Modeling
Prehistori--(セ M
Agricultu セ
Productivity in
Southwes ern
Colorado:
AGIS
Approac
Washington State University Department of Anthropology
REPORTS OF INVESTIGATlONS 67
published jointly by
WASHINGTON STATE UNIVERSITY DEPARTMENT OF ANTHROPOLOGY, Pullman, Washington
and CROW CANYON ARCHAEOLOGICAL CENTER, Cortez, Colorado
Complete bibliographic format for Repons of Investigations 67:
Van West. CarlaR.
1994 Modeling Prehistoric Agricultural Productivity in Southwestern Colorado: A GIS
Approach. Repons of Investigations 67. Department of Anthropology, Washington State
University, Pullman, and Crow Canyon Archaeological Center, Cortez, Colorado.
© 1994 Carla R. Van West
ISSN 10594000
This repon is printed on acidfree paper.
Reports of Investigations is an occasional series published by the Department of Anthropology,
Washington State University, to disseminate information resulting from research undertaken by
students, staff, and faculty associated with the department Timothy A. Kohler, General Editor.
This volume is published jointly with
Crow Canyon Archaeological center, an independent. notforprofit
organization devoted to archaeological research and education.
ISBN 0962464066
-ii-
Page Blank in Original
To my Father
Joseph B. Van West
19221987
Page Blank in Original
T ABLE OF CONTENTS
DEDICATION
iii
v
TABLE OF CON1ENTS
vii
LIST OF TABLES
LIST OF ll.LUSTR.ATIONS
.
xi
PREFACE
.
xiii
xv
ACKNOWLEDGMENTS
1.
IN1RODUCTION
1
Preview of the Model
Organization of the Study
Spatial Context
Temporal Context
Research Context
2
1
5
5
7
13
BUll.DING THE MODEL: METHODS AND DATA ACQUISmON
High Resolution Soils Data
Modem Crop Yield Data
Digital Elevation Data
Modeling Soil Moisture
TreeRing Data
GIS Technology
Archaeological Data
3.
BUll.DING THE MODEL: PRELIMINARY DATA ANALYSES
Palmer Drought Severity Index Reconstruction
Geographic Information System Processing
Linking Soil Moisture Conditions and Levels of Agricultural Productivity
4.
BUll.DING THE MODEL: FINAL DATA ANALYSES
Final Geographic Information System Analyses
Supplemental Analyses with SAS
The Complete Model
5.
RESULTS
6.
SUMMARY AND EVALUATION
The Study Area
Localities within the Study Area: The Block Survey Areas
Site Catchments within the Study Area: TreeRing Dated Sites
17
17
40
41
41
43
49
52
55
55
94
97
. 113
113
120
129
131
132
136
147
, 187
Sununary
Strengths of the Model
Limitations of the Model
Reconunendations for Future Research
-v-
187
189
190
191
BIBUOGRAPHY
195
APPENDICES
A.
Eigenvector Amplitudes Generated by a Principal Components Analysis on
Seven Expanded TreeRing Chronologies (SWOLD7)
215
Historic Commentary on Climate and Agriculture in Montezuma County,
A.D. 18941970
237
C.
Calculating Arumal Population Size: An Example for A.D. 902
241
D.
SAS Programs for Calculating Total Annual Maize Productivity
(TOTPROD) and Population Density (pOPKM)
243
Total Annual Maize Productivity and Maximum Potential Population for Three
Levels of Storage in the Study Area, A.D. 9011300
246
Total Annual Maize Productivity and Maximum Potential Population for Three
Levels of Storage in the Sand Canyon Survey Locality, A.D. 9011300
247
Total Annual Maize Productivity and Maximum Potential Population for Three
Levels of Storage in the Mockingbird Mesa Survey Locality, A.D. 9011300
248
Total Annual Maize Productivity and Maximum Potential Population for Three
Levels of Storage in the Catchment of 5MT8371, DCA Site. A.D. 9011300
249
Total Annual Maize Productivity and Maximum Potential Population for Three
Levels of Storage in the catchment of 5MT8839. Norton House,A.D. 9011300
250
Total Annual Maize Productivity and Maximum Potential Population for
Three Levels of Storage in the catchment of 5MT2433,
Aulston Pueblo. A.D. 9011300
251
Total Annual Maize Productivity and Maximum Potential Population for
Three Levels of Storage in the catchment of 5MT3834.
Mustoe Ruin, A.D. 9011300
252
Total Arumal Maize Productivity and Maximum Potential Population for
Three Levels of Storage in the catchment of 5MT6970,
Wallace Ruin. A.D. 9011300
253
Total Annual Maize Productivity and Maximum Potential Population for
Three Levels of Storage in the Catchment of 5MT1566.
I..oWty Ruin, A.D. 9011300
254
Total Annual Maize Productivity and Maximum Potential Population for
Three Levels of Storage in the Catchment of 5MT2149.
Escalante Ruin, A.D. 9011300
255
Total Annual Maize Productivity and Maximum Potential Population for
Three Levels of Storage in the Catchment of 5MT765.
Sand Canyon Pueblo, A.D. 9011300
256
B.
E.
F.
G.
R
L
J.
K.
L.
M.
N.
O.
vi-
LIST OF TABLES
2.1
Soil Conservation Service Soil Capability Class Definitions
2.2
Primary Soil Data for 113 Soils in Montezuma and Dolores Counties,
Colorado, Part 1
.
21
Primary Soil Data for 113 Soils in Montezuma and Dolores Counties,
Colorado, Part 2
.
25
.
40
.
44
2.3
_
.
19
2.4
Inter and IntraCoder Variability Rates in Soil Recording
2.5
Weather Stations Used to Reconstruct PDSI
2.6
2.7
WaterHolding Characteristics of Soils
_
.
45
TreeRing Chronologies Used in the Creation of SWOLD7
_
.
48
TreeRing Dated Sites
.
53
Soils Selected to' Represent the 11 Classes of Soil Moisture Based on
Water Held in the Upper and Lower Soil Layers
.
56
Results and Statistical Summary of the Initial Calibration, Verification,
and Final, FullPeriod Calibration Analyses for 55 Soil Moisture Groups
..
59
.
73
Descriptive Statistics Associated with 55 Fmal, Full Period Calibrations
and LongTelm Reconstructions
.
75
Occurrence of Extreme Conditions ($.4.00 or セKTNP
PDSI) in the Years
Used to Calculate PDSI for Combinations of Weather Station and
Moisture Class
_
.
87
Comparative Data from LongTelm Reconstructions in the Zuni Area and
Southwestern Colorado
.
91
.
92
2.8
3.1
3.2
3.3
"2.................•................•.....•.......................................................
Adjusted r
3.4
3.5
3.6
3.7
_
Values for 55 FullPeriod Calibrations
Comparison of Historic Comments with Annual PDSI Values for 34
LongTelm Reconstructions, A.D. 18901970
_
_
3.8
Historic Crop Yields and PDSI Values, A.D. 19311960
.
98
3.9
Distribution of Soils for Which Nonirrigated Bean Yield Data Are
Available by PDSI Reconstruction Class
.
99
.
101
3.10
Results of Linear, Polynomial, and Partial Regression Studies of the
Relationship Between PDSI and Crop Yield Values
-vii -
_
3.11
3.12
3.13
Table of Productivity Values for All Soil Types within the Study Area
Under Five Different Climatic Conditions
.
104
Nonirrigated Bean Data from Dolores and Montezuma Counties Used to
Generate a Mean Bean Yield Value for Years A.D. 19311960
.
107
Nonirrigated Maize Data from Dolores and Montezuma Counties Used to
Generate a Mean Maize Yield Value for Years A.D. 19311960
.
108
.
109
3.14
Eleven Nominal PDSI Classes
4.1
PDSI Class Correspondence Table
.
113
4.2
Productivity Table (pROD.TBL) Used in GIS Analysis
.
116
4.3
An Example of the Annual EPPL7 (GIS) Tabular Output: A COUNT File
of Agricultural Productivity in A.D. 902
;
.
119
4.4
_
Modem Nonirrigated Bean Yield for 46 Agricultural Soils in Montezuma
and Dolores Counties, Colorado
.
121
4.5
Summary of ModelBuilding Steps: Operationalizing the Model
_
.
127
5.1
Carrying Capacity Estimates at Three Levels of Storage for A.D. 9011300 for the 1470.36 km 2 Study Area
_
.
135
5.2
Periods of Greatest Occupational Attractiveness for the Study Area
5.3
Comparison of Population Values for the Study Area, Sand Canyon
Survey Locality, and Mockingbird Mesa Survey Locality, A.D. 9011300
.
143
Carrying Capacity Values (pOPKM) for the Study Area, Sand Canyon
Survey Locality, and Mockingbird Mesa Survey Locality, Based on
POnYR Estimates for A.D. 9011300
.
143
5.5
TreeRing Dates from Site 5MI'8371, DCA Site
.
148
5.6
Comparison of Population Values (pOPKM) for 5MT8371, DCA Site, for
A.D. 9011300 and A.D. 935950
_
.
149
Periods of Equal or Greater OCCUpational Attractiveness in the Catchment
of 5MT8371 (A.D. 935950)
.
150
5.4
5.7
.
136
5.8
TreeRing Dates from Site 5MT8839, Norton House
_
.
150
5.9
Comparison of Population Values (pOPKM) for 5MT8839, Norton
House, for A.D. 9011300 and A.D. 10291048
_
.
152
5.10
5.11
Periods of Equal or Greater Occupational Atuaetiveness in the Catchment
of 5MT8839 (A.D. 10291048)
TreeRing Dates from Site 5MT2433, Aulston Pueblo
-
viii-
153
_..........
155
5.12
Comparison of Population Values (POPKM) for 5MT2433, Aulston
Pueblo, for A.D. 9011300 and A.D. 10301050
156
Periods of Equal or Greater Occupational Attractiveness in the Catchment
of 5MT2433 (A.D. 10301050)
157
5.14
TreeRing Dates from Site 5MT3834, Mustoe Ruin
159
5.15
Comparison of Population Values (POPKM) for 5MT3834, Mustoe Ruin.
for A.D. 9011300 and A.D. 11731231
160
Periods of Greatest Occupational Attractiveness in the Catchment of
5MT3834 (A.D. 11731231)
160
TreeRing Dates from Site 5MT6970, Wallace Ruin (5.l7a) and from Site
5MT4126, Ida Jean Ruin (5.17b)
163
Comparison of Population Values (POPKM) for 5MT6970, Wallace Ruin,
for A.D. 9011300 and A. D. 10451125
164
Periods of Greatest Occupational Attractiveness in the Catclurient of
5MT6970(A.D.10451125)
165
5.20
TreeRing Dates from Site 5MT1566, Lowry Ruin
168
5.21
Comparison of Populations Values (pOPKM) for 5MT1566, Lowry Ruin,
for A.D. 9011300 and A.D. 10861120
171
Periods of Greatest Occupation Attractiveness in the Catchment of
5MT1566 (A.D. 10861120)
172
5.13
5.16
5.17
5.18
5.19
5.22
5.23
TreeRing Dates from Site 5MT2149, Escalante Ruin
5.24
Comparison of Population Values (pOPKM) for 5MT2149, Escalante
Ruin,for A.D. 9011300 and A.D. 11241138
5.25
Periods of Greatest Occupational Attractiveness in the Catchment of
5MT2149 (A.D. 11241138)
_..........
172
173
_..........
175
5.26
TreeRing Dates from Site 5MT765, Sand Canyon Pueblo
175
5.27
Suggested Construction Dates (A.D.) for TreeRing Dated Structures at
5M1'765
182
Comparison of Population Values (pOPKM) for 5MT765, Sand Canyon
Pueblo, for A.D. 9011300 and A.D. 12521274
_..........
183
Periods of Greatest Occupational Attractiveness in the Catchment of
5MT765 (A.D. 12521274)
185
Summary of Population Values (pOPKM) Using POP2YR Estimates for
Eight TreeRing Dated Sites, A.D. 9011300
186
5.28
5.29
5.30
-
ix-
Page Blank in Original
LIST OF ILLUSTRATIONS
1.1
Location of the study area
_..........
2
1.2
The conceptual model for reconstructing prehistoric agricultural
productivity
""""'"
4
2.1
Schematic diagram illustrating the arrangement of DEMs used in the
study
38
2.2
Location of the five weather stations
_..........
44
2.3
Location of the seven treering chronology stations
_..........
48
3.1
Distribution of 113 soil map units in the study area by amount of water
held in the upper and lower soil layers
_..........
57
Plot of the arumal mean PDSI values by PDSI class for a single group of
34 longterm reconstructions (REC34), A.D. 18901970
_..........
91
Plot of the annual standard deviations associated with the annual mean
PDSI value for a single group of 34 longterm reconstructions (REC34),
A.D. 18901970
_..........
94
3.2
3.3
3.4
Schematic diagram illustrating the relationship among reconstructed PDSI
values, their interpretive categories, and crop yield production categories
110
4.1
4.2
4.3
4.4
Diagram illustrating procedural steps used to process the final GIS
integrated analysis
..
114
Normal distribution of modern bean yield (lbs/ac) from 46 agricultural
soils in Montezuma and Dolores Counties, Colorado
..
120
EPPL7 dotplot image of prehistoric agricultural productivity in the study
area as reconstructed for July I, A.D. 902
.
122
Schematic diagram illustrating the relationships among three concepts of
carrying capacity
..
127
_
5.1
Maize production in the study area, A.D. 9011300
_
.
133
5.2
Supportable population density in the study area, A.D. 9011300
_
..
134
5.3
Maize production in the Sand Canyon Survey Locality, A.D. 9011300
5.4
Supportable population density in the Sand Canyon Survey Locality,
A.D. 9011300
-
xi-
138
_
.
139
5.5
5.6
5.7
5.8
5.9
5.10
5.11
5.12
5.13
5.14
5.15
5.16
Maize production in the Mockingbird Mesa Survey Locality, A.D. 9011300
141
Supportable population density in the Mockingbird Mesa Survey
Locality, A.D. 90]-1300
] 42
Comparison of estimates of carrying capacity for the Sand Canyon
Survey Locality generated by the model with estimates derived from
archaeological survey in the locality, A.D. 901-1300
145
Comparison of estimates of carrying capacity for the Mockingbird Mesa
Survey Locality generated by the model with estimates derived from
archaeological survey in the locality, A.D. 901-1300
146
Population density supportable within the 1.6-km-radius catchment of
5MT8371, DCA Site, A.D. 901-1300
151
Population density supportable within the 1.6-km-radius catchment of
5MT8839, Norton House, A.D. 901-1300
154
Population density supportable within the 1.6-km-radius catchment of
5Mr2433, Aulston Pueblo, A.D. 901-1300
158
Population density supportable within the 1.6-km-radius catchment of
5MT3834, Mustoe Ruin, A.D. 901-1300
161
Population density supportable within the 1.6-km-radius catchment of
5MT6970, Wallace Ruin, A.D. 901-1300
_..........
166
Population density supportable within the l.6-km-radius catchment of
5MT1566, Lowry Ruin, A.D. 901-1300
170
Population density supportable within the 1.6-km-radius catchment of
5MT2149, Escalante Ruin, A.D. 901-1300
174
Population density supportable within the 1.6-krn-radius catchment of
5MTI65, Sand Canyon Pueblo, A.D. 901-1300
184
-
xii-
PREFACE
At least since the times of Duff (1904), archaeologists have been making inferences
about how culture change in the prehistoric
Southwest has been affected by climate
change. Over the years the paleoenvironmental
data on which these inferences are based have
become increasingly refined, with the analysis
of pollen and macrobotanical materials from
sediment traps, archaeological sites, and packrat middens; the development of alluvial
chronologies; and with increasingly sophisticated analyses of tree rings (for a general
review see Gumerman, ed., 1988).
Because of the high temporal precision
they afford, and because of the relative ease
with which they can be calibrated to various
modem climatic parameters, tree rings have
played an increasingly important role Southwestern paleoclimatic studies. The Past Climate
of Arroyo Hondo New Mexico Reconstructed
from Tree Rings (Rose et al. 1981) demonstrated that both spring precipitation and annual precipitation could be estimated for the
prehistoric northern Rio Grande region. Rose
et al. (1982) took this analysis one step further
by using the reconstructed precipitation values
to estimate prehistoric values for soil moisture
based on the Palmer Drought Severity Index
(see also Sebastian 1992). Van West's work
brings these approaches two steps forward by
reconstructing soil moisture values (and,
through them, estimates for maize production)
for specific soil types at specific elevations. She
therefore succeeds, for the first time, in coupling the high temporal resolution afforded by
dendroclimatologically-based crop retrodictions with a high spatial resolution. Such
data in tum make possible detailed studies of
the relationship between prehistoric settlement
systems-and culture change in general-with
annual estimates of potential prehistoric production.
model enVan West's ー。ャ・ッセイ 、オcエゥッョ
compasses a 1816 kID area of southwestern
Colorado to examine the potential effects of
past climatic variation on dryland maize agriculture and sustainable population during the
late Mesa Verde Anasazi occupation of the
area (A.D. 901-13(0). The data she generates
are used to evaluate a question of long standing in the northern Southwest: whether climatic
variability was severe enough to disrupt agriculture and promote the abandonment of the
northern San Juan region toward the end of the
thirteenth century.
The model incorporates techniques and
data sets that have not been used together
before. Long regional dendroclimatic records
are used to retrodiet 1070 years (A.D. 9011970) of June Palmer Drought Severity Indices
(pDSI}-measures of stored soil moisture-for
area-specific soils. Reconstructed PDSls are
reexpressed in terms of their local equivalent in
potential maize yield. The integration, quantification, and visual display of these productivity values are coordinated through geographic information systems technology. The
method results in the production of annual
maps depicting the variable character of the
potential agricultural environment, and annual
values for total maize productivity, which can
be translated into the population size and
density that could be supported by such yields.
The results indicate that there was always
enough productive land somewhere in the
study area to support thousands of persons
(e.g., 31,363 persons or 21 persons/km2 minimum in the 1470.36 km 2 study area) over the
400-year period, even in the dry times of the
middle twelfth and late thirteenth centuries. It
would seem, therefore, that high-frequency
climatic variability was never so extreme in this
area that decreased agricultural production can
be cited as the sole or even the primary cause
of its thirteenth century depopulation.
Other implications of the data generated in
this study are still being worked through in a
series of papers and publications by Van West
and coworkers (Kohler and Van West 1994;
Van West and Kohler 1994; Van West and Lipe
1992; Van West 1994b, 1995). These papers
have generated a great deal of interest in the
original dissertation, which itself was awarded
Honorable Mention by the 1992 Dissertation
Prize Committee of the Society for American
Archaeology, who deemed it "well written, well
organized, convincingly argued, and a
methodological tourdeforce ... an excellent
dissenation." This excellent dissenation has
been thoroughly revised for this presentation,
and we are very pleased make it more readily
available to allow funher development of the
approaches so ably pioneered here.
-xiv -
Timothy A. Kohler
General Editor, Repons of Investigations
Pullman, WA
ACKNOWLEDGMENTS
Many individuals and institutions have assisted me in the completion of this study. My
dissertation committee, consisting of Timothy
Kohler, chair, and William Lipe, Jeffrey Dean,
and John Bodley, was always patient, supportive, and willing to provide advice. They encouraged my explorations, and frequently
suggested a method or strategy that helped me
cut a path through the often dark. forest of
confusion. lowe a special thanks to Jeff Dean
of the University of Arizona, who spent many
hours with me discussing data, methods, and
interpretations, and who, by means of his calm
and thoughtful demeanor, gave me the confidence to keep going when I often felt at a loss.
Funding for this project came from several
sources. Sigma Xi, The Scientific Research
Society, awarded me a grant-in-aid of research
for archaeological field work. in southwestern
Colorado in 1986. Amaterra, a non-profit educational and environmental service organization, defrayed the costs of keeping me and my
survey crew the field in 1986 by feeding and
sheltering us in their field camp at Sand Canyon Pueblo. The WelUler-Gren Foundation for
Anthropological Research awarded me a predoctoral grant (#4799) in 1987 to partially
cover the costs of data analysis. The Laboratory of Tree-Ring Research, University of
Arizona, paid the computing costs associated
with the dendroclimatic-based reconstructions
of Palmer Drought Severity Indices and generously provided support persolUlel to assist
me with processing of the data, 1987-1989.
The Department of Anthropology, Washington
State University, awarded me money to cover
computing costs while in residence in Pullman,
1989-1990.
Very special support has been provided by
several other institutions and individuals. Crow
Canyon Archaeological Center, Cortez, Colorado has continually provided logistical support for the project since its inception in 1985.
David Wherry, Director of the Digital Image
Analysis Laboratory at Washington State
University, was a constant source of information, logistic support, and friendship
throughout this project. Without his dedication,
talent, and interest, this project would not have
been as successful. Alan Price, currently the
Assistant State Soil Scientist, Soil Conservation
Service, Denver, Colorado, was my principal
liaison in the Cortez field of the SCS, and durinE his tenure there, kindly provided me with
much of the soils data and agricultural data
used in this study. Martin Rose, dendroecologist, formerly with the Laboratory of TreeRing Research, University of Arizona, Tucson,
and currently with the Desert Research Institute, University of Nevada, Reno, was my first
guide into the world of dendroclimatological
reconstruction. He suggested the strategy for
retrodicting highly localized Palmer Drought
Severity Indices from mapped soils data and
multiple sets of precipitation and temperature
records. He also generously supplied the treering data used in this study from his own as yet
unpublished dissertation research. Shela
McFarlin, Victoria Atkins, and Susan Baugh,
archaeologists with the Bureau of Land Management's Anasazi Heritage Center, Dolores,
Colorado granted me access to their archaeological site records and maps and provided me
a place to work. when I was gathering information for this project Assistance with computing
challenges was provided by Steve Samuels,
archaeologist and consultant, Center for North
American Archaeology, Washington State
University; Dan Vakoch, consultant, Computing Service Center, Washington State University; G. Robert Lofgren, Laboratory of
Tree-Ring Research, University of Arizona;
and Mark Westergaard, Marvin Landis, and
Karen Warren, Center for Computing and information Technology, University of Arizona.
Karla Harrington, Pima Community College, Tucson, was an able research assistant and
helped me greatly in the mapping of soil data
from aerial photographs. Ron Beckwith, Arizona State Museum, Tucson, skillfully drafted
the figures in this study. Susan Martin,
Statistical Research, Inc., created the cover and
the chapterstart artwork. Her interest and good
cheer made the design process enjoyable.
Over the course of the five years it took to
gather data, conduct field work, develop methods, process data, and write up this research,
many individuals have been there for me.
Without their support and often special gifts,
this study would have been more difficult and
less rewarding. While it is not possible to list
every person who helped in this way, I do want
to thank. the following friends and colleagues
for their special contributions: E. Charles
Adams, Michael Adler, Rick Ahlstrom, Christine Amiss, Edward Berger, Jo Berger, Dale
Black, W. Harold Bowling, Robin Bowyer,
Bruce Bradley, Sharon Bronson, Geneva
Burkhart, Barney Burns, Chiu-an Chang, Ruth
Chappell, Jim Colleran, Margie Connolly, Elysa
Crum, Norma Dever, Bella Eibensteiner,
Carlisle Ellis, Mary Etzkorn, Jim Finley, Steve
Gill, Raquel Goldsmith, Judy Han, Brian
Harvey, Megg Heath, Ann Hitchcock, Edgar
Huber, Wei-Chi Jao, Julie Jefferson, Marge
Kennedy, Roberta Leicester, June Lipe, Bill
Lipe, Ricky Lightfoot, Nancy Menthe, Shorlene Oliver, Troy Oliver, Roy Paul, Bill Robinson, Angela Schwab, Phil Scuderi, Karen
Seger, Ruth Slickman, Allan Smith, Catherine
Stanley, Ian Thompson, Micky Thurston, Bill
Thurston, Phillip Van West, Mark Varien,
Nancy Wall, Dave Whelchel, Jim Whitfield, Ric
Windmiller, and Angela Zerdavis.
I would also like to acknowledge those individuals, other than my comminee members,
who have been instrumental in my academic
and professional career, and who have contributed directly and indirectly to my appreciation for the art and science of doing archaeology: F. "Cal" Calabrese, Jonathan Gell,
Fekri Hassan, Emil Haury, Julian Hayden,
Alden Hayes, Alexander Lindsay, Peter Mehringer, and R. Gwinn Vivian. R. G. Matson
provided perceptive comments on the dissertation that I have attempted to address here.
I would like to express my gratitude to my
parents, Selma Van West and the late Joseph
Van West. Although my father died before this
study was completed, he saw me begin it and
grow through the experience. I have always
cherished their unconditional love, support,
and confidence in my ability to successfully
complete this work, and know they are largely
responsible for its outcome.
My final thanks must go to my husband,
Roger Itwin, who has seen me through an entire decade devoted to the pursuit of this degree. His tolerance, understanding, financial
and emotional support were unqualified and
abundant. Without him and his gifts born of
love, this work would not have been completed.
-xvi -
- Carla Van West
Statistical Research, Inc.
Tucson, AZ
1
INTRODUCTION
The Anasazi were a prehistoric farming
people who inhabited the Northern San Juan
Region of the American Southwest more or
less continualIy from the beginning or middle
of the first millenium B.C. through the thirteenth century A.D. In the late 1200s, however,
they permanently abandoned not only the
Northern San Juan but also much of their
former homeland in the Four Comers area.
From there they migrated south and east to
fonn the groups ancestral to the Hopi-, Zuni-,
Keres-, and Tanoan-speakers colIectively
known as the modem Pueblo Indians of
Arizona and New Mexico. That they left in the
late thirteenth century is well known; what is
not known with certainty, however, is why they
left, although much archaeological research
has been conducted to account for the density
and distribution of populations and their settlements, local and regional population movements, and local and regional abandonments.
It has long been suspected that such a
widespread and seemingly sudden depopulation of this large area must have had its
roots in an environmental crisis. Whether
precipitated by natural climatic fluctuations,
such as the Great Drought of A.D. 1276-1299
(Douglass 1929), or by humanly-induced
conditions, such as poor land management
practices (e.g. Wycoff 1977; Stiger 1979), the
immediate cause has assumed to be connected
to agricultural production and a failure to grow
adequate supplies of food. The problem is an
important one because in its solution and
understanding comes an appreciation for our
own relationship and responsibility to the
environment; the tolerances of our economic
adaptations to the nonnal low-frequency and
high-frequency fluctuations in climate; and the
resilience of our socioeconomic systems to
accommodate changing environmental conditions, particularly in semiarid areas.
The research described in this monograph
helps illuminate this important problem in
Anasazi prehistory and cultural process. In
attempting to reconstruct the agricultural
environment available to Anasazi farmers in
the heartland of the abandoned area, J have
created a quantitative, high-resolution model of
prehistoric agricultural productivity and sustainable population for a 1816 krn 2 (701 mi 2)
portion of the dry-farming region of southwestern Colorado (Figure 1.1). My immediate
goal was to develop methods, techniques, and
the database to construct the model. That portion of this study is a methodological contribution to Anasazi archaeology. However, the
tree-ring-based reconstructions resulting from
the model also represent a database that can be
used to address a large number of substantive
questions of current importance to Southwestern archaeologists, including those pertinent to the causes of the final thirteenthcentury abandonment. Insofar as I have begun
to explore a few of these questions of settlement attractiveness, agricultural productivity
and predictability, and potential human carrying capacity, this research also makes a
contribution to theory building and testing
with wider implications for American archaeology.
PREVIEW OF THE MODEL
These chapters summarize my attempt to
model past climatic variation and its effect on
agricultural productivity in a portion of the
Colorado Plateau that was home to the Mesa
Verde Anasazi during the last 400 years that
they inhabited that region, from A.D. 9001300. The model incorporates the highest
resolution environmental data available to
archaeologists working on the Plateau. It utilizes several techniques and datasets not used
. -' ...セRMN
.v
/
o
..,-'"
o
I /
.. ; . .
] ]セL
セMB
1=1
5 Miles
5 Kilometers
If'.
/
N
..
........... . ' . \,
I
e7
/
.....セ NZ ⦅
.. \ ... B セ
UT
CO
NM
Figure 1.1. Location of the study area. 1. DCA Site; 2. Lowry Ruin; 3. Norton House; 4. Aulstol"!
Pueblo; 5. Sand Canyon Pueblo; 6. Mustoe Ruin; 7. Wallace Ruin; 8. Escalante Ruin.
-2-
...
-,
...
..
....
together before, and uses stateoftheart spatial
data management and processing technology
to encode, store, retrieve, manipulate, and display data. Spatially, model resolution is 4 ha
(200 x 200 m); temporally, its resolution is one
year. One important product is a set of 400
arumal reconstructions of the potential agricultural niche available to prehistoric dryland
maize farmers. These reconstructions depict
the condition of the agricultural habitat as it
existed each year on July I, just before the
arrival of the summer rains and after the dry
spring and early summer typical of the Colorado Plateau. For any single year or series of
years, the size and productivity of the niche
can be described and quantified and the spatial
distribution of the farmlands can be displayed
cartographically. In addition, the size of the
human population that can be supported from
that yield can be estimated, given certain assumptions about dry-fanning locations and
consumption rates. Changes through time in
these annual estimates can be examined on regional, subregional, and local levels.
