong> deepest levels. The nature <strong>of</strong><br />
topstratum deposits and stratigraphic detail were similar among test pits and will be summarized<br />
here by reference to Test Pits 1 and 6, for which soil and sediment analyses are<br />
available (Chapter 7). These test pits are separated by 100 m along <strong>the</strong> axis <strong>of</strong> <strong>the</strong> levee<br />
surface, which lies about 0.4 m lower at Test Pit 6 (Figures 37, 38). The topstratum deposits<br />
<strong>in</strong>tersected are f<strong>in</strong>e gra<strong>in</strong>ed loams and clays, predom<strong>in</strong>antly gray <strong>in</strong> color, and massive<br />
to weakly lam<strong>in</strong>ated <strong>in</strong> appearance. Erosional or nondepositional breaks, paleosoils, and<br />
obtrusive cultural zones or “floors” are miss<strong>in</strong>g <strong>in</strong> <strong>the</strong>se test pit pr<strong>of</strong>iles (and o<strong>the</strong>rs) with<br />
one major exception. This exception is a th<strong>in</strong>, dark gray, organically sta<strong>in</strong>ed zone, associated<br />
with relatively dense, white novaculite debitage, at about 0.9 to 1.0 m <strong>in</strong> Test Pit 6 (Figure<br />
38). (Novaculite is def<strong>in</strong>ed <strong>in</strong> Chapter 8.) Small quantities <strong>of</strong> fire-cracked rock, fired clay, and<br />
wood charcoal, but no sherds, have been recovered from this zone. Comparison <strong>of</strong> <strong>the</strong> depth<br />
and artifact content with o<strong>the</strong>r test pits <strong>in</strong>dicate that <strong>the</strong> age <strong>of</strong> this zone is ei<strong>the</strong>r late Poverty<br />
Po<strong>in</strong>t or early Tchula period.<br />
Published soil survey data contribute some additional <strong>in</strong>sights to our observations<br />
on topstratum deposits. Marie Sal<strong>in</strong>e occupies Guyton soils, which are extensive, poorly<br />
dra<strong>in</strong>ed, slowly permeable, siliceous silt loams and silty clay loams derived from forested,<br />
Coastal Pla<strong>in</strong> uplands (Gill et al. 1979:Sheet 40). They are strongly or very strongly acid,<br />
moderate to low <strong>in</strong> organic content, and low <strong>in</strong> natural fertility. Gley<strong>in</strong>g, or reduction and<br />
transfer <strong>of</strong> iron, and subsoil mottl<strong>in</strong>g result<strong>in</strong>g from <strong>in</strong>termittent waterlogg<strong>in</strong>g are charac-
A b<br />
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 163<br />
Figure 37. Marie Sal<strong>in</strong>e (3AS329), deep test pits pr<strong>of</strong>iles. a. Test Pit 1; b. Test Pit (AAS neg.<br />
797140, 797146).
164 Hemm<strong>in</strong>gs<br />
Figure 38. Marie Sal<strong>in</strong>e (3AS329), pr<strong>of</strong>ile records for Test Pits 1 and 6 (south walls).
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 165<br />
teristic, and are evident <strong>in</strong> our pr<strong>of</strong>iles. Soil horizons are fa<strong>in</strong>t and weakly developed because<br />
sediments are <strong>of</strong> recent orig<strong>in</strong> and are deposited at frequent <strong>in</strong>tervals (Larance 1961:<br />
60).<br />
The cultural stratigraphy obta<strong>in</strong>ed by excavation <strong>of</strong> Area C test pits is excellent, but<br />
not complete <strong>in</strong> all details. Figure 39 schematically presents this stratigraphy by means <strong>of</strong><br />
artifact and debris densities <strong>in</strong> 10 cm levels. Five categories <strong>of</strong> imperishable artifacts and<br />
debris, which occur frequently <strong>in</strong> this and o<strong>the</strong>r floodpla<strong>in</strong> sites, are <strong>in</strong>dicated: sherds, stone<br />
tools, debitage, fire-cracked rock, and particles <strong>of</strong> fired clay. (We will utilize <strong>the</strong>se categories<br />
here and elsewhere <strong>in</strong> this report for comparison with<strong>in</strong> and between sites. Frequencies <strong>of</strong><br />
material <strong>in</strong> <strong>the</strong>se categories and identification <strong>of</strong> temporally or functionally dist<strong>in</strong>ctive artifact<br />
types are employed with o<strong>the</strong>r conjunctive data <strong>in</strong> <strong>the</strong>se comparisons.) An additional<br />
category <strong>of</strong> material, manuports, is omitted here (and <strong>in</strong> later comparisons) because human<br />
<strong>in</strong>troduction <strong>of</strong> <strong>the</strong>se objects, usually unmodified pebbles, cannot be established with certa<strong>in</strong>ty<br />
(see also Chapter 8). We can say that occurrence <strong>of</strong> 273 pebble “manuports” recovered<br />
<strong>in</strong> Test Pits 1-8 is strongly, but not perfectly, correlated with o<strong>the</strong>r artifacts and debris.<br />
The <strong>in</strong>terpretation <strong>of</strong> cultural stratigraphy <strong>in</strong> Area C is <strong>the</strong>refore summarized as follows:<br />
1. A preceramic zone is probably present <strong>in</strong> all test pits below about 1.1 or 1.2<br />
m and lies partly below ground water or river stage; <strong>the</strong> pr<strong>in</strong>cipal artifacts<br />
recovered are two steatite sherds <strong>in</strong> Test Pit 8 and a muscovite schist slab <strong>in</strong><br />
Test Pit 3, <strong>the</strong>se materials be<strong>in</strong>g <strong>of</strong> presumed Appalachian orig<strong>in</strong> (Figure 64a,<br />
b); novaculite flakes and fire-cracked rock are aga<strong>in</strong> associated; <strong>the</strong> dates for<br />
preceramic occupation here should lie between 1200 and 500 B.C.<br />
2. A def<strong>in</strong>ite early ceramic zone appears at about 0.8 to 1.2 m <strong>in</strong> Test Pits 1 and<br />
7 (shaded levels <strong>in</strong> Figure 39); this zone is marked by eroded, s<strong>of</strong>t paste,<br />
Tchefuncte sherds, probably <strong>in</strong>clud<strong>in</strong>g Tchefuncte Incised and o<strong>the</strong>r types<br />
(Figure 65e); <strong>the</strong> low frequency and eroded, friable condition <strong>of</strong> sherds suggest<br />
that this zone is present, but not dist<strong>in</strong>guished by us, <strong>in</strong> o<strong>the</strong>r test pits; a<br />
few novaculite flakes and considerable fire-cracked rock are associated; <strong>the</strong><br />
dates for early ceramic occupation here should lie between 500 B.C. and A.D.<br />
1.<br />
3. A late ceramic zone occurs as a diffuse scatter <strong>of</strong> Baytown Pla<strong>in</strong> sherds and a<br />
few chert and novaculite flakes from about 0.2 to 0.7 m <strong>in</strong> depth (dashed l<strong>in</strong>e<br />
<strong>in</strong> various test pits, Figure 39); a contract<strong>in</strong>g stemmed dart po<strong>in</strong>t (Figure 59b)<br />
is evidently associated with a s<strong>in</strong>gle Baytown Pla<strong>in</strong> sherd <strong>in</strong> Test Pit 5 at 75<br />
cm; <strong>the</strong> dates for sporadic occupation here dur<strong>in</strong>g <strong>the</strong> Baytown-Coles Creek<br />
period should range from about A.D. 300-1100.
166 Hemm<strong>in</strong>gs<br />
Figure 39. Cultural stratigraphy <strong>in</strong> Area C, Marie Sal<strong>in</strong>e (3AS329). West walls <strong>of</strong> Test Pits 1-8 <strong>in</strong>dicated with artifacts<br />
and debris counts by 10 cm levels. With<strong>in</strong> each test pit pr<strong>of</strong>ile <strong>the</strong> five columns <strong>of</strong> figures represent sherds,<br />
stone tools, debitage, fire-cracked rock, and fired clay particles respectively. Superscripts on artifact counts as<br />
follows: P=dart po<strong>in</strong>t; M=muscovite or talc schist slab; S=steatite sherds (<strong>the</strong>se items appear <strong>in</strong> Figures 59b, c;<br />
64a); dashed horizontal l<strong>in</strong>es equal <strong>the</strong> position <strong>of</strong> lowermost Baytown Pla<strong>in</strong> sherds. Hachured blocks <strong>in</strong> Test<br />
Pits 1 and 7 represent probable earliest ceramic occupation levels with Tchefuncte sherds (one <strong>of</strong> <strong>the</strong>se sherds<br />
appears <strong>in</strong> Figure 65e). NP=new navigation pool elevation (meters).
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 167<br />
4. The sparse Mississippi period occupation known to be present at about 0.2<br />
m (from surface f<strong>in</strong>ds) is not dist<strong>in</strong>guishable <strong>in</strong> <strong>the</strong>se test pits: dates for this<br />
occupation are believed to be late (Caney Bayou phase) or about A.D. 1400-<br />
1600.<br />
In summariz<strong>in</strong>g Area C cultural stratigraphy we have not specifically identified <strong>the</strong><br />
component represented obtrusively by organic matter and novaculite debitage <strong>in</strong> Test Pit 6<br />
at 0.9-1.0 m. On <strong>the</strong> basis <strong>of</strong> correlation among test pits, this zone may ei<strong>the</strong>r be preceramic<br />
or early ceramic <strong>in</strong> age. It has not presently produced Tchefuncte sherds. Four radiocarbon<br />
samples were selected from among <strong>the</strong> 30 small charcoal samples obta<strong>in</strong>ed <strong>in</strong> Area C. Results<br />
<strong>of</strong> age determ<strong>in</strong>ation were for <strong>the</strong> most part unacceptably early; this problem is discussed<br />
fully <strong>in</strong> Chapter 7.<br />
Two trial wet screen (1/8 <strong>in</strong>ch or 3.1 mm) samples <strong>of</strong> about 2 liters volume each were<br />
processed <strong>in</strong> an attempt to evaluate <strong>the</strong> potential for recovery <strong>of</strong> organic rema<strong>in</strong>s. These<br />
samples were collected from Test Pit 1, 1.1-1.2 m (presumed Calion phase occupation) and<br />
from Test Pit 6, 0.9-1.0 m (organic zone with debitage noted above). The results were equivocal,<br />
but not promis<strong>in</strong>g. The coherent silty clay matrix must be dispersed or deflocculated <strong>in</strong><br />
this procedure. Both samples yielded rare wood charcoal particles and organic (?) matter not<br />
positively identified, but no carbonized seeds or nuts and no animal bone. The general experience<br />
<strong>of</strong> test pit excavation is that organic rema<strong>in</strong>s are exceed<strong>in</strong>gly sparse, where present, <strong>in</strong><br />
all components represented at Marie Sal<strong>in</strong>e.<br />
In Area C, 15 auger holes were emplaced to 1.0 m depth <strong>in</strong> order to delimit area <strong>of</strong><br />
occupation. One hole produced two white novaculite flakes from about 0.8-1.0 m, while all<br />
o<strong>the</strong>rs were apparently sterile (Figure 36). S<strong>in</strong>ce some auger holes between test pits would<br />
certa<strong>in</strong>ly <strong>in</strong>tersect cultural “horizons,” we have reason to believe that 4-<strong>in</strong>ch (10 cm) auger<br />
bits are <strong>in</strong>effective <strong>in</strong> locat<strong>in</strong>g diffusely distributed artifacts and debris.<br />
Overview. Marie Sal<strong>in</strong>e (3AS329) conta<strong>in</strong>s a record <strong>of</strong> <strong>in</strong>termittent occupation and<br />
reoccupation cover<strong>in</strong>g a 3000-year span. Artifacts and debris <strong>of</strong> all k<strong>in</strong>ds are scarce by <strong>the</strong><br />
standards <strong>of</strong> upland habitation sites <strong>in</strong> this region. Table 10 below summarizes <strong>the</strong> collections<br />
<strong>of</strong> tools and imperishable materials recovered from three occupation areas by survey<br />
Table 10. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area, Marie<br />
Sal<strong>in</strong>e (3AS329).<br />
Occupation Area Sherds Stone Tools Debitage Fire-cracked Rock Fired Clay<br />
Area A 10 3 26 1 0<br />
Area B 19 8 292 2 0<br />
Area c 68 21 159 213 134<br />
Site Totals 97 32 477 216 134
168 Hemm<strong>in</strong>gs<br />
and test excavation techniques. The deep occupation zones now identified as Poverty Po<strong>in</strong>t<br />
and Tchula periods <strong>in</strong> Area C appear to represent more <strong>in</strong>tensive and extensive site use than<br />
do overly<strong>in</strong>g zones. We have <strong>in</strong>troduced <strong>the</strong> possibility that Marie Sal<strong>in</strong>e occupied a natural<br />
levee location at <strong>the</strong> mouth <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River dur<strong>in</strong>g early occupation. In fact, <strong>the</strong> <strong>in</strong>itial use<br />
<strong>of</strong> this site lies below river stage, and has not yet been <strong>in</strong>vestigated (see Appendix C).<br />
False Indigo (3AS285)<br />
The False Indigo site (3AS285) is located on <strong>the</strong> low natural levee <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River<br />
0.3 river miles (less than one-half kilometer) above <strong>the</strong> present mouth. The site name is<br />
taken from a legum<strong>in</strong>ous shrub, Amorpha fruticosa, here grow<strong>in</strong>g pr<strong>of</strong>usely on <strong>the</strong> exposed<br />
bankl<strong>in</strong>e, but absent from <strong>the</strong> floodpla<strong>in</strong> forest understory. This bankl<strong>in</strong>e reaches a maximum<br />
height <strong>of</strong> 2.1 m above river level (62.2 to 69.9 feet MSL). The topstratum deposits<br />
exposed are similar <strong>in</strong> appearance to <strong>the</strong> Marie Sal<strong>in</strong>e cutbank and test pit pr<strong>of</strong>iles.<br />
When discovered by our survey team, False Indigo produced a m<strong>in</strong>or scatter <strong>of</strong><br />
sherds and lithic debris along <strong>the</strong> steep bank and water’s edge for 40 m (Figure 40), but also<br />
produced an impressive and dense sherd cluster on <strong>the</strong> levee surface where degraded by<br />
sheet erosion. Much <strong>of</strong> this ceramic array could be attributed to <strong>the</strong> Gran <strong>Marais</strong> ceramic<br />
complex and Gran <strong>Marais</strong> phase (A.D. 1100-1400). A journal entry <strong>of</strong> that <strong>in</strong>itial discovery<br />
states that “<strong>the</strong> rema<strong>in</strong><strong>in</strong>g sherd concentration on <strong>the</strong> brow <strong>of</strong> <strong>the</strong> bank literally looks like<br />
it was dumped yesterday ra<strong>the</strong>r than five or more centuries ago.” These observations led<br />
us to plan and implement a controlled surface collection at False Indigo, focus<strong>in</strong>g on an 18<br />
m 2 area <strong>of</strong> dense artifact occurrence (Figures 40, 41). We shall return to <strong>the</strong> objectives and<br />
results <strong>of</strong> this effort below.<br />
Upstream from <strong>the</strong> dense sherd cluster, evidently separated by a few meters <strong>of</strong> sterile<br />
bankl<strong>in</strong>e, a second small sherd concentration was located <strong>in</strong> <strong>the</strong> bank at 0.6-0.8 m depth.<br />
These sherds were l<strong>in</strong>ear punctated or “jabbed <strong>in</strong>cised” on Tchefuncte paste, <strong>the</strong>refore a variety<br />
<strong>of</strong> Lake Borgne Incised (Figure 65a-c). We now dist<strong>in</strong>guish two occupation areas, Area<br />
A downstream and Area B upstream (Figure 40), which conta<strong>in</strong> younger and older components<br />
and are separated vertically by about 0.4 m <strong>of</strong> topstratum deposition; <strong>the</strong>se components<br />
represent two periods <strong>of</strong> <strong>the</strong> site use as follows:<br />
Mississippi period, Gran <strong>Marais</strong> phase (A.D. 1100-1400)<br />
Tchula period, Coon Island phase (500 B.C.-A.D. 1)<br />
Area A. The Gran <strong>Marais</strong> phase occupation <strong>in</strong> Area A was <strong>in</strong>vestigated by systematic<br />
(100%) surface collection, one judgmental test pit (1), one random test pit (15), and <strong>17</strong> shovel<br />
tests to about 0.4 m depth. This work delimited a compact, 12 x 15 m area <strong>of</strong> artifacts, midden,<br />
and debris at about 0.1 to 0.3 m depth below <strong>the</strong> levee surface, partly exposed by sheet<br />
erosion near <strong>the</strong> bankl<strong>in</strong>e. Figure 41 is compiled from two episodes <strong>of</strong> systematic surface
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 169<br />
Figure 40. False Indigo (3AS285), controlled surface collection and test excavations, November 1979.
<strong>17</strong>0 Hemm<strong>in</strong>gs<br />
Figure 41. False Indigo (3AS285) controlled surface collection <strong>in</strong> Area A with scatter map <strong>of</strong> sherds and o<strong>the</strong>r artifacts.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites <strong>17</strong>1<br />
collection and “piece plott<strong>in</strong>g” which attempted to dist<strong>in</strong>guish differential distribution <strong>of</strong><br />
artifacts and debris. Such patterns could be related to onsite behavior, for example, men’s or<br />
women’s activities <strong>in</strong> or near dwell<strong>in</strong>gs or merely refuse disposal. Our collection and plott<strong>in</strong>g<br />
activities were separated by two month’s time between site discovery and <strong>in</strong>itiation<br />
<strong>of</strong> test<strong>in</strong>g; ongo<strong>in</strong>g erosion uncovered about 100 additional items dur<strong>in</strong>g this <strong>in</strong>terval. The<br />
second surface collection effort also <strong>in</strong>volved superficial trowel<strong>in</strong>g <strong>of</strong> leaf litter and 2-3 cm<br />
<strong>of</strong> silt from <strong>the</strong> previous spr<strong>in</strong>g flood on <strong>the</strong> 18 m 2 area <strong>of</strong> densest artifact occurrence. We<br />
now th<strong>in</strong>k that sheet erosion <strong>of</strong> ra<strong>the</strong>r homogeneous midden is reflected by this distribution<br />
(Figure 41), but <strong>in</strong> fact additional spatial analyses <strong>of</strong> this scatter are possible.<br />
Test Pit 1 <strong>in</strong>tersected an area lowered by erosion, but was never<strong>the</strong>less rich <strong>in</strong> artifacts<br />
to 0.2 m depth. An Ashley arrow po<strong>in</strong>t (Figure 58 l), Gran <strong>Marais</strong> sherds, chert, and clear<br />
crystal quartz debitage, and about 2900 fired clay particles were recovered here. Fired clay<br />
on <strong>the</strong> surface had <strong>in</strong> fact suggested <strong>the</strong> existence <strong>of</strong> a hearth, but none was dist<strong>in</strong>guishable.<br />
An irregular shallow pit, Feature 1, extend<strong>in</strong>g <strong>in</strong>to <strong>the</strong> subsoil, may merely be <strong>the</strong> remnant<br />
basal level <strong>of</strong> midden. A 2-liter wet screen sample yielded many small wood charcoal fragments,<br />
modest amounts <strong>of</strong> calc<strong>in</strong>ed bone, and much imperishable residue, chiefly fired clay<br />
and debitage. The numerous fired clay particles from this test pit are variable <strong>in</strong> size up to<br />
40 mm, usually iregular <strong>in</strong> form, but <strong>in</strong>clude a few examples with smoo<strong>the</strong>d or prepared<br />
surfaces (not <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> wattle impressions <strong>of</strong> true daub).<br />
Test Pit 15, selected at random from our 18 m 2 area <strong>of</strong> debris, had little surface occurrence<br />
<strong>of</strong> artifacts, but <strong>in</strong>tersected well preserved midden and features from less than 0.1 to<br />
0.5 m depth (Figures 42, 43). Feature 2 is an irregular midden lens with a conspicuous downward<br />
protrusion <strong>in</strong>to subsoil, ei<strong>the</strong>r an odd postmold or tree root disturbance. A 2-liter wet<br />
screen sample from this lens was relatively more productive <strong>of</strong> organic rema<strong>in</strong>s than that<br />
noted above for Feature 1. This sample conta<strong>in</strong>ed abundant t<strong>in</strong>y wood charcoal fragments,<br />
burned and calc<strong>in</strong>ed bone, and imperishable residue. Fish vertebrae and sp<strong>in</strong>es, drum teeth,<br />
and gar scales are present; <strong>in</strong> fact nearly all bone fragments may be those <strong>of</strong> fish. Charred<br />
nutshell or seed fragments may be present <strong>in</strong> m<strong>in</strong>ute amounts. Wood ash is believed to be<br />
a constitutent, based on midden appearance and soil pH (5.9) contrast<strong>in</strong>g with nonmidden<br />
samples (Chapter 7, Table 8).<br />
Feature 3 <strong>in</strong> Test Pit 15 is <strong>the</strong> submidden surface, an irregular occupation floor marked<br />
by sizable sherds and buried by vary<strong>in</strong>g thicknesses <strong>of</strong> midden. From various data we<br />
believe this is <strong>in</strong>deed a structure floor or outdoor work area, not merely a locus <strong>of</strong> refuse<br />
disposal. The pr<strong>in</strong>cipal artifacts recovered <strong>in</strong> Test Pit 15 are an Ashley arrow po<strong>in</strong>t (Figure<br />
58l), chert, novaculite, and clear crystal quartz debitage, and about 600 fired clay particles <strong>of</strong><br />
small size. Traces <strong>of</strong> glossy, black, c<strong>in</strong>derlike residue from both features may perta<strong>in</strong> to <strong>the</strong><br />
burn<strong>in</strong>g process which occurred here (<strong>in</strong> addition to wood charcoal, ash, and fired clay).
<strong>17</strong>2 Hemm<strong>in</strong>gs<br />
Figure 42. False Indigo (3AS285), pr<strong>of</strong>ile with midden lens and o<strong>the</strong>r features<br />
<strong>in</strong> Test Pit 15 (AAS neg. 797097).
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites <strong>17</strong>3<br />
Figure 43. False Indigo (3AS285), Area A, show<strong>in</strong>g 18 m 2 area <strong>of</strong> dense artifact<br />
concentration after surface collection and test excavation; Sal<strong>in</strong>e<br />
River bankl<strong>in</strong>e is <strong>in</strong> <strong>the</strong> foreground (AAS neg. 797100).
<strong>17</strong>4 Hemm<strong>in</strong>gs<br />
One additional 2-liter wet screen sample from subsoil below Feature 3 was processed for<br />
organic rema<strong>in</strong>s and produced traces <strong>of</strong> calc<strong>in</strong>ed bone and wood charcoal as t<strong>in</strong>y fragments.<br />
N<strong>in</strong>e <strong>of</strong> <strong>17</strong> shovel tests <strong>in</strong> Area A produced artifacts or debris between about 0.1 and<br />
0.4 m depth; at least one <strong>of</strong> <strong>the</strong>se (Figure 40) <strong>in</strong>tersected dark colored midden at about 0.1-<br />
0.2 m. One o<strong>the</strong>r shovel test recovered a sandstone cobble anvil, <strong>the</strong> only heavy duty tool<br />
(Figure 63c). These tests toge<strong>the</strong>r <strong>in</strong>dicate an irregularly oval occupation area, unlike <strong>the</strong><br />
l<strong>in</strong>ear configuration which characterizes most riverbank sites recorded.<br />
Area A conta<strong>in</strong>ed a ceramic sample <strong>of</strong> about 1600 sherds which are entirely attributable<br />
to <strong>the</strong> Gran <strong>Marais</strong> ceramic complex. The follow<strong>in</strong>g frequencies are noted (see also<br />
Chapter 8): clay-tempered pla<strong>in</strong>, 58.4%; clay-tempered decorated, 9.3%; shell-tempered<br />
pla<strong>in</strong>, 2.5%; shell-tempered decorated, 0.3%; bone-tempered pla<strong>in</strong>, 0.1%; <strong>in</strong>determ<strong>in</strong>ate<br />
(small wea<strong>the</strong>red sherds), 29.4%.<br />
Area B. Figure 40 <strong>in</strong>dicates <strong>the</strong> position <strong>of</strong> Test Pit 19 and three shovel tests <strong>in</strong> Area B.<br />
These tests establish <strong>the</strong> presence <strong>of</strong> a Tchula period component with Lake Borgne Incised<br />
and o<strong>the</strong>r Tchefuncte sherds at about 0.6 to 0.8 m below <strong>the</strong> modern levee surface. A white<br />
novaculite dart po<strong>in</strong>t fragment and six white novaculite flakes were associated with this<br />
level <strong>in</strong> Test Pit 19. M<strong>in</strong>or amounts <strong>of</strong> Baytown Pla<strong>in</strong> sherds and chert or novaculite debitage<br />
occurred <strong>in</strong> upper levels <strong>of</strong> Test Pit 19 and two shovel tests. From <strong>the</strong>se test excavations and<br />
riverbank exposure, <strong>the</strong> Area B early ceramic component is evidently localized and sparse,<br />
lack<strong>in</strong>g <strong>the</strong> fire-cracked rock and organic rema<strong>in</strong>s seen at Marie Sal<strong>in</strong>e. Depth <strong>of</strong> this material<br />
and predom<strong>in</strong>ance <strong>of</strong> a late Tchula ceramic type may <strong>in</strong>dicate occupation late <strong>in</strong> <strong>the</strong> Coon<br />
Island phase, between about 250 B.C. and A.D. 1.<br />
Overview. False Indigo (3AS285) has early and late ceramic components separated by<br />
0.4 m <strong>of</strong> silty clay and more than a thousand years <strong>of</strong> time. The sett<strong>in</strong>g and nature <strong>of</strong> Coon<br />
Island phase occupation cannot be <strong>in</strong>terpreted from test excavation data. In contrast, <strong>the</strong><br />
Gran <strong>Marais</strong> phase occupation was clearly oriented to a Sal<strong>in</strong>e River course much like that<br />
<strong>of</strong> <strong>the</strong> present; <strong>the</strong> contents <strong>of</strong> a shallow refuse deposit <strong>in</strong>dicate fish<strong>in</strong>g activity and wet<br />
cookery. Well preserved organic midden deposits were not known to exist <strong>in</strong> floodpla<strong>in</strong> sites<br />
<strong>of</strong> our study area prior to test excavation at False Indigo. Relatively high densities <strong>of</strong> artifacts<br />
and debris <strong>in</strong>clud<strong>in</strong>g organic material, and <strong>in</strong>dications <strong>of</strong> structural rema<strong>in</strong>s (daublike<br />
material, postmold?) occur here <strong>in</strong> a spatially restricted area. An upland site <strong>of</strong> <strong>the</strong>se characteristics<br />
would be identified as a s<strong>in</strong>gle house site <strong>of</strong> some permanence. The paucity <strong>of</strong> lithic<br />
tools and debitage as compared to ceramic rema<strong>in</strong>s (Table 11), lowland sett<strong>in</strong>g, and midden<br />
content (fishbone and residue from burn<strong>in</strong>g) imply more specialized site use.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites <strong>17</strong>5<br />
Table 11. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area, False<br />
Indigo (3AS285).<br />
Occupation Area Sherds Stone Tools Debitage Fire-cracked Rock Fired Clay<br />
Entire Site Initial Collection 90 0 4 2 11<br />
Area A 1490 7 83 7 4068<br />
Area b 83 2 10 1 0<br />
Site Totals 1663 9 97 10 4079<br />
Buttonbush (3BR58)<br />
This small ceramic site is located 0.8 river miles (1.3 km) upstream from <strong>the</strong> present<br />
mouth <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River. The bankl<strong>in</strong>e is degraded by sheet erosion and slopes to <strong>the</strong><br />
water’s edge. Our survey team recorded Buttonbush as a very sparse lithic artifact scatter<br />
extend<strong>in</strong>g 98 m along <strong>the</strong> shorel<strong>in</strong>e and a small dense cluster <strong>of</strong> Baytown Pla<strong>in</strong> sherds erod<strong>in</strong>g<br />
out 1.5 m above river level near <strong>the</strong> upstream end <strong>of</strong> <strong>the</strong> site (Figure 44).<br />
Test excavation consisted <strong>of</strong> two judgmental test pits and 19 shovel tests. Test Pit 1 recovered<br />
fired clay particles to 0.1 m depth, but was o<strong>the</strong>rwise sterile. Test Pit 2 <strong>in</strong>vestigated<br />
<strong>the</strong> dense sherd cluster and was expanded to 1.5 x 1.5 m <strong>in</strong> order to encompass all additional<br />
sherds encountered. Analysis <strong>in</strong>dicates that <strong>the</strong> entire group <strong>of</strong> 454 sherds, tightly massed<br />
and nested at or just below <strong>the</strong> modern surface, is <strong>the</strong> rema<strong>in</strong>s <strong>of</strong> two Baytown Pla<strong>in</strong> vessels<br />
(Figure 45). These vessels were relatively large bowls with straight flar<strong>in</strong>g walls and flat disc<br />
bases (Figure 66h, i). Most <strong>of</strong> <strong>the</strong> rim sections are miss<strong>in</strong>g or have been wea<strong>the</strong>red beyond<br />
recognition. Test Pit 2 encountered no o<strong>the</strong>r artifacts or debris and was sterile below 0.1 m<br />
depth. A soil sample collected beneath <strong>the</strong> sherd cluster is reported <strong>in</strong> Chapter 7 (Table <strong>17</strong>).<br />
Shovel tests at upstream and downstream ends <strong>of</strong> <strong>the</strong> site were sterile except <strong>in</strong> one <strong>in</strong>stance.<br />
This test about 4 m west <strong>of</strong> <strong>the</strong> sherd cluster (Figure 44) produced charcoal at slightly<br />
over 0.2 m depth, which may better reflect <strong>the</strong> extent <strong>of</strong> alluviation follow<strong>in</strong>g site use than<br />
does <strong>the</strong> exposed sherd cluster.<br />
The pr<strong>in</strong>cipal lithic artifacts collected from <strong>the</strong> shorel<strong>in</strong>e are a small chert Gary po<strong>in</strong>t<br />
(Figure 59a), two chert and novaculite preforms, and m<strong>in</strong>or amounts <strong>of</strong> chert and novaculite<br />
debitage. Most <strong>of</strong> this material was scattered 30 to 50 m downstream from <strong>the</strong> sherd cluster,<br />
and may or may not be contemporary. S<strong>in</strong>ce <strong>the</strong> Baytown Pla<strong>in</strong> vessels can most likely be<br />
ascribed to <strong>the</strong> long span <strong>of</strong> <strong>the</strong> Baytown-Coles Creek period (A.D. 400-1100), <strong>the</strong> lithic artifacts<br />
may <strong>in</strong> fact be associated. More precise dat<strong>in</strong>g is desirable, but beyond <strong>the</strong> capability <strong>of</strong><br />
exist<strong>in</strong>g site data.
<strong>17</strong>6 Hemm<strong>in</strong>gs<br />
Figure 44. Buttonbush (3BR58), test excavations, November 1979 (four shovel tests are outside map area).
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites <strong>17</strong>7<br />
Figure 45. Buttonbush (3BR58), rema<strong>in</strong>s <strong>of</strong> two Baytown Pla<strong>in</strong> vessels <strong>in</strong> Test<br />
Pit 2 (AAS neg. 797105).
<strong>17</strong>8 Hemm<strong>in</strong>gs<br />
Buttonbush (3BR58) appears to represent a small collect<strong>in</strong>g (?) station on <strong>the</strong> Sal<strong>in</strong>e<br />
River bank whose occupants made use <strong>of</strong> large open-mou<strong>the</strong>d ceramic vessels. For comparative<br />
purposes, <strong>the</strong> total sample <strong>of</strong> imperishable artifacts and debris is recorded here: 454<br />
sherds, 3 stone tools, 8 items <strong>of</strong> debitage, 5 fire-cracked rock fragments, and 9 particles <strong>of</strong><br />
fired clay. A similar site at similar elevation, Water Elm 3 (3AS287), was recorded dur<strong>in</strong>g <strong>the</strong><br />
Ouachita River bankl<strong>in</strong>e survey (recommended for fur<strong>the</strong>r work <strong>in</strong> Appendix C). These sites<br />
now seem to consist chiefly <strong>of</strong> <strong>the</strong> rema<strong>in</strong>s <strong>of</strong> one or two cook<strong>in</strong>g or collect<strong>in</strong>g vessels broken,<br />
discarded, alluviated, and reexposed by sheet erosion near a relatively stable bankl<strong>in</strong>e.<br />
One Cypress Po<strong>in</strong>t (3AS286)<br />
One Cypress Po<strong>in</strong>t (3AS286) occupies <strong>the</strong> low natural levee <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River at <strong>the</strong><br />
entrance to a small lake, <strong>the</strong> scar <strong>of</strong> an earlier river channel (Figure 46). Many older channel<br />
scars and sloughs weave through <strong>the</strong> adjacent backswamps. In <strong>the</strong> site description and<br />
analysis that follows we will attribute site use to a term<strong>in</strong>al phase <strong>of</strong> <strong>the</strong> Mississippi period<br />
(ca A.D. 1600-<strong>17</strong>00). Government Land Office maps <strong>of</strong> 1827 and 1844 <strong>in</strong>dicate that <strong>the</strong> river<br />
channel has not shifted appreciably at this location for a century and a half. This stability<br />
suggests that <strong>the</strong> river and lake existed also at <strong>the</strong> time <strong>of</strong> occupation, or about three centuries<br />
ago, ra<strong>the</strong>r than <strong>the</strong> possibility that a cut<strong>of</strong>f lake formed dur<strong>in</strong>g this <strong>in</strong>terval.<br />
Discovery and record<strong>in</strong>g <strong>of</strong> One Cypress Po<strong>in</strong>t quickly established <strong>the</strong> dimensions and<br />
content <strong>of</strong> this important site. An elongate 5 m wide band <strong>of</strong> chert tools, debitage, and an occasional<br />
shell-tempered sherd were exposed on <strong>the</strong> surface by sheet erosion from <strong>the</strong> Sal<strong>in</strong>e<br />
River bank 60 m northward along <strong>the</strong> lakeshore (Figure 47). This band <strong>of</strong> debris extended<br />
from <strong>the</strong> levee surface nearly to <strong>the</strong> water’s edge, a vertical distance <strong>of</strong> 1.7 m (62.2-68.4 feet<br />
MSL), on a gently slop<strong>in</strong>g bankl<strong>in</strong>e. An apparent sterile zone <strong>in</strong>tervened near <strong>the</strong> midpo<strong>in</strong>t<br />
<strong>of</strong> this elongate surface scatter, and has provided <strong>the</strong> basis for dist<strong>in</strong>guish<strong>in</strong>g two occupation<br />
areas, Area A adjo<strong>in</strong><strong>in</strong>g <strong>the</strong> Sal<strong>in</strong>e River bank and Area B along <strong>the</strong> lakeshore. These occupation<br />
areas have been <strong>in</strong>vestigated by systematic surface collection and mapp<strong>in</strong>g on two occasions<br />
(site discovery and <strong>in</strong>itiation <strong>of</strong> test<strong>in</strong>g) and by means <strong>of</strong> seven test pits and 28 shovel<br />
tests. The typological and stratigraphic data now available do not justify any temporal or<br />
cultural dist<strong>in</strong>ction between Areas A and B. They are residential or functional loci <strong>of</strong> a s<strong>in</strong>gle<br />
late Mississippi or protohistoric period component.<br />
One Cypress Po<strong>in</strong>t’s stone tool assemblage is dist<strong>in</strong>ctive, previously unknown <strong>in</strong> <strong>the</strong><br />
Felsenthal region, and likely to be <strong>in</strong>trusive. The hallmarks <strong>of</strong> this assemblage are 15 Nodena<br />
arrow po<strong>in</strong>ts and fragments <strong>of</strong> two t<strong>in</strong>y endscrapers (Figure 48). These tools are entirely<br />
like <strong>the</strong>ir counterparts <strong>of</strong> <strong>the</strong> Nodena phase (A.D. 1400-<strong>17</strong>00) <strong>in</strong> nor<strong>the</strong>ast <strong>Arkansas</strong> and <strong>the</strong><br />
Quapaw phase (ca A.D. 1600-1824) <strong>in</strong> <strong>the</strong> lower <strong>Arkansas</strong> River dra<strong>in</strong>age (Ford 1961:Plate<br />
29; McGimsey 1964, 1969; Morse 1973:Figure 9; H<strong>of</strong>fman 1977:31; see also Williams 1980 on<br />
<strong>the</strong> protohistoric Armorel phase). O<strong>the</strong>r flaked stone tools are generally consistent with this
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites <strong>17</strong>9<br />
Figure 46. Aerial view <strong>of</strong> <strong>the</strong> lower Sal<strong>in</strong>e River floodpla<strong>in</strong> and <strong>the</strong> location<br />
<strong>of</strong> One Cypress Po<strong>in</strong>t (3AS286) (air photo December 1975, courtesy<br />
<strong>of</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers) (AAS neg. 814152).
180 Hemm<strong>in</strong>gs<br />
Figure 47. One Cypress Po<strong>in</strong>t (AS286), controlled surface collection and test excavations, November 1979.
g<br />
A<br />
h<br />
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 181<br />
Figure 48. Flaked stone tools and cores from One Cypress Po<strong>in</strong>t (3AS286). a. arrow po<strong>in</strong>ts;<br />
b. arrow po<strong>in</strong>t fragments; c. end scrapers; d. cyl<strong>in</strong>drical drill; e, f. bifacial preforms<br />
and fragments; g. bifacial cutt<strong>in</strong>g tool; h. perforators; i. retouched and<br />
utilized flakes and bladelike flakes; j. pebble cores (AAS neg. 813883).<br />
e<br />
f<br />
j<br />
b<br />
i<br />
c d
182 Hemm<strong>in</strong>gs<br />
Quapaw attribution, as are <strong>the</strong> shell-tempered ceramics (which are too fragmentary to<br />
classify on a specific level). No European trade items have been recovered <strong>in</strong> our work at<br />
One Cypress Po<strong>in</strong>t, but limited test<strong>in</strong>g does not rule out <strong>the</strong>ir possible existence here. We<br />
are aware <strong>of</strong> <strong>the</strong> complex problem <strong>of</strong> ascrib<strong>in</strong>g Quapaw phase archeological rema<strong>in</strong>s to <strong>the</strong><br />
historic Quapaw people (House and McKelway 1980). However, we are confident that One<br />
Cypress Po<strong>in</strong>t conta<strong>in</strong>s <strong>the</strong> rema<strong>in</strong>s <strong>of</strong> an occupation by Quapaw phase people, regardless <strong>of</strong><br />
<strong>the</strong>ir l<strong>in</strong>guistic identity and po<strong>in</strong>t <strong>of</strong> orig<strong>in</strong>. It thus becomes <strong>of</strong> great <strong>in</strong>terest to exam<strong>in</strong>e <strong>the</strong><br />
temporal and functional nature <strong>of</strong> this small settlement or camp.<br />
Area A. The larger <strong>of</strong> two activity or residence loci was exam<strong>in</strong>ed by two systematic<br />
(2, 3) and two judgmental (4, 5) test pits and by 13 shovel tests, all excavated to sterile soil at<br />
about 0.3 m depth. Test pits near <strong>the</strong> bankl<strong>in</strong>e produced chert flakes and o<strong>the</strong>r material <strong>in</strong><br />
<strong>the</strong> first level primarily, while Test Pit 2 on slightly higher terra<strong>in</strong> was most productive <strong>in</strong> <strong>the</strong><br />
second 10 cm level excavated. Nodena arrow po<strong>in</strong>t fragments, shell-tempered sherds, chert<br />
debitage, and o<strong>the</strong>r debris characterize this shallow buried zone at about 0.1-0.2 m depth.<br />
No organic rema<strong>in</strong>s and no <strong>in</strong>dications <strong>of</strong> features were uncovered, although irregular fired<br />
clay particles and sparse fire-cracked rock fragments were present <strong>in</strong> <strong>the</strong>se test pits. Chapter<br />
7, Table 18, conta<strong>in</strong>s soil test data for three samples from Test Pit 2, where soil conditions are<br />
extremely acid. Shovel tests extend <strong>the</strong> dimensions <strong>of</strong> Area A and suggest that much more<br />
<strong>of</strong> <strong>the</strong> shallow buried horizon rema<strong>in</strong>s here, unaffected by sheet or bankl<strong>in</strong>e erosion (Figure<br />
47). A significant collection <strong>of</strong> stone tools, debitage, sherds, and o<strong>the</strong>r imperishable debris<br />
is <strong>the</strong> pr<strong>in</strong>cipal result <strong>of</strong> <strong>the</strong> test<strong>in</strong>g work. This collection and that from Area B have been<br />
analyzed <strong>in</strong> considerable detail, and <strong>the</strong> results will be summarized below and <strong>in</strong> Chapter 8.<br />
Area B. Test excavation <strong>in</strong> this smaller locus consisted <strong>of</strong> two systematic test pits (1, 6),<br />
one random test pit (7), and 15 shovel tests to sterile soil at 0.3 m depth. The results obta<strong>in</strong>ed<br />
are entirely similar to those <strong>in</strong> Area B. The shallow buried horizon at 0.1-0.2 m depth is best<br />
represented <strong>in</strong> Test Pit 7 where all <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g materials were recovered: chert arrow<br />
po<strong>in</strong>t tip, chert microcore, chert utilized and retouched flakes, chert and novaculite debitage,<br />
fire-cracked sandstone, traces <strong>of</strong> burned bone, and one charred persimmon seed. Aga<strong>in</strong>,<br />
<strong>the</strong>se rema<strong>in</strong>s suggest that an occupation zone at shallow depth is partly unaffected by erosion<br />
and extends well beyond <strong>the</strong> bankl<strong>in</strong>e artifact scatter.<br />
Overview. One Cypress Po<strong>in</strong>t (3AS286) is a shallow buried site believed to have been<br />
occupied by term<strong>in</strong>al Mississippi period or Quapaw phase hunters between about A.D. 1600<br />
and <strong>17</strong>00. The settlement or camp extended over an area <strong>of</strong> about 700 m 2 , so far as we can<br />
tell, and <strong>in</strong>cluded two loci <strong>of</strong> activity on <strong>the</strong> Sal<strong>in</strong>e River bank and adjacent lakeshore. The<br />
rema<strong>in</strong>s <strong>of</strong> this camp are a rich assemblage <strong>of</strong> tools and imperishable debris, but evidently<br />
conta<strong>in</strong> few traces <strong>of</strong> structures or features and only rare organic debris. Table 12 below<br />
summarizes this site assemblage.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Site 183<br />
Table 12. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area, One<br />
Cypress Po<strong>in</strong>t (3AS286).<br />
Occupation Area Sherds Stone Tools Debitage Fire-cracked Rock Fired Clay<br />
Area A 79 133 847 5 124<br />
Area b 5 45 237 2 9<br />
Site Totals 84 <strong>17</strong>8 1084 7 133<br />
Chert and novaculite tools and toolmak<strong>in</strong>g debris from One Cypress Po<strong>in</strong>t appear<br />
to constitute <strong>the</strong> best available sample from a floodpla<strong>in</strong> site <strong>in</strong> our project area. The lithic<br />
assemblage has been analyzed and tabulated by functional categories for tools (Figure 49a)<br />
and by reduction categories for debitage (Figure 49b). Expediently utilized flakes, usually <strong>of</strong><br />
small size, with one or more cutt<strong>in</strong>g edges sheared or nibbled by use, comprise 64% <strong>of</strong> this<br />
assemblage. Bifaces <strong>of</strong> several presumed functional types, <strong>in</strong>clud<strong>in</strong>g arrow po<strong>in</strong>ts and arrow<br />
po<strong>in</strong>t preforms, comprise an additional 31%. Endscrapers (Figure 48c) which must have<br />
been employed <strong>in</strong> hidework, and o<strong>the</strong>r ma<strong>in</strong>tenance tools are sparsely represented. The sole<br />
occurrence <strong>of</strong> a ground stone tool is a hematite plummet from Area A (Figure 64c), which is<br />
quite likely out <strong>of</strong> context among <strong>the</strong> Quapaw phase materials.<br />
Although Nodena arrow po<strong>in</strong>ts predom<strong>in</strong>ate, s<strong>in</strong>gle examples <strong>of</strong> o<strong>the</strong>r styles are present<br />
(Figure 58a-c). These <strong>in</strong>clude triangular, concave-based willow leaf, and stemmed arrow<br />
po<strong>in</strong>t with barbs. The barbed arrow po<strong>in</strong>t (fragment not illustrated) is not “typical” <strong>of</strong> Nodena<br />
phase and Quapaw phase assemblages, and may be a local style or trace <strong>of</strong> earlier Mississippi<br />
period site use. Nodena po<strong>in</strong>ts are not known <strong>in</strong> <strong>the</strong> Shallow Lake site assemblage<br />
(Rol<strong>in</strong>gson and Schambach 1981), and were previously unknown <strong>in</strong> <strong>the</strong> Felsenthal region. In<br />
addition to One Cypress Po<strong>in</strong>t, we have recorded a s<strong>in</strong>gle Nodena po<strong>in</strong>t f<strong>in</strong>d at 3AS310, 1.9<br />
river miles (3 km) upstream on <strong>the</strong> Sal<strong>in</strong>e River.<br />
Figure 49b compares total samples <strong>of</strong> debitage from Areas A and B by means <strong>of</strong> cumulative<br />
frequencies for lithic reduction categories. This comparison <strong>in</strong>dicates that reduction <strong>of</strong><br />
pebble chert and novaculite <strong>in</strong> <strong>the</strong>se areas was highly similar with regard to <strong>the</strong> debris produced<br />
and discarded. Production <strong>of</strong> bifaces was an important task <strong>in</strong> both areas. The choice<br />
<strong>of</strong> raw material was also similar—chert <strong>of</strong> upland terrace orig<strong>in</strong> comprises 89.5 and 92.0% <strong>of</strong><br />
Area A and B samples <strong>of</strong> debitage (Chapter 8). F<strong>in</strong>ally, approximately 10-25% <strong>of</strong> cores and<br />
debitage from all subsamples <strong>in</strong> both areas exhibit obtrusive heat alteration and imply heat<br />
treatment <strong>of</strong> chert and novaculite raw material to improve its knapp<strong>in</strong>g quality.<br />
One Cypress Po<strong>in</strong>t furnishes an excellent sample <strong>of</strong> flaked stone tools and debris which<br />
may perta<strong>in</strong> to a s<strong>in</strong>gle brief seventeenth century occupation. Moreover, <strong>the</strong>se rema<strong>in</strong>s are
184 Hemm<strong>in</strong>gs<br />
Figure 49. One Cypress Po<strong>in</strong>t (3AS286), graphic presentation <strong>of</strong> lithic assemblage. a.<br />
pie diagram show<strong>in</strong>g relative proportions <strong>of</strong> flaked stone tools; b. cumulative<br />
frequency curves for cores and categories <strong>of</strong> debitage <strong>in</strong> Areas A<br />
and B.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 185<br />
likely to derive from <strong>the</strong> all male activity <strong>of</strong> a seasonal hunt<strong>in</strong>g camp <strong>in</strong> <strong>the</strong> Sal<strong>in</strong>e River<br />
floodpla<strong>in</strong>. This statement, is, <strong>of</strong> course, subject to future field and laboratory f<strong>in</strong>d<strong>in</strong>gs at<br />
3AS286 (see Appendix C).<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76)<br />
This small midden site is located on <strong>the</strong> “gooseneck” <strong>of</strong> Jug Po<strong>in</strong>t Bend at <strong>the</strong> entrance<br />
to a chute or slough. Dur<strong>in</strong>g ris<strong>in</strong>g and fall<strong>in</strong>g flood stages, <strong>the</strong> chute conveys a strong current<br />
across <strong>the</strong> gooseneck. Eventually neck cut<strong>of</strong>f here will produce a sizable oxbow lake,<br />
but Government Land Office maps <strong>of</strong> 1827 and 1844 aga<strong>in</strong> <strong>in</strong>dicate stability <strong>of</strong> <strong>the</strong> present<br />
Sal<strong>in</strong>e River channel dur<strong>in</strong>g <strong>the</strong> past century and a half.<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f was discovered as a small dense sherd scatter on <strong>the</strong> eroded bankl<strong>in</strong>e<br />
<strong>of</strong> <strong>the</strong> slough from <strong>the</strong> water’s edge almost to <strong>the</strong> modern levee surface, a 2.2 m vertical<br />
distance (62.2-69.1 feet MSL). The s<strong>in</strong>gle component identified from surface collection and<br />
test<strong>in</strong>g is evidently much like that present <strong>in</strong> Area A <strong>of</strong> <strong>the</strong> False Indigo site (3AS285), that is<br />
a shallow buried midden <strong>of</strong> <strong>the</strong> Mississippi period, Gran <strong>Marais</strong> phase (ca A.D. 1100-1400).<br />
Test excavation at Jug Po<strong>in</strong>t Cut<strong>of</strong>f consisted <strong>of</strong> one judgmental test pit to 0.4 m depth<br />
and 21 shovel tests rang<strong>in</strong>g from 0.4 to 0.6 m depth (Figure 50). Test Pit 1 <strong>in</strong>tersected a dense<br />
organic midden lens from about 0.1 to 0.4 m <strong>in</strong> depth, slop<strong>in</strong>g eastward slightly toward <strong>the</strong><br />
slough. Abundant mussel shell (four or more species) and alkal<strong>in</strong>e midden soil are among<br />
<strong>the</strong> chief constituents <strong>of</strong> this refuse deposit (Appendix B, Chapter 7). O<strong>the</strong>r contents recovered<br />
<strong>in</strong> Test Pit 1 are a chert Ashley arrow po<strong>in</strong>t (Figure 58i), a s<strong>in</strong>gle novaculite flake, abundant<br />
sherds, and fired clay particles.<br />
Three wet screen samples <strong>of</strong> about 2 liters volume each were taken from dense midden<br />
<strong>in</strong> Test Pit 1 and shovel tests nearby at 0.1 to 0.2 m depth. No quantitative results were<br />
obta<strong>in</strong>ed, but <strong>the</strong> qualitative results are consistent among all samples. Fishbone occurs <strong>in</strong><br />
pr<strong>of</strong>usion as small burned and calc<strong>in</strong>ed fragments, <strong>in</strong>clud<strong>in</strong>g vertebrae, sp<strong>in</strong>es, otoliths, and<br />
o<strong>the</strong>r elements. Gar scales and drum teeth are numerous. Considerable variation <strong>in</strong> size <strong>of</strong><br />
elements (and species ?) may <strong>in</strong>dicate that nets or traps were employed. O<strong>the</strong>r organic midden<br />
contents, generally <strong>of</strong> less frequency, are small animal bone, crawfish exoskeleton (?),<br />
wood charcoal, carbonized seed fragment (?), siliceous s<strong>in</strong>ter (?), herbaceous plant stem (?),<br />
and several animal coprolites with bone or fiber content. Under <strong>the</strong> b<strong>in</strong>ocular microscope,<br />
both plant and animal matter are coated with a cement or precipitate, and some bone fragments<br />
are rounded or altered by soil acidity or partial digestion. On <strong>the</strong> whole, however,<br />
this deposit has high quality preservation <strong>of</strong> fishbone and o<strong>the</strong>r organic debris.<br />
Shovel tests extend <strong>the</strong> dimensions <strong>of</strong> surface artifact scatter and subsurface midden<br />
as <strong>in</strong>dicated <strong>in</strong> Figure 50. Only three <strong>of</strong> 21 such tests were sterile; all o<strong>the</strong>rs produced sherds<br />
and o<strong>the</strong>r debris sometimes <strong>in</strong> appreciable quantity.
186 Hemm<strong>in</strong>gs<br />
Figure 50. Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76), test excavations November 1979. Hachured area is presumed extent<br />
<strong>of</strong> subsurface midden lens.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 187<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76) now appears as a shallow buried occupation area, 11 x <strong>17</strong><br />
m and irregular <strong>in</strong> plan, enclos<strong>in</strong>g a smaller circular midden deposit. The follow<strong>in</strong>g totals<br />
for imperishable artifact and debris are recorded: 250 sherds, 2 stone tools, 6 items <strong>of</strong> debitage,<br />
2 fire-cracked rock fragments, and 599 particles <strong>of</strong> fired clay. Many <strong>of</strong> <strong>the</strong> observations<br />
noted for Gran <strong>Marais</strong> occupation at False Indigo (3AS285) apply equally well to Jug Po<strong>in</strong>t<br />
Cut<strong>of</strong>f—here a shallow organic midden with high density <strong>of</strong> artifacts and diverse debris<br />
<strong>in</strong>dicates fish<strong>in</strong>g activity and wet cookery <strong>in</strong> a spatially restricted area.<br />
The Gran <strong>Marais</strong> ceramic complex is exclusively represented and <strong>the</strong> follow<strong>in</strong>g frequencies<br />
are noted: clay-tempered pla<strong>in</strong>, 56.8%; clay-tempered decorated, 3.6%; bone-tempered<br />
pla<strong>in</strong>, 1.2%; <strong>in</strong>determ<strong>in</strong>ate sherd fragments, 16.8%. Relatively more shell-tempered<br />
ceramics and o<strong>the</strong>r differences suggest that Jug Po<strong>in</strong>t Cut<strong>of</strong>f was occupied slightly later than<br />
<strong>the</strong> Gran <strong>Marais</strong> phase component at False Indigo (Chapter 8, Table 26). Independent dat<strong>in</strong>g<br />
evidence is needed to corroborate this sequence.<br />
Jug Po<strong>in</strong>t 1 (3AS306)<br />
This site is located on <strong>the</strong> low natural levee <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River at Jug Po<strong>in</strong>t Bend. It<br />
appears to be well exposed by cutbank and sheet erosion 75 m along <strong>the</strong> shorel<strong>in</strong>e, and was<br />
recorded as two discrete artifact scatters (Figure 51). Area A downstream is evidently exclusively<br />
a lithic scatter, while Area B upstream conta<strong>in</strong>s both shell-tempered sherds and lithic<br />
debris. Both artifact scatters <strong>in</strong>cluded occasional debris at <strong>the</strong> water’s edge or on <strong>the</strong> cutbank,<br />
and concentrations exposed by erosion near <strong>the</strong> levee surface. The bankl<strong>in</strong>e here has<br />
a vertical span <strong>of</strong> 1.8 m (62.2-68.1 feet MSL). An <strong>in</strong>cised shell-tempered sherd from Area B is<br />
shown <strong>in</strong> Figure 67b. The concentration <strong>in</strong> Area A also has produced <strong>the</strong> most unusual artifact<br />
now known <strong>in</strong> our project area, a perforated disc <strong>of</strong> fossil amber (?), discussed fur<strong>the</strong>r <strong>in</strong><br />
Chapter 8. We will present some data from surface collection and test excavation <strong>in</strong>dicat<strong>in</strong>g<br />
that Areas A and B are contemporaneous activity loci, and represent brief occupation dur<strong>in</strong>g<br />
<strong>the</strong> Late Mississippi period, Caney Bayou phase (A.D. 1400-1600).<br />
Area A. Test excavations <strong>in</strong> Area A consisted <strong>of</strong> one judgmental test pit and 18 shovel<br />
tests to about 0.4 m depth. Test Pit 1 was emplaced at <strong>the</strong> locus <strong>of</strong> productive shovel tests<br />
and also <strong>of</strong> an arrow po<strong>in</strong>t surface f<strong>in</strong>d. This chert arrow po<strong>in</strong>t has a dist<strong>in</strong>ctive “gar scale”<br />
silhouette, and has been provisionally described as an Eagle Creek po<strong>in</strong>t (with two o<strong>the</strong>r<br />
examples recovered at Jug Po<strong>in</strong>t 1; see Chapter 8 and Figure 58f). Chert and novaculite<br />
debitage was concentrated <strong>in</strong> Test Pit 1 at 0.1 to 0.2 m depth. Flat-ly<strong>in</strong>g flakes may mark an<br />
o<strong>the</strong>rwise imperceptible “floor” at 19 cm depth (Figure 2). No organic rema<strong>in</strong>s were noted<br />
<strong>in</strong> this test pit. Chert tools from <strong>the</strong> productive level <strong>in</strong>clude a bifacial cutt<strong>in</strong>g tool and two
188 Hemm<strong>in</strong>gs<br />
Figure 51. Jug Po<strong>in</strong>t 1 (3AS306), test excavations, November 1979.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 189<br />
Figure 52. Jug Po<strong>in</strong>t 1 (3AS206), test pit pr<strong>of</strong>iles.
190 Hemm<strong>in</strong>gs<br />
utilized flakes. Soil samples collected at and near <strong>the</strong> presumed floor are characterized <strong>in</strong><br />
Chapter 7, Table 18.<br />
Ten <strong>of</strong> <strong>the</strong> 18 shovel tests <strong>in</strong> Area A were sterile; all o<strong>the</strong>rs produced from one to seven<br />
chert or novaculite flakes at less than 0.3 m <strong>in</strong> depth. These shovel tests assisted us <strong>in</strong> def<strong>in</strong><strong>in</strong>g<br />
an elongate oval occupation zone, marked pr<strong>in</strong>cipally by debitage.<br />
Area B. A s<strong>in</strong>gle judgmental test pit was emplaced near productive shovel tests which<br />
yielded shell-tempered sherds and chert and novaculite debitage at less than 0.3 m depth.<br />
Test Pit 2 <strong>in</strong>tersected an occupation level or “floor” at <strong>17</strong>-19 cm below <strong>the</strong> levee surface and<br />
a small refuse filled pit (Feature 1) extend<strong>in</strong>g <strong>in</strong>to subsoil from this level ( Figure 52). The<br />
contents <strong>of</strong> Feature 1 <strong>in</strong>cluded <strong>the</strong> stem portion <strong>of</strong> an Eagle Creek po<strong>in</strong>t, three chert and<br />
novaculite biface fragments, five shell-tempered sherds, abundant chert and novaculite<br />
debitage, calc<strong>in</strong>ed mammal bone fragments, wood charcoal, carbonized seeds and nutshell<br />
(?), fired clay particles, and o<strong>the</strong>r residue. A 2-liter wet screen sample produced similar<br />
materials, and especially a small quantity <strong>of</strong> carbonize seeds. No fishbone has been noted <strong>in</strong><br />
Feature 1 refuse. The relatively large seeds and seed fragments are coated by a sediment film<br />
or precipitate and obscured, but are dist<strong>in</strong>ctive <strong>in</strong> size and shape and are tentatively identified<br />
as honey locust and persimmon. This suggests that Jug Po<strong>in</strong>t 1 occupation and disposal<br />
<strong>of</strong> food materials <strong>in</strong> Feature 1 occurred <strong>in</strong> fall.<br />
Only six <strong>of</strong> 14 shovel tests <strong>in</strong> Area B were sterile. O<strong>the</strong>rs yielded chert and novaculite<br />
debitage, shell-tempered sherds, and o<strong>the</strong>r debris at less than 0.3 m depth. Area B has a<br />
surface scatter <strong>of</strong> historic artifacts and debris which can be ascribed to a 1930-1940 hunter’s<br />
or fisherman’s camp (Figure 51).<br />
Overview. Two small discrete activity loci <strong>of</strong> <strong>the</strong> Caney Bayou phase (A.D. 1400-1600)<br />
are separated by about 25 m on <strong>the</strong> Sal<strong>in</strong>e River bank. Eagle Creek arrow po<strong>in</strong>ts are sparsely<br />
represented <strong>in</strong> both loci, while shell-tempered sherds, mostly pla<strong>in</strong>ware, are restricted to<br />
Area B. Shallow occupation floors were <strong>in</strong>tersected by test pits at about <strong>the</strong> same depth (<strong>17</strong>-<br />
19 cm) <strong>in</strong> each locus. Arrow po<strong>in</strong>t style and depth <strong>of</strong> alluviation suggest that <strong>the</strong>se areas are<br />
coeval.<br />
Feature 1 <strong>in</strong> Area B furnished ord<strong>in</strong>ary food refuse which is important to <strong>the</strong> <strong>in</strong>terpretation<br />
<strong>of</strong> seasonal use at Jug Po<strong>in</strong>t 1. Calc<strong>in</strong>ed mammal bone and charred seeds <strong>of</strong> honey<br />
locus and persimmon are <strong>in</strong>termixed with o<strong>the</strong>r debris <strong>in</strong> a small pit. Nutshell may also be<br />
present <strong>in</strong> small quantities. The seed pods, nuts, and fruits <strong>of</strong> <strong>the</strong>se trees were undoubtedly<br />
available <strong>in</strong> <strong>the</strong> natural levee forest dur<strong>in</strong>g fall.<br />
The content <strong>of</strong> Areas A and B, known from surface collection and limited test<strong>in</strong>g, presents<br />
an <strong>in</strong>terest<strong>in</strong>g contrast. Table 13 summarizes <strong>the</strong> occurrence <strong>of</strong> imperishable artifacts<br />
and debris <strong>in</strong> <strong>the</strong>se areas. It may be suggested that flaked stone tools and debitage <strong>in</strong> Area<br />
A represent male items and male activity, and that ceramics and food preparation refuse <strong>in</strong><br />
Area B represent female items and debris (although stone tools and debitage are present
Assessment <strong>of</strong> Floodpla<strong>in</strong> Site 191<br />
here too). A family group or task unit <strong>of</strong> this k<strong>in</strong>d would reflect quite different strategy for<br />
exploit<strong>in</strong>g floodpla<strong>in</strong> resources than that proposed earlier for One Cypress Po<strong>in</strong>t (3AS286).<br />
Table 13. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area, Jug<br />
Po<strong>in</strong>t 1 (AS306)<br />
Occupation Area Sherds Stone Tools Debitage Fire-cracked Rock Fired Clay<br />
Entire Site Surface Collection 2 4 74 1 6<br />
Area A 0 12 164 0 0<br />
Area b 60 7 439 7 107<br />
Site Totals 62 23 677 8 113<br />
F<strong>in</strong>ally, <strong>the</strong>re is <strong>the</strong> highly unusual f<strong>in</strong>d at Jug Po<strong>in</strong>t 1, Area A, <strong>of</strong> an amber (?) disc or<br />
bead among o<strong>the</strong>rwise ord<strong>in</strong>ary debitage. Detailed discussion <strong>of</strong> this specimen is presented<br />
<strong>in</strong> Chapter 8.<br />
Jug Po<strong>in</strong>t 2 (3AS307)<br />
Jug Po<strong>in</strong>t 2 (3AS307) occupies <strong>the</strong> same bank <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River about 70 m upstream<br />
from Jug Po<strong>in</strong>t 1. The bankl<strong>in</strong>e, about 1.9 <strong>in</strong> height (62.2-68.3 feet MSL), slopes gently upward<br />
from <strong>the</strong> water’s edge. This site was first recorded as a sparse artifact scatter extend<strong>in</strong>g<br />
80 m along <strong>the</strong> shorel<strong>in</strong>e, but was subsequently collected, tested, and, mapped as three<br />
discrete and closely spaced activity or residential loci (Figure 53). Area A downstream is a<br />
sparse lithic scatter, while Areas B and C upstream conta<strong>in</strong> both shell-tempered and claytempered<br />
ceramics and lithic debris. This site is <strong>the</strong>refore extremely similar <strong>in</strong> layout and<br />
content to Jug Po<strong>in</strong>t 1, and one might ask if <strong>the</strong>se two sites are a s<strong>in</strong>gle Mississippi period<br />
component consist<strong>in</strong>g <strong>of</strong> five successive bankl<strong>in</strong>e activity loci. Ceramic and o<strong>the</strong>r artifact<br />
data, however, suggest to us that Jug Po<strong>in</strong>t 2 was occupied early <strong>in</strong> <strong>the</strong> Caney Bayou phase,<br />
between about A.D. 140 and 1500, and prior to occupation at Jug Po<strong>in</strong>t 1. A s<strong>in</strong>gle radiocarbon<br />
sample taken from Area C did not provide <strong>the</strong> chronological control needed to answer<br />
this question (Chapter 7).<br />
Area A. This area is known only from surface collection. The pr<strong>in</strong>cipal artifacts recovered<br />
here are a novaculite Ashley po<strong>in</strong>t (Figure 58h), and chert and silicified wood debitage.<br />
Thirteen shovel tests to about 0.3 m depth were sterile.<br />
Area B. A s<strong>in</strong>gle judgmental test pit and 10 shovel tests to 0.3 m depth were emplaced<br />
<strong>in</strong> Area B. Test Pit 2 produced a few Baytown Pla<strong>in</strong> sherds and manuports at 0.2 to 0.3 m<br />
depth, and was o<strong>the</strong>rwise sterile. Surface f<strong>in</strong>ds from this area <strong>in</strong>clude a novaculite Ashley<br />
po<strong>in</strong>t (Figure 58h), novaculite drill (Figure 61f), chert debitage, shell-tempered sherds, and<br />
fired clay particles. One shovel test recovered a pla<strong>in</strong> shell-tempered sherd, while n<strong>in</strong>e o<strong>the</strong>rs<br />
were sterile.
192 Hemm<strong>in</strong>gs<br />
Figure 53. Jug Po<strong>in</strong>t 2 (3AS307), test excavations, November 1979.
Assessment <strong>of</strong> Floodpla<strong>in</strong> Site 193<br />
Area C. Area C is a larger and more diverse artifact scatter, and has been <strong>in</strong>vestigated<br />
by means <strong>of</strong> one judgmental test pit and 1 shovel tests to about 0.3 m depth. Test Pit 1<br />
disclosed a subtle, but visible, occupation level at 14 cm below <strong>the</strong> levee surface, extend<strong>in</strong>g<br />
4 to 8 cm <strong>in</strong>to subsoil. This th<strong>in</strong> zone is organically sta<strong>in</strong>ed and conta<strong>in</strong>s artifacts, calc<strong>in</strong>ed<br />
mammal (?) bone fragments, sparse wood charcoal, and a few carbonized seeds, probably<br />
all honey locust. A collection <strong>of</strong> charcoal fragments from <strong>the</strong> entire test square (14-22 cm<br />
organic zone) was comb<strong>in</strong>ed as one radiocarbon sample (see Chapter 8). The pr<strong>in</strong>cipal artifacts<br />
from this level are a chert Bassett arrow po<strong>in</strong>t (Figure 58g), a chert spokeshave, chert<br />
and novaculite debitage, and a small group <strong>of</strong> th<strong>in</strong> pla<strong>in</strong> body sherds (about equally divided<br />
between clay and shell temper). Characteristics <strong>of</strong> a soil sample from <strong>the</strong> occupation level<br />
are tabulated <strong>in</strong> Chapter 7.<br />
Six <strong>of</strong> 15 shovel tests <strong>in</strong> Area C produced artifacts or debris above 0.2 m depth, and<br />
<strong>the</strong>se tests extend <strong>the</strong> known dimensions <strong>of</strong> this shallow occupation level to about 8 x 15 m.<br />
Overview. Jug Po<strong>in</strong>t 2 (3AS307) consists <strong>of</strong> three small activity or residential loci with<strong>in</strong><br />
an 80 m stretch <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River bankl<strong>in</strong>e at Jug Po<strong>in</strong>t Bend. These loci differ <strong>in</strong> size and<br />
content, and may or may not be strictly contemporary. All occur at very shallow depth, less<br />
than 0.2 m, beneath <strong>the</strong> modern levee surface. The co-occurrence <strong>of</strong> Ashley arrow po<strong>in</strong>ts <strong>in</strong><br />
Areas A and B, and <strong>of</strong> shell-tempered sherds with a th<strong>in</strong> late variety <strong>of</strong> Baytown (?) pla<strong>in</strong>ware<br />
<strong>in</strong> Areas B and C leads us to place <strong>the</strong> occupation <strong>of</strong> Jug Po<strong>in</strong>t 2 early <strong>in</strong> <strong>the</strong> Caney<br />
Bayou phase, or about A.D. 1400 (see also Chapter 8:Table 26).<br />
The presumed functional contrast among activity loci noted at Jug Po<strong>in</strong>t 1 is evidently<br />
repeated at Jug Po<strong>in</strong>t 2 60 m upstream. Area A conta<strong>in</strong>s male-produced lithic items, while<br />
Areas C conta<strong>in</strong>s a mixed assemblage <strong>of</strong> ceramic and lithic materials with an <strong>in</strong>crement <strong>of</strong><br />
organic refuse. Food rema<strong>in</strong>s <strong>in</strong>clude small quantities <strong>of</strong> calc<strong>in</strong>ed mammal bone and carbonized<br />
honey locust seeds, <strong>the</strong> latter <strong>in</strong>dicat<strong>in</strong>g site use dur<strong>in</strong>g fall. Table 14 summarizes <strong>the</strong><br />
contrast between activity loci <strong>in</strong> terms <strong>of</strong> imperishable materials.<br />
Table 14. Total Recovery <strong>of</strong> Imperishable Artifacts and Debris by Occupation Area, Jug<br />
Po<strong>in</strong>t 2 (3AS307).<br />
Occupation Area Sherds Stone Tools Debitage Fire-cracked Rock Fired Clay<br />
Entire Site Surface Collection 67 2 44 1 2<br />
Area A 0 1 2 1 0<br />
Area b 10 2 0 0 2<br />
Area c 21 3 20 2 6<br />
Site Totals 98 8 66 4 10
194 Hemm<strong>in</strong>gs<br />
Mouth <strong>of</strong> Eagle Creek (3BR78)<br />
This small site lies on <strong>the</strong> low natural levee <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River at <strong>the</strong> mouth <strong>of</strong> Eagle<br />
Creek. Dur<strong>in</strong>g <strong>in</strong>itial survey, a large novaculite Gary po<strong>in</strong>t (Figure 59e), small quantities <strong>of</strong><br />
novaculite and chert debitage, and fire-cracked rock were found on <strong>the</strong> eroded bankl<strong>in</strong>e.<br />
Some recent historic debris, but no prehistoric artifacts, were found <strong>in</strong> a clear<strong>in</strong>g on <strong>the</strong> levee<br />
surface which reaches a height <strong>of</strong> 1.8 m (62.2-68.0 feet MSL) above river stage. The k<strong>in</strong>ds <strong>of</strong><br />
artifacts and debris recovered led us to believe that a relatively deep preceramic component<br />
is represented at Mouth <strong>of</strong> Eagle Creek.<br />
Test<strong>in</strong>g consisted <strong>of</strong> trowel<strong>in</strong>g and exam<strong>in</strong>ation <strong>of</strong> <strong>the</strong> bank pr<strong>of</strong>ile, emplacement <strong>of</strong><br />
two judgmental test pits (which were not excavated below 0.5 m due to rise <strong>of</strong> river level),<br />
and 15 shovel tests to 0.6 m depth. The general results <strong>of</strong> this work are <strong>in</strong>dications that a<br />
th<strong>in</strong>, charcoal-sta<strong>in</strong>ed lens or “floor” with sparse fire-cracked rock and flake is exposed at<br />
about 0.6-0.7 m depth, runn<strong>in</strong>g at least 9 m along <strong>the</strong> bank; o<strong>the</strong>r f<strong>in</strong>ds <strong>of</strong> fire-cracked rock<br />
and flakes were made from about 0.1 to 0.7 m depth at various locations with<strong>in</strong> a 50 m<br />
stretch <strong>of</strong> riverbank or <strong>in</strong> adjacent test units (Figure 54). Two <strong>of</strong> 15 shovel tests produced one<br />
flake each above 0.6 m; all o<strong>the</strong>rs were sterile. Nei<strong>the</strong>r Test Pits 1 or 2 reached <strong>the</strong> level <strong>of</strong> <strong>the</strong><br />
charcoal-sta<strong>in</strong>ed lens appear<strong>in</strong>g <strong>in</strong> <strong>the</strong> riverbank, and <strong>the</strong>refore we have <strong>in</strong>sufficient <strong>in</strong>formation<br />
about <strong>the</strong> age, content, and degree <strong>of</strong> preservation <strong>of</strong> this relatively deep occupation<br />
level.<br />
The pr<strong>in</strong>cipal occupation at Mouth <strong>of</strong> Eagle Creek (3BR78) was probably Poverty Po<strong>in</strong>t<br />
period, Calion phase (1200-500 B.C.) on slim evidence and on <strong>the</strong> basis <strong>of</strong> comparison to<br />
this early component at Marie Sal<strong>in</strong>e (3AS329). Gary dart po<strong>in</strong>ts, white novaculite debitage,<br />
and fire-cracked rock are to be expected <strong>in</strong> such a component. For comparative purposes <strong>the</strong><br />
follow<strong>in</strong>g total sample <strong>of</strong> imperishable artifacts and debris is noted: 1 stone tool, 32 items <strong>of</strong><br />
debitage, and <strong>17</strong> fire-cracked rock fragments. Very sparse rema<strong>in</strong>s at shallow depth <strong>in</strong>dicate<br />
occasional low <strong>in</strong>tensity site use at Mouth <strong>of</strong> Eagle Creek dur<strong>in</strong>g subsequent ceramic<br />
periods, although we have not presently recovered sherds or o<strong>the</strong>r diagnostic late artifacts.<br />
Completion <strong>of</strong> test<strong>in</strong>g, and perhaps extensive mitigation, are needed to <strong>in</strong>vestigate <strong>the</strong><br />
nature <strong>of</strong> <strong>the</strong> early component, obta<strong>in</strong> samples <strong>of</strong> organic material for dat<strong>in</strong>g and identification,<br />
and verify <strong>the</strong> presence <strong>of</strong> younger superposed components (see Appendix C).<br />
eVAluAtion <strong>of</strong> test excAVAtion strAtegy And tActics<br />
Test excavation was undertaken at selected sites <strong>in</strong> order to assess <strong>the</strong>ir significance as<br />
cultural resources and to acquire objective data as a basis for recommendations about management<br />
<strong>of</strong> <strong>the</strong>se resources. A large, systematic body <strong>of</strong> data has <strong>in</strong> fact resulted from this<br />
work. These data and <strong>in</strong>terpretations placed on <strong>the</strong>m are, we believe, fundamental contributions<br />
to regional archeological research and to management needs <strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area, although <strong>the</strong>y are prelim<strong>in</strong>ary <strong>in</strong> nature. The follow<strong>in</strong>g brief evaluation <strong>of</strong> our work
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 195<br />
Figure 54. Mouth <strong>of</strong> Eagle Creek (3BR78) test excavations, November 1979.
196 Hemm<strong>in</strong>gs<br />
notes only <strong>the</strong> particular successes and difficulties <strong>of</strong> <strong>the</strong> test<strong>in</strong>g stage, perceived <strong>in</strong> retrospect.<br />
1. Test<strong>in</strong>g effort concentrated on one major Ouachita River prehistoric site and<br />
a cluster <strong>of</strong> lower Sal<strong>in</strong>e River sites; <strong>the</strong>se are shown to be regionally significant,<br />
but test<strong>in</strong>g <strong>of</strong> additional prehistoric and historic sites throughout <strong>the</strong><br />
project area would greatly re<strong>in</strong>force <strong>the</strong> data base and validity <strong>of</strong> <strong>the</strong> results.<br />
2. Vertical control on <strong>in</strong>dividual sites was ma<strong>in</strong>ta<strong>in</strong>ed scrupulously so that we<br />
could differentiate superposed assemblages and relate buried components<br />
to <strong>the</strong> proposed navigation pool; however, exist<strong>in</strong>g lock stage and river level<br />
were employed as a horizontal datum (slackwater pool) when some slope<br />
<strong>of</strong> <strong>the</strong> water surface must be present, and errors <strong>in</strong> tenths <strong>of</strong> feet are likely to<br />
affect our site records; average slopes <strong>of</strong> reaches on <strong>the</strong> Ouachita and Sal<strong>in</strong>e<br />
rivers could have been determ<strong>in</strong>ed by <strong>in</strong>strument and <strong>in</strong>troduced as correction<br />
factors (Marie Sal<strong>in</strong>e site elevations refer directly to read<strong>in</strong>gs on a river<br />
gauge and thus are considered accurate).<br />
3. The basel<strong>in</strong>e method <strong>of</strong> horizontal control, used with systematic, judgmental,<br />
and random test pits, was quick and efficient on our floodpla<strong>in</strong> sites characterized<br />
by small size, elongate configuration, and shallow buried components;<br />
systematic surface collection at several levels <strong>of</strong> precision was also<br />
employed to discrim<strong>in</strong>ate activity loci and site boundaries <strong>in</strong> shallow components;<br />
s<strong>in</strong>ce none <strong>of</strong> <strong>the</strong>se sites is disturbed by plow<strong>in</strong>g (or by o<strong>the</strong>r modern<br />
processes to significant extent), we th<strong>in</strong>k a presumption <strong>of</strong> spatial pattern<strong>in</strong>g<br />
is necessary, and even more rigorous controlled collection techniques could<br />
justifiably be tried and evaluated.<br />
4. Excavation, screen<strong>in</strong>g, and wet screen<strong>in</strong>g techniques were uniformly applied,<br />
but difficult, because <strong>of</strong> relatively high clay content <strong>in</strong> floodpla<strong>in</strong> deposits<br />
(Chapter 7, Tables 8 and 19); experimentation is needed here to determ<strong>in</strong>e<br />
<strong>the</strong> most efficient and reliable techniques <strong>of</strong> recovery <strong>of</strong> materials, <strong>in</strong>clud<strong>in</strong>g<br />
delicate faunal and floral rema<strong>in</strong>s; flotation experimentation is needed so that<br />
preservation can be rigorously assessed and organic rema<strong>in</strong>s identified and<br />
quantified on a statistical basis; dispersion or deflocculation <strong>of</strong> matrix will be<br />
required when flotation sampl<strong>in</strong>g is implemented.<br />
5. The experience <strong>of</strong> both survey and test excavation is that power auger<strong>in</strong>g,<br />
while very quick, is <strong>in</strong>effective <strong>in</strong> locat<strong>in</strong>g buried components with low density<br />
<strong>of</strong> <strong>in</strong>organic and organic debris, as characterizes most floodpla<strong>in</strong> sites <strong>in</strong><br />
<strong>the</strong> project area; <strong>the</strong> slower method <strong>of</strong> shovel test<strong>in</strong>g does prove effective to<br />
about 0.5 m depth, presumably because it greatly <strong>in</strong>creases <strong>the</strong> volume <strong>of</strong>
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 197<br />
soil extracted by each test; although we have not employed deep test<strong>in</strong>g by<br />
means <strong>of</strong> core drill<strong>in</strong>g or backhoe trench<strong>in</strong>g <strong>in</strong> <strong>the</strong> project area, we believe<br />
such equipment could be effective <strong>in</strong> carefully designed modes <strong>of</strong> use.<br />
suMMAry AssessMent <strong>of</strong> floodplA<strong>in</strong> sites<br />
Results <strong>of</strong> test excavation <strong>in</strong> eight floodpla<strong>in</strong> sites (described <strong>in</strong> this chapter) and<br />
subsequent analyses <strong>of</strong> materials recovered (Chapters 7, 8, Appendixes A, B) are <strong>the</strong> basis<br />
for evaluat<strong>in</strong>g <strong>the</strong>se sites for potential eligibility and for nom<strong>in</strong>ation to <strong>the</strong> National Register<br />
<strong>of</strong> Historic Places. Our f<strong>in</strong>d<strong>in</strong>gs and recommendations are summarized <strong>in</strong> Table 16. More<br />
detailed discussion <strong>of</strong> significance, adverse effects, and proposed mitigation is conta<strong>in</strong>ed <strong>in</strong><br />
Appendix C. All sites selected for test<strong>in</strong>g are buried components with relatively high <strong>in</strong>tegrity,<br />
hav<strong>in</strong>g a degree <strong>of</strong> alteration or loss by riverbank erosion, but no significant historical<br />
disturbance. No site tested has been truncated or superficially disturbed by plow<strong>in</strong>g.<br />
Elevations cited <strong>in</strong> Table 16 are those obta<strong>in</strong>ed by methods described earlier <strong>in</strong> this<br />
chapter for prehistoric cultural rema<strong>in</strong>s <strong>of</strong> diverse k<strong>in</strong>ds <strong>in</strong> primary context (persist<strong>in</strong>g essentially<br />
at <strong>the</strong> orig<strong>in</strong>al locus <strong>of</strong> use). In every case artifacts and debris were also recovered<br />
dur<strong>in</strong>g <strong>in</strong>itial survey at <strong>the</strong> water’s edge and on eroded bankl<strong>in</strong>es at various elevations<br />
above river level. Test<strong>in</strong>g has shown lowly<strong>in</strong>g cultural rema<strong>in</strong>s to be eroded and transported<br />
(<strong>in</strong> secondary context) from contour <strong>in</strong> <strong>the</strong> case <strong>of</strong> six sites. We believe that all sites tested<br />
will be adversely affected by <strong>the</strong> 65-foot navigation pool, when cultural rema<strong>in</strong>s are eroded<br />
and removed at <strong>the</strong> shorel<strong>in</strong>e, regardless <strong>of</strong> any case where primary context occurs above 65<br />
feet (see also Appendix C).<br />
As Table 15 <strong>in</strong>dicates, all eight floodpla<strong>in</strong> sites tests are scientifically important and<br />
“have yielded or may be likely to yield <strong>in</strong>formation important <strong>in</strong> prehistory” (36 CFR<br />
1202.6). We <strong>the</strong>refore recommend seven sites be nom<strong>in</strong>ated to <strong>the</strong> National Register <strong>of</strong> Historic<br />
Places; <strong>the</strong> rema<strong>in</strong><strong>in</strong>g site (Buttonbush 3BR58) is not eligible by virtue <strong>of</strong> <strong>the</strong> fact that all<br />
known or expected cultural rema<strong>in</strong>s and related site data were recovered dur<strong>in</strong>g test<strong>in</strong>g.<br />
site functionAl hypo<strong>the</strong>ses<br />
Test excavation <strong>in</strong> floodpla<strong>in</strong> sites furnished data relevant to six <strong>of</strong> eight hypo<strong>the</strong>ses<br />
stated <strong>in</strong> <strong>the</strong> research design (Chapter 1), but we are here concerned with one key hypo<strong>the</strong>sis<br />
regard<strong>in</strong>g <strong>the</strong> functional nature <strong>of</strong> such sites. The quality <strong>of</strong> data also permits us to go<br />
beyond this <strong>in</strong>itial hypo<strong>the</strong>sis and ref<strong>in</strong>e or present new site functional hypo<strong>the</strong>ses perta<strong>in</strong><strong>in</strong>g<br />
to <strong>the</strong> Mississippi and earlier periods. Our <strong>in</strong>itial hypo<strong>the</strong>sis was stated as follows:
198 Hemm<strong>in</strong>gs<br />
Table 15. Summary Assessment <strong>of</strong> Eight Floodpla<strong>in</strong> Sites.<br />
Significance National Register General<br />
Site Name and Elevation* <strong>of</strong> Eligibility Mitigation<br />
Designation Components) Site Specific Data Research Potential Recommendation Recommendation**<br />
Marie Sal<strong>in</strong>e
Problem Doma<strong>in</strong> 1: Settlement Characteristics<br />
Assessment <strong>of</strong> Floodpla<strong>in</strong> Site 199<br />
H 1 If subsistence strategies employed wild plant and animal resources, impermanent<br />
specialized extractive sites would implement <strong>the</strong>se strategies <strong>in</strong> true floodpla<strong>in</strong><br />
environments.<br />
Test Implications: Recognition <strong>of</strong> specialized extractive sites as settlement types,<br />
dist<strong>in</strong>ct from base settlements.<br />
Data Requirements: Floodpla<strong>in</strong> site assemblages with high ratio <strong>of</strong> extractive/<br />
ma<strong>in</strong>tenance tool kits; faunal or floral <strong>in</strong>dicators <strong>of</strong> seasonality and procurement<br />
strategy; absence <strong>of</strong> permanent structures for habitation, storage, or burial.<br />
In <strong>the</strong> preced<strong>in</strong>g summaries <strong>of</strong> test excavation results, 12 components were isolated<br />
and identified <strong>in</strong> eight floodpla<strong>in</strong> sites; half <strong>of</strong> <strong>the</strong>se components represent Mississippi<br />
period occupation and <strong>the</strong> o<strong>the</strong>r half are attributed equally to Baytown-Coles Creek, Tchula,<br />
and Poverty Po<strong>in</strong>t period occupations. Data obta<strong>in</strong>ed <strong>in</strong> Mississippi period components is<br />
more diverse and more complete <strong>in</strong> several respects. It seems <strong>the</strong>n that we can best exam<strong>in</strong>e<br />
<strong>the</strong> functional nature <strong>of</strong> floodpla<strong>in</strong> sites established between about A.D. 100 and <strong>17</strong>00, draw<strong>in</strong>g<br />
on pre-Mississippi period components for comparison and contrast.<br />
Our results are potentially <strong>of</strong> great importance <strong>in</strong> <strong>the</strong> context <strong>of</strong> recent ecologically<br />
oriented studies <strong>of</strong> Mississippi period settlement and subsistence (Lewis 1974; Smith 1975;<br />
contributors <strong>in</strong> Smith 1978). Many <strong>of</strong> <strong>the</strong>se excellent studies employ site locational and size<br />
data with <strong>in</strong>ternal characteristics to derive settlement hierarchies <strong>in</strong> which farmsteads are<br />
<strong>the</strong> m<strong>in</strong>imal economic and residential units. Where very small extractive camps are identified<br />
<strong>in</strong> regional Mississippi period settlement hierarchies, <strong>the</strong> analysts seem to doubt <strong>the</strong>ir<br />
functional typologies. Green and Munson (1978:319), for example, can dist<strong>in</strong>guish relatively<br />
many Angel phase “hunt<strong>in</strong>g and/or ga<strong>the</strong>r<strong>in</strong>g camps” <strong>in</strong> <strong>the</strong> Ohio River floodpla<strong>in</strong>, terrace,<br />
and upland locations, but believe that “additional research will reveal many <strong>of</strong> <strong>the</strong> sites<br />
located <strong>in</strong> <strong>the</strong> river valley [as] farmsteads.” And Muller (1978:283) concludes for numerous<br />
small sites <strong>in</strong> <strong>the</strong> Black Bottom surround<strong>in</strong>g K<strong>in</strong>caid that “<strong>the</strong>re is really no good evidence<br />
to support an idea <strong>of</strong> widespread specialized extractive sites.” These views may be partly <strong>in</strong>consistent<br />
with <strong>the</strong> strategy <strong>of</strong> Mississippian subsistence <strong>in</strong> floodpla<strong>in</strong> environments, which<br />
had “two str<strong>in</strong>gs to its bow—wild and agricultural resources” (Peebles 1978:392).<br />
Before present<strong>in</strong>g data bear<strong>in</strong>g on Hypo<strong>the</strong>sis 1, we should amplify some <strong>of</strong> <strong>the</strong> terms<br />
which have been <strong>in</strong>corporated.<br />
1. A subsistence strategy is a schedule <strong>of</strong> decisions about, and priorities for, obta<strong>in</strong><strong>in</strong>g<br />
seasonally and spatially restricted food resources.<br />
2. Impermanent sites are loci <strong>of</strong> brief occupation or transient activity on a scale<br />
<strong>of</strong> days or weeks.
200 Hemm<strong>in</strong>gs<br />
3. Specialized extractive sites are loci <strong>of</strong> one or few procurement activity sets<br />
(not <strong>in</strong>clud<strong>in</strong>g horticultural activities) which are operated by few personnel or<br />
small task units.<br />
4. True floodpla<strong>in</strong> environments here refer to annually overflowed bottomlands<br />
which impose seasonal limits on site occupancy, as well as on seasonal availability<br />
<strong>of</strong> resources.<br />
The first data requirement is a high ratio <strong>of</strong> extractive to ma<strong>in</strong>tenance tools <strong>in</strong> site assemblages.<br />
Two pr<strong>in</strong>cipal sampl<strong>in</strong>g biases will affect our comparison: nonpreservation <strong>of</strong><br />
perishable equipment and curate behavior (conserv<strong>in</strong>g and remov<strong>in</strong>g useful tools). With regard<br />
to stone tools recovered <strong>in</strong> test excavations, we note first that <strong>the</strong>y are scarce (N= 256),<br />
rang<strong>in</strong>g from two at 3BR58 to <strong>17</strong>8 at 3AS286, and that <strong>the</strong>y are nearly all projectile po<strong>in</strong>ts,<br />
bifacial preforms, bifacial cutt<strong>in</strong>g tools, and utilized flakes with cutt<strong>in</strong>g edges. We note also<br />
that one or more drills, scrapers, perforators, beads, plummets, pitted cobbles, and steatite<br />
sherds are present, but numerically <strong>in</strong>significant <strong>in</strong> any site tested and <strong>in</strong> <strong>the</strong> group as a<br />
whole. There are no woodwork<strong>in</strong>g tools, no agricultural tools, and no gr<strong>in</strong>d<strong>in</strong>g implements.<br />
With regard to ceramic vessels, we note aga<strong>in</strong> <strong>the</strong>ir scarcity (N=2708 sherds, represent<strong>in</strong>g<br />
far fewer vessels) <strong>in</strong> ceramic components, rang<strong>in</strong>g from 62 at 3AS306 to 1663 at 3AS285,<br />
and that two Gran <strong>Marais</strong> phase components with fishbone midden (3AS285 and 3BR76)<br />
furnished 70% <strong>of</strong> <strong>the</strong> sherd sample. In <strong>the</strong>se components particularly, <strong>the</strong> systemic context<br />
<strong>of</strong> ceramic vessels may have been process<strong>in</strong>g <strong>of</strong> fish meat and oil for later distribution and<br />
consumption. Therefore we believe that extractive tool kits are dom<strong>in</strong>ant features <strong>of</strong> floodpla<strong>in</strong><br />
site assemblages, although variability exists among <strong>the</strong>se tool kits <strong>in</strong> <strong>the</strong> proportions<br />
<strong>of</strong> weapons, items associated with refurbishment <strong>of</strong> weapons, and utensils or conta<strong>in</strong>ers<br />
directly employed to butcher, render, boil, or o<strong>the</strong>rwise process foodstuffs.<br />
The second data requirement is <strong>the</strong> discovery <strong>of</strong> faunal and floral <strong>in</strong>dicators <strong>of</strong> seasonality<br />
and procurement strategy. The limited evidence can be quickly summarized as follows<br />
(see Potential Lowland Resources <strong>in</strong> Chapter 2).<br />
3AS285 rema<strong>in</strong>s <strong>of</strong> gar, drum, and o<strong>the</strong>r fish (late spr<strong>in</strong>g, summer)<br />
3AS286 animal rema<strong>in</strong>s, persimmon (fall)<br />
3BR76 rema<strong>in</strong>s <strong>of</strong> gar, drum, o<strong>the</strong>r fish, crawfish, and mussels (late spr<strong>in</strong>g, summer)<br />
3AS306 mammal rema<strong>in</strong>s, honey locust, persimmon (fall)<br />
3AS307 mammal rema<strong>in</strong>s, honey locust (fall)<br />
Suitable <strong>in</strong>dicators thus appear to be present, while quality <strong>of</strong> data is presently <strong>in</strong>adequate<br />
and cannot strongly support <strong>the</strong> hypo<strong>the</strong>sis.<br />
The last data requirement is an absence <strong>of</strong> permanent structures for habitation, storage,<br />
and burial. On <strong>the</strong> basis <strong>of</strong> bankl<strong>in</strong>e exposure and 5% subsurface <strong>in</strong>vestigation <strong>the</strong>se<br />
structural rema<strong>in</strong>s are <strong>in</strong>deed absent. We doubt whe<strong>the</strong>r any well preserved Mississippi<br />
period farmstead could be similarly tested without produc<strong>in</strong>g substantial <strong>in</strong>dications <strong>of</strong>
Assessment <strong>of</strong> Floodpla<strong>in</strong> Site 201<br />
<strong>the</strong>se k<strong>in</strong>ds <strong>of</strong> facilities. Such farmsteads are dist<strong>in</strong>guished <strong>in</strong> our view as relatively permanent<br />
base settlements, regardless <strong>of</strong> small size (Green and Munson [1978:310] rate Mississippi<br />
period farmsteads and camp similarly <strong>in</strong> respect to small size, “less than .25 ha”).<br />
An overall conclusion, qualified at various po<strong>in</strong>ts <strong>in</strong> <strong>the</strong> preced<strong>in</strong>g discussion, is that<br />
data acquired <strong>in</strong> test excavation <strong>of</strong> eight floodpla<strong>in</strong> sites tend to support <strong>the</strong> hypo<strong>the</strong>sis:<br />
impermanent specialized sites were established <strong>in</strong> <strong>the</strong> overflow bottoms <strong>of</strong> <strong>the</strong> Ouachita and<br />
Sal<strong>in</strong>e rivers to extract wild plant and animal resources.<br />
Variation Among Extractive Sites<br />
A f<strong>in</strong>al objective <strong>of</strong> our functional analysis is to consider (<strong>in</strong> brief and prelim<strong>in</strong>ary fashion)<br />
<strong>the</strong> implications <strong>of</strong> variation <strong>in</strong> extractive sites as a class. We shall look at <strong>the</strong> proportions<br />
<strong>of</strong> tools <strong>in</strong> assemblages, as W<strong>in</strong>ters (1969) and o<strong>the</strong>rs have done, but will also <strong>in</strong>troduce<br />
three categories <strong>of</strong> debris. These debris categories are presumed to reflect activities as do<br />
tool kits and are much less likely to be “curated.” Table <strong>17</strong> presents <strong>the</strong> proportional data<br />
for imperishable artifacts and debris <strong>in</strong> n<strong>in</strong>e components (six floodpla<strong>in</strong> sites) where total<br />
samples exceed about 100 items. These components are easily assigned to periods and phases,<br />
except <strong>in</strong> <strong>the</strong> case <strong>of</strong> 3AS329, Area C, which <strong>in</strong>corporates both Calion and Coon Island<br />
phase assemblages (see Figure 39). Two <strong>of</strong> our Mississippi period components <strong>in</strong>corporate<br />
discrete activity loci (3AS306 and 3AS307), while <strong>the</strong> two relatively prolific loci <strong>in</strong> 3AS286<br />
are kept separate <strong>in</strong> order to exam<strong>in</strong>e <strong>the</strong>ir relationship. The next step <strong>in</strong> exam<strong>in</strong><strong>in</strong>g <strong>the</strong>se<br />
proportional data was performed by cluster analysis us<strong>in</strong>g <strong>the</strong> Numerical Taxonomy System<br />
(NTSYS) with unweighted arithmetic averag<strong>in</strong>g an coefficients <strong>of</strong> Euclidean distance (Rohlf<br />
et al. 1972).<br />
In <strong>the</strong> result<strong>in</strong>g dendrogram, we dist<strong>in</strong>guish two clusters <strong>of</strong> components which are<br />
jo<strong>in</strong>ed at levels <strong>of</strong> distance less than 0.2 (hachured areas <strong>in</strong> Figure 55). These clusters are <strong>of</strong><br />
particular <strong>in</strong>terest <strong>in</strong> <strong>the</strong> light <strong>of</strong> total available archeological evidence from our floodpla<strong>in</strong><br />
sites. Our <strong>in</strong>ferences about <strong>the</strong>se results are as follows:<br />
1. Cluster A components were heavily <strong>in</strong>volved <strong>in</strong> hunt<strong>in</strong>g; debitage and stone<br />
tools predom<strong>in</strong>ate <strong>in</strong> <strong>the</strong>se assemblages; mammal bone and seeds <strong>of</strong> fall fruits<br />
are present.<br />
2. Cluster B components were specially, perhaps exclusively, <strong>in</strong>volved <strong>in</strong> fish<strong>in</strong>g<br />
and fish process<strong>in</strong>g; sherds and fired clay predom<strong>in</strong>ate <strong>in</strong> <strong>the</strong>se Gran <strong>Marais</strong><br />
phase assemblages which are all but identical <strong>in</strong> proportions (Table 16).<br />
3. Three components are dissimilar and <strong>in</strong>termediate to <strong>the</strong>se clusters and dissimilar<br />
among <strong>the</strong>mselves; we cannot say that <strong>the</strong>y are simply hunt<strong>in</strong>g/fish<strong>in</strong>g<br />
components; <strong>the</strong> heterogeneity <strong>in</strong> proportions <strong>of</strong> artifacts and debris may<br />
be taken as a tendency toward characteristics <strong>of</strong> base settlements; <strong>the</strong> earliest<br />
component represented (3AS329, Area C) especially should be tested more<br />
extensively as a possible base settlement (Appendix C).
202 Hemm<strong>in</strong>gs<br />
Figure 55. Dendrogram generated by cluster analysis <strong>of</strong> artifacts and debris <strong>in</strong> n<strong>in</strong>e prehistoric components<br />
(six floodpla<strong>in</strong> sites). Clusters A and B (dist<strong>in</strong>guished by hachur<strong>in</strong>g) represent extractive sites<br />
associated with hunt<strong>in</strong>g and fish<strong>in</strong>g respectively (see text discussion).
Assessment <strong>of</strong> Floodpla<strong>in</strong> Sites 203<br />
Table 16. Proportions <strong>of</strong> Five Functional Categories <strong>of</strong> Imperishable Artifacts and Debris <strong>in</strong><br />
N<strong>in</strong>e Prehistoric Components (six floodpla<strong>in</strong> sites).<br />
Percents <strong>of</strong> Assemblage Total*<br />
Fire-<br />
Provenience by Cultural Stone Cracked Fired Assemblage<br />
Site and Locus Period/Phase Sherds Tools Debitage Rock Clay Total<br />
3AS329-B Baytown-Coles 5.9 2.5 91.0 0.6 0.0 321<br />
Marie Sal<strong>in</strong>e Creek/unnamed<br />
3AS286-A Mississippi/ 6.6 11.2 71.3 4.2 10.4 1188<br />
One Cypress Po<strong>in</strong>t term<strong>in</strong>al<br />
3AS306 Mississippi/ 7.0 2.6 76.7 0.9 12.8 883<br />
Jug Po<strong>in</strong>t 1 Caney Bayou<br />
3AS286-B Mississippi/ 16.8 15.1 79.5 0.7 3.0 298<br />
One Cypress Po<strong>in</strong>t term<strong>in</strong>al<br />
3AS329-C Poverty Po<strong>in</strong>t- 11.4 3.5 26.7 35.8 22.5 595<br />
Marie Sal<strong>in</strong>e Tchula/Calion-<br />
Coon Island<br />
3AS285-B Tchula/Coon 86.5 2.1 10.4 1.0 0.0 96<br />
False Indigo Island<br />
3AS307 Mississippi/ 52.7 4.3 35.5 2.2 5.4 186<br />
Jug Po<strong>in</strong>t 2 Caney Bayou<br />
3AS285-A Mississippi/ 26.3 0.1 1.5 0.1 71.9 5655<br />
False Indigo Gran <strong>Marais</strong><br />
3BR76 Mississippi 29.1 0.2 0.7 0.2 69.7 859<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f Gran <strong>Marais</strong><br />
* Percentages rounded <strong>of</strong>f from five places; order <strong>of</strong> components <strong>in</strong> table based on results <strong>of</strong> cluster analysis<br />
(Figure 55).
204 Hemm<strong>in</strong>gs<br />
Data and results <strong>of</strong> analyses presented <strong>in</strong> this chapter can lead, we th<strong>in</strong>k, to new approaches<br />
and new conclusions regard<strong>in</strong>g early human adaptation <strong>in</strong> <strong>the</strong> Felsenthal region.<br />
Much additional, carefully designed research is needed. One necessary step will be to<br />
reformulate and test rigorously hypo<strong>the</strong>ses about floodpla<strong>in</strong> extractive site functions with<strong>in</strong><br />
particular cultural systems. Without attempt<strong>in</strong>g to characterize <strong>the</strong>se hypo<strong>the</strong>ses <strong>in</strong> every<br />
detail, we suggest <strong>the</strong> direction <strong>the</strong>y can usefully take:<br />
H 1 Transient fall camp <strong>in</strong> <strong>the</strong> overflow bottom used by male hunt<strong>in</strong>g party to<br />
procure deer and o<strong>the</strong>r game (term<strong>in</strong>al phase <strong>of</strong> <strong>the</strong> Mississippi period)<br />
Data Base: One Cypress Po<strong>in</strong>t (3AS286)<br />
H 2 Transient late spr<strong>in</strong>g and summer camp <strong>in</strong> <strong>the</strong> overflow bottom used by specialist<br />
male and female task unit to harvest and process fish on a “bulk” basis<br />
(Gran <strong>Marais</strong> phase <strong>of</strong> <strong>the</strong> Mississippi period)<br />
Data Base: False Indigo (3AS285); Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3 R76)<br />
H 3 Semisedentary late spr<strong>in</strong>g-fall base settlement <strong>in</strong> <strong>the</strong> overflow bottom used<br />
by multiple family unit to exploit riparian resources (Calion phase <strong>of</strong> Poverty<br />
Po<strong>in</strong>t period, Coon Island phase <strong>of</strong> Tchula period)<br />
Data Base: Marie Sal<strong>in</strong>e (3AS329)<br />
F<strong>in</strong>ally, we believe that results <strong>of</strong> test<strong>in</strong>g reported <strong>in</strong> this chapter affirm <strong>the</strong> research<br />
potential and quality <strong>of</strong> significance that may be present <strong>in</strong> small limited activity sites (Talmage<br />
and Chester 1977; Ward 1978). This perspective is absolutely necessary for manag<strong>in</strong>g<br />
resources <strong>in</strong> this <strong>Arkansas</strong> region, as it is elsewhere. Small sites <strong>in</strong>vestigated by us were not<br />
located <strong>in</strong> marg<strong>in</strong>al upland sett<strong>in</strong>gs and, for <strong>the</strong> most part, cannot be attributed to hunterga<strong>the</strong>rer<br />
settlement systems. In fact, our most dramatic results perta<strong>in</strong> to Mississippi period<br />
societies who were part-time agriculturalists.
Chapter 7<br />
SOILS, SEDIMENTS, AND CHRONOLOGy<br />
A limited program <strong>of</strong> soil test<strong>in</strong>g and radiocarbon age determ<strong>in</strong>ation was carried out<br />
<strong>in</strong> conjunction with test<strong>in</strong>g <strong>of</strong> floodpla<strong>in</strong> sites. Results are reported below and are applied to<br />
site transformational hypo<strong>the</strong>ses at <strong>the</strong> end <strong>of</strong> this chapter. Radiocarbon dates have proved<br />
to be an unusual problem, s<strong>in</strong>ce our results are markedly earlier than expected; we discuss<br />
this problem fully near <strong>the</strong> end <strong>of</strong> this chapter. As <strong>the</strong> chapter title implies we deal with<br />
properties <strong>of</strong> soils and sediments, so it is well to dist<strong>in</strong>guish between <strong>the</strong>se terms. Accord<strong>in</strong>g<br />
to Haynes (1978:8)1 a soil is<br />
a modification <strong>of</strong> earth material by physical, chemical, and biological agents...[it<br />
is not] regolith, which as all unconsolidated material over bedrock can support a<br />
soil [and it is not] sediment but an alteration <strong>the</strong>re<strong>of</strong> and o<strong>the</strong>r rock types as well.<br />
We po<strong>in</strong>ted out <strong>in</strong> Chapter 2 that floodpla<strong>in</strong> topstratum deposits <strong>in</strong> <strong>the</strong> Felsenthal<br />
Project area are mapped as Guyton soils which are strongly acid, siliceous, silt loams over<br />
f<strong>in</strong>er textured subsoils. Floodpla<strong>in</strong> topsoil (<strong>the</strong> A horizon or humus) is weakly developed<br />
because flood<strong>in</strong>g removes leaf litter and retards degradation by forest floor organisms, and<br />
because annual <strong>in</strong>crements <strong>of</strong> silt and clay are deposited (overbank primarily). It is generally<br />
true that buried soils or paleosols are not present, or at least are not visible as organically<br />
enriched horizons, <strong>in</strong> riverbank exposures <strong>of</strong> <strong>the</strong> Felsenthal floodpla<strong>in</strong> topstratum.<br />
An exception to <strong>the</strong> latter statement occurs where buried archeological levels are enriched<br />
(anthropogenic soils, <strong>in</strong>clud<strong>in</strong>g middens). Rich middens are recorded <strong>in</strong> only two floodpla<strong>in</strong><br />
sites, False Indigo (3AS285) and Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76); a th<strong>in</strong> anthropogenic soil zone<br />
was noted at Jug Po<strong>in</strong>t 2 (3AS307) <strong>in</strong> Test Pit 1 (Chapter 6). (Test pit pr<strong>of</strong>iles presented <strong>in</strong><br />
Chapter 6 <strong>in</strong>clude Figures 37, 38, 42, and 52.) We report soil test results from <strong>the</strong>se and o<strong>the</strong>r<br />
floodpla<strong>in</strong> sites below.<br />
selected properties <strong>of</strong> floodplA<strong>in</strong> soil sAMples<br />
Table <strong>17</strong> presents <strong>the</strong> soil test data for seven floodpla<strong>in</strong> sites which are described <strong>in</strong><br />
Chapter 6. Analyses were performed by <strong>the</strong> Soil Test<strong>in</strong>g and Research Laboratory, Agronomy<br />
Department, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>. Soil samples <strong>of</strong> about 480 cc volume were collected<br />
<strong>in</strong> standard conta<strong>in</strong>ers furnished by <strong>the</strong> Agronomy Department.
206 Hemm<strong>in</strong>gs<br />
Table <strong>17</strong>. Selected Properties <strong>of</strong> Soil Samples from Seven Floodpla<strong>in</strong> Sites.<br />
Soil Observed Organic<br />
Site Name and Horizontal Depth Below Sample Texture Munsell Soil Re- Organic Phosphorus<br />
Designation Provenience Surface (cm) Number Remarks Color 2 action (pH )3 Carbon (%) mg/g<br />
Marie Sal<strong>in</strong>e Test Pit 1 10 1415-1 vfsl 10yR6/2 5.0 0.55 70<br />
(3AS329) “ 40 1415-2 sic 10yR5/1 4.1 0.45 90<br />
“ 70 1415-3 sicl 10yR6/1 4.9 0.39 60<br />
“ 100 1415-4 c 10yR5/1 4.9 0.27 50<br />
“ 130 1415-5 fsl 10yR5/1 5.7 0.09 25<br />
“ 150 1415-6 cl 10yR5/1 5.8 0.08 10<br />
Test Pit 6 5 1415-7 l 10yR5/2 4.9 1.69 50<br />
“ 8 1415-8 l (humus) 10yR3/1 5.1 5.86 65<br />
“ 25 1415-9 sicl 10yR6/2 4.4 0.59 60<br />
“ 55 1415-10 sil 10yR6/1 4.1 0.41 15<br />
“ 88-94 1415-11 cl (charcoal) 10yR5/1 3.9 0.20 25<br />
“ 110 1415-12 cl 10yR5/1 4.0 0.20 10<br />
False Indigo Test Pit 15 3-6 1433-1 si 10yR5/2 4.7 0.98 1250<br />
(3AS285) “ 11-15 1433-2 midden 10yR2/1 5.2 1.68 100<br />
“ 25-32 1433-3 sic 10yR6/1 5.8 0.64 75<br />
“ 38-42 1433-4 c 10yR6/1 5.9 0.19 30<br />
“ 14-<strong>17</strong> 1433-5 midden 10yR2/1 5.9 1.98 750<br />
“ 29-34 1433-6 midden 10yR2/1 5.9 3.78 875<br />
Shovel Test 1 0-30 1433-7 sic 10yR7/1 5.3 2.38 350<br />
Buttonbush Test Pit 2 0-10 1435-1 sic (under sherd 10yR6/1 4.5 1.44 250<br />
(3BR58) concentration)<br />
One Cypress Test Pit 2 8 1434-1 cl (humus) 10yR5/2 4.5 2.64 165<br />
Po<strong>in</strong>t (3AS286) “ 20 1434-2 c 10yR6/2 4.4 1.10 580<br />
“ 36 1434-3 c 10yR6/2 4.4 0.84 60<br />
Jug Po<strong>in</strong>t Cut- Test Pit 1 10-20 1439-1 midden — 7.0 2.55 1500<br />
<strong>of</strong>f (3BR76) Shovel Test 16 10-20 1439-2 midden — 7.0 2.47 250<br />
Jug Po<strong>in</strong>t 1 Test Pit 1 10 1436-1 cl 10yR4/2 4.3 1.35 160<br />
(3AS306) “ 20 1436-2 sic (occupation 10yR4/2 4.2 0.83 195<br />
level)<br />
“ 30 1436-3 sic 10yR4/2 4.2 0.69 135<br />
Test Pit 2 5 1436-4 sic (humus) 10yR5/1 4.4 1.70 185<br />
“ 12 1436-5 c 10yR5/8 4.4 1.06 215<br />
“ 22 1436-6 sic (feature) 10yR4/2 4.5 0.70 75<br />
“ 30 1436-7 c 10yR5/2 4.1 0.45 90<br />
Jug Po<strong>in</strong>t 2 Test Pit 1 10-20 1437-1 sic (occupation 10yR4/1 4.4 0.69 40<br />
(3AS307) level)<br />
Notes: (1) Texture abbreviations as follows: v=very, f=f<strong>in</strong>e, c=clay, l=loam, si=silt, s=sand; see Table 19 for dimensions <strong>of</strong> particle size<br />
(texture) classes; texture estimated <strong>in</strong> <strong>the</strong> field except for all 12 samples from Marie Sal<strong>in</strong>e (3AS329) which were determ<strong>in</strong>ed <strong>in</strong><br />
<strong>the</strong> laboratory; <strong>the</strong> association <strong>of</strong> soil samples with visible natural or cultural features or zones noted <strong>in</strong> this column; association<br />
with artifacts is not <strong>in</strong>dicated, but only <strong>the</strong> follow<strong>in</strong>g samples represent “sterile” subsoil: 1433-4, 1434-3, 1436-3, 1436-7.<br />
(2) Munsell Soil Color determ<strong>in</strong>ed <strong>in</strong> <strong>the</strong> field from moist pr<strong>of</strong>iles.<br />
(3) Soil reaction scale as follows:
Soils, Sediments, and Chronology 207<br />
We chose to measure (or <strong>in</strong> some cases estimate) soil properties which can <strong>in</strong>dicate<br />
cultural activity when site stratigraphy and soil variability are reasonably well understood<br />
(color, texture, pH, organic carbon, organic phosphorus, and o<strong>the</strong>r characteristics). In this<br />
case test pit pr<strong>of</strong>iles (<strong>the</strong> source <strong>of</strong> most soil samples) have been carefully studied, but test<strong>in</strong>g<br />
was limited on all sites, and data for only 33 samples from seven sites are reported <strong>in</strong><br />
Table <strong>17</strong>.<br />
Our general conclusions from available soil test data are as follows:<br />
1. Samples from Ouachita River and Sal<strong>in</strong>e River floodpla<strong>in</strong> sites, which are not<br />
modified by cultural activity, are probably homogeneous and <strong>in</strong>dist<strong>in</strong>guishable<br />
Guyton soils; however we have not done any <strong>of</strong>fsite soil sampl<strong>in</strong>g to support<br />
this observation.<br />
2. Soil reaction (pH) values are generally <strong>in</strong> <strong>the</strong> extremely acid to strongly acid<br />
range, except where organically enriched midden deposits range from medium<br />
acid to neutral values.<br />
3. High percentages (1-4%) <strong>of</strong> organic carbon among <strong>the</strong>se samples generally<br />
dist<strong>in</strong>guish midden deposits and humic soil zones; wood charcoal is probably<br />
<strong>the</strong> pr<strong>in</strong>cipal constituent <strong>in</strong> midden samples measured by this test (ignition<br />
loss); not all visibly charcoal sta<strong>in</strong>ed features and occupation levels produced<br />
high percentages.<br />
4. High concentrations <strong>of</strong> organic phosphorus (>75 µg/g) are clearly associated<br />
with midden deposits and cultural features <strong>in</strong> Sal<strong>in</strong>e River sites; this concentration<br />
presumably reflects deposition and decay <strong>of</strong> human wastes and/or<br />
animal residues; very high concentrations <strong>of</strong> organic phosphorus (750-1500<br />
µg/g) characterize midden samples from False Indigo (3AS285) and Jug Po<strong>in</strong>t<br />
Cut<strong>of</strong>f (3BR76), which we have <strong>in</strong>terpreted as fish process<strong>in</strong>g camps; this is an<br />
<strong>in</strong>terest<strong>in</strong>g result which requires fur<strong>the</strong>r tests and comparisons.<br />
5. It is disappo<strong>in</strong>t<strong>in</strong>g to note that Marie Sal<strong>in</strong>e (3AS329 , a stratified multicomponent<br />
site on <strong>the</strong> Ouachita River, does not furnish good <strong>in</strong>dications <strong>of</strong> early<br />
cultural levels or features based on soil samples and soil properties tested<br />
(Table 18); <strong>in</strong> fact this site is notably deficient <strong>in</strong> organic matter (although we<br />
have observed and collected charcoal concentrations); <strong>the</strong>re are two medium<br />
acid pH values at 130 and 150 cm <strong>in</strong> Test Pit 1 which could reflect known early<br />
occupation levels, but no high or even moderate carbon or phosphorus values<br />
are recorded for <strong>the</strong>se same samples as would be expected. Marie Sal<strong>in</strong>e<br />
site soils, generally strongly acid, may be deficient <strong>in</strong> organic matter because<br />
cultural occupations were <strong>of</strong> low <strong>in</strong>tensity and pr<strong>in</strong>cipal occupation zones are<br />
greater than 2000 years <strong>of</strong> age.
208 Hemm<strong>in</strong>gs<br />
sediMent textures And floodplA<strong>in</strong> AlluViAtion<br />
Field observations and soil tests suggest to us that local and regional processes <strong>of</strong><br />
floodpla<strong>in</strong> development can be <strong>in</strong>vestigate and elucidated <strong>in</strong> conjunction with archeological<br />
site survey, test<strong>in</strong>g, and excavation. Based on <strong>in</strong>formation presented <strong>in</strong> Chapter 2 and elsewhere,<br />
we propose that channel patterns <strong>in</strong> <strong>the</strong> Felsenthal Project area have rema<strong>in</strong>ed stable<br />
for centuries, that <strong>the</strong> Ouachita River has carried a relatively small load <strong>of</strong> f<strong>in</strong>e suspended<br />
materials, and that overbank sedimentation, ra<strong>the</strong>r than lateral accretion, is <strong>the</strong> dom<strong>in</strong>ant<br />
process <strong>of</strong> floodpla<strong>in</strong> development. These are key concepts for understand<strong>in</strong>g why regional<br />
floodpla<strong>in</strong> sites are buried, why <strong>the</strong>y are not destroyed by lateral erosion, and why “floors,”<br />
or activity loci with arrays <strong>of</strong> artifacts and debris reflect<strong>in</strong>g human behavior many centuries<br />
ago, may persist <strong>in</strong> relatively undisturbed condition.<br />
In <strong>the</strong> case <strong>of</strong> Marie Sal<strong>in</strong>e (3AS329) we had limited opportunity (Test Pits 1-8) to look<br />
at sediments to 1.5 m depth below <strong>the</strong> modern surface <strong>of</strong> <strong>the</strong> Ouachita River levee. We know<br />
this site was occupied or reoccupied sporadically for about 3000 years, possibly more, and<br />
we have briefly suggested that <strong>Marais</strong> Sal<strong>in</strong>e Lake nearby (Figure 3 ) was <strong>the</strong> mouth <strong>of</strong> <strong>the</strong><br />
Sal<strong>in</strong>e River early <strong>in</strong> this <strong>in</strong>terval (Chapter 6). Therefore, do <strong>the</strong> sediments enclos<strong>in</strong>g cultural<br />
rema<strong>in</strong>s at Marie Sal<strong>in</strong>e record processes <strong>of</strong> floodpla<strong>in</strong> development or changes <strong>in</strong> <strong>the</strong>se<br />
processes (and <strong>in</strong> site environment) through time?<br />
The limited attempt we made to <strong>in</strong>vestigate <strong>the</strong>se questions <strong>in</strong>cluded collection <strong>of</strong><br />
sediment samples from carefully studied test pit pr<strong>of</strong>iles and particle size analysis <strong>of</strong> <strong>the</strong>se<br />
samples (among o<strong>the</strong>r tests reported earlier <strong>in</strong> this chapter). These analyses were run at <strong>the</strong><br />
Soil Characterization Laboratory, Department <strong>of</strong> Agronomy, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>, and<br />
data are presented <strong>in</strong> Table 18 and Figure 56. The provenience <strong>of</strong> <strong>the</strong> 12 samples is Test Pits 1<br />
and 6, which are 100 m apart near <strong>the</strong> modern bankl<strong>in</strong>e, and vary 40 cm <strong>in</strong> surface elevation<br />
(Figures 36, 38).<br />
Our general conclusions from data <strong>in</strong> Table 19 are as follows:<br />
1. Samples are generally f<strong>in</strong>e gra<strong>in</strong>ed loams, or even f<strong>in</strong>er silty clay or clay; <strong>the</strong><br />
two coarsest samples are sandy loams, one surficial and modern (1415-1), <strong>the</strong><br />
o<strong>the</strong>r at 130 cm depth (1415-5); all o<strong>the</strong>r samples conta<strong>in</strong> less than 52% sand,<br />
usually f<strong>in</strong>e sand or very f<strong>in</strong>e sand; on <strong>the</strong> whole <strong>the</strong>se sediments and <strong>the</strong>ir<br />
fabrics seen <strong>in</strong> pr<strong>of</strong>ile walls are attributed to overbank accretion; po<strong>in</strong>t bar<br />
deposits (variable <strong>in</strong> particle size range, sort<strong>in</strong>g and fabric) do not appear <strong>in</strong><br />
<strong>the</strong>se test pit pr<strong>of</strong>iles, so far as we can tell.<br />
2. Test Pit 6 samples (all loams) are somewhat more homogeneous <strong>in</strong> texture<br />
than Test Pit 1 samples; also no dist<strong>in</strong>ctive sediment lens or break <strong>in</strong> <strong>the</strong> sequence<br />
<strong>of</strong> sediments can be correlated between test pits (although we correlate<br />
cultural stratigraphy <strong>in</strong> Figure 39).
Table 18. Particle Size Frequencies by Depth <strong>in</strong> Sediment Samples from Test Pits 1 and 6, Area C, Marie Sal<strong>in</strong>e (3AS329).<br />
Sand (mm) Silt (um) Clay<br />
Sample VCS CS MS FS VFS TS CSI MSI FSI TSI TC TEX<br />
Depth (cm) Number 2-1 1-.5 .5-.25 .25-.1 .1-.05 2-.05 50-20 20-5 5-2 50-2
210 Hemm<strong>in</strong>gs<br />
Figure 56. Estimated rate <strong>of</strong> alluviation for <strong>the</strong> natural levees <strong>of</strong> <strong>the</strong> Ouachita and lower Sal<strong>in</strong>e<br />
rivers <strong>in</strong> <strong>the</strong> Felsenthal Project area, based on riverbank sites. The hatchured<br />
zone <strong>in</strong>cludes <strong>the</strong> best available, but slightly scattered, data po<strong>in</strong>ts, which generally<br />
describe a curve <strong>of</strong> decreas<strong>in</strong>g slope. Local conditions <strong>of</strong> alluviation may<br />
cause some scatter<strong>in</strong>g <strong>of</strong> data, but <strong>the</strong> trend is apparent.
Soils, Sediments, and Chronology 211<br />
3. There may be a slight trend for coarser gra<strong>in</strong> size upward <strong>in</strong> <strong>the</strong>se sample columns,<br />
but it is not clear or consistent; this trend, if real, could be due to chang<strong>in</strong>g<br />
fluvial processes, but it also could be due to degradation <strong>of</strong> older, more<br />
deeply buried particles; surficial samples (
212 Hemm<strong>in</strong>gs<br />
3. Most importantly, <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> and natural levee environment <strong>of</strong> <strong>the</strong><br />
Felsenthal Project area, depths are consistent with attributed ages <strong>of</strong> cultural<br />
assemblages, and depth observed or measured should predict <strong>the</strong> age <strong>of</strong> an<br />
newly discovered site or site level <strong>in</strong> this environmental context.<br />
We reach similar conclusions with regard to channel stability and floodpla<strong>in</strong> development<br />
<strong>in</strong> <strong>the</strong> Felsenthal Project area (especially for <strong>the</strong> Marie Sal<strong>in</strong>e site) as those stated by<br />
Ritter et al. (1973:375) for <strong>the</strong> Faucett site on <strong>the</strong> Delaware River.<br />
The Faucett site sequence is devoid <strong>of</strong> po<strong>in</strong>t-bar sediment...Evidence <strong>of</strong> abundant<br />
scour and fill, normal <strong>in</strong> po<strong>in</strong>t-bar deposits, has not been observed....It becomes<br />
clear that <strong>the</strong> Delaware River has ma<strong>in</strong>ta<strong>in</strong>ed its present valley-bottom position<br />
for an extremely long time [o<strong>the</strong>rwise] older sediments would have been disturbed<br />
and <strong>the</strong> earlier records <strong>of</strong> human occupation rendered un<strong>in</strong>telligible....The<br />
Faucett sequence demonstrates that a natural channel <strong>in</strong> unconsolidated material<br />
can be fixed for very long periods <strong>of</strong> time and that under those conditions overbank<br />
accretion becomes <strong>the</strong> dom<strong>in</strong>ant process <strong>in</strong>volved <strong>in</strong> <strong>the</strong> development <strong>of</strong><br />
<strong>the</strong> floodpla<strong>in</strong>.<br />
rAdiocArbon dAtes for floodplA<strong>in</strong> sites<br />
As <strong>in</strong>dicated earlier <strong>in</strong> this chapter, <strong>the</strong> radiocarbon date obta<strong>in</strong>ed from floodpla<strong>in</strong> sites<br />
gave ages much greater than expected, and <strong>the</strong> results are <strong>in</strong> fact unacceptable to us <strong>in</strong> light<br />
<strong>of</strong> present knowledge <strong>of</strong> regional prehistory. We suspect that contam<strong>in</strong>ation <strong>of</strong> samples is<br />
<strong>in</strong>volved, but <strong>the</strong> ord<strong>in</strong>ary outcome <strong>of</strong> contam<strong>in</strong>ation is to produce younger than expected<br />
ages and erratic or <strong>in</strong>consistent results where samples were obta<strong>in</strong>ed <strong>in</strong> a stratified context.<br />
The dates reported here are both too old and <strong>in</strong>consistent with stratigraphy. We have attempted<br />
to exam<strong>in</strong>e <strong>the</strong> various possibilities for contam<strong>in</strong>ation with older carbon, and discussed<br />
<strong>the</strong>se possibilities with <strong>the</strong> laboratory <strong>in</strong>volved; our queries and suggestions to <strong>the</strong><br />
laboratory scientists and <strong>the</strong>ir replies are reproduced later <strong>in</strong> this section.<br />
Table 19 characterizes <strong>the</strong> five samples assayed and presents both <strong>the</strong> laboratory determ<strong>in</strong>ations<br />
and our estimates <strong>of</strong> accurate sample age. (The radiocarbon laboratory which<br />
furnished <strong>the</strong> results was Beta Analytic, Inc., Coral Gables, Florida.) S<strong>in</strong>ce charcoal and o<strong>the</strong>r<br />
organic residues are exceed<strong>in</strong>gly sparse or rare <strong>in</strong> occurrence throughout <strong>the</strong> project area<br />
and <strong>in</strong> nearly all floodpla<strong>in</strong> sites exam<strong>in</strong>ed or tested, we submitted a small series <strong>of</strong> samples,<br />
each sample also <strong>of</strong> relatively small size. All samples were exam<strong>in</strong>ed by us under a microscope,<br />
concentrated or “picked” for extraneous matter <strong>in</strong> prelim<strong>in</strong>ary fashion, and identified<br />
as wood charcoal <strong>of</strong> one or ano<strong>the</strong>r taxon or group (see Appendix B. Those samples (1415
Table 19. Characteristics and Results <strong>of</strong> Age Determ<strong>in</strong>ation for Radiocarbon Samples from Marie Sal<strong>in</strong>e (3AS329) and Jug Po<strong>in</strong>t 2<br />
(3AS307).<br />
Age Determ<strong>in</strong>ation<br />
Site Name Project Sample Sample Laboratory (C-14 years<br />
and Designation Number Provenience Description Number B.P. uncorrected) Expected Age*<br />
Marie Sal<strong>in</strong>e 1415-5-4 Collected from riverbank adjacent Wood charcoal <strong>in</strong> 8 x 25 cm lens; Beta-1880 2945 ± 100 Early Tchula period,<br />
(3AS329) to Test Pit 4, Area C, at 1.0 m depth no direct association <strong>of</strong> artifacts (or 995 B.C.) ca. 500 B.C.±<br />
(19.4 m elevation)<br />
Marie Sal<strong>in</strong>e 1415-11-7 Test Pit 6, Area C, at 0.8 m depth Wood charcoal fragment or chunk Beta-1882 5390 ± <strong>17</strong>0 Early Tchula period,<br />
(3AS329) (19.4 m elevation) associated with novaculite flakes (or 3440 B.C.) ca. 500 B.C.±<br />
Marie Sal<strong>in</strong>e 1415-3-1 Test Pit 3, Area C, at 0.7 m depth Wood charcoal <strong>in</strong> small concentra- Beta-1879 2650 ± 215 Early Baytown-Coles<br />
(3AS329) (19.5 m elevation) tion; no association <strong>of</strong> artifacts (or 700 B.C.) Creek period, ca.<br />
A.D. 300±<br />
Soils, Sediments, and Chronology 213<br />
Marie Sal<strong>in</strong>e 1415-7-6 Test pit 7, Area C, 0.4-0.8 m depth Scattered wood charcoal associ- Beta-1881 3410 ± 200 Baytown-Coles Creek<br />
(3AS329) (19.4-19.8 m elevation) ated with novaculite flakes (or 1460 B.C.) period, ca. A.D. 500±<br />
Jug Po<strong>in</strong>t 2 1437-7-2 Test Pit 1, Area C, 0.2 m depth Wood charcoal scattered <strong>in</strong> th<strong>in</strong> Beta-1883 1540 ± 85 Mississippi period,<br />
3AS307) (20.6 m elevation) occupation level; associated with (or A.D. 410) early Caney Bayou<br />
Bassett po<strong>in</strong>t and both clay- and phase, ca. A.D. 1400±<br />
shell-tempered sherds<br />
* Expected age based on natural and cultural stratigraphy with<strong>in</strong> sites and on external correlation (Chapter 3)
214 Hemm<strong>in</strong>gs<br />
series) selected from stratified components at Marie Sal<strong>in</strong>e (3AS329) were essentially <strong>the</strong><br />
only adequate ones available here <strong>in</strong> terms <strong>of</strong> sample size and association; this site is sparse<br />
<strong>in</strong> artifact content and features and deficient <strong>in</strong> organic matter, based on results <strong>of</strong> test<strong>in</strong>g,<br />
but is potentially one <strong>of</strong> <strong>the</strong> most important multicomponent sites <strong>in</strong> <strong>the</strong> region (Chapter 6,<br />
Appendix C). We <strong>in</strong>tended <strong>the</strong> four samples from Marie Sal<strong>in</strong>e to verify stratigraphy observed<br />
and to date <strong>the</strong> major ceramic components represented here. These objectives failed,<br />
as <strong>the</strong> age determ<strong>in</strong>ations appear to run from 500 to 3,000 years too old.<br />
The sample selected from Jug Po<strong>in</strong>t 2 (3AS307) was well associated with a presumed<br />
early Caney Bayou phase occupation level and artifact assemblage. We exam<strong>in</strong>ed and concentrated<br />
this sample <strong>in</strong> <strong>the</strong> manner <strong>in</strong>dicated above, and it appeared to be quite adequate<br />
<strong>in</strong> size. Sample 1437-7-2 was <strong>in</strong>tended to provide <strong>the</strong> first date for this late Mississippi<br />
period phase <strong>in</strong> <strong>the</strong> region. In this case also, <strong>the</strong> age determ<strong>in</strong>ation appears to fall 1,000 years<br />
too early (Table 20).<br />
In <strong>the</strong> follow<strong>in</strong>g section we quote correspondence (E. T. Hemm<strong>in</strong>gs letter to Murray<br />
Tamers, December 3, 1980) and paraphrase a telephone conversation (Jerry Stipp to Hemm<strong>in</strong>gs,<br />
December 15, 1980) regard<strong>in</strong>g <strong>the</strong> problematical radiocarbon results from <strong>the</strong>se two<br />
floodpla<strong>in</strong> sites.<br />
1. All five samples are from buried levels <strong>in</strong> an overflow bottomland where<br />
deep, prolonged <strong>in</strong>undation has occurred annually for a long time; <strong>the</strong> soils<br />
are siliceous silty clay loams, gleyed or reduced, extremely acid (pH—4.0-4.5);<br />
is some unusual soil chemical process occurr<strong>in</strong>g that could <strong>in</strong>troduce old carbon?<br />
Laboratory reply: We have no specific knowledge <strong>of</strong> such processes; <strong>in</strong> one<br />
case two water-logged stumps at <strong>the</strong> same level <strong>in</strong> Mississippi River alluvium<br />
produced <strong>in</strong>consistent dates, and <strong>the</strong>se dates also could not be replicated.<br />
2. Oil and gas fields are mapped all around this bottomland area and upstream;<br />
does methane gas (or any o<strong>the</strong>r dead carbon compound) move vertically or<br />
laterally <strong>in</strong> recent alluvium, river water, or ground water? No liquid or solid<br />
petroleum residues have been noted <strong>in</strong> our work <strong>in</strong> floodpla<strong>in</strong> sediments.<br />
Laboratory reply: Aga<strong>in</strong>, we cannot specifically evaluate such processes; we<br />
do know that methane or marsh gas can be produced <strong>in</strong> wetlands and swamps<br />
by decomposition <strong>of</strong> vegetal matter. We also know <strong>of</strong> one site on <strong>the</strong> Gulf <strong>of</strong><br />
Mexico shorel<strong>in</strong>e where pitch residues produced <strong>in</strong>consistent results; <strong>in</strong> this<br />
case special extraction <strong>of</strong> <strong>the</strong>se residues was possible.<br />
3. Lignite outcrops <strong>in</strong> <strong>the</strong> riverbed some miles upstream, but we believe that<br />
all carbon materials exam<strong>in</strong>ed <strong>in</strong> our excavations (and under <strong>the</strong> microscope<br />
later) are just wood charcoal <strong>of</strong> cultural orig<strong>in</strong>. Redeposited lignite has not
Soils, Sediments, and Chronology 215<br />
been observed <strong>in</strong> our work <strong>in</strong> this floodpla<strong>in</strong>, but I suppose f<strong>in</strong>ely divided<br />
material could be present?<br />
Laboratory reply: Lignite particles could occur <strong>in</strong> alluvium and could resemble<br />
more recent wood charcoal strongly; also, hypo<strong>the</strong>tically, <strong>the</strong>re could<br />
be soluble carbonaceous extracts <strong>of</strong> lignite as a contam<strong>in</strong>ant <strong>in</strong> archeological<br />
samples.<br />
4. All five samples were collected and stored <strong>in</strong> amber colored plastic vials for<br />
a year. These vials have not been used for C-14 samples by me previously.<br />
Could <strong>the</strong> vials be degass<strong>in</strong>g some antique carbon?<br />
Laboratory reply: The few molecules <strong>of</strong> carbon released <strong>in</strong> such a process<br />
would be an <strong>in</strong>significant source <strong>of</strong> contam<strong>in</strong>ation (alum<strong>in</strong>um foil, used widely<br />
to collect and store samples has a th<strong>in</strong> oily film which is evidently <strong>in</strong>significant<br />
as a contam<strong>in</strong>ant).<br />
5. Redeposition <strong>of</strong> older charcoal or reuse <strong>of</strong> older wool is not likely to be a factor,<br />
I th<strong>in</strong>k. The local floodpla<strong>in</strong> sediments are devoid <strong>of</strong> wood and carbonized<br />
wood, except possibly below <strong>the</strong> permanent water table and river stage.<br />
Laboratory reply: Never<strong>the</strong>less, transportation and redeposition <strong>of</strong> older<br />
charcoal or wood can occur <strong>in</strong> floodpla<strong>in</strong>s; because <strong>of</strong> this possibility scattered<br />
particles or fragments are poor sampl<strong>in</strong>g material compared with obvious<br />
concentrations or features <strong>in</strong> place.<br />
6. Laboratory error result<strong>in</strong>g <strong>in</strong> older dates? I have no idea what to suspect here.<br />
Can you discuss this possibility <strong>in</strong> relation to <strong>the</strong>se samples specifically? The<br />
actual count<strong>in</strong>g times for each sample should be <strong>in</strong>cluded <strong>in</strong> my report.<br />
Laboratory reply: The samples were given our usual acid-alkali-acid pretreatment.<br />
They were all small. We obta<strong>in</strong>ed less than one gram benzene for four<br />
samples, while Beta-1883 was borderl<strong>in</strong>e (0.9676 g benzene). Count<strong>in</strong>g times<br />
were extended (2500-2900 m<strong>in</strong>utes for <strong>the</strong> smallest samples) and 2400 m<strong>in</strong>utes<br />
for Beta-1883. We have no reason to suspect laboratory error produced older<br />
dates.<br />
7. The ages reported are accurate and <strong>the</strong> archeological data are <strong>in</strong>correctly <strong>in</strong>terpreted?<br />
As I <strong>in</strong>dicated...this is not likely to be <strong>the</strong> case. We have abundant correlative<br />
age data from <strong>the</strong> Lower Mississippi Valley area and also some from<br />
this region that simply make <strong>the</strong>se dates too early.<br />
Laboratory reply: You, <strong>of</strong> course, are <strong>in</strong> <strong>the</strong> position to make this judgment;
216 Hemm<strong>in</strong>gs<br />
<strong>the</strong> laboratory always respects <strong>the</strong> <strong>in</strong>formation obta<strong>in</strong>ed from field context and<br />
regional study by <strong>the</strong> archeologist or geologist.<br />
We have <strong>in</strong>troduced <strong>the</strong>se detailed comments to show that special problems <strong>of</strong> contam<strong>in</strong>ation<br />
may exist <strong>in</strong> <strong>the</strong> Felsenthal Project area. A well designed problem oriented, chronometric<br />
study should be <strong>in</strong>tegrated with future work <strong>in</strong> floodpla<strong>in</strong> sites (Appendix C).<br />
The failure <strong>of</strong> our series <strong>of</strong> dates means that we lack f<strong>in</strong>e scaled chronometric data for<br />
some <strong>of</strong> <strong>the</strong> most important project research results, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> follow<strong>in</strong>g:<br />
1. chronometry <strong>of</strong> multiple components at Marie Sal<strong>in</strong>e (3AS329) (Chapter 6)<br />
2. chronometrically established rates <strong>of</strong> alluviation (Figure 56)<br />
3. chronometric data for recent floodpla<strong>in</strong> and river<strong>in</strong>e features (Chapter 2)<br />
4. chronometrically established arrow po<strong>in</strong>t, dart po<strong>in</strong>t, and ceramic sequences<br />
(Chapter 8)<br />
5. <strong>in</strong>dependent age data for seriation <strong>of</strong> ceramic temper types and differential<br />
use <strong>of</strong> chert/novaculite raw materials (Chapter 8).<br />
However, we have attempted to make maximum use <strong>of</strong> stratigraphy, seriation, and<br />
cross-dat<strong>in</strong>g (regional and extraregional correlation) to provide temporal order<strong>in</strong>g for sites,<br />
artifacts, and assemblages <strong>in</strong> <strong>the</strong> Felsenthal Project area, as <strong>in</strong>dicated at various po<strong>in</strong>ts <strong>in</strong> this<br />
report. This order<strong>in</strong>g will undoubtedly be tested and ref<strong>in</strong>ed by future work.<br />
site trAnsforMAtionAl hypo<strong>the</strong>ses<br />
Data presented <strong>in</strong> this chapter and earlier <strong>in</strong> this report (Chapters 5, 6) are relevant to<br />
<strong>the</strong> last problem doma<strong>in</strong> stated <strong>in</strong> our research design, site transformations. The transformations<br />
<strong>of</strong> concern here are certa<strong>in</strong> “postdepositional changes <strong>in</strong> site and artifact morphology<br />
caused by noncultural processes” (Schiffer 1976:15). In <strong>the</strong> context <strong>of</strong> floodpla<strong>in</strong> sites <strong>in</strong> our<br />
project area, we predicted <strong>the</strong> outcome <strong>of</strong> alluviation on site survey results (Hypo<strong>the</strong>sis 7)<br />
and <strong>of</strong> soil chemical processes on site content (Hypo<strong>the</strong>sis 8). These hypo<strong>the</strong>ses were stated<br />
<strong>in</strong> full as follows:<br />
Problem Doma<strong>in</strong> 4: Site transformations<br />
H 7 If geologically Recent topstratum deposits have aggraded 20 feet or more <strong>in</strong><br />
<strong>the</strong> Ouachita Valley, methods <strong>of</strong> <strong>in</strong>tensive survey <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> will fail to<br />
sample adequately early occupation sites.<br />
Test Implications: Recognition <strong>of</strong> age <strong>of</strong> sediments drowned and exposed; discovery<br />
<strong>of</strong> buried sites <strong>in</strong> datable context.<br />
Data Requirements: Significant disparity <strong>in</strong> observed and expected occurrence <strong>of</strong><br />
early sites; geochronological controls.