The model incorporates the following logical elements (Figure 1.2). First, it reconstructs
soil moisture conditions for each soil type in
the region as they existed on July 1 of each
year in the reconstructed sequence. This is accomplished by calculating Palmer Drought
Severity Indices (PDSI) for the soils of the
study area within the period of instrumented
weather records (defined by the availability of
appropriate monthly temperature and precipitation records). quantifying the relationship
between PDSI values and historic tree-ring
width values. and retrodicting prehistoric PDSI
values with the preinstrumented and prehistoric
portion of the tree-ring record. Second. it
translates the soil moisture conditions as expressed by PDSI values into predicted crop
yields. first as production in beans and finally
as production in maize. Third. it calculates total
productivity of the agricultural habitat on a
year-by-year basis and provides an estimate of
maximum annual food supply potentially
available from maize agriculture. Fourth, it estimates the maximum arumal population size
and density that could have been supported by
the annual maize supply. given certain assumptions about agricultural demand (i.e.• annual consumption rates and potential levels of
storage). Finally. it examines the entire 400-
year sequence of yield and population values
and identifies sustainable population levels
over time. expressed as maximum, critical. and
optimal carrying capacity values.
While efforts to reconstruct climatic variation and its influence on agricultural productivity. population, and settlement in the Mesa
Verde area have been attempted before (e.g..
Bums 1983; Cordell 1975; Schlanger 1986b).
none has had the opportunity to use both the
high-quality environmental data and the spatial
data management systems that are now available. Without these newly derived environmental data and this recently developed computing technology, this research could not have
been done-the data are too many. the calculations too complex. and the accurate evaluations of options too numerous. Thus. computer
technology, especially geographic information
systems (GIS). plays a crucial role in this research. Without GIS it would have been impossible to integrate the reconstructed drought
indices and the calibrated series of agricultural
yields from the actual soils with the spatial locations of these soils. By capturing, co-registering. and evaluating all data layers. as weU as by
creating new data layers from reclassifications
and transfonnations of the original data, GIS
technology made the fast. accurate. and consistent (and therefore replicable) assignments
necessary to create the model. display the results. and assess patterning across space and
through time.
Two raster (grid cell-based) programs were
used to capture. store. manipulate. and display
the spatial data of the model: mainframe software called VICAR/IBIS and microcomputer
software called EPPL7. They made possible the
investigation of the study area at a relatively
high level of spatial resolution (200 x 200 m
or 4 ha cellular units) in an area 40 x 45.4 kIn
(or 1816 k.m 2). Input to the GIS was in the
form of previously generated computer values
entered through floppy disks, digital elevation
data purchased from the U.S. Geological Survey on magnetic tape. and newly digitized
spatial data and keyboard-entered tabular data.
Output from the GIS consisted of color graphic displays on video monitors, black-and-white
image output on dot matrix printers. and tabular data that were transferred to mainframe
programs for further analysis.
-3-
Data
Collection
Preliminary
Analyses and
Paleoenvironmental d。エ セpdsi
(TreeRing Chronologies)
Elevational Data
(OEMs)
セ
Soils Data
(Types, Locations,
Depth and w。エ・イMィッャ、ゥョァセ
Capacity)
I
セ
I
Soil Productivity Data セ
(Agricultural and Natural
Plant Productivity)
Analysis セlッョァMt・ョ
Preliminary
Products
PDSI セ
Reconstructions
Elevational Data .... PDSI
Plane (s)
Reconstruction
Data Plane
Final
Analyses and
Final Modeling
Products
t
.. Annual
POST
Reconstructions
.. Annual
Productivity
Reconstructions
.. Annual セ
Population
Estimates
セ
Soils Data
Plane(s)
Crop Yield and
Natural Plant
Productivity
Studies
t
Agricultural
Productivity and
PDSI Calibration
Studies
Ethnol:raphic Data
(Agricultural Practices,
Consumption and Per Capita
Demand for Yield)
Figure 1.2. The conceptual model for reconstructing prehistoric agricultural productivity.
I
LongTenn
Population
Estimates
Equally important to the precision and
success of the model is the availability of highquality environmental data for the study area.
Most of these have only been available since
1982 and some were created as recently as
1988. These include a contiguous set of 12
7.5-minute Digital Elevation Models (DEMs)
available from the National Cartographic Information Center, U.S.G.S., and completed
between 1985 and 1987; new maps and classifications of soils in the study area developed
by the Cortez Soil Conservation Service and
usable since 1988; estimates of potential natural plant productivity under varying climatic
conditions for each described soil, as well as
modern crop-yield estimates for selected soil
types (prepared by the Cortez Soil Conservation Service and at hand in 1988); recently
published agricultural yield data that summarize formerly dispersed and unstandardized data
(Burns 1983); and unpublished tree-ring data
from Martin Rose, related to work published in
Rose et al. (1982).
A final source of data for this study is the
archaeological record. While "real-world"
observations were not essential to model
implementation, they are extremely useful in
preliminary tests of the model's utility. Consequently, two fonns of archaeological data
were needed to investigate the usefulness of the
model for addressing questions important to
settlement analysis: population estimates from
high-quality block survey data and site locational information from tree-ring dated sites.
High-quality survey data were needed from at
least two locales in the region that differed in
elevation and exhibited significant settlement
during the period of interest. These could be
compared against the modeled estimates to
explore questions of carrying capacity. Minimum requirements for tree-ring dated sites
were at least two cutting dates, at least one
cluster of dates, and occupation during a limited period within the 4OO-year range. These
data were used to explore questions relating to
site catchment characteristics as well as optimal
locations and times for site selection.
Some of these archaeological data existed
prior to the initiation of this research, others
did not. Extensive field survey undertaken by
the author and others in 1986 and 1987 near
Sand Canyon Pueblo and Goodman Point Ruin
under the auspices of the Crow Canyon
Archaeological Center, Cortez, Colorado, provided the requisite settlement data needed to
conduct a preliminary test of the model. Treering dates were obtained from the Laboratory
of Tree-Ring Research, University of Arizona,
. for sites where complete locational and other
information was available from reports, other
publications, and site files.
ORGANIZATION OF THE STUDY
The remainder of this chapter is devoted to
placing the study area and research problem in
a spatial, temporal, and research context. This
is accomplished by a brief review of Puebloan
culture history in the Mesa Verde Region; a
brief history of archaeological research in the
study area; and a brief review of methods used
to reconstruct aspects of paleoenvironment on
the Colorado Plateau. Chapters 2, 3, and 4 describe the methods used to construct the model
and focus on data acquisition, preliminary data
manipulation, and final analyses. The methods
and data used or created to build the model are
described in chapter 2. Preliminary analyses
and preliminary products are described in
chapter 3. Final analyses are described in chapter 4, and a synopsis of the model-building
process is presented. The final products-the
long-term reconstructions of annual maize
yield and sustainable population at three levels
of storage/demand-are provided in the appendices. Chapter 5 describes the results of the
model-building effort and provides several test
applications of the database created by the
model. Chapter 6 summarizes the results of the
research and assesses the effectiveness of the
model. It includes a discussion of the strengths
and weaknesses of the data and the techniques
used, and suggests future modifications to test
and enhance the reconstructions of agricultural
productivity.
SPATIAL CONTEXT
The study area is located in southwestern
Colorado, about 20 miles north of the Colorado-New Mexico border and immediately east
of the Utah-Colorado border (Figure 1.1).
Notable landmarks within its perimeter are
Sleeping Ute Mountain on the south; the
-5-
northwest escarpment of Mesa Verde on the
southeast; the northward bend of the Dolores
River on the northeast; the major easttowest
flowing drainage of McElmo Creek in the
lower portion; the southwesttrending northern
tributaries, canyons, and mesas of McElmo
Creek in the lower half; and the gently rolling
mesa tops and broad highlands of the Great
Sage Plain in the upper half. The horizons beyond but visible from the study area are almost
entirely defined by mountains or distinctive
volcanic features: the San Juan Range of the
Colorado Rocky Mountains (including the La
Platas) to the east and northeast, the Mesa
Verde highlands to the southeast, the Shiprock
in New Mexico to the south-southeast, Ute
Mountain immediately to the south, the Carrizos of Arizona to the south-southwest, the
Abajos of Utah to the west, and the La Sals of
Utah to the north-northwest. The highest elevations in the study area are generally in the
northeast. The lowest elevation in the study
area is west of Ute Mountain at the level of
McElmo Creek at 1,500 m (4,922 ft). The
highest elevation is at the peak of Ute Mountain at 3,011 m (9,879 ft).
Physiographically, the study area is located
in the Navajo section of the Colorado Plateau
Province (Hunt 1956). Most of the area is
comprised of sedimentary layers of Cretaceous
(65-136 million years), Jurassic (136-190
million years), and Jurassicffriassic Period
(190-225 million years) sandstones and shales.
With the exception of Sleeping Ute Mountain
(a volcanic laccolith) and the cuesta-like Mesa
Verde, the uppermost formations in the study
area are the erosion-resistant Dakota Sandstone
and small remnant outcrops of overlying and
highly erodable Mancos Shale. Where canyons
have eroded the highland landscape, layers of
Cretaceous-age Burro Canyon deposits, Jurassic-age Morrison and Summerville Formations, and the Jurassicffriassic-age Navajo
Sandstone have been exposed (Ekren and
Houser 1965; Haynes et al. 1972).
Reddish-hued sand- and silt-sized particles
derived from sedimentary rocks and alluvial
deposits along the lower San Juan River have
been transported by wind from sources in
Arizona and Utah to the study area (Price et al.
1988). These eolian sediments were deposited
across the entire study area but have been
eroded away at canyon rims and slopes (places
that are usually characterized by exposed
bedrock and colluvial talus). Where present,
they have mixed with weathered residual sediments and have formed rich loamy soils (Arrhenius and Bonatti 1965). These reddish soils
mantle much of the highland portions of the
study area and rest on the Dakota Sandstone
substrate. These soils are referred to as eolian,
whereas others that were formed in place by
the weathering of parent material are referred
to as residual soils, and those that were deposited by water are referred to as alluvial soils.
Almost 100 soils (or more precisely "soil map
units") have been identified to date within the
boundaries of the study area. Of these, 44 soil
map units, mostly eolian-derived soils on mesas
and highland portions of the study area, are
currently used for dry-farming of pinto beans,
alfalfa hay, and winter wheat. Additional soil
map units, principally alluvial in origin, are
located in McElmo Creek Canyon, where fruit
trees and small grains are grown on floodplain
and lower terrace locations. While mineral, gas,
and oil exploitation are components of the
modem economy in this largely rural area,
agriculture is the mainstay.
The climate of the study area is typical of
that for most places in the northern Southwest
west of the Rocky Mountains. It has been described in Koppen's classification as a cold
middle latitude, semi-arid climate (Bsk) in
which potential atmospheric evaporation exceeds the usual amounts of precipitation available (Trewartha 1954:267). The area has a
biseasonal moisture pattern, with the majority
falling as snow more evenly over the area during the late winter (January through early
March), and a lesser amount falling as rain
usually in a patchy pattern as swnmer thunderstorms (July through early September). The
Dolores River is the only naturally occurring
source of permanent surface water. The McElmo Creek flows more-or-Iess permanently
today, but was seasonally ephemeral until the
1890s when it began to collect irrigation water
that was artificially routed from the Dolores
River (Freeman 1958). Ground water is available, however, in the form of seeps and springs
that occur at the contacts between permeable
water-bearing sandstones above and impervious shale layers below. These contacts are
most apparent in canyonhead settings or within
-6-
canyon walls, and have a wide distribution
across much of the study area.
The majority of lands in the study area
support an Upper Sonoran Woodland Biotic
Pinon
Community (Lowe 1964:11, SセVIN
and juniper woodlands occur on the shallow or
rockier soils of the mesas and highlands, and
sagebrush and other deciduous shrubs occur
on the deeper and more level areas at moderate
elevations. Woodlands of oaks, Ponderosa pine,
and occasionally Douglasfir occur at higher
elevations. Great Basin desert scrub occurs at
lower and drier elevations. Riparian vegetation
is fOWld in mesic areas, particularly in canyon
bottoms near water sources. More detailed descriptions of the flora and faWla in or near the
study area may be found in Brandagee (1976),
Erdman et al. (1969), Nickens (1977), Petersen
et al. (1985), Petersen and Orcutt (1987), and
Winter (1976).
of southwestern Colorado. The 400-year period is coterminous with the Pueblo II and
Pueblo III stages of Anasazi culture (Kidder
1927). and it incorporates notable "events" in
Mesa Verde Region culture history that have
been attributed to climatic and environmental
causes. These include the local abandonment
of the Dolores River Valley in the very late
800s or early 900s and the near absence of
settlement in this valley in the tenth and early
eleventh centuries; the wide geographic dispersal of settlement in the region from the middle
eleventh to the early twelfth centuries; the demise of the Chaco presence in the San Juan
Basin ca. A.D. 1125-1150; and the final abandonment of Mesa Verde Region at the end of
the thirteenth century. These four centuries,
then. represent an excellent set of years in
which to "observe" the effects of climatic
variation on local soils and on their potential
productivity.
TEMPORAL CONTEXT
The Anasazi presence immediately adjacent to the study area can be documented as
early as the A.D. 470s (Breternitz 1986), but it
is generally accepted that the main arrival and
settlement of Formative-level Anasazi Basketmaker III populations (probably from the east)
did not occur Wltil the late 500s and early
6005. Until about A.D. 1300, Anasazi populations settled, relocated, and resettled on lands
within the study area-their numbers sometimes waxing, sometimes waning, but seemingly always present. Prior to their presence,
seasonally-mobile hWlting and gathering Archaic populations (ca. 5500/6000 B.C.- A.D.
1), and before them, migratory big-game hunting and plant gathering PaleoIndian populations (ca. 10,000-5500/6000 B.C.) exploited
the resources of the study area. After the agricultural Anasazi left, seasonally mobile foraging populations (first the Numic-speaking Utes
and later the Athabascan-speaking Navajos)
returned to the study area perhaps as early as
the 1500s.
The reconstructions of agricultural productivity and sustainable population created by
this study are confined to the four centuries
beginning in A.D. 901. The lower limit was
determined by the availability of tree-ring data
necessary to the reconstruction process. The
upper limit of A.D. 1300 was selected to correspond with the end of the Anasazi occupation
Agricultural populations did not resettle
the study area Wltil the late 1800s, with the advent of the Euro-American farmers and ranchers who initially supplied the mines and later
established an independent agricultural economy when mining failed to create a dependable demand for crops and meat. Because
mobile populations of pre- and post-Anasazi
Archaeologically, the study area is located
in the Northern San Juan (Le., Mesa Verde)
Region of the Anasazi Culture (Bullard 1962).
The majority of the study area, with the exception of the northwest portion that incorporates
highlands near the Dolores River and the river
drainage itself, is contained within a sub-regional setting called the Yellowjacket District
by Gillespie (1976), the McElmo Drainage
Unit by Eddy et al. (1984), the Core Area of
the Northern San Juan by Fuller (1988), or the
McElmo-Yellow Jacket District by Varien et al.
(1990). In the typological schemes of these
authors, the northwest portion would also be
assigned to the Yellowjacket District by Gillespie (1976) but would be distinguished as being
a portion of the Dolores Drainage Unit by
Eddy et al. (1984), the Northern Periphery by
Fuller (1988), or the Dolores Valley District by
Varien et al. (1990).
-7-
periods built impennanent shelters and often
manufactured and used perishable items, much
less of their material culture is preserved and
recovered archaeologically. Moreover, until
recently, interest in the PaleoIndian. Archaic,
Ute, and Navajo culture and land use of the
study area has taken a back seat to interest in
the betterpreserved and often spectacular
Anasazi remains. By far, the vast majority of
archaeological remains reported for the study
area (and for the Mesa Verde Region in general) derives from the sedentary, agricultural
Anasazi occupation. Consequently, the brief
review of culture history that follows is restricted to the Anasazi Culture of the Northern
San Juan (the Mesa Verde) Region, and is
provided to place the 400 years of this study in
a temporal context.
The Anasazi Developmental Sequence in the
Northern San Juan Region
Basketmaker II
This first period of the Pecos Classification
system (Kidder 1927) dates from about 500
B.C. to A.D. 450/500 (Matson 1991). Sites attributed to Basketmaker II have only recently
been reported for the study area (e.g., Winter
1977). In addition, two sites immediately adjacent to the study area-a Mancos Canyon site
and Cougar Springs Cave near the Dolores
River-are thought to be affiliated with this
period (Hallisy 1974, redocumented by Bretemitz 1986; Gross and Howes 1986). Basketmaker II sites, however, are well documented in
portions of the region to the east (e.g., Eddy
1966; Morris and Burgh 1954) and to the west
(e.g., Lipe and Matson 1971; Matson and Lipe
1978; Matson et al. 1988). During this time,
populations built shallow pit houses that they
lived in for at least a portion of the year, built
small storage facilities, and cultivated maize
and squash. In the early portion of the period,
crops may have been grown primarily in
floodplain environments, but not by dry-farming teclmiques. In the later portion of the
period, around A.D. 100, dry-fanning techniques were also used to grow crops. Habitation sites were located in the open and within
rockshelters. Campsites were used as bases for
hunting and gathering, which continued to be
important sources of food. Populations were
small and dispersed, and habitation sites were
small, consisting of one or only a few structures. Distinctive material items and traits associated with this period include sophisticated
coiled and plaited basketry (but no pottery),
atlatls, large corner- or side-notched projectile
points, basin metates and one-handed manos,
rabbit-fur blankets. and characteristic rock art
motifs depicting large anthropomorphic figures associated with shamanistic experiences
(Cole 1989; Schaafsma 1980). Dogs were the
only domesticated animals.
Basketmaker III
Just as Basketmaker II culture probably
developed out of the Archaic, Basketmaker III
cultures continued traditions established in
Basketmaker II. This developmental stage dates
roughly from A.D. 500 to 750. It is well documented in the study area (e.g., Bretemitz et al.
1986; Fetterman and Honeycutt 1987; Kuckelman and Morris 1988; Lange et al. 1986; Martin 1939; Rolm 1974, 1975; Schlanger 1985;
Wheat 1955; Winter 1977) and throughout the
Northern San Juan Region. Basketmaker III pit
houses were larger and deeper than those in
Basketmaker II times and typically included
wing-walls, clay-lined hearths, low banquettes,
and interior storage cists. Sipapus, interpreted
by archaeologists as symbolic places of emergence from the underworld, appeared for the
first time.
Accompanying pit houses were exterior
storage facilities and surface features situated
to the north and midden areas located to the
south-an internal site configuration and orientation that persisted through the Pueblo III
period. Some Basketmaker III settlements were
surrounded by wooden stockades, although the
function of such features is not clear. Sites occurred in both rockshelters and in the open.
Habitation sites were usually small and consisted of one or a few houses, although larger
villages in open settings do occur on occasion.
The greater investment in architectural facilities, greater attention to storage, more efficient
grinding equipment, and addition of beans to
the diet in this period, are regarded as signifying a greater dependence on cultivated products and year-round habitation-a trend that
continues through Pueblo III times. As with
Basketmaker II, campsites and rock art sites are
-8-
known to be associated with Baskennaker III
culture. Although sites were smal1 and moderately dispersed, there is some evidence for the
aggregation of populations in particular locations. Great kivas have been identified as being
associated with Baskennaker III sites and represent the earliest evidence for the use of public architecture to integrate dispersed populations (unless Baskennaker II rock art panels
served a similar purpose). Distinctive material
items include the first evidence of fired ceramic vessels in the form of smoothed, coilmade gray jars and bowls; the addition of the
bow, arrow, and small comer-notched projectile point (and eventual replacement of the
atlatl, dart/spear, and large comer-notched
projectile point); and the introduction of
trough metates and two-handed manos. Turkeys are believed to have been domesticated by
Baskennaker III times.
Pueblo I
The Pueblo I period dates from about A.D.
750-900. It is wel1 represented in the study
area (e.g., Breternitz et al. 1986; Fetterman and
Honeycutt 1987; Martin 1939; Varien and
Lightfoot 1989; Winter 1977) and in the Northern San Juan (e.g., Brew 1946; Debloois and
Green 1978; Haase 1983; Hayes 1964; Hayes
and Lancaster 1975; Morris 1939; Roberts
1930; Rohn 1977). Pueblo I populations
tended to be clustered in distinct locations and
exhibited definite tendencies toward living in
larger, aggregated settlements than did their
ancestors. A typical Pueblo I habitation site
consisted of one or more deep, squarish pit
houses located south of an arc or row of
contiguous surface rooms and north of a trash
midden. Subterranean pit houses continued to
have a variety of floor features as part-time
domestic structures but had roof entries and
ventilator systems rather than antechambers;
some ("protokivas") exhibited greater evidence for ceremonial use in the form of complex sipapus and lateral floor vaults (Wilshusen
1989). The surface rooms were usually arranged in two rows and built as "modular
units" with two smaller, generally better-made,
storage rooms to the north and a large living
room to the south with a south-facing door.
Both jacal and simple masonry were used to
construct walls. These modular units are considered to be the domain of a single household
unit, and when joined, are considered to represent the co-residence of several household
groups. Pit structures are believed to have been
shared by more than one household group.
Larger-than-average pit houses and great kivas
are documented for some Pueblo I settlements
(e.g., Lightfoot 1988) and are thought to symbolize attempts to unify and provide cohesion
for local populations, exceeding Baskennaker
III attempts. Other than habitations, site types
dated to this period include a variety of limited
activity locations and field houses. Distinctive
material items and traits include the deliberate
deformation of the skul1 by means of a hard
"pil1ow" positioned at the back of the head of
an infant on a cradleboard; neckbanded ceramic jars, widespread trade of redware jars and
vessels, small stemmed and tanged projectile
points, and the addition of cotton to the list of
cultivars in the warmer, better-watered part of
the Anasazi world (probably not in the study
area).
Pueblo II
Until recently, the Pueblo II period has
been dated to A.D. 900-1100, and subdivided
into an early and later period of about 100
years each. At the present time, it is not uncommon to consider that Pueblo n lasts until
about A.D. 1150 (Lipe 1989). Populations
were larger and sites of this time period were
widely distributed across the study area (e.g.,
Adler 1988; Bradley 1988a; Chandler et al.
1980; Fetterman and Honeycutt 1987; Kuckelman and Morris 1988; Martin 1938; Winter
1976) as well as in the Northern San Juan Region (e.g., Eddy 1966; Hayes 1964; Hayes and
Lancaster 1975; Matson et al. 1988; Swannack
1969). In fact, within this period the maximum
geographic range of the Anasazi culture was
reached. This period is associated with the florescence and the demise of the Chaco sociopolitico-religious expression in the American
Southwest. Several sites in and adjacent to the
study area (including Wallace, Ida Jean, Lowry,
Escalante, Yucca House, and Ansel Hall Ruins)
have architectural attributes that are believed to
be expressions of Chacoan presence or influence. These Chacoan architectural traits (i.e.,
Chaco great houses, roads, tower kivas, great
kivas, formalized trash middens, or Chaco-style
masonry, architectural features, and building
proportions) generally occur in association
-9-
with clusters of somewhat dispersed but apparently contemporary Pueblo II sites believed to
be part of a single interacting community. The
Chacoan features and associated ideology may
have acted as forces of integration for dispersed Pueblo II settlements and communities.
In contrast to late Pueblo I settlements, individual Pueblo II settlements were home to
fewer households and were dispersed more
broadly across the landscape in a wider variety
of settings. While the basic architectural arrangement and orientation (the classic "Prudden unit" fonn [Prudden 1903]) persistedroomblock to the north, pit structure in the
center, and trash midden to the southsurface
structures became more substantial and more
commonly built with spalled stone masonry.
The pit structures are recognized by fonnal
characteristics and are commonly cal1ed kivas
rather than pit houses. These fonnal characteristics include a round instead of square plan;
a cribbed roof supported on pilasters built
above a bench instead of a flat roof supported
by four central post supports; inclusion of a
banquette or bench; the development of a
southern recess in the bench, and a greater use
of stone to line the lower chamber. In addition
to field houses and limited activity sites of
various kinds, water control features (Haase
1985; Rohn 1963) including acrossslope
terraces, acrossdrainage checkdams, artificial
reservoirs, and shrines were associated with
Pueblo II culture. Distinctive items associated
with this period include the manufacture of allover corrugated jars, loomwoven textiles,
turkeyfeather blankets, and small cornernotched points with expanding stems and
convex bases.
Pueblo III
Although the Anasazi sequence continues
elsewhere in North American Southwest, this
period is the last in the Northern San Juan
Region. A dramatic regionwide depopulation
at the end of the thirteenth century has led researchers to search for a single powerful cause
that could account for this abandonment.
Pueblo III is dated from either A.D. 1100 or
1150 to A.D. 1300. Initially it was a time of
continuity of tradition in architecture and settlement patterns, with most of the population
living in small villages. However, by A.D.
1200, a dramatic coalescence of population
into fewer but larger settlements had begun.
These sites are often in close proximity to
strong springs near mesa rims and canyon
heads. In addition, the deliberate management
. of watersoil resources continues and
apparently increases in this period (Haase
1985; ROM 1963, 1972).
Sites in the study area that date at least in
part to this period include Yel10w Jacket
(Lange et al. 1986), Goodman Point (Fewkes
1919), Sand Canyon Pueblo (Bradley 1986,
1987, 1988a; Kleidon and Bradley 1989), Mud
Springs (Fewkes 1919), and the Lancaster Ruin
(Martin 1929). In other places in the Northern
San Juan, the period is represented by the famous cliff dwellings of Mesa Verde National
Park, such as Cliff Palace, Balcony House,
Spruce Tree, Long House, and Mug House
(Cattanach 1980; Rohn 1971, 1977); Lion and
Hoy House in Mancos Canyon (Nickens
1981); and sites like the Nancy Patterson Village in southeastern Utah (Thompson et al.
1986).
Pueblo III settlements generally retained
their modular configuration of roomblock, .
kiva, and trash midden with modules often
linked into agglomerated structures with structurally incorporated kivas. By the beginning of
the thirteenth century, informal plazas, enclosing walls, and towers were common features of large habitation sites. Kivas were
regularly lined with carefully shaped and
pecked masonry. New architectural features
particular to this time period include masonry
cliff dwellings; rare D.shaped, biwalled, or triwalled structures; and relatively common freestanding towers at canyonheads, such as those
of Hovenweep National Monwnent. The function of the Dshaped and bi or triwalled
structures is unknown (but see Bradley 1992).
Multiple functions have been postulated for
freestanding towers and include defense,
intersite communication, community storage,
religious or astronomical significance (Riley
1950; Schulman 1950; Williamson 1984;
Winter 1984). Distinctive items associated with
this period include slab metates, flat twohanded manos; tchamahias; new ceramic vessel
forms (e.g., canteens, mugs, flutedhandled
dippers, kiva jars); elaborate ceramic designs;
the predominant use of carbon paints rather
10-
than mineral paints to decorate vessels; and
small, sidenotched, triangular projectile points.
Previous Archaeological Research in the Study
Alli1
The Northern San Juan Region is one of
the most intensely studied archaeological areas
in the North American Southwest. Comprehensive reviews of the history of investigations
in this region have been published by Brew
(1946), Herold (1961), Nickens and Hull
(1982), and Eddy et aI. (1984). These works
contain valuable chronological summaries as
well as commentary on the significance of selected research efforts for advances in archaeological method, theory, and resource management. The brief review that follows identifies
the major investigations within the study area.
Its intent is to emphasize that the archaeological remains of the study area are abundant,
historically well-known, and continue to be investigated as a source of information on prehistoric Pueblo populations and their adaptations.
Whether classified as a cultural core area,
drainage unit, or archaeological district, the
study area has been the location of a remarkable amount of archaeological work, and a
large number of well-preserved architectural
ruins are contained within it Densities of over
40 sites/km2 are common for those places in
the study area that have been systematically
surveyed in recent years--an impressive fact
considering the intensive land modification
that has occurred in the last 100 or so years
result of agriculture. energy exploration. and
other forms of land development.
Early Observations and Investigations
The earliest observations recorded on the
archaeological remains of the area are those of
the Franciscan fathers, Escalante and Dominguez. who passed through the study area in
1776 on their WlSuccessful search for a route
to Monterey. California from Santa Fe, New
Mexico (Bolton 1972). However, almost 100
years passed before goverrunent officials and
the public became aware of the richness of this
archaeological area through the explorations
and publications of the U.S. Corps of Topo-
-11
graphic Engineers (Macomb 1876; Newberry
1876) and the Hayden Survey (Holmes 1878;
Jackson 1878). The earliest surveys and excavations in the study area are known through
the publications of Prudden (1903, 1914,
1918), Morley (1908), Morley and Kidder
(1917), Fewkes (1919), and Roberts (1925).
This is also the time when the Wetherills were
actively involved in site exploration. Although
not systematic or intensive by the standards of
today, these early investigations helped establish the basic chronology for the Mesa Verde
area (e.g., that the Baskennakers were the cultural predecessors and ancestors to the Pueblo),
identified basic architectural configurations
(e.g., the modular "Prudden unit" of Pueblo
I-III Anasazi habitations). and perceived the
enduring settlement and adaptive patterns associated with prehistoric Pueblo culture.
Privately Funded and Locality-Based
Archaeological Research
The first period of rigorous investigations
dedicated to exploring a single locality in the
study area was initiated by Paul Martin in the
Ackmen-Lowry area in 1928. Between 1928
and 1938, Martin and his associates investigated sites representative of seventh century
Basketmaker m through thirteenth century
Pueblo III occupation in the Mesa Verde
Region in the north-central portion of the
study area. These included excavations at the
well-known ruins of Herren Farm and Charnel
House (Martin 1929). Beartooth and Little Dog
(Martin 1930). Lowry (Martin 1936) and a
number of Basketmaker III through Pueblo n
sites in northern portion of the study area
(Martin 1938. 1939). Major studies of architecture (Roys 1936) and ceramics (Rinaldo
1950) resulting from the Ackmen-Lowry work.
have become part of the classic literature of
southwestern archaeology.
South and east of Lowry. in the central
portion of the study area, a long-term research
project including both survey and excavation
in the YeUowjacket area was initiated in 1954
by Joe Ben Wheat as annual field schools of
the University of Colorado. Boulder (CU).
These annual expeditions continue today and
have focused on excavation of sites in the
"suburbs" of the YelIowjacket Ruin community, the largest Anasazi ruin in the study area
(Brown 1975; Lange et al. 1986; Swedlund
1969; Wheat 1955, 1980). Prior to these CU
field schools, Hurst and Lotrich from Western
State College of Colorado did some limited
work in 1931 on the Yellowjacket Ruin itself
(Hurst and Lotrich 1932, 1933). Elsewhere in
the locality, ROM and students from Wichita
State University conducted several seasons of
work in the late 1960s and early 1970s. Many
sites were recorded on survey and a number of
sites were tested, including the stockaded Basketmaker III sites, Payne and Gilliland (ROM
1974, 1975).
Locality-based research on Mockingbird
Mesa, in the west-central portion of the study
area, was undertaken by Ives and students from
Fort Lewis College, Durango, from 1970 to
1973. Seventeen sites ranging in age from
Basketmaker III to Pueblo III were excavated
(Ives 1971a, 1971b, 1972, 1973). The Bureau
of Land Management undertook an intensive
survey of Mockingbird Mesa between 1981
and 1984 and located 684 sites on 3,976 acres
of land (Fettennan and Honeycutt 1987). This
survey provided significant data on settlement
patterns for regional analyses.
In the southwestern portion of the study
area, Riley surveyed five units of Hovenweep
National Monument (Riley 1948) and published a study of the famous towers from that
area (Riley 1950), which was complemented by
a spatially extensive study of towers by Schulman (1950) who examined these architectural
features across the northern Southwest. Later,
problem-oriented research was undertaken in
the Hovenweep-Cajon Mesa area by San Jose
State University to gain a better understanding
of Anasazi farming practices (Hammett and
Olsen 1984; Winter 1975, 1976, 1977; Woosley 1977). An offshoot of this project was the
systematic and complete survey of Cow Mesa
north of the Hovenweep area by Neily (1983).
A recent program of research was begun in
1983 in the Sand Canyon locality in the southcentral portion of the study area. In conjunction with major excavations at Sand Canyon
Pueblo, the Crow Canyon Archaeological
Center, Cortez, is conducting survey and testexcavations in late Pueblo III sites within an
approximate 200-km2 locality defined around
the Sand Canyon Pueblo and Goodman Point
Ruins (Adams 1985; Adler 1988; Bradley
1986, 1987. 1988a; Huber 1989; Huber and
Bloomer 1988; Kleidon and Bradley 1989;
Lipe 1992; Van West 1986; Van West et al.
1987; Varien 1990). Prior to Crow Canyon's
research. only two sites in the Sand Canyon
locality had been excavated, one by Colorado
Mountain College. Leadville (Bagwell, personal
communication, December 1986), and one by
Gould (1982). A little to the east of the Sand
Canyon Locality. the Crow Canyon Archaeological Center also conducted a major excavation and small locality study around the late
Pueblo I Duckfoot Site in the vicinity of Crow
and Alkali Canyons (Adams 1984; Lightfoot
1985, 1987, 1992; Lightfoot and Van West
1986; Lightfoot and Varien 1988; Varien and
Lightfoot 1989).
A final locality-based study located in the
far east-central portion of the study area was
begun in 1969 by Bradley with the partial excavation of the Wallace Ruin. one of three major sites in the Lakeview Group (the other two
being the Ida Jean Ruin and Haney Ruin). This
privately-funded research has continued intermittently since 1969 and has produced the
strongest evidence to date for the presence or
influence of Chaco culture in the heartland of
the Mesa Verde Region (Bradley 1974, 1984,
1988b).
Federally-Sponsored Archaeological Research
and Resource Management
A new genre of archaeological research
was introduced to the study area in the 1960s
with the extensive reconnaissance-level surveys
of grazing units managed by the Bureau of
Land Management. Hundreds of archaeological sites and areas of high site density were
identified during the 1965-1969 surveys (D.
Martin 1971) for the purpose of resource inventory and resource management Most archaeological research in the study area since
1960 has been in response to federal and state
laws and resource management policies concerned with the identification, protection,
and/or interpretation of cultural resources.
The Bureau of Land Management (BLM)
is the federal land-managing agency with the
most extensive holdings in the study area.
Subsequent to the reconnaissance surveys of
12 -
the 1960s, the BLM was responsible fOr the
previously mentioned intensive systematic survey conducted on Mockingbird Mesa; an
8,OOQ-acre quadrat-based survey of lands in
the Sacred Mountain Planning Unit of the
Montrose district, which produced a statistical
sample useful for infening characteristics of
the total Plamting Unit (Chandler et al. 1980);
and two systematic surveys within lower Sand
Canyon (Adler and Metcalf 1991; Gleichman
and Gleichman 1992). It has sponsored excavations in the study area at Escalante Ruin
(Hallasi 1979), Dominguez Ruin (Reed 1979),
Casa de Suenos and another small site near the
Anasazi Heritage Center (Douthit 1984), and it
has initiated a number of ruins stabilization
programs at sites under its care. Stabilization
has been undertaken at Lowry Ruin (White and
Bretemitz 1976), Escalante and Dominguez
Ruins (White and Bretemitz 1979), East Rock
Canyon (White 1976), Sand Canyon Pueblo
Cliff Dwelling (Metzger and Bretemitz-Goulding 1981), and towers in the McLean Basin
and lower Sand Canyon (Tipps 1978). In addition, a stabilization evaluation program was
conducted for 49 cliff dwellings in Sand Canyon that provided descriptive data (C. Martin
1976). Finally, the BLM has become the curator of sizable archaeological collections
resulting from the Dolores Archaeological
Project (described below), the locally renowned
Chappell Collection, and other sites on BLM
land in the study area. Their collections are
housed in the new Anasazi Heritage Center
adjacent to the Escalante and Dominguez sites
near Dolores. The Center serves as a collections
repository and an interpretive museum.
In 1972 a major program of data recovery,
which included survey, site testing, large-scale
excavation, analysis, and model building, was
initiated by the Bureau of Reclamation in order
to build the McPhee DamlReservoir and its
associated sets of canals and laterals on the Dolores River. Early survey in and adjacent to the
study area resulted in rePorts that identified the
range of cultural resources to be affected by
the project (Breternitz and Martin 1973; Kane
1975a, 1975b; Nickens 1977; Toll 1977). In
1978, a major contract was awarded to the
University of Colorado to recover a significant
sample of data from the project area prior to
the filling of the reservoir and construction of
related facilities. The project, known as the
Dolores Archaeological Project (DAP), became
one of the largest archaeological studies ever to
have taken place in the United States. The
project ran from 1978 through 1985 and
generated scores of published reports, major
analytic studies, theses, and dissenations. The
primary results of the eight-year project are
published in a series of Bureau of Reclamation
reports too numerous to cite here, with each
volume devoted to a particular set of topics or
activities. The project is summarized in a final
synthetic report (Bretemitz et al. 1986). In
addition to the Bureau of Reclamation reports,
substantive publications of contracted research
undertaken on non-reservoir portiOns of the
Dolores Project have been produced. Auxiliary
programs of project-related survey, testing, and
excavation within the study area were undertaken for the Dove Creek canal (Fuller 1985a,
1985b, 1987; Honeycutt and Fetterman 1985;
Ives and Enickson 1983; Ives and Orcutt
1981; Monis 1986b), South canal (Kuckelman
and Monis 1988), Towaoc canal (Kuckelman
1986), Fairview laterals (Walkenhorst et al.
1983; Monis 1986a), Cahone laterals (Walkenhorst 1983), and the Ute Mountain Ute inigated lands (Fuller 1988).
A great deal of contracted archaeological
research in the study area has been conducted
in conjunction with energy exploration and
resource development. The mining and processing of carbon dioxide gas in southwestern
Colorado for use as a tertiary recovery agent in
the oil fields of West Texas has resulted in surveys and excavations within proposed facilities
and along rights-of-ways for pipelines and
powerlines (e.g., CASA 1981; Dykeman 1986;
Fetterman and Honeycutt 1982a; Hammack
1983, 1984; Honeycutt and Fetterman 1982;
Whitten et al. 1986; Woods Canyon Archaeological Consultants 1985). Similarly, construction of a liquid hydrocarbon pipeline for El
Paso-Northwest Gas companies initiated many
miles of intensive survey, test excavations, and
full-scale excavations along proposed rightsof-way (Fetterman and Honeycutt 1980,
1982b).
RESEARCH CONTEXT
Paleoenvironmental research is well established in the American Southwest and may be
13-
classified as to the sources of its data: geomorphological, palynological, dendroclimatological, macrobotanical, and faunal studies (Cordell
1984:35). Of these, the first three are the major
sources of information for reconstruction of
Anasazi environments. Each of the five specialties has its strengths and limitations as well as
appropriate levels of resolution and confidence
(Cordell 1984:35-45; Dean 1988b), but
underlying all attempts at paleoenvironmental
reconstruction is the principle of uniformitarianism, which assumes that processes responsible for the variation in the natural world that
can be observed in the present also operated in
the past.
Geomorphological studies generally have
focused on identifying and dating episodes of
deposition and erosion along drainages, identifying the fluctuations in ground-water tables,
and finally suggesting the climatic and/or cultural correlates for these cutting and filling
events (e.g., Bryan 1925, 1941, 1954; Hack
1942; Hall 1977; Karlstrom 1988; Schoenwetter and Eddy 1964).
Palynological studies generally have focused on the recovery and identification of
fossil pollen and spores from cultural and noncultural contexts to reconstruct the composition and distribution of former vegetation
communities. From these data are inferred the
prevailing temperature and precipitation patterns (e.g., Dean et al. 1985; Euler et al. 1979;
Hevly 1988; Martin and Byers 1965; Petersen
1981, 1988; Petersen and Mehringer 1976;
Schoenwetter 1970; Schoenwetter and Eddy
1964) or human activities (e.g., Betancourt and
Van Devender 1981; Wycoff 1977) that likely
produced these plant assemblages. Pollen
recovered from archaeological sites is less
often used for climatic reconstruction than it
once was because the biases introduced by
human agency are now recognized, but it is
frequently used to identify the presence of
plants of economic importance and cultural
significance (Bohrer 1981).
Dendroclimatological studies in the
Southwest use knowledge of the particular annual growth characteristics of conifers such as
pifton, juniper, Douglas-fir, and ponderosa to
infer annual and seasonal patterns in precipitation and temperature that account for that
growth. Sophisticated biological and mathematical models have been developed to discriminate the macroclimatic components of
tree growth from the microclimatic and genetic
components. Tree-ring chronologies of 1,500
to 2,000 years or more have been constructed
by the merging of archaeological tree-ring
chronologies with living-tree chronologies to
create an absolutely-dated record of local and
regional climatic conditions with a temporal
resolution of one year (Dean and Robinson
1978). The modem end of these chronologies
can be calibrated with various climatic variables
to estimate (or "retrodict") values for these
same variables for the full length of the chronologies. In this way, long records of annual or
seasonal values of precipitation and temperature (Hogan 1987; Rose et al. 1981, 1982;
Sebastian 1988), stream flow (Graybill 1989;
Larson and Michaelsen 1990; Stockton 1975;
Stockton and Boggess 1980), annospheric pattern anomalies (Blasing 1975; Fritts et al.
1971), crop yields (Bums 1983), and droughts
or meteorological conditions as measured by
Palmer Drought Severity Indices (Meko et al.
1980; Mitchell et al. 1979; Rose et al. 1982)
have been created.
Macrobotanical studies generally focus on
the identification of seeds and plant parts recovered from archaeological contexts to reconstruct economic practices of former populations rather than paleoclimates or paleoenvironments (e.g., Adams 1980; Bohrer 1981;
Bohrer and Adams 1977). However, packrat
midden analysis-a type of macrobotanical
study-is increasingly being used to reconstruct the distribution of vegetation communities and their associated climatic regimes (Betancourt and Davis 1984; Van Devender and
Spaulding 1979).
Like macrobotanical studies, faunal studies
generally focus on the identification of animal
remains recovered from archaeological contexts to infer past subsistence behavior. However, knowledge of the specific ecological adaptations of certain species and observation of
changes in the frequency of one indicator species to another have been used to infer local
vegetation types and changing environmental
conditions that are likely to have been present
in the past (e.g., Flint and Neusius 1987; Harris
1970; Mackey and Holbrook 1978).
14 -
While it is known that major patterns of
climate and vegetation patterns have been
stable in the Southwest for some 8,000 years
(Van Devender and Spaulding 1979), fluctuations in temperature and precipitation that
result in "minor" shifts of water supply and
the distribution of plants and animals have
been identified. Some of the changes in environmental conditions are caused by natural
processes that occur slowly, some even regularly, over many centuries. Generally they
occur in periodicities of greater than 25 years.
Changes in alluvial water tables and episodes of
arroyo cutting and filling are often-cited
examples. Other changes in environmental
condition are caused by natural processes that
occur quickly, on the scale of years, seasons,
and even shorter periods of time. Changes in
the seasonal patterns of precipitation and
temperature, and the rapid changes in the size
and distribution of certain plant and animal
populations are examples.
Dean (1988a) has termed the former Low
Frequency Processes (LFP) and the latter High
Frequency Processes (HFP) Paleoenvironmental reconstructions drawn from geomorphology and some palynological studies generally reconstruct LFP changes, whereas paleoenvironmental reconstructions drawn from
dendroclimatology, some palynological
studies, macrobotanical studies, and faunal
studies generally reconstruct changes conditioned by HFP. Both are important to document because they have different behavioral
and adaptational consequences for human
populations. Dean contends that humans generally do not perceive LFP because they often
occur so slowly (although there are predictable
exceptions to this; for a complete discussion
see Dean 1988a), and that populations consider
these conditions to be the prevailing and stable
condition of their environment. When LFP
changes do occur, they are severe and require
major behavioral adaptations on the part of
human populations to survive.
Alternatively, he proposes that the environmental changes caused by HFP are quite
perceptible and are considered to be within the
range of normal variation experienced by the
population. Consequently, it is to these variations in climate and resource availability that
human populations regularly respond with
-
behavioral mechanisms such as water-control
technology, food-storage systems, and exchange networks. For additional discussion of
this topic, the reader is referred to Dean's full
treatment of tttis model of "Anasazi behavioral
adaptation" (Dean 1988a). The significance of
ttis distinction between LFP and HFP here,
however, is that both types of processes and
their effects on the environment and the subsistence and settlement patterns of prettistoric
populations must be reconstructed in order to
fully comprehend cultural stability and
change. Thus, the goal of the archaeological
efforts to reconstruct past climate and environments in the Southwest should be to document
as fully as possible these fluctuations and their
effects on human populations.
The research described in this study is
based on dendroclimatological data. Its goal is
to reconstruct local high-frequency variation in
regional climate, as it has been modeled by
tree-ring data, and to observe its effect on the
distribution of arable land and potential human
carrying capacity. The work has been inspired
by four studies in particular: the conceptual
model of Anasazi adaptive behavior proposed
by Dean et at. (1985; see also Euler et al.
1979; Gumerman 1988); the methods proposed in Rose et al. (1981) for high-resolution
reconstructions of past climate; Bums' (1983)
long-term reconstruction of crop yields for
Southwestern Colorado; and Petersen's reconstruction of the dryland farm belt in the Dolores Archaeological Project Area (Petersen
1988). One of the conclusions drawn by Dean
and his colleagues in the 1985 paper was that
no matter how compelling their regional reconstruction of paleoenvironment, demography, and human behavior, it is still necessary to
create local reconstructions of environment,
population, and behavioral variability to explain sociocultural stability and change on the
Colorado Plateaus (Dean et al. 1985:550).
The method used here for creating highresolution reconstructions of annual and seasonal precipitation and temperature and Palmer
Drought Severity Indices (PDSI) was described
by Rose et al. (1981) in a major study for the
southeastern portion of the Colorado Plateaus.
Of these reconstructions, those employing the
PDSI produced the best results. This was
attributed to the fact that-like tree-ring
15-
widthPDSI (an index of stored soil moisture)
is a measure that integrates the cumulative
effects of both precipitation and temperature
(Rose et al. 1981 :244). The attempts by Bums
(1983) to reconstruct annual yields of bean
and maize for southwestern Colorado at varying levels of storage and to make inferences
about specific intervals of food shortfalls and
food surplus and their correspondence to periods of major building episodes, was made possible by the existence of appropriate long-term
tree-ring chronologies. Bums' study led me to
believe that high-frequency reconstructions of
climate could have direct applications to the
interpretation of the archaeological record.
Finally, Petersen's reconstruction of changes in
the location and width of the potential dryingfanning belt in southwestern Colorado from
A.D. URSQセ
(Petersen 1988; Petersen and
Clay 1987) led him and some other DAP researchers to conclude that fluctuations in temperature and precipitation were of sufficient
magnitude and duration that populations were
forced to migrate from the area because of
their inability to grow adequate supplies of
food. This was cited as the probable cause of
the depopulation of the Dolores area in the
-
early 900s and of the entire Mesa Verde
Region in the late 1200s. Yet, the Dolores area
was relatively attractive from an agricultural
perspective after the early tenth century local
abandonment and the late thirteenth century
regional abandonment. Therefore, it appeared
to me that the role of climatic fluctuations on
potential agricultural productivity and on sustainable population size and spatial distribution
was unresolved, and that further investigation
of this relationship was warranted.
In retrospect, Dean et al. (1985) validated
the need for locally relevant, high-frequency
reconstructions of climatic variability by demonstrating climatic influence on imponant
aspects of Anasazi adaptations. Rose et al.
(1982) described how high-frequency climatic
reconstruction could be done and suggested
use of the Palmer Drought Severity Index as a
particularly effective measure of climate. Bums
(1983) demonstrated the archaeological value
of high-frequency, long-term reconstructions
of crop production in particular. Finally,
Petersen's (1988) findings challenged me to
examine this question in a new way with
recently available data and methods.
16-
2
A model is a simplified representation of
some multivariate, dynamic, realworld phenomenon. The ability of a model to faithfully
portray and ultimately predict behavior is
largely a function of two conditions. First, the
variables that account for the greatest amount
of variation in the targeted phenomenon must
be correctly identified and included. Second,
the observations or measurements taken on
those key variables must be of sufficient quality and quantity that the dynamic character of
the phenomenon can be perceived.
The phenomenon under investigation here
is agricultural productivity of lands used for
dryland fanning of maize in southwestern
Colorado. For the purposes of this research, the
critical environmental variables accounting for
the most variation in pre-industrial agricultural
production are arable soil quality, soil moisture, and adequate growing season length (see
Hogan [1987 :29--60] for a detailed discussion
of growth requirements of maize).
Arable soil quality is measured on a number of dimensions, but texture, depth, waterholding capacity, and natural productivity are
among the most telling attributes. The average
crop yield of a specific soil type is, of course,
an excellent measure of productivity, but these
data are not available for all soils. Natural productivity measures the ability of a soil to produce new native plant growth on an annual basis. It is most influenced' by plant nutrients and
moisture supply and to a lesser degree by soil
acidity, salt content. and a seasonal high water
table (Soil Conservation Service 1951). The
Soil Conservation Service measures this property on rangeland as "natural potential productivity" under favorable, normal, or unfavorable climatic conditions. It is expressed as
the weight, in pounds, of armually produced
air-dried understory vegetation per acre of
land. In this study. this natural potential productivity is used as an integrative index of a
particular soil's ability to produce plant
growth. whether native or domestic.
Soil moisture is largely a function of soil
texture and water-holding potential, coupled
with precipitation, evaporation, and runoff. It
can be calculated by the Palmer Drought Severity Index, which requires data on waterholding capacities of each subsurface layer,
total depth of the soil profile, and a source of
information on precipitation and temperature
input for that specific soil and location. Historic records of at least 30 years in length on
monthly precipitation and temperature from
various weather stations in the Four Comers
area provide the climatic data used to calculate
soil moisture values in this study.
Adequate growing season length, in this
area, is predominantly a function of elevation,
mediated through the adaptive capability of the
plant to withstand temperature extremes. Elevation is used as a proxy for the general zone
in which the number of growing days (or alternatively expressed, the number of killing
frost-free days) is sufficient to grow and mature a crop of maize. The location of each soil
type can be recorded from maps and photographs and georeferenced セゥエィ
digital elevation
maps.
mGH RESOLUTION SOILS DATA
Descriptive and Locational Infounation
Currently Available
Information on soil was obtained from the
Cortez field office of the U.S. Department of
Agriculture, Soil Conservation Service (SCS).
This office is responsible for the classification
and inventory of land in Montezuma and
Dolores Counties, southwestern Colorado. and
the provision of land use information to the
public. Field survey by SCS staff produces soil
maps and soil descriptions published in Soil
Survey manuals for specific areas. The reports
for Montezuma and Dolores Counties are not
yet available, but by 1988, when data were being gathered for this study, most of the land in
the study area had been mapped as to soil type,
and preliminary descriptions of those soils had
been written. Certain data on soil productivity
and water-holding characteristics also had been
amassed. Therefore, it was possible to record
soil location from aerial photographs and
gather data on soil attributes from materials on
file in that office.
The SCS recognizes a total of 113 different
soil map units in the two-county area. These
units are either consociations of a single soil
taxon representing a soil series (e.g., Witt
Loam, 1-3% slope) or complexes of heterogeneous soil taxa (e.g., Romberg-Cragola-Rock
Outcrop Complex, 25-80% slope) that are
given distinct names and unique field map
symbols (e.g., ROB and M2CE, respectively).
Descriptions of these soil map units include information on topography, vegetation, range or
woodland association, elevational range, average annual precipitation and temperature, and
length of the killing frost-free period. Where
relevant, all identifiable components of that
map unit are described and their relative contribution to the soil complex estimated. Information on soil depth, permeability, shrink
-swell potential, and rooting depth are given,
and all subsurface units of the individual soils
are described. The potential natural plant community on each named soil map unit is estimated in pounds per acre of arumal production
of air-dried understory vegetation for a normal
growing year. For each soil, the range of current land uses and an ev'aluation of agricultural
capability are reported. Comments on agricultural land management practices are also included.
Each soil map unit is classified on the basis
of a number of inherent and situational characteristics. Inherent characteristics include: composition, arrangement, and thickness of subsur-
face horizons; soil structure, color, texture, reaction, and consistence; content of humus, rock
fragments, carbonates, and other salts; and
mineralogical composition. These inherent
properties govern such agriculturally important
qualities as productivity, tillage, rooting depth,
and available water capacity. Situational characteristics are also used to classify soil types
and include slope, elevation, and topographic
position.
Archaeologists working in southwestern
Colorado have often used the SCS capability
classification value of a soil map unit as a
proxy for its appropriateness for prehistoric
agriculture and to a lesser extent, its agricultural productivity. While this is a convenient
measure, it may not be sufficient for identifying a soil's potential for prehistoric agricultural success. The system of capability classification was designed to indicate the suitability
of a soil for commercially growing most but
not all modem field crops using modem technology. In this typological system, soils are
evaluated on 11 dimensions (including both
physical properties and climatic characteristics)
but assigned a capability class based on their
limitations for yielding crops and resisting deterioration. In fact, a soil is assigned the value
equivalent to the lowest value on all dimensions
for which it is evaluated. Physical property limitations include: depth, surface texture, percentage of coarse fragments, maximum slope,
minimum permeability, minimum available
water capacity, minimum drainage class and .
growing season water table depth, maximum
flood hazard, maximum salinity, and maximum erosion hazard by water and wind. Climatic limitations include the combination of
average annual precipitation and the number
of killing frost-free growing season days
(Hamon 1983).
Capability classes.(Table 2.1) are minimally designated by roman numerals I through
VIII where Class I has the fewest limitations
and is the best for successful modem agriCUlture, whereas Class VIII has the most severe
limitations and is regarded as totally unsuitable for agriculture. In the Montezuma and
Dolores County area, the best nonirrigated soils
used for commercial agriculture are Class III
and IV. No Class I, II, or V nonirrigated soils
exist. Only two soil map units of the many that
18 -
Table 2.1. Soil Conservation Service Soil Capability Class Definitions.
Class
Class I
Class II
Class III
Class IV
Class V
Class VI
Class VII
Class VIII
subclass "e"
subclass "w"
subclass "s"
subclass "c"
Definition
Soils with few limitations that restrict their use.
Soils with moderate limitations that reduce the choice of plants or that require special conservation practices.
Soils with severe limitations that reduce the choice of plants or that require
special conservation practices or both.
Soils with very severe limitations that reduce the choice of plants or that require very careful management or both.
Soils not likely to erode but that have other limitations, impractical to remove,
that limit their use.
Soils with severe limitations that make them generally unsuitable for cultivation.
Soils with very severe limitations that make them unsuitable for cultivation.
Soils with limitations that nearly preclude their use for commercial crop production.
The main limitation is risk of erosion unless close-growing plant cover is
maintained.
The main limitation is water in or on the soil that interferes with plant growth.
The main limitation is shallowness or stoniness.
The main Iimhation is climate, either too cold or too dry.
are classed as VI are ever used for commercial
agriculture. No Class VII or Class VIII soil is
reported as having been used for commercial
agriculture. Often the main limitation or risk to
successful plant growth is indicated in the
capability classification system by adding a
small letter e, w, s, or c to the roman numeral
(see definitions in Table 2.1). By this method,
a soil receives class designations on each of the
11 dimensions above. By invoking "Liebig's
Law of the Minimum," the final land capability class assigned to a soil type is based on the
presence of even a single higher (Le., more
limiting) value.
.
record them for other purposes. Instead, I
modeled prehistoric agricultural potential from
known characteristics of and yield measurements associated with each named soil map
unit in the study area. PDSI modeling was used
to calculate moisture inputs and outputs to a
given soil and the soil moisture retained in the
profile. For each soil, it required historic values
on monthly precipitation and temperature and
unit specific values on the potential waterholding capacity on the uppermost six inches
of soil and the potential water-holding capacity
of the remaining lower soil section.
Whereas the capability classification system
provides broad guidelines for successful modem crop production, which is dependent on
mechanization and often on chemical applications and minimum levels of production, it is
less satisfactory for approximating prehistoric
agricultural potential. The adaptation of prehistoric crops to climatic or soil conditions, differences in the maximum allowable slope
(where mechanized equipment is not an issue),
differences in risk associated with erosion due
to (presumably) less extensive land clearance
and lack of deep plowing, are not taken into
account. Therefore, I decided not to use the
capability classes as major indices of prehistoric agricultural productivity although I did
The resulting PDSI value, calculated on an
annual basis, approximates stored soil moisture
available to a plant and incorporates the effects
of moisture surplus or deficits to a soil over
many months and even years. Thus, it can be
used as an integrative index of potential dryland productivity because it takes into consideration soil depth and texture, available waterholding capacity, moisture input (precipitation
and runoft), and moisture output (temperaturecontrolled evapotranspiration). After calculating PDSI values, potential natural plant productivity values-estimated under a variety of
climatically controlled growing conditionswere linked to actual measurements of nonirrigated agricultural yields. This resulted in a set
of estimated yield values that were quantitative
19 -
and realistic under varying cHmatic conditions.
These were used as an alternative method for
estimating prehistoric potential agricultural capability.