Soils, Sediments, and Chronology 2<strong>17</strong><br />
H 8 If differential preservation <strong>of</strong> organic matter occurs with<strong>in</strong> or between sites,<br />
soil acidity has adversely affected <strong>the</strong> archeological context.<br />
Test Implications: Recognition <strong>of</strong> preservation gradients by sampl<strong>in</strong>g soil and refuse;<br />
demonstrable loss <strong>of</strong> organic matter.<br />
Date Requirements: Translation <strong>of</strong> preservation gradient to a numerical scale; significant<br />
positive correlation with pH (low soil pH—low degree <strong>of</strong> preservation).<br />
With regard to Hypo<strong>the</strong>sis 7, we have compiled considerable data on <strong>the</strong> occurrence <strong>of</strong><br />
buried prehistoric sites and identification <strong>of</strong> components represented (Chapter 5, Appendix<br />
A). In this chapter we discussed <strong>the</strong> consistent relationship between depth <strong>of</strong> components<br />
<strong>in</strong> natural levee deposits and ages assigned by virtue <strong>of</strong> diagnostic artifacts or assemblages<br />
(i.e., concordance between stratigraphy and cross-dat<strong>in</strong>g). This concordance is limited by<br />
available site data to <strong>the</strong> upper 1.5-2.0 m <strong>of</strong> topstratum deposits and <strong>the</strong> last 2000-3000 years.<br />
Site Units 1 and 4 (described <strong>in</strong> Chapter 5) comprise 27.4 and 18.8 km reaches <strong>of</strong> <strong>the</strong><br />
Ouachita River adjo<strong>in</strong><strong>in</strong>g at river mile 254. Site Unit 1 was longer, had substantially higher<br />
banks, and had deeper exposure <strong>of</strong> topstratum deposits (because <strong>of</strong> <strong>the</strong> exist<strong>in</strong>g navigation<br />
pool) than Site Unit 4 downstream. The follow<strong>in</strong>g components were recorded <strong>in</strong> each (Chapter<br />
5):<br />
Site Unit 1<br />
Mississippi 0<br />
Baytown-Coles Creek 3<br />
Tchula/Coon Island 1<br />
Poverty Po<strong>in</strong>t/Calion 0<br />
Archaic 8<br />
Unknown 6<br />
18<br />
Site Unit 4<br />
Mississippi 5<br />
Baytown-Coles Creek 9<br />
Tchula/Coon Island 2<br />
Poverty Po<strong>in</strong>t/Calion 1<br />
Late Archaic 1<br />
Unknown 5<br />
23<br />
This compilation suggests that preceramic components (before Tchula period or about 500<br />
B.C.) are relatively more abundant <strong>in</strong> Site Unit 1 even though total components are fewer.<br />
However, unidentified components <strong>in</strong> both site units obscure this comparison. One <strong>of</strong> <strong>the</strong><br />
Archaic components <strong>in</strong> Site Unit 1 is 3UN162 (Figure 25), <strong>the</strong> deepest artifact occurrence<br />
recorded (fire-cracked rock at 2.95 m depth). This site po<strong>in</strong>ts up a weakness <strong>in</strong> our attempt
218 Hemm<strong>in</strong>gs<br />
to test Hypo<strong>the</strong>sis 7, because we identified such deeply buried components without ceramics<br />
as Archaic, even where diagnostic dart po<strong>in</strong>ts were absent. On <strong>the</strong> whole, however, we<br />
are confident that older, deeper sites are exposed <strong>in</strong> Site Unit 1, and that <strong>the</strong>ir counterparts<br />
may be present <strong>in</strong> Site 4, largely below river level and <strong>the</strong>refore unrecorded.<br />
It is also evident from site survey data, that no Paleo-Indian and Early or Middle Archaic<br />
components (i.e., sites dat<strong>in</strong>g before 4000 B.C.) were recorded <strong>in</strong> <strong>the</strong> Felsenthal floodpla<strong>in</strong>,<br />
although such components are known <strong>in</strong> adjacent upland sites. We can suggest four<br />
possible reasons for this hiatus:<br />
1. Such sites are so few <strong>the</strong>y were missed by site survey teams.<br />
2. No human occupation or use <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> occurred before 4000 B.C.<br />
3. Fluvial processes have removed such sites, or<br />
4. Sites occupied before 4000 B.C. are deeply buried and drowned by exist<strong>in</strong>g<br />
river levels.<br />
We are <strong>in</strong>cl<strong>in</strong>ed to reject <strong>the</strong> second statement and admit that all o<strong>the</strong>rs may be causative<br />
factors. The last two statements describe natural transformations <strong>of</strong> <strong>the</strong> archeological record,<br />
<strong>in</strong>clud<strong>in</strong>g <strong>the</strong> agent <strong>of</strong> ongo<strong>in</strong>g alluviation and site burial proposed <strong>in</strong> our hypo<strong>the</strong>sis. Although<br />
many <strong>of</strong> <strong>the</strong> data cited above support this hypo<strong>the</strong>sis, complex problems <strong>of</strong> <strong>in</strong>terpretation<br />
rema<strong>in</strong> to be solved (and early floodpla<strong>in</strong> sites rema<strong>in</strong> to be discovered). Among <strong>the</strong>se<br />
problems are <strong>the</strong> 1ack <strong>of</strong> reliable chronometric data, although we believe that geochronological<br />
<strong>in</strong>terpretation <strong>of</strong> floodpla<strong>in</strong> development and site burial processes has been advanced<br />
by our work.<br />
Hypo<strong>the</strong>sis 8 predicts that loss <strong>of</strong> organic materials (faunal and floral rema<strong>in</strong>s) <strong>in</strong><br />
floodpla<strong>in</strong> sites will <strong>in</strong>crease with soil acidity. Our limited soil test data <strong>in</strong>dicate that extreme<br />
acidity recurs <strong>in</strong> soil samples from floodpla<strong>in</strong> sites and that organic residues are absent or<br />
rare <strong>in</strong> such sites except where very local anthropogenic enrichment (middens, pit fills, occupation<br />
“floors”) has occurred (Table 18). Ideally, this hypo<strong>the</strong>sis should be tested by means<br />
<strong>of</strong> <strong>in</strong>tensive soil and faunal/floral sampl<strong>in</strong>g with<strong>in</strong> a s<strong>in</strong>gle component, ra<strong>the</strong>r than by few<br />
samples from many sites. Our observations from local soil test data might be summarized as<br />
follows:<br />
pH 3.9-4.3 generally no visible organic rema<strong>in</strong>s<br />
pH 4.4-5.2 traces <strong>of</strong> charred faunal or floral matter may be present<br />
pH 5.3-7.0 visibly enriched by organic material, usually charred faunal and<br />
floral matter, <strong>in</strong>clud<strong>in</strong>g “black dirt midden” deposits<br />
We cannot deal more adequately with Hypo<strong>the</strong>sis 8, but suggest that carefully designed soil<br />
test<strong>in</strong>g dur<strong>in</strong>g mitigation work (Appendix C) can elicit a better understand<strong>in</strong>g <strong>of</strong> floodpla<strong>in</strong><br />
site transformation, <strong>in</strong>clud<strong>in</strong>g anthropogenic soil processes.
chapter 8<br />
lithic And cerAMic AnAlyses<br />
In this chapter we summarize and report laboratory studies <strong>of</strong> about 9,400 lithic and<br />
3,700 ceramic specimens recovered dur<strong>in</strong>g site survey and test excavation work. Analyses <strong>of</strong><br />
<strong>the</strong>se materials were performed by <strong>the</strong> author, project staff members, and various personnel<br />
<strong>of</strong> <strong>the</strong> Contract Laboratory, <strong>Arkansas</strong> Archeological Survey, dur<strong>in</strong>g 1980 (see Acknowledgments).<br />
Clearly, we have not been able to carry <strong>the</strong>se analyses as far as we would have liked<br />
because <strong>of</strong> time constra<strong>in</strong>ts. For example, <strong>the</strong> b<strong>in</strong>ocular microscope was our basic tool <strong>of</strong><br />
lithic and ceramic study, but we have not attempted high magnification use-wear analysis<br />
<strong>of</strong> flaked stone tools nor petrographic study <strong>of</strong> sherd composition and fabric. Many useful<br />
directions for future work are suggested by our <strong>in</strong>itial studies <strong>of</strong> floodpla<strong>in</strong> site collections.<br />
Information presented <strong>in</strong> this chapter should lead to a better understand<strong>in</strong>g <strong>of</strong> <strong>the</strong><br />
k<strong>in</strong>ds <strong>of</strong> tools and debris (preserved and recoverable) which occur <strong>in</strong> floodpla<strong>in</strong> extractive<br />
sites, <strong>the</strong>ir implications for extractive site behavior, sources <strong>of</strong> raw material and artifact<br />
forms <strong>in</strong>digenous or alien to <strong>the</strong> region, and o<strong>the</strong>r functional or technological considerations.<br />
Functional analysis was <strong>the</strong> primary emphasis <strong>of</strong> laboratory study, and results obta<strong>in</strong>ed<br />
were <strong>in</strong>tegrated with site data and analyses presented elsewhere <strong>in</strong> this report, when<br />
possible, to explore <strong>the</strong> archeological dimensions <strong>of</strong> extractive sites.<br />
We have also attempted typological and related analyses <strong>of</strong> projectile po<strong>in</strong>ts and<br />
decorated ceramics as a basis for understand<strong>in</strong>g <strong>in</strong>tersite and <strong>in</strong>terregional <strong>in</strong>teractions and<br />
changes <strong>in</strong> <strong>the</strong>se <strong>in</strong>teractions through time. In fact, some fundamental data perta<strong>in</strong><strong>in</strong>g to<br />
technological and stylistic change <strong>in</strong> floodpla<strong>in</strong> site assemblages have emerged, and must<br />
now be tested <strong>in</strong> larger upland sites <strong>of</strong> <strong>the</strong> region. We know, for example, <strong>the</strong> proportions <strong>of</strong><br />
chert and novaculite debitage and <strong>of</strong> clay-tempered and shell-tempered ceramics to expect<br />
<strong>in</strong> floodpla<strong>in</strong> sites <strong>of</strong> various periods. At <strong>the</strong> end <strong>of</strong> this chapter two hypo<strong>the</strong>ses perta<strong>in</strong><strong>in</strong>g<br />
to regional culture history are briefly exam<strong>in</strong>ed and tested through artifact data. For <strong>the</strong><br />
most part, such hypo<strong>the</strong>sis test<strong>in</strong>g will require much more systematic data from sites <strong>in</strong> both<br />
upland and lowland sett<strong>in</strong>gs.<br />
lithic source study<br />
From <strong>the</strong> outset <strong>of</strong> site survey work and through later stages <strong>of</strong> test excavation and<br />
field laboratory operation, we could see that fl<strong>in</strong>t-knapp<strong>in</strong>g activity on prehistoric flood-
220 Hemm<strong>in</strong>gs<br />
pla<strong>in</strong> sites featured reduction <strong>of</strong> rounded pebbles and cobbles. Among <strong>the</strong> larger collections<br />
<strong>of</strong> debitage from Mississippi period sites, discussed later <strong>in</strong> this chapter, brown pebble<br />
chert is especially prevalent <strong>in</strong> <strong>the</strong> form <strong>of</strong> cortical flakes and occasionally also pebble cores.<br />
Among collections <strong>of</strong> debitage from Archaic and early ceramic sites, gray or white novaculite<br />
flakes, less obviously attributable to pebble orig<strong>in</strong>, are prevalent. As <strong>the</strong>se observations<br />
accumulated, we designed and carried out a small-scale study <strong>of</strong> potential local resources<br />
for stone toolmak<strong>in</strong>g. This study <strong>in</strong>volved <strong>the</strong> <strong>in</strong>vestigation <strong>of</strong> graveliferous deposits at two<br />
localities and comparison <strong>of</strong> results with archeological site data.<br />
In Chapter 2 we emphasized <strong>the</strong> f<strong>in</strong>e gra<strong>in</strong>ed texture <strong>of</strong> topstratum deposits exposed<br />
<strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>s <strong>of</strong> Ouachita and Sal<strong>in</strong>e rivers. These floodpla<strong>in</strong>s are virtually “stoneless,”<br />
without exposed gravel lenses and with small scattered pebbles exceed<strong>in</strong>gly rare. However,<br />
we know from n<strong>in</strong>eteenth century hydrographic surveys <strong>of</strong> Ouachita River that gravel<br />
bars were exposed at low water <strong>in</strong> predictable locations, especially at <strong>the</strong> crossover po<strong>in</strong>t<br />
between bends. These riffles, <strong>of</strong>ten colorfully named, were among <strong>the</strong> serious obstructions<br />
and hazards to low water navigation, but were elim<strong>in</strong>ated <strong>in</strong> this century by lock and dam<br />
construction and by dredg<strong>in</strong>g. Such channel bars represent <strong>the</strong> coarse bedload <strong>of</strong> <strong>the</strong> river,<br />
material “stored” for long periods or transported downvalley at a slow rate. Channel bars<br />
are proposed here as one important source <strong>of</strong> lithic raw materials for aborig<strong>in</strong>al <strong>in</strong>habitants<br />
<strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> <strong>in</strong> this region. Also, <strong>the</strong>re may well be a close association between locations<br />
<strong>of</strong> certa<strong>in</strong> k<strong>in</strong>ds <strong>of</strong> prehistoric sites and past locations <strong>of</strong> gravel bars. Not only toolmak<strong>in</strong>g<br />
materials, but also rocks for stone boil<strong>in</strong>g and earth oven cookery could be easily obta<strong>in</strong>ed<br />
on gravel bars.<br />
A second potential source <strong>of</strong> pebbles or cobbles comprises <strong>the</strong> series <strong>of</strong> Pleistocene<br />
terraces adjo<strong>in</strong><strong>in</strong>g <strong>the</strong> project area (Chapter 2), especially where substratum materials are<br />
exposed or concentrated by erosion. In general, older, more wea<strong>the</strong>red graveliferous deposits<br />
are exposed at <strong>in</strong>creas<strong>in</strong>g distances and elevations from <strong>the</strong> modern river channel. One<br />
may suggest that terrace pebbles or cobbles collected and <strong>in</strong>troduced <strong>in</strong>to <strong>the</strong> floodpla<strong>in</strong> for<br />
aborig<strong>in</strong>al use would be obta<strong>in</strong>ed from a proximal, accessible source. However, <strong>the</strong> <strong>in</strong>tricate<br />
series <strong>of</strong> floodpla<strong>in</strong> dra<strong>in</strong>ages makes many terrace edges and many upland areas accessible<br />
by water travel, and it is likely that numerous expedient sources were exploited <strong>in</strong> this manner.<br />
Our Ouachita River Valley localities were selected as study and sampl<strong>in</strong>g stations<br />
pr<strong>in</strong>cipally by virtue <strong>of</strong> accessibility to <strong>the</strong> Felsenthal Project area. We were unsuccessful <strong>in</strong><br />
locat<strong>in</strong>g a gravel deposit <strong>in</strong> or near <strong>the</strong> lower Sal<strong>in</strong>e River floodpla<strong>in</strong>, where many prehistoric<br />
sites have been recorded, and where some sites have been tested and analyzed <strong>in</strong> detail<br />
(Chapters 5, 6). The place-names at Gravel Ridge, which we have not visited, and Goulett<br />
Island (galet = pebble, gravel, or gravel bank <strong>in</strong> Mississippi Valley French) where we have
Lithic and Ceramic Analyses 221<br />
done m<strong>in</strong>imal survey work, suggest possible lithic sampl<strong>in</strong>g stations for future work <strong>in</strong> <strong>the</strong><br />
lower Sal<strong>in</strong>e River area.<br />
Station 1, Ouachita River Bedload Material. This locality consists <strong>of</strong> a pipel<strong>in</strong>e cross<strong>in</strong>g<br />
at river mile 262.6 where dredged bedload material has been deposited on <strong>the</strong> left bank<br />
(SW 1/4, Sec. 30, T<strong>17</strong>S, R11W). The hydrographic survey <strong>of</strong> 1873 (Ouachita River Survey,<br />
Sheet No. 11) shows “Cany Mary Shoals” upstream and unnamed shoals, gravel shoals, and<br />
rapids extend<strong>in</strong>g from this location downstream to Five Mile Bend. The dredged spoil exam<strong>in</strong>ed<br />
by us <strong>in</strong>cluded chiefly novaculite pebbles or cobbles and a few sandstone boulders;<br />
quartz, quartzite, shale, and silicified wood materials were also present, but little or no chert.<br />
Station 2, Pleistocene Terrace Alluvium. This locality consists <strong>of</strong> a gravel pit at <strong>the</strong> <strong>in</strong>ner<br />
edge <strong>of</strong> <strong>the</strong> Prairie Terrace (Fleetwood 1969) about 10 km east <strong>of</strong> <strong>the</strong> present Ouachita<br />
River channel (NW 1/4, Sec. 34, T18S, R9W). The pit walls expose a deep red, sandy clay<br />
subsoil with gravel lenses <strong>in</strong>creas<strong>in</strong>g <strong>in</strong> extent with depth. These lenses conta<strong>in</strong>ed primarily<br />
chert, sandstone, and quartzite pebbles or cobbles; novaculite, oolitic chert, silicified wood,<br />
and banded agate were present <strong>in</strong> lesser quantities or as traces. Chert and o<strong>the</strong>r pebbles<br />
tended to be far more wea<strong>the</strong>red than those from Station 1, hav<strong>in</strong>g thick cortical zones <strong>of</strong><br />
alteration.<br />
Methods <strong>of</strong> Collection and Analysis<br />
At each station a 1 x 1 m square was placed at any surface area where coarse material<br />
was visibly concentrated or well represented, and all surface particles <strong>of</strong> 40 mm or greater<br />
diameter were collected without excavat<strong>in</strong>g. A second square was similarly placed, but at a<br />
m<strong>in</strong>imum distance <strong>of</strong> 10 m from <strong>the</strong> first. Then a grab sample <strong>of</strong> about 30 pebbles <strong>of</strong> any size<br />
was collected <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity, choos<strong>in</strong>g apparently high quality lithic materials or unusual<br />
materials. This collect<strong>in</strong>g procedure was <strong>in</strong>tended to reproduce potential aborig<strong>in</strong>al use <strong>of</strong><br />
surficial deposits <strong>in</strong> two respects. First, gravel bars, or lag gravel concentrated on upland<br />
surfaces may have f<strong>in</strong>e particles w<strong>in</strong>nowed away, leav<strong>in</strong>g a pavement <strong>of</strong> coarse particles<br />
highly accessible. Second, <strong>the</strong> m<strong>in</strong>imum functional size <strong>of</strong> pebbles for production <strong>of</strong> most<br />
stone tools and useful flakes is about 40 mm, so far as we can tell from actual pebble cores<br />
and tools <strong>in</strong> site collections. (The factor <strong>of</strong> pebble or cobble shape was not considered <strong>in</strong> our<br />
procedure.)<br />
In <strong>the</strong> laboratory each pebble was measured on <strong>the</strong> long axis, broken with a rock hammer;<br />
identified as to silica m<strong>in</strong>eral (Frondel 1962) or rock type (Travis 1955), and subjectively<br />
evaluated as to knapp<strong>in</strong>g quality. The size frequency data and composition for conchoidally<br />
fractur<strong>in</strong>g materials are dist<strong>in</strong>guished for each sampl<strong>in</strong>g station <strong>in</strong> Figure 57. The 1 x 1 m<br />
subsamples collected at each station were found to be essentially identical <strong>in</strong> size and com-
222 Hemm<strong>in</strong>gs<br />
position, and have been presented as a s<strong>in</strong>gle sample (N=75 <strong>in</strong> each); grab samples (N=30 <strong>in</strong><br />
each) are not <strong>in</strong>cluded <strong>in</strong> Figure 57).<br />
The follow<strong>in</strong>g operat<strong>in</strong>g def<strong>in</strong>itions were employed <strong>in</strong> <strong>the</strong> sort<strong>in</strong>g and identification<br />
procedure:<br />
Chert is a silica m<strong>in</strong>eral commonly orig<strong>in</strong>at<strong>in</strong>g as nodules or beds <strong>in</strong> mar<strong>in</strong>e<br />
limestone or o<strong>the</strong>r sedimentary rocks; generally it is light colored with matte<br />
to lustrous surfaces; m<strong>in</strong>ute fossil rema<strong>in</strong>s are common constituents; cortex on<br />
wea<strong>the</strong>red chert pebbles is variable <strong>in</strong> color and texture (Frondel 1962:221).<br />
Novaculite is a f<strong>in</strong>e-gra<strong>in</strong>ed silica rock derived from chert by <strong>the</strong>rmal metamorphism;<br />
milky white to gray colors and matte gra<strong>in</strong>y surfaces predom<strong>in</strong>ate; fossil<br />
rema<strong>in</strong>s are generally not preserved; “polygonal triple-po<strong>in</strong>t texture” is characteristic<br />
when novaculite is exam<strong>in</strong>ed by scann<strong>in</strong>g electron microscope (Frondel<br />
1962:222; Keller et al. 1977).<br />
Pebbles and cobbles are 2-64 mm and 64-256 mm <strong>in</strong> maximum diameter respectively<br />
<strong>in</strong> terms <strong>of</strong> <strong>the</strong> Wentworth size scale (Compton 1962: 213).<br />
Results <strong>of</strong> Analysis<br />
The Ouachita River bedload and Pleistocene terrace samples are similar with regard<br />
to prevalence <strong>of</strong> pebbles and small cobbles <strong>of</strong> conchoidally fractur<strong>in</strong>g siliceous materials <strong>in</strong><br />
<strong>the</strong> 40-85 mm size range, but <strong>the</strong> latter sample has larger modal sizes and poorer sort<strong>in</strong>g.<br />
The composition <strong>of</strong> <strong>the</strong>se samples is drastically different <strong>in</strong> that <strong>the</strong> first sample is entirely<br />
novaculite and <strong>the</strong> second largely chert (when sorted accord<strong>in</strong>g to def<strong>in</strong>itions and attributes<br />
given above). Many <strong>of</strong> <strong>the</strong> freshly broken pebbles exhibit color, texture, and o<strong>the</strong>r raw material<br />
characteristics which can be matched (on a gross visual comparative basis) with flaked<br />
stone tools or debitage <strong>in</strong> collections from prehistoric floodpla<strong>in</strong> sites <strong>in</strong> <strong>the</strong> project area.<br />
On <strong>the</strong> o<strong>the</strong>r hand, dist<strong>in</strong>ctive novaculite, chert, silicified wood, and crystal quartz varieties<br />
which recur <strong>in</strong> several archeological collections are not represented <strong>in</strong> <strong>the</strong> samples collected<br />
and described above. On a subjective basis each sample <strong>of</strong> 75 pebbles and small cobbles <strong>in</strong>cluded<br />
only one or two relatively unwea<strong>the</strong>red, homogeneous specimens which a fl<strong>in</strong>tknapper<br />
might select for reduction. The novaculite tended to be laced with ve<strong>in</strong>s and fractures,<br />
while <strong>the</strong> chert was deeply wea<strong>the</strong>red with heterogeneous <strong>in</strong>clusions. We conclude from our<br />
analysis <strong>of</strong> both samples that similar local gravel deposits may well have been exploited for<br />
toolmak<strong>in</strong>g raw materials, but that selectivity, by means <strong>of</strong> trial flak<strong>in</strong>g tests, would have<br />
been employed by <strong>the</strong> aborig<strong>in</strong>al fl<strong>in</strong>tknapper. We make <strong>the</strong> presumption that our dredged<br />
and gravel pit samples are representative <strong>of</strong> past channel bar and Pleistocene terrace gravels<br />
widely distributed and exposed <strong>in</strong> <strong>the</strong> project area.
Lithic and Ceramic Analyses 223<br />
Figure 57. Histograms <strong>of</strong> size frequency <strong>in</strong> two samples <strong>of</strong> pebbles and small<br />
cobbles from Ouachita Valley localities.
224 Hemm<strong>in</strong>gs<br />
If this is so, numerous late prehistoric floodpla<strong>in</strong> camps conta<strong>in</strong> pebble chert brought<br />
from upland source areas. In contrast, Archaic and early ceramic floodpla<strong>in</strong> camps could obta<strong>in</strong><br />
novaculite toolmak<strong>in</strong>g raw material and granular rocks for cookery from channel bars,<br />
and might be expected to favor riverbank locations adjacent to <strong>the</strong>se bars (Chapter 6).<br />
lithic AnAlysis And results<br />
Analyses <strong>of</strong> stone tools and a variety <strong>of</strong> <strong>in</strong>organic materials proceeded <strong>in</strong> conjunction<br />
with o<strong>the</strong>r studies <strong>of</strong> Felsenthal site collections. Only about 400 stone tools or fragments<br />
were collected, but many <strong>of</strong> those were recovered <strong>in</strong> s<strong>in</strong>gle component sites and <strong>in</strong> stratigraphic<br />
context. We have studied <strong>the</strong>se tools <strong>in</strong>tensively, employ<strong>in</strong>g functional concepts and<br />
a classification similar to House’s (1975) Cache Project lithic analysis. Approximately 3,000<br />
flakes and 6,000 o<strong>the</strong>r items <strong>of</strong> <strong>in</strong>organic debris have also been studied from this functional<br />
po<strong>in</strong>t <strong>of</strong> view. A series <strong>of</strong> artifact data sheets were devised to meet <strong>the</strong> particular needs and<br />
goals <strong>of</strong> analysis, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> follow<strong>in</strong>g:<br />
Lithic Analysis (Flaked Stone Tools)<br />
Lithic Analysis (Ground Stone Tools)<br />
Lithic Analysis (Debitage)<br />
Inorganic Debris Analysis<br />
These records are reta<strong>in</strong>ed at <strong>the</strong> <strong>Arkansas</strong> Archeological Survey, and results are summarized<br />
and reported <strong>in</strong> this section.<br />
In addition to functional analyses <strong>of</strong> artifacts and site collections, projectile po<strong>in</strong>ts were<br />
exam<strong>in</strong>ed typologically, and projectile po<strong>in</strong>t data were recorded and analyzed. We propose<br />
sequences or partial sequences <strong>of</strong> arrow po<strong>in</strong>t and dart po<strong>in</strong>t types for <strong>the</strong> region <strong>in</strong> this<br />
chapter, based on about 70 specimens from floodpla<strong>in</strong> sites. These sequences are sketchy<br />
and <strong>in</strong>complete, and clearly require fur<strong>the</strong>r work, but we hope <strong>the</strong>y will be fundamentally<br />
useful for organiz<strong>in</strong>g regional archeological data.<br />
Arrow Po<strong>in</strong>ts<br />
Arrow po<strong>in</strong>ts are bilaterally symmetrical bifaces with a po<strong>in</strong>ted tip at one end and a<br />
haft modification (except <strong>in</strong> some triangular and ovate forms) at <strong>the</strong> opposite end (House<br />
1975:60). In large samples <strong>of</strong> American Indian arrow po<strong>in</strong>ts, both ethnographic and archeological<br />
<strong>in</strong> orig<strong>in</strong>, <strong>the</strong> range <strong>of</strong> weight is from about 0.3 to 4.0 g, although heavier examples<br />
occur (Fenenga 1953; Thomas 1978). The majority <strong>of</strong> archeological specimens <strong>of</strong> appropriate<br />
morphology and size/weight can be presumed to have been hafted and propelled by bow,<br />
and were <strong>the</strong>refore associated with hunt<strong>in</strong>g, bow fish<strong>in</strong>g, or warfare. Sociotechnic variants
Lithic and Ceramic Analyses 225<br />
<strong>of</strong> arrow po<strong>in</strong>t styles, dist<strong>in</strong>ct <strong>in</strong> morphology, size, raw material, and craftmanship, are also<br />
known <strong>in</strong> much <strong>of</strong> North America, usually from cache or burial contexts (e.g., <strong>in</strong> <strong>the</strong> Great<br />
Bend <strong>of</strong> Red River and <strong>Arkansas</strong> Valley Caddoan regions).<br />
Introduction <strong>of</strong> <strong>the</strong> bow and arrow <strong>in</strong> various parts <strong>of</strong> eastern North America, <strong>in</strong>clud<strong>in</strong>g<br />
<strong>Arkansas</strong>, occurred with<strong>in</strong> a few centuries <strong>of</strong> A.D. 500. In <strong>the</strong> Felsenthal Project area, this<br />
<strong>in</strong>troduction must have occurred dur<strong>in</strong>g <strong>the</strong> Baytown-Coles Creek period, but no s<strong>in</strong>gle or<br />
multiple component floodpla<strong>in</strong> site yet <strong>in</strong>vestigated has produced evidence which could<br />
identify or date this event. Most arrow po<strong>in</strong>t styles which recur here <strong>in</strong> floodpla<strong>in</strong> sites are<br />
good markers <strong>of</strong> Mississippi period extractive camps.<br />
A brief study <strong>of</strong> arrow po<strong>in</strong>t manufacture, based on Shallow Lake site (3UN9/52) lithic<br />
materials, is reported by Flenniken (<strong>in</strong> Stacy 1976:Appendix A). Many aspects <strong>of</strong> this study<br />
require updat<strong>in</strong>g. However, Flenniken experimentally replicated six arrow po<strong>in</strong>ts <strong>of</strong> generalized<br />
Alba and Ashley type, beg<strong>in</strong>n<strong>in</strong>g with 42-70 mm long chert and novaculite pebbles<br />
collected on <strong>the</strong> site, produc<strong>in</strong>g flakes and flake preforms by percussion and pressure flak<strong>in</strong>g,<br />
and f<strong>in</strong>ally produc<strong>in</strong>g completed arrow po<strong>in</strong>ts by pressure flak<strong>in</strong>g. Each arrow po<strong>in</strong>t required<br />
about six m<strong>in</strong>utes to complete from a detached flake (one failure occurred <strong>in</strong> <strong>the</strong> f<strong>in</strong>al<br />
stage <strong>of</strong> pressure flak<strong>in</strong>g). We suspect that some or even many arrow po<strong>in</strong>ts recovered <strong>in</strong> <strong>the</strong><br />
Felsenthal region were produced by bifacial reduction from pebble cores, while some arrow<br />
po<strong>in</strong>ts were certa<strong>in</strong>ly produced on flake blanks <strong>in</strong> <strong>the</strong> manner <strong>of</strong> Flenniken’s experiment.<br />
More experimentation <strong>of</strong> this k<strong>in</strong>d is clearly needed to understand regional lithic technologies.<br />
Our arrow po<strong>in</strong>t sample (N = 44) <strong>in</strong>cludes 28 complete or nearly complete specimens<br />
as well as basal and tip fragments. Three-quarters <strong>of</strong> <strong>the</strong> identifiable specimens are from<br />
floodpla<strong>in</strong> sites which have been tested, so that assignment to a component is generally possible<br />
(Chapter 6). We are able to propose an arrow po<strong>in</strong>t sequence which extends throughout<br />
<strong>the</strong> Mississippi period (A.D. 1100-<strong>17</strong>00); only a few arrow po<strong>in</strong>ts can be attributed to <strong>the</strong><br />
preced<strong>in</strong>g Baytown-Coles Creek period and <strong>the</strong>se few present certa<strong>in</strong> problems <strong>of</strong> classification<br />
and context. Table 20 is <strong>the</strong> first attempt to organize a sequence <strong>of</strong> arrow po<strong>in</strong>t styles for<br />
<strong>the</strong> Felsenthal region, and this sequence may well be supplemented and modified by future<br />
work. Figure 58 portrays this sequence by means <strong>of</strong> actual specimens recovered <strong>in</strong> site<br />
survey or test<strong>in</strong>g work. The recent report on Shallow Lake site (3UN9/52) by Rol<strong>in</strong>gson and<br />
Schambach (1981) <strong>in</strong>dicates that Catahoula and Rockwall arrow po<strong>in</strong>ts may also occur <strong>in</strong> <strong>the</strong><br />
regional sequence, but no examples were obta<strong>in</strong>ed <strong>in</strong> our work.<br />
The follow<strong>in</strong>g paragraphs provide brief summaries <strong>of</strong> descriptive data, typological<br />
discussion, selected references, and o<strong>the</strong>r comments for each <strong>of</strong> <strong>the</strong> arrow po<strong>in</strong>t types enumerated<br />
<strong>in</strong> Table 21. From among more extensive records <strong>of</strong> our projectile po<strong>in</strong>t study, only<br />
selected attributes are recorded here: N = sample size; L = maximum length or range ob-
226 Hemm<strong>in</strong>gs<br />
Table 20. Proposed Sequence <strong>of</strong> Arrow Po<strong>in</strong>t Types <strong>in</strong> <strong>the</strong> Felsenthal Region.<br />
Regional Type Number <strong>of</strong><br />
(Extraregional Type) Cultural Period/Phase Floodpla<strong>in</strong> Sites Terrace/Island Sites Specimens<br />
Nodena Term<strong>in</strong>al Mississippi One Cypress Po<strong>in</strong>t (3AS286) 6<br />
Term<strong>in</strong>al Mississippi No name (3AS310) 1<br />
Nodena, Banks var. (Shetley) Term<strong>in</strong>al Mississippi One Cypress Po<strong>in</strong>t (3AS286) 1<br />
Madison (Fresno) Term<strong>in</strong>al Mississippi One Cypress Po<strong>in</strong>t (3AS286) 1<br />
Term<strong>in</strong>al Mississippi Marie Sal<strong>in</strong>e (3AS329) 1<br />
Term<strong>in</strong>al Mississippi Lapile Creek 4 (3UN158) 1<br />
Washita (Cahokia) Mississippi/Caney Bayou Marie Sal<strong>in</strong>e (3AS329) 1<br />
Eagle Creek Mississippi/Caney Bayou Jug Po<strong>in</strong>t 1 (3AS306) 3<br />
Bassett Mississippi/Caney Bayou Jug Po<strong>in</strong>t 2 (3AS307) 1<br />
Ashley (Colbert) Mississippi/Caney Bayou Jug Po<strong>in</strong>t 2 (3AS307) 2<br />
Ashley (Colbert) Mississippi/Gran <strong>Marais</strong> Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76) 1<br />
Mississippi/Gran <strong>Marais</strong> False Indigo (3AS285) 2<br />
Mississippi/Gran <strong>Marais</strong> Jones Lake North (3UN81) 1<br />
Mississippi/Gran <strong>Marais</strong> No name (3AS326) 1<br />
Alba Mississippi/Gran <strong>Marais</strong> False Indigo (3AS285) 1<br />
Homan Baytown/Coles Creek* Goulett Island (3BR8) 1<br />
Baytown/Coles Creek* P<strong>in</strong>e Ridge 1 (3UN<strong>17</strong>7) 1<br />
Honey Creek Baytown/Coles Creek* Small Cane 1 (3BR72) 1<br />
Baytown/Coles Creek* One Cypress Po<strong>in</strong>t (3AS286) 1<br />
* Homan and Honey Creek arrowpo<strong>in</strong>ts may perta<strong>in</strong> to unnamed phase(s) late <strong>in</strong> this period; four sites listed here presently lack adequate<br />
evidence for Coles Creek components.
e<br />
Lithic and Ceramic Analyses 227<br />
A b c d<br />
f<br />
g h<br />
i j k l M<br />
n o p Q<br />
Figure 58. Arrow po<strong>in</strong>ts from <strong>the</strong> Felsenthal Project area <strong>in</strong> proposed chronological<br />
sequence. a. Nodena (3AS286); b. Nodena, Banks variant (3AS286);<br />
c. Madison (3AS286); d. Madison (3UN158); e. Washita (3AS329); f.<br />
Eagle Creek (3AS306); g. Bassett (3AS307); h. Ashley (3AS307); i. Ashley<br />
(3BR76); j. Ashley (3UN81); k. Ashley (3AS326); l. Ashley (3AS285); m.<br />
Alba (3AS285); n. Homan (3BR8); o. Homan (3UN<strong>17</strong>7); p. Honey Creek<br />
(3BR72); q. Honey Creek (3AS286) (AAS neg. 813868).
228 Hemm<strong>in</strong>gs<br />
served; HL = haft length or range; W = maximum width or range; WT = weight or range; RM<br />
= raw material count; TA = <strong>the</strong>rmal alteration observed; ( ) = measurement <strong>of</strong> slightly <strong>in</strong>complete<br />
specimen, not an estimate if specimen were complete. Dimensions are <strong>in</strong> millimeters<br />
and weights <strong>in</strong> grams and tenths. All arrow po<strong>in</strong>ts recorded measured between 3 and 7 mm<br />
<strong>in</strong> maximum thickness.<br />
Nodena (Bell 1958:64; Morse 1973:Figure 9; H<strong>of</strong>fman 1977:Figure 8). N = 6, L =<br />
(<strong>17</strong>)-(32), W = 12-14, WT = (0.7)-(2.2), RM = 6 chert, TA = 0.<br />
Slender, willow leaf shaped po<strong>in</strong>ts without haft modification, associated particularly<br />
with late prehistoric/protohistoric cultures <strong>in</strong> eastern <strong>Arkansas</strong>, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> historic Quapaw.<br />
The pr<strong>in</strong>cipal occurrence <strong>in</strong> our project area is at One Cypress Po<strong>in</strong>t (3AS286) on <strong>the</strong><br />
Sal<strong>in</strong>e River, evidently a Quapaw (?) site-unit <strong>in</strong>trusion, ca A.D. 1600-<strong>17</strong>00. No Nodena<br />
po<strong>in</strong>ts have been reported previously from <strong>the</strong> Felsenthal region.<br />
Nodena, Banks variant (Morse 1973:Figure 9c). N = 1, L = 22, W = 12, WT = 1.0,<br />
RM = chert, TA = 0.<br />
This specimen is a variant <strong>of</strong> Nodena with concave basal edge, associated with <strong>the</strong> One<br />
Cypress Po<strong>in</strong>t (3AS286) assemblage. See also Per<strong>in</strong>o (1971:92) on <strong>the</strong> Caddoan Shetley po<strong>in</strong>t.<br />
Madison (Per<strong>in</strong>o 1968:52). N = 3, L = <strong>17</strong>-24, W = 12-<strong>17</strong>, WT = 0.5-2.5, RM = 2 chert,<br />
1 novaculite, TA = 2.<br />
Widely distributed triangular arrow po<strong>in</strong>t <strong>of</strong> <strong>the</strong> Mississippi period. See also Bell<br />
(1960:44) on <strong>the</strong> Fresno po<strong>in</strong>t. One <strong>of</strong> our chert specimens is associated with <strong>the</strong> Nodena<br />
po<strong>in</strong>t assemblage at One Cypress Po<strong>in</strong>t (3AS286) and is comparable <strong>in</strong> raw material, symmetry,<br />
and flak<strong>in</strong>g technique.<br />
Washita (Bell 1958:98; Brown 1976:105). N = 1, W = 19, RM = chert, TA = 1.<br />
Widely distributed, side-notched, triangular po<strong>in</strong>t <strong>of</strong> <strong>the</strong> Mississippi period. See also<br />
Per<strong>in</strong>o (1968:12) on <strong>the</strong> Cahokia po<strong>in</strong>t and variants.<br />
Eagle Creek (provisional type described here). N = 3, L = 21-15, HL = 8-9, W = 10-<br />
13, WT = (0.5)-1.3, RM = 3 chert, TA = 0.<br />
A small slender po<strong>in</strong>t with contract<strong>in</strong>g stem, convex blade edges, and slight shoulders.<br />
Our three specimens are from a s<strong>in</strong>gle site, Jug Po<strong>in</strong>t 4 (3AS306) on <strong>the</strong> Sal<strong>in</strong>e River,<br />
assigned to <strong>the</strong> Caney Bayou phase (Chapter 6). At least one contract<strong>in</strong>g stemmed arrow<br />
po<strong>in</strong>t recovered at Shallow Lake (3UN9/52) may also belong <strong>in</strong> this provisional type (Rol<strong>in</strong>gson<br />
and Schambach 1981:Figure 15n). A small series <strong>of</strong> contract<strong>in</strong>g stemmed arrow po<strong>in</strong>ts,<br />
unidentified to type, was recovered at <strong>the</strong> Salsbury site (16OU15) on Ouachita River below<br />
Monroe, Louisiana (Price and Heartfield 1977:78). No contract<strong>in</strong>g stem arrow po<strong>in</strong>t type<br />
presently def<strong>in</strong>ed subsumes <strong>the</strong> attributes noted above, although <strong>the</strong> silhouette resembles<br />
modified gar scale arrow po<strong>in</strong>ts shown by Suhm and Krieger (1954: Plate 134E). The specimens<br />
from Jug Po<strong>in</strong>t 1 were evidently made on bifacial preforms ra<strong>the</strong>r than flakes, are
Lithic and Ceramic Analyses 229<br />
relatively thick even on haft elements, and could have been <strong>in</strong>serted <strong>in</strong> a socket ra<strong>the</strong>r than<br />
a split or notched shaft. Haft element length (HL) approximates one third <strong>of</strong> <strong>the</strong> total length<br />
(L) <strong>in</strong> <strong>the</strong>se po<strong>in</strong>ts.<br />
Bassett (Bell 1958:10; Webb 1980b:3). N = 1, L = 41, HL = 5, W = 22, WT = 3.0, RM =<br />
chert, TA = 1.<br />
Basally modified, triangular arrow po<strong>in</strong>t with small po<strong>in</strong>ted stem isolated by notch<strong>in</strong>g,<br />
distributed pr<strong>in</strong>cipally <strong>in</strong> <strong>the</strong> Caddoan Red River region. Our ra<strong>the</strong>r large specimen from<br />
Jug Po<strong>in</strong>t 2 (3AS307) may date early <strong>in</strong> <strong>the</strong> Caney Bayou phase or about A.D. 1400-1500, contemporary<br />
with latest arrow po<strong>in</strong>ts <strong>of</strong> this type <strong>in</strong> <strong>the</strong> Caddoan area.<br />
Ashley (Rol<strong>in</strong>gson 1971; Brown 1976:79). N = 7, L = (16)-31, HL = 4-6, W = 13-15,<br />
WR = (0.6)-1.3, RM = 4 chert, 3 novaculite, TA = 2.<br />
Triangular bladed arrow po<strong>in</strong>t, commonly with recurved blade edges and t<strong>in</strong>y graverlike<br />
tip; prom<strong>in</strong>ent, expand<strong>in</strong>g, “bulbous” stem or haft element is one fourth total length<br />
(less <strong>in</strong> long slender forms). This arrow po<strong>in</strong>t type is a marker for <strong>the</strong> Bartholomew phase,<br />
A.D. 1100-1400, <strong>in</strong> sou<strong>the</strong>ast <strong>Arkansas</strong> and <strong>the</strong> contemporary Gran <strong>Marais</strong> phase <strong>in</strong> <strong>the</strong><br />
Felsenthal region, but persists <strong>in</strong>to <strong>the</strong> early Caney Bayou phase <strong>in</strong> <strong>the</strong> project area so far<br />
as we can tell. Brown (1976:79) extends <strong>the</strong> type name <strong>in</strong>to <strong>the</strong> <strong>Arkansas</strong> Valley Caddoan<br />
region (Harlan phase) and notes <strong>the</strong> confusion with Per<strong>in</strong>o’s (1968:34) description <strong>of</strong> earlier<br />
Homan po<strong>in</strong>ts. Ashley po<strong>in</strong>ts <strong>in</strong> <strong>Arkansas</strong> are evidently contemporary with Colbert po<strong>in</strong>ts<br />
<strong>in</strong> northwest Louisiana and overlap morphologically with, or are derived from, earlier Caddoan<br />
arrow po<strong>in</strong>t styles, <strong>in</strong>clud<strong>in</strong>g Homan (see Wood 1963; Webb 1980a, 1980b).<br />
Alba (Bell 1958:8; Brown 1976:61; Webb 1980b:2). N = 1, L;= 23, HL = 6, W = 14,<br />
WT = 0.8, RM = chert, TA = 0.<br />
Small arrow po<strong>in</strong>t, commonly hav<strong>in</strong>g <strong>the</strong> blade form described above (Ashley) but<br />
with essentially rectangular stem; widely distributed <strong>in</strong> <strong>the</strong> Caddoan area eastward to <strong>the</strong><br />
Lower Mississippi Valley. Our specimen from False Indigo (3AS285) may be later than <strong>the</strong><br />
A.D. 750-1100 time range commonly ascribed this type, s<strong>in</strong>ce it is apparently associated with<br />
Ashley po<strong>in</strong>ts <strong>in</strong> a Gran <strong>Marais</strong> phase assemblage.<br />
Homan (Wood 1963; Per<strong>in</strong>o 1968:34; Brown 1976:92; Webb 1980b:4). N = 2, L = 14-<br />
21, HL = 6-7, W = 13-14, WT = 0.3=0.9, RM = 1 chert, 1 novaculite, TA = 1.<br />
Deeply and relatively broadly corner-notched arrow po<strong>in</strong>ts hav<strong>in</strong>g convex basal edge<br />
and relatively broad triangular blade; blade edges straight, slightly convex, or slightly recurved.<br />
This arrow po<strong>in</strong>t style is quite early, ca A.D. 800-900, <strong>in</strong> <strong>the</strong> Ouachita Mounta<strong>in</strong> and<br />
Red River Caddoan regions, but Brown <strong>in</strong>dicates a later date <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong> Valley. The<br />
typological confusion discussed by Brown was noted above. Our two specimens were recovered<br />
on upland sites without known stratigraphic context, and also may not be representative<br />
<strong>of</strong> late Coles Creek or Fourche Mal<strong>in</strong>e arrow po<strong>in</strong>ts elsewhere attributed to this type.
230 Hemm<strong>in</strong>gs<br />
Honey Creek (Derley 1979). N = 2, L = (24)-27, W = 16-20, WT = (1.2)-(1.9), RM = 2<br />
chert, TA = 0.<br />
This is a provisional type, m<strong>in</strong>imally described from <strong>the</strong> Honey Creek site (3PR28)<br />
near <strong>the</strong> lower White River. Numerous specimens which may be assigned to this type have<br />
also been recovered at Toltec Mounds (3LN42), Dawson Mound (3CE68), and <strong>the</strong> Alexander<br />
site (3CN1<strong>17</strong>) <strong>in</strong> <strong>the</strong> central <strong>Arkansas</strong> Valley, and at Roland Mound (3AR30) on <strong>the</strong> lower<br />
White River. Honey Creek po<strong>in</strong>ts are th<strong>in</strong>, “lozenge-shaped” bifaces with straight or slightly<br />
concave blade edges and a convex base without haft<strong>in</strong>g modification. In sites enumerated<br />
above <strong>the</strong>y are most likely associated with a Coles Creek occupation, between about A.D.<br />
700 and 1100, and some or all forms may be advanced preforms for arrow po<strong>in</strong>ts <strong>of</strong> Scallorn-<br />
Rockwall style. In <strong>the</strong> Felsenthal region th<strong>in</strong> well-made bifaces <strong>of</strong> Honey Creek form can<br />
ei<strong>the</strong>r be completed arrow po<strong>in</strong>ts or advanced preforms for arrow po<strong>in</strong>ts, lack<strong>in</strong>g only <strong>the</strong><br />
notches. Only additional work on larger better controlled collections will resolve <strong>the</strong> functional<br />
and classificatory problems posed by <strong>the</strong>se bifaces.<br />
dart po<strong>in</strong>ts<br />
Dart po<strong>in</strong>ts have <strong>the</strong> symmetry and haft modification (except <strong>in</strong> relatively few lanceolate<br />
or ovate forms) noted above for arrow po<strong>in</strong>ts, but almost <strong>in</strong>variably exceed 4.0 g <strong>in</strong><br />
weight. Archeological specimens generally date before A.D. 500 <strong>in</strong> eastern United States,<br />
and can be presumed to have been fixed to a dart foreshaft or simple, light spear shaft which<br />
was cast by means <strong>of</strong> an atlatl. Some special forms may have been cast javel<strong>in</strong>-style or thrust<br />
as lances. Dart po<strong>in</strong>ts were thus components <strong>of</strong> equipment associated pr<strong>in</strong>cipally with<br />
hunt<strong>in</strong>g and occasionally, perhaps, with warfare. As <strong>in</strong> <strong>the</strong> case <strong>of</strong> arrow po<strong>in</strong>ts, sociotechnic<br />
variants may also occur (e.g., Ohio Valley Adena stemmed po<strong>in</strong>ts <strong>in</strong> mortuary context).<br />
Moreover, dart po<strong>in</strong>ts <strong>of</strong> many styles were quite <strong>of</strong>ten refurbished, or modified and recycled,<br />
largely by virtue <strong>of</strong> <strong>the</strong> large mass <strong>of</strong> raw material still available <strong>in</strong> damaged po<strong>in</strong>ts.<br />
Our sample <strong>of</strong> classifiable dart po<strong>in</strong>ts is limited (N=26), chiefly because <strong>of</strong> <strong>the</strong> relatively<br />
few preceramic and early ceramic components recorded or <strong>in</strong>vestigated <strong>in</strong> <strong>the</strong> project<br />
area (Chapter 5). In Table 22 we summarize data for 20 complete or nearly complete dart<br />
po<strong>in</strong>ts <strong>in</strong> typological categories and <strong>in</strong> general temporal group<strong>in</strong>gs. Figure 59 illustrates<br />
this sample by means <strong>of</strong> specimens recovered <strong>in</strong> site survey and test<strong>in</strong>g work. Most <strong>of</strong> our<br />
specimens are readily classified <strong>in</strong> types well known from <strong>the</strong> Lower Mississippi Valley<br />
(Ford and Webb 1956) or from <strong>the</strong> West Gulf Coastal Pla<strong>in</strong> and Sou<strong>the</strong>rn Pla<strong>in</strong>s <strong>in</strong> Louisiana<br />
and Texas (Suhm and Krieger 1954; Webb 1980b; see also Bell 1958, 1960; Per<strong>in</strong>o 1968, 1971).<br />
In general, precise dates for <strong>the</strong>se types have not been obta<strong>in</strong>ed and some types (e.g., Gary)<br />
persisted for centuries or even several millenia. Our broad temporal group<strong>in</strong>gs <strong>in</strong> Table 21<br />
are thus based on extraregional data <strong>of</strong> little precision, but also, <strong>in</strong> a few cases noted below,<br />
on stratigraphic or site assemblage data from our own floodpla<strong>in</strong> sites:
Table 21. Metric and Nonmetric Data for Dart Po<strong>in</strong>ts and Proposed General Sequence <strong>of</strong> Dart Po<strong>in</strong>t Types <strong>in</strong> <strong>the</strong> Felsenthal Region.<br />
Descriptive Data* Provenience<br />
Regional Types L HL W T WT RM TA G Floodpla<strong>in</strong> Sites Terrace/Island Sites<br />
Ceramic (ca. 500 B.C. - A.D. 500)<br />
Gary (small) 37 13 <strong>17</strong> 11 5.0 Ch Buttonbush (3BR58)<br />
Langtry (38) 13 28 9 (8.9) Ch X Marie Sal<strong>in</strong>e (3AS329)<br />
Steuben 48 15 24 10 9.1 Ch X Marie Sal<strong>in</strong>e (3AS329)<br />
Preceramic/Ceramic (ca. 2000 B.C. - A.D. 1)<br />
Gary - 19 (27) 9 (6.4) No X X Marie Sal<strong>in</strong>e (3AS329)<br />
Gary (52) 16 33 7 (10.7) No X Mouth <strong>of</strong> Eagle Creek (3BR78)<br />
Gary (39) - 24 9 (7.7) Ch X Lapile Creek 4 (3UN158)<br />
Gary (35) 14 21 7 (5.1) Ch Lapile Creek 5 (3UN157)<br />
Gary (37) - 20 a (5.6) Ch X Lapile Creek 5 (3UN157)<br />
Gary 47 20 25 9 8.5 Ch Lapile Creek 7 (3UN151)<br />
Gary (43) <strong>17</strong> 32 9 (14.5) No X X Bold<strong>in</strong>g Road (3UN<strong>17</strong>5)<br />
Ellis 60 9 33 10 16.8 No X No name (3AS324)<br />
Ellis 37 11 26 8 6.2 Ch Burnie Creek (3UN132)<br />
Ellis 36 9 27 8 6.0 Ch Cross Road near Burnt Bridge (3UN72)<br />
Marcos (32) 7 30 10 (5.8) Ch Kelly Spr<strong>in</strong>g Creek (3UN<strong>17</strong>3)<br />
Delhi 52 7 29 10 10.9 No X X Potlatch Field (3AS160)<br />
Macon 77 18 28 12 21.0 No S Small Cane 2 (3BR73)<br />
Lithic and Ceramic Analyses 231<br />
Preceramic (ca. 3000 -1000 B.C.)<br />
Big Creek (Williams) 48 15 37 10 14.0 Ch CoulettIsland (3BR8)<br />
yarbrough 50 20 27 11 14.2 No X Marie Sal<strong>in</strong>e (3AS329)<br />
Morhiss - <strong>17</strong> 28 11 (15.5) No X MarieSal<strong>in</strong>e (3AS329)<br />
Carrollton 34 11 22 8 5.0 Ch X Lapile Creek 7 (3UN151)<br />
* L = maximum length <strong>in</strong> mm; HL = haft length <strong>in</strong> mm; W = maximum width <strong>in</strong> mm; T = maximum thickness <strong>in</strong> mm; WT = weight <strong>in</strong> grams; RM = raw material, ei<strong>the</strong>r chert (Ch) or<br />
novaculite (No); TA = <strong>the</strong>rmal alteration observed; G = gr<strong>in</strong>d<strong>in</strong>g on basal edge, stem edges, or notches
232 Hemm<strong>in</strong>gs<br />
A b c d e<br />
f<br />
g<br />
j k l M<br />
o<br />
Figure 59. Dart po<strong>in</strong>ts from <strong>the</strong> Felsenthal Project area <strong>in</strong> proposed chronological<br />
sequence. Specimens are shown <strong>in</strong> <strong>the</strong> same order as Table 22.<br />
(AAA neg. 813869)<br />
h<br />
p Q r s<br />
i<br />
n
Lithic and Ceramic Analyses 233<br />
1. A small Gary po<strong>in</strong>t surface f<strong>in</strong>d (Figure 59a) from Buttonbush (3BR58) may be<br />
associated with Baytown Pla<strong>in</strong> ceramics and may represent occupation early<br />
<strong>in</strong> <strong>the</strong> Baytown-Coles Creek period (A.D. 300-1100); this specimen compares<br />
closely with Schambach’s (1970:200) Gary var. Camden assigned to <strong>the</strong> Oak<br />
Grove phase (300 B.C.-A.D. 300) <strong>in</strong> <strong>the</strong> Mid-Ouachita region (see also Rol<strong>in</strong>gson<br />
and Schambach 1981:99).<br />
2. A small dart po<strong>in</strong>t, essentially <strong>the</strong> Langtry type but with markedly slender<br />
stem (Figure 59b), was recovered at Marie Sal<strong>in</strong>e (3AS329) <strong>in</strong> a deep undisturbed<br />
level (70-80 cm) with one Baytown Pla<strong>in</strong> sherd, and may also represent<br />
occupation early <strong>in</strong> <strong>the</strong> Baytown/Coles Creek period; Langtry po<strong>in</strong>ts are related<br />
to <strong>the</strong> variable cluster <strong>of</strong> Gary po<strong>in</strong>t forms which persist <strong>in</strong>to <strong>the</strong> first millenium<br />
A.D. <strong>in</strong> Oklahoma (Baerreis et al. 1958) and elsewhere.<br />
3. An expand<strong>in</strong>g stemmed po<strong>in</strong>t (Figure 59c), highly similar to <strong>the</strong> Steuben type,<br />
was recovered <strong>in</strong> yet a deeper level (90-100 cm) at Marie Sal<strong>in</strong>e (3AS329) without<br />
o<strong>the</strong>r diagnostic artifacts directly associated, and could perta<strong>in</strong> to an o<strong>the</strong>rwise<br />
“<strong>in</strong>visible” Marksville period (ca A.D. 1-300) occupation (on <strong>the</strong> basis <strong>of</strong><br />
depth alone we would assign this po<strong>in</strong>t to <strong>the</strong> preced<strong>in</strong>g Tchula period); Steuben<br />
po<strong>in</strong>ts were described by Morse (1963) <strong>in</strong> late Middle Woodland context<br />
(ca A.D. 500) <strong>in</strong> Ill<strong>in</strong>ois, but none have been identified previously <strong>in</strong> <strong>Arkansas</strong>.<br />
4. A large, well-made Gary po<strong>in</strong>t (Figure 59e) surface f<strong>in</strong>d at Mouth <strong>of</strong> Eagle<br />
Creek (3BR78) may perta<strong>in</strong> to preceramic Poverty Po<strong>in</strong>t period (1200-500 B.C.)<br />
occupation.<br />
5. A crude corner-notched Ellis po<strong>in</strong>t (Figure 59j) was recovered <strong>in</strong> place at 2.0 m<br />
depth without associated artifacts at riverbank site 3AS324, and may perta<strong>in</strong> to<br />
<strong>the</strong> Late Archaic period (ca 4000-1200 B.C.).<br />
6. A long, well-made Macon po<strong>in</strong>t (Figure 59o) recovered on <strong>the</strong> surface with<br />
fire-cracked rock at Small Cane 2 (3BR73) may perta<strong>in</strong> to Poverty Po<strong>in</strong>t or Late<br />
Archaic period occupation.<br />
7. A Big Creek (or Williams) corner-notched dart po<strong>in</strong>t (Figure 59p) from <strong>the</strong><br />
surface at Goulett Island (3BR8) may perta<strong>in</strong> to Poverty Po<strong>in</strong>t or Late Archaic<br />
period occupation; Big Creek po<strong>in</strong>ts are estimated to date at about 750 B.C. and<br />
Williams po<strong>in</strong>ts at about 2500 B.C. (Morse 1970; Schambach 1970:152; Per<strong>in</strong>o<br />
1971:10; Rol<strong>in</strong>gson and Schambach 1981:99). There seems to be little basis for<br />
dist<strong>in</strong>guish<strong>in</strong>g <strong>the</strong>se as regionally dist<strong>in</strong>ct types <strong>of</strong> great age difference.<br />
Rol<strong>in</strong>gson and Schambach (1981:Table 15) describe a series <strong>of</strong> dart po<strong>in</strong>ts from <strong>the</strong><br />
Shallow Lake site (3UN9/52) <strong>in</strong> a private collection and also provide a tentative sequence <strong>of</strong>
234 Hemm<strong>in</strong>gs<br />
dart po<strong>in</strong>t types for <strong>the</strong> Felsenthal region. There is relatively little overlap <strong>in</strong> types identified<br />
and ordered between <strong>the</strong>ir work and ours, partly because <strong>of</strong> sampl<strong>in</strong>g error and partly<br />
because we have necessarily dealt with <strong>the</strong> last 3500 years <strong>of</strong> such a sequence. However, it is<br />
clear that a detailed regional sequence is emerg<strong>in</strong>g.<br />
In Figure 60 we show flaked stone tools with <strong>the</strong> haft modification <strong>of</strong> dart po<strong>in</strong>ts. Two<br />
specimens represent asymmetrical blades adapted to cutt<strong>in</strong>g tasks, one has a scrap<strong>in</strong>g edge<br />
prepared on a large thick blade, and two specimens have blades markedly reduced by resharpen<strong>in</strong>g.<br />
Bifacial Preforms and O<strong>the</strong>r Bifaces<br />
Bifacial artifacts o<strong>the</strong>r than projectile po<strong>in</strong>ts (N = 88) were frequently recovered <strong>in</strong> site<br />
survey and test<strong>in</strong>g work. One site collection has an excellent series <strong>of</strong> <strong>the</strong>se items (3AS286),<br />
but most were scattered f<strong>in</strong>ds <strong>in</strong> small sites <strong>of</strong> all time periods (or <strong>of</strong> unknown age). A particular<br />
problem <strong>of</strong> biface analysis is <strong>the</strong> high frequency <strong>of</strong> fragments for which only limited<br />
observations can be made. We can dist<strong>in</strong>guish preforms as an important functional category,<br />
but o<strong>the</strong>r classes <strong>of</strong> bifacial tools and also multipurpose tools are present <strong>in</strong> floodpla<strong>in</strong> sites.<br />
Preforms are symmetrical, generally ovate or trianguloid bifaces which lack <strong>the</strong> f<strong>in</strong>e<br />
marg<strong>in</strong>al/facial retouch and haft modification <strong>of</strong> projectile po<strong>in</strong>ts, and also lack prepared<br />
and/or utilized edges <strong>of</strong> bifacial cutt<strong>in</strong>g tools. Cortex and thick cross section may dist<strong>in</strong>guish<br />
early from advanced stages <strong>of</strong> preform reduction. Aborted preforms commonly exhibit raw<br />
material irregularities and breakage (House 1975:67). The presence <strong>of</strong> preforms <strong>in</strong> archeological<br />
assemblages <strong>in</strong>dicates manufacture, or at least <strong>in</strong>itial stages <strong>of</strong> manufacture, <strong>of</strong> bifacial<br />
tools such as projectile po<strong>in</strong>ts.<br />
Bifacial cutt<strong>in</strong>g tools may or may not exhibit bilateral symmetry and haft modification,<br />
but will commonly have s<strong>in</strong>uous to regular lateral edges with edge angles less than 35° and<br />
evidence <strong>of</strong> use-wear on <strong>the</strong>se edges (Wilmsen 1968:156). Depend<strong>in</strong>g on results <strong>of</strong> use-wear<br />
analysis, assemblages with such cutt<strong>in</strong>g tools may reflect butchery activity.<br />
Bifacial choppers are relatively heavy bifaces or cobbles with one or more cutt<strong>in</strong>g<br />
edges, hav<strong>in</strong>g edge angles <strong>in</strong> <strong>the</strong> 45-65° range and evidence <strong>of</strong> pronounced edge wear or impact<br />
damage (House 1975:62). Such tools are likely to have been produced and employed <strong>in</strong><br />
expedient fashion for multiple tasks, <strong>in</strong>clud<strong>in</strong>g food process<strong>in</strong>g and work<strong>in</strong>g <strong>of</strong> tough plant<br />
or animal materials.<br />
Figure 60 illustrates bifacial preforms and o<strong>the</strong>r bifacial tools from various sites <strong>in</strong> <strong>the</strong><br />
project area. In Chapter 6 <strong>the</strong> sample <strong>of</strong> flaked stone tools from One Cypress Po<strong>in</strong>t (3AS286)
Lithic and Ceramic Analyses 235<br />
Figure 60. Bifacial tools, preforms, and cores. a. hafted knife, Shallow Lake West<br />
(3UN154); b. hafted knife, Lapile Creek 8 (3UN152); c. hafted scraper,<br />
Small Cane 1 (3BR72); d. resharpened dart po<strong>in</strong>t, Lapile Creek 2 (3UN160);<br />
e. resharpened dart po<strong>in</strong>t, Goulett Island (3BR8); f. arrow po<strong>in</strong>t preform,<br />
Goulett Island (3BR8); g. arrow po<strong>in</strong>t preform, Persimmon 3 (3AS315); h.<br />
dart po<strong>in</strong>t preform, Lapile Creek 4 (3UN158); i. dart po<strong>in</strong>t preform, Lapile<br />
Creek 8 (3UN152); j. preform, Willow Oak (3AS312); k. preform, Persimmon<br />
2 (3AS314); l. preform/chopper, unnamed site (3BR88); m. chert pebble core,<br />
Goulett Island (3BR8); n. novaculite pebble core, Marie Sal<strong>in</strong>e (3AS329); o.<br />
silicified wood tabular core, Goulett Island (3BR8) (AAS neg. 813870).
236 Hemm<strong>in</strong>gs<br />
A<br />
h<br />
l<br />
b<br />
i<br />
c d e<br />
n<br />
M<br />
j<br />
f g<br />
o<br />
k
Lithic and Ceramic Analyses 237<br />
<strong>in</strong>cludes arrow po<strong>in</strong>t preforms and o<strong>the</strong>r bifaces perta<strong>in</strong><strong>in</strong>g to a s<strong>in</strong>gle term<strong>in</strong>al Mississippi<br />
period component (Figure 48).<br />
O<strong>the</strong>r Flaked Stone Tools<br />
A variety <strong>of</strong> o<strong>the</strong>r modified flaked stone tools, both bifacial and unifacial, was recovered<br />
<strong>in</strong> site survey and test<strong>in</strong>g work; <strong>the</strong> number from any s<strong>in</strong>gle site is small and <strong>the</strong> total<br />
number considered here is also relatively small (N = 41). The collection from One Cypress<br />
Po<strong>in</strong>t (3AS286) aga<strong>in</strong> conta<strong>in</strong>s a greater variety <strong>of</strong> specialized flaked stone tools, examples <strong>of</strong><br />
which are shown <strong>in</strong> Figure 48. A generalization we can draw from all floodpla<strong>in</strong> site collections<br />
is that projectile po<strong>in</strong>ts, po<strong>in</strong>t fragments, and preforms which reflect use, manufacture,<br />
or refurbish<strong>in</strong>g <strong>of</strong> weapons are relatively more numerous than ma<strong>in</strong>tenance types <strong>of</strong> tools<br />
(B<strong>in</strong>ford and B<strong>in</strong>ford 1966). Four functional types <strong>of</strong> flaked stone tools are described below.<br />
Scrapers are edged tools which may or may not have haft modification, but <strong>in</strong>clude<br />
a moderately steep work<strong>in</strong>g edge, generally <strong>in</strong> <strong>the</strong> 45-55° edge angle range (Wilmsen<br />
1968:156). Placement and shape <strong>of</strong> work<strong>in</strong>g edge def<strong>in</strong>e certa<strong>in</strong> common scraper<br />
types: end scraper, side scraper, spokeshave, plane, and o<strong>the</strong>r (see House 1975: 62).<br />
Also, both bifacial and unifacial scraper forms occur, although plano-convex cross<br />
sections and unifacially prepared work<strong>in</strong>g edges predom<strong>in</strong>ate. Scrapers are associated<br />
with a variety <strong>of</strong> process<strong>in</strong>g and ma<strong>in</strong>tenance activities, <strong>in</strong>clud<strong>in</strong>g hide work<strong>in</strong>g,<br />
rett<strong>in</strong>g <strong>of</strong> fibrous vegetal material, and shav<strong>in</strong>g <strong>of</strong> wood, bone, and antler. The<br />
few identified scrapers or scraperlike tools are listed below:<br />
1. Two small chert endscrapers associated with <strong>the</strong> term<strong>in</strong>al Mississippi period<br />
component at One Cypress Po<strong>in</strong>t (3AS286), and considered to be diagnostic <strong>of</strong><br />
Nodena or Quapaw phases <strong>in</strong> o<strong>the</strong>r areas (Chapter 6, Figure 48c).<br />
2. Side scraper on a chert flake, Marie Sal<strong>in</strong>e (3AS329), probably Poverty Po<strong>in</strong>t or<br />
Tchula period (Figure 61a).<br />
3. Discoidal scraper on a chert flake, Eagle Creek 6 (3BR79) probably Mississippi<br />
period (Figure 61b).<br />
4. Spokeshave on a novaculite flake, P<strong>in</strong>e Island Road (3AS290), probably Archaic<br />
(Figure 61c).<br />
5. Plane or heavy scrap<strong>in</strong>g tool on chert cobble, Mud Lake Bend 2 (3UN166),<br />
probably Archaic.<br />
Drills and perforators are bifacial or sometimes m<strong>in</strong>imally retouched flake tools<br />
with an elongate bit, usually bear<strong>in</strong>g traces <strong>of</strong>f rotary wear. Bit wear is a function<br />
<strong>of</strong> drill<strong>in</strong>g implement hardness relative to raw material hardness and also duration<br />
<strong>of</strong> implement use; transverse bit breakage is frequently observed. Drills
238 Hemm<strong>in</strong>gs<br />
commonly have haft modification and a bit approach<strong>in</strong>g cyl<strong>in</strong>drical form; perforators<br />
generally lack haft modification and have a short, contract<strong>in</strong>g bit with<br />
alternately beveled edges. This class <strong>of</strong> tools can be associated with fabricat<strong>in</strong>g<br />
tasks, <strong>in</strong>clud<strong>in</strong>g pierc<strong>in</strong>g or ream<strong>in</strong>g wood, bone, shell, and o<strong>the</strong>r moderately<br />
hard materials. Only three drills and five perforators were recovered <strong>in</strong> site<br />
survey and test<strong>in</strong>g work and most <strong>of</strong> <strong>the</strong>se are from One Cypress Po<strong>in</strong>t (3AS286).<br />
The follow<strong>in</strong>g specimens can be assigned to particular components:<br />
1. One cyl<strong>in</strong>drical chert drill and a cyl<strong>in</strong>drical chert drill tip associated with <strong>the</strong><br />
term<strong>in</strong>al Mississippi period component at One Cypress Po<strong>in</strong>t (3AS286); <strong>the</strong><br />
cyl<strong>in</strong>drical drill form without haft modification appears to be diagnostic <strong>of</strong><br />
Nodena or Quapaw phases <strong>in</strong> o<strong>the</strong>r areas (Chapter 6, Figure 48d).<br />
2. Three perforators on chert flakes associated with <strong>the</strong> same component at<br />
3AS286 (Figure 48h).<br />
3. Perforator on a chert flake; Caney Bayou phase component at Jug Po<strong>in</strong>t Cut<strong>of</strong>f<br />
(3BR76) (Figure 61d).<br />
4. An expanded base drill on a novaculite flake associated with <strong>the</strong> Caney Bayou<br />
phase component (ca A.D. 1400-1500) at Jug Po<strong>in</strong>t 2 (3AS307) (Figure 61f).<br />
Gravers are m<strong>in</strong>imally retouched flakes with one or more small sharp po<strong>in</strong>ts or<br />
spurs isolated and formed by unifacial retouch. Haft<strong>in</strong>g modification is absent.<br />
Use-wear, when dist<strong>in</strong>guishable, consists <strong>of</strong> back-and-forth striae or rotary wear<br />
or both (cf. House 1975:64). Gravers are <strong>the</strong>refore associated with fabricat<strong>in</strong>g<br />
tasks, <strong>in</strong>clud<strong>in</strong>g engrav<strong>in</strong>g, <strong>in</strong>cis<strong>in</strong>g, or punctat<strong>in</strong>g moderately hard raw materials<br />
such as wood, bone, and shell. Only one grav<strong>in</strong>g tool, a small chert flake with<br />
two delicate spurs (Figure 61g) was recovered <strong>in</strong> our work. This specimen is<br />
tentatively associated with a Caney Bayou phase component at Overcup Oak 3<br />
(3AS305).<br />
Retouched flakes have one or more marg<strong>in</strong>s trimmed or regularly flaked to<br />
produce a work<strong>in</strong>g edge or to refurbish a previously dulled edge. Such edges<br />
may have been employed <strong>in</strong> a variety <strong>of</strong> cutt<strong>in</strong>g/scrap<strong>in</strong>g activities, may show<br />
heavy to undetectable use-wear, and may be especially difficult to dist<strong>in</strong>guish<br />
from un<strong>in</strong>tentional edge-wear which takes a regular form (shear<strong>in</strong>g). Moderately<br />
steep work<strong>in</strong>g edges <strong>in</strong>tergrade with those <strong>of</strong> tools identified as scrapers. Retouched<br />
flakes are most reliably associated with process<strong>in</strong>g or fabricat<strong>in</strong>g tasks<br />
when a large series is available for study with<strong>in</strong> a s<strong>in</strong>gle site assemblage. Our<br />
sample (N = 22) does not meet this requirement <strong>in</strong> that specimens were recovered<br />
<strong>in</strong>frequently <strong>in</strong> widely scattered sites. Examples are illustrated <strong>in</strong> Figure 48i and<br />
Figure 61g-i.
A b c d e<br />
f<br />
g<br />
h<br />
Lithic and Ceramic Analyses 239<br />
Figure 61. Miscellaneous flaked stone tools. a. side scraper, Marie Sal<strong>in</strong>e (3AS329); b.<br />
discoidal scraper, Eagle Creek 6 (3BR79); c. spokeshave, P<strong>in</strong>e Island Road<br />
(3AS290); d. perforator, Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76); e. perforator, Redeye Lake<br />
Prairie (3AS293); f. expanded base drill, Jug Po<strong>in</strong>t 2 (3AS307); g. graver,<br />
Overcup Oak 3 (3AS305); h. retouched flake, P<strong>in</strong>e Island Road (3AS290); i.<br />
retouched blade or bladelike flake, Marie Sal<strong>in</strong>e (3AS329); j. utilized flakes,<br />
Lapile Creek 7 (3UN151). (AAS neg. 813871)<br />
i<br />
j
240 Hemm<strong>in</strong>gs<br />
Utilized flakes are characterized by irregular edge damage and/or use-wear<br />
which cannot be attributed to purposeful modification <strong>of</strong> tool form. Utilization<br />
can be so brief and traces <strong>of</strong> this utilization so m<strong>in</strong>imal that presumed expedient<br />
tools <strong>in</strong>tergrade with waste flakes modified by cultural or natural processes not<br />
directly related to tool use, e.g., postdepositional dump<strong>in</strong>g and trampl<strong>in</strong>g (Flenniken<br />
and Haggarty 1979). Our sample <strong>of</strong> utilized flakes (N = 188) is large, variable,<br />
and difficult to associate with particular tasks. Utilized flakes constituted<br />
64% <strong>of</strong> <strong>the</strong> flaked stone tool assemblage recovered at One Cypress Po<strong>in</strong>t (3AS286)<br />
where small th<strong>in</strong> chert flakes or bladelike flakes with nibbled edges are abundant<br />
(Chapter 6, Figure 48i). We can characterize <strong>the</strong>se tools from <strong>the</strong> term<strong>in</strong>al Mississippi<br />
period component by type <strong>of</strong> flake selected as follows (see also Cores and<br />
Debitage below):<br />
primary flakes 19<br />
secondary flakes 47<br />
<strong>in</strong>terior flakes 24<br />
flakes <strong>of</strong> bifacial retouch 24<br />
Total 114<br />
These utilized flakes are generally <strong>of</strong> markedly small size (maximum dimension<br />
= 11-30 mm, mean maximum dimension = 18.9 mm). These and o<strong>the</strong>r data for <strong>the</strong><br />
One Cypress Po<strong>in</strong>t lithic assemblage suggest that flakes were randomly selected<br />
from <strong>the</strong> population <strong>of</strong> waste flakes, and expediently employed as delicate, disposable,<br />
cutt<strong>in</strong>g tools.<br />
Utilized flakes were isolated by <strong>in</strong>spection dur<strong>in</strong>g sort<strong>in</strong>g and were exam<strong>in</strong>ed under a<br />
b<strong>in</strong>ocular microscope. Detailed study at high magnification <strong>of</strong> use-wear on tool edges was<br />
beyond <strong>the</strong> scope <strong>of</strong> this study. O<strong>the</strong>r examples <strong>of</strong> utilized flakes or bladelike flakes from<br />
floodpla<strong>in</strong> sites are shown <strong>in</strong> Figure 61j. The dist<strong>in</strong>ction between utilization and postdepositional<br />
damage among <strong>the</strong>se specimens requires additional microscopic exam<strong>in</strong>ation and<br />
analysis.<br />
Cores and Debitage<br />
Lithic waste result<strong>in</strong>g from flaked stone tool production or refurbish<strong>in</strong>g is termed debitage,<br />
and any substantial assemblage <strong>of</strong> this material may <strong>in</strong>clude three subcategories:<br />
Cores or masses <strong>of</strong> conchoidally fractured raw material which have two or more<br />
flake removals, recur <strong>in</strong> a variety <strong>of</strong> dist<strong>in</strong>ctive forms, and exhibit core preparation<br />
but not use-wear; such cores occur <strong>in</strong> archeological assemblages as <strong>the</strong> result<br />
<strong>of</strong> discard (rejected and exhausted cores) or merely through loss (House 1975:65).
Lithic and Ceramic Analyses 241<br />
Waste flakes <strong>of</strong> conchoidally fractured raw material which have one or more attributes<br />
associated with controlled flake removal (platforms, dorsal and ventral<br />
surfaces, etc.); such flakes may occur complete or as fragments, but lack evidence<br />
<strong>of</strong> use-wear; an assemblage <strong>of</strong> waste flakes attests to fl<strong>in</strong>tknapp<strong>in</strong>g activity <strong>in</strong><br />
general, and variability with<strong>in</strong> this assemblage may adduce <strong>the</strong> range <strong>of</strong> raw<br />
material types <strong>in</strong> use, forms undergo<strong>in</strong>g reduction, and sequence <strong>of</strong> knapp<strong>in</strong>g<br />
techniques employed (House 1975:67).<br />
Shatter, or blocky fragments <strong>of</strong> conchoidally fractur<strong>in</strong>g raw material which lack<br />
all flake characteristics; such fragments are ord<strong>in</strong>arily present as a m<strong>in</strong>or fraction<br />
<strong>of</strong> any substantial assemblage <strong>of</strong> debitage and are presumed to orig<strong>in</strong>ate dur<strong>in</strong>g<br />
pretreatment or reduction <strong>of</strong> raw material.<br />
Debitage (N = 3256) was nearly ubiquitous on floodpla<strong>in</strong> sites and o<strong>the</strong>r prehistoric<br />
sites exam<strong>in</strong>ed dur<strong>in</strong>g survey and test<strong>in</strong>g work (Appendix A, Table A-1). We will attempt to<br />
characterize this material generally, not<strong>in</strong>g that more than half <strong>of</strong> <strong>the</strong> sample was recovered<br />
<strong>in</strong> two late Mississippi period sites (3AS286 and 3AS306), and that relatively little material<br />
attributable to Archaic components was collected.<br />
Cores are chiefly pebbles or small cobbles reduced bifacially or <strong>in</strong> irregular fashion,<br />
but some tabular specimens also occur (Figure 60m-o). In <strong>the</strong> best available series, 12 pebble<br />
cores from One Cypress Po<strong>in</strong>t (3AS286), maximum dimensions ranged between 28 and 66<br />
mm, <strong>the</strong> smaller specimens seem<strong>in</strong>gly split and rejected (Figure 48j). At least half <strong>of</strong> <strong>the</strong>se<br />
pebble cores were <strong>the</strong>rmally altered prior to <strong>the</strong> flake removal. Eleven cores are brown chert<br />
and one is novaculite. Direct freehand percussion and, perhaps more rarely, bipolar percussion<br />
were techniques used to reduce <strong>the</strong>se cores. Bifacially flaked pebble cores and early<br />
bifacial preforms <strong>in</strong>tergrade <strong>in</strong> this assemblage (Figure 48). Among <strong>the</strong> many o<strong>the</strong>r pebble<br />
cores from scattered sites, we could f<strong>in</strong>d no dist<strong>in</strong>ctive differences from this group, although<br />
examples from Archaic or early ceramic sites were generally lack<strong>in</strong>g.<br />
Debitage and cores were classified as heat altered when some comb<strong>in</strong>ation <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g<br />
attributes occurred: (1) discoloration and differential discoloration, generally toward<br />
p<strong>in</strong>k or gray, (2) waxy or vitreous luster where flakes are detached after heat<strong>in</strong>g, (3) craz<strong>in</strong>g,<br />
crenated fracture, or unopened fractures, and (4) pot lids or spalls. Heat alteration may<br />
imply heat treatment to improve knapp<strong>in</strong>g quality; however, any collection may conta<strong>in</strong><br />
<strong>in</strong>advertently heat altered specimens or heat treated specimens not dist<strong>in</strong>guished by <strong>the</strong><br />
obtrusive attributes cited above.