Descriptive and Locational Infonnation
Recorded
Soils data acquired from the SCS included
descriptions of each soil type in the two-county
region and information on the distribution of
each of the mapped soil types in the study
area. Most relevant for building the model
were data on soil depth, available water-holding
capacity, potential natural plant productivity
estimates under varying climatic conditions, associated crop yields (when available), and
whether or not the soil unit was artificially wet
or saline due to irrigation. Other information
included data on nonirrigated and irrigated
land capability classification, elevational range,
and range/woodland affiliation as they were reported in the individual soil descriptions on file
in the Cortez SCS office (Tables 2.2 and 2.3).
Soil scientists in the Cortez office have located these soil map units on aerial photographs enlarged to plainly identify local natural and cultural landmarks. The central portion of each photograph depicts a horizontal
ground distance of one mile on four inches of
photograph (approximately 1: 15,625). Section
comer markers and section numbers have been
identified on each photo.
Final soil maps printed as soil distributions
overlaid on 7.5-minute orthophoto quads are
not yet available from the SCS. Had they been,
the mapped soil units could have been digitized from these as vector data sets and encoded directly into a computer file to be converted to raster data. Since they were not, I was
forced to take a different approach to recording soil distributions. This consisted of manually creating a raster data set, cell by cell, from
the aerial photographs using each photograph
in the study area. In order to do this, several
interrelated decisions had to be made. The first
decision was how to record the soil locational
data, either in vector format (Le., as a multitude
of irregular polygons created by linking lines
and nodes and points) or in raster format (Le.,
as a uniform set of cells created by the regular
gridding of the study area into rows and
columns of a predetermined size). Given that
the GIS systems available to me use the cell as
the basic unit of analysis, and given that the
DEMs were already in the form of a raster-like
array, it seemed logical to record the locations
of the soils in a raster format directly. without
having to convert them for analysis. At this
point I also firmly established the bound-aries
of the study area; decided on a coordinate
system to provide the geographic reference
points; determined the cell size that would be
the basic unit of analysis and gridded the entire
study area into these cellular units; located
each cell on the orthophoto quad and its equivalent position on the aerial photograph; ascertained the dominant soil type within each
cell; and recorded the map unit symbol so that
each map unit code is linked to its Cartesian
coordinates.
Boundary Delimitation
Several factors were taken into account in
determining the size of the study area. The region had to be large enough to allow significant spatial variability in soils and human settlement patterns. However, the effective study
area had to be comprised of areas for which
DEMs and soil mapping and descriptions were
available. A rectangular study area incorporating highlands and lowlands and archaeological areas of potential interest (Mockingbird
Mesa, Sand Canyon Survey LoCality, and the
Dolores Archaeological Project, among others)
satisfied these requirements (Figure 1.1). Its
maximum extent is a rectangle that surrounds
the perimeter of a mosaic of 12 7.5-minute
DEMs arranged in a pattern of four maps
(west-to-east) by three maps (north-to-south).
It incorporates some 1,943.12 km 2 (750.24
mi 2) representing 30 minutes of long-itude
(the four 7.5-minute maps joined west-to-east)
and 22.5 minutes of latitude (the three 7.5minute maps joined north-to-south) de-icted
on the 12 DEMs (Figure 2.1).
Coordinate System
DEMs are georeferenced in both latitude/longitude and UTM systems. Soils are
referenced to township, range, and section
-20-
Table 2.2. Primary Soli Data for 113 Solis In Montezuma and Dolores Counties, Colorado. Part 1.
!'.)
Soil
Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Soli Name and Percent Slopes
Torrifluvents Fluvaquents complex. 0-3%
Sill silty clay loam, 0-3%
Recapture variant sandy loam, 0-6%
Panitchen loam. 0--3%
Panitchen clay loam, 0-3%
Ustic Torrlfluvents, gravelly substratum, 0--3%
Youngston clay loam, 0-6%
Sill silty clay loam, 3-6%
Sill silty clay loam, 6-12%
sm variant silty clay loam, 3-9%
Umbarg-Winn-Tesajo complex, 0-2%
Lillings silt loam, sodic, 1--3%
Lillings silty clay loam, 1--3%
L1l1ings silty clay loam, 3-8010
Poganeab loam, 0-2%
Torrlorthents-Badland complex, 25-100%
Zyme variant channery clay loam, 3-25%
Zyme gravelly clay loam, dry, 3-12%
Zyme-Sm complex. 25-65%
Zyme very channery clay loam dry, 12-65%
lies loam, moist, 3-12%
lies loam, moist, 12-25%
Uzona-Zwlcker-Claysprings complex, 3-12%
SIII·Zyme complex. 3-25%
NI
Classb
6w
3s
6s
3c
3c
6s
6c
3e
4e
4e
4w
6s
6s
6s
6w
8e
6e
6s
7e
7e
4e
6e
6s
6e
I
Soil Map Unita
A1B
A2B (A2A)
A3B
A4
A41
A4A
A4B
A5C
A5D (C9D)
A5Wd
A6A (J1B)
AOAB
AOB
AOC
AOd
BD
C2CE
C2D
C2F (C7E)
C2V
C3CD (S3C)
C3E
C5C
C7D
Class c
6w
3s
6s
3c
3c
4s
3e
3e
4e
4e
4w
6s
4s
4s
6w
8e
6e
6s
7e
7e
4e
6e
6s
6e
Low
Elev (tt)
5,000
5,400
5,000
5,400
5,400
5,400
5,000
5,400
5,400
5,400
5,400
5,400
5,400
5,400
5,400
5,000
5,400
5,400
5,400
5,400
7,100
7,100
5,000
5,400
High
Elev (tt)
7,400
7,400
5,700
7,400
7,400
7,400
5,700
7,400
7,400
7,400
7,400
7,400
7,400
7,400
7,400
7,400
7,400
6,200
7,400
6,200
8,500
8,500
5,700
7,400
25
26
27
28
29
30
31
32
CP
DOCE
GP
H1CD
H1DC
H6D (H6CD)
M1CE
M2C
Shaly Pits
Circleville variant cobbly loam, 3-25%
Gravelly Pits
Sill clay loam, shale substratum, 3-12%
Belmear silty clay, 3-12%
Romberg very stony loam. 6-25%
Falcon gravelly fine sandy loam, stony, 3-25%
Romberg-Cragola. comj)lex,very stony. 6-25%
8e
7s
8e
4e
4e
7s
6e
7s
8e
7s
8e
4e
4e
7s
6e
7s
5,000
7,100
5,000
5,400
7,100
5,400
7,100
5,400
8,000
8,500
8,500
7,400
8,500
7,400
8,500
7,400
Range/Woodland
Classification
semi-wet salt streambank
alkali bottom
alkali bottom
loamy bottom
loamy bottom
loamy bottom
alkali bottom
clayey foothills
clayey foothills
salt meadow
wet meadow
alkali bottom
alkali bottom
alkali bottom
wet meadow
NA
piFlon-juniper
shaly knobs
pinon-juniper
shaly knobs
mountain loam
mountain loam
clayey salt desert
pinon-juniper and clayey
foothills
pit
ponderosa pine
pit
clayey foothills
ponderosa pine
pinon-juniper
pine grassland
pinon-juniper
Table 2.2 (Continued).
Nt
Low
I
Soli Name and Percent Slopes
Classb Class c Elev (tt)
Romberg-Cragola-Rock Outcrop complex, 25-80% 7e
7e
5,400
High
Elev (ft)
7,400
Gladel-Pulpit complex, 3-9%
6s
6s
6,200
7,400
Gladel-Pulpit complex, dry, 3-9%
6s
6s
5,400
6,200
36
M2001
(ROHC1)
M5E (S52)
lIes-Granath Ioams, 2-6%
4c
4c
7,100
8,500
37
M6C
Ormiston very stony-Nortez complex, 3-12%
7s
7s
7,100
8,500
38
M7E
Ormiston variant-Granath complex, 3-12%
7s
7s
7,100
8,500
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
M80 (5S5, 54C)
OAB
PiC
RiC
RiO
R1HB
R1HC
R1HO
R3B
R3B1
R3C (R9B, R5C)
R3C1
R30
R301 (ROO1)
R4C (R4B)
R4C1 (R4B1,
R7C1, ROLC1)
R40
R7C (RC)
R70
R701 (ROH01)
R8B
IIes-Pramlss-Faicon complex, 3-20%
Binton clay loam, 0-3%
Sheppard fine sand, 1-6%
Granath loam, 3-6%
Granath loam, 6-12%
Rlcot loam, 1-3%
Ricot loam, 3-6%
Ricot loam, 6-12%
Cahona very fine sandy loam, 1-3%
Cahona very fine sandy loam, dry, 1-3%
Cahona very fine sandy loam, 3-6%
Cahona very fine sandy loam, dry, 3-6%
Cahona very fine sandy loam, 6-12%
Cahona very fine sandy loam, dry, 6-12%
Cahona-Sharps-Witt complex, 2-6%
Cahona-Sharps-Witt complex, dry, 2-6%
6e
7s
7s
4c
4e
4c
4c
4e
3c
4c
3e
4c
4e
4e
3e
4c
6e
4s
4s
4c
4e
4c
4c
4e
3c
2e
3e
3e
4e
4e
3e
3e
7,100
5,000
5,000
7,100
7,100
7,100
7,100
7,100
6,200
5,400
6,200
5,400
6,200
5,400
6,200
5,400
8,500
5,700
5,700
8,500
8,500
8,500
8,500
8,500
7,400
6,200
7,400
6,200
7,400
6,200
7,400
6,200
Range/Woodland
Classification
pinon-juniper and rock
outcrop
pinon-juniper and loamy
foothills
pinon-juniper and semidesert loam
mountain clay and mountain
loam
ponderosa pine and pine
grassland
pine grassland and mountain
loam
ponderosa pine
alkali bottom
desert sand
mountain loam
mountain loam
mountain loam
mountain loam
mountain loam
loamy foothills
semi-desert loam
loamy foothills
semi-desert loam
loamy foothills
semi-desert loam
loamy foothills
semi-desert loam
Sharps-Cahona complex, 6-12%
Sharps-Pulpit loams, 2-6%
Sharps-Pulpit loams, 6-12%
Sharps-Pulpit loams, dry, 6-12%
Recapture fine sandy loam, 0-6%
4e
3e
4e
4e
6s
4e
3e
4e
4e
6s
6,200
6,200
6,200
5,400
5,000
7,400
7,400
7,400
6,200
5,700
loamy foothills
loamy foothills
loamy foothills
semi-desert loam
alkali flat
Soil
Code
33
34
35
N
N
55
56
57
58
59
Soil Map Unita
M2CE (H5E,
H50E)
M200
Table 2.2 (Continued).
Soli
Code
60
61
62
63
64
65
66
67
68
69
70
71
72
N
w
Soil Map Un/ta
Nt
Classb
4e
8e
3c
4c
3e
4c
4e
3e
4e
3e
4e
4e
6c
I
Soli Name and Percent Slopes
Cahona-Pulpit complex, 3-9%
Rock Outcrop
Witt loam, 1-3%
Witt loam, dry, 1-3%
Witt loam, 3-6%
Witt loam, dry, 3-6%
Witt loam, 6-12%
Pulpit loam, 3-6%
Pulpit loam, 6-12%
Sharps loam, 3-6%
Sharps loam, 6-12%
Sharps loam, dry, 6-12%
Mack-Redlands fine sandy loams, 0-6%
Class c
4e
8e
3c
2e
3e
3e
4e
3e
4e
3e
4e
4e
3e
Low
Elev (ft)
6,200
NA
6,200
5,400
6,200
5,400
6,200
6,200
6,200
6,200
6,200
5,400
5,000
High
Elev (ft)
7,400
NA
7,400
6,200
7,400
6,200
7,400
7,400
7,400
7,400
7,400
6,200
5,700
Range/Woodland
Classification
loamy foothills
rock outcrop
loamy foothills
semi-desert loam
loamy foothills
semi-desert loam
loamy foothills
loamy foothills
loamy foothills
loamy foothills
loamy foothills
semi-desert loam
alkali flat
Acree loam, sandstone substratum, 0-6%
Acree loam, 2-6%
lIes-Granath Ioams, 6-12%
4c
4c
4e
4c
4c
4e
7,100
7,100
7,100
8,500
8,500
8,500
4e
7s
4e
7s
7,100
7,100
8,500
8,500
4c
4c
7,100
8,500
mountain loam
mountain loam
mountain clay and mountain
loam
mountain loam
mountain clay and mountain
loam
mountain clay and mountain
loam
mountain clay
pine grassland and mountain
loam
mountain clay and mountain
loam
loamy foothills
pine grassland
NA
NA
loamy foothills
loamy foothills
loamy foothills
73
74
75
RHC
RL
ROB
ROB1
ROC
ROC1
ROD
ROHC
ROHD
ROLC (C4C)
ROLD (C4D)
ROLD1 (R4D1)
S1C (S2B, S2C,
MD,DD,MOS)
S5(S5C)
S51
S53
76
77
S5D
S6(S60,S6C)
78
S6C.
Acree loam, 6-12%
Ormiston-very stony Acree, sandstone substratum
complex, 1-12%
Pramlss-Acree Ioams, 3-9%
79
80
S6E
S7
Pramlss very cobbly loam, very stony, 6-25%
Nortez-Acree sandstone substratum loams, 0-6%
7s
4c
7s
4c
7,100
7,100
8,500
8,500
81
S8C
lIes-Pramiss Granath Ioams, 2-6%
4c
4c
7,100
8,500
82
83
84
85
86
87
88
T1B (T1C)
T2C
T3E
T4B
T6
T6B
T6C
Weber loam, 1-3%
Ormiston variant very cobbly loam, 0-9%
Torrlorthents,very stony, 12-65%
Purcella loam, 0-3%
Collbran clay loam, 3-6%
Collbran-Collbran variant complex, 0-2%
Collbran-Collbran variant 」ッューャ・クLセᄚ
____
3c
7s
7e
6s
3e
3c
3e
3c
7s
7e
6s
3e
3c
3e
5,400
7,100
5,000
5,400
5,400
5,400
5,400
7,400
8,500
7,400
7,400
7,400
7,400
7,400
Table 2.2 (Concluded).
Soil
Code
89
90
91
92
93
94
95
96
97
IV
セ
98
99
100
101
102
103
104
105
106
107
Soil Map Unita
T6D
V1B
V1C
V1D
V2B (W1B)
V2C (V3C)
V2S (VOAB)
V2Vd (W1BB)
v2wd (W1CB,
AOBB, VOBB)
V4C
VOB
VOC
VOCB
VOCC (VOBC)
WOB
WOBBd
WOC
XC6D (C5E)
XC6E (XM3F)
108
109
110
XH2D
XH2E
XH3E (M4E)
111
XM1F
Fughes loam, 2-9%
Yarts fine sandy loam, 1-3%
Yarts fine sandy loam, 3-6%
Yarts clay loam, 1-3%
Yarts clay loam, 3-6%
Ackmen loam, 1-3%
IraJ< loam, G-3%
Ackmen loam, 3-6%
Claysprings clay loam,very stony, 12-65%
Typic Torriorthents-bouldery Rock Outcrop
complex, 12-80%
Sheek-Archuleta complex,very stony, 6-25%
Sheek-Archuleta Rock Outcrop complex, 25-80%
Sheek-Archuleta-Rock Outcrop complex, 25-80%,
north slopes
Falcon-Rock Outcrop complex, 25-65%
112
XM3D
Farb-Rock Outcrop complex, 3-12%
Soli Name and Percent Slopes
Collbran clay loam, 6-12%
Hesperus loam, 1-3%
Hesperus loam, 3-6%
Hesperus loam, 6-12%
Mldm clay loam, 1-3%
Miklm loam, 3-6%
Miklm clay loam, sodle, G-3%
Miklm clay loam, slightly wet, G-3%
Midm variant clay loam, 0-3%
NI
Class b
4e
4c
4c
4e
4s
4s
6s
4s
7s
I
Class c
4e
4c
4c
4e
3s
3e
6s
3s
7s
Low
Elev (ft)
5,400
7,100
7,100
7,100
5,400
5,400
5,400
5,400
5,400
High
Elev (ft)
7,400
8,500
8,500
8,500
7,400
7,400
7,400
7,400
7,400
4e
3c
3e
3c
3e
3c
3w
3e
7e
7e
4e
3c
3e
3c
3e
3c
3w
3c
7e
7e
7,100
5,400
5,400
5,400
5,400
5,400
5,400
5,400
5,000
5,000
8,500
7,400
7,400
7,400
7,400
7,400
7,400
7,400
5,700
5,700
7s
7e
7e
7s
7e
7e
7,100
7,100
7,100
7e
7e
7,100
7s
7s
5,000
7e
5,400
Ustic Torriorthents-Gullled Land complex, 25-60% 7e
Z
113
aEarlier, alternative soil map unit symbols, now officially subsumed under a final soil map unit designation, are shown
bNonirrigated land SCS Capability Class; Roman numerals converted to Arabic numbers.
Clrrigated land SCS Capability Class; Roman numerals converted to Arabic numbers.
dArtificial1y wet and possibly saline from irrigation water.
Range/Woodland
Classification
loamy foothills
loamy pari<
loamy pari<
brushy loam
alkali flat
alkali flat
alkali flat
NA
salt meadow
loamy pari<
semi-desert loam
semi-desert loam
semi-desert loam
semi-desert loam
loamy bottom
wet meadow
loamy bottom
salt desert breaks
salt desert breaks and rock
outcrop
8,500
ponderosa and Douglas-fir
8,500
ponderosa and Douglas-fir
8,500
ponderosa pine, Douglas-fir,
and rock outcrop
8,500
ponderosa pine and rock
outcrop
5,700
shallow desert and rock
outcrop
7,400
gullied land
in parentheses.
Table 2.3. Primary Soil Data for 113 Soils In Montezuma and Dolores Counties, Colorado. Part 2.
Soli
Code
1
Soil Map
Unit
A1B
2
A2B
Soil Unit and Percent a
Torrifluvents (62.5)
Fluvaquents (37.5)
Sill (100)
3
A3B
Recapture (100)
4
A4
Panitchen (100)
5
A41
Panitchen (100)
6
A4A
Ustic Torrifluvents (100)
7
A4B
Youngston (100)
8
A5C
Sill (100)
9
A5D·
Sill (100)
10
A5We
Sill Variant (100)
N
UI
SoJI Unit
Depth (in)
NA
NA
0-6
6-60
0-6
6-13
13-17
17-38
38-60
0-3
3-60
0-3
3-60
0-3
3-11
11-27
27-60
0-10
10-43
43-60
0-6
6-60
0-6
6-60
0-4
4-32
32-60
AWC
(inlln)b
NA
NA
.18-.20
.16-.18
.08-.10
.15-.17
.08-.10
.12-.14
.07-.09
.15-.18
.14-.17
.14-.17
.13-.16
.05-.06
.13-.15
.04-.08
.08-.10
.15-.18
.15-.18
.15-.18
.18-.20
.16-.18
.18-.20
.16-.18
.15-.17
.13-.16
.13-.15
NI Beand
(Ib/ac)
Favorable
NPpc (Ib/ac)
2,000
2,500
875
Normal NPpc
(Ib/ac)
1,200
2,000
625
Unfavorable
NPpc (Ib/ac)
500
1,500
450
900
650
450
1,200
900
700
1,200
900
700
950
800
600
900
700
500
1,200
900
600
275
1,200
900
600
250
2,500
(1,200)
2,000
(900)
1,500
(600)
Table 2.3 (Continued).
Soil
Code
11
Soil Map
Unn
A6A
Soil Unn and Percent a
Umbarg (40)
Winn (35)
Tesajo (25)
to)
12
AOAB
L1l1ings (100)
13
AOB
L1l1ings (100)
14
AOC
L1l1ings (100)
15
AOe
Poganeab (100)
16
BO
17
C2CE
Torriorthents (62.5)
Badlands (37.5)
Zyme (100)
18
C20
Zyme (100)
19
C2F
Zyme (67.5)
0\
Sill (32.5)
20
C2V
Zyme (100)
21
C3CO
lies (100)
Soil Unn
.Depth (in)
0-2
2-5
5-12
12-42
42-60
0-4
4-31
31-60
0-3
3-36
36-60
0-2
2-60
0-2
2-60
0-2
2-60
0-2
2-60
0-60
0-60
0-1
1-19
0-1
1-19
0-1
1-19
0-6
6-60
0-1
1-19
0-2
2-22
22-60
AWC
(inlin)b
.15-.18
.17-.20
.17-.20
.15-.18
.08-.11
.18-.20
.18-.20
.07-.10
.11-.13
.04-.06
.04-.06
.18-.20
.18-.20
.18-.20
.18-.20
.18-.20
.18-.20
.15-.18
.17-.20
.00-.00
.00-.00
.08-.10
.15-.17
.12-.14
.15-.17
.08-.10
.15-.17
.18-.20
.16-.18
.09-.10
.15-.17
.08-.18
.19-.21
.14-.16
Favorable
NPpc (Ib/ac)
3,000
Normal NPpc
(Ib/ac)
2,600
Unfavorable
NPpc (Ib/ac)
2,300
3,000
2,600
2,300
2,500
2,000
1,500
775
600
450
1,000
700
500
1,000
700
500
Nt Beand
(Ib/ac)
2,000
(667)
200
4,000
(1,333)
500
0
600
3,000
(1,000)
350
0
400
500
300
200
600
400
300
700
500
300
500
300
200
1,800
1,500
1,200
a
300
400
Table 2.3 (Continued).
Soil
Code
22
Soil Map
Unit
C3E
Soil Unit and Percent a
lies (100)
23
C5C
Uzona (40)
Zwicker (35)
Claysprings (25)
24
C7D
Sill (52.5)
Zyme (47.5)
N
-...I
25
26
CP
DOCE
Pits (100)
Circleville Variant (100)
27
28
GP
H1CD
sm (100)
29
H1DC
Belmear (100)
30
H6D
Romberg (100)
31
M1CE
Falcon (100)
32
M2C
Romberg (52.5)
Pits (100)
Cragola (47.5)
33
M2CE
Romberg (40)
Cragola (35)
Rock Outcrop (25)
Soil Unit
Depth (in)
0-2
2-22
22-60
0-1
1-45
45-60
0-1
1-4
4-32
0-3
3-18
0-6
6-60
0-1
1-19
0-60
0-9
9-24
0-6
0-6
6-42
0-10
10-30
0-2
2-20
0-4
4-13
0-2
2-60
0-2
2-18
0-2
2-60
0-2
2-18
0-60
AWC
(inlin)b
.16-.18
.19-.21
.14-.16
.16-.18
.15-.17
.17-.19
.17-.21
.17-.21
.14-.21
.10-.12
.16-.18
.18-.20
.16-.18
.08-.10
.15-.17
.00-.00
.10-.12
.10-.12
.00-.00
.17-.19
.14-.16
.14-.16
.12-.15
.12-.14
.07-.08
.07-.10
.06-.08
.12-.14
.07-.08
.06-.09
.07-.10
.12-.14
.07-.08
.06-.09
.07-.10
.00-.00
Favorable
NPpc (Ib/ac)
1,800
Normal NPpc
(Iblac)
1,500
Unfavorable
NPpc (Ib/ac)
1,200
625
450
250
500
350
200
375
250
150
1,200
900
600
600
400
300
0
700
0
500
0
300
0
1,100
0
800
0
500
900
500
300
500
350
200
900
650
550
500
350
200
400
200
100
500
350
200
400
200
100
0
0
0
NI Beand
(Ib/ac)
250
Table 2.3 (Continued).
Soil
Code
34
Soil Map
Unit
M2DD
Soil Unit and PercentS
Gladel (55)
Pulpit (45)
35
M2DD1
Gladel (55)
Pulpit (45)
36
M5E
lies (67.5)
Granath (32.5)
N
00
37
M6C
Ormiston (52.5)
Nortez (47.5)
38
M7E
Ormiston Variant (57.5)
Granath (42.5)
39
M8D
lies (40)
Pramlss (35)
Falcon (25)
Soil Unit
Depth (In)
0-5
5-15
0-10
10-20
20-36
0-5
5-15
0-10
10-20
20-36
0-2
2-22
22-60
0-13
13-60
60-80
0-3
3-7
7-32
32-44
0-2
2-6
6-31
0-9
9-13
13-60
0-13
13-60
60-80
0-2
2-22
22-60
0-3
3-31
0-4
4-13
AWC
(inJin)b
.11-.13
.09-.12
.16-.18
.19-.21
.16-.18
.11-.13
.09-.12
.16-.18
.19-.21
.16-.18
.16-.18
.19-.21
.14-.16
.18-.20
.18-.20
.15-.19
.06-.11
.07-.13
.05-.11
.11-.17
.15-.18
.17-.20
.15-.18
.06-.09
.09-.11
.07-.10
.18-.20
.18-.20
.15-.19
.16-.18
.19-.21
.14-.16
.07-.12
.16-.18
.07-.10
.06-.08
Favorable
NPPC (Ib/ac)
500
Normal NPpc
(Ib/ac)
350
Unfavorable
NPpc (Ib/ac)
200
NIBeand
(lb/ac)
300
1,125
800
600
500
350
200
750
600
450
250
1,300
1,100
800
450
1,700
1,400
1,100
450
1,300
1,000
800
1,400
1,200
900
1,200
900
750
1,700
1,400
1,100
1,500
1,000
800
1,100
800
550
900
650
550
Table 2.3 (Continued).
Soil
Code
40
Soil Map
Unit
OAB
Soil Unit and PercentS
Blnton (100)
41
P1C
Sheppard (100)
42
R1C
Granath (100)
43
R1D
Granath (100)
44
R1HB
Rlcot (100)
45
R1HC
Rlcot (100)
46
R1HD
Rlcot (100)
47
R3B·
Cahona (100)
48
R3B1
Cahona (100)
49
R3C
Cahona (100)
50
R3C1
Cahona (100)
IV
\0
Soli Unit
Depth (in)
0-9
9-60
0-7
7-60
0-13
13-60
60-80
0-13
13-60
60-80
0-8
8-16
16-34
34-60
0-8
8-16
16-34
34-60
0-8
8-16
16-34
34-60
0-11
11-24
24-60
0-11
11-24
24-60
0-11
11-24
24-60
0-11
11-24
24-60
AWC
(Inlin)b
.14-.17
.12-.15
.05-.07
.06-.08
.18-.20
.18-.20
.15-.19
.18-.20
.18-.20
.15-.19
.16-.18
.16-.18
.12-.15
.10-.14
.16-.18
.16-.18
.12-.15
.10-.14
.16-.18
.16-.18
.12-.15
.10-.14
.14-.16
.15-.17
.13-.16
.14-.16
.15-.17
.13-.16
.14-.16
.15-.17
.13-.16
.14-.16
.15-.17
.13-.16
NI Beand
(Ib/ac)
Favorable
NPpc (Ib/ac)
800
Normal NPpc
(Ib/ac)
600
Unfavorable
NPpc (Ib/ac)
400
900
550
350
1,700
1,400
1,100
450
1,700
1,400
1,100
400
1,800
1,500
1,200
1,800
1,500
1,200
1,800
1,500
1,200
1,500
1,000
800
350
800
700
400
275
1,500
1,000
800
350
800
700
400
275
Table 2.3 (Continued).
Soil
Code
51
Soil Map
Unit
R3D
Soil Unit and
Cahona (100)
52
R3D1
Cahona (100)
53
R4C
Cahona (40)
Percent 8
Sharps (35)
Witt (25)
w
54
R4C1
Cahona (40)
0
Sharps (35)
Witt (25)
55
R4D
Sharps (52.5)
Cahona (47.5)
Soli Unit
Depth (in)
0-11
11-24
24-60
0-11
11-24
24-60
0-11
11-24
24-60
0-9
9-19
19-30
0-7
7-48
48-60
0-11
11-24
24-60
0-9
9-19
19-30
0-7
7-48
48-60
0-9
9-19
19-30
0-11
11-24
24-60
AWC
(inJin)b
.14-.16
.15-.17
.13-.16
.14-.16
.15-.17
.13-.16
.14-.16
.15-.17
.13-.16
.16-.18
.15-.17
.13-.15
.15-.18
.18-.21
.16-.19
.14-.16
.15-.17
.13-.16
.16-.18
.15-.17
.13-.15
.15-.18
.18-.21
.16-.19
.16-.18
.15-.17
.13-.15
.14-.16
.15-.17
.13-.16
NI Beand
(Ib/ac)
300
Favorable
NPpc (Ib/ac)
1,500
Normal NPpc
(Ib/ac)
1,000
Unfavorable
NPpe (Ib/ac)
800
800
700
400
250
1,500
1,000
800
350
1,200
900
700
350
1,500
1,200
800
450
800
700
400
275
750
600
450
250
1,200
750
500
300
1,200
900
700
300
1,500
1,000
800
300
Table 2.3 (Continued).
Soil
Code
56
Soil Map
Unit
R7C
Soli Unit and Percent a
Sharps (52.5)
Pulpit (47.5)
57
R7D
Sharps (52.5)
Pulpit (47.5)
58
R7D1
Sharps (52.5)
Pulpit (47.5)
....
VJ
59
R8B
Recapture (100)
60
RHC
Cahona (57.5)
Pulpit (42.5)
61
62
RL
ROB
Rock Outcrop (100)
Witt (100)
63
ROB1
Witt (100)
Soli Unit
Depth (in)
0-9
9-19
19-30
0-10
10-20
20-36
0-9
9-19
19-30
0-10
10-20
20-36
0-9
9-19
19-30
0-10
10-20
20-36
0-7
7-26
26-60
0-11
11-24
24-60
0-7
7-25
25-36
0-60
0-7
7-48
48-60
0-7
7-48
48-60
AWC
(inlin)b
.16-.18
.15-.17
.13-.15
.16-.18
.19-.21
.16-.18
.16-.18
.15-.17
.13-.15
.16-.18
.19-.21
.16-.18
.16-.18
.15-.17
.13-.15
.16-.18
.19-.21
.16-.18
.13-.15
.15-.17
.14-.16
.14-.16
.15-.17
.13-.16
.16-.18
.19-.21
.16-.18
.00-.00
.15-.18
.18-.21
.16-.19
.15-.18
.18-.21
.16-.19
NI Beand
(Ib/ac)
350
Normal NPpc
(Iblac)
900
Unfavorable
NPpc (Ib/ac)
700
1,125
800
600
350
1,200
900
700
300
1,125
800
600
300
750
600
450
250
750
600
450
250
900
650
450
1,500
1,000
800
350
1,125
800
600
350
0
1,500
0
1,200
0
800
450
1,200
750
500
300
Favorable
NPpc (Ib/ac)
1,200
セ
Table 2.3 (Continued).
Soli
Code
64
Soil Map
Unit
ROC
Soil Unit and PercentS
Witt (100)
65
ROC1
Witt (100)
66
ROD
Witt (100)
67
ROHC
Pulpit (100)
68
ROHD
Pulpit (100)
69
ROLC
Sharps (100)
70
ROLD
Sharps (100)
71
ROLD1
Sharps (100)
72
S1C
Mack (52.5)
N
Redlands (47.4)
73
S5
Acree (100)
74
S51
Acree (100)
Soil Unit
Depth (In)
0-7
7-48
48-60
0-7
7-48
48-60
0-7
7-48
48-60
0-10
10-20
20-36
0-10
10-20
20-36
0-9
9-19
19-30
0-9
9-19
19-30
0-9
9-19
19-30
0-13
13-60
0-7
7-46
46-60
0-10
10-42
42-50
0-10
10-42
42-60
AWC
(inJln)b
.15-.18
.18-.21
.16-.19
.15-.18
.18-.21
.16-.19
.15-.18
.18-.21
.16-.19
.16-.18
.19-.21
.16-.18
.16-.18
.19-.21
.16-.18
.16-.18
.15-.17
.13-.15
.16-.18
.15-.17
.13-.15
.16-.18
.15-.17
.13-.15
.13-.18
.13-.16
.13-.16
.16-.18
.15-.17
.16-.18
.15-.21
.16-.18
.17-.21
.13-.16
.13-.16
Favorable
NPpc (Ib/ac)
1,500
NI Beand
(Ib/ac)
450
Normal NPpc
(Iblac)
1,200
Unfavorable
NPpc (Ib/ac)
800
1,200
750
500
300
1,500
1,200
800
400
1,125
800
600
350
1,125
800
600
300
1,200
900
700
350
1,200
900
700
300
750
600
450
250
800
700
500
800
700
500
1,800
1,500
1,200
450
2,000
1,600
1,400
450
Table 2.3 (Continued).
Soli
Code
75
Soil Map
Unit
S53
Soli Unit and Percent a
lies (67.5)
Granath (32.5)
76
S5D
Acree (100)
77
S6
Ormiston (57.5)
セ
Acree (42.5)
78
S6C
Pramlss (55)
Acree (45)
79
S6E
Pramiss(100)
80
S7
Nortez(52.5)
Acree (47.5)
Soil Unit
Depth (in)
0-2
2-22
22-60
0-13
13-60
60-80
0-10
10-42
42-60
0-3
3-7
7-32
32-44
0-10
10-42
42-50
0-3
3-31
0-10
10-42
42-60
0-3
3-31
0-1
1-9
9-25
0-10
10-42
42-50
AWC
(inlin)b
.16-.18
.19-.21
.14-.16
.18-.20
.18-.20
.15-.19
.17-.21
.13-.16
.13-.16
.06-.11
.07-.13
.05-.11
.11-.17
.16-.18
.15-.21
.16-.18
.15-.18
.15-.18
.17-.21
.13-.16
.13-.16
.07-.12
.16-.18
.15-.18
.17-.20
.15-.18
.16-.18
.15-.21
.16-.18
Favorable
NPPC (Ib/ac)
1,300
Nonnal NPpc
(Iblac)
1,100
Unfavorable-------NIB-eand
NPpc (Ib/ac)
(Ib/ac)
800
400
1,700
1,400
1,100
400
2,000
1,600
1,400
400
1,200
900
600
1,800
1,500
1,200
1,100
900
700
400
2,000
1,600
1,400
450
1,100
800
550
1,400
1,200
900
400
1,800
1,500
1,200
450
Table 2.3 (Continued).
Soil
Code
81
Soil Map
Unit
S8C
Soil Unit and Percent a
lies (40)
Pramiss (35)
Granath (25)
82
11B
Weber (100)
83
T2C
Ormiston Variant (100)
84
T3E
Torrlorthents (100)
85
T4B
Purcella (100)
86
T6
Collebran (100)
87
T6B
Collebran (52.5)
Yo)
.f>.
Collebran Variant (47.5)
Soil Unit
Depth (in)
0-2
2-22
22-60
0-3
3-31
0-13
13-60
60-80
0-8
8-32
32-60
0-9
9-13
13-60
0-3
3-7
7-16
0-4
4-11
11-41
41-60
0-10
10-40
40-60
0-10
10-40
40-60
0-2
2-8
8-45
45-60
AWC
(inlin)b
.16-.18
.19-.21
.14-.16
.15-.18
.15-.18
.18-.20
.18-.20
.15-.19
.16-.19
.16-.20
.03-.06
.06-.09
.09-.11
.07-.10
.04-.06
.07-.08
.07-.08
.15-.18
.12-.15
.05-.08
.03-.05
.18-.20
.16-.19
.13-.15
.18-.20
.16-.19
.13-.15
.15-.18
.18-.21
.15-.18
.07-.10
NI Beand
(Ib/ac)
450
Normal NPpc
(Ib/ac)
1,100
Unfavorable
NPpc (Ib/ac)
800
1,100
900
700
400
1,700
1,400
1,100
450
1,500
1,200
800
1,200
900
750
500
350
200
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Favorable
NPpc (Ib/ac)
1,300
Table 2.3 (Continued).
Soil
Code
88
SoitMap
Unit
T6C
Soil Unit and Pereent a
Collebran (52.5)
Collebran Variant (47.5)
w
VI
89
T6D
Collebran (100)
90
V1B
Hesperus (100)
91
V1C
Hesperus (100)
92
V1D
Hesperus (100)
93
V2B
Miklm (100)
94
V2C
Mikim (100)
95
V2S
Mikim (100)
96
V2Ve
Mikim (100)
Soil Unit
Depth (in)
0-10
10-40
40-60
0-2
2-8
8-45
45-60
0-10
10-40
40-60
0-11
11-44
44-60
0-11
11-44
44-60
0-11
11-44
44-60
0-3
3-15
15-32
32-60
0-3
3-15
15-32
32-60
0-3
3-15
15-32
32-60
0-3
3-15
15-32
32-60
AWC
(inlin)b
.18-.20
.16-.19
.13-.15
.15-.18
.18-.21
.15-.18
.07-.10
.18-.20
.16-.19
.13-.15
.16-.19
.16-.19
.16-.19
.16-.19
.16-.19
.16-.19
.16-.19
.16-.19
.16-.19
.14-.17
.14-.17
.14-.17
.14-.17
.12-.15
.14-.17
.14-.17
.14-.17
.12-.15
.12-.15
.12-.15
.12-.15
.15-.17
.15-.17
.15-.17
.15-.17
NI Beand
(Ib/ae)
Favorable
NPpc (Iblae)
NA
Normal NPpc
(Ib/ae)
NA
Unfavorable
NPpc (Ib/ae)
NA
NA
NA
NA
NA
NA
NA
2,500
1,800
1,000
500
2,500
1,800
1,000
500
3,000
2,000
1,500
450
875
625
450
875
625
450
675
475
350
2,000
(875)
1,600
(625)
1,200
(450)
Table 2.3 (Continued).
Vo)
Soil
Code
97
Soil Map
Unit
V2We
Soil Unit and PercentS
Miklm Variant (100)
98
V4C
Fughes (100)
99
VOB
Yarts (100)
100
vac
Yarts (100)
101
VOCB
Yarts (100)
102
VOCC
Yarts (100)
103
WOB
Ackmen (100)
104
WOBBe
IraJ< (100)
105
woe
Ackmen (100)
106
XC6D
Claysprlngs (100)
107
XC6E
Typic Torrlorthents (67.5)
0\
Rock Outcrop (32.5)
108
XH2D
Sheek (57.5)
Archuleta (42.5)
Soil Unit
Depth (In)
0-8
8-35
35-60
0-4
4-50
50-60
0-9
9-13
13-60
0-9
9-13
13-60
0-9
9-13
13-60
0-9
9-13
13-60
0-20
20-60
0-8
8-60
0-20
20-60
0-3
3-18
0-3
3-7
7-16
0-60
0-4
4-18
18-60
0-8
8-17
AWC
(inlin)b
.15-.17
.09-.15
.09-.15
.16-.18
.15-.17
.14-.16
.13-.15
.11-.12
.11-.12
.13-.15
.11-.12
.11-.12
.18-.20
.11-.12
.11-.12
.18-.20
.11-.12
.11-.12
.17-.19
.17-.19
.14-.17
.15-.18
.17-.19
.17-.19
.10-.12
.16-.18
.04-.06
.07-.09
.07-.09
.00-.00
.06-.08
.11-.14
.11-.14
.08-.10
.16-.18
Favorable
NPpc (Ib/ac)
1,800
(675)
Normal NPpc
(Ib/ac)
1,600
(475)
Unfavorable
NPpc (Ib/ac)
1,200
(350)
NI Beand
(Ib/ac)
2,300
1,800
1,100
900
700
500
900
700
500
900
700
500
900
700
500
1,000
800
600
500
3,000
(1,000)
1,000
2,500
(800)
800
2,000
(600)
600
250
500
350
200
500
350
200
0
500
0
400
0
200
500
350
200
500
500
Table 2.3 (Concluded).
Soil
Code
109
Soli Map
Unit
XH2E
Soil Unit and Percent a
Sheek (40)
Archuleta (30)
110
XH3E
Rock Outcrop (20)
Sheek (45)
Archuleta (30)
w
111
XM1F
Rock Outcrop (25)
Falcon (62.5)
112
XM3D
Rock Outcrop (37.5)
Farb (62.5)
113
z
....,J
Rock Outcrop (37.5)
Ustlc Toniorthents (52.5)
Gullied Land (47.5)
Soil Unit
Depth (in)
04
418
1860
08
817
060
04
418
1860
08
817
060
04
413
060
03
316
060
05
560
060
AWC
(in/in)b
.06.08
.11.14
.11.14
.08.10
.16.18
.00.00
.06.08
.11.14
.11.14
.08.10
.16.18
.00.00
.07.10
.06.08
.00.00
.08.13
.06.13
.00.00
.10.13
.06.18
.00.00
Favorable
NPpc (Ib/ac)
500
Normal NPpc
(Ib/ac)
400
Unfavorable
NPpc (Ib/ac)
200
500
350
200
0
700
0
550
0
300
700
500
300
0
900
0
650
0
550
0
550
0
400
0
275
0
500
0
300
0
200
0
0
0
NI Beand
(Ib/ac)
aThis value is used as a weighted multiplier in computing total available water capacity per soil profile. It is derived from estimates of soil type composition provided by
the SCS. Each soil type Is actually composed of the major components listed plus some small fraction, usually 20 percent or less, of other named and unnamed soil
types. The multiplier provided above is the sum of the actual percentage of that component, as estimated by the SCS, plus a portion of the remaining percent
representing the unspecified soils. For example, Soil Type A6A Is comprised of three major components. SCS soil descriptions indicate that approximately 35% of this
"soil type" is Umbarg, 30% is Winn, 20% is Tesajo, and 15% are other less prevalent soil types. This remaining 15% is divided equally among the dominant soil types to
croduce the weighting of 40:35:25 used to calculate soil moisture for this study.
Available Water Capacity per subsoil unit given in inches of water per inch of soil (SCS 6/21/88).
CPotential Natural Plant Productivity estimate provided for a soil under favorable, normal, and unfavorable growing season conditions (SCS 3/3/89).
dAverage yield per given soil type for pinto beans grown on nonirrigated soils in Montezuma and Dolores Counties (SCS 3/3/89).
eArtificially wet and possibly saline from irrigation water. Values reduced to reflect this; guidelines suggested by SCS Soil Scientist, Alan Price (personal communication,
March 1989).
lI
NW Corner
e
ll
":'
(
101. 37°37'0"
long. 109°0'0"
4165886 m N
676499 m E
11
RUIN
CANYON
(CEDAR 3 SW)
All
NEGRO
CANYON
(MOQUI NW)
I
YELLOWJACKET
TRIMBLE
POINT
WOODS
CANYON
ARRIOLA
DOLORES
WEST
37° 15'0"
long. 109°0'0"
4166944 m N
720630 m E
II
11
0
11
S
I1
CORTEZ
"\;)
SW Corner V
101
37°37'30"
long 108°30'0"
101.
PLEASANT
VIEW
BOWDISH
BATT LEROCK
MUD CREEK
CANYON
(MOQUI SE)
(MOQUI SW)
NE Corner
SE Corner
101
37° 15' 0"
long. 108°30'0"
4125332 m N
721734 m E
4124277 m N
677382 m E
Figure 2.1. Schematic diagram illustrating the arrangement of OEMs used in this study.
boundary lines depicted on the available aerial
photographs. Whereas the Public Land Survey
System of township, range, section, quarter
section and 40acre parcels proves to be useful
in legal matters of ownership and property decription, it is not useful for consistent georeferencing because of its low resolution, often deformed grid, and inconsistent depiction on
standard topographic or orthophoto quads.
Given that both latirudenongitude and UTM
systems work well to accurately locate grid
cells, it was decided that UTM would be preferable because it is a plane grid rather than a
spherical grid, and the other forms of descriptive measures--distances and elevations-were
in meters rather than in degree, minutes, and
seconds. Thus, the UTM coordinate system was
selected for georeferencing.
Cell Size
Raster-based spatial analysis systems demand that cells (or "pixels", short for picture
elements) be rectangular units described by
their location within columns and rows in a
numeric array. Although they need not be,
these cells are generally square. The entire area
contained within the boundary of the cell is
assigned a single value, regardless of the size of
the area being represented. Thus, the issue of
cell size is important (Wehde 1982). If the cell
is too large, it underrepresents the true variability of the phenomenon being considered
and reduces the accuracy of the model. If the
cell is too small, however, it may impede recording or storing information on an area
large enough for meaningful analysis. Given
these competing considerations, the precision
and accuracy of the soil mapping, and the
general goals of the research, a cell of 200 x
200 m (4 har-believed to be the smallest cell
possible given the large site of study area-was
selected as the standard unit of observation.
This cell size would require 48,578 observations (214 rows by 227 columns of cells) to
cover the study area.
38 -
Gridding the Study Area and Locating Cells
Accurately locating the individual 200 x
200 m cells on photographs using the UTM
system was the next step. This was facilitated
by hypothetically gridding the entire study
area into 2 x 2 kID blocks and identifying each
by the UTM coordinate of its northwest comer
(e.g., Block I = UTM Zone 12,4,166,000 m N,
676,000 m E; Block 2 = 4,166,000 m N,
678,000 m E; etc.). Each block was subdivided
into 10 rows and 10 columns depicting 100
200 x 200 m cells.
I therefore created 2 x 2 kID grids on large,
clear mylar sheets that completely overlaid
each 7.5minute orthophoto quad. The first 2
x 2 kIn block that encompassed the northwestenunost areas on the northwestenunost orthophoto quad was placed so that its northwestern comer fell on U.S.G.S-ploned UTM tick
marks, which were easy to track across the entire study area. A second, smaller mylar grid
scaled to fit the 2 x 2 kID blocks on the orthophoto quad was subdivided into 10 rows and
10 columns representing the 100 200 x 200 m
cells. The exact location of each cell was then
easily determined by reference to the coordinates of its northwest comer in relation to that
of the larger 2 x 2 kIn block. A third, small
mylar sheet was gridded into the 100 cell format but drawn at the scale of the aerial photograph so that the area bounded by one 200 x
200 m cell on the orthophoto quad was equivalent to the area bounded by a cell on the photograph. In this way, 502 2 x 2 kIn blocks were
examined across the full extent of the study
area, representing a maximum of 50,200 200 x
200 m georeferenced cell locations. Observations in excess of 48,578 (214 rows by 227
columns) were made to create a soils data plane
for the study. Ninety-eight different soil map
units were recorded within the study area,
representing 87% of the 113 soil map units
recorded for the combined Montezuma and
Dolores County area.
a computer data entry sheet. Township, range,
and section values associated with that 2 x 2
kID block were used as general guides to selecting the proper aerial photograph from the coded set of photos on file with the SCS. Once
found, distinct cultural and natural features
visible on the orthophoto quad were used to
position the lOO-cell mylar grid on the aerial
photo and match it with its 100-celI mylar grid
on the orthophoto. The evaluation process began with the cell in the first row and first column, and proceeded eastward then southward
until all cells within the 2 x 2 kID block were
examined. The dominant soil type within the
cell was determined and its soil map unit code
(Table 2.2) entered in its proper geographic
position within the rows and columns on the
data entry sheet. Many times, one soil type
completely filled a cell, but frequently, two,
three, four, and even five different soils fell
within a single cell. In these situations subjective judgment was exercised to select one soil
type as representing the greatest area in the
cell. Of the 502 2 x 2 kIn blocks covering the
entire study area, 76 were nearly or completely
empty (Le., no soil data were available), and 32
had fewer than 50 cells with soil information.
This left a total of 394 2 x 2 kIn blocks that
were completely recorded.
The replicability of the subjective method
was tested informally by comparing my "call"
with that of my recording assistant after we had
spent approximately seven 4O-hour weeks performing these evaluations. As a test, 36 out of
the 394 (9.4% of the total number of blocks)
were reexamined. The person who had made
the original assigrunent checked the record
sheets and the alternate evaluator reestablished
the placement of the 100-cell grid on the aerial
photograph and determined the dominant soil
type in each cell. The precision rate (Le., the
number of identical calls) for anyone block
ranged from 77-99% with rates per 7.5-minute
orthophoto quad ranging from 85.3-98.5%,
for an overall mean recorder precision for the
36-block sample of 92.9 ± 5.0% (Table 2.4).
Determining and Recording Soil Types
Once this system was established, the actual
recording of the soils could begin. First, a 2 x
2 kIn block was identified on one of the 12
7.5-minute orthophoto quads. Its number and
northwest comer coordinates were recorded on
-
In this way, 50,200 cells were located and
evaluated as to predominant soil type and their
associated values entered into a computer file.
Eventually the original image, which was comprised of the set of 12 mosaiced OEMs (214
rows x 227 columns representing 1943 kIn 2 or
39-
Table 2.4. Inter- and Intra-Coder Variability Rates in Soil Recording.
Orthophoto Quad
Map Name
Ruin Canyon
Negro Canyon
Bowdish Canyon
Pleasant View
Woods Canyon
Battle Rock
Yellowjacket
Arriola
Mud Creek
Trimble Point
Dolores West
Cortez
Resampled Block Numbers and
Percent of Cells in Agreement
10=95%, 11 =96%, 20=87%, 33=93%, 39=85%
57=95%, 61=94% 74=95%
92=96%, 105=96%, 122=88%
144=93%, 146=91 %, 162=92%
178=96%,199=91%,211=87%
225=96%, 227=91 %, 229=93%
275=96%, 286=92%, 291=95%
300=77%, 301=89%, 317=80%, 325=95%
341=99%, 369=98%
389=95%, 402=99%, 414=94%
421 =95%, 438=97%
468=96%, 487=99%
750 mi 2) with 48,578 cells, was trimmed along
its southern margin to reflect the absence of
soil data in that portion of the study area. The
final image contained 45,400 cells (200 rows
by 227 columns representing 1,816 km 2 or
701 mi 2). However, 7,175 cells were "empty"
(Le., no soils were mapped in those locations).
Some cells lacked information on natural plant
productivity or waterholding capacity. Others
were located on the margins of the soil plane
where information on the elevation plane was
lacking. Ultimately, some 8,641 cells could not
be analyzed, leaving a final total of 36,759
cells within the 200 x 227 image that could be
analyzed, representing 1,470 km 2 or 567 mi 2
of data
MODERN CROP YIELD DATA
To determine what effect soil moisture
would have on agricultural productivity, data
were sought on historic yields for dryland
crops grown in the study area. These data were
available from two sources: records of historic
bean and com yield in Montezuma and Dolores counties compiled by Burns (1983:313,
315, 388), and dry bean crop yield values for
46 of the 113 soil types in Dolores and Montezuma Counties available from the Cortez area
Soil Conservation Service (see Table 2.3, column 9). Corroborative data were found in environmental studies undertaken by the Dolores
Archaeological Project (petersen 1987:220,
221), and an early study undertaken by the
Mean Percentage
in Agreement Per
Quad
91.2%
94.7%
93.3%
92.0%
91.3%
93.3%
94.3%
85.3%
98.5%
96.0%
96.0%
97.5%
Colorado State College (University) Agricultural Experimental Station (Brown 1938).
These data helped to solve two methodological problems. The first was to establish a
relation between PDSI values and agricultural
yield. What is the shape, direction, and strength
of any relationships between bean yield and
PDSI, and between maize yield and PDSI? The
second problem was to assign crop yield values
to soil type and PDSI values and thereby create
a quantitative "calibration" between soil type,
soil moisture conditions, and agricultural productivity. Solving the second problem required
solving the first.
In order to address the first problem, relationships between historic bean yield, actual
PDSI values, and reconstructed PDSI values;
the relationships between com yield, actual
PDSI values, and reconstructed PDSI values;
and the relationship between bean yield and
com yield needed to be explored. In order to
address the second, a correlational study
needed to be conducted on historic crop yield
data available for selected soils with natural
plant growth yield data obtained under a
variety of climatic conditions for those same
soils within the study area. This would establish
a transfer function that could be used to assign
potential crop production values to all soils,
regardless of the availability of modem yield
data. In sum, the crop yield and PDSI data and
regression studies established the direction and
strength of the relationship, and the crop
-40-
yon 1965 [1971 PI] (formerly Moqui NE
1955), Arriola 1965 [1973 PI], Dolores West
1965, Bowdish Canyon 1979 (formerly Moqui
SW 1955), Battle Rock 1979 (formerly Moqui
SE 1955), Mud Creek 1979 (formerly Cortez
SW 1955), and Cortez 1965.
yield/soil type/potential natural productivity
data and correlation studies provided distributional information and the values used to
calibrate absolute crop yield values with PDSI
moisture classes.
DIOITAL ELEVATION DATA
MODELINO SOIL MOISTURE
Digital Elevation Models
Palmer Drought Severity Index
Digital maps are numeric representations
of cartographic data used in the analysis of
spatially distributed phenomena Particularly
useful are digital maps encoding elevation
within the area of a 7.5minute (1 :24,000
scale) U.S.O.S. topographic map. Such digital
data sets, called Digital Elevation Models
(DEMs), have become standard cartographic
products of the U.S.O.S. National Mapping
Program. They consist of an array of numeric
values spaced at 30m intervals georeferenced
to the UTM system. This results in each numeric value having three locational attributes, an
x, y, and z coordinate, representing easting and
northing (horizontal) values and an elevation
above sea level (vertical) value. They are often
created as byproducts of orthophoto quads
from aerial photographs taken at an altitude of
12,192 m or 40,000 ft (l:80,OOO scale). Each
digital map is tested for critical accuracy before it is included in the National Digital
Cartographic Data Base and made available to
the public. Vertical accuracy must be within ±
7 m of the true elevation for that point when
created by orthophotographic methods
(McEwen et al. 1983; Elassal and Caruso
1983) or within onehalf of a contour interval
when created from topographic maps (U.S.
Bureau of the Budget 1979).
The Palmer Drought Severity Index (PDSI)
is a temporally sensitive measure of soil moisture. It was designed as an index of meteorological drought, defined as "a period of prolonged and abnormal moisture deficiency"
(palmer 1965:1). Its creator, Wayne Palmer, a
meteorologist working for the U.S. Weather
Bureau, developed a quantitative method for
calculating moisturedeparture values that can
be compared through time and across space.
(It is important to note that PDSI values represent departures from the longterm mean
condition of a place as recorded by a single
weather station. As such, the absolute amount
of water in a soil at one place is not being compared directly to the absolute amount of water
in another soil in a different place; rather it is
the magnitude of the departure from normal
local conditions that is being compared. Thus,
a PDSI indicating moderate drought for a soil
in Iowa may be based on more soil moisture
than a PDSI indicating a wet period for a soil
in southwestern Colorado).
Palmer assumed that local economies were
adapted to the longterm mean climatic conditions of a place and that it was only Significant departures from this mean that presented
adaptational problems. His immediate goal was
to evaluate the frequency, duration, and magnitude of prolonged periods of abnormally dry
or wet weather in the past and ultimately to
forecast the likelihood of either condition in
the immediate future. Today the calculation,
reporting, and graphic depiction of the PDSI
has become standard for the U.S. Departments
of Commerce and Agriculture (e.g., NOAA's
Weekly Weather and Crop Bulletin) because it
is a reasonably accurate and integrative measure of the effects of past temperature and
precipitation on current soil moisture conditions. Whereas it is primarily a meteorological
For many portions of the U.S., DEMs may
be purchased from the U.S.O.S. National
Cartographic Information Center. They are
stored on magnetic tape and various formatting
options may be specified by the user; I used an
unlabeled 9track tape, at a density of 6250
BPI, with a fixed blocksize of 7200, in ASCII
character mode. This study required purchase
of 12 adjacent DEMs: Ruin Canyon 1979 (formerly Cedar 3 SW 1955), Pleasant View 1965
[1971 PI], Yellowjacket 1965 [1973 PI],
Trimble Point 1965 [1973 PI], Negro Canyon
1979 (formerly Moqui NW 1955), Woods Can
41
index, it has significant implications for plant
growth. hydrological supplies, and human economic systems. including agriculture.
Palmer utilized a water-balance or hydrological accounting approach to modeling soil
moisture conditions (palmer 1965:6). Both
sup-ply and demand of water are considered.
Water supply to a given soil is modeled by precipitation (from snow. rain. and recharge from
other water sources) and water already stored
in a soil profile. Water demand for a given soil
is modeled by potential evapotranspiration (the
process of transferring soil moisture to the atmosphere by evaporation of water and transpiration from plants). by potential moisture
needed to recharge soils to their normal waterholding levels. and by runoff required to
maintain water in rivers. lakes. and reservoirs.
Calculation of the PDSI requires specific
climatic. locational. and soil moisture data
These include monthly total precipitation and
monthly mean temperature values consistently
recorded at known stations near the area of interest. These climatic records must also incorporate enough time to embrace characteristic
trends as we)) as significant variation occurring
in the area so that reasonably accurate mean
climatic values can be computed. The PDSI
calculations also require information on the
global location of the study area. in the form
of monthly daylight adjustment values, so that
potential evapotranspiration rates can be computed. Fmally, data on the available waterholding capacity of soils used in the study area
must be estimated or provided. Palmer's model
divides a soil profile into an upper soil layer
equivalent to the plow zone, and a lower soil
layer that includes the effective root zone of
economic plants. In general, researchers who
derive PDSI consider this upper zone to be six
inches thick and assume it contains one inch of
water at field capacity. The depth of the remaining lower soil section normally is not specified but it is assumed to contain five inches
of water.
A schematic diagram of the Palmer waterbalance model would portray incoming moisture filling the upper soil layer until it reaches
its field capacity, then filling or recharging the
lower layer until it reaches its field capacity.
Only then can excess water run off the ground
surface. Similarly. evapotranspiration takes
place first at the surface layer and removes all
of its stored moisture before any moisture is
withdrawn from the lower layer.