242 Hemm<strong>in</strong>gs<br />
Waste flakes from all sites were sorted by raw material (see chert and novaculite<br />
above) and by reduction categories, def<strong>in</strong>ed as follows:<br />
Primary flakes have 50% or more <strong>of</strong> outer (dorsal) surface covered by cortex, and<br />
generally have hard hammer flak<strong>in</strong>g characteristics (prom<strong>in</strong>ent bulb, ripples, etc.)<br />
on <strong>in</strong>ner (ventral) surface;<br />
Secondary flakes have less than 50% cortex on dorsal surface;<br />
Interior flakes lack cortex; and<br />
Flakes <strong>of</strong> bifacial retouch are th<strong>in</strong> expand<strong>in</strong>g flakes without cortex, generally hav<strong>in</strong>g<br />
s<strong>of</strong>t hammer characteristics (diffuse bulb, lipped lenticular platform, etc.) on<br />
<strong>the</strong> ventral surface.<br />
These categories reflect <strong>in</strong> a general way <strong>the</strong> reduction sequence from pebble cores<br />
through several stages <strong>of</strong> bifacial preforms to completed bifacial tools, represented schematically<br />
<strong>in</strong> Figure 62 (see also Shafer 1973:43; House 1975:Figure 1). In Chapter 6 (Figure 49b)<br />
we compared horizontally separated subsamples from One Cypress Po<strong>in</strong>t (3AS286) <strong>in</strong> terms<br />
<strong>of</strong> <strong>the</strong>se general reduction categories for cores, waste flakes, and shatter.<br />
Almost all site survey collections conta<strong>in</strong> less than 50 items <strong>of</strong> debitage, and most<br />
conta<strong>in</strong> only a few waste flakes (Appendix A). Qualitative observations from <strong>the</strong>se small collections<br />
suggested that brown pebble chert reduction was highly associated with Mississippi<br />
period sites and gray or white novaculite reduction with Tchula period and all preceramic<br />
sites. Table 22 exam<strong>in</strong>es <strong>the</strong>se observations for three site collections <strong>of</strong> adequate size (from<br />
comb<strong>in</strong>ed surface collections and test excavation) and <strong>of</strong> reasonably well known age; see<br />
Chapter 6). For <strong>the</strong>se samples <strong>the</strong>re are <strong>in</strong> fact strik<strong>in</strong>g differences <strong>in</strong> chert/novaculite ratios<br />
and also <strong>in</strong> proportions <strong>of</strong> flakes <strong>in</strong> various reduction categories. The earlier sample <strong>in</strong> this<br />
comparison, Marie Sal<strong>in</strong>e, Area C (3AS329), stands out by virtue <strong>of</strong> <strong>the</strong> high frequency <strong>of</strong><br />
novaculite and low frequency <strong>of</strong> primary and secondary flakes (which bear cortex). Evidently<br />
novaculite bifaces or bifacial preforms were be<strong>in</strong>g reduced here, ra<strong>the</strong>r than pebble cores,<br />
dur<strong>in</strong>g <strong>the</strong> early occupation at Marie Sal<strong>in</strong>e. We suggest that <strong>the</strong>se raw material preferences<br />
and reduction strategies have regional significance, and should be tested more thoroughly<br />
<strong>in</strong> future work, with larger site assemblages.
Lithic and Ceramic Analyses 243<br />
Figure 62. Schematic l<strong>in</strong>ear model for production <strong>of</strong> a bifacial tool and discard materials
244 Hemm<strong>in</strong>gs<br />
Table 22. Comparison <strong>of</strong> Debitage by Reduction Category and Raw Material Frequency<br />
among Three Floodpla<strong>in</strong> Site Assemblages.<br />
Reduction Categories<br />
Bifacial Chert/<br />
Sample Primary Secondary Interior Retouch Novaculite<br />
Identification <strong>of</strong> Component Size Cores Flakes Flakes Flakes Flakes Shatter Ratio<br />
One Cypress Po<strong>in</strong>t (3AS286) 1084 12 222 343 <strong>17</strong>3 318 16 9:1<br />
(ca. A.D. 1600-<strong>17</strong>00)<br />
Jug Po<strong>in</strong>t 1 (3AS306) 677 4 100 185 193 180 15 3:1<br />
(ca. A.D. 1400-1600)<br />
Marie Sal<strong>in</strong>e, Area C (3AS329) 159 5 2 5 24 111 12 1:10<br />
(ca. 1200 B.C.-A.D. 1)<br />
The follow<strong>in</strong>g list <strong>of</strong> conchoidally fractur<strong>in</strong>g raw materials describes generally all<br />
types which occur among debitage and have been dist<strong>in</strong>guished <strong>in</strong> Felsenthal Project site<br />
collections.<br />
Major or abundant types: chert (pale brown, brown, yellowish brown, reddened);<br />
novaculite (white, gray, yellowish brown, p<strong>in</strong>k, reddened).<br />
M<strong>in</strong>or or poorly represented types: chert (oolitic); novaculite (translucent and<br />
dark ve<strong>in</strong>ed); clear crystal quartz; quartz (milky, rose); quartzite (<strong>of</strong>f-white, gray);<br />
silicified wood (varicolored and gra<strong>in</strong>ed).<br />
Some rarely occurr<strong>in</strong>g, but dist<strong>in</strong>ctive, raw materials are associated with particular site assemblages,<br />
and may eventually prove to have regional cultural or temporal significance.<br />
Clear crystal quartz, for example, may have been acquired from <strong>the</strong> central <strong>Arkansas</strong>-<br />
Ouachita Mounta<strong>in</strong> area through trade or long range procurement.<br />
Cobble and Heavy Duty Tools<br />
This class <strong>of</strong> tools <strong>in</strong>cludes rounded to tabular cobbles modified by specific modes <strong>of</strong><br />
use, result<strong>in</strong>g <strong>in</strong> (1) battered or pitted edges and surfaces produced by percussion, (2) abraded,<br />
faceted, or grooved edges and surfaces produced by gr<strong>in</strong>d<strong>in</strong>g, and (3) comb<strong>in</strong>ations <strong>of</strong><br />
<strong>the</strong>se. The use-wear patterns are dist<strong>in</strong>ctive for certa<strong>in</strong> tool forms, such as hammerstones<br />
and pitted cobbles or anvils, which are easily separated from naturally modified cobbles <strong>in</strong><br />
most cases. However, where use-wear is dubious, we have assigned <strong>the</strong> specimen to a
Lithic and Ceramic Analyses 245<br />
manuport category (see Inorganic Debris below). Cobble tools were evidently employed <strong>in</strong> a<br />
variety <strong>of</strong> “tough” extractive and ma<strong>in</strong>tenance tasks. Some pitted cobbles, for example, may<br />
be associated with nut process<strong>in</strong>g and o<strong>the</strong>rs with bipolar core reduction (House 1975:71).<br />
In spite <strong>of</strong> <strong>the</strong> accessibility <strong>of</strong> cobbles and <strong>the</strong> apparent utility <strong>of</strong> cobbles tools, we<br />
found such tools to be exceed<strong>in</strong>gly rare <strong>in</strong> <strong>the</strong> project area. In fact, <strong>the</strong>re are only 18 examples<br />
<strong>of</strong> hammerstones and o<strong>the</strong>r modified cobbles <strong>in</strong> all site survey and test excavation<br />
collections. For <strong>the</strong> floodpla<strong>in</strong> sites <strong>in</strong> our sample, mostly late prehistoric extractive camps,<br />
expedient cobbles tools were ei<strong>the</strong>r unimportant, were conserved for fur<strong>the</strong>r use at a base<br />
settlement, or were not sufficiently modified by use to permit differentiation from manuports.<br />
We have <strong>the</strong> impression that relatively more cobble and heavy duty tools occur <strong>in</strong> Archaic<br />
sites, but no presently available data can be relied upon to substantiate this observation. The<br />
follow<strong>in</strong>g examples illustrate <strong>the</strong> range <strong>of</strong> cobble tools:<br />
1. battered novaculite hammerstone with abraded facets on oppos<strong>in</strong>g ends, Mud<br />
Lake Bend 2 (3UN166), probably Archaic (Figure 63a).<br />
2. edge-abraded and facially abraded sandstone cobble mano(?), unnamed site<br />
3UN169 (Figure 63b).<br />
3. facially abraded and pitted sandstone cobble anvil, False Indigo (3AS285), associated<br />
with Gran <strong>Marais</strong> component (Figure 63c).<br />
4. two edge-battered and facially pitted sandstone cobble anvils from a buried<br />
feature at Swamp Privet (3AS327), probably Baytown-Coles Creek period (Figure<br />
63d).<br />
Ground Stone Tools<br />
This group <strong>of</strong> seven artifacts is very small and diverse. Some warrant little more than<br />
brief mention, but several are highly <strong>in</strong>formative <strong>in</strong> a particular site context or are extremely<br />
unusual f<strong>in</strong>ds <strong>in</strong> <strong>the</strong> region.<br />
1. Two steatite vessel sherds <strong>of</strong> small size (Figure 64a) were recovered at Marie<br />
Sal<strong>in</strong>e (3AS329), Test Pit 8, 120 cm below <strong>the</strong> surface. These sherds are <strong>17</strong><br />
mm thick, rough <strong>in</strong> appearance with tool marks pla<strong>in</strong>ly visible, on s<strong>of</strong>t, gray,<br />
coarsely schistose steatite. Although no o<strong>the</strong>r diagnostic artifacts were directly<br />
associated, stratigraphic context <strong>in</strong>dicates that this deep level perta<strong>in</strong>s to a<br />
Poverty Po<strong>in</strong>t period occupation, ca 1200-500 B.C. Steatite vessels are well represented<br />
at <strong>the</strong> Poverty Po<strong>in</strong>t site <strong>in</strong> Louisiana, and <strong>the</strong> source <strong>of</strong> steatite has
246 Hemm<strong>in</strong>gs<br />
A<br />
Figure 63. Cobble tools. a. hammerstone, Mud Lake Bend 2 (3UN166); b. mano?,<br />
unnamed site (3UN169); c. pitted cobble or anvil, False Indigo<br />
(3AS285); d. pitted cobble anvils, Swamp Privet (3AS327). (AAS neg.<br />
813872)<br />
b<br />
d e<br />
c
A<br />
c<br />
b<br />
Lithic and Ceramic Analyses 247<br />
f<br />
e<br />
d<br />
Figure 64. Ground stone tools and pigment. a. steatite sherds, Marie Sal<strong>in</strong>e (3AS329); b. schist slab, Marie Sal<strong>in</strong>e<br />
(3AS329); c. hematite plummet, One Cypress Po<strong>in</strong>t (3AS286); d. shale gorget fragment, Persimmon 1<br />
(3AS313); e. amber(?) disk fragment, Jug Po<strong>in</strong>t 1 (3AS306); f. red ochre or earthy hematite, Marie Sal<strong>in</strong>e<br />
(3AS329). Length <strong>of</strong> c is 62 mm (note millimeter scale for item e) (AAS neg. 813873, 806895)
248 Hemm<strong>in</strong>gs<br />
been attributed to <strong>the</strong> Sou<strong>the</strong>rn Appalachian region (Ford and Webb 1956).<br />
The <strong>in</strong>frequent occurrence <strong>of</strong> steatite vessels and fragments <strong>in</strong> sou<strong>the</strong>rn <strong>Arkansas</strong><br />
has been summarized by Schambach (1974). It is clear that <strong>the</strong>se vessels<br />
did move long distances through a Poverty Po<strong>in</strong>t exchange network.<br />
2. A tabular section <strong>of</strong> muscovite schist or talc schist (Figure 64b) was also recovered<br />
at Marie Sal<strong>in</strong>e (3AS329), Test Pit 3, 110-120 cm below <strong>the</strong> surface.<br />
This specimen is about 18 mm thick and is probably modified <strong>in</strong> shape, but<br />
lacks dist<strong>in</strong>guish<strong>in</strong>g tool marks. It is conceivably a vessel fragment. Aga<strong>in</strong>, we<br />
believe that Poverty Po<strong>in</strong>t occupants <strong>in</strong>troduced this object, <strong>the</strong> raw material<br />
probably be<strong>in</strong>g orig<strong>in</strong>ally derived from <strong>the</strong> Sou<strong>the</strong>rn Appalachian region.<br />
3. A plummet <strong>of</strong> massive hematite (Figure 64c) was a surface f<strong>in</strong>d at One Cypress<br />
Po<strong>in</strong>t (3AS286). Although associated with term<strong>in</strong>al Mississippi period occupation<br />
debris, it is stylistically and technologically a much older artifact, strongly<br />
resembl<strong>in</strong>g magnetite and hematite plummets from Poverty Po<strong>in</strong>t (Ford and<br />
Webb 1956:Figure 33). The form is an elongated ovoid, broken through <strong>the</strong><br />
perforation at one end and subsequently smoo<strong>the</strong>d over. The surface is wea<strong>the</strong>red<br />
and etched. An important source area for this m<strong>in</strong>eral is at Magnet Cove<br />
on <strong>the</strong> headwaters <strong>of</strong> <strong>the</strong> Ouachita River <strong>in</strong> central <strong>Arkansas</strong>. Ford and Webb<br />
attribute plummets to use <strong>of</strong> <strong>the</strong> bolas <strong>in</strong> hunt<strong>in</strong>g fowl and small game. The<br />
late Mississippi period occupants <strong>of</strong> One Cypress Po<strong>in</strong>t need not have reused<br />
<strong>the</strong> plummet <strong>in</strong> this manner, but perhaps were merely “curat<strong>in</strong>g” <strong>the</strong> mass <strong>of</strong><br />
hematite for o<strong>the</strong>r eventual use.<br />
4. A small fragment <strong>of</strong> shale gorget (Figure 64d) was retrieved from <strong>the</strong> surface<br />
at Persimmon 1 (3AS313), but may derive from a buried Poverty Po<strong>in</strong>t period<br />
occupation. The fragment reta<strong>in</strong>s a rounded edge, ground and polished face,<br />
and two remnants <strong>of</strong> drill holes. Raw material is a s<strong>of</strong>t, fissile, gray shale <strong>of</strong><br />
unknown orig<strong>in</strong>.<br />
5. A t<strong>in</strong>y chip or flake from an unidentified ground stone tool was recovered on<br />
<strong>the</strong> surface at Mud Lake Bend 1 (3UN167), a probable Archaic component. The<br />
material is light colored siltstone.<br />
6. A perforated disc fragment, possibly a bead <strong>of</strong> fossil amber (Figure 64e), was<br />
found on <strong>the</strong> surface at Jug Po<strong>in</strong>t 1, Area A (3AS306), and is attributed to<br />
Caney Bayou phase occupation (ca A.D. 1400-1600) at that site. This artifact is<br />
clearly a unique f<strong>in</strong>d and warrants detailed treatment, but only some prelim<strong>in</strong>ary<br />
observations can be presented here <strong>in</strong> regard to outward appearance and
Lithic and Ceramic Analyses 249<br />
<strong>in</strong>dications <strong>of</strong> raw material and source. F<strong>in</strong>al results <strong>of</strong> a complex analysis <strong>of</strong><br />
<strong>the</strong> amber (?) are not yet available, but some prelim<strong>in</strong>ary f<strong>in</strong>d<strong>in</strong>gs and methods<br />
used will be noted.<br />
7. The rema<strong>in</strong><strong>in</strong>g fragment represents about one third <strong>of</strong> a disc 24 mm <strong>in</strong> diameter<br />
and 4.5 mm <strong>in</strong> maximum thickness, centrally perforated as shown <strong>in</strong><br />
Figure 64e. The disc was not flat, but gently concave; that is, <strong>the</strong> obverse face<br />
shown is concave and <strong>the</strong> reverse face is parallel and convex. The obverse face<br />
bears <strong>the</strong> only o<strong>the</strong>r modification evident, and this is a broad shallow groove<br />
(0.9 mm deep) closely encircl<strong>in</strong>g <strong>the</strong> central perforation. The effect <strong>of</strong> this<br />
groove is to leave a raised rim about <strong>the</strong> <strong>in</strong>terior and exterior marg<strong>in</strong>s. The<br />
form is thus rem<strong>in</strong>iscent <strong>of</strong> European style buttons, but matches none that we<br />
have seen <strong>in</strong> all details. Nor does it reproduce <strong>in</strong> detail <strong>the</strong> form <strong>of</strong> any shell<br />
disc bead we have seen.<br />
The raw material is s<strong>of</strong>t, markedly lightweight, amber colored, and translucent,<br />
although a wea<strong>the</strong>red crust makes it appear opaque <strong>in</strong> reflected light.<br />
No manufactur<strong>in</strong>g marks are visible, perhaps because <strong>of</strong> <strong>the</strong> crust. The broken<br />
edges do not reveal much <strong>of</strong> <strong>the</strong> nature <strong>of</strong> fracture <strong>in</strong> fresher material. We easily<br />
melted a corner with a quick pass <strong>of</strong> a match’s flame. Most <strong>of</strong> <strong>the</strong>se qualitative<br />
observations led us to make <strong>the</strong> comparison with amber or fossil res<strong>in</strong><br />
which can be characterized as follows (Langenheim 1964, 1969):<br />
(a) Honey colored to molasses colored commonly, but nearly colorless, ruby<br />
red, and even “blue” varieties are well known.<br />
(b) Specific gravity commonly 1.05-1.08 (not quite float<strong>in</strong>g, but easily transported<br />
by water).<br />
(c) Hardness 1 to 3 Mohs scale (scratched by f<strong>in</strong>gernail to calcite).<br />
(d) Melt<strong>in</strong>g po<strong>in</strong>t 150° to 420° C (a variable quality like color).<br />
(e) Consist<strong>in</strong>g <strong>of</strong> res<strong>in</strong> acids, res<strong>in</strong> alcohols, and resene (<strong>in</strong>soluble <strong>in</strong> water<br />
and decay-resistant <strong>in</strong> alluvial environments, although progressive oxidation<br />
and polymerization occur).<br />
Langenheim and Beck (1968) have published a catalogue <strong>of</strong> <strong>in</strong>frared spectra <strong>of</strong> fossil<br />
ambers, which chemically “f<strong>in</strong>gerpr<strong>in</strong>ts” fossil and modern res<strong>in</strong>s and assists <strong>in</strong> identify<strong>in</strong>g<br />
botanical sources. An important occurrence <strong>of</strong> amber <strong>in</strong> Eocene Gulf Coastal Pla<strong>in</strong> lignitic<br />
sediments at Malvern, <strong>Arkansas</strong>, has been described by Saunders et al. (1974).<br />
Our approach to analyz<strong>in</strong>g <strong>the</strong> disc fragment has <strong>in</strong>volved <strong>in</strong>frared spectrophotometric<br />
study (at <strong>the</strong> Department <strong>of</strong> Chemistry, Vassar College, by Pr<strong>of</strong>. Curt W. Beck, and also at<br />
<strong>the</strong> <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>). At <strong>the</strong> present time we have identified <strong>the</strong> material as a res<strong>in</strong>,<br />
but with an admixture <strong>of</strong> silica not characteristic <strong>of</strong> known amber sources. Certa<strong>in</strong>ly, our
250 Hemm<strong>in</strong>gs<br />
specimen does not match <strong>the</strong> published Malvern, <strong>Arkansas</strong>, amber spectra. If we are able<br />
to confirm <strong>the</strong> artifact from Jug Po<strong>in</strong>t 1 as fossil amber, it will represent <strong>the</strong> first record <strong>of</strong><br />
aborig<strong>in</strong>al use <strong>of</strong> this material <strong>in</strong> sou<strong>the</strong>astern United States, and one <strong>of</strong> relatively few such<br />
records for <strong>the</strong> cont<strong>in</strong>ent.<br />
M<strong>in</strong>eral Pigments<br />
Scattered f<strong>in</strong>ds <strong>of</strong> earthy hematite and limonite <strong>in</strong> <strong>the</strong> project area <strong>in</strong>dicate that at<br />
least <strong>the</strong>se two potential sources <strong>of</strong> m<strong>in</strong>eral pigment were locally available. Yellow-brown<br />
limonite, <strong>in</strong> particular, occurs as concretions, coat<strong>in</strong>gs, and m<strong>in</strong>or lenses <strong>in</strong> floodpla<strong>in</strong> sediments.<br />
Dark red ocherous masses <strong>of</strong> hematite were found more rarely, and may have been<br />
<strong>in</strong>troduced by aborig<strong>in</strong>al occupants <strong>of</strong> floodpla<strong>in</strong> sites, although none was dist<strong>in</strong>guished by<br />
facets or striations from pigment preparation. The Magnet Cove area was cited above as a<br />
potential source <strong>of</strong> hematite. Two red ocher masses, illustrated <strong>in</strong> Figure 64f, were excavated<br />
<strong>in</strong> a deep level at Marie Sal<strong>in</strong>e (3AS329) and were presumably associated with <strong>the</strong> Poverty<br />
Po<strong>in</strong>t period occupation at that site.<br />
Inorganic Debris<br />
Inorganic waste materials produced and discarded on an aborig<strong>in</strong>al site may furnish<br />
important functional data s<strong>in</strong>ce <strong>the</strong>y (1) are resistant to decay, (2) were rarely removed or recycled<br />
compared to many classes <strong>of</strong> tools, and (3) can sometimes be associated with specific<br />
k<strong>in</strong>ds <strong>of</strong> activity. In analyses <strong>of</strong> site assemblages presented <strong>in</strong> Chapter 6 we <strong>in</strong>cluded considerations<br />
<strong>of</strong> debris, as well as ceramics and stone tools. Three categories <strong>of</strong> debris which recur<br />
repeatedly <strong>in</strong> floodpla<strong>in</strong> sites, but <strong>in</strong> significantly vary<strong>in</strong>g frequency, are discussed here.<br />
Fire-cracked rock <strong>in</strong>cludes granular and sometimes conchoidally fractur<strong>in</strong>g<br />
cobbles and cobble fragments <strong>in</strong> an archeological assemblage which are characterized<br />
by (1) irregular or created fracture surfaces and craz<strong>in</strong>g or <strong>in</strong>complete<br />
fractures, (2) p<strong>in</strong>k, red, or gray discoloration and differential discoloration with<strong>in</strong><br />
a cobble, and (3) absence <strong>of</strong> breakage features which can be attributed to percussion.<br />
House and Smith (1975) report on <strong>the</strong> experimental replication <strong>of</strong> firecracked<br />
rock, and problem <strong>of</strong> dist<strong>in</strong>guish<strong>in</strong>g it from o<strong>the</strong>r debris. Fire-cracked<br />
rock from floodpla<strong>in</strong> sites <strong>in</strong> <strong>the</strong> Felsenthal Project area <strong>in</strong>cludes sandstone, novaculite,<br />
chert, quartzite, and quartz. We noted a high frequency <strong>of</strong> fire-cracked<br />
rock <strong>in</strong> preceramic sites and low frequency or absence <strong>in</strong> ceramic sites. This<br />
observation is emphasized <strong>in</strong> <strong>the</strong> analysis and comparison <strong>of</strong> excavated Poverty<br />
Po<strong>in</strong>t/Tchula period levels at Marie Sal<strong>in</strong>e (3AS329, Area C). Abundant firecracked<br />
rock <strong>in</strong> preceramic floodpla<strong>in</strong> sites is believed to reflect <strong>the</strong> employment<br />
<strong>of</strong> stone boil<strong>in</strong>g and earth oven cookery as pr<strong>in</strong>cipal food preparation techniques.
Lithic and Ceramic Analyses 251<br />
Fired clay particles are hardened clay masses <strong>of</strong> irregular form and variable, but<br />
generally small size, which are light colored (oxidized), dark colored (reduced),<br />
or clouded due to contact with wood-fueled fires. Specimens from floodpla<strong>in</strong><br />
sites are few and scattered <strong>in</strong> occurrence, and very few have <strong>the</strong> dist<strong>in</strong>ctive<br />
vegetal impressions <strong>of</strong> daub or o<strong>the</strong>r prepared surfaces. No modeled “Poverty<br />
Po<strong>in</strong>t objects” are known from our sites, but good examples are reported from<br />
o<strong>the</strong>r sites <strong>in</strong> <strong>the</strong> region (Rol<strong>in</strong>gson and Schambach 1981:<strong>17</strong>9). Most clay particles<br />
exam<strong>in</strong>ed by us can probably be attributed to <strong>in</strong>advertent fir<strong>in</strong>g <strong>of</strong> floodpla<strong>in</strong><br />
silty clays <strong>in</strong> and around hearths, ra<strong>the</strong>r than to burned structures or prepared<br />
clay features. Only close association with artifacts and o<strong>the</strong>r cultural debris separates<br />
<strong>the</strong> majority <strong>of</strong> <strong>the</strong>se items from naturally occurr<strong>in</strong>g fired clay fragments.<br />
We noted an unusually high frequency <strong>of</strong> fired clay particles <strong>in</strong> two Mississippi<br />
period extractive camps (3AS285 and 3BR76), which are believed to be bulk fish<br />
process<strong>in</strong>g stations (Chapter 6, Table <strong>17</strong>). Among <strong>the</strong> mostly irregular masses or<br />
lumps from <strong>the</strong>se sites are a series <strong>of</strong> small ovoid pellets and several daublike<br />
specimens which warrant fur<strong>the</strong>r study. Substantial amounts <strong>of</strong> daub or daublike<br />
material were recovered on <strong>the</strong> surface at Lapile Creek 4 (3UN158), a multicomponent<br />
ceramic site.<br />
Manuports are pebbles, cobbles, blocks, or slabs <strong>of</strong> any rock material which are<br />
evidently unmodified or exhibit simple breakage <strong>of</strong> unknown orig<strong>in</strong>, but which<br />
occur <strong>in</strong> a context which implies human <strong>in</strong>troduction and manipulation. In<br />
our site survey work many such objects were recorded and collected from f<strong>in</strong>e<br />
gra<strong>in</strong>ed floodpla<strong>in</strong> deposits where o<strong>the</strong>r artifacts or suspected artifact were present.<br />
At Marie Sal<strong>in</strong>e (3AS329), 273 pebble manuports were positively correlated<br />
with o<strong>the</strong>r artifact classes <strong>in</strong> excavated levels with<strong>in</strong> Area C (Chapter 6). Pebble<br />
raw materials are those known to be present <strong>in</strong> <strong>the</strong> bedload gravels <strong>of</strong> <strong>the</strong> Ouachita<br />
River, although some chert pebbles may have been <strong>in</strong>troduced from Pleistocene<br />
Terrace gravels. Manuports presently appear to be <strong>the</strong> least useful class <strong>of</strong><br />
debris for functional analysis, although <strong>the</strong>y can contribute to site identification<br />
(Chapter 5).<br />
cerAMic AnAlysis And results<br />
Felsenthal Project ceramic collections <strong>in</strong>clude about 3,700 sherds from 60 sites. A s<strong>in</strong>gle<br />
floodpla<strong>in</strong> site, False Indigo (3AS285) produced 45% <strong>of</strong> <strong>the</strong> total collection, and ano<strong>the</strong>r,<br />
Buttonbush (3BR58), produced 454 sherds attributed to two Baytown Pla<strong>in</strong> vessels. Most site<br />
collections are markedly small, conta<strong>in</strong><strong>in</strong>g less than 15 sherds, and many sherds are highly<br />
wea<strong>the</strong>red, even when recovered from buried context by test excavation. Our analysis took<br />
<strong>the</strong> form <strong>of</strong> a detailed <strong>in</strong>ventory and comparison <strong>in</strong> <strong>the</strong> case <strong>of</strong> small sherd samples, but <strong>in</strong><br />
<strong>the</strong> case <strong>of</strong> False Indigo we have attempted a more complex stylistic analysis (see Gran
252 Hemm<strong>in</strong>gs<br />
<strong>Marais</strong> complex below). Ceramic data were recorded on a standard form entitled “Ceramic<br />
Analysis (Sherds),” except that additional records were generated for <strong>the</strong> stylistic analysis<br />
noted above. These records are reta<strong>in</strong>ed at <strong>the</strong> <strong>Arkansas</strong> Archeological Survey.<br />
We were fortunate to have Schambach’s unpublished ceramic analysis and detailed<br />
notes for <strong>the</strong> Shallow Lake site (3UN9/52) as a current source for regional ceramic data and<br />
classification (report now <strong>in</strong> pr<strong>in</strong>t; see Rol<strong>in</strong>gson and Schambach 1981). Dr. Schambach<br />
served as a ceramic consultant and exam<strong>in</strong>ed samples <strong>of</strong> sherds as <strong>the</strong> project proceeded.<br />
Regional site reports, published before 1979 or unpublished, are few and conta<strong>in</strong> ceramic<br />
descriptions and data which are <strong>in</strong>consistent and must be used with caution. Schambach’s<br />
study provides a new set <strong>of</strong> methods and concepts, and also <strong>in</strong>dicates that systematic ceramic<br />
analysis <strong>in</strong> <strong>the</strong> Felsenthal region has much potential for future work.<br />
In <strong>the</strong> follow<strong>in</strong>g sections we briefly characterize, compare, and illustrate four ceramic<br />
complexes best known from floodpla<strong>in</strong> site collections. “Complexes” here recognized may<br />
correspond with phases, periods, or even greater temporal units (cf. Phillips 1970:30). There<br />
are no floodpla<strong>in</strong> site data for diagnostic Marksville ceramic types, little data for def<strong>in</strong>itive<br />
Coles Creek ceramics and few examples <strong>of</strong> obviously <strong>in</strong>troduced Caddoan ceramic types.<br />
Presently, it seems clear that floodpla<strong>in</strong> extractive sites lack <strong>the</strong> ceramic abundance and diversity<br />
<strong>of</strong> upland habitation sites <strong>in</strong> <strong>the</strong> region. However, many <strong>of</strong> our small sherd samples<br />
are useful “time capsules” and <strong>in</strong>dicators <strong>of</strong> site activity, as well as sources <strong>of</strong> technological<br />
and stylistic ceramic data.<br />
Tchefuncte Complex<br />
The earliest known ceramics <strong>in</strong> <strong>the</strong> Felsenthal region are low fired, untempered or<br />
clay-tempered wares with dist<strong>in</strong>ctive “contorted lam<strong>in</strong>ated texture” (Phillips 1970:163),<br />
which may be <strong>in</strong>dist<strong>in</strong>guishable from Lower Mississippi Valley Tchefuncte ceramics. Schambach<br />
(1981:181) notes <strong>the</strong> presence <strong>of</strong> Tchefuncte Pla<strong>in</strong>, Tchefuncte Incised, Tchefuncte<br />
Stamped, and Lake Borgne Incised types at Shallow Lake (3UN9/52) and several o<strong>the</strong>r sites<br />
<strong>in</strong> <strong>the</strong> Felsenthal region. Our small sample <strong>of</strong> Tchefuncte ceramics is taken from five newly<br />
recorded floodpla<strong>in</strong> sites, as follows (see also Figure 65):<br />
False Indigo (3AS285)<br />
Lake Borgne Incised 20 sherds<br />
Tchefuncte Incised 18 sherds<br />
Tchefuncte Pla<strong>in</strong> ? 9 sherds<br />
47 sherds<br />
River Birch 2 (3AS321)<br />
Tchefuncte Stamped 1 sherd<br />
Tchefuncte Incised ? 1 sherd<br />
2 sherds
A<br />
A<br />
b<br />
c<br />
Lithic and Ceramic Analyses 253<br />
f<br />
d e<br />
Figure 65. Tchefuncte complex ceramics. a. Lake Borgne Incised, False Indigo (3AS285); b. Tchefuncte Incised, False Indigo<br />
(3AS285); c. Tchefuncte Pla<strong>in</strong>?, False Indigo (3AS285); d. Tchefuncte Stamped, River Birch 2 (3AS321); e.<br />
Tchefuncte Incised?, Marie Sal<strong>in</strong>e (3AS329); f. Tchefuncte Stamped, Hunter’s Swan (3BR70). (AAS neg. 813875)
254 Hemm<strong>in</strong>gs<br />
Marie Sal<strong>in</strong>e (3AS329)<br />
Tchefuncte Incised 14 sherds<br />
Eagle Creek (3BR65)<br />
Tchefuncte Stamped 1 sherd<br />
Hunter’s Swan (3BR70)<br />
Tchefuncte Stamped 2 sherds<br />
Local varieties <strong>of</strong> <strong>the</strong>se Tchefuncte types have not been named, and we cannot make<br />
such dist<strong>in</strong>ctions based on <strong>the</strong> few uneroded sherds available. We can suggest that an early<br />
Coon Island phase component (ca 500-250 B.C.) is relatively well represented at Marie<br />
Sal<strong>in</strong>e (3AS329) from 80-120 cm depth, and that a later Coon Island phase component (ca<br />
250 B.C.-A.D. 1) is represented at False Indigo (3AS285) from 60-0 cm depth. Tchefuncte<br />
Incised may predom<strong>in</strong>ate <strong>in</strong> <strong>the</strong> earlier component, and Lake Borgne Incised <strong>in</strong> <strong>the</strong> later<br />
one. There are no depth data for Tchefuncte Stamped sherds <strong>in</strong> our floodpla<strong>in</strong> site sample.<br />
Phillips (1970:165) and o<strong>the</strong>r workers <strong>in</strong> Louisiana (We<strong>in</strong>ste<strong>in</strong> and Rivet 1978:50) recognize<br />
<strong>the</strong> temporal priority <strong>of</strong> rocker stamped over <strong>in</strong>cised or drag-and-jab Tchefuncte decorative<br />
treatments. Two or more Tchula period phases with dist<strong>in</strong>ctive ceramic assemblages may<br />
eventually be identified <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong> portion <strong>of</strong> <strong>the</strong> Felsenthal region.<br />
Baytown Complex<br />
A long-lived clay-tempered pla<strong>in</strong>ware tradition <strong>in</strong> <strong>the</strong> Felsenthal region has been identified<br />
most closely with <strong>the</strong> Baytown culture <strong>of</strong> <strong>the</strong> Lower Mississippi Alluvial Valley ra<strong>the</strong>r<br />
than with pre-Caddo or Fourche Mal<strong>in</strong>e cultures to <strong>the</strong> west. Accord<strong>in</strong>g to Schambach (<strong>in</strong><br />
Rol<strong>in</strong>gson and Schambach 1981:149), <strong>the</strong> Fourche Mal<strong>in</strong>e pla<strong>in</strong>ware type, Williams Pla<strong>in</strong>,<br />
has a low ratio <strong>of</strong> rim sherds to basal sherds <strong>in</strong> certa<strong>in</strong> site samples, while Baytown Pla<strong>in</strong><br />
has a higher rim/base ratio at Shallow Lake and o<strong>the</strong>r sites <strong>in</strong> <strong>the</strong> Felsenthal region. Presumably,<br />
this ratio reflects <strong>the</strong> prevalence <strong>of</strong> narrow-mou<strong>the</strong>d beaker forms <strong>in</strong> Williams Pla<strong>in</strong><br />
and open-mou<strong>the</strong>d jars and bowls <strong>in</strong> Baytown Pla<strong>in</strong>. Both pla<strong>in</strong>wares have square and disc<br />
bases, although <strong>the</strong> frequency <strong>of</strong> <strong>the</strong>se may also vary among sites and regions. Disc bases<br />
are predom<strong>in</strong>ant <strong>in</strong> our collections.<br />
Baytown Pla<strong>in</strong> utility vessels were obviously important items <strong>of</strong> equipment <strong>in</strong> floodpla<strong>in</strong><br />
extractive sites. Numerous rim sherds and body sherds are recorded for sites <strong>of</strong> several<br />
time periods. One site alluded to earlier, Buttonbush (3BR58) on <strong>the</strong> Sal<strong>in</strong>e River, consisted<br />
pr<strong>in</strong>cipally <strong>of</strong> <strong>the</strong> fragments <strong>of</strong> two wide-mou<strong>the</strong>d, Baytown Pla<strong>in</strong> vessels with disc bases,<br />
broken and concentrated <strong>in</strong> a small sherd cluster (Figure 45). A second similar site on <strong>the</strong><br />
Ouachita River, Water Elm 3 (3AS287), conta<strong>in</strong>ed a cluster <strong>of</strong> 298 pla<strong>in</strong> and four <strong>in</strong>cised<br />
sherds, represent<strong>in</strong>g one or two Baytown vessels with disc bases. Marie Sal<strong>in</strong>e (3AS329),<br />
Mo-Pac (3UN121) and o<strong>the</strong>r riverbank sites have produced Baytown Pla<strong>in</strong> sherds <strong>in</strong> place at
Lithic and Ceramic Analyses 255<br />
80-90 cm and shallower depths. A selection <strong>of</strong> Baytown Pla<strong>in</strong> sherds from <strong>the</strong>se and o<strong>the</strong>r<br />
floodpla<strong>in</strong> sites is illustrated <strong>in</strong> Figure 66.<br />
Baytown pla<strong>in</strong>ware appears to dom<strong>in</strong>ate ceramic assemblages from about A.D. 1 to<br />
1400, to dim<strong>in</strong>ish <strong>in</strong> proportion to decorated wares after this date, and to recur <strong>in</strong> trace<br />
amounts with shell-tempered ceramics after A.D. 1500. We have been unable to identify<br />
Marksville period sites because no diagnostic decorated Marksville ceramic types have been<br />
recovered (Schambach reports a site collection with 1,532 pla<strong>in</strong> and 18 decorated sherds<br />
upstream from <strong>the</strong> project area; cf. Rol<strong>in</strong>gson and Schambach 1981:182). We also have been<br />
generally unable to differentiate ceramic collections from Baytown (A.D. 300-700) and Coles<br />
Creek periods (A.D. 700-1100) because <strong>of</strong> <strong>the</strong> uniformity <strong>of</strong> pla<strong>in</strong>wares and rarity <strong>of</strong> decorated<br />
wares. It is probable that Baytown pla<strong>in</strong>wares varied <strong>in</strong> temper admixtures, surface<br />
f<strong>in</strong>ishes, and vessel forms and sizes through <strong>the</strong> 1500-year span <strong>of</strong> this ceramic tradition,<br />
but no thoroughgo<strong>in</strong>g analysis <strong>of</strong> pla<strong>in</strong>wares has been attempted for <strong>the</strong> Felsenthal region.<br />
Depth data for floodpla<strong>in</strong> sites with Baytown pla<strong>in</strong>ware (Appendix A, Table A-1) may eventually<br />
provide <strong>the</strong> key to dist<strong>in</strong>guish<strong>in</strong>g early and late phases with<strong>in</strong> a Baytown-Coles Creek<br />
cont<strong>in</strong>uum.<br />
Floodpla<strong>in</strong> sites recorded by us at about 30-50 cm depth may possibly represent Coles<br />
Creek occupations, ca A.D. 700-1100. Such sites which conta<strong>in</strong>ed ceramics <strong>in</strong>cluded only<br />
pla<strong>in</strong>ware and wea<strong>the</strong>red, <strong>in</strong>determ<strong>in</strong>ate sherds. We are <strong>the</strong>refore unable to describe a Coles<br />
Creek ceramic complex, <strong>the</strong> obvious l<strong>in</strong>k with earlier Baytown pla<strong>in</strong>ware and with later<br />
ceramic complexes hav<strong>in</strong>g abundant, diverse and decorative treatment. The reader may<br />
f<strong>in</strong>d an <strong>in</strong>troductory statement on <strong>the</strong> regional Coles Creek ceramic complex <strong>in</strong> Rol<strong>in</strong>gson<br />
and Schambach (1981:182-189). Figure 67a illustrates a Coles Creek Incised rim sherd from<br />
Eutaw Rapids (3BR75) which was recovered on <strong>the</strong> Ouachita River shorel<strong>in</strong>e (a buried shell<br />
lens nearly 25 cm <strong>in</strong> depth may have been <strong>the</strong> place <strong>of</strong> orig<strong>in</strong>). This large bowl sherd has two<br />
lip l<strong>in</strong>es and a haphazard l<strong>in</strong>e at <strong>the</strong> base <strong>of</strong> <strong>the</strong> rim fold, resembl<strong>in</strong>g Phillips’s (1970:72) Yazoo<br />
Bas<strong>in</strong>, var. Ely. Much <strong>of</strong> <strong>the</strong> Coles Creek decorative treatment is strongly cont<strong>in</strong>ued and<br />
elaborated <strong>in</strong> <strong>the</strong> Gran <strong>Marais</strong> ceramic complex, discussed below.<br />
Gran <strong>Marais</strong> Complex<br />
Late <strong>in</strong> <strong>the</strong> Coles Creek period or at <strong>the</strong> outset <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> phase, Mississippi<br />
period (A.D. 1100-1400), marked changes occurred with<strong>in</strong> <strong>the</strong> <strong>in</strong>digenous Baytown-Coles<br />
Creek ceramic tradition. Schambach characterizes <strong>the</strong>se changes as follows (Rol<strong>in</strong>gson and<br />
Schambach 1981: 107):<br />
<strong>the</strong>y were foreign to <strong>the</strong> Lower Mississippi Valley and <strong>the</strong>y resulted <strong>in</strong> a ceramic<br />
tradition that defies classification accord<strong>in</strong>g to <strong>the</strong> Phillips system. The ma<strong>in</strong> <strong>in</strong>novations<br />
were to decorate vessel bodies as well as <strong>the</strong> rims or shoulders
256 Hemm<strong>in</strong>gs<br />
A<br />
f<br />
c<br />
h i<br />
Figure 66. Baytown complex pla<strong>in</strong>ware. a. rim, Mo-Pac (3UN121); b. bowl rims, False<br />
Indigo (3AS285); c. car<strong>in</strong>ated bowl rims, False Indigo (3AS285); d. rim castellation,<br />
False Indigo (3AS285); e. body sherd with shoulder, False Indigo<br />
(3AS285); f. disk base, False Indigo (3AS285); g. base and wall, Water Elm<br />
3 (3AS287); h. body sherd, Buttonbush (3BR58); i. disk base, Buttonbush<br />
(3BR58). (AAS neg. 813876)<br />
d<br />
b<br />
g<br />
e
c<br />
A b<br />
Lithic and Ceramic Analyses 257<br />
Figure 67. Miscellaneous sherds from floodpla<strong>in</strong> sites. a. Coles Creek<br />
Incised rim, Eutaw Rapids (3BR75); b. shell-tempered <strong>in</strong>cised<br />
body, Jug Po<strong>in</strong>t 1 (3AS306); c, d. S<strong>in</strong>ner L<strong>in</strong>ear Punctated rim<br />
and neck, Water Elm 1 (3AS322). (AAS neg. 813877)<br />
d
258 Hemm<strong>in</strong>gs<br />
(probably this was a Caddoan idea), and to treat rims and bodies as different<br />
decorative fields. . .this is more than an occasional occurrence; it is a general characteristic<br />
<strong>of</strong> <strong>the</strong> Felsenthal region pottery <strong>of</strong> <strong>the</strong> Mississippi period.<br />
Schambach conv<strong>in</strong>c<strong>in</strong>gly shows why <strong>the</strong>se Mississippi period ceramics, when <strong>the</strong>y occur as<br />
sherds or as whole vessels, <strong>of</strong>ten cannot correctly be classified <strong>in</strong> <strong>the</strong> Lower Valley or Caddoan<br />
typologies (Suhm and Krieger 1954; Phillips 1970). He <strong>the</strong>n sets out <strong>in</strong> detail <strong>the</strong> basis<br />
for a new “descriptive classification system” for rim-and-body decorated ware, especially<br />
relevant to clay-tempered ceramics <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> phase.<br />
At this po<strong>in</strong>t we must refer <strong>the</strong> close reader to Schambach’s ceramic presentation <strong>in</strong><br />
<strong>the</strong> Shallow Lake site report (Rol<strong>in</strong>gson and Schambach 1981). However, three hierarchical<br />
concepts are necessary to classify sherds or vessels <strong>in</strong> <strong>the</strong> descriptive system, and <strong>the</strong>y are<br />
briefly summarized here as a basis for discussion <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> complex below:<br />
Class is a general category based on a major motif and decorative technique (e.g.,<br />
Class B <strong>in</strong>cludes rectil<strong>in</strong>ear or curvil<strong>in</strong>ear, <strong>in</strong>cised l<strong>in</strong>e designs placed horizontally,<br />
or composed <strong>of</strong> horizontal l<strong>in</strong>es). Five classes (lettered A through E) are<br />
def<strong>in</strong>ed by Schambach presently.<br />
Patterns are groups <strong>of</strong> similar designs with<strong>in</strong> classes, and are based on surface<br />
f<strong>in</strong>ish, l<strong>in</strong>e quality, and contrasts <strong>in</strong> decorative technique. Numerous patterns<br />
are dist<strong>in</strong>guished by Schambach with<strong>in</strong> Classes A to C, and are named “Afton,<br />
Anthony, Butler, Concord, Cornell,” etc.<br />
Designs are variations with<strong>in</strong> patterns, usually m<strong>in</strong>or variations or modes, such<br />
as lip notch<strong>in</strong>g, lip l<strong>in</strong>es, etc., and are numbered after <strong>the</strong> pattern name, e. g.,<br />
“Afton 1, Afton 2.”<br />
Also a notation system for sherds and vessels consists <strong>of</strong> rim pattern or design notation<br />
separated from body pattern or design notation by <strong>the</strong> symbol “::” (aga<strong>in</strong>, <strong>the</strong> reader should<br />
consult Rol<strong>in</strong>gson and Schambach 1981:106-131). Depend<strong>in</strong>g on completeness and preservation,<br />
all sherds can be descriptively classified at some level <strong>of</strong> specificity, for example, “Afton<br />
:: pla<strong>in</strong>, Butler 1 :: Coker 2, Butler 15 :: miss<strong>in</strong>g, miss<strong>in</strong>g :: Alma 3,” and so on.<br />
As <strong>the</strong> result <strong>of</strong> site survey <strong>in</strong> <strong>the</strong> Felsenthal Project area (Chapter 5) we recorded many<br />
small Gran <strong>Marais</strong> phase components, and have classified many small sherd samples from<br />
<strong>the</strong>se sites. One very important and relatively large collection was obta<strong>in</strong>ed from False Indigo<br />
(3AS285) on <strong>the</strong> Sal<strong>in</strong>e River. This site conta<strong>in</strong>s a 15 x 18 m artifact scatter (Area A) and<br />
small midden deposited dur<strong>in</strong>g a brief <strong>in</strong>terval with<strong>in</strong> <strong>the</strong> Gran <strong>Marais</strong> phase. The ceramic<br />
sample (1,130 identifiable sherds) from False Indigo, Area A, thus represents a “pure” as-
Lithic and Ceramic Analyses 259<br />
semblage which can be used to characterize, analyze, and compare Gran <strong>Marais</strong> complex<br />
pottery used <strong>in</strong> a floodpla<strong>in</strong> extractive site.<br />
Schambach’s analysis <strong>of</strong> Shallow Lake ceramics <strong>in</strong>cludes a sample from a structure<br />
(Unit X) underly<strong>in</strong>g Mound C, which has furnished an important basis for describ<strong>in</strong>g Gran<br />
<strong>Marais</strong> complex pottery used <strong>in</strong> a ceremonial site (Rol<strong>in</strong>gson and Schambach 1981:Table 20).<br />
This structure is well dated at about A.D. 1250-1300. In <strong>the</strong> follow<strong>in</strong>g analysis we implement<br />
Schambach’s descriptive classification system and compare Shallow Lake and False Indigo ceramics.<br />
We know that variability among <strong>the</strong> two samples could reflect sociological and functional<br />
differences between sites, and could also reflect temporal change <strong>in</strong> <strong>the</strong> Gran <strong>Marais</strong><br />
complex. The two samples are sufficiently large and “unmixed,” so far as we can determ<strong>in</strong>e,<br />
to furnish a valid comparative basis. They may be generally characterized as follows:<br />
Pla<strong>in</strong> Decorated<br />
Shallow Lake, Mound C, 1,148 (81.6%) 258 (18%)<br />
Structure Unit X<br />
(N=1,406)<br />
False Indigo, Area A 976 (86.4%) 154 (13%)<br />
(N=1,130)<br />
Tables 23 and 24 present <strong>the</strong> raw sherd counts <strong>in</strong> terms <strong>of</strong> <strong>the</strong> descriptive classification<br />
system. We note that <strong>the</strong>re is general similarity <strong>in</strong> <strong>the</strong> proportion <strong>of</strong> pla<strong>in</strong>ware and, by <strong>in</strong>ference,<br />
<strong>in</strong> vessel form among <strong>the</strong> two samples, except that polished pla<strong>in</strong>ware is not present<br />
at False Indigo. Among decorated wares <strong>the</strong>re is also general similarity, except that proportions<br />
<strong>of</strong> class vary as follows:<br />
Shallow Lake, Mound C, False Indigo,<br />
Structure Unit X Area A<br />
Class A 26.8% 24.2%<br />
Class A or B 29.4% 21.4%<br />
Class B 11.6% 29.2%<br />
Class C 31.8% 21.4%<br />
All o<strong>the</strong>r 0.4% 3.8%<br />
100.0% 100.0%<br />
The contrast is greatest for classes B and C (Class B <strong>in</strong>cludes horizontally placed, <strong>in</strong>cised<br />
elements and Class C <strong>in</strong>cludes punctated, p<strong>in</strong>ched, and notched elements). One <strong>in</strong>terpretation<br />
<strong>of</strong> this difference is that False Indigo is an earlier assemblage, reta<strong>in</strong><strong>in</strong>g more <strong>of</strong> <strong>the</strong><br />
Coles Creek decorative style, and that Shallow Lake, Mound C, is a later assemblage with<br />
<strong>in</strong>creas<strong>in</strong>g use <strong>of</strong> rim and body punctation. This <strong>in</strong>terpretation assumes that functional
260 Hemm<strong>in</strong>gs<br />
Table 23. Descriptive Classification <strong>of</strong> Pla<strong>in</strong>ware from Two Sites <strong>of</strong> <strong>the</strong> Gran<br />
<strong>Marais</strong> Phase, Mississippi Period.<br />
Shallow Lake (3UN9/52) False Indigo (AS285)<br />
Structure Unit X Area A<br />
Clay Tempered<br />
Pla<strong>in</strong> Rims 82 76<br />
Pla<strong>in</strong> Body Sherds 960 843<br />
Polished Pla<strong>in</strong> Body Sherds 85 0<br />
Lugs and Suspension Holes 2 1<br />
Car<strong>in</strong>ated Bases 4 0<br />
Convex Bases 1 9<br />
Flat Circular Bases 2 12<br />
Indeterm<strong>in</strong>ate Bases 9 0<br />
Shell Tempered<br />
Pla<strong>in</strong> Rims 0 2<br />
Pla<strong>in</strong> Body Sherds 3 38<br />
Bone Tempered<br />
Pla<strong>in</strong> Rims 0 0<br />
Pla<strong>in</strong> Body Sherds 0 1<br />
TOTALS 1,148 982
Lithic and Ceramic Analyses 261<br />
Table 24. Descriptive Classification <strong>of</strong> Decorated Sherds from Two Sites <strong>of</strong> <strong>the</strong><br />
Gran <strong>Marais</strong> Phase, Mississippi Period.<br />
Shallow Lake (3UN9/52), False Indigo (3AS285),<br />
Structure Unit X Area A<br />
Afton 1 :: miss<strong>in</strong>g 1 miss<strong>in</strong>g :: Adelphi 5<br />
Afton 2 :: miss<strong>in</strong>g 1 Afton :: miss<strong>in</strong>g 4<br />
Afton 10 :: miss<strong>in</strong>g 1 miss<strong>in</strong>g :: Alma 4 1<br />
miss<strong>in</strong>g :: Alma 3 1 miss<strong>in</strong>g :: Alma 3<br />
Anthony 1 :: pla<strong>in</strong> 2 Anthony 1 :: miss<strong>in</strong>g 2<br />
Anthony 1 :: miss<strong>in</strong>g 2 Anthony :: class A 1<br />
Anthony 3 :: miss<strong>in</strong>g 2 Anthony :: miss<strong>in</strong>g 1<br />
Anthony 4 :: miss<strong>in</strong>g 1 miss<strong>in</strong>g :: Ashland 1<br />
Anthony :: miss<strong>in</strong>g 11<br />
Class A rims/bodies 47 Class A rims/bodies 19<br />
Class A or B rims/bodies 75 Class A or B bodies 33<br />
Boise 10 :: miss<strong>in</strong>g 1 Barnard 3 :: miss<strong>in</strong>g 1<br />
Boise :: miss<strong>in</strong>g 2 Barnard 4 :: miss<strong>in</strong>g 2<br />
Butler 1 or 2 :: pla<strong>in</strong> 1 Barnard :: pla<strong>in</strong> 3<br />
Butler 1 or 2 :: Coker 1 Barnard :: miss<strong>in</strong>g 3<br />
Butler 2 :: miss<strong>in</strong>g 1 Barr<strong>in</strong>gton :: miss<strong>in</strong>g 2<br />
Butler 3 or 5 :: pla<strong>in</strong> 1 Bates :: miss<strong>in</strong>g 1<br />
Butler 3 or 5 :: Coker 1 miss<strong>in</strong>g :: Buffalo 1<br />
Butler 3 or 5 :: miss<strong>in</strong>g 3 Butler 1 :: Coker 2 2<br />
Butler 3 or <strong>17</strong> :: miss<strong>in</strong>g 1 Butler 1 :: miss<strong>in</strong>g 7<br />
Butler 15 :: miss<strong>in</strong>g 1 Butler 4 :: miss<strong>in</strong>g 4<br />
Butler 1, 8, or 9 :: miss<strong>in</strong>g 10 Butler <strong>17</strong> :: miss<strong>in</strong>g 2<br />
Butler :: miss<strong>in</strong>g 6 Butler :: pla<strong>in</strong> 1<br />
Butler :: miss<strong>in</strong>g 2<br />
Class B rims/bodies 1 Class B rims/bodies 14<br />
Class B or C rims/bodies 2<br />
Canisius :: or :: Creston 1 Cameron 2 :: miss<strong>in</strong>g 2<br />
miss<strong>in</strong>g :: Casper 2 miss<strong>in</strong>g :: Caney 10<br />
pla<strong>in</strong> :: Castleton 2 1 miss<strong>in</strong>g :: Carver 1<br />
:: Catawba or Caldwell :: 3 Cedar 2 :: miss<strong>in</strong>g 1<br />
Cisco 5 :: miss<strong>in</strong>g 1 Cedar :: miss<strong>in</strong>g 1<br />
Cisco :: or :: Clafl<strong>in</strong> 1 miss<strong>in</strong>g :: Coker 1<br />
(cont<strong>in</strong>ued)
262 Hemm<strong>in</strong>gs<br />
Table 24. (concluded) Descriptive Classification <strong>of</strong> Decorated Sherds from Two<br />
Sites <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> Phase, Mississippi Period.<br />
Shallow Lake (3UN9/52), False Indigo (3AS285),<br />
Structure Unit X Area A<br />
miss<strong>in</strong>g :: Clafl<strong>in</strong> 2 Concord 6 :: miss<strong>in</strong>g 1<br />
miss<strong>in</strong>g :: Coker 1 3 Cornell 2 :: pla<strong>in</strong> 1<br />
miss<strong>in</strong>g :: Coker 4 1 Cornell :: pla<strong>in</strong> 1<br />
miss<strong>in</strong>g :: Coker 5 1<br />
miss<strong>in</strong>g :: Coker 6<br />
Concord :: or :: Coker 30<br />
Concord :: miss<strong>in</strong>g 1<br />
cornell 1 :: pla<strong>in</strong> 8<br />
Cornell 2 :: pla<strong>in</strong> 9<br />
cornell 3 :: pla<strong>in</strong> 1<br />
Culver :: miss<strong>in</strong>g 1<br />
miss<strong>in</strong>g :: Curry 2 1<br />
miss<strong>in</strong>g :: Curry 3<br />
Curtis 1 :: miss<strong>in</strong>g 1<br />
Curtis 2 :: miss<strong>in</strong>g 2<br />
Curtis 3 :: miss<strong>in</strong>g 2<br />
Class C rims/bodies 1 Class C rims/bodies 14<br />
Class D rims/bodies 1 Class D rims/bodies 2<br />
Class E rims/bodies 1 Class E rims/bodies 2<br />
TOTALS 258 154
Lithic and Ceramic Analyses 263<br />
differences among contemporary assemblages is not <strong>the</strong> significant factor, but <strong>of</strong> course we<br />
do not know <strong>the</strong> truth <strong>of</strong> this statement. A reliable set <strong>of</strong> dates for False Indigo occupation<br />
could test this <strong>in</strong>terpretation <strong>of</strong> ceramic differences. At this po<strong>in</strong>t we regard <strong>the</strong> variation as<br />
likely to reflect both functional and temporal differences between ceramic samples.<br />
S<strong>in</strong>ce some rim-and-body classifications designated by Schambach can be translated<br />
<strong>in</strong>to types and varieties, we will also attempt to compare assemblages on this basis. The follow<strong>in</strong>g<br />
list <strong>of</strong> occurrence is m<strong>in</strong>imal, and more complete sherds or vessels would undoubtedly<br />
add a few o<strong>the</strong>r types and varieties (SL=present or probably present <strong>in</strong> Shallow Lake,<br />
Mound C, Structure Unit X, sample; FI=present or probably present <strong>in</strong> False Indigo, Area A,<br />
sample):<br />
Avoyelles Punctated, var. Wolf Slough SL<br />
Baytown Pla<strong>in</strong>, var. unspecified SL, FI<br />
Baytown Pla<strong>in</strong>, var. Shallow Lake SL, FI<br />
Coles Creek Incised, var. Kemp SL<br />
Evansville Punctated, var. P<strong>in</strong>e Island SL, FI<br />
Harrison Bayou Incised, var. Strong SL<br />
Kiam Incised, var. Redeye Lake SL, FI<br />
Kiam Incised, var. Sowell SL, FI<br />
Lapile Incised, var. unspecified SL<br />
Pargoud Incised, var. Monroe FI<br />
This list comprises a large part <strong>of</strong> <strong>the</strong> <strong>in</strong>digenous pottery types now known to characterize<br />
<strong>the</strong> Gran <strong>Marais</strong> complex <strong>in</strong> <strong>the</strong> Felsenthal region. Figures 68 and 69 illustrate decorated<br />
sherds <strong>of</strong> <strong>the</strong> Gran <strong>Marais</strong> complex from False Indigo, Area A. Baytown Pla<strong>in</strong> sherds<br />
from this site are illustrated <strong>in</strong> Figure 66.<br />
Shell-tempered Complex<br />
Shell-tempered ceramics were undoubtedly present throughout <strong>the</strong> Mississippi period<br />
<strong>in</strong> <strong>the</strong> Felsenthal region, but <strong>the</strong> use <strong>of</strong> crushed shell prevailed, and eventually nearly<br />
excluded clay temper, dur<strong>in</strong>g <strong>the</strong> Caney Bayou phase (A.D. 1400-1600). Table 25 records this<br />
transition from <strong>the</strong> standpo<strong>in</strong>t <strong>of</strong> ceramic samples obta<strong>in</strong>ed <strong>in</strong> small floodpla<strong>in</strong> extractive<br />
sites, which appear <strong>in</strong> presumed chronological order from early to late Mississippi period<br />
(see also Chapter 6).<br />
Although we have many sherd samples perta<strong>in</strong><strong>in</strong>g to <strong>the</strong> Caney Bayou phase, and<br />
one perta<strong>in</strong><strong>in</strong>g to term<strong>in</strong>al Mississippi period occupation at One Cypress Po<strong>in</strong>t (3AS286),<br />
<strong>the</strong>se samples conta<strong>in</strong> only small, wea<strong>the</strong>red, and <strong>of</strong>ten leached and exfoliated, sherds. We<br />
<strong>the</strong>refore cannot characterize shell-tempered ceramics except to say that pla<strong>in</strong> and <strong>in</strong>cised<br />
vessels were present <strong>in</strong> floodpla<strong>in</strong> extractive sites, especially those occupied after about A.D.<br />
1400. Schambach def<strong>in</strong>es and illustrates <strong>the</strong> Caney Bayou ceramic complex at Shallow Lake<br />
(Rol<strong>in</strong>gson and Schambach 1981:193-195). Indigenous pla<strong>in</strong> and decorated ceramics, as well
264 Hemm<strong>in</strong>gs<br />
Figure 68. Decorated rim sherds, Gran <strong>Marais</strong> ceramic complex, False Indigo (3AS285) (AAS neg. 813879).
Lithic and Ceramic Analyses 265<br />
Figure 69. Decorated body sherds, Gran <strong>Marais</strong> ceramic complex, False Indigo (3AS285) (AAS neg. 813881).
266 Hemm<strong>in</strong>gs<br />
as various Caddoan types and W<strong>in</strong>terville Incised, are well represented <strong>in</strong> this complex. Figure<br />
67b illustrates one relatively well preserved shell-tempered body sherd from Jug Po<strong>in</strong>t 1<br />
(3AS306). This sherd appears to represent a vessel with diagonally opposed, <strong>in</strong>cised l<strong>in</strong>es on<br />
<strong>the</strong> body just below <strong>the</strong> rim panel.<br />
Table 25. Proportions <strong>of</strong> Shell and Clay Temper<strong>in</strong>g <strong>in</strong> Ceramic Samples from Four<br />
Floodpla<strong>in</strong> Sites <strong>of</strong> <strong>the</strong> Mississippi Period.<br />
Archeological Sample Shell Clay<br />
Site Name and Designation Phase Size Temper Temper<br />
Jug Po<strong>in</strong>t 1 (3AS306) Caney Bayou (late) 62 98.4% 1.6%<br />
Jug Po<strong>in</strong>t 2 (3AS307) Caney Bayou (early) 98 74.5% 25.5%<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76) Gran <strong>Marais</strong> (late) 250 22.8% 76.0%<br />
False Indigo, Area A (3AS285) Gran <strong>Marais</strong> (early) 1130 4.0% 95.9%<br />
Caddoan Imports<br />
Caddoan vessels (or sherds) which can most likely be attributed to manufacture<br />
outside <strong>the</strong> Felsenthal region or to manufacture by ethnic Caddoan potters are evidently<br />
present at Shallow Lake and o<strong>the</strong>r upland habitation sites. These ceramics <strong>in</strong>clude various<br />
polished, engraved, pseudol<strong>in</strong>ear punctated, stamped, and brushed types. Sherds from such<br />
vessels are exceed<strong>in</strong>gly rare <strong>in</strong> our floodpla<strong>in</strong> site samples, and isolated small sherds are<br />
difficult to identify or dist<strong>in</strong>guish from local wares. Figure 67c, d illustrates rim and body<br />
sherds from a s<strong>in</strong>gle vessel <strong>of</strong> S<strong>in</strong>ner L<strong>in</strong>ear Punctated, var. S<strong>in</strong>ner, recovered at Water Elm<br />
1 (3AS322) on <strong>the</strong> Ouachita River. The sherds were erod<strong>in</strong>g out <strong>of</strong> <strong>the</strong> riverbank at about 40<br />
cm depth and were not associated with o<strong>the</strong>r artifacts or debris. It is possible that Water Elm<br />
1 represents a Caddo Indian transient camp or bivouac on <strong>the</strong> riverbank, but Schambach’s<br />
comment on <strong>the</strong> frequency <strong>of</strong> var. S<strong>in</strong>ner <strong>in</strong>dicates that it may also be locally manufactured<br />
(Rol<strong>in</strong>gson and Schambach 1981:<strong>17</strong>0).<br />
culture historicAl hypo<strong>the</strong>ses<br />
Analysis <strong>of</strong> artifacts from site survey and test excavation has provided data relevant to<br />
two hypo<strong>the</strong>ses stated <strong>in</strong> <strong>the</strong> research design. In general we have not obta<strong>in</strong>ed <strong>the</strong> quantity<br />
or quality <strong>of</strong> data which could produce a rigorous test <strong>of</strong> <strong>the</strong>se hypo<strong>the</strong>ses, and we shall<br />
advance only a prelim<strong>in</strong>ary statement and discussion <strong>of</strong> possibilities. The culture historical<br />
hypo<strong>the</strong>ses were framed as follows:
Problem Doma<strong>in</strong> 3: Culture History<br />
Lithic and Ceramic Analyses 267<br />
H 3 If upland mound complexes represent a dom<strong>in</strong>ant regional culture, numerous<br />
sites <strong>of</strong> this culture should also be identified <strong>in</strong> <strong>the</strong> lowland sett<strong>in</strong>g.<br />
Test Implications: Discovery <strong>of</strong> culturally related sites <strong>in</strong> study area.<br />
Data Requirements: Common artifact styles, especially ceramics.<br />
H 4 If aborig<strong>in</strong>al Felsenthal populations occupied a frontier region, local styles <strong>of</strong><br />
artifacts and settlement rema<strong>in</strong>s will occur mixed with extraregional styles.<br />
Test Implications: Recognition <strong>of</strong> local and alien styles.<br />
Data Requirements: Typological and comparative analyses disclos<strong>in</strong>g <strong>in</strong>troduced<br />
styles, especially <strong>in</strong> ceramics.<br />
Hypo<strong>the</strong>sis 3 seeks to provide <strong>the</strong> physical l<strong>in</strong>k between upland population centers<br />
and transient floodpla<strong>in</strong> camps through shared artifact styles. We have <strong>in</strong> fact identified<br />
numerous “culturally related sites” on <strong>the</strong> basis <strong>of</strong> ceramics primarily, but also on <strong>the</strong> basis<br />
<strong>of</strong> projectile po<strong>in</strong>ts and lithic raw materials. Our comparisons are almost always constra<strong>in</strong>ed<br />
by <strong>the</strong> small, and sometimes wea<strong>the</strong>red, artifact samples furnished by floodpla<strong>in</strong> sites and<br />
by <strong>the</strong> dearth <strong>of</strong> controlled and detailed artifact data for upland sites <strong>in</strong> <strong>the</strong> region. In <strong>the</strong><br />
preced<strong>in</strong>g section <strong>of</strong> this chapter a detailed ceramic comparison was made for Gran <strong>Marais</strong><br />
phase assemblages at Shallow Lake (3UN9/52) and False Indigo (3AS285). These assemblages<br />
are markedly similar and must derive from a s<strong>in</strong>gle decorative style, and probably<br />
also from a s<strong>in</strong>gle technological tradition. Stylistic similarities are present among simple elements<br />
or techniques and also among complex designs and patterns. M<strong>in</strong>or variation is likely<br />
to reflect functional or temporal differences among <strong>the</strong>se dissimilar site units, and behavioral<br />
differences among social units.<br />
Although we cannot prove Hypo<strong>the</strong>sis 3 <strong>in</strong> any formal sense, it would seem to be well<br />
supported by available data. We could easily envision <strong>the</strong> work party, which deposited its<br />
ceramic waste <strong>in</strong> <strong>the</strong> False Indigo midden, actually be<strong>in</strong>g deployed from <strong>the</strong> Shallow Lake<br />
center or its satellite communities. Ceramic analysis has not achieved <strong>the</strong> level <strong>of</strong> analytical<br />
sophistication which could justify such a statement, but perhaps that level is atta<strong>in</strong>able<br />
through stylistic and technological approaches now be<strong>in</strong>g developed.
268 Hemm<strong>in</strong>gs<br />
Hypo<strong>the</strong>sis 4 also relies on stylistic analysis <strong>of</strong> ceramics and o<strong>the</strong>r rema<strong>in</strong>s to determ<strong>in</strong>e<br />
whe<strong>the</strong>r settlements were characteristic <strong>of</strong> a frontier region. Floodpla<strong>in</strong> sites studied<br />
by us do not appear to furnish significant and relevant test data. If <strong>the</strong> broken l<strong>in</strong>ear punctated<br />
vessel from Water Elm 1 (3AS322) is <strong>in</strong> fact an “alien” Caddo artifact, no <strong>in</strong>digenous<br />
materials were associated or “mixed” with it. O<strong>the</strong>r examples <strong>of</strong> imports are too few and<br />
too uncerta<strong>in</strong>. We believe that upland sites <strong>of</strong> several types may furnish evidence <strong>of</strong> frontier<br />
character when more detailed analytical work is accomplished <strong>in</strong> this region.
Chapter 9<br />
huMAn AdAptAtion <strong>in</strong> <strong>the</strong> grAnd MArAis lowlAnd<br />
This conclud<strong>in</strong>g chapter does not attempt to summarize all important archeological<br />
and historical results <strong>of</strong> <strong>the</strong> Felsenthal Project; it does expand and discuss two topics <strong>of</strong> cont<strong>in</strong>u<strong>in</strong>g<br />
research <strong>in</strong>terest. From development <strong>of</strong> research design through all successive stages<br />
<strong>of</strong> work, we emphasized <strong>the</strong> potential and distribution <strong>of</strong> floodpla<strong>in</strong> resource and <strong>the</strong> physical<br />
and spatial record <strong>of</strong> prehistoric human activity <strong>in</strong> this overflow bottomland (extractive<br />
sites <strong>in</strong> <strong>the</strong>ir environmental context).<br />
Our conclud<strong>in</strong>g comments deal first with <strong>the</strong> archeological and systemic qualities <strong>of</strong><br />
extractive sites and <strong>the</strong>n with a hypo<strong>the</strong>tical seasonal model for resource use <strong>in</strong> an overflow<br />
bottomland environment. These observations stem from regional environmental data (<strong>the</strong><br />
<strong>Grand</strong> <strong>Marais</strong> Lowland and its floodpla<strong>in</strong>) and from our most complete site data (specialized<br />
extractive sites <strong>of</strong> <strong>the</strong> Mississippi period, A.D. 1100-<strong>17</strong>00), but <strong>the</strong>re are broad implications<br />
for subsistence-settlement studies <strong>in</strong> all forested river valleys <strong>of</strong> <strong>the</strong> eastern United<br />
States and <strong>in</strong> all periods.<br />
on <strong>the</strong> nAture <strong>of</strong> extrActiVe sites<br />
While a concern with large multiple activity or complex sites has prevailed <strong>in</strong> American<br />
archeology, many students <strong>of</strong> settlement systems dist<strong>in</strong>guish a class, occasionally a functionally<br />
differentiated series, <strong>of</strong> small limited activity sites. Small site studies frequently, but<br />
not always, perta<strong>in</strong> to hunter-ga<strong>the</strong>rer cultural systems (Hayden 1965; W<strong>in</strong>ters 1969; Moseley<br />
and Mackey 1972; Dillehay 1973; Talmage and Chester 1977; Ward 1978; B<strong>in</strong>ford 1980).<br />
Table 26 compares <strong>the</strong> term<strong>in</strong>ology and site characteristics developed <strong>in</strong> three relevant studies;<br />
hierarchical classifications <strong>of</strong> Mississippian settlement types were omitted here because<br />
<strong>the</strong>y generally emphasize complex centers and large habitation sites, group<strong>in</strong>g all small site<br />
units as “farmsteads” or “camps.” We noted examples <strong>of</strong> this emphasis on small Mississippi<br />
period sites <strong>in</strong> Chapter 6 (e.g., Muller 1978:283). Kl<strong>in</strong>ger (1975:51) at least recognizes <strong>the</strong> potential<br />
<strong>of</strong> such sites for <strong>the</strong> Park<strong>in</strong> phase (ca A.D. 1400-1600) <strong>in</strong> nor<strong>the</strong>ast <strong>Arkansas</strong>: “<strong>the</strong> fifth<br />
order <strong>of</strong> settlement is composed <strong>of</strong> small hamlets and farmsteads.... O<strong>the</strong>r activity loci such<br />
as quarry sites, butcher<strong>in</strong>g stations or overnight campsites would be <strong>in</strong>cluded [but none]<br />
have been recognized <strong>in</strong> <strong>the</strong> St. Francis Bas<strong>in</strong>.” It is <strong>of</strong> <strong>in</strong>terest to note that a recent <strong>in</strong>tensive<br />
survey <strong>of</strong> a 1 km wide area around <strong>the</strong> Park<strong>in</strong> site on <strong>the</strong> St. Francis River did not confirm<br />
<strong>the</strong> existence <strong>of</strong> fifth order sites as proposed (see Morse 1981:41).