Details of the method used to derive PDSI
are contained in Palmer (1965), but in broad
outline the procedure is as follows. PDSls are
expressed as negative (dry) and positive (wet)
real numbers (e.g.• -3.23, +1.75) and result
from a number of steps. First, all relevant values in the hydrological accounting are calculated on a monthly basis for an adequately
long and representative period. The minimum
number of years is not specified by Palmer. but
30 years is considered minimum by researchers who regularly use PDSI in paleoenvironmental reconstructions (Martin Rose, personal
com-munication, April 1989). Second, certain
der-ived mean values are compared to create
ratios or coefficients that weight the key hydrologic variables in such a way that subsequent
values that are "Climatically Appropriate For
Existing Conditions" (CAFEC) are produced.
These CAFEC values reflect the long-term
mean conditions for evapotranspiration. recharge, runoff, and so on, and therefore reflect
the average climate of the area. Third, the entire time series is reanalyzed using these coefficients and CAFEC values to derive a CAFEC
value for overall moisture required to match
"normal" climatic conditions for each month
in the time series. Fourth, departures (d) from
the CAFEC value are calculated by subtracting
the CAFEC value for precipitation for a given
month (now used as the theoretical or expected
value) from the actual value of precipitation
recorded for that month. Fifth, these departures
are converted to indices of moisture anomaly
(z) by multiplying this quantity by a climatic
constant (k) that represents the ratio of local
moisture demand to local moisture supply for
a given time period. Finally, these moisture
anomaly indices are converted into Palmer
Drought Severity Indices (x) by incorporating
the effects of the previous month's soil moisture index on the current month's index
through an equation that simulates lag time. In
this way, the effects of antecedent meteorological and soil moisture conditions are reflected in subsequent conditions.
-42-
While PDSI is used frequently by dendro-
climatologists (Cook 1982; Cook andJacoby
1977; Meko et al. 1980; Mitchell et al. 1979)
and dendroecologists (Graybill 1989; Zahner
et al. 1989), it has only recently been used in
archaeological applications (Rose et al. 1982).
The research described in this dissertation is a
new use of the PDSI to link longtenn treering
chronologies and estimates of agricultural productivity in the prehistoric Southwest.
Data Sources for Calculating the PDSI
Palmer's PDSI procedure has been implemented as a 800+ line FORTRAN computer
program calIed PDSI or DROUGHT by the
Laboratory of Tree-Ring Research, University
of Arizona This program was used to generate
the PDSI values used in this study.
The requisite climatic data were obtained
from Climatological Data Annual Summaries
for Colorado and Utah, published by the National Oceanic and Atmospheric Administration (l987a, 1987b). Instead of using a single
weather station to model the precipitation and
temperature conditions in the study area, a
region exhibiting significant relief and ranging
in elevation from 1,500-3,012 m (4,921-9,882
ft), I decided to use multiple stations with each
representing a different range of elevations in
the study area. Each selected station had to be
near the study area, have precipitation and
temperature records of adequate length, and be
representative of the regional climate of the
southeastern portion of the Colorado Plateau.
Data from all of the stations initially considered were statistically examined by the
Laboratory of Tree-Ring Research prior to the
commencement of this research for the purpose of detecting "inhomogeneities in the
records and unusual trends" (Graybill
1989:9). Some (e.g., YelIowjacket) were found
to be too short; some (e.g., Dolores and Mancos) possessed either precipitation or temperature records but not the other; and others
(e.g., Durango) had significantly deviant
trends. For the usable data sets any missing
values were estimated with linear regression
(Graybill 1989:9). Eventually, five stations
were selected on the basis of contrasting elevation, proximity to the center of the study
area, and the proportion of the study area that
could be modeled by that weather station. Dis-
play capability and tabular data output functions of the GIS made this an easy task. While
several combinations ranging from three to
five stations were acceptable, the five stations
fmally selected were Ft Lewis, Mesa Verde,
Ignacio, and Cortez, alI in Colorado, and Bluff,
Utah (Table 2.5, Figure 2.2).
Daylight adjustment values appropriate to
these five stations were calculated by dividing
the "mean possible monthly duration of sunlight in the northern hemisphere expressed in
units of 12 hours" (Mather 1977:364, Table
19) by the number of days in a month.
Available water capacity values (AWC) for
the upper six inches and the lower soil section
for each soil were calculated for all 113 soils
within SCS jurisdiction (Table 2.6). SCS data
on average soil depth and water-holding capacity per subsoil unit (Table 2.3) were used to
derive the values required by the PDSI model.
This derivation of actual AWC values is generally not done by the meteorologists, paleoclimatologists, and dendrochronologists who
use the PDSI to model soil moisture. Instead
they usually use the one inch and five inch approximations (e.g., Meko et al. 1980) suggested by Palmer.
I hoped that more accurate estimates of
AWC would lead to more realistic modeling of
moisture stored and withdrawn from individual
soils in the study area and that these could be
portrayed easily with GIS technology. This
seemed to be especially desirable since there
are marked differences in soil quality, texture,
depth and distribution across the area Further,
the purpose of reconstructing PDSI values for
the study area was not only to identify any
episodes of abnormally dry or wet weather but
also to identify the places where these occurred
to form the basis for future models of settlement This ーイッ」・、オセキゥエィ
actual waterholding values computed for soils in the study
area--produced markedly different PDSls than
would have been obtained if only the standard
one inch and five inches of water were assumed
for the upper and lower soil layers.
-43-
TREE-RING DATA
The tree-ring data were eigenvector
Table 2.5. Weather Stations Used to Reconstruct PDSI.
NOAAa
Index
No.
3016
Distance to
Precip.
Temp.
Weather
Study Area
Data
Data
Elevation Center
Longitude
Begins
Latitude
Station
Begins
03'
2,316
m
95km
1912
108
Ft. Lewis, La 37" 14'
1921
(7,600 ft)
(GO mi)
Plata Co.,
CO
2,169 m
40km
1923
108 29'
37 12'
1923
5531
Mesa
(7,115
ft)
(25
mi)
Verde,
Montezuma
Co., CO
168 km
1,969 m
1914
107 38'
1914
37 08'
4250
Ignacio, La
(105
mi)
(6,460
ft)
Plata Co.,
CO
18 km
1,893 m
1931
108 33'
1930
37" 22'
Cortez,
1886
(11
mi)
(6,212
ft)
Montezuma
Co., CO
1,315 m
105 km
1927
1927
109 33'
37 17'
Bluff, San
0788
(4,315
ft)
(66
mi)
Juan Co.,
UT
aNational Ocenanic and Atmospheric Administration (U.S. Environmental Data and Information Service, National
Climatic Data Center).
0
0
0
0
0
0
0
0
UT
co
+
+
c
++
NM
Figure 2.2. Location of the five weather stations.
Lewis.
a, Bluff; C, Cortez; I, Ignacio; M, Mesa Verde; F, Ft.
-44-
Table 2.6. Water-Holding Characteristics of Soils.
Soil
Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Soil Map
Unit
A1B
A2B
A3B
A4
A41
A4A
A4B
A5C
A5D
A5Wa
A6A
AOAB
AOB
AOC
AQa
BD
C2CE
C2D
C2F
C2V
C3CD
C3E
C5C
C7D
CP
DOCE
GP
H1CD
H1DC
H6D
M1CE
M2C
M2CE
M2DD
M2DD1
M5E
M6C
M7E
M8D
OAB
P1C
R1C
R1D
R1HD
R1HC
R1HD
R3B
R3B1
Est. Soil
Depth (in.)
.2:,60
.2:,60
セVP
PVセ
.2:,60
セVP
.2:,60
.2:,60
PVセ
PVセ
.2:,60
セVP
.2:,60
.2:,60
PVセ
0-60
6-20
6-20
6-60
6-20
PVセ
セTP
18-60
10-60
NA
20-40
NA
40-60
20-40
PVセ
20-40
10-60
10-60
10-40
10-40
セXP
40-60
40-80
10-60
セVP
セXP
PXセ
PXセ
PVセ
PVセ
PVセ
.2:,60
.&:.60
Upper Soil
Layer (6 in.)
NA
1.14
0.54
0.96
0.90
0.57
0.99
1.14
1.14
0.93
0.96
1.14
1.14
1.14
1.07
0.00
0.89
0.93
0.97
0.89
1.14
1.14
0.99
1.02
0.00
0.66
0.00
1.08
0.90
0.56
0.48
0.53
0.40
0.85
0.85
1.14
0.80
0.80
0.85
0.93
0.36
1.14
1.14
1.02
1.02
1.02
0.90
0.90
Lower Soil Layer Total Field
(6-80 in.)
Capacity (in.)
NA
NA
9.18
10.32
5.97
6.51
8.37
9.33
7.83
8.73
4.63
5.20
8.91
9.90
9.18
10.32
9.18
10.32
7.69
8.62
6.31
7.27
10.26
11.40
10.26
11.40
10.26
11.40
9.99
11.06
0.00
NA
2.08
2.97
2.08
3.01
4.39
5.36
2.08
2.97
8.90
10.04
8.90
10.04
5.68
6.67
5.81
6.83
0.00
NA
1.98
2.64
0.00
NA
5.40
6.48
3.24
4.14
4.05
4.61
0.53
1.01
2.61
3.14
1.98
2.38
2.95
3.80
2.95
3.80
10.45
11.59
3.94
4.74
7.98
8.78
5.17
6.02
7.35
8.28
3.77
4.13
13.66
14.80
13.66
14.80
7.25
8.27
7.25
8.27
7.25
8.27
8.05
8.95
8.05
8.95
45 -
AWCClass
(1-11 )
NA
9
1
8
7
1
8
9
9
7
6
10
10
10
9
NA
2
2
4
2
8
8
5
5
NA
1
NA
5
3
1
1
1
1
2
2
10
3
7
5
7
1
11
11
7
7
7
8
S
Table 2.6. (Continued).
Soil
Code
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
Soil Map
Unit
R3C
R3C1
R3D
R3D1
R4C
R4C1
R4D
R7C
R7D
R7D1
R8B
RHC
RL
ROB
ROB1
ROC
ROC1
ROD·
ROHC
ROHD
ROLC
ROLD
ROLD1
S1C
S5
S51
S53
S5D
S6
S6C
S6E
S7
S8C
T1B
T2C
T3E
T4B
T6
T6B
T6C
T6D
V1B
V1C
V1D
V2B
V2C
V2S
V2Va
V2Wa
Est. Soil
Depth (in.)
2,60
2,60
2,60
2,60
20-60
20-60
20-60
20-40
20-40
20-40
2,60
20-60
NA
2,60
2,60
2,60
2,60
2,60
20-40
20-40
20-40
20-40
20-40
2,60
40-60
2,60
2,80
2,60
40-60
20-60
20-40
20-60
20-80
2,60
2,60
10-60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
2,60
Upper Soil
Layer (6 in.)
0.90
0.90
0.90
0.90
0.97
0.97
0.96
1.02
1.02
1.02
0.84
0.95
0.00
0.99
0.99
0.99
0.99
0.99
1.02
1.02
1.02
1.02
1.02
0.90
1.02
1.14
1.14
1.14
0.75
1.06
0.80
1.06
1.09
1.05
0.45
0.38
0.93
1.14
1.13
1.13
1.14
1.05
1.05
1.05
0.93
0.87
0.81
0.96
0.96
Lower Soil Layer Total Field
(6-80 in.)
Capacity (in.)
8.05
8.95
8.05
8.95
8.05
8.95
8.05
8.95
7.06
8.03
7.06
8.03
5.74
6.70
4.48
5.50
4.48
5.50
4.48
5.50
8.28
9.12
6.92
7.87
0.00
0.00
10.26
11.25
10.26
11.25
10.26
11.25
11.25
10.26
10.26
11.25
5.40
6.42
5.40
6.42
4.67
3.65
4.67
3.65
4.67
3.65
9.33
8.43
7.80
8.82
9.15
8.01
10.45
11.59
8.01
9.15
5.51
6.93
5.87
6.93
4.25
5.05
5.40
6.46
9.51
8.42
6.98
5.93
5.07
4.62
0.75
1.13
3.39
4.32
8.81
9.95
9.45
8.32
9.45
8.32
8.81
9.95
9.45
10.50
10.50
9.45
9.45
10.50
9.30
8.37
8.37
9.24
8.10
7.29
8.47
9.43
7.52
6.56
-46-
AWCClass
(1-11)
8
8
8
8
7
7
5
4
4
4
8
6
NA
10
10
10
10
10
5
5
3
3
3
8
7
8
10
8
5
5
4
5
8
5
1
1
3
8
8
8
8
9
9
9
8
8
7
8
6
Table 2.6. (Concluded).
Soil
Code
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
Soil Map
Unit
V4C
vas
VOC
vocs
VOCC
was
wassa
wac
XC6D
XC6E
XH2D
XH2E
XH3E
XM1F
XM3D
Z
Est. Soil
Depth (in.)
セVP
セVP
セVP
セVP
PVセ
セVP
セVP
セVP
10-20
4-60
PVセ
10-60
10-60
10-20
10-20
セVP
Upper Soil
Layer (6 in.)
1.00
0.84
0.84
1.14
1.14
1.08
0.93
1.08
0.84
0.39
0.53
0.37
0.40
0.30
0.38
0.70
Lower Soil Layer Total Field
(6-80 in.)
Capacity (in.)
8.54
9.54
6.29
7.13
6.29
7.13
6.44
7.58
6.44
7.58
9.72
10.80
8.89
9.82
9.72
10.80
2.04
2.88
0.80
1.19
4.61
5.14
3.58
3.21
3.55
3.95
0.31
0.61
0.97
0.59
7.18
6.48
AWCClass
(1-11)
8
6
6
6
6
9
8
9
2
1
1
1
1
1
1
1
aArtificially wet and possibly saline from irrigation water.
amplitudes (factor scores) that resulted from a
principal components analysis of seven expanded tree-ring chronologies from the Colorado Plateau. The data set, "SWOLD7"
(Southwest Old Seven), was created by Martin
Rose in conjWlction with research Wldertaken
by the Laboratory of Tree-Ring Research. The
original data, as well as the data manipulation
and reduction techniques used to create the
107D-year (row) by seven variable (column)
amplitude matrix, are described by Rose et al.
(1982).
The tree-ring chronologies used by Rose et
al. (1982) are called "expanded chronologies" because they result from merging chronologies created with archaeological specimens
with chronologies created from living trees
(Dean and Robinson 1977). The merging was
possible because of a temporal overlap between
the recent end of the archaeological series and
the older end of the living tree series. Rigorous
statistical tests were performed on the period of
overlap to ascertain whether the archaeological
.and living-tree sequences represented a single
statistical population, and the tests confirmed
the acceptability of merging these chronologies (Rose et al. 1982).
Methods for constructing tree-ring chronologies are discussed in detail in Stokes and
Smiley (1968); Fritts (1976); Hughes et al.
(1982). The tree-ring chronologies themselves
are already standardized to facilitate comparison and to emphasize the macroclimatic
factors that affect tree growth. Standardization
is accomplished in several ways. Dated, annual
ring-width measurements are fitted to growth
curves, removing the biological effects of
aging on growth rings, and are then converted
to departures from the growth curve for that
sample. Thereafter, ring-width indices for all
samples for a particular site or locale are averaged to create a mean index value per year for
that location. Thus, a tree-ring chronology,
from living trees or archaeological samples, is
comprised of the standardized, averaged, ringwidth indices for a location or region; each
value represents the mean index for all specimens included in that chronology for a given
year.
Rose examined many chronologies before
settling on seven (Table 2.7, Figure 2.3). A
battery of statistical techniques and tests was
used to identify chronologies that would emphasize tree growth controlled by the macroclimatic variables of precipitation and temperature over growth affected by nonclimatic
sources of variation. Rose then subjected the
tree-ring indices of the seven selected chronologies to a principal components analysis
-47-
Table 2.7. Tree-Ring Chronologies Used in the Creation of SWOLD7 (Rose et al. 1982; years AD).
Archaeoloaical Chronologies
End
Number
Date
Name
799900
1930
Santa Fe
890009
1864
Jemez
Mountains
1751
803999
Gobernador
Cibola
400999
1900
Chama Valley
900009
1835
Mesa Verde
799199
1276
Cebolleta
Mesa
666003
1885
Living-Tree Chronologies
Beginning &
Number
End Dates
Name
1556-1972
Glorieta Mesa 343000
Paliza Canyon 383000
1658-1972
Pueblito
Canyon
Turkey
Springs
Echo
Amphitheatre
Bobcat
Canyon and
Schulman Old
Tree
Cebolleta
Mesa
Expanded Tree-Ring
Chronoloaies
073000
1594-1971
Name
Santa Fe
Jemez
Mountains
Gobemador
272000
1595-1972
Cibola
400009
171000
1362- 1972
900005
061099
and
641000
1390-1971
and
1200-1963
Chama
Valley
Mesa Verde
303000
1662-1972
UT
Cebolleta
Mesa
Number
700000
890000
803000
772100
666000
co
+
+
2
+
+
3
+5
6
+i
+7
NM
Figure 2.3. Location of the seven tree-ring chronology stations. 1, Mesa Verde; 2, Gobernador; 3,
Cibola; 4, Cebolleta Mesa; 5, Chama; 6, Jemez; 7, Santa Fe.
-48-
(PCA), following conventions established by
the Laboratory of TreeRing Research (Stockton 1975; Fritts 1976; Fritts et al. 1981). This
was done to further reduce the data and create
a set of statistically independent (orthogonal)
predictor variables that could be used for subsequent climatic reconstructions. PCA is a type
of factor analysis that reexpresses original
multivariate data as axes ("principal components" or "eigenvectors") of variability in the
same number as the original variables. It is
done in such as way that each axis is orthogonal to the others and the first axis accounts
for as much of the total variance as possible,
with subsequent axes accounting for decreasing amounts of the total variability. Given that
the seven chronologies used were all created
with an emphasis on macroclimate, and that
they are all from a fairly restricted geographic
region, Rose anticipated that the original variables (the seven chronologies) would be correlated to some degree.
large data sets and for displaying results were
essential to this research. The term "geographic information system" refers to a particular
type of computer system incorporating software programs and hardware to capture, store,
manipulate, analyze, retrieve, and display data
referenced to geographic locations (Avery and
Berlin 1985; Berry 1989a, 1989b; Burrough
1986; Clarke 1986; Kvamme 1988; McDonald
and Crain 1985; Marble 1984; Parker 1988;
Tomlinson et al. 1976). In contrast to computer-automated mapping and design programs (Cowen 1988; Dueker 1987), which
generally focus on the manipulation and display of a single image, GIS have the ability to
interrelate multiple spatial and nonspatial data
sets concurrently; to create new information
through combinations or transformations of
original data (Collins and Moon 1981; Monmonier 1982:76-79); and to produce analytic
products such as tables, charts, and graphs in
addition to map-like images.
Therefore, performing a PCA would
produce a set of uncorrelated axes that would
not only meet the statistical requirements of
later manipulations but also would likely produce a few axes accounting for most of the
variance and a larger number of axes that
would account for small amounts of variance.
Later, the axes depicting the majority of the
variability could be used in stepwise, multiple
regression-based climatic reconstructions and
the lesser contributors ignored. In this way, a
reduced and more powerful data base is produced, particularly since it has been found that
the largest axes are generally those that can be
directly attributed to climate (Stockton 1975;
Fritts et al. 1981).
There are two basic types of GIS, which are
classified on the basis of how the geographic
information is spatially referenced and organized. Vector-based systems encode spatially
referenced data as points (e.g.• a spring), lines
(e.g., a stream), or polygons (e.g., the perimeter of a lake) to defme the specific location and
boundaries of spatial phenomenon. The boundaries of the polygons imply uniformity of
their interiors (Berry 1989b:34). Recording
data in vector format facilitates production of
images that resemble ordinary maps; they are
ideal for bounding discrete spatial data; they
use less computer storage space than do rasterbased systems; and they are easy to "rasterize" at a variety of cell sizes. However, they
require more complex algorithms to encode
and manipulate data and hence operate more
slowly, are less appropriate for representing
continuous spatial data (e.g., elevations) and
often must be converted to raster format to
perform even simple analyses, such as determining area (Maffini 1987).
Appendix A contains the final output
generated for the PCA on SWOLD7. The
values listed are eigenvector amplitudes, "the
products of eigenvectors and the data from
which the eigenvectors were derived expressing
the importance of each eigenvector in each
observation set" (Fritts 1976:531). This table
was the source of tree-ring data for this study.
GIS TECHNOLOGY
Geographic Information System hardware
and software for managing and analyzing these
Raster systems, on the other hand, encode
spatially referenced data to specific cell locations on a regular grid of columns and rows
arbitrarily placed over a study region. Raster
systems approximate the location of spatially
distributed information in a probabilistic fashion with a dominant or mean value assigned to
-49-
the entire cell. In contrast to vectorbased systems, raster systems depict interiors and only
imply boundaries (Berry 1989b:34). Consequently, they are better at representing continuous spatial data that in the real world have no
visible edge (e.g., slope, climatic zones). Because raster systems are cellular, the map-like
images they produce are less "realistic" representations than vector display images, they do
not convert accurately to vector form, and they
require more computer storage space. However, they require much simpler mathematics to
store, overlay, and analyze data, and they more
readily produce statistical and numeric output
than do vector systems (Maffini 1987).
GIS technology has existed since the 1960s
but it has only become a mainstream data management tool since the mid-1980s for land-,
facilities-, resource-, and records-management
agencies and as a scientific tool for the professional community. At least 10 recent fulllength texts and reference books have been
written about the history and use of geographic
information systems (Aronoff 1989; Bracken
et al. 1989; Burrough 1986; Forrest et al.
1990; GIS World 1990; Huxold 1990; Ripple
1987, 1989; Star and Estes 1989; Tomlin
1990) and many articles are published arumally in a variety of professional journals, conference proceedings, and trade publications
that report on the analytic capabilities and limitations of the technology, new methods of
modeling, and the substantive results of specific studies.
GIS technology is increasingly being used
by archaeologists in the management and
analysis of archaeological and environmental
data. The first book-length treatment of GIS
has been published (Allen et al. 1990). Recent
review articles by Kvamme (1986a, 1986b,
1989) and Kvamme and Kohler (1988) detail
the history of use and the potential future uses
of this technology in archaeological research.
Additional examples of GIS use in cultural resource inventories and database management
can be found in, for example, Briuer (1988),
Brown and Rubin (1982), Calamia (1986),
Harris (1986, 1988); Hasenstab (1983a), and
Parker (1986). For predictive models of site
location and land use, see Bailey et al. (1985),
Brown and Rubin (1982), Carmichael and
Christensen (1988), Creamer (1985), Farley et
al. (1990), Hasenstab (1983b), Kvamme
(1983a, 1983b, 1984, 1985, 1986b, 1988),
Kvamme and Jochim (1989), Marozas and
Zack (1990), Martin and Scurry (1988), Parker
(1985), Parker et al. (1986), Scholtz (1981),
Wansleeben (1988), Warren (1988), and
Warren et al. (1987). GIS approaches are also
well-suited to simulation models focused on
archaeological methods, as can be seen for the
study of sampling strategies and statistical
models in Kvamme (1986b); studies of surface
artifact distributions (Gill and Howes 1985);
and for simulation models focused on behavioral processes, such as movement and size
of populations (Zubrow 1990) or agricultural
productivity (Darsie 1982).
This study might be described as a simulation of prehistoric agricultural potential. A dynamic model of climatic variation and soil
moisture conditions has been applied to a set
of spatially distributed environmental variables
in order to simulate the variation in agricultural
production over time and across space in order
to "observe" its effect on human populations.
The model affords "dynamism" by virtue of
the time-depth of the annual tree-ring data
from the study region. The model suggests
methods and generates data that allow the size
and shape of the annual agricultural niche to
be delimited on a number of criteria, and it
permits sustainable population size to be estimated in a number of ways. Consequently, a
series of scenarios or "submodels" could be
generated and explored for their congruence
with known archaeological data, although this
is not done here. What the simulation model
has been able to do is provide insight into the
role that climatic variation plays in determining
soil moisture and plant growth conditions,
which in tum affect the spatial and temporal
distribution of farmable lands and set limits on
the size of human populations over given periods of time. Without GIS it would have been
impossible to model these phenomena, depict
them spatially at anyone point in time, or simulate them dynamically from the perspective of
a specific geographic location over time.
VICAR/IBIS and EPPL7
Washington State University (WSU) is
among a number of institutions that has devel-
-50-
mental) status for the upgraded Release 2.0. It
runs under MSDOS and was developed by the
Minnesota State Planning Agency of S1. Paul.
Minnesota as a commercial product for IBMcompatible machines with either EGA or VGA
display capabilities. EPPL7 has four major
program components that permit encoding and
capture of digital data (DIGITIZE), analysis
and manipulation of spatial data (EPPL). display of spatial data (DISPLAY) and output of
spatial data (DOTPLOT). It can also import
files from and export files to other computer
systems. It is easier to use than VICAR and has
more GIS capability in most functional areas.
Although normally an interactive system,
EPPL7 can also perform batchtype operations
in which lengthy instruction sets are read from
disk. In EPPL7, data can be entered into the
system as stored files on diskettes, entered directly from the keyboard, or entered via digitizing programs linked to external digitizing
equipment on a mouse. Data are stored as files
and may be manipulated and transformed in a
variety of ways by evoking various executable
EPPL7 commands tailored to user requirements through specification of parameters and
options. Spatial images can be immediately
displayed on color monitors and modified interactively. Tabular data output and digital images may be plotted on standard personal
computing printing devices. Although no internal photographic equipment is installed to
record spatial images in color on either EPPL7
or VICAR, it is quite easy to photograph images on a color monitor screen using standard
camera and tripod setups.
oped a considerable facility devoted to image
processing and geographic analysis. WSU's
Digital Image Analysis Laboratory (DIAL) has
obtained a number of image processing and
spatial analysis systems for which instruction
and research support are available. DIAL was
established in the late 1970s with the installation of a single, powerful mainframe batchprocessing system developed at NASA's Jet
Propulsion Laboratory in Pasadena California
called VICAR/IBIS (Hart and Wherry 1984).
These are acronyms for "Video Image Communication and Retrieval" and a subset appli.cation called "Image Based Information Systern." Since that time. DIAL has upgraded its
capabilities for interactive image analysis and
GIS software and hardware, but at the time of
this study, only VICAR/IBIS and EPPL7 were
available. From DIAL's system offerings, an
IBM DOSbased GIS software package was selected to be the primary system for use in this
research. EPPL7 was selected because of its
greater accessibility (it is installed on multiple
IBMcompatible machines), its simplicity, and
its appropriateness to the tasks required.
VICAR/IBIS was used only to process and mosaic the OEMs.
VICAR/IBIS is a set of computer programs
designed specifically to process digital image
data It is rasterbased. A special set of programs called ffiIS was added later to analyze
previously processed earth imagery (Hart and
Wherry 1984:1). It is considered a useroriented system in which FORTRAN and Assembler language programs are invoked by specifying VICAR control language statements and
program options. VICAR runs on IBMenvironment mainframe computers and at the present time resides within WSU's IBM 3090200
system. Input to the system may take the form
of digital data stored on tape or entered by
keyboard. Manipulation of data takes place
through batch programming. Image display is
accomplished by transmitting the image data to
image monitors such as the International Imaging Systems (I2S) color monitor. Permanent
output is routed to the mainframe printers and
related plotting devices. Data storage is accomplished through disk and tape on the IBM
mainframe.
VICAR and EPPL7 not only made this research possible by their ability to capture, store,
and manipulate vast quantities of spatial and
nonspatial data, but they also established the
conditions under which the data had to be
structured. Three examples illustrate this second point First, being raster or cellbased systems, both VICAR and EPPL7 ultimately must
use data in raster format. TIlis was an important
consideration in the way soils data were collected. Second, OEMs when purchased from
the USGS are stored digitally on magnetic
tape. This necessitates access to a tape drive
that can load and transmit data to the appropriate program. VICAR is a mainframe program with preestablished connections to mainframe data transfer programs and tape drives;
EPPL7 is a rasterbased GIS that, at the
time of use, was in beta release (i.e., develop-
51
this was clearly the system of choice with regard to capturing digital elevation data and the
converting them to fonnats readable by
EPPL7. Third, EPPL7 runs on widely available
personal computers and is much easier to use
than VICAR, but it has limitations on the number of classes of data in anyone thematic layer
that can be manipulated at one time. The maximum number of classes is 255. Therefore,
some data needed to be reclassified in order to
be analyzed and displayed; for example, elevations depicted on the mosaiced set of 12
OEMS ranged from a low of 1,500 m to a high
of 3,012 m, representing 1,512 different elevation values. In order to use the full range and
display these data, these 1,512 different values
had to be reduced to fewer than 255. In the
end, 189 elevation classes were established with
each class representing eight meters.
In sum, geographic infonnation systems
were an essential aspect of this research. The
capabilities and limitations of the available
systems were also important. EPPL7 was selected as the major system of spatial analysis
with VICAR providing auxiliary assistance in
the initial capture and organization of the elevation and soil data.
EPPL7 was relatively easy to learn and was
more accessible given its residence on multiple
personal computer systems. Sufficient as a
spatial analysis system, it has limitations as a
database management and statistical analysis
system. Consequently, large data sets containing nonspatial tabular information created by
EPPL7 were transferred to an ffiM mainframe
for editing and statistical analysis. The use of
GIS technology depended on the existence of
other computer systems for the creation of input data sets and for subsequent analysis of
data using statistical software.