270 Hemm<strong>in</strong>gs<br />
Table 26. Three Classifications <strong>of</strong> Settlement Type which Emphasize or Dist<strong>in</strong>guish Small<br />
Extractive Sites.<br />
W<strong>in</strong>ters (1969): Wabash Valley Fitzhugh (1972): Coastal Labrador B<strong>in</strong>ford (1980): Collectors/Foragers<br />
1. Settlement–2-3 acres, per- 1. Ga<strong>the</strong>r<strong>in</strong>g Site–large size; sea- 1. Residential Base–one <strong>of</strong> multiple camps<br />
manent houses, storage pits, sonally reoccupied by band. occupied annually by mobile forag<strong>in</strong>g or<br />
burials, many fabricat<strong>in</strong>g, collect<strong>in</strong>g group, “archeological visibility”<br />
process<strong>in</strong>g, and domestic tools 2. Base Camp–smaller size; occupied varies with group size, duration <strong>of</strong> stay<br />
by several families for extended and reoccupation <strong>of</strong> site; process<strong>in</strong>g,<br />
2. Transient Camp–2-3 acres; period <strong>in</strong> a resource area. manufactur<strong>in</strong>g, and ma<strong>in</strong>tenance activifew<br />
or no houses, storage pits ties; daily forag<strong>in</strong>g orig<strong>in</strong>ates here.<br />
or burials; many weapons and 3. Exploitation Camp, <strong>in</strong>tensive–a<br />
process<strong>in</strong>g tools. s<strong>in</strong>gle dwell<strong>in</strong>g occupied by family 2. location–a place where extractive tasks<br />
or extended family dur<strong>in</strong>g several are exclusively carried out by foragers or<br />
3. Base Camp–2 acres; no weeks <strong>of</strong> hunt<strong>in</strong>g or fish<strong>in</strong>g. collectors; “archeological visibility” is low<br />
houses, few storage pits because <strong>of</strong> low bulk procurement, brief<br />
or burials; many weapons 4. Exploitation Camp, light–s<strong>in</strong>gle site use, <strong>in</strong>frequent reoccupation; expediand<br />
fabricat<strong>in</strong>g tools. dwell<strong>in</strong>g or hearth attributed to ent extractive tools obta<strong>in</strong>ed and abana<br />
family briefly engaged <strong>in</strong> food doned.<br />
4. Hunt<strong>in</strong>g Camp–small area; ga<strong>the</strong>r<strong>in</strong>g or hunt<strong>in</strong>g.<br />
only weapons and process<strong>in</strong>g 3. Field Camp–temporary camp occupied<br />
tools. 5. Bivouac–transient camp with m<strong>in</strong>i- by collectors’ task group to procure and<br />
mal activities; few structures or process particular resources; “archeo-<br />
5. Ga<strong>the</strong>r<strong>in</strong>g Camp–small area; tools. logical visibility” may be high because <strong>of</strong><br />
process<strong>in</strong>g tools. bulk procurement and reoccupation <strong>of</strong><br />
6. Specialized Camp, <strong>in</strong>ternal– preferred location.<br />
6. Bivouac–about .01 acres or limited activity site with<strong>in</strong> band<br />
less; debitage territory. 4. station–<strong>in</strong>formation ga<strong>the</strong>r<strong>in</strong>g po<strong>in</strong>ts,<br />
such as hunt<strong>in</strong>g stand, used by logistically<br />
7. Specialized Camp, external– organized collectors’ task group.<br />
trad<strong>in</strong>g or procurement sites<br />
outside band territory. 5. cache–temporary field storage facility<br />
used by collectors’ task group.
<strong>Human</strong> <strong>Adaptation</strong> 271<br />
We th<strong>in</strong>k B<strong>in</strong>ford’s (1980:10) discussion <strong>of</strong> “logistically organized” resource procurement<br />
strategies has relevance to Mississippi period subsistence and settlement <strong>in</strong> our study<br />
area (even though he generalizes from <strong>the</strong> Nuniamut Eskimo).<br />
Logistical strategies...are accomodations to <strong>the</strong> situation where consumers are<br />
near to one critical resource but far from ano<strong>the</strong>r equally critical resource.... Special<br />
task groups may leave a residential location and establish a field camp or station<br />
from which food procurement operations may be planned and executed....<br />
The obta<strong>in</strong>ed food may be field processed to facilitate transport and <strong>the</strong>n moved<br />
to <strong>the</strong> consumers....Logistically organized task groups are generally small and<br />
composed <strong>of</strong> skilled and knowledgeable <strong>in</strong>dividuals.<br />
This is a fair description, we th<strong>in</strong>k, <strong>of</strong> <strong>the</strong> Mississippi period subsistence-settlement system<br />
described from site survey data <strong>in</strong> Chapter 5. In our system a peripheral r<strong>in</strong>g <strong>of</strong> mound-village<br />
complexes adjo<strong>in</strong>s one critical resource—farmland on upland terraces—and is distant<br />
from ano<strong>the</strong>r—fish and o<strong>the</strong>r aquatic/riparian species <strong>in</strong> <strong>the</strong> <strong>in</strong>terior <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong>. We<br />
called this a “centripetal pattern,” and we th<strong>in</strong>k “logistically organized” task groups operated<br />
<strong>the</strong> Mississippi period extractive sites (“field camps” and “locations” <strong>in</strong> Table 28) concentrated<br />
on <strong>the</strong> lower Sal<strong>in</strong>e River.<br />
If we distill from our experience with floodpla<strong>in</strong> sites and from various studies cited<br />
earlier, we can suggest a series <strong>of</strong> qualities which should dist<strong>in</strong>guish extractive sites from<br />
o<strong>the</strong>r settlement types, and fur<strong>the</strong>r, should dist<strong>in</strong>guish variations among extractive sites<br />
which may occur <strong>in</strong> <strong>the</strong> archeological record <strong>of</strong> a region. Our presentation will provide<br />
<strong>in</strong>formation about <strong>the</strong> archeological context, i.e.,, “artifacts, featues, and residues...no longer<br />
participat<strong>in</strong>g <strong>in</strong> a behavioral system,” and <strong>the</strong> systemic context, <strong>the</strong> past <strong>in</strong>ferred behavioral<br />
system which contributed to this observed array <strong>of</strong> materials (Schiffer 1976:28). In o<strong>the</strong>r<br />
words, we will attempt to establish some correlates <strong>of</strong> human behavior on extractive sites<br />
(but see Schiffer [1976] and Wood and Johnson [1978] on site disturbance processes).<br />
1. Site size. Although we frequently refer to “small” extractive sites, one might expect<br />
considerable variation <strong>in</strong> this quality, depend<strong>in</strong>g on task group size, nature <strong>of</strong> extractive,<br />
process<strong>in</strong>g, and ma<strong>in</strong>tenance activities, nature <strong>of</strong> discard behavior, duration <strong>of</strong> occupation,<br />
“redundancy” <strong>of</strong> site use, and o<strong>the</strong>r factors. In our sample <strong>of</strong> floodpla<strong>in</strong> extractive sites,<br />
which admittedly are not well known <strong>in</strong> regard to extent because <strong>of</strong> site burial, area <strong>of</strong> artifact<br />
and debris scatter ranged from 1-900 m 2 and maximum l<strong>in</strong>ear dimensions rarely exceeded<br />
60 m. (About half <strong>of</strong> our floodpla<strong>in</strong> sites had maximum l<strong>in</strong>ear dimensions less than 20<br />
m.) These sites are a fraction <strong>of</strong> a hectare and should be dist<strong>in</strong>guishable from “hamlets” and<br />
“farmsteads” on size alone.<br />
Without a doubt much more extensive extractive sites occur <strong>in</strong> <strong>the</strong> archeological record,<br />
e.g., Ouachita Mounta<strong>in</strong> novaculite quarries studied by Baker (1974, 1975) and a mass<br />
bison kill studied by Wheat (1972). The first case presumably results chiefly from redundant
272 Hemm<strong>in</strong>gs<br />
site use, but <strong>the</strong> latter represents a s<strong>in</strong>gle event where extractive activity featured high bulk<br />
procurement.<br />
Correlates: Extractive site size relates to multiple variables; <strong>in</strong> <strong>the</strong> case <strong>of</strong> discrete activity<br />
loci, size, number <strong>of</strong> personnel, and <strong>in</strong>tensity <strong>of</strong> task-related behavior should be positively<br />
correlated.<br />
2. Site Configuration. In very general terms, an extractive site is observed archeologically<br />
as a cont<strong>in</strong>uous distribution <strong>of</strong> artifacts and rema<strong>in</strong>s hav<strong>in</strong>g some outer limit which<br />
def<strong>in</strong>es its shape, or as an apparent cluster <strong>of</strong> artifact scatters (loci) arranged <strong>in</strong> some configuration.<br />
Loci <strong>the</strong>mselves may vary <strong>in</strong> shape. Aga<strong>in</strong>, considerable variation <strong>in</strong> configuration<br />
is to be expected, depend<strong>in</strong>g on task group size and composition, nature <strong>of</strong> target<br />
resources, differentiation or repetition <strong>of</strong> tasks, nature <strong>of</strong> discard behavior, redundancy <strong>of</strong><br />
site use, terra<strong>in</strong> characteristics, and o<strong>the</strong>r factors. In our floodpla<strong>in</strong> site sample, many extractive<br />
sites have pronounced l<strong>in</strong>ear configuration, whe<strong>the</strong>r observed as cont<strong>in</strong>uous scatters or<br />
as successive discrete loci. These configurations are congruent with narrow levees and river<br />
bankl<strong>in</strong>es, and imply that performance <strong>of</strong> tasks and discard <strong>of</strong> wastes were strongly determ<strong>in</strong>ed<br />
by terra<strong>in</strong> characteristics. Discrete l<strong>in</strong>ear scatters observed by us may also frequently<br />
result from s<strong>in</strong>gle brief episodes <strong>of</strong> occupation, s<strong>in</strong>ce artifact and debris density was light. In<br />
contrast, we noted compact oval or more irregular riverbank artifact scatters with centrally<br />
located midden lenses at 3AS285 and 3BR76, where bulk fish process<strong>in</strong>g tasks and more<br />
prolonged site occupation are <strong>in</strong>ferred. At this po<strong>in</strong>t, we might suggest that defensibility<br />
could also affect extractive site configuration (if we are correct about sites 3AS285 and 3BR76<br />
as bulk facilities, <strong>the</strong>y might require protection).<br />
O<strong>the</strong>r variations <strong>in</strong> extractive site configuration are likely to be many (e.g., fishweirs,<br />
quarries, salt-mak<strong>in</strong>g stations, and butcher<strong>in</strong>g sites) and we will not attempt to expand our<br />
discussion <strong>of</strong> this quality.<br />
Correlates: Extractive site configuration relates to many variables; specific configurations<br />
<strong>of</strong> discrete loci should correlate with specific tasks, e.g., fan-shaped and o<strong>the</strong>r arrays <strong>of</strong><br />
debris left by a fl<strong>in</strong>tworker (Callahan 1980; Newcomer and Sievek<strong>in</strong>g 1980).<br />
3. Site Assemblage. Assemblage here refers to tools and debris from toolmak<strong>in</strong>g,<br />
refurbish<strong>in</strong>g, or recycl<strong>in</strong>g, and also facilities for extractive or process<strong>in</strong>g tasks, which persist<br />
<strong>in</strong> archeological context. Extractive site assemblages can vary greatly <strong>in</strong> quantity <strong>of</strong> such<br />
materials and <strong>in</strong> proportions <strong>of</strong> various functional categories <strong>of</strong> material. This variation is<br />
affected by size and composition <strong>of</strong> <strong>the</strong> task group, nature <strong>of</strong> target resources, <strong>in</strong>tensity <strong>of</strong><br />
extractive, process<strong>in</strong>g, or ma<strong>in</strong>tenance activity, duration <strong>of</strong> occupation, redundancy <strong>of</strong> site<br />
use, discard and abandonment behavior, seasonality, and o<strong>the</strong>r factors. In our floodpla<strong>in</strong> site<br />
sample, substantial permanent structures for habitation, storage, and burial appeared to be<br />
absent, but o<strong>the</strong>r task-related or ma<strong>in</strong>tenance facilities were occasionally identified (hearths,
<strong>Human</strong> <strong>Adaptation</strong> 273<br />
pits, floors, post structures?). These characteristics may be associated with a “f<strong>in</strong>e gra<strong>in</strong>ed<br />
assemblage,” or <strong>the</strong> rema<strong>in</strong>s <strong>of</strong> specific events that “accumulated over a short period <strong>of</strong> time,<br />
for example, a two-day camp” (B<strong>in</strong>ford 1980:<strong>17</strong>). Also, <strong>the</strong> numbers <strong>of</strong> tools and task-related<br />
items <strong>in</strong> our sample ranged from one to about 5,000 items. In Chapter 6 (Site Functional Hypo<strong>the</strong>ses)<br />
we noted <strong>the</strong> small size <strong>of</strong> assemblages recovered from sites tested, and also lack<br />
<strong>of</strong> diversity <strong>in</strong> <strong>the</strong>se assemblages (high ratio <strong>of</strong> extractive/ma<strong>in</strong>tenance tool kits). The floodpla<strong>in</strong><br />
sites we studied may be highly biased toward small task groups, specificity <strong>of</strong> target<br />
resources, and transient site use dur<strong>in</strong>g seasonal low river levels. These sites have unusually<br />
low “archeological visibility.”<br />
Very much larger, more <strong>in</strong>tensively used, and reoccupied extractive sites are well<br />
known <strong>in</strong> <strong>the</strong> archeological record <strong>of</strong> many regions, e.g., fl<strong>in</strong>t quarries and salt-mak<strong>in</strong>g stations.<br />
B<strong>in</strong>ford’s (1980:<strong>17</strong>) “coarse gra<strong>in</strong>ed” dist<strong>in</strong>ction applies to <strong>the</strong>se assemblages <strong>in</strong> <strong>the</strong><br />
sense that “<strong>the</strong> resolution between archeological rema<strong>in</strong>s and specific events is poor.”<br />
Correlates: Multiple variables determ<strong>in</strong>e <strong>the</strong> size and content <strong>of</strong> an extractive site<br />
assemblage; <strong>in</strong> f<strong>in</strong>e gra<strong>in</strong>ed assemblages differentiation <strong>in</strong> task group personnel and tasks<br />
performed should be directly related to assemblage variability (e.g., male vs female tasks,<br />
hunt<strong>in</strong>g vs fish<strong>in</strong>g activity, and so on).<br />
4. Site Residue. Under favorable conditions <strong>of</strong> preservation, <strong>the</strong> rema<strong>in</strong>s <strong>of</strong> extractive<br />
sites <strong>in</strong>clude organic or <strong>in</strong>organic residues which are <strong>the</strong> byproducts <strong>of</strong> extractive or process<strong>in</strong>g<br />
activity. These residues may be expected to vary with seasonality, target resources,<br />
process<strong>in</strong>g techniques, and discard behavior, among o<strong>the</strong>r factors. Target resources may<br />
be concentrated or diffuse <strong>in</strong> occurrence. Also, quantities <strong>of</strong> residue vary with <strong>the</strong> nature <strong>of</strong><br />
process<strong>in</strong>g tasks, e.g., where high bulk procurement strategies are employed.<br />
In our floodpla<strong>in</strong> site sample, residues were usually associated with burn<strong>in</strong>g. Many<br />
extractive, process<strong>in</strong>g, or ma<strong>in</strong>tenance tasks required heat or flame, and a wide range <strong>of</strong><br />
residues resulted (wood charcoal, ash, c<strong>in</strong>der, charred faunal and floral matter, heat-altered<br />
chert and novaculite, fire-cracked rock, and fired clay). Analysis <strong>of</strong> <strong>the</strong>se residues may furnish<br />
critical data for <strong>in</strong>terpret<strong>in</strong>g site function, s<strong>in</strong>ce <strong>the</strong>y are much less likely to be “curated”<br />
than tools or tool fragments. However, where target resources were foodstuffs, analysis<br />
<strong>of</strong> faunal and floral residues provides empirical evidence for site function.<br />
Correlates: Extractive site residues may vary <strong>in</strong> identity and quantity, but an adequate<br />
site sample should relate directly to target resources processed on that site and to procurement<br />
<strong>in</strong> low bulk or high bulk amounts.<br />
5. Site Location. The f<strong>in</strong>al and <strong>of</strong>ten most difficult quality <strong>of</strong> extractive sites to <strong>in</strong>terpret<br />
is location. Location <strong>in</strong>cludes <strong>the</strong> spatial relationship <strong>of</strong> an extractive site with its contemporary<br />
residential base and with its contemporary target resources or zone <strong>of</strong> resources.
274 Hemm<strong>in</strong>gs<br />
S<strong>in</strong>ce an extractive site may be a po<strong>in</strong>t <strong>of</strong> assembly and process<strong>in</strong>g <strong>of</strong> bulk resources for<br />
delivery to base settlements (Hypo<strong>the</strong>sis 2 <strong>in</strong> Chapter 5), access routes can be an important<br />
locational factor. O<strong>the</strong>r factors <strong>in</strong>clude <strong>the</strong> spac<strong>in</strong>g <strong>of</strong> contemporary base settlements<br />
and <strong>the</strong>ir satellite communities and extractive camps with<strong>in</strong> <strong>the</strong> region. In our study <strong>of</strong><br />
floodpla<strong>in</strong> sites, we had <strong>in</strong>sufficient data to establish contemporaneity among upland base<br />
settlements, lowland extractive site , and access routes (presumably <strong>the</strong> latter refers to river<br />
channels, bayous, and backswamp tributaries which are not static features <strong>of</strong> <strong>the</strong> landscape).<br />
As <strong>in</strong>dicated at <strong>the</strong> end <strong>of</strong> Chapter 5 (Figure 34) and earlier <strong>in</strong> this chapter, we have made a<br />
substantial advance <strong>in</strong> del<strong>in</strong>eat<strong>in</strong>g a “centripetal pattern” <strong>of</strong> Mississippi period settlement <strong>in</strong><br />
<strong>the</strong> Felsenthal Project area, but much work rema<strong>in</strong>s to be done here.<br />
One aspect <strong>of</strong> extractive site location has received little attention. Not only base settlements,<br />
but small seasonal camps <strong>the</strong>mselves, have catchment areas (Vita-F<strong>in</strong>zi and Higgs<br />
1970). These areas encompass resources which can be efficiently exploited from a central<br />
site locus; archeological models <strong>of</strong> site catchments are usually presented as circles <strong>of</strong> a few<br />
kilometers radius or as concentric r<strong>in</strong>gs. Flannery (1976:116) notes with regard to <strong>the</strong> early<br />
Mesoamerican village and its subsistence system that “one must consider not only its own<br />
immediate catchment area, but those <strong>of</strong> its associated seasonal camps.” We suggest that extractive<br />
site catchment analysis is a fruitful area for future work <strong>in</strong> our study area. However,<br />
where l<strong>in</strong>earity <strong>of</strong> floodpla<strong>in</strong> and meander belt features prevails and where canoe travel<br />
may be <strong>the</strong> most efficient means <strong>of</strong> access to resources, circular catchment areas may be<br />
totally <strong>in</strong>appropriate for extractive site analyses.<br />
Correlates: Extractive site location is related to multiple environmental and cultural<br />
variables; among o<strong>the</strong>r factors, location is highly related to <strong>the</strong> distribution <strong>of</strong> target resources.<br />
Once aga<strong>in</strong> we would affirm <strong>the</strong> research potential and quality <strong>of</strong> significance that may<br />
dist<strong>in</strong>guish small extractive sites. Our brief review <strong>of</strong> <strong>the</strong>ir attributes and potential as archeological<br />
resources leads to <strong>the</strong> follow<strong>in</strong>g conclusions.<br />
1. These sites may represent brief time <strong>in</strong>tervals or events and can contribute to studies<br />
emphasiz<strong>in</strong>g temporal order<strong>in</strong>g (seriation, cross-dat<strong>in</strong>g)<br />
2. Site assemblages may be culturally “pure” or unadulterated by redundant site use,<br />
and contribute to taxonomic studies (typology, stylistic analysis).<br />
3. Artifacts and o<strong>the</strong>r rema<strong>in</strong>s may perta<strong>in</strong> directly and exclusively to exploitative<br />
activity (environmental archeology, subsistence and settlement systems).
<strong>Human</strong> <strong>Adaptation</strong> 275<br />
4. Task group personnel may have been culturally or ethnically homogeneous (social<br />
organization, ethnohistory).<br />
5. The array <strong>of</strong> tools, facilities, and debris may perta<strong>in</strong> to specific activities and to<br />
discard and abandonment processes (behavioral archeology, functional analysis).<br />
6. Small extractive sites are underrepresented <strong>in</strong> current archeological studies (research<br />
design, sampl<strong>in</strong>g strategies).<br />
7. Small sites are manageable units <strong>in</strong> terms <strong>of</strong> cost for comprehensive data recovery<br />
or for alternative mitigation measures (cultural resource management studies).<br />
A Model for resource use <strong>in</strong> An oVerflow bottoM<br />
In order to understand <strong>the</strong> abundance and nature <strong>of</strong> archeological resources <strong>in</strong> <strong>the</strong><br />
Felsenthal Project area we have necessarily exam<strong>in</strong>ed <strong>the</strong> salient environmental characteristics<br />
<strong>of</strong> an overflow bottomland (Chapter 2). In this floodpla<strong>in</strong> environment, potential human<br />
food resources are predom<strong>in</strong>antly aquatic and riparian species, complexly distributed or<br />
concentrated <strong>in</strong> space and responsive to seasonal changes, especially <strong>the</strong> annual hydroperiod.<br />
Prehistoric Native Americans <strong>of</strong> this region developed and implemented subsistence<br />
strategies which were at least partly determ<strong>in</strong>ed by upland-lowland environmental contrast<br />
and by annual prolonged flood<strong>in</strong>g <strong>of</strong> <strong>the</strong> lowlands. Intensive archeological research <strong>in</strong> <strong>the</strong><br />
Felsenthal region has not yet adduced <strong>the</strong> subsistence-settlement system for any prehistoric<br />
culture or period (Chapter 3). Certa<strong>in</strong>ly regional substantive data for Mississippi period<br />
populations are approach<strong>in</strong>g <strong>the</strong> level needed to atta<strong>in</strong> this goal (Rol<strong>in</strong>gson and Schambach<br />
1981, and this report).<br />
In Figure 70 we present a general seasonal model for extractive activity <strong>in</strong> <strong>the</strong> <strong>Grand</strong><br />
<strong>Marais</strong> Lowland. This model perta<strong>in</strong>s most closely to <strong>the</strong> Mississippi period (A.D. 1100-<br />
<strong>17</strong>00), but may be applicable on a general level to earlier cultural periods when lowland<br />
environments were similar. Not only Felsenthal Project data, but also Comeaux’s (1972:<br />
Chapter 10) discussion <strong>of</strong> Indian and Euramerican “swamp life” <strong>in</strong> <strong>the</strong> Atchafalaya Bas<strong>in</strong>,<br />
south-central Louisiana, are <strong>in</strong>corporated <strong>in</strong> Figure 70. The Atchafalaya River overflows<br />
its banks <strong>in</strong> spr<strong>in</strong>g, <strong>in</strong>undat<strong>in</strong>g <strong>the</strong> entire bas<strong>in</strong>, and here fish<strong>in</strong>g has always dom<strong>in</strong>ated<br />
“swamp life.”<br />
Our model emphasizes <strong>the</strong> dual nature <strong>of</strong> floodpla<strong>in</strong> activity. Dur<strong>in</strong>g <strong>the</strong> prolonged<br />
w<strong>in</strong>ter-spr<strong>in</strong>g hydroperiod, fish<strong>in</strong>g occurs not only <strong>in</strong> deep channels or lakes, but “<strong>in</strong> <strong>the</strong><br />
woods” where concentrated forag<strong>in</strong>g and spawn<strong>in</strong>g species are available. Some trapp<strong>in</strong>g<br />
and collect<strong>in</strong>g activity focuses on <strong>the</strong> marg<strong>in</strong>s <strong>of</strong> <strong>the</strong> flooded bas<strong>in</strong>. When <strong>the</strong> floodpla<strong>in</strong><br />
emerges <strong>in</strong> summer-fall, fish<strong>in</strong>g cont<strong>in</strong>ues <strong>in</strong> permanent channels and lakes, but is supplemented<br />
markedly by hunt<strong>in</strong>g and collect<strong>in</strong>g activity. We noted earlier <strong>the</strong> potential for
276 Hemm<strong>in</strong>gs<br />
Figure 70. A seasonal model for subsistence activity and resource use <strong>in</strong> <strong>the</strong><br />
<strong>Grand</strong> <strong>Marais</strong> Lowland. The bold outer r<strong>in</strong>g represents <strong>the</strong> annual<br />
hydroperiod. The more labor <strong>in</strong>tensive and productive extractive<br />
activities are shown at <strong>in</strong>creas<strong>in</strong>g distance from <strong>the</strong> center.
<strong>Human</strong> <strong>Adaptation</strong> 277<br />
mass summer harvest<strong>in</strong>g <strong>of</strong> fish <strong>in</strong> dw<strong>in</strong>dl<strong>in</strong>g lakes and pools as floodwaters receded (Chapter<br />
2). The counterpart condition, ris<strong>in</strong>g floodwaters <strong>in</strong> late fall temporarily concentrated<br />
terrestrial animals on ridges and islands, where <strong>the</strong>y also were susceptible to mass capture<br />
techniques. It <strong>the</strong>refore seems likely that extractive activity <strong>in</strong>tensified dur<strong>in</strong>g <strong>the</strong>se episodes<br />
<strong>of</strong> ris<strong>in</strong>g and fall<strong>in</strong>g floodwaters.<br />
O<strong>the</strong>r times <strong>of</strong> <strong>in</strong>tensive activity would be determ<strong>in</strong>ed by seasonal migration, movement,<br />
rutt<strong>in</strong>g, spawn<strong>in</strong>g, feed<strong>in</strong>g, forag<strong>in</strong>g, and o<strong>the</strong>r behavior <strong>of</strong> important food animals<br />
and by ripen<strong>in</strong>g <strong>of</strong> floodpla<strong>in</strong> nut or seed crops. Buffalo fish<strong>in</strong>g <strong>in</strong> August may serve here as<br />
an example; an early historical account states ra<strong>the</strong>r colorfully, “I have seen <strong>the</strong> Sal<strong>in</strong>e river<br />
black with Buffalo fish <strong>in</strong> August, <strong>the</strong>ir float<strong>in</strong>g time, which [Jack Fogle, a Frenchman born<br />
<strong>in</strong> <strong>the</strong> area <strong>in</strong> 1818] would kill with a gig to feed his hogs on, and give us all we wanted...”<br />
(W. T. Mart<strong>in</strong> <strong>in</strong> E<strong>the</strong>ridge 1959:xviii).<br />
Although <strong>the</strong> floodpla<strong>in</strong> <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland had dist<strong>in</strong>ctive annual emergent<br />
(summer-fall) and <strong>in</strong>undated (w<strong>in</strong>ter-spr<strong>in</strong>g) phases, which could be nearly <strong>of</strong> equal<br />
length, it is clear from our model that extractive activity could be cont<strong>in</strong>uous throughout <strong>the</strong><br />
year. Dur<strong>in</strong>g w<strong>in</strong>ter-spr<strong>in</strong>g all such activity was implemented by dugout canoe or by foot<br />
along <strong>the</strong> accessible marg<strong>in</strong>s <strong>of</strong> <strong>the</strong> bas<strong>in</strong>.<br />
Moreover, extractive sites <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong>, concentrated along <strong>the</strong> low narrow levees<br />
<strong>of</strong> <strong>the</strong> rivers and <strong>the</strong>ir tributaries, can only represent summer-fall camps (although dry<br />
years with reduced hydroperiod could extend <strong>the</strong>se occupations). The prehistoric extractive<br />
camps <strong>in</strong> <strong>the</strong> Felsenthal Project area, at about 20 m above sea level, were just as dist<strong>in</strong>ctively<br />
seasonal as <strong>the</strong>ir functional counterparts <strong>in</strong> a high mounta<strong>in</strong> range.
References Cited<br />
Allen, J. R. L.<br />
1965 A Review <strong>of</strong> <strong>the</strong> Orig<strong>in</strong> and Characteristics <strong>of</strong> Recent Alluvial Sediments. Sedimentology<br />
5:89-191.<br />
<strong>Arkansas</strong> Archeological Survey<br />
1978 Proposal for a Cultural Resources Survey and Evaluation: A) Calion Lock and<br />
Dam Site; B) Felsenthal Navigation Pool, South-central <strong>Arkansas</strong>. Ms. on file, <strong>Arkansas</strong><br />
Archeological Survey, Fayetteville.<br />
<strong>Arkansas</strong> Department <strong>of</strong> Plann<strong>in</strong>g<br />
1974 <strong>Arkansas</strong> Natural Area Plan. <strong>Arkansas</strong> Department <strong>of</strong> Plann<strong>in</strong>g, Little Rock.<br />
<strong>Arkansas</strong> Gazette<br />
Newspaper published at Little Rock, <strong>Arkansas</strong>. Available on micr<strong>of</strong>ilm at <strong>the</strong> <strong>Arkansas</strong><br />
History Commission, Little Rock.<br />
Asch, David L., Kenneth B. Farnsworth, and Nancy B. Asch<br />
1979 Woodland Subsistence and Settlement, West Central Ill<strong>in</strong>ois. Ms. on file, Foundation<br />
for Ill<strong>in</strong>ois Archeology.<br />
Baker, Charles M.<br />
1974 A Study <strong>of</strong> Aborig<strong>in</strong>al Novaculite Exploitation <strong>in</strong> <strong>the</strong> Ouachita Mounta<strong>in</strong>s <strong>of</strong> Southcentral<br />
<strong>Arkansas</strong>. Unpublished M. A. <strong>the</strong>sis, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>.<br />
1975 Site Abandonment and <strong>the</strong> Archeological Record: An Empirical Case for Anticipated<br />
Return. <strong>Arkansas</strong> Academy <strong>of</strong> Science Proceed<strong>in</strong>gs 29:10-11.<br />
Baerris, David A., Joan E. Freeman, and James V. Wright<br />
1958 The Contract<strong>in</strong>g Stem Projectile Po<strong>in</strong>t <strong>in</strong> Eastern Oklahoma. Bullet<strong>in</strong> <strong>of</strong> <strong>the</strong> Oklahoma<br />
Anthropological Society 6:61-82.<br />
Barth, Fredrick (editor)<br />
1969 Ethnic Groups and Boundaries. Little-Brown, Boston.<br />
Bartlett, Charles S. Jr.<br />
1963 The Tom’s Brook Site–3JO1: A Prelim<strong>in</strong>ary Report. In <strong>Arkansas</strong> Archeology, 1962,<br />
edited by Charles R. McGimsey III, pp. 15-67. <strong>Arkansas</strong> Archeological Society, Fayetteville.
280<br />
Bell, Robert E.<br />
1958 Guide to <strong>the</strong> Identification <strong>of</strong> Certa<strong>in</strong> American Indian Projectile Po<strong>in</strong>ts. Oklahoma<br />
Anthropological Society, Special Bullet<strong>in</strong> 1.<br />
1960 Guide to <strong>the</strong> Identification <strong>of</strong> Certa<strong>in</strong> American Indian Projectile Po<strong>in</strong>ts. Oklahoma<br />
Anthropological Society, Special Bullet<strong>in</strong> 2.<br />
Belmont, John S., and Stephen Williams<br />
1981 Pa<strong>in</strong>ted Pottery Horizons <strong>in</strong> <strong>the</strong> Sou<strong>the</strong>rn Mississippi Valley. Geoscience and Man<br />
22:19-42.<br />
B<strong>in</strong>ford, Lewis R.<br />
1980 Willow Smoke and Dogs’ Tails: Hunter-ga<strong>the</strong>rer Settlement-Systems and Archeological<br />
Site Formation. American Antiquity 45:4-20.<br />
B<strong>in</strong>ford, Lewis R., and Sally R. B<strong>in</strong>ford<br />
1966 A Prelim<strong>in</strong>ary Analysis <strong>of</strong> Functional Variability <strong>in</strong> <strong>the</strong> Mousterian <strong>of</strong> Levallois<br />
Facies. In Recent Studies <strong>in</strong> Paleoanthropology, edited by J. D. Clarkland and F. C. Howell,<br />
pp. 238-295. American Anthropologist 68(2).<br />
Bolton, Herbert Eugene<br />
1914 Athanase de Mezieres and <strong>the</strong> Louisiana Texas Frontier, <strong>17</strong>68-<strong>17</strong>80 (Vol. 2). Arthur J.<br />
Clarke, Cleveland.<br />
Bradley County<br />
County Court Records. Bradley County Courthouse, Warren.<br />
Bra<strong>in</strong>, Jeffry P.<br />
1978 Late Prehistoric Settlement Pattern<strong>in</strong>g <strong>in</strong> <strong>the</strong> Yahoo Bas<strong>in</strong> and Natchez Bluffs<br />
Regions <strong>of</strong> <strong>the</strong> Lower Mississippi Valley. In Mississippian Settlement Patterns, edited by<br />
Bruce D. Smith, pp. 331-368. Academic Press, New York.<br />
Branner, John C.<br />
1899 Some Old French Place Names <strong>in</strong> <strong>Arkansas</strong>. Modern Language Notes 14:33-40.<br />
Brown, James A.<br />
1976 Spiro Studies: The Artifacts (Vol. 4). Stovall Museum <strong>of</strong> Science and History and<br />
<strong>University</strong> <strong>of</strong> Oklahoma Resarch Institute, Norman.
Brown, Sandra, Mark M. Br<strong>in</strong>son, and Ariel E. Lugo<br />
1979 Structure and Function <strong>of</strong> Riparian Wetlands. In Strategies for Protection and Management<br />
<strong>of</strong> Floodpla<strong>in</strong> Wetlands and O<strong>the</strong>r Riparian Ecosystems, coord<strong>in</strong>ated by R. Roy<br />
Johnson and J. Frank McCormick, pp. <strong>17</strong>-31. U.S. Department <strong>of</strong> Agriculture, General<br />
Technical Report WO-12.<br />
Brunsden, D., and R. Kesel<br />
1974 Slope Development on a Mississippi River Bluff <strong>in</strong> Historic Time. Journal <strong>of</strong> Geology<br />
81:576-598.<br />
Bryson, Reid A., David A. Baerreis, and Wayne M. Wendland<br />
1970 The Character <strong>of</strong> Late Glacial and Post-glacial Climatic Changes. In Pleistocene<br />
and Recent Environments <strong>of</strong> <strong>the</strong> Central Great Pla<strong>in</strong>s, edited by Wakefield Dort, Jr. and J.<br />
Knox Jones, Jr. Department <strong>of</strong> Geology, <strong>University</strong> <strong>of</strong> Kansas Special Publication 3:53-<br />
74.<br />
Buchanan, Thomas M.<br />
1973 Key to <strong>the</strong> Fishes <strong>of</strong> <strong>Arkansas</strong>. <strong>Arkansas</strong> Game and Fish Commission, Little Rock.<br />
1974 Threatened Native Fishes <strong>of</strong> <strong>Arkansas</strong>. In <strong>Arkansas</strong> Natural Area Plan, pp. 67-92.<br />
<strong>Arkansas</strong> Department <strong>of</strong> Plann<strong>in</strong>g, Little Rock.<br />
Buckner, John W.<br />
1979 Wilderness Lady: A History <strong>of</strong> Crossett, <strong>Arkansas</strong>. Rose Publish<strong>in</strong>g, Little Rock.<br />
Byrd, Kathleen M., and Robert H. Neuman<br />
1978 Archaeological Data Relative to Prehistoric Subsistence <strong>in</strong> <strong>the</strong> Lower Mississippi<br />
Alluvial Valley. Geoscience and Man 19:9-21.<br />
Ca<strong>in</strong>, Robert H.<br />
1974 Pimple Mounds: A New Viewpo<strong>in</strong>t. Ecology 55:<strong>17</strong>8-182<br />
Callahan, Errett<br />
1980 Spatial Organization <strong>of</strong> <strong>the</strong> Work Areas <strong>of</strong> Three Contemporary Fl<strong>in</strong>tknappers.<br />
Quarterly Bullet<strong>in</strong> <strong>of</strong> <strong>the</strong> Archeological Society <strong>of</strong> Virg<strong>in</strong>ia 35:101-108.<br />
Camden <strong>in</strong> 1855<br />
1969 Unidentified newspaper clipp<strong>in</strong>g. Repr<strong>in</strong>ted <strong>in</strong> <strong>the</strong> Ouachita County Historical<br />
Quarterly 1(4):10-12.<br />
281
282<br />
Chomko, Stephen A., and Gary W. Crawford<br />
1978 Plant Husbandry <strong>in</strong> Prehistoric Eastern North America: New Evidence for its<br />
Development. American Antiquity 44:405-408.<br />
Comeaux, Malcolm L.<br />
1972 Atchafalaya Swamp Life, Settlement and Folk Occupations. Geoscience and Man 2.<br />
Compton, Robert R.<br />
1962 Manual <strong>of</strong> Field Geology. John Wiley and Sons, New York.<br />
Conner, William H., and John W. Day<br />
1976 Productivity and Composition <strong>of</strong> a Bald Cypress-Water Tupelo Site and a Bottomland<br />
Hardwood Site <strong>in</strong> a Louisiana Swamp. American Journal <strong>of</strong> Botany 63(10):1354-<br />
1364.<br />
Connaway, John M., Samuel O. Brooks, and Samuel O. McGahey<br />
1977 The Denton Site: A Middle Archaic Occupation <strong>in</strong> <strong>the</strong> Nor<strong>the</strong>rn Yazoo Bas<strong>in</strong>, Mississippi.<br />
Mississippi Department Of Archives and History, Archaeological Report 4.<br />
Cordell, Anna H.<br />
1951 Champagnolle: A Pioneer River Town. <strong>Arkansas</strong> Historical Quarterly 10:37-47.<br />
Cordell, Mrs. Robert V. M.<br />
1947 Lapile: A Pioneer Community. <strong>Arkansas</strong> Historical Quarterly 6:273-285.<br />
Crockett, John W.<br />
1904 Biennial Report <strong>of</strong> <strong>the</strong> Secretary <strong>of</strong> State. Little Rock, <strong>Arkansas</strong>.<br />
Curry, Corliss C.<br />
1960 Early Timber Operations <strong>in</strong> Sou<strong>the</strong>ast <strong>Arkansas</strong>. <strong>Arkansas</strong> Historical Quarterly<br />
19:111-118.<br />
D’Anemours, Chevalier Paul<strong>in</strong>ier<br />
1803 Memoire sur le District de Ouachita dans la Prov<strong>in</strong>c de la Louisianne. Ms. on file,<br />
American Philosophical Society Library, Philadelphia.<br />
Davis, E. Mott, editor<br />
1961 Proceed<strong>in</strong>gs <strong>of</strong> <strong>the</strong> Fifth Conference on Caddoan Archeology. Texas Archeological<br />
Society Bullet<strong>in</strong> 31:77-143.<br />
1970 Texas. In Archeological and Historical Resources <strong>of</strong> <strong>the</strong> Red Red River Bas<strong>in</strong>, edited by<br />
Hester A. Davis. <strong>Arkansas</strong> Archeological Survey Research Series 1:27-68.
Davis, Hester A.<br />
1981 A State Plan for <strong>the</strong> Conservation <strong>of</strong> Archeological Resources <strong>in</strong> <strong>Arkansas</strong>. Volume<br />
II: Research Study Units. Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
Deeter, E. B., W. J. Latimer, and C. E. Born<br />
1925 Soil Survey <strong>of</strong> Bradley County, <strong>Arkansas</strong>. U.S. Department <strong>of</strong> Agriculture, Soil Survey<br />
Report 7.<br />
DeJarnette, David L., Edward B. Kurjack, and James W. Cambron<br />
1962 Stanfield-Worley Bluff Shelter Excavations. Journal <strong>of</strong> Alabama Archaeology 8:1-111.<br />
Derley, Jim<br />
1979 The Honey Creek Po<strong>in</strong>t: A Prairie County Orig<strong>in</strong>al? Field Notes, Newsletter <strong>of</strong> <strong>the</strong><br />
<strong>Arkansas</strong> Archeological Society <strong>17</strong>0:5-6.<br />
Dick<strong>in</strong>son, S. D., and Harry J. Lemley<br />
1939 Evidences <strong>of</strong> <strong>the</strong> Marksville and Coles Creek Complexes at <strong>the</strong> Kirkham Site,<br />
Clark County, <strong>Arkansas</strong>. Texas Archeleogical and Paleontological Society Bullet<strong>in</strong> 11:139-<br />
189. Repr<strong>in</strong>ted 1963 <strong>in</strong> The <strong>Arkansas</strong> Archeologist 8(4):57-76.<br />
Dillehay, Tom D.<br />
1973 Small Archeological Site Investigations for Interpretation <strong>of</strong> Site Activities. Texas<br />
Archeological Society Bullet<strong>in</strong> 44:169-<strong>17</strong>7.<br />
E<strong>the</strong>ridge, Y. W.<br />
1959 History <strong>of</strong> Ashley County, <strong>Arkansas</strong>. <strong>Arkansas</strong> Historical Series 8. Press-Argus, Van<br />
Buren.<br />
Ewel, Ka<strong>the</strong>r<strong>in</strong>e Carter<br />
1979 Riparian Ecosystems: Conservation <strong>of</strong> Their Unique Characteristics. In Strategies<br />
for Protection and Management <strong>of</strong> Floodpla<strong>in</strong> Wetlands and O<strong>the</strong>r Riparian Ecosystems,<br />
coord<strong>in</strong>ated by R. Roy Johnson and J. Frank McCormick, pp. 56-62. U.S. Department <strong>of</strong><br />
Agriculture, General Technical Report WO-12.<br />
Felsenthal Press<br />
Newspaper published at Felsenthal, <strong>Arkansas</strong>. Available on micr<strong>of</strong>ilm at <strong>the</strong> <strong>Arkansas</strong><br />
History Commission, Little Rock.<br />
Fenenga, Frankl<strong>in</strong><br />
1953 The Weights <strong>of</strong> Chipped Stone Po<strong>in</strong>ts: A Clue to Their Functions. Southwestern<br />
Journal <strong>of</strong> Anthropology 9:309-323.<br />
283
284<br />
Fisk, Harold N.<br />
1944 Geological Investigation <strong>of</strong> <strong>the</strong> Alluvial Valley <strong>of</strong> <strong>the</strong> Lower Mississippi River.<br />
War Department, U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Publication 52.<br />
1947 F<strong>in</strong>e-gra<strong>in</strong>ed Alluvial Deposits and Their Effects on Mississippi River Activity.<br />
War Department, Corps <strong>of</strong> Eng<strong>in</strong>eers, Waterways Experiment Station, Vol. 1.<br />
Fitzhugh, William W.<br />
1972 Environmental Archeology and Cultural Systems <strong>in</strong> Hamilton Inlet, Labrador: A<br />
Survey <strong>of</strong> <strong>the</strong> Central Labrador Coast from 3000 B.C. to <strong>the</strong> Present. Smithsonian Contributions<br />
to Anthropology 16.<br />
Flannery, Kent V.<br />
1968 Archaeological Systems Theory and Early-Mesoamerica. In Anthropological Archeology<br />
<strong>in</strong> <strong>the</strong> Americas, edited by Betty J. Meggers, pp. 67-87. Anthropological Society <strong>of</strong><br />
Wash<strong>in</strong>gton, Wash<strong>in</strong>gton D. C.<br />
1976 The Early Mesoamerican Village. Academic Press, New York. (editor)<br />
Fleetwood, A. R.<br />
1969 Geological Investigation <strong>of</strong> <strong>the</strong> Ouachita River Area, Lower Mississippi Valley.<br />
U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Technical Report 81-69-2. Vicksburg.<br />
Flenniken, J. Jeffrey, and James C. Haggarty<br />
1979 Trampl<strong>in</strong>g as an Agency <strong>in</strong> <strong>the</strong> Formation <strong>of</strong> Edge Damage: An Experiment <strong>in</strong><br />
Lithic Technology. Northwest Anthropological Research Notes 13(2):208-214.<br />
Ford, James A.<br />
1935 Ceramic Decoration Sequence at an Old Indian Village Site near Sicily Island,<br />
Louisiana. Louisiana Department <strong>of</strong> Conservation, Louisiana Geological Survey Anthropological<br />
Study 1.<br />
1936 Analysis <strong>of</strong> Indian Village Site Collections from Louisiana and Mississippi. Department<br />
<strong>of</strong> Conservation, Louisiana Geological Survey, Anthropological Study 2.<br />
1951 Greenhouse: A Troyville-Coles Creek Period Site <strong>in</strong> Avoyelles Parish, Louisiana.<br />
American Museum <strong>of</strong> Natural History, Anthropological Papers 44(Part 1):1-132.
1961 Menard Site: The Quapaw Village <strong>of</strong> Osotouy on <strong>the</strong> <strong>Arkansas</strong> River. Anthropological<br />
Papers <strong>of</strong> <strong>the</strong> American Museum <strong>of</strong> Natural History 48:Part 2.<br />
Ford, James A., and George I. Quimby, Jr.<br />
1945 The Tchefuncte Culture, an Early Occupation <strong>of</strong> <strong>the</strong> Lower Mississippi Valley.<br />
Society for American Archaeology Memoirs 2. (Kraus Repr<strong>in</strong>t 1974).<br />
Ford, James A., and Clarence H. Webb<br />
1956 Poverty Po<strong>in</strong>t, a Late Archaic Site <strong>in</strong> Louisiana. Anthropological Papers <strong>of</strong> <strong>the</strong> American<br />
Museum <strong>of</strong> Natural History 46:Part 1.<br />
Ford, James A., and Gordon R. Willey<br />
1940 Crooks Site, a Marksville Period Burial Mound <strong>in</strong> La Salle Parish, Louisiana. Department<br />
<strong>of</strong> Conservation, Louisiana Geological Survey, Anthropological Study 3.<br />
Foster, James P.<br />
1864 Report to Adm. David D. Porter, April 15, 1864. In Official Records <strong>of</strong> <strong>the</strong> Union and<br />
Confederate Navies <strong>in</strong> <strong>the</strong> War <strong>of</strong> <strong>the</strong> Rebellion, Series 1, 26(Part 1):236-237.<br />
Fowke, Gerard<br />
1928 Archeological Investigations II. Bureau <strong>of</strong> American Ethnology, Annual Report<br />
44:399-540.<br />
Fowler, Melv<strong>in</strong> L.<br />
1959 Summary Report <strong>of</strong> Modoc Rock Shelter. Ill<strong>in</strong>ois State Museum Report <strong>of</strong> Investigations<br />
8.<br />
Frondel, Clifford<br />
1962 The System <strong>of</strong> M<strong>in</strong>erology <strong>of</strong> James Dwight Dana and Edward Salisbury Dana, Yale <strong>University</strong><br />
1837-1892, Vol. III, Silica M<strong>in</strong>erals. John Wiley and Sons, New York.<br />
General Land Office<br />
1827 Plat maps: T<strong>17</strong>S R11W; T<strong>17</strong>S R10W; T18S R10W; T18S R9W. General Land Office,<br />
Little Rock.<br />
1842 Plat maps: T19S R10W; T19S R9W. General Land Office, Little Rock.<br />
1843 Plat maps: T<strong>17</strong>S R10W; T18S R9W. General Land Office, Little Rock.<br />
1844 Plat maps: T<strong>17</strong>S R11W; T18S R11W; T18S R10W; T19S R11W; T19S R9W. General<br />
Land Office, Little Rock.<br />
285
286<br />
1845 Plat maps: T<strong>17</strong>S R11W; T<strong>17</strong>S R10W; T18S R11W; T19S R11W; T19S R10W; T19S<br />
R9W. General Land Office, Little Rock.<br />
Giard<strong>in</strong>o, Marco J.<br />
1979 Cultural Resources Survey and Evaluation <strong>of</strong> <strong>the</strong> Cane Creek R. C. & D. Measure,<br />
L<strong>in</strong>coln County, <strong>Arkansas</strong>. New World Research Report <strong>of</strong> Investigations 16.<br />
Gill, Hiram V., Don C. Avery, Fred C. Larance, and Charles L. Fultz<br />
1979 Soil Survey <strong>of</strong> Ashley County, <strong>Arkansas</strong>. U.S. Department <strong>of</strong> Agriculture, Soil Conservation<br />
Service.<br />
Goodspeed<br />
1890 Biographical and Historical Memoirs <strong>of</strong> Sou<strong>the</strong>rn <strong>Arkansas</strong>. Goodspeed Publish<strong>in</strong>g,<br />
Chicago.<br />
Goodyear, Albert C., John H. House, and Neal W. Ackerly<br />
1979 Laurens-Anderson: An Archeological Study <strong>of</strong> <strong>the</strong> Inter-river<strong>in</strong>e Piedmont. Institute<br />
<strong>of</strong> Archeology and Anthropology, <strong>University</strong> <strong>of</strong> South Carol<strong>in</strong>a Anthropological Studies<br />
4.<br />
Gordon, M. E., L. Russert Kraemer, and A. V. Brown<br />
1979 Unionaceae <strong>of</strong> <strong>Arkansas</strong>: Historical Review, Checklist, and Observations on Distributional<br />
Patterns. Bullet<strong>in</strong> <strong>of</strong> <strong>the</strong> American Malecological Union 1979:31-37.<br />
Gossel<strong>in</strong>k, J. G., and R. E. Turner<br />
1978 The Role <strong>of</strong> Hydrology <strong>in</strong> Freshwater Wetland Ecosystems. In Freshwater Wetlands:<br />
Ecological Processes and Management Potential, edited by Ralph E. Good, Dennis<br />
E. Whigham, and Robert L. Simpson, pp. 63-77. Academic Press, New York.<br />
Green, Juanita Whitaker<br />
1954 The History <strong>of</strong> Union County. (Mimeograph copies) Jack Lee, Eldorado, <strong>Arkansas</strong>.<br />
Green, Thomas J., and Cheryl A. Munson<br />
1978 Mississippian Settlement Patterns <strong>in</strong> Southwestern Indiana. In Mississippian Settlement<br />
Patterns, edited by Bruce D. Smith, pp. 293-330. Academic Press, New York.<br />
Green, W. D.<br />
1864 Letter to Col. Powell Clayton, April 20, 1864. In War <strong>of</strong> <strong>the</strong> Rebellion, <strong>the</strong> Official<br />
Records <strong>of</strong> <strong>the</strong> Union and Confederate Armies, Series 1, 34(Part 3):236-237.<br />
Greene, Glen S., Ernest R. Williams, and G. R. Dennis Price<br />
1975 The Search for Fort Miro: A Colonial Spanish Fort <strong>in</strong> <strong>the</strong> Ouachita Valley. Nor<strong>the</strong>ast<br />
Louisiana Archeological Society Popular Series 1.
Greengo, Robert E.<br />
1964 Issaquena: An Archeological Phase <strong>in</strong> <strong>the</strong> Yazoo Bas<strong>in</strong> <strong>of</strong> <strong>the</strong> Lower Mississippi<br />
Valley. Memoirs <strong>of</strong> <strong>the</strong> Society for American Archaeology 18.<br />
Halls, Lowell K. (editor)<br />
1977 Sou<strong>the</strong>rn Fruit-produc<strong>in</strong>g Woody Plants Used by Wildlife. U.S. Department <strong>of</strong><br />
Agriculture, Sou<strong>the</strong>rn Forest Experiment Station, General Technical Report SO-16.<br />
Hally, David J.<br />
1967 Post-Coles Creek Cultural Development <strong>in</strong> <strong>the</strong> Upper Tensas Bas<strong>in</strong> <strong>of</strong> Louisiana.<br />
Proceed<strong>in</strong>gs <strong>of</strong> <strong>the</strong> Twenty-third Sou<strong>the</strong>astern Archaeological Conference, SEAC Bullet<strong>in</strong><br />
6:36-39.<br />
Hardesty, Donald L.<br />
1980 The Use <strong>of</strong> General Ecological Pr<strong>in</strong>ciples <strong>in</strong> Archaeology. In Advances <strong>in</strong> Archaeological<br />
Method and Theory (Volume 3), edited by Michael B. Schiffer, pp. 158-188. Academic<br />
Press, New York.<br />
Harr<strong>in</strong>gton, Mark R.<br />
1920 Certa<strong>in</strong> Caddo Sites <strong>in</strong> <strong>Arkansas</strong>. Museum <strong>of</strong> <strong>the</strong> American Indian Heye Foundation,<br />
Indian Notes and Monographs 10.<br />
Hayden, Julian D.<br />
1965 Fragile Pattern Areas. American Antiquity 31:272-276.<br />
Haynes, C. Vance<br />
1973 Soil Redef<strong>in</strong>ed. Geotimes 18:8.<br />
Heartfield, Lorra<strong>in</strong>e, and G. R. Dennis Price<br />
1976 An Archeological Survey <strong>of</strong> Portions <strong>of</strong> <strong>the</strong> Monroe to Sandy Bayou Levee,<br />
Ouachita River Levees, Louisiana. Ms. on file, U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg.<br />
Heartfield, Price, and Greene, Inc.<br />
1980 A Cultural Resources Reconnaissance <strong>of</strong> <strong>the</strong> Felsenthal National Wildlife Refuge,<br />
<strong>Arkansas</strong>. Ms. on file, Interagency Archeological Services, Heritage Conservation and<br />
Recreation Service, Atlanta.<br />
Hemm<strong>in</strong>gs, E. Thomas<br />
1969 Excavation and Analysis <strong>of</strong> Liv<strong>in</strong>g Floors <strong>in</strong> Alluvial Sites. Sou<strong>the</strong>astern Archaeological<br />
Conference Bullet<strong>in</strong> 11:41-45.<br />
287
288<br />
1975 An Archaeological Survey <strong>of</strong> <strong>the</strong> South Prong <strong>of</strong> <strong>the</strong> Alafia River, Florida. Florida<br />
Anthropologist 28:41-52.<br />
1979 The Felsenthal Project: Archeological Survey and Test<strong>in</strong>g <strong>in</strong> <strong>the</strong> Ouachita Valley,<br />
South <strong>Arkansas</strong>. Paper presented at <strong>the</strong> 36th Annual Meet<strong>in</strong>g <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>astern<br />
Archaeological Conference, Atlanta.<br />
1980 Spirit Lake (3LA83): Test Excavations <strong>in</strong> a Late Caddo Site on <strong>the</strong> Red River,<br />
Southwest <strong>Arkansas</strong>. Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
Hemm<strong>in</strong>gs, E. Thomas, and Ka<strong>the</strong>r<strong>in</strong>e D<strong>in</strong>nel<br />
1980 Analysis <strong>of</strong> a Quapaw Hunt<strong>in</strong>g Camp on <strong>the</strong> Sal<strong>in</strong>e River, Sou<strong>the</strong>ast <strong>Arkansas</strong>.<br />
Paper presented at <strong>the</strong> 37th Annual Meet<strong>in</strong>g <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>astern Archaeological Conference,<br />
New Orleans.<br />
Hilliard, Jerry E.<br />
1981 Selection and Use <strong>of</strong> Acorn Species by Late Prehistoric Ozark Inhabitants. Paper<br />
presented at <strong>the</strong> 46th Annual Meet<strong>in</strong>g <strong>of</strong> <strong>the</strong> Society for American Archaeology, San<br />
Diego.<br />
Hodges, T. L., and Mrs. T. L. Hodges<br />
1945 Suggestions for <strong>the</strong> Identification <strong>of</strong> Certa<strong>in</strong> Mid-Ouachita Pottery as Cah<strong>in</strong>nio<br />
Caddo. Texas Archeological and Paleontological Society Bullet<strong>in</strong> 16:98-116. Repr<strong>in</strong>ted 1963<br />
<strong>in</strong> The <strong>Arkansas</strong> Archeologist 4(8):1-12.<br />
H<strong>of</strong>fman, Michael P.<br />
1970a <strong>Arkansas</strong>. In Archeological and Historical Resources <strong>of</strong> <strong>the</strong> Red River Bas<strong>in</strong>, edited<br />
by Hester A. Davis. <strong>Arkansas</strong> Archeological Survey Research Series 1:137-187.<br />
1970b Three Sites <strong>in</strong> Millwood Reservoir: The Hutt Site, <strong>the</strong> Stark Site, and <strong>the</strong> Beard<br />
Lake Site. <strong>Arkansas</strong> Archeological Survey Research Series 2.<br />
1977 The K<strong>in</strong>kead-Ma<strong>in</strong>ard Site, 3PU2: A Late Prehistoric Quapaw Phase Site near<br />
Little Rock, <strong>Arkansas</strong>. The <strong>Arkansas</strong> Archeologist 16-18:1-41.<br />
Holmes, W. H.<br />
1903 Aborig<strong>in</strong>al Pottery <strong>of</strong> <strong>the</strong> Eastern United States. Bureau <strong>of</strong> American Ethnology, Annual<br />
Report 20:1-237.<br />
House, John H.<br />
1975 A Functional Typology for Cache Project Surface Collections. In <strong>the</strong> Cache River<br />
Archeological Project: An Experiment <strong>in</strong> Contract Archeology, edited by Michael B.<br />
Schiffer and John H. House. <strong>Arkansas</strong> Archeological Survey Research Series 8:55-73.
1980 Archaic Occupation <strong>in</strong> <strong>the</strong> <strong>Arkansas</strong> River Lowland. <strong>Arkansas</strong> Archeological Society,<br />
Field Notes <strong>17</strong>1:5-10.<br />
1981 Evolution <strong>of</strong> Complex Society <strong>in</strong> East-central <strong>Arkansas</strong>: An Overview <strong>of</strong> Environments<br />
and Regional Databases. In <strong>Arkansas</strong> Archeology <strong>in</strong> Review: A Symposium, edited<br />
by Marv<strong>in</strong> D. Jeter and Neal L. Trubowitz. <strong>Arkansas</strong> Archeological Survey Research<br />
Series (<strong>in</strong> press).<br />
House, John H., and Marv<strong>in</strong> D. Jeter (editors)<br />
1981 Archeological Investigations <strong>in</strong> <strong>the</strong> Boydell Locality, Sou<strong>the</strong>ast <strong>Arkansas</strong>. Ms. on<br />
file, <strong>Arkansas</strong> Archeological Survey, Monticello.<br />
House, John H., and Henry McKelway<br />
1980 Mississippian and Quapaw on <strong>the</strong> Lower <strong>Arkansas</strong>. Ms. on file, <strong>Arkansas</strong> Archeological<br />
Survey, Fayetteville.<br />
House, John H., and Michael B. Schiffer<br />
1975 Archeological Survey <strong>in</strong> <strong>the</strong> Cache River Bas<strong>in</strong>. In The Cache River Archeological<br />
Project: An Experiment <strong>in</strong> Contract Archeology, edited by Michael B. Schiffer and<br />
John H. House. <strong>Arkansas</strong> Archeological Survey Research Series 8:37-53.<br />
House, John H., and James W. Smith<br />
1975 Experiments <strong>in</strong> Replication <strong>of</strong> Fire-cracked Rock. In The Cache River Archeological<br />
Project: An Experiment <strong>in</strong> Contract Archeology, edited by Michael B. Schiffer and<br />
John H. House. <strong>Arkansas</strong> Archeological Survey Research Series 8:75-80.<br />
Hrdlicka, Ales<br />
1909 Report on an Additional Collection <strong>of</strong> Skeletal Rema<strong>in</strong>s from <strong>Arkansas</strong> and Louisiana.<br />
Journal <strong>of</strong> <strong>the</strong> Academy <strong>of</strong> Natural Sciences <strong>of</strong> Philadelphia (second series) 14:<strong>17</strong>1-<br />
249.<br />
Huffman, R. T.<br />
1976 The Relation <strong>of</strong> Flood Duration Patterns to Dom<strong>in</strong>ant Forest Species Associations Occurr<strong>in</strong>g<br />
on Selected First Bottom Sites <strong>of</strong> <strong>the</strong> Ouachita River Dra<strong>in</strong>age Bas<strong>in</strong> <strong>in</strong> Sou<strong>the</strong>rn <strong>Arkansas</strong>.<br />
Unpublished Ph.D. dissertation, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>.<br />
Huxtable, J., M. J. Aitken, and J. C. Weber<br />
1972 Thermolum<strong>in</strong>escent Dat<strong>in</strong>g <strong>of</strong> Baked Clay Balls <strong>of</strong> <strong>the</strong> Poverty Po<strong>in</strong>t Culture.<br />
Archaeometry 14(2):269-275.<br />
James, Frances C.<br />
1974 Threatened Native Birds <strong>of</strong> <strong>Arkansas</strong>. In <strong>Arkansas</strong> Natural Area Plan, pp. 107-122.<br />
<strong>Arkansas</strong> Department <strong>of</strong> Plann<strong>in</strong>g, Little Rock.<br />
289
290<br />
Jeter, Marv<strong>in</strong> D.<br />
1980 The Archeology <strong>of</strong> Sou<strong>the</strong>ast <strong>Arkansas</strong>: An Overview for <strong>the</strong> 1980s. Paper presented<br />
at <strong>the</strong> 45th Annual Meet<strong>in</strong>g <strong>of</strong> <strong>the</strong> Society for American Archaeology, Philadelphia.<br />
Jeter, Marv<strong>in</strong> D., David B. Kelley, and George P. Kelley<br />
1979 The Kelley-Grimes Site: A Mississippi Period Burial Mound, Sou<strong>the</strong>ast <strong>Arkansas</strong>,<br />
Excavated <strong>in</strong> 1936. <strong>Arkansas</strong> Archeologist 20:1-51.<br />
Johnson, Roland<br />
1979 Interview at Felsenthal, <strong>Arkansas</strong>, September 21, by Beverly Watk<strong>in</strong>s and E.<br />
Thomas Hemm<strong>in</strong>gs. Transcribed tape on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
Kellam, Robert F.<br />
1970 Diary, September 1-December 31, 1860. Ouachita County Historical Quarterly 2(2):6-<br />
18.<br />
1971 Diary, January 1-March 31, 1861. Ouachita County Historical Quarterly 2(3):11-20.<br />
Keller, W. D. George W. Viele, and Clayton H. Johnson<br />
1977 Texture <strong>of</strong> <strong>Arkansas</strong> Novaculite Indicates Thermally Induced Metamorphism.<br />
Journal <strong>of</strong> Sedimentary Petrology 47:834-843.<br />
Kelley, David B.<br />
1979 Cultural Resources Assessment <strong>of</strong> <strong>the</strong> Calion Lock and Dam, Union County, <strong>Arkansas</strong>.<br />
Ms. on file. <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
Kelley, Lucretia S.<br />
1979 Animal Resource Exploitation by Early Cahokia Populations on <strong>the</strong> Merrill Tract.<br />
Ill<strong>in</strong>ois Archaeological Survey, Circular 4.<br />
K<strong>in</strong>g, James E., and W. H. Allen, Jr.<br />
1977 A Holocene Vegetation Record from <strong>the</strong> Mississippi River Valley, Sou<strong>the</strong>astern<br />
Missouri. Quarternary Research 8:307-323.<br />
Kl<strong>in</strong>ger, Timothy C. (assembler)<br />
1975 Mississippian Communities <strong>in</strong> <strong>the</strong> St. Francis Bas<strong>in</strong>: A Central Place Model. <strong>Arkansas</strong><br />
Academy <strong>of</strong> Science Proceed<strong>in</strong>gs 29:50-51.<br />
1979 Hampton: An Archeological and Historical Overview <strong>of</strong> a Proposed Strip M<strong>in</strong>e<br />
Tract <strong>in</strong> South Central <strong>Arkansas</strong>. <strong>Arkansas</strong> Archeological Survey Research Reports 19.
Krieger, Alex D.<br />
1946 Culture Complexes and Chronology <strong>in</strong> Nor<strong>the</strong>astern Texas. <strong>University</strong> <strong>of</strong> Texas<br />
Publication 4640.<br />
Lafferty, Robert H. III, Jeffrey L. Ot<strong>in</strong>ger, Sandra Clements S. Scholtz, W. Fredrick Limp, Beverly<br />
Watk<strong>in</strong>s, and Robert D. Jones<br />
1981 Settlement Prediction <strong>in</strong> Sparta. <strong>Arkansas</strong> Archeological Survey Research Series 14 (<strong>in</strong><br />
press).<br />
Langenheim, Jean H.<br />
1964 Present Status <strong>of</strong> Botanical Studies <strong>of</strong> Ambers. Harvard <strong>University</strong> Botanical Museum<br />
Leaflets 20(8):225-28.<br />
1969 Amber: A Botanical Inquiry. Science 163:1157-1169.<br />
Langenheim, Jean H., and Curt W. Beck<br />
1968 Catalogue <strong>of</strong> Infrared Spectra <strong>of</strong> Fossil Res<strong>in</strong>s (Ambers), I. North and South<br />
America. Harvard <strong>University</strong> Botanical Museum Leaflets 22(3):65-120.<br />
Larance, Red C.<br />
1961 Soil Survey <strong>of</strong> Bradley County, <strong>Arkansas</strong>. U.S. Department <strong>of</strong> Agriculture, Soil Conservation<br />
Services Series 1958(1).<br />
Lemley, Harry J., and S. D. Dick<strong>in</strong>son<br />
1937 Archeological Investigations on Bayou Macon <strong>in</strong> <strong>Arkansas</strong>. Texas Archeological and<br />
Paleontological Society Bullet<strong>in</strong> 9:11-47. Repr<strong>in</strong>ted 1964 <strong>in</strong> <strong>Arkansas</strong> Archeologist 5(2):21-39.<br />
Leopold, Luna B., M. Gordon Wolman, and John P. Miller<br />
1964 Fluvial Processes <strong>in</strong> Geomorphology. W. H. Freeman, San Francisco.<br />
Lewis, R. Barry<br />
1974 Mississippian Exploitative Strategies: A Sou<strong>the</strong>ast Missouri Example. Missouri<br />
Archeological Society Research Series 11.<br />
Lewis, Thomas M. N., and Madel<strong>in</strong>e Knebeg Lewis<br />
1961 Eva: An Archaic Site. <strong>University</strong> <strong>of</strong> Tennessee Press, Knoxville.<br />
Limp, W. Fredrick, and Van A. Reidhead<br />
1979 An Economic Evaluation <strong>of</strong> <strong>the</strong> Potential <strong>of</strong> Fish Utilization <strong>in</strong> River<strong>in</strong>e Environments.<br />
American Antiquity 44:70-78.<br />
Lischka, Joseph J.<br />
1973 Prelim<strong>in</strong>ary Report on Test Excavations <strong>of</strong> Prehistoric Sites <strong>in</strong> <strong>the</strong> Felsenthal National<br />
Wildlife Refuge, 1972. Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
291
292<br />
Lloyd, James T.<br />
1856 Lloyd’s Steamboat Directory and Disasters on <strong>the</strong> Western Rivers. Jas. T. Lloyd,<br />
C<strong>in</strong>c<strong>in</strong>nati.<br />
Maisenhelder, Louis C.<br />
1958 Understory Plants <strong>of</strong> Bottomland Forests. U.S. Department <strong>of</strong> Agriculture, Sou<strong>the</strong>rn<br />
Forest Experiment Station, Occasional Paper 165.<br />
Marquardt, William H., and Patty Jo Watson<br />
1976 Excavation and Recovery <strong>of</strong> Biological Rema<strong>in</strong>s from Two Archaic Shell Middens<br />
<strong>in</strong> Western Kentucky. Paper presented at <strong>the</strong> Sou<strong>the</strong>astern Archaeological Conference,<br />
Tuscaloosa, Alabama.<br />
Marsh, Daniel L.<br />
1977 The Distribution <strong>of</strong> Cane, Arund<strong>in</strong>aria gigantia (Poaceae: Bambusoideae). <strong>Arkansas</strong><br />
Academy <strong>of</strong> Science Proceed<strong>in</strong>gs 31:72-74.<br />
Mart<strong>in</strong>, W. T.<br />
1936 Early History. Works Projects Adm<strong>in</strong>istration, Federal Writers Project. Ms. on file,<br />
Ashley County Place File, <strong>Arkansas</strong> History Commission, Little Rock.<br />
McDermott, John F.<br />
1941 A Glossary <strong>of</strong> Mississippi Valley French, 1673-1850. Wash<strong>in</strong>gton <strong>University</strong> Studies<br />
Language and Literature 12.<br />
1963 The Western Journals <strong>of</strong> Dr. George Hunter, <strong>17</strong>96-1805. Transactions <strong>of</strong> <strong>the</strong> American<br />
Philosophical Society 53(Part 4).<br />
McGimsey, Charles R. III<br />
1964 Quapaw Sites on <strong>the</strong> Lower <strong>Arkansas</strong> River. Ms. on file, <strong>Arkansas</strong> Archeological<br />
Survey, Fayetteville.<br />
1969 Indians <strong>of</strong> <strong>Arkansas</strong>. <strong>Arkansas</strong> Archeological Survey Popular Series 1.<br />
McMahan, Capta<strong>in</strong> Jesse T.<br />
[1913] Memories <strong>of</strong> Steamboat Life on <strong>the</strong> Ouachita. Clipp<strong>in</strong>g <strong>in</strong> Gr<strong>in</strong>stead Chapter History,<br />
United Daughters <strong>of</strong> <strong>the</strong> Confederacy, compiled by Mr. J. T. Sifford, pp. 85-85.<br />
Public Library <strong>of</strong> Camden and Ouachita County, Camden, <strong>Arkansas</strong>.<br />
McManus, J. W.<br />
1963 The McElroy Site: A Coles Creek-Caddoan Transitional Site <strong>in</strong> Union County,<br />
<strong>Arkansas</strong>. <strong>Arkansas</strong> Archeologist 4(3):1-5.
Meanley, Brooke<br />
1972 Swamps, River Bottoms, and Canebrakes. Barre Publications, Barre, Massachusetts.<br />
Michie, James L.<br />
1980 An Archeological Survey <strong>of</strong> Congaree Swamp: Cultural Resources Survey and<br />
Assessment <strong>of</strong> a Bottomland Environment <strong>in</strong> Central South Carol<strong>in</strong>a. Institute <strong>of</strong> Anthropology<br />
and Archeology, <strong>University</strong> <strong>of</strong> South Carol<strong>in</strong>a, Research Manuscript Series 163.<br />
Mills, Gary B.<br />
1978 Of Men and Rivers: The Story <strong>of</strong> <strong>the</strong> Vicksburg District. U.S. Army Eng<strong>in</strong>eer District,<br />
Vicksburg.<br />
Mitchell, Jennie O’Kelly, and Robert Dabney Calhoun<br />
1937 The Marquis de Maison Rouge, <strong>the</strong> Baron de Bastrop, and Colonel Abraham Morhouse<br />
— Three Ouachita Valley Soldiers <strong>of</strong> Fortune. The Maison Rouge and Bastrop<br />
Spanish Land “Grants.” Louisiana Historical Quarterly 20:289-462.<br />
Moore, Clarence B.<br />
1909 Antiquities <strong>of</strong> <strong>the</strong> Ouachita Valley. Journal <strong>of</strong> <strong>the</strong> Academy <strong>of</strong> Natural Sciences <strong>of</strong><br />
Philadelphia 14.<br />
1913 Some Aborig<strong>in</strong>al Sites <strong>in</strong> Louisiana and <strong>in</strong> <strong>Arkansas</strong>. Journal <strong>of</strong> <strong>the</strong> Academy <strong>of</strong><br />
Natural Sciences <strong>of</strong> Philadelphia 16(Part 1).<br />
Moore, Dwight M.<br />
1960 Trees <strong>of</strong> <strong>Arkansas</strong>. <strong>Arkansas</strong> Forestry Commission, Little Rock.<br />
Morse, Dan F.<br />
1963 The Steuben Village and Mounds: A Multicomponent Late Hopewell Site <strong>in</strong> Ill<strong>in</strong>ois.<br />
Museum <strong>of</strong> Anthropology, <strong>University</strong> <strong>of</strong> Michigan, Anthropological Papers 21.<br />
1970 The Big Creek Po<strong>in</strong>t. Central States Archaeological Journal <strong>17</strong>:20-23.<br />
1973 Nodena: An Account <strong>of</strong> 75 Years <strong>of</strong> Archeological Investigation <strong>in</strong> Sou<strong>the</strong>ast Mississippi<br />
County, <strong>Arkansas</strong>. <strong>Arkansas</strong> Archeological Survey Research Series 4. (editor)<br />
1975 Research Potential <strong>in</strong> Terms <strong>of</strong> Questions <strong>of</strong> Regional Prehistory. In The Cache<br />
River Archeological Project: An Experiment <strong>in</strong> Contract Archeology, assembled by<br />
Michael B. Schiffer and John H. House. <strong>Arkansas</strong> Archeological Survey Research Series<br />
8:187-198.<br />
293
294<br />
1980a Cultural Problems. In Excavation, Data Interpretation and Report on <strong>the</strong> Zebree<br />
Homestead Site, Mississippi County, <strong>Arkansas</strong>, edited by Dan F. Morse and Phyllis A.<br />
Morse (Chapter 4). Ms. on file. Memphis District, U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers.<br />
1980b The Big Lake Household and Community. In Excavation, Data Interpretation and<br />
Report on <strong>the</strong> Zebree Homestead Site, Mississippi County, <strong>Arkansas</strong>, edited by Dan<br />
F. Morse and Phyllis A. Morse (Chapter 21). Ms. on file. Memphis District, U.S. Army<br />
Corps <strong>of</strong> Eng<strong>in</strong>eers.<br />
1981 Regional Overview <strong>of</strong> Nor<strong>the</strong>ast <strong>Arkansas</strong>. In <strong>Arkansas</strong> Archeology <strong>in</strong> Review.<br />
<strong>Arkansas</strong> Archeological Survey Research Series (<strong>in</strong> press).<br />
Morse, Dan F., Daniel Wolfman, and Herbert Haas<br />
1980 Radiocarbon Dat<strong>in</strong>g. In Excavation, Data Interpretation and Report on <strong>the</strong> Zebree<br />
Homestead Site, Mississippi County, <strong>Arkansas</strong>, edited by Dan F. Morse and Phyllis A.<br />
Morse (Chapter 11). Ms. on file. Memphis District, U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers.<br />
Moseley, M. E., and C. J. Mackey<br />
1972 Peruvian Settlement Pattern Studies and Small Site Methodology. American Antiquity<br />
37:67-81.<br />
Muller, Jon<br />
1978 The K<strong>in</strong>kead System: Mississippian Settlement <strong>in</strong> <strong>the</strong> Environs <strong>of</strong> a Large Site.<br />
In Mississippian Settlement Patterns, edited by Bruce D. Smith, pp. 269-292. Academic<br />
Press, New York.<br />
Newcomer, M. H., and G. de G. Sievek<strong>in</strong>g<br />
1980 Experimental Flake-scatter Patterns: A New Interpretive Technique. Journal <strong>of</strong> Field<br />
Archeology 7:345-352.<br />
Ouachita Telegraph<br />
Newspaper published at Monroe, Louisiana. Available on micr<strong>of</strong>ilm, Nor<strong>the</strong>ast Louisiana<br />
<strong>University</strong>, Monroe.<br />
Owen, David Dale<br />
1860 Second Report <strong>of</strong> a Geological Reconnaissance <strong>of</strong> <strong>the</strong> Middle and Sou<strong>the</strong>rn Counties <strong>of</strong><br />
<strong>Arkansas</strong>. C. Sherman and Son, Philadelphia.<br />
Owens, Harry P.<br />
1979 An Assessment <strong>of</strong> Historic Period Cultural Resources along <strong>the</strong> Yazoo River<br />
Between River Miles 75.6 and 273.0, Includ<strong>in</strong>g Tchula Lake and Honey Island, Mississippi.<br />
Ms. on file, Center for Archaeological Resources, <strong>University</strong> <strong>of</strong> Mississippi.