ARCHAEOLOGICAL DATA
Block Survey Data
Archaeological information from extensive
block surveys in the study area was sought to
generate popUlation estimates. These archaeological estimates of actual population would be
compared with the environmentally-conditioned estimates of potential population derived
from the model. The comparison would permit
an assessment of how close the prehistoric population estimated from the archaeological data
was to the periodic and long-term values for
optimum and maximum carrying capacity estimated from data generated by the environmental model. This would provide a preliminary test of the model's utility for providing
insight into issues relating to population/resource relationships.
Within the study area, three large-scale,
contiguous block surveys had taken place: on
Mockingbird Mesa (survey conducted 19811984, reported by Fetterman and Honeycutt
[1987]); in the Dolores Archaeological Project
area (conducted 1972-1973 and 1978-1981,
reported by Breternitz and Martin [1973],
Orcutt and Goulding [1986], Schlanger
[1986a], and Schlanger and Harden [1986]);
and in the Sand Canyon Locality (1986-1987,
reported by Van West [1986], Van West et al.
[1987], and Adler [1988]).
Other programs of intensive block surveys
having portions of their survey areas within the
study area were considered but not used. Sur- \
vey has been conducted in the Hovenweep
National Monument area on NPS and BLM
monument parcels and within samples of 40acre quadrats between Montezuma Creek/Cross
Canyon on the west and McElIno Canyon/Yellow Jacket Canyon on the east (Winter 1975,
1976). Contiguous block survey has occurred
on Squaw Point and Cow Mesa between Papoose Canyon on the west and Ruin Canyon
on the east (Neily 1983). Survey also has taken
place in 100 8o-acre quadrats selected by a
stratified random sample of locations within
the BLM's Sacred Mountain Planning Unit in
Dolores and Montezuma Counties (Chandler et
al. 1980). These last three datasets were rejected as sources of data for this research because
of lack of complete records on sites or survey
coverage, lack of data on room counts or rubble area, lack of precise dating of archaeological components, noncontiguous survey
coverage accompanying quadrat-based sample
surveys, and redundant elevational range coverage.
The Dolores Archaeological Project data
set was also excluded because so few of the
more than 1,600 sites recorded were assigned
-52-
to the A.D. 9001300 period of interest, and
so few of these provided data useful for estimating population (Schlanger 1985: 142, 152154, 158-164, 168, 174, 181, 193-195). Given
the large number of sites on both Mockingbird
Mesa and in the Sand Canyon Locality dating
to the A.D. 900-1300 period for which population estimates have already been derived, and
the contrasting elevational settings of these two
areas-Mockingbird Mesa at 1,859-1,951 m
(6,lO()",{),400 ft) and Sand Canyon area at
2,012-2,064 m (6,600-7,100 ft)-it was decided that these two surveys would provide
sufficient data for the preliminary test (Figure
1.1). Schlanger's estimates of population in
the Mockingbird Mesa Survey Locality
(Schlanger 1985:194, 199) and Adler's estimates of population for the Sand Canyon
Survey Locality (Adler, personal communication 1989; Adler 1990) were used as given.
セjdゥrM・ イt
Dated Sites
The files of the Laboratory of Tree-Ring
Research were searched for evidence of well-
documented and excavated habitation sites
dating to the A.D. 900-1300 period in the
study area that had produced sets of tree-ring
dates that included at least two cutting dates
and at least one cluster of dates. In addition,
supporting evidence that the site was likely to
have been occupied during the tree-ring dated
interval was sought from discussions of chronology in published reports. These sites would
provide locational and temporal information
that might be useful in exploring the role that
agricultural potential in the site's immediate
"catchmnent area" might have played in the
decision to establish or expand a habitation in
that location at a given time.
The study area is entirely encompassed
within the Colorado "V" quadrangle designated by the Laboratory of Tree-Ring Research. Within the 1816 km 2 study area, 46
archaeological sites associated with datable
tree-ring samples were identified. Twelve of
these were reported in Robinson and Harrill
(1974) (TRL# 1, 2,3,4,5,7,15,21,22,24,
125, 126); the remaining 34 (TRL# 169, 170,
176, 177, 178, 179, 181, 182, 183, 184, 187,
Table 2.8. Tree-Ring Dated Sites.
TRL
No.a
231
237
Site No.
5MT8371
5MT8839
235
5MT2433
182
5MT4126 b
183
5MT6970 b
2
5MT1566
170
5MT2149
187
233
5MT3834
5MT765 c
Site Name
None
Norton
House
Aulston
Pueblo
Ida Jean
Ruin
Wallace
Ruin
Lowry Ruin
Escalante
Ruin
Mustoe Site
Sand
Canyon
Pueblo
Occupation
Period
(Years A.D.)
935-950
1029-1048
No. of
Tree-Ring
Samples
5
19
No. of
Cutting
Dates
2
8
25
5
32
24
1045-1071
44
15
1086-1120
53
33
1124-1138
25
9
1173-1231
1252-1274
24
329
9
124
References
Dykeman 1986
Fuller 1987; Kuckelman
1988
Kane 1975a; Monis
1030-1050
1986a
Brisbin and Brisbin 1973 1124
Bradley 1974, 1984,
1988b
Martin 1936; White and
Breternitz 1976
Hallasi 1979; White and
Bretemitz 1979
Gould 1982
Adams 1985; Bradley
1986, 1987, 1988a,
1992; Kleidon and
Bradley 1989
aSite number of the Laboratory of Tree-Ring Research, University of Arizona, Tucson.
bSMT4126 and SMT6970 are combined as a single site in a later step to represent an A.D. 1045-1124 occupation.
CData as of April 1990; SMT76S currently under investigation by Crow Canyon Archaeological Center, Cortez, CO.
Note: Tree-ring dates are listed by specimen for each site in Tables 5.5, 5.8, 5.11, 5.14, 5.17, 5.20, 5.23, and S.26.
-53 -
188, 194, 195,211, 216, 217, 228, 231, 232,
233,234,235,236,237,238,239,240,241,
242, 243, 247, 253, 254) are documented in
unpublished files at the lab. For the most part,
the earliest cutting date from a site was used as
the beginning occupation date. The last date,
regardless of its status as a cutting date or noncutting date, was used as the end date of the
occupation/use period. Within this bracketed
period, a cluster of dates was sought. Nine of
the 46 sites met all these criteria. Four of these
may be characterized as small villages, hamlets,
or homesteads; (5MT8371, 5MT8839,
5MT2433, and 5MT3834), and five as larger
villages or "central places" (5MT6970,
5MT4l26, 5MT1566, 5MT2149, and
5M1765) (Table 2.8, Figure 1.1).
It was hoped that these sites would possess
the following characteristics. They would be
associated with occupational dates that are reasonably representative of the occupational
history of the region. Also, they would be dispersed geographically and located in various
elevational settings. For the most part, these
expectations were met. The early 900s were
represented by a single site, the middle 1000s
by three, the late 10005 by two, the early 1100s
by three, the late 1100s by one, the early 1200s
by one, and the middle 1200s by one site.
Lacking representation are the middle and late
900s, the early WOOs, middle 1100s, and the
very late 1200s. The sites were somewhat dispersed across the study area, although all were
essentially mesa top sites, and none was located
in deeply entrenched canyons. None of the
nine sites was located in the lowest elevational
range modeled by Bluff climatic reconstructions «1,605 m/5,266 ft, representing 2.7% of
the study area), or in the highest elevational
range modeled by Ft. Lewis climatic reconstructions (>2,244
,362 ft, representing
2.4% of the study area). Three of the nine were
located within the elevational range modeled
by Cortez climatic reconstructions (1,6051,932 m/4,26(r-6,339 ft, representing 41.0% of
the study area), two were within the elevations
modeled by Ignacio climatic reconstructions
(1,933-2;068 m or 6,442-6,785 ft, representing 33.4% of the study area), and four were
within the elevational range modeled by Mesa
Verde climatic reconstructions (2,069-2,244
mor 6,788-7,363 ft, representing 20.5% of the
study area). While all sites contained tree-ring
samples with cutting dates, the clusters of dates
were not always as clear as would have been
liked; however, architectural and artifactual
data also support the dates assigned in Table
2.8.
mn
In sum, the nine sites that met the minimum criteria include four small habitations
and five larger habitations. Collectively they
are reasonably representative of the occupational history of the area. Whereas the highest
and lowest elevations taking in only 5% of the
study area are not represented in the sample of
sites, the middle ranges that take in some 95%
of the elevations in the study area are represented. Finally, the sites are somewhat dispersed geographically.
-54-
3
BUILDING THE MODEL:
PRELIMINARY DATA ANALYSES
This chapter explains the process of reconstructing Palmer Drought Severity Indices
for representative soils in the study area using
modem precipitation and temperature data and
expanded treering chronologies. It also describes the processing of the elevation and soils
data for use in the GISintegrated analysis.
Finally, it details the assumptions and methods
used to link reconstructed annual soil moisture
levels, as modeled by the POSI, with levels of
agricultural production.
which the historic record is divided into two
blocks of time. One block is used for building
the regression model and the other is used for
testing the initial regression equation. This
second step represents the "initial calibration"
of the POSI to the treering data.
PALMER DROUGHT SEVERITY INDEX
RECONSTRUcnON
Third, the "verification" of the initial calibration equation takes place. Correlation coefficients and probability tests are used to assess the strength of the initial equations to
faithfully predict the actual values generated
by the instrumented POSI data for years not
used in the creation of the original calibration
equations.
The reconstruction of soil moisture by the
Palmer Drought Severity Index (POSI) is a
multistep process (Fritts 1976; Graybill 1989;
Hughes et al. 1982; Meko et al. 1980; Rose et
al. 1982). Subsequent to data acquisition, the
process begins with calculation of POSI values
for specific soils using historic climatic data
from a specific weather station. In this step,
actual POSI values are determined for every
month and every year in the instrumented
series.
Fourth, a "full calibration" period regression equation is created for the entire period of
instrumented record that overlaps with the
modem end of the expanded treering data.
The product of this step is the final transfer
function to be used to retrodict the preinstrumented POSI values. Because this regression
uses more years of data and often incoIporates
more climatic variation, it always produces
higher correlation coefficients, than does the
initial calibration.
Second, POSI values from a selected month
are correlated with treering values for a common period of time in order to generate an
initial multiple regression equation that can be
used as a transfer function to predict (or in this
case, "retrodict") POSI values in the prehistoric and preinstrumented time period. Here
the treering data are treated as independent
variables and the POSI value as the dependent
variable. In this second step, a portion of the
historic record is used to create the initial regression equation. Both splitrecord and random year methods have been used, but the
more common is the splitrecord approach in
Last, the retrodiction of the entire POSI series is accomplished by applying the transfer
function to the full set of treering values. Here
the treering data are used as the predictor or
independent variables and the POSI is the predicted or dependent variable.
PDSI Calculation
Input parameters for program "POSI"
include a) the total precipitation and mean
temperature values expressed as inches and
average degrees Fahrenheit per month for each
Table 3.1. Soils Selected to Represent the 11 Classes of Soil Moisture Based on Water Held
in the Upper and Lower Soil Layers.
Representative Soil
Type
XH2D
M2DD
ROLC
R7C
ROHC
VOCC
S5
V4C
V1B
ROB
R1C
Nonirrigated
Capability
Upper Soil
Class
Layer (in.)
7s
.53
6s
.85
3e
1.02
3e
1.02
1.02
3e
1.14
3c
1.02
4c
1.00
4e
1.05
4c
0.99
3c
1.14
4c
Lower Soil
Layer (in.)
4.61
2.95
3.65
4.48
5.40
6.44
7.80
8.54
9.45
10.26
13.66
weather station used, b) the total number of
years to be used in calculating PDSI values, c)
the monthly daylight adjustment values for the
latitude of the weather station, and d) the
amounts of water held in the upper soil layer
and lower soil layer expressed as inches. Experimentation with varying the amount of
water potentially held in the upper and lower
soil layers, using actual values for available
water capacity for each of the 113 soil map
units, revealed that quite different results would
obtain by using estimates other than the one
inch and five inch approximations suggested
by Palmer (1965). Therefore, it was decided to
use actual values for a reduced number of soils
rather than the standard model. This was done
by assigning each soil to a class based on water
held in the upper and lower layers, plotting
these values graphically, and selecting the single soil that fell closest to the center of its class
as being the representative of that group of
soils. In this manner, a total of 11 different
soils was selected to represent the 113 original
soils recorded in the Montezuma-Dolores
County area (Table 3.1, Figure 3.1).
The PDSI program was run 55 times,
calculating PDSI values for each of the 11
AWC soil classes with each of the five sets of
climatic data (Bluff, Ignacio, Cortez, Mesa
Verde, and Ft. Lewis).
An example output from one of these 55
PDSI calculations is presented in the original
study as appendix B (Van West 1990:384-
Upper Soil
Class (in.)
0.50 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
1.00 ±.25
Lower Soil
Class (in.)
0.00-6.50
2.00-2.99
3.00-3.99
4.00-4.99
5.00-5.99
6.00-6.99
7.00-7.99
8.00-8.99
9.00-9.99
10.00-10.99
13.00-13.99
AWC
Class
1
2
3
4
5
6
7
8
9
10
11
408). A single monthly PDSI table from each
of these programs is used as input to the next
program described below.
Calibration. verification. and Creation of the
Transfer Function
The program that performs the initial splitperiod calibration and verification as well as
the final full-period calibration, called "TERIFY," was developed by Donald Graybill and
Michael McCarthy of the Laboratory of TreeRing Research. Input parameters include: a)
identification of the named source of tree-ring
values used as the independent data set, b) a
beginning and ending date (year) for the independent data, which will also be the range of
the retrodicted data set, c) identification of
which variables in the tree-ring data set will be
used, d) specification of the maximum number
of independent variables from the tree-ring
data set that could be incorporated into multiple regression equations, e) specification of the
percentage of explained variance that must be
present in a variable before it is allowed to enter a multiple regression equation, f) identification of the named instrumented data source
of PDSI values used as the dependent data set
and which month is to be reconstructed, g)
identification of the beginning and ending
dates for the calibration process, h) selection of
the verification method to be used-split-record or random, and i) selection of the number
of years to be used in the initial calibration.
-56-
Available Water Capacity Of
The Lower Soil Layer (6-80 inches)
0.0"
1.0"
1.25"
2.0"
3.0"
4.0"
/2
/3
14
I
I
I
I
I
I
I
I
I
I
,
I
I
I
I
5.0"
6.0"
5
I6
I
I ROLe
I R7C
I ROLO
I RRTTOOI
I
I
I
; 100"
Q)
..c
o g
0
0.
o ....
U
I-C20
I
I C2eE
,C600
I
I
I
0
I
,0 C2v
セキ
·u
I
rolセQ
r_hァセ」ッ
e2F
I
1
セァ
I
hicセ
5T-
tibセQ
AOHe e
56C
I,
.1
I
I
I -HIDe I
\
1
I
I
I
I
I oM80
I
MGC.'
セ
.S6E
I
セ __ッウセ
セ
ᄚQセP
I
I
/
Mセ
A28
·S8C I
V4C
I
I
..
Qᄋ sセSQ
I
ZァセN
A41· 1•
I セッカRc
bRセ
12.0"
13.0"
AD8 I
RIC
RIO
mセ{i
0
I
..
I
I
"'48 VIO
b セG[
I"
I
I
I
I
セァャ
]ァセL
AOO
·SIC ,
I
I
,
I
I
I
I
oR88 I
I
I
I
I
I
I
I
VlS I
-M7E-
R)B I
A3Bl
I
_! セLュ
oZ
14.0"
"V18 jAO
VIC I
I A4: V2V I
OAS GキセZ
ZァセXG
I
I aセo
sセ・
I
ッセァ
1'
セNa c
I
.1 -R4CI
RHCj
160
r
11.0"
/0
I
I
I
I セ
R4C
10.0"
9
cS{セG
I
BXP」hセGZr
•
I
CT,O V2W
I
M2r%
M2001'
I
8
I
1
ウセᄋ
(t
9.0"
I
I
R40· A6A-·
_148
8.0"
I
I
_____ l __ J. __
.75"
e
セ
70"
-- -- ---- _!
Mセ
I
セ
J
_
R30,
Q)
>.
.COCE
.... 0
セNj
0_
セ
·0
Q)
Cf)
..0
....
o
セxhRP
.50"
eAIB
.MICE
Q)
0.
0.
·0
.M2CE
>:::>
<1
• PIC
• XMIF
Q)
セ
.XH)[
XH2E-
.25"
80
CP
GP
00"
RL
Soil AlB not plotted ..
/-/1
no dalo available
Soi I used 10 represent Qroup
Class desiQnatlon
Figure 3.1. Distribution of 113 soil map units in the study area by amount of water held in the upper and
lower soil layers.
This program was run 55 times, once for
each of data sets generated by the PDSI program. Except for specification of the name of
the weather station (Bluff, Cortez, Ignacio,
Mesa Verde, FtLewis) and the first year of
dependent data .available from the station,
(1928, 1931, 1915, 1924, 1922, respectively)
all other parameters were the same. All used
SWOLD7, the matrix of eigenvector amplitude
values, for the treeringbased independent
data. The full range of values available for
SWOLD7 was applied to the reconstruction
process, A.D. 9011970. Only axes 15 (Appendix A, columns 48) were considered since
they collectively represented the vast majority
of the variation in the PDSI data, where axis 1
(variable 5) is contained in the last column on
the right, and axis 5 (variable 1) is contained
by the fifth colwnn from the right (IMSL program output of principal component analysis).
In all cases, a variable was allowed to enter the
stepwise multiple regression equation only if it
explained a minimum of five percent of variance in the dependent variable, otherwise, it
57-
was dropped. In all cases, a splitrecord calibration was selected where the most recent 30
years were used to derive an initial calibration
and the remaining portion was used to verify
the results. The full period calibration had a
beginning date equivalent to the earliest date
used for the PDSI calculations and an end date
always at 1970, equivalent to the end date for
the tree-ring data.
Example output from TERIFY for one of
the 55 programs is presented in the original
study as appendix C (Van West 1990:409447). The regression coefficients and Y
intercept for the final full-period calibration
output by TERIFY are used as input for the
final program that predicts PDSI values.
Evaluating the Calibrations
Table 3.2 summarizes the results of the
calibration process following conventional
procedures (Graybill 1989). The topmost section of the three parallel sections summarizes
the results of the initial calibration period. It
provides the parameters generated by the multiple regression (the regression coefficients for
the variables used in the equation and the Y
intercept), the correlation coefficients (the
Pearson's product-moment correlation coefficient. r; the coefficient of detennination, r2;
and a more conservative coefficient of determination, the adjusted ,.1., "adjusted" for degrees of freedom), and the F value with its associated probability that the correlation between the two sets of data is zero (Le., there is
no correlation). The r value can range from
-1.00 to 1.00. High absolute values indicate a
strong relationship between the predictor treering series and the predicted PDSI series. A
positive r indicates that they covary positively,
a negative r indicates that they are inversely
related. Both ,.1. and adjusted ,.2 are measures of
how much variance in the PDSI values is explained by the tree-ring values, and thus their
power to predict the PDSI values. Adjusted r 2
values are always lower than unadjusted ,.1.
values and provide a more conservative estimate of the explained variance. The higher
these values, the better the fit between the
estimated modern PDSI values for particular
soil types and weather stations and the modern
climate data, as filtered through the tree-ring
data and reduced through principal compon-
ents analysis. The F value is a ratio of the explained variance to the unexplained variance,
adjusted for degrees of freedom. This F value
is interpreted using the probability of F(P > F)
which provides the probability that the true
correlation is zero. The smaller this value, the
better the chance that the locally estimated
PDSI values and the filtered. reduced regional
tree-ring data are in fact related in the population from which these samples were drawn.
The utility of the equation is questioned if the
Probability of F (P > F) value is greater than
.01.
The middle section of Table 3.2 summarizes the verification period tests of the initial
calibration. It assesses the ability of the initial
calibration model to estimate the actual PDSI
values for a portion of the historic record not
used in the creation of the initial calibration.
Thus, it is an independent test of the reliability
of the initial regression equation. The mean of
the actual PDSI values during that period (ACT
mean) and the mean reconstructed by the initial regression equation (REC mean) are provided, as are the correlation coefficients (r, r 2,
and the Probability of r, the P > r) associated
with the reconstruction during the verification
period. High r and ,.2 values are associated with
low probability (p > r)of r values, meaning
they have a low probability of not being correlated. Conversely, low rand r 2 values are
associated with probability of r (P > r) values
that are higher, indicating a greater chance that
the reconstructed PDSI values and the actual
PDSI values are uncorrelated. Reconstructions
with a (P > r) value greater than .05 (greater
than one chance in 20 that the data sets are
uncorrelated) indicate a poor correlation, probably caused by the small number of years used
in the verification period relative to the calibration period as well as by markedly different
climatic conditions characteristic of these two
periods.
Another test is the T-test for differences in
means. It compares the mean of the actual
PDSI values (ACT mean) and the mean of the
reconstructed PDSI values (REC mean) using a
T statistic to assess the probability (P >
that
both series derive from the same statistical
population. However, it may be questioned if
the sample size is small « 30). The smaller the
T value and the larger the P > T value, the
-58-
n
Table 3.2. Results and Statistical Summary of the Initial Calibration, Verification, and Final, FullPeriod Calibration Analyses for 55 Soil Moisture Groups.
RECONSTRUCTION 1 B 0.53 4.61 *
1941-1970
Calibration Period
30 (70%)
No. of Years I Percent
Regression
-.974
Coefficients : V5
-1.036
V4
V3
V2
V1
-1.060
.479
Y Intercept
.778
f
,2.606
.591
Adjusted ,213.325
F
P>F
<.000
RECONSTRUCTION 1 B. 0.53 4.61
Verification Period
1928-1940
No of Years I Percent
13 (30%)
.588
f
.346
,2P>f
.033
-.57
ACT mean
REC mean
-.22
T
.536
P>T
.607
Ws
-.594
P>Z
.276
RECONSTRUCTION 1 B. 0.53 4.61
1928-1970
Calibration Period
No. of Years I Percent
43 (100%)
Regression
-.954
Coefficients : V5
V4
-.900
V3
-1.560
V2
V1
Y Intercept
-.556
.781
.610
,2.600
Adjusted ,2F
20.352
P>F
<.000
,
Represe ntation
in Study Area
1.52%
INITIAL CALIBRATION
2 B 0.85 2.95 3 B 1.02 3.65
1941-1970
1941-1970
30 (70%)
30 (70%)
4 B 1.02 4.48
1941-1970
30 (70%)
-.932
-1.290
-.898
-.930
-.921
-.849
-1.116
-1.232
-.422
.793
.628
.614
14.656
<.000
-.524
.743
.552
.535
16.626
<.000
-.607
.769
.591
.575
12.525
<.000
VERIFICATION
2 B 0.85 2.95 3 B 1.02 3.65
1928-1940
1928-1940
13 (30%)
13 (30%)
.485
.702
.493
.235
.090
.007
-1.30
-.62
-.18
-.46
1.932
.194
.075
.842
-1.852
-.314
.377
.032
4 B 1.02 4.48
1928-1940
13 (30%)
.812
.659
.001
-.48
-.58
.115
.905
-.384
.350
FINAL CALIBRATION
2 B 0.85 2.95 3 B 1.02 3.65
1928-1970
1928-1970
43 (100%)
43 (100%)
4 B 1.02 4.48
1928-1970
43 (100%)
-.739
-1.419
-.937
-.912
-1.504
-.948
-.865
-1.647
-.787
.719
.517
.505
21.387
<.000
-.559
.770
.606
.596
20.011
<.000
-.571
.781
.610
.600
20.359
<.000
0.42
-59-
NA
NA
Table 3.2 (Continued).
RECONSTRUCTION 5 B 1.02 5.40·
Calibration Period
19411970
No. of Years / Percent
30 (70%)
Regression
.883
Coefficients : V5
.888
V4
V3
V2
V1
.631
Y Intercept
.742
f
.551
,2
.535
Adjusted ,2
16.578
F
<.000
P>F
RECONSTRUCTION
Verification Period
No of Years / Percent
f
,2
P>f
ACT mean
REC mean
T
P>T
Ws
P>Z
5 B 1.02 5.40
19281940
13 (30%)
.672
.451
.012
.41
.56
.170
.861
.035
.486
RECONSTRUCTION 5 B 1.02 5.40
19281970
Calibration Period
43 (100%)
No. of Years / Percent
Regression
.952
Coefficients : V5
V4
1.916
V3
V2
V1
.481
Y Intercept
.746
f
.557
,2.546
Adjusted ,2
F
22.125
P>F
<.000
Representation
in Study Area
.27%
INITIAL CALIBRATION
6 B 1.14 6.44
7 B 1.027.80
19411970
19411970
30 (70%)
30 (70%)
8 B1.008.54
19411970
30 (70%)
.872
.849
.883
.766
1.051
.874
.753
1.046
.636
.742
.551
.534
16.555
<.000
.652
.767
.588
.572
12.382
<.000
.664
.767
.589
.573
12.401
<.000
VERIFICATION
7 B 1.027.80
6 B 1.14 6.44
19281940
19281940
13 (30%)
13 (30%)
.738
.640
.410
.545
.004
.018
.19
.39
.56
.62
.463
.196
.653
.841
1.013
.035
.486
.155
8 B1.008.54
19281940
13 (30%)
.722
.521
.005
.16
.63
.502
.626
1.013
.155
FINAL CALIBRATION
6 B1.14 6.44
7 B 1.027.80
19281970
19281970
43 (100%)
43 (100%)
8 B 1.008.54
19281970
43 (100%)
.935
.910
.898
1.929
1.916
1.924
.484
.742
.551
.540
24.546
<.000
.440
.733
.538
.526
23.254
<.000
.442
.729
.532
.520
22.700
<.000
NA
-60-
.04%
.47%
Table 3.2 (Continued).
RECONSTRUCTION
9 B 1.05 9.45*
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
1941-1970
30 (70%)
f
,2Adjusted ,2F
P>F
-.864
-.742
-1.036
-.856
-.733
-1.027
-.904
-.831
-1.184
-.677
.767
.588
.572
12.373
<.000
-.683
.767
.588
.572
12.368
<.000
-1.072
-.685
.814
.663
.650
12.306
<.000
RECONSTRUCTION
9 B 1.05 9.45
Verification Period
No of Years I Percent
1928-1940
13 (30%)
.699
.489
.008
-.12
f
,2-
P>f
ACT mean
REC mean
T
P>T
Ws
P>Z
-.64
.555
.592
-1.013
.155
RECONSTRUCTION
9B1.059.45
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
1928·1970 .
43 (100%)
f
,2Adjusted ,2-
F
P>F
Representation
in Study Area
INITIAL CALIBRATION
11 B 1.14
10 B 0.99
10.26
13.66
1941-1970
1941-1970
30 (70%)
30 (70%)
VERIFICATION
11B1.14
10 B 0.99
10.26
13.66
1928-1940
1928-1940
13 (30%)
13 (30%)
.411
.682
.169
.465
.010
.160
-.09
.05
-.65
-.47
.596
.496
.565
.630
-1.013
-.314
.155
.377
FINAL CALIBRATION
10 B 0.99
11 B 1.14
10.26
13.66
1928-1970
1928-1970
43 (100%)
43 (100%)
12 C 0.53 4.61
1941-1970
30 (75%)
-.858
-1.254
-.073
.766
.587
.572
19.188
<.000
12 C 0.53 4.61
1931-1940
10 (25%)
.600
.361
.065
-.47
-.05
.602
.568
-.866
.193
12 C 0.53 4.61
1931·1970
40 (100%)
-.884
-.875
-.835
-1.929
-1.932
-1.924
-.441
.722
.522
.510
21.821
<.000
-.436
.717
.514
.502
21.169
<.000
-.443
.692
.479
.466
18.403
<.000
-1.167
.736
.541
.529
21.848
<.000
NA
NA
NA
18.38%
-
61-
-.852
-1.130
Table 3.2 (Continued).