Parmalee, Paul W., and Walter E. Klippel<br />
1974 Freshwater Mussels as a Prehistoric Food Resource. American Antiquity 39:421-<br />
434.<br />
Patterson, James L.<br />
1971 Floods <strong>in</strong> <strong>Arkansas</strong>, Magnitude and Frequency Characteristics through 1968. <strong>Arkansas</strong><br />
Geological Commission, Water Resources Circular 11.<br />
Peebles, Christopher S.<br />
1978 Determ<strong>in</strong>ants <strong>of</strong> Settlement Size and Location <strong>in</strong> <strong>the</strong> Moundville Phase. In Mississippian<br />
Settlement Patterns, edited by Bruce D. Smith, pp. 369-416. Academic Press,<br />
New York.<br />
Per<strong>in</strong>o, Gregory<br />
1968 Guide to <strong>the</strong> Identification <strong>of</strong> Certa<strong>in</strong> American Indian Projectile Po<strong>in</strong>ts. Oklahoma<br />
Anthropological Society, Special Bullet<strong>in</strong> 3.<br />
1971 Guide to <strong>the</strong> Identification <strong>of</strong> Certa<strong>in</strong> American Indian Projectile Po<strong>in</strong>ts. Oklahoma<br />
Anthropological Society, Special Bullet<strong>in</strong> 4.<br />
Phillips, Philip<br />
1970 Archeological Survey <strong>in</strong> <strong>the</strong> Lower Yazoo Bas<strong>in</strong>, Mississippi, 1949-1955. Papers <strong>of</strong><br />
<strong>the</strong> Peabody Museum <strong>of</strong> Archaeology and Ethnology 60. Harvard <strong>University</strong>.<br />
Phillips, Philip, James A. Ford, and James B. Griff<strong>in</strong><br />
1951 Archaeological Survey <strong>in</strong> <strong>the</strong> Lower Mississippi Alluvial Valley, 1940-1947. Papers<br />
<strong>of</strong> <strong>the</strong> Peabody Museum <strong>of</strong> American Archaeology and Ethnology 25. Harvard <strong>University</strong>.<br />
Price, G. R. Dennis<br />
1979 Cultural Resources Survey Report: Appendix to Design Memorandum No. 58,<br />
N<strong>in</strong>e-foot Navigation Channel, Jonesville and Columbia Pools. Ms. on file, U.S. Army<br />
Corp <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg.<br />
Price, G. R. Dennis, and Lorra<strong>in</strong>e Heartfield<br />
1977 Archeological Test Excavation <strong>of</strong> Site 16OU15 Located on <strong>the</strong> Monroe to Sandy<br />
Bayou Levee, Ouachita River Levees, Ouachita Parish, Louisiana. Ms. on file, U.S.<br />
Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg.<br />
Price, G. R. Dennis et al.<br />
1975 Excavation at <strong>the</strong> Site <strong>of</strong> Historic Fort Miro (16OU3) <strong>in</strong> Monroe, Louisiana. Ms. on<br />
file, U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg District.<br />
295
296<br />
Quimby, George I,.<br />
1951 The Fedora Site, West Baton Rouge Parish, Louisiana. Field Museum <strong>of</strong> Natural History<br />
Anthropological Series 24(2):81-135.<br />
Raab, L. Mark<br />
1976a A Proposal for Archeological Investigations <strong>in</strong> <strong>the</strong> Felsenthal Project Area, <strong>Arkansas</strong>.<br />
Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
1976b A Report <strong>of</strong> Archeological Reconnaissance <strong>in</strong> <strong>the</strong> Felsenthal Project Area, <strong>Arkansas</strong>.<br />
Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
Raynolds, Lt. Col. W. F.<br />
1871 Topographical Sketch <strong>of</strong> Ouachita River <strong>in</strong> <strong>Arkansas</strong> from Arkadelphia to <strong>the</strong> Sou<strong>the</strong>rn<br />
State L<strong>in</strong>e. U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg.<br />
Reagan, Douglas P.<br />
1974 Threatened Native Reptiles <strong>of</strong> <strong>Arkansas</strong>. In <strong>Arkansas</strong> Natural Area Plan, pp. 101-<br />
105. <strong>Arkansas</strong> Department <strong>of</strong> Plann<strong>in</strong>g, Little Rock.<br />
Reaves, Robert G., and Burney McClurkan<br />
1970 The Dansbury Creek Site (3CV24). <strong>Arkansas</strong> Archeological Society Field Notes 70:3-4.<br />
Reynolds, R.<br />
1980 The Crossett Story: The Beg<strong>in</strong>n<strong>in</strong>g <strong>of</strong> Forestry <strong>in</strong> Sou<strong>the</strong>rn <strong>Arkansas</strong> and Nor<strong>the</strong>rn<br />
Louisiana. U.S. Department <strong>of</strong> Agriculture, Sou<strong>the</strong>rn Forest Experiment Station, General<br />
Technical Report SO-32.<br />
Ritter, Dale F., W. Fred K<strong>in</strong>sey, and Marv<strong>in</strong> E. Kaufman<br />
1973 Overbank Sedimentation <strong>in</strong> <strong>the</strong> Delaware River Valley Dur<strong>in</strong>g <strong>the</strong> Last 6000<br />
Years. Science <strong>17</strong>9:374-375.<br />
Robertson, Philip A., and George T. Weaver<br />
1979 Vegetation and Tree Species Patterns near <strong>the</strong> Nor<strong>the</strong>rn Term<strong>in</strong>us <strong>of</strong> <strong>the</strong> Sou<strong>the</strong>rn<br />
Floodpla<strong>in</strong> Forest. Ecological Monographs 48:249-267.<br />
Rob<strong>in</strong>son, Henry W.<br />
1975 New Distributional Records <strong>of</strong> Fishes from <strong>the</strong> Lower Ouachita River System <strong>in</strong><br />
<strong>Arkansas</strong>. <strong>Arkansas</strong> Academy <strong>of</strong> Science Proceed<strong>in</strong>gs 29:54-56.<br />
Rohlf, F. James, John Kishpaugh, and David Kirk<br />
1972 Numerical Taxonomy System <strong>of</strong> Multivariate Statistical Programs. State <strong>University</strong> <strong>of</strong><br />
New York at Stony Brook.
Rol<strong>in</strong>gson, Martha A.<br />
1970 Archeological Investigation <strong>of</strong> Bayou Bartholomew, 1969. <strong>Arkansas</strong> Academy <strong>of</strong> Science<br />
Proceed<strong>in</strong>gs 24:13-16.<br />
1971 The Ashley Po<strong>in</strong>t. <strong>Arkansas</strong> Archeologist 12(3):50-52.<br />
1972 Report on <strong>the</strong> Prelim<strong>in</strong>ary Site Survey <strong>of</strong> <strong>the</strong> Felsenthal National Wildlife Refuge,<br />
1971. Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
1973 Initial Draft <strong>of</strong> <strong>the</strong> Bartholomew Sequence. Ms. on file, <strong>Arkansas</strong> Archeological<br />
Survey, Fayetteville.<br />
1976a The Bartholomew Phase: A Plaquem<strong>in</strong>e <strong>Adaptation</strong> <strong>in</strong> <strong>the</strong> Mississippi Valley.<br />
In Cultural Cont<strong>in</strong>uity and Change: Essays <strong>in</strong> Honor <strong>of</strong> James Bennett Griff<strong>in</strong>, edited by<br />
Charles C. Cleland, pp. 99-119. Academic Press, New York.<br />
1976b Forests and Floods: Clues to Mississippi Valley Floodpla<strong>in</strong> Environment <strong>in</strong> Sou<strong>the</strong>astern<br />
<strong>Arkansas</strong> and Significance for Prehistoric Settlement. Paper presented at <strong>the</strong><br />
33rd Sou<strong>the</strong>astern Archaeological Conference, Tuscaloosa, Alabama.<br />
1979 Emerg<strong>in</strong>g Cultural Patterns at <strong>the</strong> Toltec Site. Paper presented at <strong>the</strong> 36th Sou<strong>the</strong>astern<br />
Archaeological Conference, Atlanta, Georgia, November 1979.<br />
1981 SE Study Units 3, 4, 5 and 6: Tchula, Marksville, Baytown and Nor<strong>the</strong>rn Coles<br />
Creek. In A State Plan for <strong>the</strong> Conservation <strong>of</strong> Archeological Resources <strong>in</strong> <strong>Arkansas</strong>,<br />
Volume II: Research Study Units. Ms. on file, <strong>Arkansas</strong> Archeological Survey.<br />
Rol<strong>in</strong>gson, Martha A., and Frank F. Schambach<br />
1981 The Shallow Lake Site (3UN9/52) and Its Place <strong>in</strong> Regional Prehistory. <strong>Arkansas</strong><br />
Archeological Survey Research Series 12.<br />
Rowland, Eron (editor)<br />
1930 Life, Letters and Papers <strong>of</strong> William Dunbar, <strong>17</strong>49-1810. National Society <strong>of</strong> Colonial<br />
Dames, Jackson, Mississippi.<br />
Saucier, Roger T.<br />
1974 Quaternary Geology <strong>of</strong> <strong>the</strong> Lower Mississippi Valley. <strong>Arkansas</strong> Archeological Survey<br />
Research Series 6.<br />
Saucier, Roger T., and A. R. Fleetwood<br />
1970 Orig<strong>in</strong> and Chronologic Significance <strong>of</strong> Late Quaternary Terraces, Ouachita River,<br />
<strong>Arkansas</strong> and Louisiana. Bullet<strong>in</strong> <strong>of</strong> <strong>the</strong> Geological Society <strong>of</strong> America 81:869-890.<br />
297
298<br />
Saunders, W. B., R. H. Mapes, F. M. Carpenter, and W. C. Elsik<br />
1974 Fossiliferous Amber from <strong>the</strong> Eocene (Claiborne) <strong>of</strong> <strong>the</strong> Gulf Coastal Pla<strong>in</strong>. Geological<br />
Society <strong>of</strong> America Bullet<strong>in</strong> 85:979-984.<br />
Schambach, Frank F.<br />
1970 Pre-Caddoan Cultures <strong>in</strong> <strong>the</strong> Trans-Mississippi South: A Beg<strong>in</strong>n<strong>in</strong>g Sequence. Unpublished<br />
Ph.D. dissertation, Harvard <strong>University</strong>.<br />
1974 A Unique Engraved Steatite Bowl from Southwest <strong>Arkansas</strong>. <strong>Arkansas</strong> Archeologist<br />
15:40-44.<br />
1976 Some Suggestions for 1976-77 Operations on <strong>the</strong> Calion-Felsenthal Project. Ms. on<br />
file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
1979 Summary <strong>of</strong> Local and Regional Archeology. In Hampton: An Archeological and<br />
Historical Overview <strong>of</strong> a Proposed Strip M<strong>in</strong>e Tract <strong>in</strong> South Central <strong>Arkansas</strong>, assembled<br />
by Timothy C. Kl<strong>in</strong>ger. <strong>Arkansas</strong> Archeological Survey Research Report 19:21-30.<br />
Schiffer, Michael B.<br />
1976 Behavioral Archeology. Academic Press, New York.<br />
Scholtz, James A.<br />
1967 An Archeological Survey <strong>of</strong> <strong>the</strong> Beaver Reservoir <strong>in</strong> Northwest <strong>Arkansas</strong>. M. A. <strong>the</strong>sis,<br />
<strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>, Fayetteville.<br />
Scott, Walter<br />
1940 Survey<strong>in</strong>g <strong>the</strong> Ouachita. <strong>Arkansas</strong> Gazette Sunday Magaz<strong>in</strong>e, July 14, Little Rock.<br />
Sealander, John A.<br />
1979 A Guide to <strong>Arkansas</strong> Mammals. River Road Press, Conway, <strong>Arkansas</strong>.<br />
Setzler, Frank<br />
1933 Pottery <strong>of</strong> <strong>the</strong> Hopewell Type from Louisiana. U.S. National Museum, Proceed<strong>in</strong>gs<br />
28(art. 22).<br />
Shafer, H. G.<br />
1973 Lithic Technology at <strong>the</strong> George C. Davis Site, Cherokee County, Texas. Unpublished<br />
Ph.D., dissertation, <strong>University</strong> <strong>of</strong> Texas. <strong>University</strong> Micr<strong>of</strong>ilms, Ann Arbor.<br />
Sifford, Mrs. J. T. (compiler)<br />
n.d. History <strong>of</strong> Camden scrapbooks. Public Library <strong>of</strong> Camden and Ouachita County,<br />
Camden, <strong>Arkansas</strong>.
Smith, Bruce D.<br />
1975 Middle Mississippi Exploitation <strong>of</strong> Animal Populations. Museum <strong>of</strong> Anthropology,<br />
<strong>University</strong> <strong>of</strong> Michigan, Anthrpological Papers 57.<br />
1978 Variation <strong>in</strong> Mississippian Settlement Patterns. In Mississippian Settlement Patterns,<br />
edited by Bruce D. Smith, pp. 479-503. Academic Press, New York.<br />
Smith, J. Russell<br />
1953 Tree Crops: A Permanent Agriculture. Dev<strong>in</strong>-Adair, New York.<br />
Society <strong>of</strong> American Foresters<br />
1954 Forest Cover Types <strong>of</strong> North America (exclusive <strong>of</strong> Mexico). Society <strong>of</strong> American Foresters,<br />
Wash<strong>in</strong>gton, D. C.<br />
Spencer, Annie L.<br />
1953 The Blacks <strong>of</strong> Union County. <strong>Arkansas</strong> Historical Quarterly 12:225-252.<br />
Sprague, Marshall<br />
1974 So Vast So Beautiful a Land: Louisiana and <strong>the</strong> Purchase. Little, Brown, Boston.<br />
Spr<strong>in</strong>ger, James W.<br />
1980 An Analysis <strong>of</strong> Prehistoric Food Rema<strong>in</strong>s from <strong>the</strong> Bruly St. Mart<strong>in</strong> Site, Louisiana,<br />
with a Comparative Discussion <strong>of</strong> Mississippi Valley Faunal Studies. Midcont<strong>in</strong>enta1<br />
Journal <strong>of</strong> Archaeology 5:193-224.<br />
Stacy, V. K. Pheriba<br />
1976 Summary Report on Archeological Survey and Excavations <strong>in</strong> <strong>the</strong> Felsenthal National<br />
Wildlife Refuge. Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
State Land Office<br />
Tract Books. State Land Office, Little Rock.<br />
Suhm, Dee Ann, and Edward B. Jelks (editors)<br />
1962 Handbook <strong>of</strong> Texas Archeology: Type Descriptions. Texas Archeological Society,<br />
Special Publication 1 and Texas Memorial Museum, Bullet<strong>in</strong> 4.<br />
Suhm, Dee Ann, and Alex D. Krieger<br />
1954 An Introductory Handbook <strong>of</strong> Texas Archeology. Bullet<strong>in</strong> <strong>of</strong> <strong>the</strong> Texas Archeological<br />
Society 25.<br />
Swanton, John R.<br />
1911 Indian Tribes <strong>of</strong> <strong>the</strong> Lower Mississippi Valley and Ajacent Coast <strong>of</strong> <strong>the</strong> Gulf <strong>of</strong><br />
Mexico. Bureau <strong>of</strong> American Ethnology Bullet<strong>in</strong> 43.<br />
299
300<br />
Travis, Russell B.<br />
1955 Classification <strong>of</strong> Rocks. Quarterly <strong>of</strong> <strong>the</strong> Colorado School <strong>of</strong> M<strong>in</strong>es 50(1).<br />
Tucker, Gary E.<br />
1974 Threatened Native Plants <strong>of</strong> <strong>Arkansas</strong>. In <strong>Arkansas</strong> Natural Area Plan, pp. 39-65.<br />
<strong>Arkansas</strong> Department <strong>of</strong> Plan<strong>in</strong>g, Little Rock.<br />
Union County<br />
County Court Records. Union County Courthouse, El Dorado, <strong>Arkansas</strong>.<br />
U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers<br />
1871 Topographic Sketch <strong>of</strong> Ouachita River <strong>in</strong> <strong>Arkansas</strong> from Arkadelphia to <strong>the</strong> Sou<strong>the</strong>rn<br />
State L<strong>in</strong>e; Made under <strong>the</strong> Orders <strong>of</strong> Lieut. Colonel W. F. Raynolds by Just<strong>in</strong> Straszer, C. E.<br />
(6 sheets; scale pr<strong>in</strong>ted on maps). Vicksburg.<br />
1873 Ouachita River Survey (1:10,000; 34 sheets). Vicksburg.<br />
1875 Letters received by <strong>of</strong>ficer <strong>in</strong> charge <strong>of</strong> Ouachita River improvements stationed at<br />
Monroe, Louisiana, Vicksburg, Mississippi, and Memphis, Tennessee. PI entry <strong>17</strong>19,<br />
RG77, Federal Archives and Records Center, East Po<strong>in</strong>t, Georgia.<br />
1890 Map <strong>of</strong> Ouachita River Surveyed under <strong>the</strong> Direction <strong>of</strong> Gen. W. F. Raynolds, U.S. A., by<br />
Clement Smith, C. E. (1:30,000; 7 sheets). Vicksburg.<br />
1894 Bas<strong>in</strong> <strong>of</strong> Ouachita River and Tributaries from Arkadelphia to Red River (1:200,000; 3<br />
sheets). Vicksburg.<br />
1895- Ouachita River Survey (1:10,000; 48 sheets). Vicksburg.<br />
1897<br />
1896 Ouachita River Hydrographic Survey Maps. Vicksburg.<br />
1897 Index Map <strong>of</strong> Ouachita River Survey, Louisiana and <strong>Arkansas</strong>, Camden to Mouth <strong>in</strong> Five<br />
Sheets (scale pr<strong>in</strong>ted on maps). Vicksburg.<br />
1979 Feasibility Report, Ouachita-Black Rivers N<strong>in</strong>e-foot Navigation Project, Crossett Harbor,<br />
Ashley County, <strong>Arkansas</strong>. Vicksburg.<br />
1980 Cultural Resources Survey Report: Appendix to Design Memorandum No. 58.<br />
N<strong>in</strong>e foot Navigation Channel, Jonesville and Columbia Pools. Vicksburg District.<br />
n.d. Prelim<strong>in</strong>ary Plan for Slackwater Navigation, Ouachita River and Tributaries (1:300,000).<br />
Vicksburg.
U.S. Bureau <strong>of</strong> Mar<strong>in</strong>e Inspection and Navigation<br />
Enrollment Certificates, RG41. National Archives, Wash<strong>in</strong>gton D.C.<br />
U.S. Congress<br />
1872 Survey <strong>of</strong> Certa<strong>in</strong> Rivers <strong>in</strong> <strong>the</strong> Sou<strong>the</strong>rn States. 2nd Congress, 2nd Session, House<br />
Executive Document 25 .<br />
1874 Ouachita River. 43rd Congress, 1st Session, House Executive Document 259.<br />
1876 Report <strong>of</strong> <strong>the</strong> Chief <strong>of</strong> Eng<strong>in</strong>eers. 44th Congress, 2nd Session, House Executive<br />
Document 1(1):Part 2.<br />
1877 Report <strong>of</strong> <strong>the</strong> Chief <strong>of</strong> Eng<strong>in</strong>eers. 45th Congress, 2nd Session, House Executive<br />
Document 1(2):Part 2.<br />
1900 Ouachita and Black Rivers, <strong>Arkansas</strong> and Louisiana. 56th Congress, 1st Session,<br />
House Document 480.<br />
U.S. Department <strong>of</strong> Agriculture<br />
1951 Soil Survey Manual, Handbook No. 18. U.S. Government Pr<strong>in</strong>t<strong>in</strong>g Office, Wash<strong>in</strong>gton<br />
D. C.<br />
1968 General Soil Map, Union County, <strong>Arkansas</strong> (1:126, 20). U.S. Department <strong>of</strong> Agriculture,<br />
Soil Conservation Service.<br />
1978 Special Report, Fish and Wildlife Ouachita River Bas<strong>in</strong> Study, <strong>Arkansas</strong> and Louisiana.<br />
U.S. Department <strong>of</strong> Agriculture, Alexandria, Louisiana.<br />
U.S. Geological Survey<br />
1939 Felsenthal, <strong>Arkansas</strong>-Louisiana (1:62,500). U.S. Geological Survey, Wash<strong>in</strong>gton, D. C.<br />
1978 Felsenthal NE, NW, SE, and SW, Advance Workpr<strong>in</strong>ts (1:24,000). U.S. Geological Survey,<br />
Reston.<br />
U.S. Post Office<br />
n.d. Appo<strong>in</strong>tments <strong>of</strong> Postmasters, <strong>Arkansas</strong>. Ms. on micr<strong>of</strong>ilm, <strong>Arkansas</strong> History<br />
Commission, Little Rock.<br />
Veatch, A. C.<br />
1906 Geology and Underground Water Resources <strong>of</strong> Nor<strong>the</strong>rn Louisiana and Sou<strong>the</strong>rn<br />
<strong>Arkansas</strong>. U.S. Geological Survey Pr<strong>of</strong>essional Paper 46.<br />
Vita-F<strong>in</strong>zi, C., and E. S. Higgs<br />
1970 Prehistoric Economy <strong>in</strong> <strong>the</strong> Mount Carmel Area <strong>of</strong> Palest<strong>in</strong>e: Site Catchment<br />
Analysis. Proceed<strong>in</strong>gs <strong>of</strong> <strong>the</strong> Prehistoric Society 36:1-37.<br />
301
302<br />
Walker, W<strong>in</strong>slow<br />
1936 The Troyville Mounds, Catahoula Parish, Louisiana. Bureau <strong>of</strong> American Ethnology,<br />
Bullet<strong>in</strong> 133.<br />
Watk<strong>in</strong>s, Beverly<br />
1981 History and Historical Documentation. In Settlement Prediction <strong>in</strong> Sparta: A Location<br />
Analysis and Cultural Resource Assessment <strong>in</strong> <strong>the</strong> Uplands <strong>of</strong> Calhoun County,<br />
<strong>Arkansas</strong>, by Robert H. Lafferty III et al., <strong>Arkansas</strong> Archeological Survey Research<br />
Series 14.<br />
Weber, J. Cynthia<br />
1973 The Paw-Paw Site. Ms. on file, <strong>Arkansas</strong> Archeological Survey, Fayetteville.<br />
Webb, Clarence H.<br />
1959 The Belcher Mound: A Stratified Caddoan Site <strong>in</strong> Caddo Parish, Louisiana. Society<br />
for American Archaeology, Memoirs 11.<br />
1968 The Extent and Content <strong>of</strong> Poverty Po<strong>in</strong>t Culture. American Antiquity 33:297-321.<br />
1977 The Poverty Po<strong>in</strong>t Culture. Geoscience and Man 12.<br />
1980 Stone Tools and Po<strong>in</strong>ts <strong>of</strong> Northwestern Louisiana. Ms. on file with author,<br />
Shreveport.<br />
1981 Review <strong>of</strong> <strong>the</strong> Shallow Lake Site (3UN9/52) and Its Place <strong>in</strong> Regional Prehistory<br />
by Martha A. Rol<strong>in</strong>gson and Frank F. Schambach, Appendix C. <strong>Arkansas</strong> Archeological<br />
Survey Research Series 12.<br />
Webb, Clarence H., Joel Sh<strong>in</strong>er, and E. Wayne Roberts<br />
1971 The John Pearce Site (16CD56): A San Patrice Site <strong>in</strong> Caddo Parish, Louisiana. Bullet<strong>in</strong><br />
<strong>of</strong> <strong>the</strong> Texas Archeological Society 42:1-47.<br />
We<strong>in</strong>ste<strong>in</strong>, Richard A., and Philip G. Rivet<br />
1978 Beau Mire: A Late Tchula Period Site <strong>of</strong> <strong>the</strong> Tchefuncte Culture, Ascension Parish,<br />
Louisiana. Louisiana Archeological Survey and Antiquities Commission, Anthropological<br />
Report 1.<br />
Welcomme, Rob<strong>in</strong> L.<br />
1979 Fisheries Ecology <strong>of</strong> Floodpla<strong>in</strong> Rivers. Longmans, New York.<br />
Wharton, Charles H., and Mark M. Br<strong>in</strong>son<br />
1979 Characteristics <strong>of</strong> Sou<strong>the</strong>astern River Systems. Strategies for Protection and<br />
Management <strong>of</strong> Floodpla<strong>in</strong> Wetlands and O<strong>the</strong>r Riparian Ecosystems, coord<strong>in</strong>ate by<br />
R. Roy Johnson and J. Frank McCormick, pp. 32-40. U.S. Department <strong>of</strong> Agriculture,<br />
General Technical Report WO-12.
Wheat, Joe Ben<br />
1972 The Olsen-Chubbuck Site: A Paleo-Indian Bison Kill. Society for American Archaeology<br />
Memoir 26.<br />
White, Patsy<br />
1970 Investigation <strong>of</strong> <strong>the</strong> Cemetery at <strong>the</strong> Gee’s Land<strong>in</strong>g Site, 3DR<strong>17</strong>. <strong>Arkansas</strong> Archeologist<br />
11(1):1-20.<br />
Willey, Gordon R., and Philip Phillips<br />
1958 Method and Theory <strong>in</strong> American Archaeology. <strong>University</strong> <strong>of</strong> Chicago Press.<br />
Williams, Stephen<br />
1980 Armorel: A Very Late Phase <strong>in</strong> <strong>the</strong> Lower Mississippi Valley. Sou<strong>the</strong>astern Archaeological<br />
Conference Bullet<strong>in</strong> 22:105-110.<br />
Williamson, Frederick W., and Lillian Herron Williamson<br />
1939 Nor<strong>the</strong>ast Louisiana, a Narrative History <strong>of</strong> <strong>the</strong> Ouachita River Valley and <strong>the</strong> Concordia<br />
Country. Historical Record Association, Monroe, Louisiana.<br />
Wilmsen, Edw<strong>in</strong> H.<br />
1968 Lithic Analysis <strong>in</strong> Paleoanthropology. Science 161:982-987.<br />
W<strong>in</strong>ters, Howard Dalton<br />
1969 The Riverton Culture: A Second Millenium Occupation <strong>in</strong> <strong>the</strong> Central Wabash<br />
Valley. Ill<strong>in</strong>ois Archaeological Survey, Monograph 1, and Ill<strong>in</strong>ois State Museum, Reports <strong>of</strong><br />
Investigations 13.<br />
Wood, W. Dean<br />
1976 A Sampl<strong>in</strong>g Scheme for Subsurface Archeological Survey. Sou<strong>the</strong>astern Archaeological<br />
Conference Bullet<strong>in</strong> 19:39-42.<br />
Wood, W. Raymond<br />
1963 Two New Projectile Po<strong>in</strong>ts: Homan and Agee Po<strong>in</strong>ts. <strong>Arkansas</strong> Archeologist 4(2):1-<br />
6.<br />
Wood, W. Raymond, and Donald Lee Johnson<br />
1978 A Survey <strong>of</strong> Disturbance Processes <strong>in</strong> Archeological Site Formation. In Advances <strong>in</strong><br />
Archaeological Method and Theory Volume 1, edited by Michael B. Schiffer, pp. 315-381.<br />
Academic Press, New York.<br />
Wood, W. Raymond, and R. Bruce McMillan<br />
1976 Prehistoric Man and His Environments: A Case Study <strong>in</strong> <strong>the</strong> Ozark Highlands. Academic<br />
Press, New York.<br />
303
304<br />
WPA Author<br />
n.d. Ouachita County History by Works Project Adm<strong>in</strong>istration Federal Writers<br />
Project. Ms. on file, Ouachita County Place File, <strong>Arkansas</strong> History Commission, Little<br />
Rock.
Appendix A<br />
site descriptiVe And AnAlyticAl dAtA<br />
This appendix presents a summary tabulation <strong>of</strong> field data and results <strong>of</strong> analyses<br />
for 144 archeological sites recorded or revisited by <strong>the</strong> 1979 Felsenthal Project. These<br />
data are drawn from <strong>Arkansas</strong> Archeological Site Survey Forms which run to seven or<br />
more page for each <strong>of</strong> <strong>the</strong> 144 sites, and entries <strong>the</strong>refore represent <strong>the</strong> briefest characterization<br />
and assessment <strong>of</strong> each site.<br />
Table A-1 summarizes 132 prehistoric sites <strong>in</strong> <strong>the</strong> order <strong>in</strong>troduced <strong>in</strong> Chapter 5;<br />
that is, by seven dra<strong>in</strong>age and geomorphic units, and from upstream to downstream<br />
locations with<strong>in</strong> each unit. Table A-2 summarizes 15 historic sites <strong>in</strong> Unit 8, also <strong>in</strong> order<br />
from upstream to downstream location. Three sites hav<strong>in</strong>g both prehistoric and historic<br />
components appear <strong>in</strong> both tables (3BR8, 3BR81, 3UN92). The follow<strong>in</strong>g explanatory<br />
notes may assist <strong>the</strong> use <strong>of</strong> <strong>the</strong>se tables (also see Chapter 5).<br />
TABLE A-1, PREHISTORIC SITES<br />
Column (1): Tr<strong>in</strong>omial site designations assigned by <strong>Arkansas</strong> Archeological<br />
Survey; sites marked by asterisk were previously recorded and <strong>the</strong>n revisited<br />
by <strong>the</strong> 1979 Felsenthal Project.<br />
Column (2): Site names from previous records or assigned by project except<br />
to s<strong>in</strong>gle f<strong>in</strong>ds; some apparent clusters <strong>of</strong> sites <strong>in</strong>dicated by a name and consecutive<br />
numbers, e.g., Jug Po<strong>in</strong>t 1, Jug Po<strong>in</strong>t 2.<br />
Column (3): For all buried sites, field observations and 1978 topographic<br />
maps were comb<strong>in</strong>ed to provide <strong>the</strong> best estimate <strong>of</strong> elevation for upper<br />
and lower boundaries <strong>of</strong> cultural levels; elevations followed by <strong>the</strong> letter “I”<br />
<strong>in</strong>dicate that cultural rema<strong>in</strong>s are known only to derive from some unknown<br />
level or levels with<strong>in</strong> this <strong>in</strong>terval; for all sites <strong>in</strong> Unit 7 map elevations are<br />
given for <strong>the</strong> surface extent, regardless <strong>of</strong> buried rema<strong>in</strong>s that may be present.<br />
Column (4): River channel mileages above mouths <strong>of</strong> Black or Sal<strong>in</strong>e rivers to<br />
nearest tenth as <strong>in</strong>terpolated on USGS Felsenthal 15’ Ark-La (1939).<br />
Column (5): See discussions <strong>of</strong> landforms and channel geometry <strong>in</strong> Chapters<br />
2 and 5.
A-2<br />
Column (6): Dimensions <strong>of</strong> artifact scatters recorded <strong>in</strong> <strong>the</strong> field, rounded to<br />
nearest meter; for riverbank sites a width perpendicular to bankl<strong>in</strong>e is reported<br />
first, followed by a length parallel to bankl<strong>in</strong>e; no reliable dimensions<br />
were available for large complex sites previously recorded (Unit 7).<br />
Column (7): Depths <strong>of</strong> artifacts or cultural levels, measured <strong>in</strong> place from<br />
modern surface, rounded to nearest tenth <strong>of</strong> a meter.<br />
Column (8): Project numbers assigned <strong>in</strong> <strong>the</strong> field to artifacts scatters or s<strong>in</strong>gle<br />
f<strong>in</strong>ds, later comb<strong>in</strong>ed as sites <strong>in</strong> some cases.<br />
Columns (9)-(13): Totals for prehistoric artifact and debris categories, discussed<br />
<strong>in</strong> Chapters 6 and 8, and site totals.<br />
Column (14): Organic materials observed <strong>in</strong> newly recorded sites or <strong>in</strong>dicated<br />
<strong>in</strong> exist<strong>in</strong>g records for previously recorded sites.<br />
Column (15): Best estimate <strong>of</strong> component or components represented (see<br />
Chapter 5).<br />
Column (16): Recommended status for nom<strong>in</strong>ation to <strong>the</strong> National Register <strong>of</strong><br />
Historic Places, based on 1979 Felsenthal Project survey and test<strong>in</strong>g; one site,<br />
3UN121 <strong>in</strong> Site Unit 4, is listed as “<strong>in</strong>determ<strong>in</strong>ate” because it has been tested<br />
for significance subsequent to our work and results are not yet available (see<br />
Appendix C).<br />
TABLE A-2, HISTORIC SITES<br />
Column (1): Tr<strong>in</strong>om<strong>in</strong>al site designations assigned by <strong>Arkansas</strong> Archeological<br />
Survey; sites marked by asterisk were previously recorded and revisited by<br />
1979 Felsenthal Project.<br />
Column (2): Site names from previous records or assigned by project; historic<br />
placenames used where firm identifications made; e.g., Marie Sal<strong>in</strong>e Land<strong>in</strong>g.<br />
Column (3): Elevations to nearest foot for surface extent <strong>of</strong> artifacts scatters<br />
and structures obta<strong>in</strong>ed from 1978 topographic maps; underwater rema<strong>in</strong>s<br />
<strong>in</strong>dicated by elevations below 62.0 feet.<br />
Column (4): River channel mileage as <strong>in</strong> Table A-1.<br />
Column (5): Topographic locations as <strong>in</strong> Table A-1.<br />
Column (6): Dimensions <strong>of</strong> artifact scatters or, <strong>in</strong> some cases, stand<strong>in</strong>g structures<br />
recorded <strong>in</strong> <strong>the</strong> field, rounded to nearest meter.
Column (7): Project numbers assigned <strong>in</strong> <strong>the</strong> field to artifact scatters and<br />
structures, later comb<strong>in</strong>ed as sites <strong>in</strong> some cases.<br />
Columns (8)-(12): Totals for historic artifact and debris categories and site<br />
totals.<br />
Column (13): Organic materials observed <strong>in</strong> field.<br />
Column (14): Historic components with best estimate <strong>of</strong> occupation or use<br />
on a decade-by-decade basis; more precise dates are available <strong>in</strong> a few cases<br />
(Chapter 5).<br />
Column (15): Recommended status for nom<strong>in</strong>ation to <strong>the</strong> National Register <strong>of</strong><br />
Historic Places based on 1979 survey results ( (see Appendix C).<br />
A-3
A-4<br />
Table A-1. Prehistoric Sites.<br />
Unit 1: Ouachita River Above Mile 254<br />
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)<br />
State Topographic L<strong>in</strong>ear Prehistoric Component(s) National<br />
Site Elevation Location or River Dimensions Depth Project Stone Inorganic Organic Identified (cultural Register<br />
Number Site Name (feet MSL) (river mileage) Channel Location (meters) (meters) Number(s) Ceramics Tools Debitage Debris Total Rema<strong>in</strong>s period/phase) Eligibility<br />
3UN167 Mud Lake Bend 1 68.0-76.01 270.1 right concave bank 1 x 5 093 0 1 4 3 8 Archaic Not eligible<br />
3UN166 Mud Lake Bend 2 71.1-72.4 269.9 right concave bank 1 x 50 1.1-1.5 094, 095 0 3 6 52 61 Archaic Eligible<br />
3UN165 Mud Lake Bend 3 63.0-76.01 269.7 right concave bank 1 x 100 096 0 0 0 7 7 Archaic Not eligible<br />
3UN164 Stormhole Bend 62.0-65.01 268.9 right concave bank 1 x 1 104 0 0 0 2 2 unknown Not eligible<br />
3BR71 Parrigee<strong>the</strong> Shoals 1 63.0-70.01 268.2 left concave bank 1 x 2 105 0 0 0 21 21 Archaic Not eligible<br />
3BR33* Parrigee<strong>the</strong> Shoals 2 63.0-75.01 268.1 left concave bank 1 x 25 106 0 1 0 6 7 unknown Not eligible<br />
3UN163 – 71.5 267.7 right convex bank 1 x 1 1.0 121 0 0 1 0 1 Archaic Not eligible<br />
3BR75 Eutaw Rapids 74.0 267.5 left concave bank 1 x 13 0.3 111 1 0 0 0 1 shell lens Baytown-Coles Creek Not eligible<br />
3UN162 – 65.3 267.3 right concave bank 1 x 1 3.0 123 0 0 0 1 1 Archaic Not eligible<br />
3BR73 Small Cane 2 71.0 266.2 left concave bank 2 x 23 1.2 109 0 1 1 10 12 late Archaic Eligible<br />
3BR74 – 72.0-75.01 266.1 left concave book 1 x 1 110 0 0 1 0 1 unknown Not eligible<br />
3BR72 Small Cane 1 70.0-73.51 266.0 left concave bank 2 x 35 107, 108 0 3 1 11 15 charcoal Archaic Eligible<br />
3UN161 Fletchers Land<strong>in</strong>g 74.0 265.4 right concave bank 1 x 12 0.3 124 1 0 1 1 3 charcoal Baytown-Coles Creek Not eligible<br />
3BR70 Hunter’s Swan 70.0-73.01 265.0 left convex bank 1 x 1 091 8 0 0 0 8 Tchula/Coon Island Not eligible<br />
3BR57 Little Bay Rapids 63.0-70.01 261.8 left concave bank 1 x 6 012 0 0 1 1 2 unknown Not eligible<br />
3UN188 Five Mile Bend 62.0-70.01 257.2 right concave bank 1 x 5 007 5 0 0 1 6 Baytown-Coles Creek Not eligible<br />
3UN185 – 63.0-71.01 255.0 right concave bank 1 x 1 023 0 1 0 0 1 unknown Not eligible<br />
3BR77 Lower Sal<strong>in</strong>e Cut<strong>of</strong>f 67.0-70.07 254.3 left concave bank 2 x 42 128 0 0 1 4 5 unknown Not eligible
Table A-1 (cont<strong>in</strong>ued). Prehistoric Sites.<br />
Unit 2: Lower Sal<strong>in</strong>e River<br />
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)<br />
State Topographic L<strong>in</strong>ear Prehistoric Component(s) National<br />
Site Elevation Location or River Dimensions Depth Project Stone Inorganic Organic Identified (cultural Register<br />
Number Site Name (feet MSL) (river mileage) Channel Location (meters) (meters) Number(s) Ceramics Tools Debitage Debris Total Rema<strong>in</strong>s period/phase) Eligibility<br />
3BR92 Water Locust 65.0-67.0 9.9 left concave bank 2 x 105 0.8-1.1 233 0 0 4 37 41 Archaic Eligible<br />
3BR91 – 63.0-70.01 9.1 left concave bank 1 x 1 216 0 0 0 1 1 unknown Not eligible<br />
3BR90 Round Turn 63.0-66.01 8.3 right concave bank 1 x 40 212, 213 0 0 1 1 2 Archaic Not eligible<br />
3BR89 Mouth <strong>of</strong> Miller Creek 65.0-70.01 8.0 right concave bank 1 x 20 211 1 0 0 1 2 Baytown-Coles Creek Not eligible<br />
3BR88 – 67.0-70.01 7.8 left concave bank 1 x 1 150 0 1 0 0 1 unknown Not eligible<br />
3BR87 Horton Island Land<strong>in</strong>g 68.0-70.01 7.3 right concave bank 2 x 19 149 8 0 2 5 15 Miss./Gran <strong>Marais</strong> Not eligible<br />
3BR86 Sweetgum 2 67.5 7.0 right concave bank 10 x 200 0.2-0.3 146, 147 15 2 3 32 52 Mississippi Eligible<br />
3BR85 Sweetgum 1 63.0-69.01 6.6 right convex bank 1 x 1 145 0 1 1 1 3 Archaic Not eligible<br />
3AS318 Hawthorn 3 63.0-70.01 6.3 left concave bank 2 x 24 144 0 0 2 2 4 unknown Not eligible<br />
3AS3<strong>17</strong> Hawthorn 2 63.5-64.5 6.0 left concave bank 2 x 10 1.7-1.9 142 0 1 0 7 7 Archaic Not eligible<br />
3AS316 Hawthorn 1 64.0-67.1 5.7 left concave bank 2 x 27 0.4-1.3 141 9 0 0 1 10 charcoal Baytown-Coles Creek Eligible<br />
3BR81 Prairie Island Land<strong>in</strong>g 66.0-68.01 5.1 right concave bank 1 x 1 137 0 0 1 2 3 unknown Not eligible<br />
3BR80 Prairie Island Bend 67.0-68.0 5.0 right concave bank 2 x 40 0.2 136 1 0 2 13 16 Mississippi Not eligible<br />
3AS315 Persimmon 3 68.0-70.01 4.9 levee near left bank 3 x 90 135 0 2 24 29 55 Mississippi Eligible<br />
3BR82 Bitter Pecan 1 68.0-69.0 4.4 right convex bank 3 x 45 0.3 138 10 0 4 12 26 charcoal/bone Baytown-Coles Creek Eligible<br />
3AS314 Persimmon 2 63.0-70.01 4.4 left concave bank <strong>17</strong> x 20 134 0 2 2 20 24 Mississippi Not eligible<br />
3BR84 Bitter Pecan 3 63.0-69.01 4.4 levee near right bank 1 x 5 140 1 1 0 1 3 unknown Not eligible<br />
3BR83 Bitter Pecan 2 63.0-69.01 4.2 right straight bank 2 x 50 139 0 0 0 3 3 unknown Not eligible<br />
3AS313 Persimmon 1 63.0-66.01 4.1 left concave bank 3 x 75 133 0 2 7 4 13 Poverty Po<strong>in</strong>t/Calion Not eligible<br />
3BR78 Mouth <strong>of</strong> Eagle Creek 63.2-68.0 3.9 right concave bank 5 x 49 0.1-0.7 131 0 1 32 <strong>17</strong>* 50 charcoal Poverty Po<strong>in</strong>t/Calion Eligible (tested)<br />
3AS312 Willow Oak 66.0-70.01 3.7 left concave bank 1 x 2 130 0 1 0 1 2 unknown Not eligible<br />
3AS311 Tupelo Gum 64.0-68.01 2.9 levee near left bank 2 x 20 129 0 0 0 5 5 unknown Not eligible<br />
3AS309 Jug Po<strong>in</strong>t 4 63.0-68.01 2.7 left concave bank 3 x 20 1<strong>17</strong> 0 0 4 0 4 unknown Not eligible<br />
3AS310 – 67.5 2.7 left concave bank 1 x 1 0.1 118 0 1 0 0 1 Miss./Term<strong>in</strong>al Not eligible<br />
3AS308 Jug Po<strong>in</strong>t 3 66.0-67.01 2.6 left convex bank 3 x 18 116 4 0 16 1 21 Miss./Caney Bayou Not eligible<br />
3AS307 Jug Po<strong>in</strong>t 2 67.2-68.3 2.6 left concave bank 8 x 80 0.1-0.3 115 98 8 66 14* 186 charcoal/bone Miss./Caney Bayou Eligible (tested)<br />
3AS306 Jug Po<strong>in</strong>t 1 67.2-68.1 2.5 left concave bank 9 x 75 0.1-0.3 114 62 23 677 121* 883 charcoal/bone Miss./Caney Bayou Eligible (tested)<br />
3BR76 Jug Po<strong>in</strong>t Cut<strong>of</strong>f 67.6-68.4 2.1 right concave bank 11 x <strong>17</strong> 0.1-0.4 119 250 2 6 601* 859 midden Miss./Gran <strong>Marais</strong> Eligible (tested)<br />
3AS305 Overcup Oak 3 66.0-67.07 1.6 left concave bank 13 x 51 103 35 5 84 15 139 Miss./Caney Bayou Eligible<br />
3AS304 Overcup Oak 2 66.0-67.01 1.3 left convex bank 3 x 15 102 12 0 7 0 10 Miss./Gran <strong>Marais</strong> Not eligible<br />
3AS303 Overcup Oak 1 66.0-67.01 1.1 left convex bank 5 x 60 101 14 5 27 10 56 Miss./Gran <strong>Marais</strong> Not eligible<br />
3AS302 Sycamore 66.0-67.01 0.9 levee near left bank 1 x 2 100 0 0 1 1 2 unknown Not eligible<br />
3AS286 One Cypress Po<strong>in</strong>t 66.6-68.4 0.8 levee near left bank 16 x 58 0.1-0.2 024 84 <strong>17</strong>8 1084 140* 1486 floral/bone Miss./Term<strong>in</strong>al Eligible (tested)<br />
3BR59 Possumhaw 64.0-66.01 0.8 levee near right bank 2 x 3 026 0 0 3 0 3 unknown Not eligible<br />
3BR58 Buttonbush 67.2-67.3 0.8 right convex bank 4 x 98 0.2 025 454 3 8 14* 479 Baytown-Coles Creek Not eligible<br />
(tested)<br />
3BR60 – 61.0-64.01 0.6 right straight bank 1 x 1 027 0 0 0 1 1 unknown Not eligible<br />
3BR61 – 63.0-64.01 0.4 levee near right bank 1 x 1 028 0 0 1 0 1 unknown Not eligible<br />
3BR62 Mouth <strong>of</strong> Long Slough 65.0-66.0 0.4 right convex bank 7 x 10 0.1-0.3 029 13 0 1 2 16 charcoal/bone Miss./Gran <strong>Marais</strong> Eligible<br />
3AS285 False Indigo 66.9-69.9 0.3 left straight bank 12 x 40 0.1-0.5 021 1663 9 97 4089* 4848 midden Tchula/Coon Island Eligible (tested)<br />
Miss./Gran <strong>Marais</strong><br />
3AS284 Mouth <strong>of</strong> Sal<strong>in</strong>e River 63.0-67.01 0.0 left convex bank 2 x 10 019 0 0 1 4 5 unknown Not eligible<br />
A-5
A-6<br />
Table A-1 (cont<strong>in</strong>ued). Prehistoric Sites.<br />
Unit 3: Lower Eagle Creek<br />
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)<br />
State Topographic L<strong>in</strong>ear Prehistoric Component(s) National<br />
Site Elevation Location or River Dimensions Depth Project Stone Inorganic Organic Identified (cultural Register<br />
Number Site Name (feet MSL) (river mileage) Channel Location (meters) (meters) Number(s) Ceramics Tools Debitage Debris Total Rema<strong>in</strong>s period/phase) Eligibility<br />
3BR68 – 66.0-68.01 – right concave bank 1 x 1 065 0 1 0 0 1 unknown Not eligible<br />
3BR67 – 66.0-68.01 – right concave hank 1 x 1 084 1 0 0 0 1 Mississippi Not eligible<br />
38866 Eagle Creek 4 66.0-68.01 – right convex bank 15 x 15 083 56 0 0 41 97 charcoal/bone Miss./Gran <strong>Marais</strong> Eligible<br />
3BR69 Eagle Creek 5 66.0-68.01 – left concave bank 5 x 50 086 4 1 1 0 6 Mississippi Not eligible<br />
3BR65 Eagle Creek 3 67.5 – right convex bank 15 x 80 0.1 058 16 0 15 29 60 charcoal Miss./Caney Bayou Eligible<br />
38864 Eagle Creek 2 63.0-69.01 – right concave bank 1 x 40 056, 057 0 0 2 0 2 unknown Not eligible<br />
38863 Eagle Creek 1 63.0-69.01 – right convex bank 3 x 125 055 5 1 9 7 21 Mississippi Not eligible<br />
3BR79 Eagle Creek 6 65.0 – left concave bank 3 x 20 0.1-0.3 132 0 3 26 12 41 Mississippi Eligible<br />
Unit 4: Ouachita River Below Mile 254<br />
3UN184 – 68.0-70.01 254.1 right concave bank 1 x 1 030 1 0 0 0 1 Baytown-Coles Creek Not eligible<br />
3UN92* Brown Camp below 71.01 252.9 levee near right bank 022 4 1 4 0 9 Mississippi Not eligible<br />
3AS329 Marie Sal<strong>in</strong>e 62.0-67.5 252.5 left concave bank 14 x 200 0.0-1.5 001 97 32 477 350• 956 charcoal Poverty Po<strong>in</strong>t/Calion Eligible<br />
Tchula/Coon Island (tested)<br />
Baytown-Coles Creek<br />
Miss./Caney Bayou<br />
3AS326 – 68.0-69.01 252.4 left concave bank 1 x 1 231 0 1 0 0 1 Miss./Gran <strong>Marais</strong> Not eligible<br />
3A5301 Honey Locust 66.0-69.01 252.0 left concave bank 2 x 20 090, 219 4 0 0 2 6 Baytown-Coles Creek Not eligible<br />
3AS319 – 66.0 251.9 left concave bank 1 x 1 0.3 220 0 0 0 1 1 unknown Not eligible<br />
3AS320 River Birch 1 63.0-67.01 251.5 left concave bank 1 x 3 222 0 0 0 4 4 unknown Not eligible<br />
3AS321 River Birch 2 63.0-67.01 251.4 left straight bank 2 x 40 223 8 1 1 1 11 Tchula/Coon Island Eligible<br />
3AS288 Red Oak 67.0-68.01 250.5 levee near left bank 2 x 4 032 0 0 0 2 2 unknown Not eligible<br />
3AS322 Water Elm 1 67.5 250.4 left concave bank 1 x 1 0.4 225 8 0 0 0 8 Mississippi/Caddo 2 Eligible<br />
3AS332 Water Elm 2 63.0-70.01 250.4 left concave bank 2 x 50 018 0 0 1 3 4 unknown Not eligible<br />
3A5287 Water Elm 3 67.0 250.2 left concave bank 3 x 5 0.1 031 277 0 1 4 302 Baytown-Coles Creek Eligible<br />
3AS323 Water Elm 4 68.0 250.0 left convex bank 1 x 1 0.5-0.6 226 52 0 0 0 52 Baytown-Coles Creek Not eligible<br />
3A5324 – 63.0 247.7 left convex bank 1 x 1 2.0 227 0 1 0 0 1 late Archaic Not eligible<br />
3UN123* Mouth <strong>of</strong> Lapile Creek 62.0-65.01 245.4 right concave bank 2 x 30 077 27 0 1 0 28 Baytown-Coles Creek Not eligible<br />
3AS327 Swamp Privet 66.0 245.0 levee near left bank 2 x <strong>17</strong> 0.5 232 1 3 16 1 21 Baytown-Coles Creek Eligible<br />
3A5325 – 65.5 244.8 left concave bank 1 x 1 0.6 229 1 0 0 0 1 Baytown-Coles Creek Not eligible<br />
3UN169 – 64.0-65.01 244.4 levee near right bank 1 x 1 078 0 1 0 0 1 unknown Not eligible<br />
3UN168 Coon Glory Bend 63.0-65.01 244.2 levee near right bank 20 x 25 080 14 1 67 1 83 Mississippi/Gran Eligible<br />
3UN121* Mo-Pac 67.5 242.3 right concave bank 1 x 10 0.7-0.9 230 5 0 0 2 7 charcoal Baytown-Coles Creek Indeterm<strong>in</strong>ate
Table A-1 (cont<strong>in</strong>ued). Prehistoric Sites.<br />
Unit 5: Lower Lapile Creek<br />
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)<br />
State Topographic L<strong>in</strong>ear Prehistoric Component(s) National<br />
Site Elevation Location or River Dimensions Depth Project Stone Inorganic Organic Identified (cultural Register<br />
Number Site Name (feet MSL) (river mileage) Channel Location (meters) (meters) Number(s) Ceramics Tools Debitage Debris Total Rema<strong>in</strong>s period/phase) Eligibility<br />
3UN<strong>17</strong>8 Lapile Creek 1 68.0-72.01 – left concave bank 30 x 60 059 1 2 11 0 14 unknown Not eligible<br />
3UN160 Lapile Creek 2 68.0-72.01 – left concave bank 10 x 320 151-154, <strong>17</strong>5 1 3 16 2 22 Baytown-Coles Creek Not eligible<br />
3UN159 Lapile Creek 3 68.0-72.01 – opposite concave/ 10 x 160 155, 168-169 25 0 1 0 26 Baytown-Coles Creek Not eligible<br />
convex banks<br />
3UN158 Lapile Creek 4 68.0-71.01 – opposite concave/ 10 x 300 156-159, <strong>17</strong>0, 5 6 29 8 48 Baytown-Coles Creek, Not eligible<br />
convex banks <strong>17</strong>4 Mississippi<br />
3UN157 Lapile Creek 5 65.0-69.01 – opposite concave/ 10 x 165 160-161, <strong>17</strong>1 26 4 10 5 45 Baytown-Coles Creek Eligible<br />
– convex banks<br />
3UN156 Lapile Creek 6 65.0-68.01 – opposite concave/ 165 x 190 162-164, <strong>17</strong>2 0 1 24 15 40 unknown Not eligible<br />
convex banks<br />
3UN152 Lapile Creek 8 64.0-67.01 – right concave bank 1 x 3 <strong>17</strong>3 0 2 1 0 3 Archaic Not eligible<br />
3UN151 Lapile Creek 7 64.0-67.01 – left concave bank 5 x 200 165-167 105 5 33 2 145 Archaic, Baytown- Eligible<br />
Coles Creek<br />
Unit 6: Oxbow Lakes and Backswamp<br />
3UN187 St. Mary’s Lake North 69.5 – oxbow lake bank 5 x 40 0.4 013 0 0 4 5 9 unknown Not eligble<br />
3UN186 – 64.0-71.01 – oxbow lake bank 1 x 1 014 0 0 1 0 1 unknown Not eligible<br />
3UN<strong>17</strong>0 Borrow Pit near 63.0-65.01 – back swamp 1 x 10 076 0 0 3 2 5 unknown Not eligible<br />
Lapoile Creek<br />
3UN183 Lapoile Creek 62.0-64.01 – opposite concave/ 1 x 2 046, 047 0 1 1 1 3 unknown Not eligible<br />
– convex banks<br />
3UNI82 Fishtrap Lake 1 63.0-65.01 – right concave bank 5 x 5 048 0 1 4 1 6 unknown Not eligible<br />
3UN180 Fishtrap Lake 2 63.0-66.01 – left concave bank 2 x 3 049 4 0 1 1 6 Baytown-Coles Creek Eligible<br />
3UN181 Fishtrap Lake 3 63.0-65.01 – right convex bank 1 x 100 050 0 0 0 3 3 unknown Not eligible<br />
3UN<strong>17</strong>9 Fishtrap Lake 4 63.0-68.01 – left concave bank 2 x 2 051 7 0 0 1 8 Mississippi Eligible<br />
3AS295 Borrow Pit near 82 68.0-71.01 – backswamp/terrace 3 x 30 043, 044 4 2 4 2 12 Baytown-Coles Creek Not eligible<br />
Bridge 1 edge<br />
3AS294 Borrow Pit near 82 62.0-70.01 – backswamp/terrace 2 x 3 045 0 0 5 3 8 unknown Not eligible<br />
Bridge 2 edge<br />
3AS292 Wildcat Lake 62.0-63.01 – oxbow lake bank 5 x 15 019 0 0 4 10 14 unknown Not eligible<br />
A-7
A-8<br />
Table A-1 (concluded). Prehistoric Sites.<br />
Unit 7: Pleistocene Terraces and Islands<br />
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)<br />
State Topographic L<strong>in</strong>ear Prehistoric Component(s) National<br />
Site Elevation Location or River Dimensions Depth Project Stone Inorganic Organic Identified (cultural Register<br />
Number Site Name (feet MSL) (river mileage) Channel Location (meters) (meters) Number(s) Ceramics Tools Debitage Debris Total Rema<strong>in</strong>s period/phase) Eligibility<br />
3BR4* Eagle Lake Mounds 85.0-95.0 – Deweyville 3 Terrace 237 midden Mississippi (see Appen-<br />
3BR8* Goulett Island 63.0-86.0 – Deweyville 2 Terrace 2<strong>17</strong>, 218 midden Archaic-Mississippi dix C).<br />
3UN81* Jones Lake North 66.0-90.0 – Deweyville 2/3 Terrace 081, 082 midden Archaic-Mississippi<br />
3AS296 Brushy Creek 1 73.0-74.0 – Deweyville 2 Terrace 3 x 10 072 0 0 5 1 6 unknown<br />
3AS297 Brushy Creek 2 67.0-70.0 – Deweyville 3 Terrace 3 x 200 073 1 0 12 1 13 Baytown-Coles Creek<br />
3AS298 Brushy Creek 3 68.0 – Deweyville 2/3 Terrace 3 x 12 074 0 0 3 4 7 unknown<br />
3UN18* WattsField 80.0-90.0 – Deweyville 2 Terrace 180 midden Archaic-Mississippi<br />
3UN8* Locust Ridge 90.0 – Deweyville 2 Terrace <strong>17</strong>9 midden Archaic-Mississippi<br />
3AS328 Greenbrier 115.0 – Deweyville 1 Terrace 60 x120 234 45 3 7 5 60 midden Miss./Gran <strong>Marais</strong><br />
3AS289 Parker Ridge 70.0 – Deweyville 3 Terrace 15 x 150 034 9 3 20 16 48 Baytown-Coles Creek<br />
3AS160* Potlatch Field 69.0-74.0 – Deweyville 3 Terrace 037, 038 Archaic-Mississippi<br />
3AS293 Redeye Lake Prairie 73.0-80.0 – Deweyville 2/3 Terrace 125 x 200 041, 042 91 6 108 42 247 Miss./Gran <strong>Marais</strong><br />
3UN72* Crossroad near Burnt 75.0-83.0 – Deweyville 2 Terrace 3 x 150 065 2 2 6 2 12 Archaic, Baytown-<br />
Bridge Coles Creek<br />
3UN<strong>17</strong>7 P<strong>in</strong>e Ridge 1 85.0-86.0 – Deweyville 2 Terrace 3 x 150 061 1 3 7 1 12 Miss./Gran <strong>Marais</strong><br />
3UN<strong>17</strong>6 P<strong>in</strong>e Ridge 2 85.0 – Deweyville 2 Terrace 3 x 50 062 1 2 12 0 15 unknown<br />
3UN<strong>17</strong>5 Bold<strong>in</strong>g Road 80.0 – Deweyville 2 Terrace 3 x 10 064 0 1 0 1 2 late Archaic<br />
3UN9/52* Shallow Lake 85.0-87.0 – Deweyville 2 Terrace 054, 063, 067 midden Archaic-Mississippi<br />
3UN154 Shallow Lake West 85.0 – Deweyville 2 Terrace 5 x 20 <strong>17</strong>7 0 1 2 0 3 Archaic<br />
3UN34* Pendleton Camp 76.0-82.0 – Deweyville 2 Terrace 052, 053 midden Archaic-Mississippi<br />
3AS291 Black Slough 67.0-68.0 – Deweyville 3 Terrace 3 x 5 036 0 1 0 2 3 Archaic<br />
3AS290 P<strong>in</strong>e Island Road 73.0-74.0 – Deweyville 3 Terrace 20 x 110 035 0 8 47 41 96 Archaic<br />
3UN<strong>17</strong>3 Kelly Spr<strong>in</strong>g Creek 1 65.0-66.0 – Deweyville 3 Terrace 3 x 90 068, 075 2 2 41 2 47 Baytown-Coles Creek,<br />
3UN<strong>17</strong>2 Kelly Spr<strong>in</strong>g Creek 2 73.0 – Deweyville 2 Terrace 3 x 10 069 1 0 3 1 5 Mississippi<br />
y unknown-Coles Creek<br />
3UN<strong>17</strong>1 Gum Ridge Road 76.0 – Deweyville 2 Terrace 3 x 25 070 0 0 5 1 6 unknown<br />
3UN132* Burnie Creek 73.0-78.0 – Deweyville 2 Terrace 071 Archaic-Mississippi<br />
3AS1* Sulphur Spr<strong>in</strong>gs Mound 96.0 – Prairie Terrace 236 Baytown-Coles Creek,<br />
Mississippi<br />
3A56* Big Mound Ridge 73.0-75.0 – Deweyville 3 Terrace 238 midden Archaic-Mississippi
Table A-2. Historic Sites.<br />
Unit 8: Historic Sites<br />
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)<br />
State L<strong>in</strong>ear Historic National<br />
Site Elevation Location River Dimensions Project Organic Component(s) Register<br />
Number Site Name (feet MSL) (river mileage) Channel Location (meters) Number(s) Ceramics Glass Metal Lithic Total Rema<strong>in</strong>s Identified Eligibility<br />
3BR32* Ganer Mound 75.0-87.0 267.9 levee near left bank 50 x 50 097 0 0 1 0 1 oyster shell American 1910s Not eligible<br />
3BR56 Caney Marie Land<strong>in</strong>g 74.0-76.0 263.5 left concave bank 1 x 1 008 1 0 1 1 3 American late 1800s-early 1900s Eligible<br />
3BR8* Goulett Island 63.0-86.0 9.4 Deweyville 2 Terrace 2<strong>17</strong>, 218 1 1 1 1 4 structures, American 1850s-present Eligible<br />
refuse<br />
3BR81 Prairie Island Land<strong>in</strong>g 70.0 5.1 right concave bank 137 0 0 0 0 0 American 1920s-1930s Not eligible<br />
3BR55 Keelboat Brake Sawmill 66.0-68.0 – backswamp 004 0 0 1 0 1 charcoal American late 1800s-early 1900s Eligible<br />
3UN155 Ouachita Belle Land<strong>in</strong>g 70.0 257.8 right concave bank <strong>17</strong>8 0 0 0 0 0 American 1870s Not eligible<br />
3UN88* Fletchers Land<strong>in</strong>g and 63.0-75.0 265.7 right concave bank 40 x 60 125, 126 0 0 0 6 6 shell American 1870s-1890s Not eligible<br />
Woodyard<br />
3UN92* Brown Camp 71.0-90.0 252.9 levee near right bank 022 0 0 0 0 0 American 1920s-present Not eligible<br />
3AS330 First Slough 64.0-65.0 252.3 left concave bank 4 x 5 002 1 1 14 5 21 shell American late 1800s-early 1900s Not eligible<br />
3AS331 Marie Sal<strong>in</strong>e Ferry 62.0-68.0 252.2 left concave bank 2 x 4 003 1 0 0 0 1 American late 1800s-early 1900s Not eligible<br />
3AS300 South <strong>of</strong> Highway 82 Bridge 69.0-70.0 252.0 left concave bank 1 x 40 088, 089 1 1 0 0 2 shell American late 1800s-early 1900s Not eligible<br />
3AS299 Marie Sal<strong>in</strong>e Land<strong>in</strong>g 61.0-69.0 251.9 left concave bank 5 x 200 087 0 0 0 0 0 American 1850s-early 1900s Eligible<br />
3UN<strong>17</strong>4 Artesian Well 65.0 – backswamp 2 x 2 066 0 0 0 0 0 American early 1900s Not eligible<br />
3UN122* <strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g 64.0-65.0 244.0 right concave bank 25 x 40 079 0 0 0 0 0 wood American late 1800s Eligible<br />
3UN153 Wreck <strong>of</strong> <strong>the</strong> Lotawanna 34.0-38.0 243.9 river bottom <strong>17</strong>6 0 0 0 0 0 American 1860s-1870s Eligible<br />
A-9
A-10<br />
Table A-3. Occurrence <strong>of</strong> Temporally or Culturally Diagnostic Artifacts <strong>in</strong> Prehistoric Floodpla<strong>in</strong> Sites.<br />
Ceramic Complexes Projectile Po<strong>in</strong>ts<br />
O<strong>the</strong>r Dist<strong>in</strong>ctive<br />
Number Site Name Tchefuncte Baytown Gran <strong>Marais</strong> Shell tempered O<strong>the</strong>r Ceramics Arrowpo<strong>in</strong>ts Dart Po<strong>in</strong>ts Artifacts or Debris<br />
Unit l: Ouachita River Above Mile 254<br />
3UN167 Mud Lake Bend 1 1 ground stone tool fragment<br />
3UN166 Mud Lake Bend 2 1 heavy scraper or plane,<br />
36 fire-cracked rock<br />
3UN165 Mud Lake Bend 5 fire-cracked rock<br />
3UN164 Stormhole Bend 1 fire-cracked rock<br />
3BR71 Parrigee<strong>the</strong> Shoals 1 12 fire-cracked rock<br />
3BR33 Parrigee<strong>the</strong> Shoals 2 3 fire-cracked rock<br />
3UN163 –<br />
3BR75 Eutaw Rapids 1 Coles Creek<br />
Incised rim<br />
3UN162 – 1 fire-cracked rock<br />
3BR73 Small Cane 2 1 Macon 10 fire-cracked rock<br />
3BR74 –<br />
3BR72 Small Cane 1 1 Honey Creek 1 hafted scraper,<br />
1 hammerstone,<br />
5 fire-cracked rock<br />
3UN161 Fletchers Land<strong>in</strong>g 1 Baytown Pla<strong>in</strong> 1 fire-cracked rock<br />
body<br />
3BR70 Hunter’s Swan 2 Tchefuncte<br />
Stamped body<br />
3BR57 Little Bay Rapids<br />
3UN188 Five Mile Bend 3 Baytown Pla<strong>in</strong> 1 untyped pla<strong>in</strong> 1 untyped 1 fire-cracked rock<br />
body body <strong>in</strong>cised rim<br />
3UN185 – 1 abrader<br />
3BR77 Lower Sal<strong>in</strong>e Cut<strong>of</strong>f 4 fire-cracked rock<br />
Unit 2: Lower Sal<strong>in</strong>e River<br />
3BR92 Water Locust 35 fire-cracked rock<br />
3BR91 – 1 fire-cracked rock<br />
3BR90 Round Turn 1 fire-cracked rock<br />
3BR89 Mouth <strong>of</strong> Miller Creek 1 Baytown Pla<strong>in</strong> 1 fire-cracked rock<br />
body<br />
3BR88 – 1 bifacial preform<br />
or chopper
Table A-3 (cont<strong>in</strong>ued). Occurrence <strong>of</strong> Artifacts.<br />
Ceramic Complexes Projectile Po<strong>in</strong>ts<br />
O<strong>the</strong>r Dist<strong>in</strong>ctive<br />
Number Site Name Tchefuncte Baytown Gran <strong>Marais</strong> Shell tempered O<strong>the</strong>r Ceramics Arrowpo<strong>in</strong>ts Dart Po<strong>in</strong>ts Artifacts or Debris<br />
3BR87 Horton Island Land<strong>in</strong>g 2 Baytown Pla<strong>in</strong> body 1 fire-cracked rock<br />
3BR86 Sweetgum 2 8 Baytown Pla<strong>in</strong> body 1 Madison 10 fire-cracked rock<br />
3BR85 Sweetgum 1 1 hammerstone<br />
3AS318 Hawthorn 3 1 fire-cracked rock<br />
3AS3<strong>17</strong> Hawthorn 2 7 fire-cracked rock<br />
3AS316 Hawthorn l 8 Baytown Pla<strong>in</strong> body<br />
3BR81 Prairie Island Land<strong>in</strong>g 2 fire-cracked rock<br />
3BR80 Prairie Island Bend 1 untyped 3 fire-cracked rock<br />
<strong>in</strong>cised body<br />
3AS315 Persimmon 3 1 untyped arrow- 1 arrowpo<strong>in</strong>t preform,<br />
po<strong>in</strong>t fragment 26 fire-cracked rock<br />
3BR82 Bitter Pecan 1 5 Baytown Pla<strong>in</strong> body, 9 fire-cracked rock<br />
4 Baytown Pla<strong>in</strong> rim<br />
3AS314 Persimmon 2 1 untyped arrow- <strong>17</strong> fire-cracked rock<br />
po<strong>in</strong>t fragment<br />
3BR84 Bitter Pecan 3 1 hammerstone<br />
3BR83 Bitter Pecan 2 3 fire-cracked rock<br />
3AS313 Persimmon 1 1 gorget fragment,<br />
1 hammerstone,<br />
4 fire-cracked rock<br />
3BR78 Mouth <strong>of</strong> Eagle Creek 1 Gary 3 fire-cracked rock<br />
(see also Chapter 6)<br />
3AS312 Willow Oak 1 bifacial preform<br />
3AS311 Tupelo Gum 2 fire-cracked rock<br />
3AS309 Jug Po<strong>in</strong>t 4<br />
3AS310 – 1 Nodena<br />
3AS308 Jug Po<strong>in</strong>t 3 4 untyped pla<strong>in</strong><br />
body<br />
3AS307 Jug Po<strong>in</strong>t 2 25 Baytown Pla<strong>in</strong> body 73 untyped pla<strong>in</strong> 2 Ashley, 1 expanded base drill<br />
body 1 Bassett 4 fire-cracked rock<br />
(see also Chapter 6)<br />
3AS306 Jug Po<strong>in</strong>t 1 1 Baytown Pla<strong>in</strong> body 58 untyped pla<strong>in</strong> 2 Eagle Creek, 1 amber (?) bead,<br />
body 1 untyped 8 fire-cracked rock<br />
arrowpo<strong>in</strong>t fragment (see also Chapter 6)<br />
3BR76 Jug Po<strong>in</strong>t Cut<strong>of</strong>f <strong>17</strong>1 Baytown Pla<strong>in</strong> body, 13 untyped <strong>in</strong>- 45 untyped pla<strong>in</strong> 2 untyped bone 1 Ashley 1 perforator,<br />
3 Baytown Pla<strong>in</strong> rim cised or punctated body, 2 untyped tempered pla<strong>in</strong> 2 fire-cracked rock<br />
body, 3 untyped pla<strong>in</strong> rim, 10 un- body (see also Chapter 6)<br />
<strong>in</strong>cised rim typed <strong>in</strong>cised or<br />
punctated body<br />
A-11
A-12<br />
Table A-3 (cont<strong>in</strong>ued). Occurrence <strong>of</strong> Artifacts.<br />
Ceramic Complexes Projectile Po<strong>in</strong>ts<br />
O<strong>the</strong>r Dist<strong>in</strong>ctive<br />
Number Site Name Tchefuncte Baytown Gran <strong>Marais</strong> Shell tempered O<strong>the</strong>r Ceramics Arrowpo<strong>in</strong>ts Dart Po<strong>in</strong>ts Artifacts or Debris<br />
3AS305 Overcup Oak 3 1 Baytown Pla<strong>in</strong> 4 untyped pla<strong>in</strong> 1 untyped arrow- 1 graver<br />
body body po<strong>in</strong>t fragment<br />
3AS304 Overcup Oak 2 7 Baytown Pla<strong>in</strong><br />
body<br />
3AS303 Overcup Oak 1 5 Baytown Pla<strong>in</strong> 1 untyped <strong>in</strong>- 1 hammerstone<br />
body cised body 6 fire-cracked rock<br />
3AS302 Sycamore<br />
3AS286 One Cypress Po<strong>in</strong>t 5 Baytown Pla<strong>in</strong> 78 untyped Pla<strong>in</strong> 14 Nodena, 1 Nodena 2 end scrapers<br />
body body, 5 untyped Banks var., 1 Madison, 1 cyl<strong>in</strong>drical drill,<br />
<strong>in</strong>cised body 1 Honey Creek, 1 un- 1 plummet<br />
typed arrowpo<strong>in</strong>t frag. (see also Chapter 6)<br />
3BR59 Possumhaw<br />
3BR58 Buttonbush 454 Baytown Pla<strong>in</strong> 1 Gary 5 fire-cracked rock<br />
body Small<br />
3BR60 –<br />
3BR61 –<br />
3BR62 Mouth <strong>of</strong> Long Slough 13 Baytown Pla<strong>in</strong><br />
body<br />
3AS285 False Indigo 20 Lake Borgne 965 Baytown Pla<strong>in</strong> 106 untyped <strong>in</strong>- 38 untyped pla<strong>in</strong> 1 untyped bone 2 Ashley, 1 pitted cobble,<br />
Incised body, body, 76 Baytown cised or punctated body, 2 untyped tempered pla<strong>in</strong> 1 Alba 10 fire-cracked rock<br />
18 Tchefuncte Pla<strong>in</strong> rim body, 48 untyped pla<strong>in</strong> rim body (see also Chapter 6)<br />
Incised body, <strong>in</strong>cised or punc-<br />
9 Tchefuncte tated rim (see<br />
Pla<strong>in</strong> body also Chapter 8)<br />
3AS284 Mouth <strong>of</strong> Sal<strong>in</strong>e River 3 fire-cracked rock<br />
Unit 3: Lower Eagle Creek<br />
3BR68 –<br />
3BR67 – 1 Baytown Pla<strong>in</strong><br />
body<br />
3BR66 Eagle Creek 4 54 Baytown Pla<strong>in</strong> 1 untyped punc- 1 fire-cracked rock<br />
body, 3 Baytown tated body, 1<br />
Pla<strong>in</strong> rim untyped p<strong>in</strong>ched<br />
body
Table A-3 (cont<strong>in</strong>ued). Occurrence <strong>of</strong> Artifacts.<br />
Ceramic Complexes Projectile Po<strong>in</strong>ts<br />
O<strong>the</strong>r Dist<strong>in</strong>ctive<br />
Number Site Name Tchefuncte Baytown Gran <strong>Marais</strong> Shell tempered O<strong>the</strong>r Ceramics Arrowpo<strong>in</strong>ts Dart Po<strong>in</strong>ts Artifacts or Debris<br />
3BR69 Eagle Creek 5 1 Baytown Pla<strong>in</strong> body 1 untyped pla<strong>in</strong> 1 hammerstone<br />
body<br />
3BR65 Eagle Creek 3 1 Tchefuncte 11 untyped pla<strong>in</strong> 3 fire-cracked rock<br />
Stamped body body, 2 untyped<br />
pla<strong>in</strong> rim<br />
3BR64 Eagle Creek 2<br />
3BR63 Eagle C:reek 1 1 Baytown Pla<strong>in</strong> body 1 untyped pla<strong>in</strong> 4 fire-cracked rock<br />
body<br />
3BR79 Eagle Creek 6 1 discoidal scraper,<br />
10 fire-cracked rock<br />
Unit 4: Ouachita River Below Mile 254<br />
A-13<br />
3UN184 – 1 Baytown Pla<strong>in</strong> body<br />
3UN184 Brown Camp 2 Baytown Pla<strong>in</strong> body 1 untyped <strong>in</strong>- 1 untyped pla<strong>in</strong> 1 Washita, 1 Morhiss, 2 steatite sherds,<br />
cised body body 1 Madison 1 yarbrough, 1 schist slab, 216<br />
1 Gary, 1 Steuben, fire-cracked rock<br />
1 Langtry (see (see also Chapter 6)<br />
Chapter 8)<br />
3AS329 Marie Sal<strong>in</strong>e 14 Tchefuncte 66 Baytown Pla<strong>in</strong> body 2 untyped pla<strong>in</strong><br />
Incised body, body<br />
15 untyped body<br />
3AS126 – 1 Ashley<br />
3AS301 Honey Locust 4 Baytown Pla<strong>in</strong> body<br />
3AS319 – 1 fire-cracked rock<br />
3AS320 River Birch 1 3 fire-cracked rock<br />
3AS321 River Birch 2 1 Tchefuncte 1 hammerstone<br />
Stamped body,<br />
1 Tchefuncte<br />
Incised body, 6<br />
untyped body<br />
3AS288 Red Oak 2 fire-cracked rock<br />
3AS322 Water Elm 1 7 S<strong>in</strong>ner L<strong>in</strong>ear<br />
Punctated body,<br />
1 s<strong>in</strong>ner l<strong>in</strong>ear<br />
Punctated rim<br />
3AS332 Water Elm 2 2 fire-cracked rock<br />
3AS287 Water Elm 3 239 Baytown Pla<strong>in</strong> 2 untyped <strong>in</strong>body,<br />
34 Baytown cised body, 2 un-<br />
Pla<strong>in</strong> rim typed <strong>in</strong>cised rim<br />
3AS323 Water Elm 4 52 Baytown Pla<strong>in</strong> body
A-14<br />
Table A-3 (cont<strong>in</strong>ued). Occurrence <strong>of</strong> Artifacts.<br />
Ceramic Complexes Projectile Po<strong>in</strong>ts<br />
O<strong>the</strong>r Dist<strong>in</strong>ctive<br />
Number Site Name Tchefuncte Baytown Gran <strong>Marais</strong> Shell tempered O<strong>the</strong>r Ceramics Arrowpo<strong>in</strong>ts Dart Po<strong>in</strong>ts Artifacts or Debris<br />
3AS324 – 1 Ellis<br />
3UN123 Mouth <strong>of</strong> Lapile 27 Baytown Pla<strong>in</strong><br />
Creek body<br />
3AS327 Swamp Privet 1 Baytown Pla<strong>in</strong> 2 cobble anvils<br />
body (see Chapter 8)<br />
3AS325 – 1 Baytown Pla<strong>in</strong><br />
body<br />
3UN169 – 1 cobble mano<br />
3UN168 Coon Glory Bend 14 Baytown Pla<strong>in</strong><br />
body<br />
3UN121 Mo-Pac 4 Baytown Pla<strong>in</strong><br />
body, 1 Baytown<br />
Pla<strong>in</strong> rim<br />
Unit 5: Lower Lapile Creek<br />
3UN<strong>17</strong>8 Lapile Creek 1 1 Baytown Pla<strong>in</strong> 1 bifacial preform<br />
body<br />
3UN160 Lapile Creek 2 1 Baytown Pla<strong>in</strong> 1 untyped<br />
body stemmed po<strong>in</strong>t<br />
3UN159 Lapile Creek 3 25 Baytown Pla<strong>in</strong><br />
body<br />
3UN158 Lapile Creek 4 5 Baytown Pla<strong>in</strong> 1 Madison 1 Gary 1 bifacial preform<br />
body<br />
3UN157 Lapile Creek 5 26 Baytown Pla<strong>in</strong> 2 Gary 1 bifacial preform,<br />
body 1 hammerstone,<br />
5 fire-cracked rock<br />
3UN156 Lapile Creek 6 5 fire-cracked rock<br />
3UN152 Lapile Creek 8 1 hafted knife,<br />
1 bifacial preform<br />
3UN151 Lapile Creek 7 104 Baytown Pla<strong>in</strong> 1 Carrollton, 1 fire-cracked rock<br />
body, 1 Baytown 1 Gary<br />
Pla<strong>in</strong> rim<br />
Unit 6: Oxbow Lakes and Backswamp<br />
3UN187 St. Mary’s Lake 2 fire-cracked rock<br />
north<br />
3UN186 –<br />
3UN<strong>17</strong>0 Borrow Pit near 2 fire-cracked rock<br />
Lapoile Creek
Table A-3 (cont<strong>in</strong>ued). Occurrence <strong>of</strong> Artifacts.<br />
Ceramic Complexes Projectile Po<strong>in</strong>ts<br />
O<strong>the</strong>r Dist<strong>in</strong>ctive<br />
Number Site Name Tchefuncte Baytown Gran <strong>Marais</strong> Shell tempered O<strong>the</strong>r Ceramics Arrowpo<strong>in</strong>ts Dart Po<strong>in</strong>ts Artifacts or Debris<br />
3UN183 Lapoile Creek<br />
3UN182 Fishtrap Lake 1<br />
3UN180 Fishtrap Lake 2 4 Baytown Pla<strong>in</strong><br />
body<br />
3UN181 Fishtrap Lake 3<br />
3UN<strong>17</strong>9 Fishtrap Lake 4 7 untyped pla<strong>in</strong><br />
body<br />
3AS295 Borrow Pit near 4 Baytown Pla<strong>in</strong> 3 fire-cracked rock<br />
82 Bridge 1 body<br />
3AS294 Borrow Pit near 2 fire-cracked rock<br />
82 Bridge 2<br />
3AS292 Wildcat Lake 6 fire-cracked rock<br />
A-15
Appendix B<br />
fAunAl And florAl AnAlysis<br />
This appendix describes, primarily by means <strong>of</strong> tables, faunal and floral samples<br />
from floodpla<strong>in</strong> sites <strong>in</strong>vestigated dur<strong>in</strong>g site survey and test<strong>in</strong>g work (Chapters 5, 6).<br />
Also, <strong>the</strong> occurrence <strong>of</strong> wood and subfossil wood from floodpla<strong>in</strong> deposits or localities,<br />
not represent<strong>in</strong>g archeological contexts but relevant to paleoenvironmental study (Chapter<br />
2), is discussed briefly below.<br />
VertebrAte fAunA<br />
Table B-1 characterizes <strong>the</strong> small sample <strong>of</strong> vertebrate rema<strong>in</strong>s from eight prehistoric<br />
floodpla<strong>in</strong> sites. Six <strong>of</strong> <strong>the</strong>se sites were tested and appear <strong>in</strong> <strong>the</strong> order <strong>the</strong>y are<br />
reported <strong>in</strong> Chapter 6. The o<strong>the</strong>r two samples represent <strong>the</strong> rare recovery <strong>of</strong> animal rema<strong>in</strong>s<br />
from site survey work done. None <strong>of</strong> <strong>the</strong> historic sites recorded <strong>in</strong> <strong>the</strong> project area<br />
produced vertebrate rema<strong>in</strong>s, although none has been tested presently.<br />
As a general observation, <strong>the</strong> extremely acid soils and annual flood<strong>in</strong>g regime are<br />
not favorable to preservation <strong>of</strong> animal rema<strong>in</strong>s <strong>in</strong> <strong>the</strong>se floodpla<strong>in</strong> sites. The archeological<br />
specimens <strong>in</strong> Table B-1 are almost all calc<strong>in</strong>ed or burned to some extent; probably<br />
<strong>in</strong>organic bone constituents are persist<strong>in</strong>g <strong>in</strong> <strong>the</strong>se specimens, with some loss or replacement<br />
<strong>of</strong> organic content. However, none <strong>of</strong> <strong>the</strong> bones or fragments exam<strong>in</strong>ed actually<br />
appear m<strong>in</strong>eralized. Two sites <strong>of</strong> particular <strong>in</strong>terest, False Indigo (3AS285) and Jug Po<strong>in</strong>t<br />
Cut<strong>of</strong>f (3BR76), have midden deposits or anthropogenic soils with medium acid to neutral<br />
pH values (Chapter 7, Table 19), and some unburned bone may be preserved <strong>in</strong> <strong>the</strong>se<br />
deposits (especially 3BR76). Unfortunately, <strong>the</strong> stratified multiple component site, Marie<br />
Sal<strong>in</strong>e (3AS329), tested extensively to 1.5 m depth, is essentially devoid <strong>of</strong> animal bone<br />
so far as we can tell.<br />
Faunal samples were exam<strong>in</strong>ed under a b<strong>in</strong>ocular microscope and identifications<br />
made by <strong>the</strong> project archeologist. Typically, <strong>the</strong> specimens recovered are m<strong>in</strong>ute, fragmentary,<br />
and calc<strong>in</strong>ed, so that specific identifications are impossible, and even broad<br />
classification (mammal, bird, turtle, fish, and so on) was <strong>of</strong>ten difficult. Among a total<br />
<strong>of</strong> 10 2-liter soil samples from occupation levels or features washed through graded f<strong>in</strong>e<br />
screens (1/16-<strong>in</strong>ch or .16 mm mesh was <strong>the</strong> f<strong>in</strong>est used), only samples from midden soil<br />
at 3AS285 and 3BR76 were truly productive (Table B-1). In <strong>the</strong>se Mississippi period sites<br />
on <strong>the</strong> lower Sal<strong>in</strong>e River (Chapter 6), fish<strong>in</strong>g activity and fish process<strong>in</strong>g are <strong>in</strong>dicated<br />
by an abundance <strong>of</strong> sp<strong>in</strong>es, scales, teeth, otoliths, and o<strong>the</strong>r elements. Only drumfish and<br />
gar have been identified, but a dozen otoliths from 3BR76, potentially identifiable, are<br />
likely to represent additional species <strong>in</strong> <strong>the</strong> 5-30 kg size range (Chapter 2, Table 3).
B-2<br />
Table B-1. Vertebrate Rema<strong>in</strong>s from Prehistoric Floodpla<strong>in</strong> Sites.<br />
Site Number Site Name Sample Number Provenience Faunal Content Taxa Identified Recovery Technique<br />
3AS285 False <strong>in</strong>digo 1433-1-2-6 Test Pit 1, 0-10 cm 4 small calc<strong>in</strong>ed bone fragments dry screened<br />
1433-1-5-7 Test Pit 1, 10-20 cm 4 small calc<strong>in</strong>ed bone fragments dry screened<br />
1433-14-2 Test Pit 15, 0-10 cm many small burned or calc<strong>in</strong>ed bone Aplod<strong>in</strong>otus wet screened<br />
fragments, <strong>in</strong>clud<strong>in</strong>g fish sp<strong>in</strong>es; 1 fish grunniens<br />
vertebra; 1 drumfish tooth<br />
1433-14-7 Test Pit 15, Feature 2 several small burned or calc<strong>in</strong>ed bone wet screened<br />
fragments; 1 fish sp<strong>in</strong>e; 1 mammal?<br />
bone fragment<br />
1433-15-4 Shovel Test 1, 0-30 cm trace <strong>of</strong> calc<strong>in</strong>ed bone dry screened<br />
718-4-3 Grid Square 2, surface 3 small calc<strong>in</strong>ed bone fragments controlled surface collection<br />
1433-13-1-2 Grid Square 14, surface 1 calc<strong>in</strong>ed mammal? bone fragment controlled surface collection<br />
3BR58 Buttonbush 1435-1-1-2 Test Pit 1, 0-10 cm trace <strong>of</strong> calc<strong>in</strong>ed bone dry screened<br />
3A5286 One Cypress Po<strong>in</strong>t 1434-18-3-6 Test Pit 7, 10-20 cm 2 small calc<strong>in</strong>ed bone fragments dry screened<br />
3BR76 Jug Po<strong>in</strong>t Cut<strong>of</strong>f 1439-16-5-6 Test Pit 1, 10-20 cm 1 fish otolith dry screened<br />
1439-16-2-5 Test Pit 1, 10-20 cm many small burned or calc<strong>in</strong>ed bone Aplod<strong>in</strong>otus wet screened<br />
fragments, <strong>in</strong>clud<strong>in</strong>g fish sp<strong>in</strong>es and grunniens,<br />
vertebrae, drum teeth, gar scales Lepisosteus sp.<br />
1439-16-2-5 Test Pit 1, 15-20 cm many small burned or calc<strong>in</strong>ed bone wet screened<br />
fragments, <strong>in</strong>clud<strong>in</strong>g fish sp<strong>in</strong>es and<br />
vertebrae; 3 fish otoliths<br />
1439-16-2-6 Test Pit 1, 20-30 cm 1 fish otolith; 1 fish vertebra dry screened<br />
1439-16-4 Test Pit 1, 20-30 cm 3 fish otoliths; 3 fish vertebrae; 4 dry screened<br />
large bird? limb bone fragments; 2<br />
mammal? bone fragments; ca 8 sections<br />
<strong>of</strong> animal coprolites?<br />
1439-16-3 Test Pit 1, depth unknown 1 fragmentary large animal <strong>in</strong>cisor hand picked<br />
tooth
Table B-1 (concluded). Vertebrate Rema<strong>in</strong>s.<br />
Site Number Site Name Sample Number Provenience Faunal Content Taxa Identified Recovery Technique<br />
1439-5-4 Shovel Test 5, 13-30 cm 1 fish otolith, many small calc<strong>in</strong>ed dry screened<br />
bone fragments<br />
1439-5-4-3-2 Shovel Test 5, 13-30 cm many small burned or calc<strong>in</strong>ed bone Lepisosteus sp. wet screened<br />
fragments <strong>in</strong>clud<strong>in</strong>g fish sp<strong>in</strong>es and<br />
vertebrae, gar scales; 2 fish otoliths;<br />
1 animal coprolite or dropp<strong>in</strong>g?<br />
1439-13-5 Shovel Test 16, 13-30 cm 2 fish otoliths; 2 fish sp<strong>in</strong>es; 1 small dry screened<br />
animal long bone fragment<br />
1439-<strong>17</strong>-4 Surface 1 charred, fragmentary, small animal surface collection<br />
bone<br />
3AS306 Jug Po<strong>in</strong>t 1 1436-20-1-7 Test Pit 2, 10-20 cm trace <strong>of</strong> calc<strong>in</strong>ed bone dry screened<br />
1436-20-3-5 Test Pit 2, Feature 1 ca. 200 small burned or calc<strong>in</strong>ed bone dry screened<br />
fragments, <strong>in</strong>clud<strong>in</strong>g large or medium<br />
sized mammal<br />
1436-20-5-5 Test Pit 2, Feature 1 2 small calc<strong>in</strong>ed bone fragments wet screened<br />
1436-1-1-6 Surface 1 calc<strong>in</strong>ed turtle shell fragment surface collection<br />
3AS307 Jug Po<strong>in</strong>t 2 1437-7-2-6 Test Pit 1, 10-20 cm ca 12 small calc<strong>in</strong>ed bone fragments dry screened<br />
3BR62 Mouth <strong>of</strong> Long Slough 1270-4 Surface 4 calc<strong>in</strong>ed turtle shell fragments surface collection<br />
3BR66 Eagle Creek 4 1379-4 Surface 2 calc<strong>in</strong>ed mammal? bone fragment surface collection<br />
B-3
B-4<br />
Among all prehistoric floodpla<strong>in</strong> sites, <strong>the</strong>se two Mississippi period components have<br />
<strong>the</strong> greatest potential for <strong>in</strong>tensive faunal analysis (Appendix C).<br />
By way <strong>of</strong> contrast, <strong>the</strong> faunal rema<strong>in</strong>s from Jug Po<strong>in</strong>t 1 (3AS306) <strong>in</strong>clude mammal<br />
bone <strong>in</strong> fragmentary, calc<strong>in</strong>ed condition; no fish elements have been identified <strong>in</strong> dry or<br />
wet screen samples (Table B-1). On present evidence, this site must be associated with<br />
hunt<strong>in</strong>g activity and fall occupation (Chapter 6).<br />
MolluscAn fAunA<br />
Table B-2 presents ra<strong>the</strong>r limited archeological data for mollusc rema<strong>in</strong>s <strong>in</strong> both<br />
historic and prehistoric floodpla<strong>in</strong> sites. The modern sample consists <strong>of</strong> about 110 mussels,<br />
usually pairs <strong>of</strong> valves, collected as dead shells along <strong>the</strong> shorel<strong>in</strong>es <strong>of</strong> <strong>the</strong> Ouachita<br />
and Sal<strong>in</strong>e rivers and <strong>in</strong> backswamp areas dur<strong>in</strong>g 1979 site survey work. These most<br />
likely represent animals stranded by lowered river levels earlier <strong>in</strong> that year, or recent<br />
death through some o<strong>the</strong>r cause. The dredged sample <strong>in</strong> Table B-2 represents ano<strong>the</strong>r six<br />
mussels collected from dredged riverbottom sediment at a pipel<strong>in</strong>e cross<strong>in</strong>g (mile 262.5)<br />
on <strong>the</strong> Ouachita River. These animals may have died as <strong>the</strong> result <strong>of</strong> dredg<strong>in</strong>g here <strong>in</strong><br />
about 1974-1975. Both modern comparative samples are deposited <strong>in</strong> <strong>the</strong> <strong>University</strong> <strong>of</strong><br />
<strong>Arkansas</strong> Museum collection (accession numbers 80-99 and 80-100).<br />
Identification <strong>of</strong> modern and archeological molluscs <strong>in</strong> Table B-2 was made by<br />
Mark E. Gordon, Department <strong>of</strong> Zoology, <strong>University</strong> <strong>of</strong> <strong>Arkansas</strong>, and <strong>the</strong> classification<br />
follows a recent checklist <strong>of</strong> Unionacea for this and o<strong>the</strong>r regions <strong>in</strong> <strong>Arkansas</strong> (Gordon et<br />
al. 1979).<br />
As <strong>in</strong> <strong>the</strong> case <strong>of</strong> animal bone, mollusc shell is not likely to be preserved <strong>in</strong> floodpla<strong>in</strong><br />
sites. Specimens collected here are usually fragmentary or eroded, and may be<br />
calc<strong>in</strong>ed. Archeological samples <strong>in</strong> Table B-2 consist <strong>of</strong> 4-30 valves <strong>in</strong> <strong>the</strong> case <strong>of</strong> historic<br />
sites 3UN88 and 3AS330, and 1-2 valves plus fragments <strong>in</strong> <strong>the</strong> case <strong>of</strong> prehistoric sites<br />
3BR75 and 3BR76. Unidentifiable traces <strong>of</strong> mussel and snail shell were also noted <strong>in</strong> deep<br />
excavated levels at Marie Sal<strong>in</strong>e (3AS329), <strong>in</strong>clud<strong>in</strong>g Test Pit 5, 100-110 cm, and Test Pit<br />
7, 80-90 cm.<br />
Four species <strong>of</strong> mussels have been identified <strong>in</strong> prehistoric site samples (Table B-2)<br />
and presumably represent supplementary food resources (Parmalee and Klippel 1974).<br />
Five species <strong>of</strong> mussels have been identified <strong>in</strong> historic sites samples and may have been<br />
eaten or used as a bait source. No mussel shell or “pearl” button <strong>in</strong>dustry is <strong>in</strong>dicated by<br />
historic floodpla<strong>in</strong> sites <strong>in</strong> this area. All <strong>of</strong> <strong>the</strong> archeological species are known or expected<br />
<strong>in</strong> <strong>the</strong> region today, although <strong>the</strong> modern molluscan fauna <strong>of</strong> <strong>the</strong> lower Ouachita<br />
River and <strong>the</strong> Sal<strong>in</strong>e River <strong>in</strong> <strong>Arkansas</strong> rema<strong>in</strong>s poorly known (Gordon et al. 1979).
Table B-2. Molluscs Identified <strong>in</strong> Prehistoric and Historic Site Samples and a Modern<br />
Sample from Felsenthal National Wildlife Refuge.<br />
Modern Dredged<br />
Sample Sample Historic Sites Prehistoric Sites<br />
Molluscan Species (1979) (1974-1975) 3UN88 3AS330 3BR75 3BR76<br />
Unionidae<br />
Fusconaia flava x x<br />
Fusconaia ebena x x x x x<br />
Amblema plicata x x x x x<br />
Quadrula pustulosa x x x?<br />
Quadrula quadrula x<br />
Quadrula nodulata<br />
Quadrula metanerva x<br />
Quadrula cyl<strong>in</strong>drica x<br />
Tritogonia verrucosa x<br />
Plectomerus dombeyanus x<br />
Pleurobema clava x x?<br />
Anodonta grandis x<br />
Anodonta imbecilis x<br />
Ptychobranchus occidentalis x<br />
Cyprogenia aberti<br />
Obovaria jacksoniana x<br />
Truncilla truncata<br />
Leptodea fragilis x<br />
Proptera purpurata x<br />
Caruncul<strong>in</strong>a texasensis x<br />
Ligumia subrostrata x<br />
Lampsilis anodontoides x<br />
Lampsilis hydiana x x?<br />
Lampsilis ovata<br />
Corbicula cf. flum<strong>in</strong>ea x<br />
Viviparidae<br />
Viviparus <strong>in</strong>tertextus x<br />
Campeloma subsolidum x<br />
Planorbidae<br />
Helisoma trivolvis x<br />
B-5
B-6<br />
cArboniZed plAnt reMA<strong>in</strong>s<br />
Identifications and o<strong>the</strong>r data for charcoal and carbonized seeds from prehistoric<br />
floodpla<strong>in</strong> sites are presented <strong>in</strong> Table B-3. Charcoal specimens were prepared and<br />
exam<strong>in</strong>ed under a b<strong>in</strong>ocular microscope by David W. Stahle, dendrochronologist and<br />
Research Assistant at <strong>Arkansas</strong> Archeological Survey. One historic wood specimen from<br />
<strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g (3UN122) was similarly exam<strong>in</strong>ed and is reported <strong>in</strong> Table B-3.<br />
Charcoal or wood taxa were identifed by reference to a wood characteristics manual<br />
(Pansh<strong>in</strong> and de Zeeuw 1970). About two dozen charcoal samples from floodpla<strong>in</strong> sites,<br />
not tabulated, were too small or too poorly preserved for identification <strong>of</strong> taxa. The few<br />
carbonized seeds recovered were identified by <strong>the</strong> project archeologist us<strong>in</strong>g a seed manual<br />
(U.S. Department <strong>of</strong> Agriculture 1948) and modern specimens collected <strong>in</strong> <strong>the</strong> project<br />
area. Surpris<strong>in</strong>gly, charred nutshell occurred only <strong>in</strong> trace amounts <strong>in</strong> <strong>the</strong>se samples, and<br />
prehistoric use <strong>of</strong> floodpla<strong>in</strong> nut crops rema<strong>in</strong>s an open question.<br />
The prehistoric charcoal specimens, presumably represent<strong>in</strong>g fuelwood, could,<br />
for <strong>the</strong> most part, be derived from bottomland hardwood and swamp forest species<br />
found <strong>in</strong> <strong>the</strong> project area today (Chapter 2, Table 1). Red oak species <strong>in</strong>clude cherrybark,<br />
Nuttall, water, and willow oak; <strong>the</strong> only white oak <strong>of</strong> this overflow bottomland noted<br />
dur<strong>in</strong>g site survey was overcup. Bald cypress is present <strong>in</strong> channels, sloughs, and lakes<br />
throughout <strong>the</strong> area. P<strong>in</strong>e charcoal and pitch or tar, tentatively identified <strong>in</strong> a midden<br />
sample from False Indigo (3AS285), may represent wood <strong>in</strong>troduced from an upland or<br />
terrace location, s<strong>in</strong>ce no p<strong>in</strong>es occur <strong>in</strong> <strong>the</strong> bottomland forest as <strong>the</strong> result <strong>of</strong> overflow<br />
conditions. This <strong>in</strong>troduction could be <strong>in</strong>tentional or could merely represent use <strong>of</strong> p<strong>in</strong>e<br />
occurr<strong>in</strong>g as driftwood. At Gran <strong>Marais</strong> Land<strong>in</strong>g (3UN122), a split, worked section <strong>of</strong><br />
p<strong>in</strong>e most likely represents n<strong>in</strong>eteenth century activity at this riverboat land<strong>in</strong>g.<br />
The few carbonized seeds <strong>of</strong> honey locust and persimmon recovered at One Cypress<br />
Po<strong>in</strong>t (3AS286), Jug Po<strong>in</strong>t 1 (3AS306), and Jug Po<strong>in</strong>t 2 (3AS307) most likely represent<br />
foodstuffs <strong>in</strong> use on <strong>the</strong>se sites <strong>of</strong> <strong>the</strong> Late Mississippi period (Chapter 6). Both trees<br />
grow today on <strong>the</strong> same natural levees <strong>of</strong> <strong>the</strong> Sal<strong>in</strong>e River where <strong>the</strong>se sites are located.<br />
Their seedpods and fruits ripen <strong>in</strong> fall, and <strong>the</strong> carbonized rema<strong>in</strong>s should be an <strong>in</strong>dicator<br />
<strong>of</strong> fall occupation, especially where overflow may extend through w<strong>in</strong>ter and spr<strong>in</strong>g<br />
(Chapter 9, Figure 70).<br />
subfossil wood<br />
Table B-4 summarizes characteristics <strong>of</strong> several wood samples from floodpla<strong>in</strong><br />
deposits or nonarcheological localities <strong>of</strong> environmental <strong>in</strong>terest. These specimens were<br />
prepared, exam<strong>in</strong>ed, and identified by David W. Stahle <strong>in</strong> conjunction with study <strong>of</strong><br />
archeological charcoal specimens. They are <strong>in</strong>cluded here to show that organic material<br />
can persist <strong>in</strong> floodpla<strong>in</strong> sediments under cont<strong>in</strong>uously saturated conditions. Such specimens<br />
are <strong>of</strong> paleoenvironmental importance if <strong>the</strong>y can be identified and dated by
Table B-3. Floral Rema<strong>in</strong>s from Prehistoric and Historic Floodpla<strong>in</strong> Sites.<br />
Site O<strong>the</strong>r Carbonized<br />
Number Site Name Sample Number Provenience Sample Description Charcoal or Wood Taxa Floral Material Remarks<br />
3AS329 Marie Sal<strong>in</strong>e 1415-2-9-2 Test Pit 2, 114 cm several small woody fragments red oak group — r<strong>in</strong>g porous species<br />
1415-10-8-5 Test Pit 5, 100-110 cm many small woody fragments cf. baldcypress — diffuse porous species<br />
1415-11-8-5 Test Pit 6, 90-100 cm many small woody fragments, red oak group — Feature1 (see Chapter<br />
(wet screened) trace <strong>of</strong> nutshell? 6)<br />
3AS285 False Indigo 1433-4-3 Test Pit 15, 10-18 cm one small woody fragment cf. p<strong>in</strong>e pitch or tar? Feature 1 (see Chapter<br />
adher<strong>in</strong>g 6)<br />
1433-15-3 Shovel Test 1, 0-30 cm several small woody fragments cf. bald cypress or — diffuse and r<strong>in</strong>g porous<br />
p<strong>in</strong>e and red oak group species<br />
1433-21-5 Shovel Test 8, 0-20 cm several small woody fragments red oak group — r<strong>in</strong>g porous species<br />
3AS286 One Cypress Po<strong>in</strong>t 1434-18-3-5 Test Pit 7, 10-20 cm one seed ca. 2 x 6 x 14 mm — honey locust (see Chapter 6)<br />
3BR76 Jug Po<strong>in</strong>t Cut<strong>of</strong>f 1439-8-3 Shovel Test 8, 0-50 cm many small woody fragments red oak group —<br />
1439-16-2-5 Test Pit 1, 10-20 cm several small woody fragments cf. baldcypress — associated with midden<br />
(wet screened) (see Chapter 6)<br />
3AS306 Jug Po<strong>in</strong>t 1 1436-20-1-6 Test Pit 2, 10-20 cm 2 small woody fragments, 1 seed, red oak group honeylocust<br />
o<strong>the</strong>r seed fragments<br />
1436-20-3-4 Test Pit 2, 19-30 cm several small woody fragments, white oak group honey locust, Feature 1<br />
ca. 3 seeds, trace <strong>of</strong> nutshell? persimmon (see Chapter 6)<br />
1436-20-5-5 Test Pit 2, Feature 1 fill 2 seeds ca. 2 x 8 x 12mm — honey locust (see Chapter 6)<br />
(wet screened)<br />
3AS307 Jug Po<strong>in</strong>t 2 1437-7-2-5 Test Pit 1, 10-20 cm 5 seeds ca. 2 x 8 x 12mm — honey locust (see Chapter 6)<br />
3UN122 <strong>Grand</strong> <strong>Marais</strong> 1372-3 unnumbered split section <strong>of</strong> wood p<strong>in</strong>e — (not charred)<br />
Land<strong>in</strong>g shovel test, 0-40 cm ca. 3 x 3 x 37 cm<br />
Note: All specimens listed above are carbonized except sample 1372-3. Samples recovered by dry screen<strong>in</strong>g except as <strong>in</strong>dicated.<br />
B-7
B-8<br />
Table B-4. Wood and Subfossil Wood Specimens from Environmental Localities.<br />
Project<br />
Description <strong>of</strong> Locality Sample Number Description <strong>of</strong> Specimens Wood Taxa Identified<br />
Eroded cypress butts, 295-120 2 sections <strong>of</strong> light, dry wood; bald cypress<br />
Ouachita River bankl<strong>in</strong>e 3 x 6 x 26 cm and 3 x 7 x 53 cm;<br />
at mile 266.6 sensitive annual r<strong>in</strong>g series may be<br />
(see Figure 7a) present; diffuse porous<br />
Dredged spoil from — 1 dark-colored, wood section; white oak group<br />
pipel<strong>in</strong>e cross<strong>in</strong>g, peel<strong>in</strong>g <strong>in</strong> th<strong>in</strong> sheets along rays;<br />
Ouachita River bottom 4 x 6 x 54 cm; r<strong>in</strong>g porous<br />
at mile 262.5<br />
— 1 dark-colored wood section, white oak group<br />
2 x 9 x 22 cm; slightly distorted<br />
r<strong>in</strong>g porous<br />
— 1 dark-colored, hard, wood section; white oak group<br />
6 x 7 x 11 cm; highly distorted;<br />
r<strong>in</strong>g porous<br />
Excavated spoil from new — 6 x 9 x 10 cm; undistorted; about<br />
Ouachita River floodpla<strong>in</strong> 60 r<strong>in</strong>gs present; r<strong>in</strong>g porous<br />
at miles 242-243<br />
— 1 dark-colored, wood section, cf. white oak group<br />
3 x 4 x 6 cm, distorted, hard, and<br />
brittle; r<strong>in</strong>g porous
dendrochronological or radiocarbon methods. In this group <strong>of</strong> samples white oak predom<strong>in</strong>ates,<br />
<strong>in</strong> contrast to results <strong>of</strong> charcoal identification (Table B-3). The cause <strong>of</strong> this<br />
prevalence is not known, and too few specimens have been collected presently. All <strong>of</strong> <strong>the</strong><br />
white oak specimens are discolored, distorted, or o<strong>the</strong>rwise altered by burial and subsequent<br />
exposure.<br />
references cited<br />
Gordon, M. E., L. Russert Kraemer, and A. V. Brown<br />
1979 Unionaceae <strong>of</strong> <strong>Arkansas</strong>: Historical Review, Checklist, and Observations on<br />
Distributional Patterns. Bullet<strong>in</strong> <strong>of</strong> <strong>the</strong> American Malecological Union 1979:31-37.<br />
Pansh<strong>in</strong>, A. J., and Carl de Zeeuw<br />
1970 Textbook <strong>of</strong> Wood Technology. Volume 1 (third ed.). McGraw-Hill, New York.<br />
U.S. Department <strong>of</strong> Agriculture<br />
1948 Woody-plant Seed Manual. Miscellaneous Publication 654. Wash<strong>in</strong>gton, D.C.<br />
B-9
Appendix C<br />
recoMMended progrAM for MitigAtion <strong>of</strong><br />
iMpActs on ArcheologicAl resources <strong>in</strong> <strong>the</strong><br />
felsenthAl nAVigAtion pool<br />
<strong>in</strong>troduction<br />
The Felsenthal Navigation Project survey, carried out by <strong>the</strong> <strong>Arkansas</strong> Archeological<br />
Survey dur<strong>in</strong>g <strong>the</strong> last half <strong>of</strong> 1979, recorded 126 previously unknown archeological<br />
sites and revisited 18 o<strong>the</strong>rs known from prior work. These 144 sites are summarized <strong>in</strong><br />
Chapter 5 and Appendix A, both <strong>of</strong> which are organized as eight “site units.” Seven <strong>of</strong><br />
<strong>the</strong>se site units <strong>in</strong>clude prehistoric and historic sites with<strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> <strong>of</strong> <strong>the</strong> Ouachita<br />
and Sal<strong>in</strong>e rivers; <strong>the</strong> o<strong>the</strong>r site unit (Pleistocene Terraces and Islands) <strong>in</strong>cludes 27<br />
sites above <strong>the</strong> floodpla<strong>in</strong>.<br />
In <strong>the</strong> mitigation proposal presented here we address all known floodpla<strong>in</strong> sites<br />
(recorded or revisited), not<strong>in</strong>g <strong>the</strong>ir relationship to <strong>the</strong> 65-foot contour (or prospective<br />
navigation pool elevation). The reader should refer back to <strong>the</strong> discussion <strong>of</strong> <strong>the</strong> 65-foot<br />
contour (as a constra<strong>in</strong>t on survey area) <strong>in</strong> Chapter 1 and to <strong>the</strong> narrative <strong>of</strong> methodology<br />
presented <strong>in</strong> Chapter 5. The po<strong>in</strong>t we emphasize here is that a number <strong>of</strong> floodpla<strong>in</strong><br />
sites evaluated and recommended for mitigation <strong>in</strong> this appendix lie partly or wholly<br />
above <strong>the</strong> 65-foot contour. Also our appraisal <strong>of</strong> project impacts is not limited strictly to<br />
<strong>the</strong> new navigation pool, but <strong>in</strong>cludes floodpla<strong>in</strong> areas above 65 feet elevation. In deal<strong>in</strong>g<br />
with all known floodpla<strong>in</strong> sites and floodpla<strong>in</strong> site impacts, we have departed from<br />
<strong>the</strong> project Scope <strong>of</strong> Services <strong>in</strong> certa<strong>in</strong> respects which have been <strong>the</strong> subject <strong>of</strong> previous<br />
discussion with <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg District.<br />
Previous site data and maps from <strong>the</strong> Felsenthal Project were forwarded to <strong>the</strong><br />
Vicksburg District <strong>in</strong> 1980 as an aid to <strong>the</strong>ir plann<strong>in</strong>g process (E. T. Hemm<strong>in</strong>gs, letter to<br />
Shelia Lewis, 5 August 1980). We were cautious about <strong>the</strong> use <strong>of</strong> site data <strong>in</strong> advance <strong>of</strong><br />
completed analyses and full discussion <strong>in</strong> a f<strong>in</strong>al report, stat<strong>in</strong>g some <strong>of</strong> <strong>the</strong> limitations<br />
as follows:<br />
Most, but not all, <strong>of</strong> <strong>the</strong> 740 sites [now 144 sites] plotted on this map<br />
are with<strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> project area. O<strong>the</strong>r sites at higher elevations were<br />
recorded or revisited when survey crews were enter<strong>in</strong>g and exit<strong>in</strong>g <strong>the</strong> floodpla<strong>in</strong>.<br />
This map represents <strong>the</strong> 1979 work <strong>of</strong> our project and does not beg<strong>in</strong> to<br />
<strong>in</strong>clude all known sites at higher elevations under Fish and Wildlife jurisdiction.<br />
It is complete through 1979 for <strong>the</strong> floodpla<strong>in</strong>, given <strong>the</strong> limitations <strong>of</strong>
C-2<br />
survey <strong>in</strong> that environment to be discussed and evaluated <strong>in</strong> our report to<br />
you. It cannot, <strong>of</strong> course, be used as an archeological site map <strong>of</strong> <strong>the</strong> entire<br />
Felsenthal National Wildlife Refuge. Lastly, avoidance <strong>of</strong> buried archeological<br />
sites <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong> cannot be assured by use <strong>of</strong> this map, as for example <strong>in</strong><br />
construct<strong>in</strong>g bendways and boat channels.<br />
Hav<strong>in</strong>g completed all analyses <strong>of</strong> site data and site significance, we th<strong>in</strong>k it is advisable<br />
to state <strong>the</strong>se limitations <strong>of</strong> project results once aga<strong>in</strong> <strong>in</strong> succ<strong>in</strong>ct form, as an aid to management<br />
decisions:<br />
1. All prehistoric sites, and probably all or parts <strong>of</strong> some historic sites, are buried<br />
<strong>in</strong> <strong>the</strong> Ouachita and Sal<strong>in</strong>e River floodpla<strong>in</strong> study area.<br />
2. No agricultural development has taken place <strong>in</strong> this floodpla<strong>in</strong> area, which<br />
is an overflow bottomland, and no sites have been, or can be, discovered as<br />
<strong>the</strong> result <strong>of</strong> such activity (Chapter 5).<br />
3. Subsurface survey methods <strong>in</strong> <strong>the</strong> bottomland hardwood and swamp forest<br />
environments, constitut<strong>in</strong>g about 85% <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> area, are deficient,<br />
and buried cultural resources are likely to rema<strong>in</strong> undiscovered (Chapter 5).<br />
4. Underwater survey methods were not employed <strong>in</strong> <strong>the</strong> river channel, open<br />
lake, and bayou environments which constitute about 15% <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong><br />
area; underwater cultural resources are likely to rema<strong>in</strong> undiscovered (<strong>the</strong><br />
steamboat Lotawanna is a notable exception—see Chapter 4, 5).<br />
5. No large scale, systematic, cultural resource survey has been carried out on<br />
floodpla<strong>in</strong> terra<strong>in</strong> over 65 feet elevation; this report does identify resources<br />
above 65 feet <strong>in</strong> certa<strong>in</strong> transect and levee locations, and more importantly,<br />
concludes that a markedly low density <strong>of</strong> sites can be expected to occur <strong>in</strong><br />
<strong>the</strong> extensive backswamp zones <strong>of</strong> Ouachita and Sal<strong>in</strong>e rivers.<br />
6. Site specific surveys <strong>of</strong> certa<strong>in</strong> floodpla<strong>in</strong> areas (new Felsenthal Lock and<br />
Dam construction area and new Crossett Harbor Facility area, among o<strong>the</strong>rs)<br />
are, we believe, now scheduled; <strong>the</strong> results <strong>of</strong> <strong>the</strong>se surveys are not known<br />
to us and are not considered <strong>in</strong> <strong>the</strong> report.<br />
7. The results <strong>of</strong> <strong>the</strong> 1979 Felsenthal Project deal briefly with sites above<br />
floodpla<strong>in</strong> level, but no systematic survey <strong>of</strong> such terra<strong>in</strong> was required or<br />
performed by us; recommendations <strong>of</strong> a general nature for management <strong>of</strong><br />
nonfloodpla<strong>in</strong> resources <strong>in</strong> <strong>the</strong> Felsenthal National Wildlife Refuge are presented<br />
at <strong>the</strong> end <strong>of</strong> this appendix.<br />
In addition to site data and maps forwarded to <strong>the</strong> Corps <strong>of</strong> Eng<strong>in</strong>eers <strong>in</strong> 1980,<br />
some specific recommendations with regard to sites endangered or possibly endangered
y construction were also furnished by <strong>Arkansas</strong> Archeological Survey. These recommendations<br />
were based on prelim<strong>in</strong>ary assessment <strong>of</strong> site significance and project<br />
impacts, and <strong>the</strong>y do not differ from f<strong>in</strong>al assessments presented <strong>in</strong> this appendix except<br />
<strong>in</strong> <strong>the</strong> matter <strong>of</strong> dist<strong>in</strong>ction we make between terra<strong>in</strong> above and below 65 feet elevation.<br />
The essential portions <strong>of</strong> <strong>the</strong>se recommendations are reproduced here, as <strong>the</strong>y cont<strong>in</strong>ue<br />
to be relevant to <strong>the</strong> mitigation proposal presented below.<br />
<strong>the</strong> follow<strong>in</strong>g <strong>in</strong>formation and recommendations should be forwarded<br />
to you prior to completion <strong>of</strong> our contract. One previously recorded site,<br />
<strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g (3UN122) and one newly recorded site, Wreck <strong>of</strong> <strong>the</strong><br />
Lotawanna (3UN153), are located with<strong>in</strong> or near <strong>the</strong> upstream end <strong>of</strong> Felsenthal<br />
Lock and Dam channel construction....As far as we can determ<strong>in</strong>e now,<br />
both sites will be adversely affected. Also, based on present knowledge both<br />
sites have regional historical significance and are potentially eligible for<br />
nom<strong>in</strong>ation to <strong>the</strong> National Register <strong>of</strong> Historic Places. However, <strong>the</strong> determ<strong>in</strong>ation<br />
<strong>of</strong> eligibility here should <strong>in</strong>clude systematic <strong>in</strong>formation on <strong>the</strong> extent<br />
and <strong>in</strong>tegrity <strong>of</strong> historic rema<strong>in</strong>s, <strong>in</strong> one case below ground and <strong>in</strong> <strong>the</strong> o<strong>the</strong>r<br />
underwater. Clearly, this <strong>in</strong>formation should be obta<strong>in</strong>ed before construction<br />
proceeds <strong>in</strong>to <strong>the</strong> area <strong>of</strong> <strong>the</strong> land<strong>in</strong>g or steamboat wreck.<br />
We recommend that <strong>the</strong> Corps arrange to have a qualified diver-archeologist<br />
<strong>in</strong>spect <strong>the</strong> underwater Lotawanna wreck site, p<strong>in</strong>po<strong>in</strong>t <strong>the</strong> location<br />
and extent, and assess <strong>the</strong> <strong>in</strong>tegrity <strong>of</strong> rema<strong>in</strong>s.<br />
We also recommend limited subsurface test<strong>in</strong>g <strong>of</strong> <strong>the</strong> riverbank area<br />
recorded as <strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g (3UN122), <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> problematical<br />
physical features exam<strong>in</strong>ed and recorded by us <strong>in</strong> 1976 and 1979, as well as<br />
sites <strong>of</strong> structures nearby <strong>in</strong>dicated only by n<strong>in</strong>eteenth century river maps (C.<br />
R. McGimsey III, letter to Shelia Lewis, 29 August 1980).<br />
...actual construction rights-<strong>of</strong>-way and impacts are not known to us <strong>in</strong> detail.<br />
I have <strong>in</strong> hand <strong>the</strong> Crossett Harbor Feasibility Report (September 1979) and<br />
a Lock and Dam construction map (April 1976). On <strong>the</strong> basis <strong>of</strong> <strong>the</strong>se, six<br />
sites on record are probably subject to direct or <strong>in</strong>direct impacts <strong>of</strong> construction.<br />
Those sites are 3AS159, 3AS299, 3AS300, 3AS301, and 3AS319 at or<br />
near Crossett Harbor and 3UN121 at <strong>the</strong> closure dam....Both Crossett Harbor<br />
and <strong>the</strong> closure dam will largely impact terra<strong>in</strong> above 65 feet which has not<br />
been <strong>in</strong>tensively surveyed by us or anyone else. With regard to <strong>the</strong> Crossett<br />
Harbor area,...Hester Davis, <strong>Arkansas</strong> State Archeologist,...noted that buried<br />
sites would be expected to fall <strong>in</strong> this area and that a more thorough survey<br />
with shovel test<strong>in</strong>g should be carried out. We cont<strong>in</strong>ue to hold this view, and<br />
moreover we call attention now to <strong>the</strong> probability that significant historic<br />
rema<strong>in</strong>s <strong>of</strong> Marie Sal<strong>in</strong>e Land<strong>in</strong>g (3AS299) are likely to be present and undiscovered<br />
<strong>in</strong> terra<strong>in</strong> areas above 65 feet. The 200-meter long canal at Marie<br />
Sal<strong>in</strong>e Land<strong>in</strong>g was recorded <strong>in</strong> detail by our survey team because it lies at<br />
and below 65 feet elevation. We know that obtrusive historic rema<strong>in</strong>s are not<br />
present on <strong>the</strong> surface near <strong>the</strong> canal, but buried rema<strong>in</strong>s may be present.<br />
C-3
C-4<br />
Also <strong>the</strong> canal itself may have a submerged bottom deposit with artifacts<br />
perta<strong>in</strong><strong>in</strong>g to <strong>the</strong> 19th century construction and use. More work is certa<strong>in</strong>ly<br />
needed <strong>in</strong> this vic<strong>in</strong>ity. See also <strong>the</strong> site records for four o<strong>the</strong>r sites that may<br />
be affected.<br />
With regard to <strong>the</strong> closure dam area, <strong>the</strong> one site on record here,<br />
3UN121, is a buried prehistoric site <strong>of</strong> apparent small size, with<strong>in</strong> <strong>the</strong> right<strong>of</strong>-way<br />
accord<strong>in</strong>g to my maps. This site lies above <strong>the</strong> 65 foot contour and was<br />
not tested by our survey team. Subsurface test<strong>in</strong>g should be carried out here<br />
prior to construction (E. T. Hemm<strong>in</strong>gs, letter to Joyce Williams, 10 April 1981).<br />
All <strong>of</strong> <strong>the</strong> floodpla<strong>in</strong> sites mentioned <strong>in</strong> <strong>the</strong>se excerpts have now been evaluated<br />
on <strong>the</strong> basis <strong>of</strong> all available project <strong>in</strong>formation. Recommendations for certa<strong>in</strong> <strong>of</strong> <strong>the</strong>se<br />
which appear below simply build upon <strong>the</strong>se earlier appraisals. Also, we would note<br />
briefly mitigative actions by <strong>the</strong> Corps which have been taken recently with regard to<br />
two <strong>of</strong> <strong>the</strong> sites <strong>of</strong> concern:<br />
1. Wreck <strong>of</strong> <strong>the</strong> Lotawanna (3UN153) on <strong>the</strong> Ouachita River bottom has<br />
probably been p<strong>in</strong>po<strong>in</strong>ted by magnetometry and bathymetry, as noted <strong>in</strong><br />
Chapter 5; additional recommendations appear below.<br />
2. Mo-Pac site (3UN121) <strong>in</strong> <strong>the</strong> closure dam right-<strong>of</strong>-way has now been<br />
tested by Historic Preservation Associates <strong>of</strong> Fayetteville; no additional<br />
recommendations appear below.<br />
significAnce <strong>of</strong> floodplA<strong>in</strong> sites<br />
The concept <strong>of</strong> significance <strong>in</strong> cultural resource management has been scrut<strong>in</strong>ized<br />
and debated for over a decade. Several dimensions <strong>of</strong> significance have been proposed<br />
and discussed, among which scientific significance, or <strong>the</strong> quality <strong>of</strong> research potential,<br />
is most relevant to our evaluation <strong>of</strong> floodpla<strong>in</strong> sites (Schiffer and Gumerman 1977:241;<br />
Advisory Council on Historic Preservation 1980:6). Although this concept is very useful,<br />
it must be adapted to <strong>the</strong> particular cultural-environmental context <strong>of</strong> <strong>the</strong> Felsenthal<br />
Project area. In contrast to many o<strong>the</strong>r archeological localities and regions <strong>in</strong> North<br />
America, our floodpla<strong>in</strong> sites are:<br />
1. markedly small <strong>in</strong> size or extent<br />
2. <strong>of</strong>ten markedly sparse <strong>in</strong> content <strong>of</strong> artifacts and debris<br />
3. buried by overbank silty clays<br />
4. never plowed, and probably never dug or collected, prior to our work <strong>in</strong><br />
1979, and<br />
5. set <strong>in</strong> an overflow bottomland and annually overflowed by several meters<br />
<strong>of</strong> water.<br />
We have dealt with <strong>the</strong> research potential <strong>of</strong> small sites <strong>in</strong> Chapter 6, and aga<strong>in</strong> <strong>in</strong><br />
Chapter 9. As Schiffer and Gumerman (1977:242) note:
<strong>the</strong>re is no a priori reason to th<strong>in</strong>k that [<strong>in</strong>formation potential] would correlate<br />
directly with site size....With <strong>the</strong> advent <strong>of</strong> settlement system analysis<br />
and o<strong>the</strong>r approaches <strong>in</strong> modern archeology <strong>the</strong> smallest site <strong>in</strong> a region may<br />
be as important or more so than <strong>the</strong> largest....<br />
The summary assessment <strong>of</strong> eight prehistoric floodpla<strong>in</strong> sites, tested extensively by<br />
us, <strong>in</strong>dicates that specific relevant site data were adduced <strong>in</strong> all <strong>the</strong> follow<strong>in</strong>g doma<strong>in</strong>s<br />
(Chapter 6, Table 16):<br />
1. cultural and technological change<br />
2. <strong>in</strong>terregional contact<br />
3. paleoecology<br />
4. procurement technology<br />
5. specialized procurement technology (focus on one or a few target resources)<br />
6. settlement/subsistence systems<br />
7. trade<br />
Scientific Significance and Research Potential<br />
Us<strong>in</strong>g this list as a start<strong>in</strong>g po<strong>in</strong>t, and consider<strong>in</strong>g now all known floodpla<strong>in</strong> sites<br />
as a data base, we can suggest also a series <strong>of</strong> more specific research problems <strong>of</strong> current<br />
methodological and <strong>the</strong>oretical <strong>in</strong>terest, but oriented to representative samples <strong>of</strong> specialized<br />
extractive sites and transient camps:<br />
1. energetics <strong>of</strong> an overflow bottomland environment <strong>in</strong> relation to human<br />
adaptations (prehistoric and historic)<br />
2. formation and transformation processes <strong>of</strong> alluvial sites <strong>in</strong> an overflow<br />
bottomland sett<strong>in</strong>g<br />
3. recent geohydrological history <strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland<br />
4. geochronological problems and chronometric potential <strong>of</strong> alluvial sites <strong>in</strong><br />
an overflow bottomland sett<strong>in</strong>g<br />
5. archeological context and systemic context <strong>in</strong> specialized extractive sites<br />
(prehistoric)<br />
6. material correlates <strong>of</strong> <strong>in</strong>trasite extractive tasks and ma<strong>in</strong>tenance tasks (prehistoric)<br />
7. social or ethnic units which operated extractive sites (and variability<br />
among <strong>the</strong>se units)<br />
8. variability <strong>in</strong> discard and curate behavior among prehistoric floodpla<strong>in</strong><br />
site assemblages<br />
9. locational analysis <strong>of</strong> riverbank/levee extractive sites or transient camps,<br />
and <strong>of</strong> historic land<strong>in</strong>gs and stations<br />
10. catchment analysis <strong>of</strong> specialized extractive sites (prehistoric)<br />
11. fisheries ecology, seasonality, fish procurement, and fish process<strong>in</strong>g <strong>in</strong> <strong>the</strong><br />
<strong>Grand</strong> <strong>Marais</strong> Lowland (aborig<strong>in</strong>al and Euramerican)<br />
12. structure and function <strong>of</strong> a n<strong>in</strong>eteenth century riverboat land<strong>in</strong>g (especially<br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g)<br />
C-5
C-6<br />
13. site discovery strategies and <strong>in</strong>vestigation methodologies <strong>in</strong> an overflow<br />
bottomland environment (<strong>in</strong>clud<strong>in</strong>g remote sens<strong>in</strong>g)<br />
14. small site methodologies<br />
15. lithic behavioral systems <strong>in</strong> specialized extractive sites (prehistoric)<br />
16. ceramic behavioral systems <strong>in</strong> specialized extractive sites (especially <strong>in</strong><br />
regard to fish<strong>in</strong>g and fish process<strong>in</strong>g)<br />
<strong>17</strong>. Baytown ceramic technology and variability<br />
18. <strong>Grand</strong> <strong>Marais</strong> ceramic styles as sociobehavioral correlates<br />
19. submerged deposits and preservation <strong>of</strong> saturated materials (prehistoric<br />
and historic)<br />
20. impacts <strong>of</strong> artificially modified and controlled <strong>in</strong>undation on cultural resources<br />
<strong>of</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland<br />
This latter list emphasizes, and even dramatizes, <strong>the</strong> broad range <strong>of</strong> relevant data<br />
and current problems which we associate with <strong>the</strong> range <strong>of</strong> floodpla<strong>in</strong> sites identified,<br />
and analyzed <strong>in</strong> Chapters 5 and 6 (see also Appendix A). Floodpla<strong>in</strong> sites, <strong>in</strong> our view,<br />
can and do have qualities <strong>of</strong> anthropological and historical significance which supplement<br />
or exceed regional research problems; we will also refer to public significance later<br />
<strong>in</strong> this chapter. Many archeologists and cultural resource managers will know, as we<br />
do, that this list is <strong>in</strong>complete. Also, it is important to note that <strong>the</strong> list <strong>of</strong> specific research<br />
problems emphasizes what we perceive as <strong>the</strong> “particular cultural-environmental<br />
context” <strong>of</strong> <strong>the</strong> Felsenthal Project area. F<strong>in</strong>ally, <strong>the</strong> quality <strong>of</strong> research potential among<br />
known floodpla<strong>in</strong> sites ranges from very little to very great.<br />
We never<strong>the</strong>less have attempted rigorous, if difficult, identification <strong>of</strong> significant<br />
floodpla<strong>in</strong> sites which warrant mitigation recommendations <strong>of</strong> one k<strong>in</strong>d or ano<strong>the</strong>r. As<br />
Schiffer and House (1977:252) noted <strong>in</strong> <strong>the</strong> case <strong>of</strong> Cache Bas<strong>in</strong> cultural resources “we<br />
cannot now, or perhaps ever, assess <strong>the</strong> substantive significance [<strong>of</strong> <strong>the</strong>se resources] mechanically.”<br />
Our process <strong>of</strong> assessment <strong>in</strong>volved review <strong>of</strong> 1<strong>17</strong> floodpla<strong>in</strong> sites (plus one<br />
upland site, Goulett Island, 3BR8, which has extensive refuse downslope to floodpla<strong>in</strong><br />
level). This process featured:<br />
1. evaluation <strong>of</strong> all site survey, test excavation, and analytical data conta<strong>in</strong>ed <strong>in</strong><br />
relevant records and <strong>in</strong> this report<br />
2. review <strong>of</strong> <strong>the</strong> 1979 project research design (Appendix D), <strong>the</strong> research design<br />
outl<strong>in</strong>e for proposed mitigation presented below, <strong>the</strong> <strong>Arkansas</strong> State Plan, Volume<br />
II, and o<strong>the</strong>r regional or general sources, and<br />
3. consultation among <strong>the</strong> Project Archeologist, Pr<strong>in</strong>cipal Investigator, Project<br />
Historian, Station Archeologists, and o<strong>the</strong>r staff members at <strong>the</strong> <strong>Arkansas</strong> Archeological<br />
Survey.<br />
We now identify 37 floodpla<strong>in</strong> sites as scientifically significant, and among <strong>the</strong>se<br />
some are also anthropologically or historically significant. One site, Buttonbush (3BR58),
has been tested extensively, and because <strong>of</strong> small size and sparse content, is not believed<br />
to conta<strong>in</strong> additional cultural data (Chapter 6). We <strong>the</strong>refore state <strong>in</strong> Table 16 for Buttonbush:<br />
“no additional work [is needed]; mitigated by test<strong>in</strong>g.” (It follows that this site<br />
is not eligible for <strong>the</strong> National Register by reason <strong>of</strong> its removal.) A second site, Mo-Pac<br />
(3UN121), was recommended for test<strong>in</strong>g, and has <strong>in</strong> fact now been tested for significance<br />
by Historic Preservation Associates (see comments at beg<strong>in</strong>n<strong>in</strong>g <strong>of</strong> this appendix). In this<br />
case we have no additional data, and cannot properly comment fur<strong>the</strong>r on significance,<br />
eligibility, or requirements for mitigation. We are <strong>the</strong>refore recommend<strong>in</strong>g 35 (<strong>of</strong> 37 significant<br />
floodpla<strong>in</strong> sites identified) as eligible for nom<strong>in</strong>ation to <strong>the</strong> National Register <strong>of</strong><br />
Historic Places, as discussed fur<strong>the</strong>r below. These 35 sites are enumerated <strong>in</strong> Appendix A<br />
and <strong>in</strong> Tables C-1 and C-2 below.<br />
Table C-1. Six Felsenthal Project Floodpla<strong>in</strong> Sites Recommended for Extensive and<br />
Intensive Investigation.<br />
Site Elevation Location by<br />
Number Site Name (feet MSL) Dra<strong>in</strong>age Topographic Location Impacts<br />
3AS329 Marie Sal<strong>in</strong>e 62.0-67.5 Ouachita River riverbank/levee near open lake 1, 2, 3, 4<br />
3AS285 False Indigo 66.9-69.9 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
3AS286 One Cypress Po<strong>in</strong>t 66.6-68.4 Sal<strong>in</strong>e River riverbank/levee and old 2, 3, 4<br />
meander scar<br />
3BR76 Jug Po<strong>in</strong>t Cut<strong>of</strong>f 67.6-68.4 Sal<strong>in</strong>e River riverbank/levee near chute 2, 3, 4<br />
3AS306 Jug Po<strong>in</strong>t 1 67.2-68.1 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
3AS307 Jug Po<strong>in</strong>t 2 67.2-68.0 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
Note: Impacts are described on pages C-15 to C-<strong>17</strong>.<br />
Through <strong>the</strong> same process <strong>of</strong> site data review, evaluation, and consultation, <strong>the</strong><br />
rema<strong>in</strong><strong>in</strong>g 81 floodpla<strong>in</strong> sites were found to be not scientifically, anthropologically, or<br />
historically significant, and not eligible for nom<strong>in</strong>ation to <strong>the</strong> National Register. Prehistoric<br />
floodpla<strong>in</strong> sites <strong>in</strong> this group generally lacked <strong>the</strong>se k<strong>in</strong>ds <strong>of</strong> research potential:<br />
C-7<br />
1. useful samples <strong>of</strong> artifacts, debris, or organic rema<strong>in</strong>s,<br />
2. <strong>in</strong> situ features, structures, occupation levels, or multiple occupation levels, and<br />
3. items or data unique to <strong>the</strong> locality or region.<br />
This statement is based on site survey methods (Chapter 5) and data recovered (Appendix<br />
A).
C-8<br />
Table C-2. Twenty-n<strong>in</strong>e Felsenthal Project Floodpla<strong>in</strong> Sites Recommended for Twostage<br />
Mitigation.<br />
Site Elevation Location by Topographic Probable<br />
Number Site Name (feet MSL) Dra<strong>in</strong>age Location Impacts<br />
Group 1 - Prehistoric Floodpla<strong>in</strong> Sites, Gran <strong>Marais</strong> and Caney Bayou Phases, Mississippi Period<br />
3AS305 Overcup Oak 3 66.0-67.01 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
3BR62 Mouth <strong>of</strong> Long Slough 65.0-66.0 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
38865 Eagle Creek 3 67.5 Eagle Creek creekbank/levee 2, 3, 4<br />
3BR66 Eagle Creek 4 66.0-68.01 Eagle Creek creekbank/levee 2, 3, 4<br />
3UN168 Coon Glory Bend 63.0-65.01 Ouachita River levee near riverbank 1, 2<br />
Group 2 - Prehistoric Floodpla<strong>in</strong> Sites, Unknown Phases, Mississippi Period<br />
3AS315 Persimmon 3 68.0-70.01 Sal<strong>in</strong>e River levee near riverbank 3, 4<br />
3AS322 Water Elm 1 67.5 Ouachita River riverbank/levee 2, 3, 4<br />
3BR79 Eagle Creek 6 65.0 Eagle Creek creekbank/levee 2, 3, 4<br />
3BR86 Sweetgum 2 67.5 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
3UN<strong>17</strong>9 Fishtrap Lake 4 63.0-68.01 Lapile Bayou bayou bank 2, 3, 4<br />
Group 3 - Prehistoric Floodpla<strong>in</strong> Sites, Unknown Phases, Baytown-Coles Creek Period<br />
3AS287 Water Elm 3 67.0 Ouachita River riverbank/levee 2, 3, 4<br />
3AS316 Hawthorn 1 64.0-67.0 Sal<strong>in</strong>e River riverbank/levee 1, 2, 3, 4<br />
3AS327 Swamp Privet 66.0 Ouachita River levee near riverbank 2, 3, 4<br />
3BR82 Bitter Pecan 1 68.0-69.0 Sal<strong>in</strong>e River riverbank/levee 3, 4<br />
3UN151 Lapile Creek 7 64.0-67.01 Lapile Creek creekbank 2, 3, 4<br />
3UN157 Lapile Creek 5 65.0-69.01 Lapile Creek creekbank 2, 3, 4<br />
3UN180 Fishtrap Lake 2 63.0-66.01 Lapile Bayou bayou bank 2, 3, 4<br />
Group 4 - Prehistoric Floodpla<strong>in</strong> Sites, Coon Island Phase, Tchula Period<br />
3AS321 River Birch 2 63.0-67.01 Ouachita River riverbank/levee 2, 3, 4<br />
Group 5 - Prehistoric Floodpla<strong>in</strong> Sites, Calion Phase, Poverty Po<strong>in</strong>t Period<br />
3BR78 Mouth <strong>of</strong> Eagle Creek 63.2-68.0 Sal<strong>in</strong>e River riverbank/levee 1, 2, 3, 4<br />
Group 6 - Prehistoric Floodpla<strong>in</strong> Sites, Unknown Phases, Archaic Periods<br />
3BR72 Small Cane 1 70.0-73.51 Ouachita River riverbank/levee 3, 4<br />
3BR73 Small Cane 2 71.0 Ouachita River riverbank/levee 3, 4<br />
3BR92 Water Locust 65.0-67.0 Sal<strong>in</strong>e River riverbank/levee 2, 3, 4<br />
3UN166 Mud Lake Bend 2 71.1-72.4 Ouachita River riverbank/levee 3, 4<br />
Group 7 - Historic Floodpla<strong>in</strong> and River Channel Sites, N<strong>in</strong>eteenth and Early Twentieth Century American<br />
3AS299 Marie Sal<strong>in</strong>e Land<strong>in</strong>g 61.0-69.0 Ouachita River riverbank/levee 1, 2, 4, 5<br />
3BR8 Goulett Island 63.0-86.0 Sal<strong>in</strong>e River Deweyville 2 Terrace 1, 2, 3, 4<br />
3BR55 Keelboat Brake Sawmill 66.0-68.0 Ouachita River backswamp 2, 3, 4<br />
3UN122 <strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g 64.0-65.0 Ouachita River riverbank/levee 1, 2, 5<br />
3UN153 Wreck <strong>of</strong> Lotawanna 34.0-38.0 Ouachita River river bottom 1, 5<br />
3BR56 Caney Marie Land<strong>in</strong>g 74.0-76.0 Ouachita River riverbank/levee 3, 4<br />
Note: Impacts are described on pages C-15 to C-<strong>17</strong>.
Historic floodpla<strong>in</strong> sites among <strong>the</strong> 81 sites not found to be significant generally lacked<br />
<strong>the</strong> follow<strong>in</strong>g k<strong>in</strong>ds <strong>of</strong> research potential:<br />
1. useful samples <strong>of</strong> artifacts, debris, or organic rema<strong>in</strong>s<br />
2. stand<strong>in</strong>g structures or subsurface features<br />
3. data bear<strong>in</strong>g on historic research problems <strong>in</strong> any <strong>of</strong> our research design<br />
sources, and<br />
4. association with an important historic event or person.<br />
This statement is based on site survey methods (Chapter 5), data recovered (Appendix<br />
A), and historical or archival research perta<strong>in</strong><strong>in</strong>g to specific sites, localities, and <strong>the</strong><br />
Felsenthal Region (Chapter 4).<br />
The determ<strong>in</strong>ation <strong>of</strong> “signficant/not significant” has been applied here <strong>in</strong> a poly<strong>the</strong>tic<br />
analytical framework (cf. Clarke 1968:35ff). In short, no s<strong>in</strong>gle criterion was both<br />
sufficient and necessary for a site to be classed as significant. Ra<strong>the</strong>r specific poly<strong>the</strong>tic<br />
sets <strong>of</strong> <strong>the</strong> 26 criteria, as presented <strong>in</strong> Table C-3, determ<strong>in</strong>ed whe<strong>the</strong>r or not <strong>the</strong> site is<br />
assessed as significant. Fur<strong>the</strong>r, it is recognized that <strong>the</strong>re are patterned regularities <strong>in</strong><br />
site types and properties. Given <strong>the</strong>se regularities, it is possible to recognize only part <strong>of</strong><br />
a set <strong>of</strong> similar sites as representative <strong>of</strong> <strong>the</strong> entire set. Additional study is not needed <strong>in</strong><br />
members <strong>of</strong> a given group<strong>in</strong>g—<strong>in</strong>stead exam<strong>in</strong>ation <strong>of</strong> representative sites will ga<strong>the</strong>r<br />
sufficient <strong>in</strong>formation to mitigate <strong>the</strong> impacts at all <strong>the</strong> sites. In a sense while we can<br />
state that all members <strong>of</strong> <strong>the</strong> group are “significant,” for <strong>the</strong> purposes <strong>of</strong> <strong>the</strong> management<br />
<strong>of</strong> those particular resources, selected members <strong>of</strong> <strong>the</strong> group can be and have been<br />
selected as representative. It is for <strong>the</strong>se that nom<strong>in</strong>ation documentation has been prepared.<br />
The rema<strong>in</strong>der <strong>of</strong> <strong>the</strong> group are <strong>the</strong>n seen, from a resource management perspective,<br />
as redundant sites and <strong>the</strong>reby deemed not eligible for <strong>the</strong> National Register <strong>of</strong><br />
Historic Places.<br />
Of <strong>the</strong> various factors considered and enumerated <strong>in</strong> Table C-3, a relatively large<br />
number <strong>of</strong> artifacts (diagnostic ceramics present) factors #1 and #2 proved to be particularly<br />
critical to determ<strong>in</strong>ation <strong>of</strong> a prehistoric site significance. Only three sites<br />
determ<strong>in</strong>ed to be significant did not have #1, all <strong>of</strong> which had #2 plus o<strong>the</strong>r important<br />
factors; only eight sites considered significant had #1 but did not have #2 and all but one<br />
<strong>of</strong> <strong>the</strong>se had o<strong>the</strong>r important factors. That is, all sites considered significant had #1 or #2<br />
(two-thirds had both) and all had o<strong>the</strong>r factors to add to <strong>the</strong> decision mak<strong>in</strong>g equation.<br />
Of <strong>the</strong>se o<strong>the</strong>r factors, five proved most critical: diagnostic lithics, carbonized materials,<br />
<strong>in</strong> situ deposits, features present, and dist<strong>in</strong>ctive environmental association. All prehistoric<br />
sites which did have <strong>the</strong> above characteristics, but were not considered significant,<br />
were excluded because <strong>the</strong>y were believed to be redundant to o<strong>the</strong>rs which had similar<br />
characteristics and which displayed <strong>the</strong>m more obtrusively or <strong>in</strong> greater quantity, or had<br />
o<strong>the</strong>r comb<strong>in</strong>ations <strong>of</strong> important factors present.<br />
The historic sites developed no patterns with respect to <strong>the</strong> importance <strong>of</strong> particular<br />
criteria or comb<strong>in</strong>ations <strong>of</strong> criteria <strong>in</strong> determ<strong>in</strong><strong>in</strong>g significance. Each site was more or<br />
less unique, perhaps because <strong>of</strong> <strong>the</strong> relatively small number <strong>of</strong> sites.<br />
C-9
C-10<br />
Table C-3 summarizes <strong>the</strong> qualities <strong>of</strong> significance found to be present or lack<strong>in</strong>g<br />
<strong>in</strong> each <strong>of</strong> 1<strong>17</strong> floodpla<strong>in</strong> sites surveyed or tested by <strong>the</strong> Felsenthal Project <strong>in</strong> 1979. These<br />
sites are listed by site unit <strong>in</strong> <strong>the</strong> order <strong>the</strong>y appear <strong>in</strong> Chapter 5 and Appendix A, Table<br />
A-1. Although this table gives <strong>the</strong> appearance <strong>of</strong> a mechanical assessment <strong>of</strong> significance,<br />
<strong>the</strong> preced<strong>in</strong>g discussion has expla<strong>in</strong>ed <strong>the</strong> more complex process by which significance,<br />
or lack <strong>of</strong> significance, was identified.<br />
nAtionAl register eVAluAtion<br />
The 35 sites found to be scientifically significant are recommended for nom<strong>in</strong>ation<br />
to <strong>the</strong> National Register <strong>of</strong> Historic Places (36 CFR 1202 and 1204). Each can be said to<br />
“have yielded or [is] likely to yield <strong>in</strong>formation important <strong>in</strong> prehistory or history.” This<br />
National Register criterion is frequently applied to archeological properties, but warrants<br />
more detailed treatment, as <strong>in</strong> our discussion <strong>of</strong> significance.<br />
In compliance with <strong>the</strong> contract Scope <strong>of</strong> Services, <strong>in</strong>dividual National Register<br />
nom<strong>in</strong>ations have been prepared for 16 significant sites that have archeological deposits<br />
at or below <strong>the</strong> 65 foot elevation.<br />
Recommendation for Multiple Resource Area Nom<strong>in</strong>ation<br />
We suggest that a Multiple Resource Area, called “Cultural Resources <strong>of</strong> <strong>the</strong> <strong>Grand</strong><br />
<strong>Marais</strong> Floodpla<strong>in</strong> Area, <strong>Arkansas</strong>,” be nom<strong>in</strong>ated to <strong>the</strong> National Register (Interim<br />
Guidel<strong>in</strong>es: How to Complete National Register Multiple Resource Nom<strong>in</strong>ation Forms,<br />
1977). This MRA shall <strong>in</strong>clude <strong>the</strong> 35 known eligible properties, <strong>in</strong>clud<strong>in</strong>g 29 prehistoric<br />
and six historic floodpla<strong>in</strong> sites. (Goulett Island, 3BR8, has significant prehistoric and<br />
historic components above and below floodpla<strong>in</strong> level, but we are not attempt<strong>in</strong>g here<br />
to nom<strong>in</strong>ate <strong>the</strong> entire elevated “island” itself because <strong>of</strong> survey and study limitations;<br />
our survey data established that prehistoric and historic artifacts and refuse extended<br />
downslope to floodpla<strong>in</strong> and river level, and <strong>the</strong>se deposits are identified as significant).<br />
The geographical extent <strong>of</strong> <strong>the</strong> proposed MRA is all <strong>of</strong> <strong>the</strong> Recent floodpla<strong>in</strong> <strong>of</strong><br />
Ouachita and Sal<strong>in</strong>e rivers and <strong>the</strong>ir tributaries with<strong>in</strong> <strong>the</strong> Felsenthal National Wildlife<br />
Refuge. (This floodpla<strong>in</strong> and some adjacent lowly<strong>in</strong>g Pleistocene terrace surface are an<br />
environmental unit we have <strong>in</strong>formally called “overflow bottomland.”) We can suggest<br />
and provide map series which will spatially def<strong>in</strong>e <strong>the</strong> proposed MRA. Based on our<br />
<strong>in</strong>tensive cultural resources survey and test<strong>in</strong>g <strong>in</strong> floodpla<strong>in</strong> terra<strong>in</strong> (but not comprehensive<br />
survey undertaken to identify all resources), <strong>the</strong> MRA <strong>in</strong>cludes 35 sites deemed<br />
eligible, 82 sites presently deemed not eligible, and one site (3UN121) await<strong>in</strong>g eligibility<br />
recommendation. It is probable that more sites <strong>of</strong> significance will be discovered and<br />
added at a future time. It may <strong>the</strong>refore be appropriate to append <strong>the</strong> notation “partial<br />
<strong>in</strong>ventory” to our MRA.
Table C-3. Summary <strong>of</strong> Site Characteristics Relevant to Research Potential and<br />
Scientific or Historic Significance for 1<strong>17</strong> Floodpla<strong>in</strong> Sites.<br />
Content Context Significance Determ<strong>in</strong>ation<br />
Site Number Site Name (see explanatory note at end <strong>of</strong> Table)<br />
UNIT 1: Ouachita River above Mile 254<br />
3UN167 Mud Lake Bend 1 B<br />
3UN166 Mud Lake Bend 2 1 7 A<br />
3UN165 Mud Lake Bend 3 B<br />
3UN164 Stormhole Bend B<br />
3BR71 Parrigee<strong>the</strong> Shoals 1 1 F<br />
3BR33 Parrigee<strong>the</strong> Shoals 2 B<br />
3UN163 — B<br />
3BR75 Eutaw Rapids 2, 5 7 F<br />
3UN162 — B<br />
3BR73 Small Cane 2 1, 3 7 A<br />
3BR74 — B<br />
3BR72 Small Cane 1 1, 3, 4 A<br />
3UN161 Fletchers Land<strong>in</strong>g 4 7 C<br />
3BR70 Hunter’s Swan 2 C<br />
3BR57 Little Bay Rapids B<br />
3UN188 Five Mile Bend 2 C<br />
3UN185 — B<br />
3BR77 Lower Sal<strong>in</strong>e Cut<strong>of</strong>f E<br />
UNIT 2: Lower Sal<strong>in</strong>e River<br />
3BR92 Water Locust 1 7 A<br />
3BR91 — B<br />
3BR90 Round Turn B<br />
3BR89 Mouth <strong>of</strong> Miller Creek 2 C<br />
3BR88 — B<br />
3BR87 Horton Island Land<strong>in</strong>g 2 C<br />
3BR86 Sweetgum 2 1, 2 7 A<br />
3BR85 Sweetgum 1 B<br />
3AS318 Hawthorn 3 B<br />
3AS3<strong>17</strong> Hawthorn 2 B<br />
3AS316 Hawthorn 1 2, 4 7, 10 A<br />
3BR81 Prairie Island Land<strong>in</strong>g B<br />
3BR80 Prairie Island Bend 2 7 F<br />
3AS315 Persimmon 3 1 A<br />
3br82 Bitter Pecan 1 1, 2, 4, 5 7, 10 A<br />
3AS314 Persimmon 2 1 F<br />
3BR84 Bitter Pecan 3 B<br />
3br83 bitter pecan 2 b<br />
3AS313 Persimmon 1 B<br />
3BR78 Mouth <strong>of</strong> Eagle Creek 1, 3, 4 7, 8, 10 A<br />
3AS312 Willow Oak B<br />
3AS311 Tupelo Gum B<br />
3AS309 Jug Po<strong>in</strong>t 4 B<br />
3AS310 — B<br />
3AS308 Jug Po<strong>in</strong>t 3 1, 2 F<br />
3AS307 Jug Po<strong>in</strong>t 2 1, 2, 3, 4, 5 7, 8 A<br />
3AS306 Jug Po<strong>in</strong>t 1 1, 2, 3, 4, 5, 6 7, 8, 10 A<br />
3BR76 Jug Po<strong>in</strong>t Cut<strong>of</strong>f 1, 2, 3, 4, 5 7, 8, 11, 12 A<br />
3AS305 Overcup Oak 3 1, 2 A<br />
3AS304 Overcup Oak 2 2 C<br />
3AS303 Overcup Oak 1 1, 2 F<br />
3AS302 Sycamore B<br />
3AS286 One Cypress Po<strong>in</strong>t 1, 2, 3, 4, 5 7, 8, 12 A<br />
3BR59 Possumhaw B<br />
3BR58 Buttonbush 1, 2, 3 7, 10 A<br />
3BR60 — B<br />
3BR61 — B<br />
3BR62 Mouth <strong>of</strong> Long Slough 2, 4, 5 7 A<br />
3AS285 False Indigo 1, 2, 3, 4, 5 7, 8, 10, 11 A<br />
3AS284 Mouth <strong>of</strong> Sal<strong>in</strong>e River B<br />
C-11
C-12<br />
Table C-3 (cont<strong>in</strong>ued). Summary <strong>of</strong> Site Characteristics Relevant to Research Potential<br />
and Scientific or Historic Significance for 1<strong>17</strong> Floodpla<strong>in</strong> Sites.<br />
Content Context Significance Determ<strong>in</strong>ation<br />
Site Number Site Name (see explanatory note at end <strong>of</strong> Table)<br />
UNIT 3: Lower Eagle Creek<br />
3BR68 — B<br />
3BR67 — 2 C<br />
3BR66 Eagle Creek 4 1, 2, 4, 5 10 A<br />
3BR69 Eagle Creek 5 2 C<br />
3BR65 Eagle Creek 3 1, 2, 4 7 A<br />
3BR64 Eagle Creek 2 B<br />
3BR63 Eagle Creek 1 1, 2 F<br />
3BR79 Eagle Creek 6 1 7 A<br />
UNIT 4: Ouachita River below Mile 254<br />
3UN184 — 2 B<br />
3UN92 Brown Camp 2 C<br />
3AS329 Marie Sal<strong>in</strong>e 1, 2, 3, 4, 6 7, 8, 9, 10, 12 A<br />
3AS326 — 3 C<br />
3AS301 Honey Locust 2 C<br />
3AS319 — 7 C<br />
3As320 river birch 1 b<br />
3As321 River Birch 2 1, 2 A<br />
3AS288 Red Oak B<br />
3As322 Water Elm 1 2 7 A<br />
3AS332 Water Elm 2 B<br />
3AS287 Water Elm 3 1, 2 7, 10 A<br />
3AS323 Water Elm 4 1, 2 7 F<br />
3AS324 — 3 7 C<br />
3UN123 Mouth <strong>of</strong> Lapile Creek 1, 2 F<br />
3AS327 Swamp Privet 1, 2 7, 10 A<br />
3AS325 — 2 7 F<br />
3UN169 — B<br />
3UN168 Coon Glory Bend 1, 2 A<br />
3UN121 Mo-Pac 2, 4 7 F<br />
UNIT 5: Lower Lapile Creek<br />
3UN<strong>17</strong>8 Lapile Creek 1 B<br />
3UN160 Lapile Creek 2 2 C<br />
3UN159 Lapile Creek 3 2 C<br />
3UN158 Lapile Creek 4 1, 2 F<br />
3UN157 Lapile Creek 5 1, 2, 3 A<br />
3UN156 Lapile Creek 6 1 F<br />
3UN152 Lapile Creek 8 B<br />
3UN151 Lapile Creek 7 1, 2, 3 A<br />
UNIT 6: Oxbow Lakes and Backswamp<br />
3UN187 St. Mary’s Lake North 1 7, 12 C<br />
3UN186 — B<br />
3UN<strong>17</strong>0 Borrow Pit near Lapoile Creek D<br />
3UN183 Lapoile Creek B<br />
3UN182 Fishtrap Lake 1 1 12 F<br />
3un180 Fishtrap Lake 2 1 12 A<br />
3UN181 Fishtrap Lake 3 12 C<br />
3UN<strong>17</strong>9 Fishtrap Lake 4 1, 2 12 A<br />
3AS295 Borrow Pit near 82 Bridge 1 1, 2 D<br />
3AS294 Borrow Pit near 82 Bridge 2 1 D<br />
3AS292 Wildcat Lake 1 12 F
Table C-3 (cont<strong>in</strong>ued). Summary <strong>of</strong> Site Characteristics Relevant to Research Potential<br />
and Scientific or Historic Significance for 1<strong>17</strong> Floodpla<strong>in</strong> Sites.<br />
Content Context Significance Determ<strong>in</strong>ation<br />
Site Number Site Name (see explanatory note at end <strong>of</strong> Table)<br />
UNIT 8: Historic Sites<br />
3BR32 Ganer Construction Mound 15, <strong>17</strong> 20, 23 G<br />
3BR56 Caney Marie Land<strong>in</strong>g 13, 14 19, 20 A<br />
3br8 Goulett Island 13, 14, 16 20, 22, 23 A<br />
3BR81 Prairie Island Land<strong>in</strong>g B<br />
3BR55 Keelboat Brake Sawmill 15, 16 20, 22 A<br />
3UN155 Ouachita Belle Land<strong>in</strong>g B<br />
3UN88 Fletchers Land<strong>in</strong>g and Woodyard 13, <strong>17</strong> C<br />
3UN92 Brown Camp 20 C<br />
3AS330 First Slough 13, <strong>17</strong> 22 G<br />
3AS331 Marie Sal<strong>in</strong>e Ferry 13, 14 23 G<br />
3AS300 South <strong>of</strong> Highway 82 Bridge 14, <strong>17</strong> 22 G<br />
3AS299 Marie Sal<strong>in</strong>e Land<strong>in</strong>g 20, 23 A<br />
3UN<strong>17</strong>4 Artesian Well B<br />
3un122 <strong>Grand</strong> <strong>Marais</strong> Land<strong>in</strong>g 16 19, 21, 23 A<br />
3UN153 Wreck <strong>of</strong> <strong>the</strong> Lotawanna 23, 24 A<br />
Notes: Historic sites 3UN81 and 3UN92 have prehistoric components and appear twice <strong>in</strong> this table. The follow<strong>in</strong>g explanatory<br />
notes perta<strong>in</strong> to columns 3-5 <strong>in</strong> this table:<br />
prehistoric content<br />
(1) Sample <strong>of</strong> artifacts and/or debris recovered was substantial relative to site size, exposure, extent <strong>of</strong> test<strong>in</strong>g, or o<strong>the</strong>r factors.<br />
(2) One or more temporally or culturally diagnostic ceramic sherds were recovered.<br />
(3) One or more temporally or culturally disgnostic projectile po<strong>in</strong>ts were recovered.<br />
(4) Carbonized seeds or charcoal were recovered.<br />
(5) Animal bone or mussel shell were recovered.<br />
(6) Artifacts evidently unique to <strong>the</strong> locality or region were recovered.<br />
prehistoric context<br />
(7) One or more buried artifacts or items <strong>of</strong> debris were recorded <strong>in</strong> place.<br />
(8) A visible buried occupation level marked by artifacts and/or organic sta<strong>in</strong> was observed.<br />
(9) Superposed buried occupation levels were observed.<br />
(10) One or more subsurface features were recorded.<br />
(11) A midden or refuse deposit was recorded.<br />
(12) Association <strong>of</strong> site with a relict floodpla<strong>in</strong> feature or o<strong>the</strong>r dist<strong>in</strong>ctive environmental feature was observed.<br />
historic content<br />
(13) Sample <strong>of</strong> artifacts and/or debris recovered was substantial relative to site size, exposure, or o<strong>the</strong>r factors.<br />
(14) One or more temporally diagnostic artifacts were recovered.<br />
(15) One or more technologically diagnostic artifacts were recovered.<br />
(16) Wood or charcoal was recovered.<br />
(<strong>17</strong>) Animal bone, mussel shell, or oyster shell was recovered.<br />
(18) Artifacts evidently unique to <strong>the</strong> locality or region were recovered.<br />
historic context<br />
(19) One or more buried artifacts or items <strong>of</strong> debris were recorded <strong>in</strong> place.<br />
(20) Above-ground structural rema<strong>in</strong>s were observed.<br />
(21) One or more subsurface features were recorded.<br />
(22) A refuse deposit was recorded.<br />
(23) Site is associated with archival documents or secondary sources.<br />
(24) Site is associated with an important event or person.<br />
C-13
C-14<br />
Table C-3 (concluded). Summary <strong>of</strong> Site Characteristics Relevant to Research Potential<br />
and Scientific or Historic Significance for 1<strong>17</strong> Floodpla<strong>in</strong> Sites.<br />
significAnce deterM<strong>in</strong>Ation<br />
(A) Significant (See discussion on page C- 4).<br />
(B) Not significant because no diagnostic criteria present.<br />
(C) Not significant because <strong>in</strong>sufficient criteria present.<br />
(D) Not significant because riverbank erosion or some o<strong>the</strong>r natural disturbance is believed to have destroyed a large part <strong>of</strong> <strong>the</strong><br />
site.<br />
(E) Not significant because historic or modern cultural disturbance is believed to have destroyed a large part <strong>of</strong> <strong>the</strong> site.<br />
(F) Not significant because redundant—o<strong>the</strong>r similar sites determ<strong>in</strong>ed eligible have greater potential for produc<strong>in</strong>g significant<br />
data.<br />
(G) Not significant because does not conform to National Register guidel<strong>in</strong>es.<br />
With regard to <strong>the</strong> proposed MRA, all 35 eligible properties are believed to be<br />
under federal ownership presently (ei<strong>the</strong>r <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers or <strong>the</strong> U.S.<br />
Fish and Wildlife Service), but some properties deemed not eligible and some <strong>in</strong>trusive<br />
properties not recorded as archeological or historical sites are privately owned presently<br />
(e.g., modern fish<strong>in</strong>g camps and boat land<strong>in</strong>gs).<br />
F<strong>in</strong>ally, we believe that choice <strong>of</strong> <strong>the</strong> Multiple Resource Area concept will lead to<br />
practical management <strong>of</strong> <strong>the</strong> nom<strong>in</strong>ation process, now and <strong>in</strong> <strong>the</strong> future, and that National<br />
Register status <strong>of</strong> <strong>the</strong> proposed “<strong>Grand</strong> <strong>Marais</strong> Floodpla<strong>in</strong> Area, <strong>Arkansas</strong>” will<br />
truly enhance recognition and wise management <strong>of</strong> significant cultural resources <strong>in</strong> this<br />
area and region.<br />
AssessMent <strong>of</strong> site iMpActs<br />
Felsenthal Project contract requirements <strong>in</strong>clude discussion <strong>of</strong> probable impact on<br />
cultural resources through proposed work <strong>of</strong> <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers (Scope<br />
<strong>of</strong> Services, item 7b). Schiffer and Gumerman (1977:291), among o<strong>the</strong>rs, po<strong>in</strong>t out that<br />
responsible and <strong>in</strong>formed forecast <strong>of</strong> impacts is <strong>of</strong> extreme importance to sponsors as<br />
resource managers. Through review <strong>of</strong> various construction maps provided to us, <strong>in</strong>formal<br />
discussions with <strong>the</strong> Contract<strong>in</strong>g Officer’s representatives, and review <strong>of</strong> current<br />
archeological and management documents perta<strong>in</strong><strong>in</strong>g to reservoir <strong>in</strong>undation, we have<br />
prepared a statement <strong>of</strong> probable impacts which can be adjudged on a site-by-site basis.<br />
In fact, <strong>in</strong> Tables C-1 and C-2 below we do predict one or more impacts <strong>in</strong> <strong>the</strong> case <strong>of</strong><br />
each site deemed eligible for <strong>the</strong> National Register <strong>of</strong> Historic Places.
C-15<br />
Our statement <strong>of</strong> probable impacts recognizes that <strong>in</strong>undation studies for <strong>the</strong> purpose<br />
<strong>of</strong> conserv<strong>in</strong>g or mitigat<strong>in</strong>g effects on cultural resources require expert assistance<br />
from eng<strong>in</strong>eers, geologists, hydrologists, and o<strong>the</strong>r scientists. We would <strong>the</strong>refore welcome<br />
review and assistance <strong>in</strong> this area. For present purposes we have relied on certa<strong>in</strong><br />
<strong>in</strong>undation studies, many <strong>of</strong> which are general models or prelim<strong>in</strong>ary reports <strong>of</strong> work <strong>in</strong><br />
particular reservoirs on few sites (Garrison 1975, 1977; Carell et al. 1976; Schnell and Tyler<br />
1976; Garrison et al. 1977; Lenihan et al. 1977). As <strong>in</strong> <strong>the</strong> case <strong>of</strong> assess<strong>in</strong>g significance,<br />
we believe <strong>the</strong>re is a particular cultural-environmental context <strong>in</strong> <strong>the</strong> Felsenthal Project<br />
area, aga<strong>in</strong>st which impacts should be assessed. Some prelim<strong>in</strong>ary conclusions <strong>in</strong>clude<br />
<strong>the</strong> follow<strong>in</strong>g:<br />
1. All floodpla<strong>in</strong> sites now known undergo prolonged submergence each<br />
year as <strong>the</strong> result <strong>of</strong> previous artificially raised river level (Lock and Dam<br />
No. 6), as well as <strong>the</strong> natural annual hydroperiod (Chapter 2).<br />
2. Many known floodpla<strong>in</strong> sites lie on riverbanks and levees along meander<strong>in</strong>g<br />
channels which are stable, or have a markedly low rate <strong>of</strong> lateral erosion<br />
compared to o<strong>the</strong>r sou<strong>the</strong>ast U.S. rivers (Chapter 2).<br />
3. The net effect <strong>of</strong> <strong>in</strong>creased or permanent <strong>in</strong>undation caused by new navigation<br />
pool and wildlife pool levels will be adverse on floodpla<strong>in</strong> site<br />
content and context, not merely preservation under a water column (see<br />
<strong>in</strong>undation studies cited above).<br />
4. A variety <strong>of</strong> chemical, mechanical, and biological effects <strong>of</strong> <strong>in</strong>undation on<br />
archeological resources have been studied or methods <strong>of</strong> study are be<strong>in</strong>g<br />
developed (e.g., Lenihan et al. 1977).<br />
5. We are consider<strong>in</strong>g also non<strong>in</strong>undation impacts, such as construction <strong>in</strong><br />
floodpla<strong>in</strong> areas and recreational effects, but our basis for assessment <strong>of</strong><br />
<strong>the</strong>se impacts is <strong>in</strong>sufficient.<br />
On <strong>the</strong> basis <strong>of</strong> <strong>the</strong>se comments and considerations, <strong>the</strong> follow<strong>in</strong>g five categories <strong>of</strong><br />
probable impacts are presented (<strong>in</strong> special reference to variation among floodpla<strong>in</strong> areas<br />
and sites by elevation).<br />
1. Permanent Inundation Impacts<br />
All sites or portions <strong>of</strong> sites ly<strong>in</strong>g at or below 65 feet elevation will be permanently<br />
<strong>in</strong>undated. Direct mechanical, chemical, or biological impacts may be expected to affect<br />
all such sites <strong>in</strong> vary<strong>in</strong>g degrees. On a site-by-site basis impacts may vary <strong>in</strong> <strong>the</strong> follow<strong>in</strong>g<br />
dimensions:
C-16<br />
Site location relative to section <strong>of</strong> new navigation pool (upstream, downstream,<br />
midpool)<br />
Site location relative to exist<strong>in</strong>g channels, bends, and floodpla<strong>in</strong> features<br />
Site physical conditions, <strong>in</strong>clud<strong>in</strong>g sediment and vegetation cover, and exist<strong>in</strong>g<br />
exposure<br />
Site content, <strong>in</strong>clud<strong>in</strong>g soil matrix, features, artifacts, organic and <strong>in</strong>organic debris<br />
Depth <strong>of</strong> submergence by new navigation pool<br />
2. Navigation Pool Shorel<strong>in</strong>e Impacts<br />
All sites or portion <strong>of</strong> sites ly<strong>in</strong>g at or near 65 feet elevation will be subject to shorel<strong>in</strong>e<br />
disturbance processes, <strong>in</strong>clud<strong>in</strong>g any or all <strong>of</strong> <strong>the</strong> follow<strong>in</strong>g:<br />
Wave current erosion produced by large commercial watercraft or smaller pleasure<br />
craft, and also by w<strong>in</strong>d<br />
Scour<strong>in</strong>g by downstream flow and by seasonal or occasional changes <strong>in</strong> flow<br />
Biological activity, <strong>in</strong>clud<strong>in</strong>g <strong>the</strong> addition or removal <strong>of</strong> plant and animal populations<br />
which <strong>in</strong>habit or visit <strong>the</strong> shorel<strong>in</strong>e<br />
These impacts will vary among sites located immediately upon a river or bayou<br />
bankl<strong>in</strong>e and sites located away from such bankl<strong>in</strong>es.<br />
3. Wildlife Pool Shorel<strong>in</strong>e Impacts<br />
All sites or portions <strong>of</strong> sites from 65 feet to 70 feet elevation and for some distance<br />
above 70 feet will be subject to <strong>the</strong> effects <strong>of</strong> a temporary or seasonal pool level and<br />
seasonally reduced pool level. Chemical, mechanical, and biological impacts associated<br />
with <strong>in</strong>undation and shorel<strong>in</strong>e disturbance will vary among sites, depend<strong>in</strong>g on location,<br />
elevation, context, and o<strong>the</strong>r factors.<br />
4. Recreational Activity Impacts<br />
Increased or shift<strong>in</strong>g recreational use <strong>of</strong> shorel<strong>in</strong>es, levees adjo<strong>in</strong><strong>in</strong>g channels,<br />
backswamp areas, and elevated “islands” will affect floodpla<strong>in</strong> terra<strong>in</strong> at and near <strong>the</strong><br />
new pool levels and shorel<strong>in</strong>es. Based on exist<strong>in</strong>g uses, <strong>the</strong> mechanical impacts on sites<br />
<strong>in</strong> <strong>the</strong>se areas could <strong>in</strong>clude boat land<strong>in</strong>g and launch<strong>in</strong>g, four-wheel drive vehicle operation,<br />
bankl<strong>in</strong>e fish<strong>in</strong>g, hunt<strong>in</strong>g, and camp<strong>in</strong>g. Collect<strong>in</strong>g <strong>of</strong> artifacts exposed by erosion<br />
or digg<strong>in</strong>g <strong>in</strong> floodpla<strong>in</strong> sites may be stimulated by raised and lowered pool levels and<br />
<strong>the</strong>ir mechanical effects.
5. Direct Construction Impacts<br />
C-<strong>17</strong><br />
All sites with<strong>in</strong> or, <strong>in</strong> some cases near, construction rights-<strong>of</strong>-way will be impacted<br />
by excavation, spoil deposit, heavy equipment operation, and o<strong>the</strong>r similar factors.<br />
These impacts are anticipated primarily for <strong>the</strong> new Felsenthal Lock and Dam structures<br />
and for <strong>the</strong> Crossett Harbor facility. Future construction <strong>of</strong> bendways, boat launch<strong>in</strong>g<br />
facilities, access roads, boat channels, or o<strong>the</strong>r structures may affect floodpla<strong>in</strong> sites.<br />
recoMMended progrAM <strong>of</strong> MitigAtion<br />
Our recommended program <strong>of</strong> mitigation is based on several general considerations.<br />
First, that we have assembled and <strong>in</strong>terpreted an adequate and sound body<br />
<strong>of</strong> data regard<strong>in</strong>g cultural resources <strong>in</strong> <strong>the</strong> Felsenthal Project area (despite <strong>the</strong> many<br />
limitations discussed throughout this report). Second, that we have made an adequate<br />
appraisal <strong>of</strong> site significance, based on current regional knowledge and broad based<br />
research goals. Third, that 35 floodpla<strong>in</strong> sites recommended for nom<strong>in</strong>ation to <strong>the</strong> National<br />
Register <strong>of</strong> Historic Place (through <strong>the</strong> mechanism <strong>of</strong> a Multiple Resource Area)<br />
will all be adversely affected to some degree by Felsenthal Navigation Pool structures<br />
or modifications <strong>of</strong> floodpla<strong>in</strong> terra<strong>in</strong>. And fourth, that mitigative actions proposed here<br />
will need to proceed dur<strong>in</strong>g an advanced stage <strong>of</strong> project implementation (construction<br />
activity) ra<strong>the</strong>r than early <strong>in</strong> <strong>the</strong> plann<strong>in</strong>g process. Obviously alternatives for mitigation<br />
<strong>of</strong> impact must be considered <strong>in</strong> <strong>the</strong> light <strong>of</strong> all <strong>the</strong>se sources <strong>of</strong> <strong>in</strong>formation and practical<br />
considerations.<br />
Mitigation by Data Recovery and Alternatives<br />
Lipe (1974), Schiffer and Gumerman (1977), and o<strong>the</strong>rs generally rank mitigative<br />
measures <strong>in</strong> terms <strong>of</strong> <strong>the</strong> goals <strong>of</strong> modern conservation archeology (see also Advisory<br />
Council on Historic Preservation 1980). We concur with this rank<strong>in</strong>g, and shall briefly<br />
here exam<strong>in</strong>e <strong>the</strong> feasibility <strong>of</strong> alternatives.<br />
Avoidance. The shorel<strong>in</strong>es at 65 and 70 feet are taken as <strong>in</strong>variable, and all 35 sites<br />
deemed eligible are ei<strong>the</strong>r below or very near <strong>the</strong>se elevations. Avoidance is here considered<br />
<strong>in</strong>feasible (but see General Recommendations below).<br />
Preservation-<strong>in</strong>-Place. Many techniques for site protection with pavement, plastic,<br />
ripraps, c<strong>of</strong>ferdams, etc., have been suggested (e.g., Garrison 1975:290). Examples <strong>of</strong>
C-18<br />
active preservation are <strong>in</strong>creas<strong>in</strong>g <strong>in</strong> all parts <strong>of</strong> <strong>the</strong> country, although long term benefits<br />
are unknown. In <strong>the</strong> case <strong>of</strong> <strong>in</strong>undated sites, mechanical impacts can be reduced and<br />
perhaps nearly elim<strong>in</strong>ated, but chemical, analytical, and contextual impacts <strong>of</strong> submersion<br />
are not well understood and cannot be controlled or elim<strong>in</strong>ated presently (Lenihan<br />
and o<strong>the</strong>rs 1977). Any decision to preserve sites <strong>in</strong> place must contend with a high cost<br />
factor. We are not prepared to recommend this alternative for even <strong>the</strong> smallest site <strong>of</strong><br />
significance, although we approve <strong>of</strong> experimentation on active preservation <strong>in</strong> o<strong>the</strong>r<br />
cases. Also, we would value discussion with eng<strong>in</strong>eers and stabilization experts on <strong>the</strong><br />
feasibility <strong>of</strong> preserv<strong>in</strong>g floodpla<strong>in</strong> sites <strong>in</strong> place <strong>in</strong> <strong>the</strong> Felsenthal Project area.<br />
Mitigation by Data Recovery. Scientific excavation <strong>in</strong>volv<strong>in</strong>g specific, appropriate,<br />
and rigorous data recovery techniques is <strong>the</strong> last alternative, and <strong>the</strong> one we can recommend<br />
<strong>in</strong> this case without reservation. We turn now to a summary <strong>of</strong> <strong>the</strong> proposed<br />
mitigation program which is designed to maximize data recovery <strong>in</strong> <strong>the</strong> “particular cultural-environmental<br />
context” <strong>of</strong> <strong>the</strong> Felsenthal Project area. Mitigation costs, presented<br />
as recommended man-day expenditures for field activity, are presented on Tables C-3,<br />
C-4, and C-5. It is recommended that <strong>the</strong>se activities be conducted over two field seasons<br />
due to limits <strong>of</strong> optimal time for fieldwork (July-November).<br />
I. Research Design Outl<strong>in</strong>e (Phase 1)<br />
The research design shall be <strong>in</strong>terdiscipl<strong>in</strong>ary <strong>in</strong> scope and shall <strong>in</strong>corporate a flexible,<br />
multistage program <strong>of</strong> data recovery. Broad-based research doma<strong>in</strong>s shall <strong>in</strong>clude,<br />
but not necessarily be limited to, significant, relevant problems identified <strong>in</strong> <strong>the</strong> 1979-<br />
1980 Felsenthal Project:<br />
Doma<strong>in</strong> 1—Floodpla<strong>in</strong> geoarcheology<br />
a. fluvial geomorphology <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong> Lowland (processes and change)<br />
b. geochronometry <strong>in</strong> alluvial sites and localities (problems and potential)<br />
Data Base: <strong>Marais</strong> Sal<strong>in</strong>e Lake, Marie Sal<strong>in</strong>e site (3AS329), Goulett Island, and<br />
o<strong>the</strong>r terrace and floodpla<strong>in</strong> localities.<br />
Doma<strong>in</strong> 2—Floodpla<strong>in</strong> environmental archeology<br />
a. function and variability <strong>in</strong> Mississippi period extractive sites (<strong>in</strong>clud<strong>in</strong>g locational<br />
analysis and comparative ethnohistorical research)<br />
b. function and variability <strong>in</strong> Woodland period extractive sites<br />
c. function and variability <strong>in</strong> extractive/habitation sites <strong>of</strong> <strong>the</strong> Archaic period<br />
d. overflow bottomland focal resources<br />
e. locational and catchment analysis <strong>of</strong> extractive sites<br />
Data Base: False Indigo (3AS285), Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76), and o<strong>the</strong>r sites listed<br />
<strong>in</strong> Tables C-1 and C-2.
Table C-4. Mitigation Budget Estimate (<strong>in</strong> field man days): PHASE 2: Mitigation <strong>of</strong> Six Prehistoric Floodpla<strong>in</strong> Sites.<br />
Number Number Total Volume <strong>of</strong> Average Sump Water O<strong>the</strong>r<br />
Site Crew <strong>of</strong> <strong>of</strong> Manpower Site to Excavated Backhoe Pump Screen<strong>in</strong>g Required<br />
Number Size* Supervisors Weeks Effort** be Excavated Per Day Time Needed Flotation Expertise<br />
(<strong>in</strong> days) (cu meters) (cu meters) (hours)<br />
3AS329 18 2 16 1600 806 .504 40 + + geomorphology,<br />
chronometrics<br />
3AS285 7 1 3 120 54 .450 + zooarcheology<br />
3AS286 7 1 14 560 278 .496 + ethnohistory<br />
3BR76 7 1 4 160 75 .469 + —<br />
3AS306 7 1 6 240 129 .538 + —<br />
3AS307 7 1 2 80 37 .463 + —<br />
— 1379 .487<br />
2760<br />
man days<br />
* a total <strong>of</strong> three crews (one <strong>of</strong> 18 members and two <strong>of</strong> seven members) could conduct this work <strong>in</strong> a s<strong>in</strong>gle four month field period. This model assumes<br />
an average excavation effort <strong>of</strong> ca .5 sq m per person per day<br />
** does not <strong>in</strong>clude Project Archeologist<br />
C-19
C-20<br />
Table C-5. Mitigation Budget Estimate (<strong>in</strong> field man days): PHASE 3: Two-Stage Mitigation, 23 Prehistoric Floodpla<strong>in</strong><br />
Sites.<br />
∑<br />
Crew Number <strong>of</strong> Number <strong>of</strong> Number <strong>of</strong> Number <strong>of</strong> Surface Area 5% Average Total Volume Average<br />
Size Supervisors Crews Weeks Man Days 23 Sites Sample Depth to be Excavated cu m/man day<br />
Stage One 3 1 5 6 600 8,693 m2 435 .5 2<strong>17</strong> m3 .36*<br />
Excavate 5%<br />
sample <strong>of</strong> 23<br />
sites<br />
Stage Two 7 1 6 240 Total manpower effort 130 m3 .54<br />
Excavate ex- per site per site**<br />
tensively at 1440<br />
least six sites<br />
* greater per day manpower cost due to boat travel and logistics associated with <strong>the</strong> dispersed nature <strong>of</strong> <strong>the</strong> sites<br />
** Site 3AS306 is considered representative <strong>of</strong> site size and mitigation effort required
Table C-6. Mitigation Budget Estimate (<strong>in</strong> field man days): PHASE 3: Two Stage Mitigation—Six Historic Sites.<br />
Surface Metal Controlled Subsurface<br />
Site Historic Crew Number <strong>of</strong> Manpower Effort Survey, Detector Surface Test<strong>in</strong>g Backhoe Underwater<br />
Number Archeologist Size Weeks <strong>in</strong> Man Days Site Type Mapp<strong>in</strong>g Survey Collection (hand) Test<strong>in</strong>g Study<br />
Stage One<br />
3AS299 1 3 2 40 canal/land<strong>in</strong>g x x x x x<br />
3BR8 1 3 1 20 dump x x x<br />
3BR55 1 3 1 20 sawmill x x x<br />
3UN122 1 3 2 40 land<strong>in</strong>g x x x<br />
3BR56 1 3 1 10 land<strong>in</strong>g x x x<br />
3UN153 1 3 1 30 wreck x<br />
1 3 7 140<br />
Stage Two: additional archival work and fur<strong>the</strong>r field <strong>in</strong>vestigations may be required on Stage Two mitigation; <strong>the</strong>se work requirements cannot be<br />
estimated until after Stage One results are evaluated.<br />
C-21
C-22<br />
Doma<strong>in</strong> 3—Floodpla<strong>in</strong> behavioral archeology<br />
a. <strong>in</strong>terassemblage variability among floodpla<strong>in</strong> sites<br />
b. material correlates <strong>of</strong> low bulk and high bulk procurement sites<br />
c. material correlates <strong>of</strong> transient site use, reuse, and anticipated return (discard and<br />
curate behavior)<br />
d. activity sets and activity areas<br />
Data Base: Especially those sites listed <strong>in</strong> Table C-1.<br />
Doma<strong>in</strong> 4—Ethnohistorical analysis<br />
a. composition and identity <strong>of</strong> social group occupy<strong>in</strong>g an extractive site (tested<br />
aga<strong>in</strong>st a Quapaw model)<br />
b. seasonality and function <strong>of</strong> an extractive site (tested aga<strong>in</strong>st a Quapaw model)<br />
Data Base: One Cypress Po<strong>in</strong>t (3AS286).<br />
Doma<strong>in</strong> 5—Historic site analysis<br />
a. structure and function <strong>of</strong> a n<strong>in</strong>eteenth century riverboat land<strong>in</strong>g<br />
b. structure and function <strong>of</strong> lowland sawmills, logg<strong>in</strong>g facilities, and logg<strong>in</strong>g techniques<br />
(prior to 1930)<br />
c. depositional and postdepositional history <strong>of</strong> a sunken steamboat<br />
d. functional and socioeconomic analysis <strong>of</strong> riverbank, riverbottom, and canal bottom<br />
refuse<br />
Data Base: Marie Sal<strong>in</strong>e Land<strong>in</strong>g (3AS329), Keelboat Brake Sawmill (3BR55), Goulett<br />
Island (3BR8), Lotawanna (3UN153) and o<strong>the</strong>r historic sites <strong>in</strong> Table<br />
C-2.<br />
Doma<strong>in</strong> 6—Floodpla<strong>in</strong> site transformations<br />
a. mechanical, chemical, and biological effects <strong>of</strong> <strong>in</strong>undation<br />
b. small site visibility and methodology<br />
Data Base: All sites <strong>in</strong> Tables C-1 and C-2.<br />
II. Extensive and Intensive Investigation <strong>of</strong> Six Floodpla<strong>in</strong> Sites (Phase 2)<br />
Six <strong>of</strong> eight sites tested extensively <strong>in</strong> 1979 and reported <strong>in</strong> Chapter 6 are known<br />
to have relatively great research potential. Data relevant to a variety <strong>of</strong> research problems<br />
cited <strong>in</strong> this research design outl<strong>in</strong>e are adduced for <strong>the</strong>se sites, as well as a variety<br />
<strong>of</strong> adverse impacts (Table C-1). A brief statement <strong>of</strong> proposed research scope and field<br />
methodology is presented below for each site, but <strong>the</strong> reader must refer to Chapter 6 for<br />
rationale.<br />
Marie Sal<strong>in</strong>e (3AS329)<br />
1. Stratigraphic excavation <strong>of</strong> two deep blocks, each about 15 x 15 m <strong>in</strong> size, at upstream<br />
and downstream ends <strong>of</strong> Area C, emphasiz<strong>in</strong>g cultural levels from about<br />
0.5 m depth to below river stage or sterile soil; extensive use <strong>of</strong> water screen<strong>in</strong>g <strong>in</strong><br />
all levels and pump<strong>in</strong>g <strong>of</strong> deep levels (below about 1.5 m).
2. Precisely controlled excavation <strong>of</strong> deep occupation floors, if identified<br />
3. Specially designed chronometric sampl<strong>in</strong>g and dat<strong>in</strong>g <strong>of</strong> deep levels<br />
4. Backhoe trench<strong>in</strong>g primarily for geological stratigraphy and soil sediment sampl<strong>in</strong>g<br />
keyed to pr<strong>of</strong>iles<br />
5. M<strong>in</strong>imal <strong>of</strong>fsite test<strong>in</strong>g or test<strong>in</strong>g <strong>in</strong> Areas A and B<br />
6. Geomorphological study <strong>of</strong> site environs, <strong>in</strong>clud<strong>in</strong>g <strong>Marais</strong> Sal<strong>in</strong>e Lake<br />
False Indigo (3AS285)<br />
1. Total excavation <strong>of</strong> a 12 x 15 m midden <strong>in</strong> Area A to about 0.3 m depth; water<br />
screen<strong>in</strong>g and/or flotation <strong>of</strong> all midden levels<br />
2. Precisely controlled excavation <strong>of</strong> features or traces <strong>of</strong> structures, if identified<br />
3. Chronometric sampl<strong>in</strong>g and dat<strong>in</strong>g <strong>of</strong> midden<br />
4. M<strong>in</strong>imal <strong>of</strong>fsite test<strong>in</strong>g or test<strong>in</strong>g <strong>in</strong> Area B<br />
One Cypress Po<strong>in</strong>t (3AS286)<br />
1. Precisely controlled horizontal excavation <strong>of</strong> a 16 x 58 m area (which may circumscribe<br />
two activity or residential loci) to sterile soil at 0.3 m depth<br />
2. Precisely controlled excavation <strong>of</strong> features, if identified<br />
3. Water screen<strong>in</strong>g and/or flotation on a sampl<strong>in</strong>g basis<br />
4. Dat<strong>in</strong>g, if suitable material recovered<br />
5. Comparative ethnohistorical research<br />
Jug Po<strong>in</strong>t Cut<strong>of</strong>f (3BR76)<br />
1. Total excavation <strong>of</strong> an 11 x <strong>17</strong> m occupation area and small midden to about 0.4 m<br />
depth; water screen<strong>in</strong>g and/or flotation <strong>of</strong> all midden levels<br />
2. Precisely controlled excavation <strong>of</strong> features or traces <strong>of</strong> structures, if identified<br />
3. Chronometric sampl<strong>in</strong>g and dat<strong>in</strong>g <strong>of</strong> midden<br />
4. M<strong>in</strong>imal <strong>of</strong>fsite test<strong>in</strong>g<br />
Jug Po<strong>in</strong>t 1 (3AS306)<br />
1. Precisely controlled horizontal excavation <strong>of</strong> two discrete activity or residential<br />
loci, Area A (8 x 20 m) and Area B (9 x 30 m), to 0.3 m depth<br />
2. Precisely controlled excavation <strong>of</strong> features, if identified<br />
3. Water screen<strong>in</strong>g and/or flotation on a sampl<strong>in</strong>g basis<br />
4. M<strong>in</strong>imal <strong>of</strong>fsite test<strong>in</strong>g<br />
5. Dat<strong>in</strong>g, if suitable material recovered<br />
C-23<br />
Jug Po<strong>in</strong>t 2 (3AS307)<br />
1. Precisely controlled horizontal excavation <strong>of</strong> three discrete activity loci, Area A (3 x<br />
8 m), Area B (2 x 5 m), and Area C (6 x 15 m), to about 0.3 m depth<br />
2. Precisely controlled excavation <strong>of</strong> features, if identified<br />
3. Water screen<strong>in</strong>g and/or flotation on a sampl<strong>in</strong>g basis<br />
4. M<strong>in</strong>imal <strong>of</strong>fsite test<strong>in</strong>g<br />
5. Dat<strong>in</strong>g, if suitable material recovered
C-24<br />
III. Two-stage mitigation <strong>of</strong> 29 floodpla<strong>in</strong> sites (Phase 3)<br />
This part <strong>of</strong> our proposal represents a much more flexible approach to mitigat<strong>in</strong>g<br />
impacts on certa<strong>in</strong> floodpla<strong>in</strong> sites, and also represents we believe, an efficient means <strong>of</strong><br />
atta<strong>in</strong><strong>in</strong>g stated research goals. Schiffer and Gumerman (1977:325), among o<strong>the</strong>rs, suggest<br />
that <strong>the</strong> two-stage mitigation strategy is well suited to programs <strong>of</strong> <strong>the</strong> scope we are<br />
propos<strong>in</strong>g. In this strategy exploratory test<strong>in</strong>g, and additional survey if needed, precede<br />
selection <strong>of</strong> sites for <strong>in</strong>tensive excavation and data recovery. Two-stage mitigation<br />
assumes that among a large group <strong>of</strong> significant sites, test<strong>in</strong>g will identify those with<br />
greatest research potential and probably also those with maximum predicted impact.<br />
Sites can be stratified on <strong>the</strong>se (or o<strong>the</strong>r) bases and one or more optimal sites with<strong>in</strong> each<br />
stratum <strong>in</strong>tensively mitigated. As <strong>in</strong> any multistage field program, <strong>the</strong> best choices are<br />
based on maximal, current site data and assessment.<br />
In <strong>the</strong> case <strong>of</strong> Felsenthal Project floodpla<strong>in</strong> sites, we have identified 29 that are well<br />
<strong>in</strong>vestigated on <strong>the</strong> basis <strong>of</strong> site survey methodology (Chapter 5), and that are recommended<br />
for nom<strong>in</strong>ation to <strong>the</strong> National Register (Table C-2). We know, for example,<br />
much about <strong>the</strong> location, elevation, extent, depth, content, <strong>in</strong>tegrity, and <strong>of</strong>ten <strong>the</strong> cultural<br />
period or phase for each (Appendix A). However, we also expect that research<br />
potential varies among <strong>the</strong>se 29 sites, both prehistoric and historic, and that redundancy<br />
<strong>in</strong> research potential may be encompassed by this group. S<strong>in</strong>ce all prehistoric sites, and<br />
all or parts <strong>of</strong> historic sites are buried, and s<strong>in</strong>ce subsurface exploration is limited to<br />
shovel or auger holes and exam<strong>in</strong>ation <strong>of</strong> riverbank exposures, we cannot know <strong>the</strong> full<br />
research potential <strong>in</strong> all cases.<br />
We <strong>the</strong>refore propose to stratify or group <strong>the</strong>se sites by cultural period (Table C-2),<br />
and to conduct limited test<strong>in</strong>g with<strong>in</strong> each group on <strong>the</strong> most flexible and efficient basis<br />
to identify those sites with greatest research potential. Intensive excavation or o<strong>the</strong>r appropriate<br />
mitigation measures for selected sites can <strong>the</strong>n proceed as rapidly as possible.<br />
While recommend<strong>in</strong>g flexibility and responsibility on <strong>the</strong> part <strong>of</strong> <strong>the</strong> project archeologist<br />
and his field staff, we also propose <strong>the</strong> follow<strong>in</strong>g guidel<strong>in</strong>es for plann<strong>in</strong>g and cost estimation<br />
purposes:<br />
1. With<strong>in</strong> each designated group <strong>of</strong> prehistoric floodpla<strong>in</strong> sites, one or more sites<br />
(and perhaps all) shall be tested by appropriate subsurface techniques, not to<br />
exceed 5% <strong>of</strong> mapped site area (we believe limited test<strong>in</strong>g <strong>of</strong> this extent was<br />
demonstrably effective dur<strong>in</strong>g our 1979 site test<strong>in</strong>g stage).<br />
2. With<strong>in</strong> each designated group <strong>of</strong> prehistoric floodpla<strong>in</strong> sites at least one site<br />
(and ord<strong>in</strong>arily no more) shall be extensively mitigated by appropriate techniques,<br />
<strong>in</strong>clud<strong>in</strong>g stratigraphic or horizontal excavation strategies, and water<br />
screen<strong>in</strong>g and/or flotation (as <strong>in</strong> Phase 2 above).<br />
3. Appropriate test<strong>in</strong>g <strong>of</strong> historic riverboat land<strong>in</strong>gs shall <strong>in</strong>clude <strong>in</strong> <strong>the</strong> case <strong>of</strong><br />
Marie Sal<strong>in</strong>e Land<strong>in</strong>g (3AS299), <strong>the</strong> follow<strong>in</strong>g:
C-25<br />
a. shovel test<strong>in</strong>g and metal detection <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> <strong>the</strong> 200 m long canal<br />
b. limited backhoe or dredge tests <strong>of</strong> <strong>the</strong> submerged canal-bottom deposit<br />
c. backhoe trenches <strong>in</strong> certa<strong>in</strong> accessible areas <strong>in</strong> <strong>the</strong> vic<strong>in</strong>ity <strong>of</strong> <strong>the</strong> canal, if<br />
<strong>in</strong>dicated by o<strong>the</strong>r subsurface or documentary evidence.<br />
4. Appropriate test<strong>in</strong>g at Goulett Island (3BR8) should <strong>in</strong>clude<br />
a. sampl<strong>in</strong>g <strong>the</strong> riverbank refuse deposits (>50 years)<br />
b. geomorphological study <strong>of</strong> <strong>the</strong> Goulett Island riverbank exposure (see Research<br />
Design outl<strong>in</strong>e)<br />
c. see General Recommendations below for island resources above floodpla<strong>in</strong><br />
level<br />
5. Wreck <strong>of</strong> <strong>the</strong> Lotawanna (3UN153) shall be located, <strong>in</strong>spected, and studied by a<br />
qualified diver-archeologist<br />
6. Extensive and <strong>in</strong>tensive mitigation <strong>of</strong> historic sites <strong>in</strong>clud<strong>in</strong>g additional archival<br />
research, shall be designed and implemented on <strong>the</strong> basis <strong>of</strong> Phase 3 test<strong>in</strong>g<br />
results—if necessary.<br />
7. The project archeologist shall regularly and formally consult with Contract<strong>in</strong>g<br />
Officer’s representative and Pr<strong>in</strong>cipal Investigator as two-stage mitigation<br />
proceeds<br />
IV. Analysis (Phase 4)<br />
A field laboratory will operate concurrently with Phases 2 and 3, accomplish<strong>in</strong>g all<br />
process<strong>in</strong>g <strong>of</strong> collections and samples and <strong>in</strong>itiat<strong>in</strong>g analysis. Both rout<strong>in</strong>e and specialized<br />
analyses follow<strong>in</strong>g <strong>the</strong> field phases shall employ experts <strong>in</strong> <strong>the</strong> follow<strong>in</strong>g fields:<br />
prehistoric archeology, geomorphology, ecology, historical archeology, ethnohistory,<br />
statistics, and computer science.<br />
V. Report <strong>of</strong> Results (Phase 5)<br />
A comprehensive technical report <strong>of</strong> mitigation program results, <strong>in</strong>clud<strong>in</strong>g all<br />
<strong>in</strong>terdiscipl<strong>in</strong>ary efforts shall be prepared and submitted to <strong>the</strong> sponsor <strong>in</strong> a timely manner.<br />
After a review process, <strong>the</strong>se results shall be revised or supplemented and published<br />
for <strong>the</strong> use <strong>of</strong> <strong>the</strong> scientific and resource management communites. Interpretation and<br />
dissem<strong>in</strong>ation <strong>of</strong> results for <strong>the</strong> lay community is discussed below (Recommendations<br />
for Interpretive Displays).
C-26<br />
Logistics and Schedul<strong>in</strong>g<br />
Field phases proposed above must proceed dur<strong>in</strong>g low water and relatively dry<br />
wea<strong>the</strong>r, normally late summer and fall, especially when Ouachita and Sal<strong>in</strong>e River levels<br />
rema<strong>in</strong> at present lock stage (62.2 feet elevation).<br />
Multiple field crews must operate concurrently <strong>in</strong> order to accomplish data recovery<br />
phases <strong>in</strong> two field seasons. We recommend that Phase 2 (sites <strong>of</strong> demonstrated<br />
research potential) be accomplished dur<strong>in</strong>g <strong>the</strong> first field season, and Phase 3 (two-stage<br />
mitigation) be accomplished dur<strong>in</strong>g a second field season. However, establishment <strong>of</strong><br />
<strong>the</strong> navigation pool at 65 feet elevation will prevent access to some sites, and <strong>the</strong> construction<br />
schedule must be considered here.<br />
Extensive boat travel is required to reach many, or even most, floodpla<strong>in</strong> sites here<br />
designated for fieldwork, and several types or sizes <strong>of</strong> boats will be required for personnel<br />
and equipment transport, as well as for special needs <strong>of</strong> collaborat<strong>in</strong>g scientists.<br />
Underwater exploration and study <strong>of</strong> Lotawanna (3UN153) clearly requires specialized<br />
div<strong>in</strong>g and record<strong>in</strong>g equipment. Water pumps, water screens, and flotation devices are<br />
required for each <strong>of</strong> <strong>the</strong> <strong>in</strong>tensive mitigation tasks.<br />
The use <strong>of</strong> backhoe trench<strong>in</strong>g on a limited basis <strong>in</strong> particular site contexts (Marie<br />
Sal<strong>in</strong>e, 3AS329, and Marie Sal<strong>in</strong>e Land<strong>in</strong>g, 3AS299) is anticipated.<br />
Recommendations for Interpretive Displays<br />
The Felsenthal Project Scope <strong>of</strong> Services requires that we furnish recommendations<br />
for <strong>in</strong>terpretive displays, and such recommendations are appropriate to <strong>the</strong> philosophy<br />
<strong>of</strong> mitigation shared by conservation archeologists. Schiffer and Gumerman (1977:326),<br />
for example, state that “an excavation program [can lead] not only to scientific results,<br />
but also to enhancement <strong>of</strong> public educational and recreational values.” We believe that<br />
systematic scientific <strong>in</strong>formation derived from our survey and test<strong>in</strong>g project, earlier<br />
work <strong>in</strong> or near <strong>the</strong> project area, and our proposed archeological mitigation program can<br />
lead to a beneficial public <strong>in</strong>terpretation program, equal <strong>in</strong> quality to any exist<strong>in</strong>g <strong>in</strong> <strong>the</strong><br />
state <strong>of</strong> <strong>Arkansas</strong>. In this section we propose two k<strong>in</strong>ds <strong>of</strong> <strong>in</strong>formation: <strong>the</strong>mes which are<br />
appropriate to <strong>the</strong> cultural-environmental context <strong>of</strong> our study area, and forms or <strong>in</strong>terpretive<br />
media through which <strong>the</strong>se <strong>the</strong>mes can be transmitted.<br />
However, we should acknowledge first that <strong>the</strong> floodpla<strong>in</strong> area and resources<br />
studied by us are circumscribed by <strong>the</strong> Felsenthal National Wildlife Refuge and that a<br />
master plan for wildlife and public use <strong>of</strong> this area has been designed by <strong>the</strong> U.S. Fish<br />
and Wildlife Service (1979). It seems clear that recommendations we make here should<br />
be compatible with objectives <strong>of</strong> <strong>the</strong> Wildlife Refuge, and also that <strong>in</strong>terpretive efforts
(as well as resource management efforts) should be closely coord<strong>in</strong>ated by <strong>the</strong> U.S.<br />
Army Corps <strong>of</strong> Eng<strong>in</strong>eers and <strong>the</strong> U.S. Fish and Wildlife Service.<br />
C-27<br />
The master plan referred to above (<strong>the</strong> version available to us is a general refuge<br />
map with limited, but useful, plann<strong>in</strong>g <strong>in</strong>formation) estimates that annual visitors to<br />
<strong>the</strong> Wildlife Refuge will <strong>in</strong>crease from <strong>the</strong> 1979 level <strong>of</strong> 50,000, summarizes <strong>the</strong> wildlife<br />
management program which <strong>in</strong>cludes <strong>in</strong>creased levels <strong>of</strong> hunt<strong>in</strong>g and fish<strong>in</strong>g, and lists<br />
various planned facilities for public use, <strong>in</strong>clud<strong>in</strong>g a visitor station, boat access sites,<br />
roads, foot trails and canoe trails, and support structures. The plan states briefly that<br />
“<strong>the</strong> refuge also <strong>of</strong>fers protection to some <strong>of</strong> <strong>the</strong> most significant archeological resources<br />
located <strong>in</strong> <strong>the</strong> state <strong>of</strong> <strong>Arkansas</strong>.” All pr<strong>of</strong>essional archeologists knowledgeable about <strong>the</strong><br />
Felsenthal region, with whom we have conferred, agree that numerous sites <strong>of</strong> great scientific<br />
and public significance exist with<strong>in</strong> <strong>the</strong> boundaries <strong>of</strong> Felsenthal National Wildlife<br />
Refuge. It has been our task to locate and evaluate floodpla<strong>in</strong> sites, but some <strong>of</strong> our<br />
<strong>in</strong>formation clearly perta<strong>in</strong>s to sites and environmental features throughout <strong>the</strong> refuge<br />
area.<br />
Themes. The follow<strong>in</strong>g <strong>the</strong>mes <strong>of</strong> various scope are suggested as particularly relevant<br />
to <strong>the</strong> study area and suitable for public educational and recreational presentation.<br />
Some <strong>the</strong>mes refer to Indian prehistory or ethnohistory, some to European or American<br />
presence, and some to unique ecological or geohydrological features (an unusual culturalenvironmental<br />
context has been emphasized <strong>in</strong> this report):<br />
1. Native Americans <strong>of</strong> <strong>the</strong> Ouachita Valley: Ways <strong>of</strong> Life <strong>in</strong> <strong>the</strong> <strong>Grand</strong> <strong>Marais</strong><br />
Lowland<br />
2. Hunters and Fishermen <strong>in</strong> <strong>the</strong> Ouachita River Valley before You<br />
3. Indian Earthworks, Mounds, and Mound Centers<br />
4. Indian Villages, Farmsteads, and Camps<br />
5. Prehistoric Trade Networks<br />
6. Caddo, Koroa, and Quapaw Indians <strong>in</strong> <strong>the</strong> Ouachita Valley<br />
7. DeSoto’s Expedition: First Europeans to Enter <strong>the</strong> Ouachita Valley<br />
8. French Hunters and Trappers <strong>in</strong> <strong>the</strong> <strong>17</strong>00s<br />
9. Dunbar-Hunter Exploration, 1804-1805<br />
10. Early American Settlements on Ouachita and Sal<strong>in</strong>e Rivers<br />
11. Improv<strong>in</strong>g <strong>the</strong> Ouachita River for Navigation<br />
12. Steamboat<strong>in</strong>g on <strong>the</strong> Ouachita and Sal<strong>in</strong>e Rivers<br />
13. Riverboat Land<strong>in</strong>gs and Commerce <strong>in</strong> <strong>the</strong> 1800s<br />
14. Growth <strong>of</strong> <strong>the</strong> Lumber<strong>in</strong>g Industry and Company Towns<br />
15. What is a Bottomland Hardwood Forest? Hydric Tree and Shrub Associations<br />
16. Energy Flow and Nutrient Cycl<strong>in</strong>g <strong>in</strong> <strong>the</strong> Ouachita River Floodpla<strong>in</strong><br />
<strong>17</strong>. Floodpla<strong>in</strong> Forest Biomass: It’s Enormous<br />
18. Wildlife <strong>in</strong> Bayous, Brakes, Backswamps, and River Channels<br />
19. Wildlife <strong>in</strong> Oak-P<strong>in</strong>e Forests and Lowland Prairies<br />
20. Preserv<strong>in</strong>g Endangered Species: Alligator, Eagle, and Red-cockaded Woodpecker
C-28<br />
21. Spanish Moss <strong>in</strong> <strong>Arkansas</strong><br />
22. Native Fishes and Molluscs <strong>of</strong> <strong>the</strong> Ouachita River System<br />
23. Poisonous Snakes <strong>of</strong> <strong>the</strong> Ouachita Valley<br />
24. What is an Oxbow Lake? Meander<strong>in</strong>g River Processes<br />
25. Wetlands <strong>in</strong> <strong>Arkansas</strong>: The Ecological Imperative<br />
Forms. The means <strong>of</strong> <strong>in</strong>terpretation best suited, <strong>in</strong> our op<strong>in</strong>ion, to present<strong>in</strong>g local<br />
cultural and environmental <strong>the</strong>mes are noted below, with examples <strong>of</strong> specific resources<br />
or optimal locations for <strong>in</strong>terpretive displays:<br />
1. Visitor Center. Relatively detailed <strong>the</strong>mes and related or sequential <strong>the</strong>mes can<br />
best be presented <strong>in</strong> <strong>the</strong> specially designed exhibit area <strong>of</strong> a visitor center. Many excellent<br />
visitor centers <strong>in</strong> this country were built near, and designed to harmonize with,<br />
major Indian mound sites (for example at Toltec Mounds State Park, <strong>Arkansas</strong>). It is essential<br />
that <strong>in</strong>tensive archeological survey be a part <strong>of</strong> site plann<strong>in</strong>g so that visitor center<br />
build<strong>in</strong>gs, roads, park<strong>in</strong>g lots, and trails do not destroy portions <strong>of</strong> <strong>the</strong> cultural resource<br />
<strong>the</strong>y are <strong>in</strong>terpret<strong>in</strong>g and protect<strong>in</strong>g. The most accessible, and also <strong>the</strong> most thoroughly<br />
studied, major mound site <strong>in</strong> <strong>the</strong> refuge is Shallow Lake (3UN9/52), but two less accessible<br />
mound groups are far more dramatic <strong>in</strong> extent and locale (Watts Field, 3UN18, and<br />
Eagle Lake Mounds, 3BR4). State archeologists are deeply concerned with protect<strong>in</strong>g <strong>the</strong><br />
<strong>in</strong>tegrity <strong>of</strong> all three sites mentioned above.<br />
There are no monumentally constructed or dramatic prehistoric sites <strong>in</strong> <strong>the</strong> floodpla<strong>in</strong><br />
which could serve as <strong>the</strong> focus <strong>of</strong> a visitor center, and we doubt <strong>the</strong> feasibility <strong>of</strong><br />
develop<strong>in</strong>g a lowly<strong>in</strong>g site because <strong>of</strong> annual <strong>in</strong>undation. (Big Mound Ridge, 3AS6, is<br />
outside <strong>the</strong> refuge, but is an important lowly<strong>in</strong>g mound site.)<br />
2. Stand<strong>in</strong>g Historic Structures. The few historic structures known to us <strong>in</strong>clude<br />
twentieth century mounds at 3BR32 and 3UN92. The former is related to early lock and<br />
dam construction (Theme 11 above), but <strong>the</strong> latter was not found to be historically or<br />
regionally sigificant (Chapter 5). These sites do not appear to merit development and<br />
<strong>in</strong>terpretation.<br />
Goulett Island (3BR8) <strong>in</strong>cludes one or more stand<strong>in</strong>g structures related to n<strong>in</strong>eteenth<br />
and early twenteeth century farm<strong>in</strong>g and logg<strong>in</strong>g activity (we briefly describe a<br />
primitive log barn <strong>in</strong> Chapter 5). Unfortunately, nei<strong>the</strong>r historic or prehistoric resources<br />
<strong>of</strong> Goulett Island have been systematically surveyed or recorded <strong>in</strong> modern times—we<br />
have emphasized <strong>the</strong> potential significance <strong>of</strong> Goulett Island at several po<strong>in</strong>ts <strong>in</strong> <strong>the</strong><br />
report (see Chapters 2, 5, and General Recommendations below). This locality may have<br />
both cultural and natural values which should be protected and <strong>in</strong>terpreted, but it is <strong>in</strong> a<br />
remote area <strong>of</strong> <strong>the</strong> Refuge.<br />
3. In-Place Site Exhibits. Such exhibits normally <strong>in</strong>clude prehistoric or historic<br />
archeological rema<strong>in</strong>s, exposed <strong>in</strong> <strong>the</strong> ground, stabilized, and <strong>in</strong>terpreted by display
panels or markers. This type <strong>of</strong> exhibit could be developed at major mound sites and at<br />
numerous o<strong>the</strong>r sites <strong>in</strong> <strong>the</strong> refuge (e.g., Jones Lake, 3UN81, and Locust Ridge, 3UN8)<br />
which are located near boat access facilities or trails. Such development requires a carefully<br />
designed program <strong>of</strong> excavation, and also an <strong>in</strong>formed decision to excavate ra<strong>the</strong>r<br />
than preserve <strong>the</strong> site <strong>in</strong>tact (i.e., public significance is thought to be great). Site exhibits<br />
also present major security problems, unless cont<strong>in</strong>uously guarded. This form <strong>of</strong> <strong>in</strong>terpretation<br />
does not seem to be feasible for floodpla<strong>in</strong> sites that will be overflowed annually.<br />
C-29<br />
4. Outdoor Exhibit Panels and Kiosks. S<strong>in</strong>gle or multiple display panels at important<br />
visitor locations or at special po<strong>in</strong>ts <strong>of</strong> <strong>in</strong>terest are currently a major form <strong>of</strong><br />
<strong>in</strong>terpretation <strong>in</strong> outdoor recreation areas. Many <strong>of</strong> <strong>the</strong> <strong>the</strong>mes suggested above are<br />
adapted to panel display. Such displays must be vandal-pro<strong>of</strong>, and must not <strong>in</strong>clude real<br />
prehistoric or historic artifacts, s<strong>in</strong>ce <strong>the</strong>y will not be attended. Although we believe that<br />
such outdoor exhibits are well adapted to <strong>the</strong> future <strong>in</strong>terpretive program <strong>in</strong> <strong>the</strong> refuge,<br />
we strongly recommend that <strong>the</strong>y do not provide specific archeological site location<br />
<strong>in</strong>formation. We presume that certa<strong>in</strong> environmental features require similar protection.<br />
Aga<strong>in</strong>, this form <strong>of</strong> <strong>in</strong>terpretation is not feasible for overflow bottomland locations.<br />
5. Canoe and Foot Trails. Evidently some marked trails <strong>of</strong> <strong>the</strong>se k<strong>in</strong>ds are planned<br />
for <strong>the</strong> refuge, and will <strong>in</strong>tersect po<strong>in</strong>ts <strong>of</strong> <strong>in</strong>terest <strong>in</strong> upland and floodpla<strong>in</strong> areas. Some<br />
<strong>of</strong> <strong>the</strong> <strong>the</strong>mes enumerated above could be <strong>in</strong>dicated, but not explicated, by markers, e.g.,<br />
“The explorers Wm. Dunbar and George Hunter camped near here on November 16,<br />
1804.” The general presence and nature <strong>of</strong> prehistoric floodpla<strong>in</strong> sites (but not specific<br />
locations), which we have <strong>in</strong>vestigated <strong>in</strong> this report, could be <strong>in</strong>dicated by markers<br />
along riverbanks or bayous, e.g., “Numerous prehistoric camps <strong>of</strong> hunters and fishermen<br />
have been located by archeologists along <strong>the</strong> lower Sal<strong>in</strong>e River.” Such markers<br />
need to be vandal-pro<strong>of</strong> and resistant to <strong>the</strong> effects <strong>of</strong> annual flood<strong>in</strong>g.<br />
6. Popular Publications, Brochures, or Leaflets. All <strong>of</strong> <strong>the</strong> <strong>the</strong>mes suggested by us<br />
can be made available to <strong>the</strong> public <strong>in</strong> this form, but without <strong>the</strong> impact <strong>of</strong> exhibits discussed<br />
above. We recommend that <strong>the</strong> array <strong>of</strong> important scientific <strong>in</strong>formation from our<br />
project and o<strong>the</strong>r work be made availabe <strong>in</strong> popular published form, cover<strong>in</strong>g unique<br />
cultural and ecological characteristics <strong>of</strong> <strong>the</strong> lower Ouachita Valley <strong>in</strong> <strong>Arkansas</strong>.<br />
generAl recoMMendAtions<br />
The follow<strong>in</strong>g recommendations are not specific to <strong>the</strong> Felsenthal Project area, and<br />
are not properly a part <strong>of</strong> our program <strong>of</strong> mitigation as required by contract with <strong>the</strong><br />
U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers.
C-30<br />
1. We recommend that a comprehensive management plan for cultural resources<br />
<strong>in</strong> <strong>the</strong> Felsenthal National Wildlife Refuge be prepared by <strong>the</strong> U.S. Fish and Wildlife<br />
Service with <strong>the</strong> assistance and coord<strong>in</strong>ation <strong>of</strong> <strong>the</strong> U.S. Army Corps <strong>of</strong> Eng<strong>in</strong>eers, Vicksburg<br />
District. Such a management plan would require more extensive site survey and<br />
site assessment <strong>in</strong> certa<strong>in</strong> areas than has now been accomplished. These areas should<br />
<strong>in</strong>clude Goulett Island, Prairie Island, Horton Island, P<strong>in</strong>e Island, Cedar Ridge, and adjacent<br />
floodpla<strong>in</strong> areas at about 70 feet elevation <strong>in</strong> <strong>the</strong> nor<strong>the</strong>rn portion <strong>of</strong> <strong>the</strong> refuge. The<br />
management plan should also summarize and appraise <strong>the</strong> results <strong>of</strong> <strong>the</strong> several large<br />
scale and small scale survey and test<strong>in</strong>g projects which have been performed with<strong>in</strong> refuge<br />
boundaries to date. We suggest that <strong>the</strong> Multiple Resource Area concept, which was<br />
proposed for significant floodpla<strong>in</strong> sites <strong>in</strong> this report, can be expanded to o<strong>the</strong>r terra<strong>in</strong><br />
or all terra<strong>in</strong> with<strong>in</strong> <strong>the</strong> refuge. Comprehensive <strong>in</strong>ventory and management <strong>of</strong> cultural<br />
resources <strong>in</strong> <strong>the</strong> Felsenthal National Wildlife Refuge would protect one <strong>of</strong> <strong>the</strong> richest and<br />
least disturbed archeological zones <strong>in</strong> <strong>the</strong> United States.<br />
2. We recommend that federal agencies responsible for terra<strong>in</strong> and waterways<br />
<strong>in</strong> <strong>the</strong> lower Ouachita River Valley <strong>in</strong> <strong>Arkansas</strong> consider <strong>the</strong> acquisition <strong>of</strong> significant<br />
cultural properties just outside refuge boundaries, and privately owned. Among <strong>the</strong>se<br />
are Big Mound Ridge (3AS6) and Sulphur Spr<strong>in</strong>gs Mound (3AS1) which are about 3.5<br />
and 8 km respectively east <strong>of</strong> <strong>the</strong> new lock and dam (Appendix A). These major mound<br />
sites are with<strong>in</strong> timber lands and may be threatened by total destruction <strong>in</strong> <strong>the</strong> next few<br />
years.
eferences cited<br />
C-31<br />
Advisory Council on Historic Preservation<br />
1980 Treatment <strong>of</strong> Archeological Properties: A Handbook. Advisory Council on Historic<br />
Preservation, Wash<strong>in</strong>gton D.C.<br />
Carrell, Toni, Sandra Rayl, and Daniel Lenihan<br />
1976 The Effects <strong>of</strong> Freshwater Inundation <strong>of</strong> Archeological Sites Through Reservoir Construction:<br />
A Literature Search. National Park Service, U.S. Department <strong>of</strong> <strong>the</strong> Interior.<br />
Clarke, David L.<br />
1968 Analytical Archaeology. Methuen, London.<br />
Garrison, Ervan G.<br />
1975 A Qualitative Model for Inundation Studies <strong>in</strong> Archeological Research and Resource<br />
Conservation: An Example from <strong>Arkansas</strong>. Pla<strong>in</strong>s Anthropologist 20:279-296.<br />
1977 Model<strong>in</strong>g Inundation Effects for Plann<strong>in</strong>g and Prediction. In Conservation<br />
Archaeology: A Guide for Cultural Resource Management Studies, edited by Michael B.<br />
Schiffer and George J. Gumerman, pp. 151-156. Academic Press, New York.<br />
Garrison, Ervan G., Alan May, Jeffrey Newsom, and Alf Sjoberg<br />
1977 Progress Report on <strong>the</strong> Effects <strong>of</strong> Inundation on Cultural Resources: Table<br />
Rock Reservoir, Missouri. Ms. on file, Department <strong>of</strong> Anthropology, <strong>University</strong> <strong>of</strong><br />
Missouri, Columbia.<br />
Lenihan, Daniel J., Toni L. Carrell, Thomas S. Hopk<strong>in</strong>s, A. Wayne Prokopetz, Sandra L.<br />
Rayl, and Cathryn S. Tarasovic<br />
1977 The Prelim<strong>in</strong>ary Report <strong>of</strong> <strong>the</strong> National Reservoir Inundation Study. National Park<br />
Service, U.S. Department <strong>of</strong> <strong>the</strong> Interior.<br />
Lipe, W. D.<br />
1974 A Conservation Model for American Archaeology. Kiva 39(4):213-245.<br />
National Park Service<br />
1977 How to Complete National Register Multiple Resource Nom<strong>in</strong>ation Forms:<br />
Interim Guidel<strong>in</strong>es. National Park Service, U.S. Department <strong>of</strong> <strong>the</strong> Interior.<br />
Schiffer, Michael B., and George J. Gumerman (editors)<br />
1977 Conservation Archaeology: A Guide for Cultural Resource Management Studies.<br />
Academic Press, New York.<br />
Schiffer, Michael B., and John H. House<br />
1977 An Approach to Assess<strong>in</strong>g Scientific Significance. In Conservation Archaeology:<br />
A Guide for Cultural Resource Management Studies, edited by Michael B. Schiffer and<br />
George J. Gumerman, pp. 249-257. Academic Press, New York.
C-32<br />
Schnell, Frank, and Jack Tyler<br />
1976 Hydrology and Archaeological Site Conservation. Sou<strong>the</strong>ast Archaeological Conference<br />
Bullet<strong>in</strong> 19:37-38.