INITIAL CALIBRATION
RECONSTRUCTION 13 C .85 2.95* 14 C 1.02 3.65 15 C 1.02 4.48 16 C 1.02 5.40
1941-1970
Calibration Period
1941-1970
1941-1970
1941-1970
30 (75%)
No. of Years I Percent
30 (75%)
30 (75%)
30 (75%)
Regression
-.850
-.874
-.732
Coefficients : V5
-.871
-1.266
-1.090
V4
-1.286
-1.065
V3
V2
V1
-.118
-.074
-.306
-.128
Y Intercept
.751
.770
.751
.743
f
.593
.563
.564
.553
,2.578
.547
.548
.536
Adjusted ,219.637
17.419
17.491
16.685
F
<.000
<.000
<.000
<.000
P>F
VERIFICATION
RECONSTRUCTION 13 C 0.85 2.95 14 C 1.02 3.65 15 C 1.02 4.48 16 C 1.02 5.40
1931-1940
1931-1940
1931-1940
1931-1940
Verification Period
10 (25%)
10 (25%)
10 (25%)
10 (25%)
No of Years I Percent
.598
.609
f
.473
.590
.370
.348
.224
.358
t2
.060
.066
.070
.165
P>f
-.47
-.44
-.38
ACT mean
-.70
-.06
-.05
-.06
-.33
REC mean
.592
.541
.498
.426
T
.607
.634
.574
.681
P>T
-.561
-.459
-.866
-.459
Ws
.193
.323
.288
.323
P>Z
FINAL CALIBRATION
RECONSTRUCTION 13 C 0.85 2.95 14 C 1.02 3.65 15 C 1.02 4.48 16 C 1.02 5.40
1931-1970
1931-1970
1931-1970
1931-1970
Calibration Period
40 (100%)
40 (100%)
40 (100%)
40 (100%)
No. of Years I Percent
Regression
-.718
-.844
-.867
-.868
Coefficients : V5
V4
-1.128
-1.138
-.981
-.948
V3
V2
V1
-.165
-.203
-.195
-.383
Y Intercept
.739
.724
f
.703
.714
.494
.546
.524
,2.510
.534
.511
.480
.497
Adjusted ,222.238
20.358
F
18.030
19.269
<.000
<.000
P>F
<.000
<.000
Representation
in Study Area
7.41%
.67%
-62-
.50%
1.23%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2F
P>F
17 C 1.14
6.44*
1941-1970
30 (75%)
-.928
-.915
-.907
-.899
.225
.677
.459
.439
23.730
<.000
.191
.663
.440
.420
21.967
<.000
.180
.660
.435
.415
21.577
<.000
.118
.657
.432
.412
21.322
<.000
RECONSTRUCTION 17 C 1.14 6.44
Verification Period
No of Years / Percent
f
,2P>f
ACT mean
REC mean
T
P>T
Ws
P>Z
1931-1940
10 (25%)
.619
.384
.055
-.29
.55
.829
.575
-1.172
.121
RECONSTRUCTION 17 C 1.14 6.44
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2F
P>F
Representation
in Study Area
INITIAL CALIBRATION
18 C 1.02
19 C 1.00 8.54 20 C 1.05 9.45
7.80
1941-1970
1941-1970
1941-1970
30 (75%)
30 (75%)
30 (75%)
1931·1970
40 (100%)
VERIFICATION
18 C 1.02
19 C 1.00 8.54 20 C 1.05 9.45
7.80
1931·1940
1931-1940
1931-1940
10 (25%)
10 (25%)
10 (25%)
.594
.584
.573
.353
.341
.329
.068
.074
.081
-.25
-.22
-.18
.51
.43
.50
.744
.698
.592
.526
.502
.569
-.968
-.866
-.663
.166
.193
.254
FINAL CALIBRATION
18 C 1.02
19 C 1.00 8.54 20 C 1.05 9.45
7.80
1931-1970
1931-1970
1931-1970
40 (100%)
40 (100%)
40 (100%)
-.921
-.915
-.908
-.902
-1.563
-1.410
-1.383
-1.350
.025
.716
.512
.499
12.607
<.000
.017
.682
.466
.451
16.113
<.000
.017
-.018
.674
.454
.439
15.395
<.000
-.n3
.57%
2.40%
-63 -
.6n
.459
.444
15.678
<.000
5.11%
.67%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2
Adjusted ,2
F
P>F
RECONSTRUCTION
Verification Period
No of Years I Percent
f
,2
P>f
ACT mean
REC mean
T
P>T
Ws
P>Z
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
,
,2Adjusted ,2F
P>F
Representation
in Study Area
21 C .99
10.26*
1941-1970
30 (75%)
INITIAL CALIBRATION
22 C 1.14
23 I 0.53 4.61
13.66
1941-1970
1941-1970
30 (75%)
30 (54%)
24 I 0.85 2.95
1941-1970
30 (54%)
-.889
-.855
-.724
-1.226
-1.219
-.692
.091
.654
.428
.407
20.911
<.000
.033
.646
.417
.397
20.056
<.000
-.453
.770
.593
.578
19.686
<.000
-.446
.772
.596
.581
19.892
<.000
21 C 0.99
10.26
1931-1940
10 (25%)
.565
.319
.087
-.15
.40
.531
.609
.663
.254
VERIFICATION
22 C 1.14
23 I 0.53 4.61
13.66
1931-1940
1915-1940
10 (25%)
26 (46%)
.541
.677
.292
.459
.104
.000
.06
.21
.33
.36
.260
.229
.794
.815
-.357
-.343
.361
.366
21 C 0.99
10.26
1931-1970
40 (100%)
FINAL CALIBRATION
22 C 1.14
23 I 0.53 4.61
13.66
1931-1970
1915-1970
40 (100%)
56 (100%)
-.864
-.051
.632
.399
.383
25.212
<.000
4.03%
-.842
.645
24 I 0.85 2.95
1915-1940
26 (46%)
.663
.440
.000
.07
.34
.421
.679
-.622
.267
24 I 0.85 2.95
1915-1970
56 (100%)
-1.729
-.612
-1.633
-1.217
-1.233
-.037
.624
.389
.373
24.228
<.000
-.548
.782
.611
.604
27.263
<.000
-.574
.778
.606
.598
26.606
<.000
NA
10.69%
-64-
5.97%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2F
P>F
25 I 1.02
3.65*
1941-1970
30 (54%)
INITIAL CALIBRATION
26 I 1.02 4.48 27 I 1.02 5.40
1941-1970
30 (54%)
1941-1970
30 (54%)
1941-1970
30 (54%)
-.698
-1.186
-.710
-1.132
-.715
-1.062
-.714
-1.048
-.384
.754
.568
.552
17.763
<.000
-.392
.743
.553
.536
16.685
<.000
-.496
.723
.523
.505
14.790
<.000
-.624
.718
.516
.498
14.381
<.000
VERIFICATION
RECONSTRUCTION 25 I 1.02 3.65 26 I 1.02 4.48 27 I 1.02 5.40
1915-1940
1915-1940
1915-1940
Verification Period
26 (46%)
26 (46%)
26 (46%)
No of Years / Percent
.674
.676
.678
f
.454
.457
.459
,2.000
.000
.000
P>f
ACT mean
.16
.20
.28
.40
.27
REC mean
.38
.274
.017
T
.365
.717
.781
.984
P>T
-.597
Ws
-.394
-.063
.275
.347
.475
P>Z
FINAL CALIBRATION
RECONSTRUCTION 25 I 1.02 3.65 26 I 1.02 4.48 27 I 1.02 5.40
Calibration Period
1915-1970
1915-1970
1915-1970
No. of Years / Percent
56 (100%)
56 (100%)
56 (100%)
Regression
-.620
-.636
Coefficients : V5
-.656
V4
-1.703
-1.693
-1.676
V3
V2
-1.236
-1.204
-1.172
V1
-.522
Ylntercept
-.516
-.546
r
.773
.767
.759
.598
,2
.589
.576
.590
.581
.567
Adjusted ,2
F
25.753
24.790
23.503
P>F
<.000
<.000
<.000
Representation
in Study Area
28 I 1.14 6.44
.35%
.33%
-65 -
1.22%
28 I 1.14 6.44
1915-1940
26 (46%)
.671
.450
.000
.27
.13
.196
.840
-.140
.444
28 I 1.14 6.44
1915-1970
56 (100%)
-.691
-1.448
-.653
.711
.505
.496
27.056
<.000
.90%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2
Adjusted ,2
F
P>F
29 I 1.02
7.80·
1941-1970
30 (54%)
,2
P>f
ACT mean
REC mean
T
P>T
Ws
P>Z
-.697
.719
.517
.499
14.447
<.000
1915-1940
26 (46%)
.653
.427
.000
.42
.06
.523
.610
.648
.259
RECONSTRUCTION 29 I 1.02 7.80
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2F
P>F
Representation
in Study Area
1941-1970
30 (54%)
1941-1970
30 (54%)
32 10.99
10.26
1941-1970
30 (54%)
-.704
-.980
-.834
-.888
-1.562
-1.859
-1.250
-.349
.722
.521
.504
14.712
<.000
-.195
.746
.557
.540
10.907
<.000
-.713
-1.032
RECONSTRUCTION 29 , 1.02 7.80
Verification Period
No of Years / Percent
f
INITIAL CALIBRATION
30 I 1.008.54 31 I 1.05 9.45
1915-1970
56 (100%)
-.701
.707
.500
.481
13.483
<.000
VERIFICATION
30 I 1.00 8.54 31 I 1.05 9.45
1915-1940
26 (46%)
.641
.411
.000
.45
.03
.610
.553
-.698
.242
1915-1940
26 (46%)
.572
.327
.002
.47
.14
.483
.637
-.292
.385
FINAL CALIBRATION
30 I 1.008.54 31 I 1.05 9.45
1915-1970
56 (100%)
1915-1970
56 (100%)
32 10.99
10.26
1915-1940
26 (46%)
.544
.296
.004
.44
.18
.348
.729
-.394
.347
32 10.99
10.26
1915-1970
56 (100%)
-.805
-1.202
-1.576
-1.016
-.728
-1.597
-1.284
-1.057
-.771
-1.362
-1.712
-1.203
-.627
.742
.550
.541
21.178
<.000
-.565
.768
.590
.582
18.361
<.000
-.473
.738
.545
.537
20.784
<.000
-.382
.749
.560
.552
16.254
<.000
-.656
-1.768
1.32%
1.18%
66-
.56%
10.93%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2F
P>F
RECONSTRUCTION
Verification Period
No of Years I Percent
f
,2P>f
ACT mean
REC mean
T
P>T
Ws
P>Z
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2-
F
P>F
Representation
in Study Area
33 I 1.14
13.66·
1941-1970
30 (54%)
-.866
INITIAL CALIBRATION
34 M 0.53 4.61 35 M 0.85 2.95 36 M 1.02 3.65
1941-1970
30 (64%)
1941-1970
30 (64%)
1941-1970
30 (64%)
-.660
-1.194
-.598
-1.097
-.651
-1.182
-.317
.824
.680
.668
28.654
<.000
-.239
.766
.587
.572
19.178
<.000
-.292
.828
.686
.674
29.482
<.000
-1.822
-1.298
.289
.753
.568
.551
11.375
<.000
33 I 1.14
13.66
1915-1940
26 (46%)
.509
.259
.008
.45
.07
.526
.607
-.521
.301
33 I 1.14
13.66
1915-1970
56 (100%)
VERIFICATION
34 M 0.53 4.61 35 M 0.85 2.95 36 M 1.02 3.65
1924-1940
17 (36%)
.619
.383
.008
-.27
.00
.374
.712
-.497
.310
1924-1940
17 (36%)
.604
.365
.010
-.32
.04
.538
.601
-.686
.246
1924-1940
17 (36%)
.606
.367
.010
-.29
.02
.432
.673
-.450
.326
FINAL CALIBRATION
34 M 0.53 4.61 35 M 0.85 2.95 36 M 1.02 3.65
1924-1970
47 (100%)
1924-1970
47 (100%)
1924-1970
47 (100%)
-.749
-1.260
-1.627
-1.136
-.622
-1.406
-.566
-1.286
-.609
-1.393
-.403
.728
.530
.521
14.364
<.000
-.454
.747
.558
.548
27.787
<.000
-.410
.703
.494
.482
21 WX セ
<.000
-.443
.742
.551
.541
27.007
<.000
NA
2.94%
-67-
1.63%
.11%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2
Adjusted ,2
F
P>F
37 M 1.02
4.48·
1941-1970
30 (64%)
INITIAL CALIBRATION
38 M 1.02 5.40
39 M 1.14
6.44
1941-1970
1941-1970
30 (64%)
30 (64%)
-.674
-1.183
-.692
-1.086
-.751
-1.152
-.810
-1.042
-.264
.832
.691
.680
30.251
<.000
-.218
.819
.671
.659
27.569
<.000
-.220
.851
.724
.714
35.382
<.000
-.201
.865
.748
.739
40.050
<.000
VERIFICATION
39 M 1.14
RECONSTRUCTION 37 M 1.02 4.48 38 M 1.02 5.40
6.44
1924-1940
1924-1940
1924-1940
Verification Period
17 (36%)
17 (36%)
17 (36%)
No of Years / Percent
.628
.628
f
.622
.394
.395
,2
.387
.007
.007
P>f
.007
-.36
-.34
-.28
ACT mean
.12
.09
.05
REC mean
.614
.676
.461
T
.512
.551
P>T
.653
-1.065
-1.018
.497
Ws
.143
.154
P>Z
.310
FINAL CALIBRATION
39 M 1.14
RECONSTRUCTION 37 M 1.02 4.48 38 M 1.02 5.40
6.44
1924-1970
1924-1970
1924-1970
Calibration Period
47 (100%)
47 (100%)
47 (100%)
No. of Years / Percent
Regression
-.625
-.661
-.628
Coefficients : V5
-1.442
-1.510
V4
-1.428
V3
V2
V1
.444
-.431
-.467
Y Intercept
.754
.781
.753
f
.567
.569
.610
,2
.601
.557
.559
Adjusted ,2
34.340
28.989
28.785
F
<.000
<.000
P>F
<.000
Representation
in Study Area
40 M 1.02
7.80
1941-1970
30 (64%)
.70%
2.14%
68 -
.57%
40 M 1.02
7.80
1924-1940
17 (36%)
.612
.375
.009
-.38
.14
.727
.520
-1.207
.114
40 M 1.02
7.80
1924-1970
47 (100%)
-.704
-1.454
-.474
.790
.624
.616
36.535
<.000
.95%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
r
,2
Adjusted ,2
F
P>F
RECONSTRUCTION
Verification Period
No of Years I Percent
r
,2-
P>r
ACT mean
REC mean
T
P>T
Ws
P>Z
41 M 1.00
8.54*
1941-1970
30 (64%)
r
,2Adjusted ,2-
F
P>F
Representation
in Study Area
44 M 1.14 13.66
1941-1970
30 (64%)
-.798
-1.033
-.766
-1.057
-.765
-1.045
-.733
-.961
-.321
.844
.712
.702
33.414
<:.000
-.393
.823
.677
.665
28.292
<:.000
-.419
.820
.672
.659
27.616
<:.000
-.410
.758
.575
.559
18.233
<:.000
41 M 1.00
8.54
1924-1940
17 (36%)
.615
.378
.008
-.37
.02
.539
.600
-1.018
.154
RECONSTRUCTION 41 M 1.00 8.54
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
INITIAL CALIBRATION
42 M 1.05
43 M. 0.99
9.45
10.26
1941-1970
1941-1970
30 (64%)
30 (64%)
1924-1970
47 (100%)
VERIFICATION
42 M 1.05
43 M. 0.99 10.26 44 M 1.14 13.66
9.45
1924-1940
1924-1940
1924-1940
17 (36%)
17 (36%)
17 (36%)
.624
.620
.651
.389
.384
.424
.007
.008
.005
-.37
-.37
-.17
-.06
-.09
-.10
.433
.397
.096
.672
.696
.921
-.876
-.876
-.308
.191
.191
.379
FINAL CALIBRATION
42 M 1.05
43 M. 0.99
9.45
10.26
1924-1970
1924-1970
47 (100%)
47 (100%)
-.6n
-.700
-1.449
-1.480
-.545
.780
.609
.600
34.250
<:.000
-.589
.769
.591
.582
31.838
<:.000
1.28%
.73%
-69-
-.675
-.1482
-.609
.767
.588
.578
31.352
<:.000
9.13%
44 M 1.14 13.66
19221924-1970
47 (100%)
-.686
-1.471
-.530
.734
.539
.528
25.696
<:.000
.28%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years I Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
f
,2Adjusted ,2F
P>F
45 F 0.53
4.61*
1941-1970
30 (61%)
INITIAL CALIBRATION
46 F 0.85 2.95 47 F 1.02 3.65 48 F 1.02 4.48
1941-1970
30 (61%)
1941-1970
30 (61%)
1941-1970
30 (61%)
-.613
-.811
-.585
-.751
-.604
-.785
-.614
-.806
.100
.672
.451
.431
11.088
<.000
.075
.671
.450
.430
11.059
<.000
.100
.675
.455
.435
11.271
<.000
.118
.674
.454
.433
11.211
<.000
VERIFICATION
RECONSTRUCTION 45 F 0.53 4.61 46 F 0.85 2.95 47 F 1.02 3.65 48 F 1.02 4.48
1922-1940
1922-1940
1922-1940
1922-1940
Verification Period
19 (39%)
19 (39%)
19 (39%)
19 (39%)
No of Years I Percent
.478
.483
.483
f
.486
.233
.228
.237
.233
,2.035
.037
.035
P>f
.033
-.31
-.38
-.37
-.38
ACT mean
.28
.31
REC mean
.32
.33
1.273
1.940
T
1.963
1.983
.210
.065
P>T
.063
.060
-1.730
-2.012
Ws
-1.972
-2.012
.042
.022
.024
.022
P>Z
FINAL CALIBRATION
RECONSTRUCTION 45 F 0.53 4.61 46 F 0.85 2.95 47 F 1.02 3.65 48 F 1.02 4.48
1922-1970
1922-1970
1922-1970
1922-1970
Calibration Period
49 (100%)
49 (100%)
49 (100%)
No. of Years I Percent
49 (100%)
Regression
-.548
-.521
-.538
Coefficients : V5
-.547
-.644
-.665
V4
-.630
-.638
V3
V2
V1
-.148
-.149
-.150
-.133
Ylntercept
.607
.609
.608
.609
f
.370
.369
.371
.370
,2.355
.357
.356
.357
Adjusted ,213.524
13.454
F
13.486
13.574
<.000
<.000
P>F
<.000
<.000
Representation
in Study Area
.29%
NA
-70-
.03%
.37%
Table 3.2 (Continued).
RECONSTRUCTION
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
(
,2Adjusted ,2F
P>F
49 F 1.02
5.40·
1941-1970
30 (61%)
-.636
-.832
-.680
-1.010
-.693
-1.026
-.703
-1.014
.083
.674
.454
.434
11.239
<.000
-.002
.703
.495
.476
13.227
<.000
-.022
.697
.486
.467
12.751
<.000
0.036
.696
.484
.465
12.661
<.000
RECONSTRUCTION 49 F 1.02 5.40
Verification Period
No of Years / Percent
(
,2-
P>(
ACT mean
REC mean
T
P>T
Ws
P>Z
1922-1940
19 (39%)
.471
.222
.040
-.36
.31
1.902
.070
-1.972
.024
RECONSTRUCTION 49 F 1.02 5.40
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
(
,2
Adjusted ,2
F
P>F
Representation
in Study Area
INITIAL CALIBRATION
50 F 1.14
51 F 1.027.80 52 F 1.008.54
6.44
1941-1970
1941-1970
1941-1970
30 (61%)
30 (61%)
30 (61%)
1922-1970
49 (100%)
VERIFICATION
50 F 1.14
51 F 1.027.80 52 F 1.008.54
6.44
1922-1940
1922-1940
1922-1940
19 (39%)
19 (39%)
19 (39%)
.469
.477
.482
.220
.228
.233
.041
.037
.035
-.32
-.34
-.32
.24
.26
.24
1.529
1.629
1.499
.117
.140
.148
-1.690
-1.650
-1.610
.045
.049
.054
FINAL CALIBRATION
50 F 1.14
51 F 1.027.80 52 F 1.008.54
6.44
1922-1970
1922-1970
1922-1970
49 (100%)
49 (100%)
49 (100%)
-.557
-.646
-.593
-.769
-.604
-.783
-.608
-.785
-.155
.604
.371
.358
13.585
<.000
-.200
.636
.405
.392
15.657
<.000
-.206
.635
.403
.390
15.533
<.000
-.217
.635
.404
.391
15.566
<.000
.11%
NA
-71-
.04%
.57%
Table 3.2 (Concluded).
INITIAL CALIBRATION
RECONSTRUCTION 53 F 1.04 9.45* 54 F 0.99 10.26 55 F 1.14 13.66
Calibration Period
1941-1970
1941-1970
1941-1970
No. of Years / Percent
30 (61%)
30 (61%)
30 (61%)
Regression
-.707
Coefficients : V5
-.706
-.697
V4
-1.000
-.986
-.930
V3
V2
V1
-.055
-.045
-.077
Y Intercept
.686
.690
.665
f
.477
.470
,2.442
.451
.457
.422
Adjusted ,211.988
10.701
F
12.296
<.000
P>F
<.000
<.000
53 F 1.04 9.45
1922-1940
19 (39%)
.456
f
,2.208
P>f
.048
ACT mean
-.32
REC mean
.22
T
1.459
P>T
.159
Ws
-1.569
P>Z
.058
RECONSTRUCTION
Verification Period
No of Years / Percent
RECONSTRUCTION
Calibration Period
No. of Years / Percent
Regression
Coefficients : V5
V4
V3
V2
V1
Y Intercept
r
,2
Adjusted ,2
F
P>F
Representation
in Study Area
53 F 1.04 9.45
1922-1970
49 (100%)
VERIFICATION
54 F 0.99 10.26 55 F 1.14 13.66
1922-1940
1922-1940
19 (39%)
19 (39%)
.435
.300
.189
.090
.060
.209
-.26
-.34
.20
.17
1.217
1.276
.238
.216
-1.248
-1.167
.106
.122
FINAL CALIBRATION
54 F 0.99 10.26 55 F 1.14 13.66
1922-1970
1922-1970
49 (100%)
49 (100%)
-.606
-.760
-.604
-.748
-.559
-.687
-.224
.626
.392
.378
14.798
<.000
-.205
.620
.384
.371
14.334
<.000
-.250
.575
.331
.316
11.361
<.000
.05%
.86%
.
.11 %
*indicates Awe values used in reconstruction; the letter represents the station used and the numbers
represent the potential inches of water that could be stored in upper and lower soil layers. NA. not
appropriate, reconstruction does not actually occur in final study area.
-72-
better (Le.• the closer the means are to each
other. the higher the probability that they both
derive from the same statistical population of
values and the better the equation that generated the reconstructed mean). If the P > T
value is less the .05 then it is considered to be a
poor reconstruction. An alternative test that is
less sensitive to sample size is the Wilcoxon
match-pairs signed-rank test (Ws). The magnitude of the Ws statistic is uninterpretable
without its probability. the Probability of Z (P
> Z.) The larger the P > Z value. the better the
calibration equation is at replicating the actual
mean of verification period. With this test. an
initial calibration equation can be rejected as
being inadequate to predict PDSI values if the
P > Z value is less than .05 (Le.• the REC mean
is statistically different than the ACT mean). In
sum. it is the results of the three tests. the probabilities of r, T, and Z, along with a consideration of the number of years included in the
verification period, that are used to assess the
reliability of the initial model.
The lowermost section of Table 3.2 summarizes the results of the final, full-period calibration. The same values are presented here as
for the initial calibration period and are interpreted by the same criteria. Because more
years are incorporated in the sample and they
embody a more representative set of climatic
Table 3.3. Adjusted セ
conditions, the correlation coefficients are consistently higher. their probabilities of not being
uncorrelated are greater. and their ability to
predict actual PDSI values are more substantial.
Likewise, their F values are consistently higher
than their initial calibration counterparts.
Results
All 55 equations produced adequate initial
calibrations (P > F of less than .000 were obtained without exceptions). When tested with
independent data during the verification period, the vast majority (45/55 or 62%) passed all
three tests for replicative adequacy. None
failed all three tests and only one failed two
tests (Reconstruction 2. Bluff .534.61). This is
the single reconstruction with the lowest total
available water capacity (AWe) of any of the
55. Nine failed to meet one of the three probability levels suggested as acceptable margins
of error (Reconstructions 11, 12, LYセUT
54,
55), although none seriously. In each of these
cases, the reconstruction represents extremes in
water-holding capacity for its elevation. either
as very low amounts of water or very great
amounts of water in the combined upper and
lower soil layers. The reconstructions representing the middle range of AWC values consistently passed every verification period in all
five elevational ranges.
Values for 55 Full-Period Calibrations.
Bluff
AWC AWC Values
Cortez Ignacio Mesa Verde Ft. Lewis
1315 m 1893 m 1969 m 2169 m
Class (Su +SI)a
2317 m
60
53
55
0.53 4.61
60
36
1
51
48
0.85
2.95
48
36
2
60
1.02 3.65
60
53
59
54
3
36
4
4.48
60
51
58
56
1.02
36
55
50
57
5
1.02 5.40
56
36
54
50
50
1.14 6.44
60
6
39
53
45
54
1.02 7.80
62
7
39
44
1.00 8.54
52
8
58
60
39
44
51
54
1.05 9.45
58
9
38
10.26
50
55
10
0.99
38
58
37
47
1.14 13.66
52
11
37
53
32
54.82
46.63
Mean
56.09
56.36
36.73
S.D.
5.19
5.55
3.33
3.91
2.05
aAvaiiable water-.holding capacity of upper six inches of soil and lower soil profile.
-73 -
Mean±S.D.
52.80 ±9.88
48.60 ± 8.59
52.40 ±9.66
52.20 ±9.65
50.80 ±8.70
50.60 ± 7.67
50.60 ±8.85
50.60 ±8.99
49.00 ±8.00
47.60 ±9.66
44.20 ±9.31
49.95
8.49
Finally, all 55 fullperiod calibrations,
without exception, produced significant correlations, again with the probability being less
than .000 that the correlation was zero. Table
3.3 lists the conservative estimate for the explained variance (the adjusted ,2) for each of
these 55 regression equations. In this data set,
the adjusted ,2 consistently is about 1% less
than the unadjusted ,2. As a group, they range
from 32-62% with an overall mean of 50%.
This indicates that about 50% of the variation
in a PDSI value can be explained by the treering data, which is here considered to be a
proxy for stochastic variation in climate.
Summarized by elevation group, the five
sets had adjusted ,2 values that ranged from
37-56%. Ignacio and Mesa Verde reconstructions proved to be the best predictors of PDSI
as these were the stations that had the longest
sets of climatic records (56 and 47 years, respectively). Bluff is next, not only by virtue of
the number of years in the sample (43 years),
but also because it is the lowest elevation and
tree-ring based reconstructions in the American Southwest are most sensitive in arid settings. Cortez placed fourth primarily because
of the limited time depth of its climatic records
(40 years) and Ft Lewis placed last, not unexpectedly, because as the highest elevational
zone it is less sensitive. Summarized by the 11
classes of soil moisture, the adjusted Tl ranged
from 44-53%. The middle and lower groufs
produced equations with higher adjusted,
than the highest groups, not surprisingly.
a mean of 58% (Rose et all 1982:252).
Retrodiction of PDSI Values A.D. 901-1970
The program that reconstructs the entire
time series of PDSI values from A.D. 901 to
970 is called "RECON" and was written by G.
Robert Lofgren of the Laboratory of TreeRing Research at the University of Arizona. In
the case of the present study, it uses the 1070year series of tree-ring values contained in
SWOLD7 as the annually varying independent
variables and the regression coefficients and
the Y intercept from the full-period calibration
derived by TERIFY as parameters.
It also generates some descriptive statistics
as a check on both input and output data and
creates a file usable for plotting reconstructed
values. The output from this program was formatted in such a way that it would be readily
acceptable to the EPPL7 GIS program. Consequently, the output files were created with 55
rows, each representing one of the long-term
PDSI reconstructions, and 20 columns, each
column representing a single year (e.g., A.D.
901-920, 921-940, 941-960, etc.). A total of
54 files were created, all with the 55 row by 20
column format with the exception of the last
file, which contains the final 10 years of data
(A.D. 1961-1970). These data are presented in
the original study (Van West 1990:448-555)
as appendix D.
Evaluating the Reconstructions
This range compares favorably with adjusted ,2 values obtained by Rose in a reconstruction of July PDSI values for the Zuni area
(using Black Rock and the combined Southwest Mountains weather stations: Ft Wingate,
EI Morro, Bluewater, Grants, San Fidel, and
Marquez) and Southwestern Colorado (using
the combined Southwest Colorado weather
stations: Cortez, Mancos, Mesa Verde, Ft
Lewis, Ignacio, and Pagosa Springs) using
similar tree ring data for essentially the same
l,07Q-year period. For the Zuni area Rose
obtained adjusted ,2 values ranging from 4554% with a mean of 50% (Martin Rose personal communication, April 1989) and for the
combined reconstructions of Southwestern
Colorado he obtained a range of 48-71 % with
Following conventions established at the
Laboratory of Tree-Ring Research (Graybill
1989), Table 3.4 summarizes for each of the
55 reconstructions the descriptive statistics associated with the final, full-period calibrations
and the long-term reconstructions. The final,
full-period calibrations include statistics for the
actual June PDSI values calculated by the original PDSI programs as well as statistics for the
reconstructed series generated by the final
mUltiple regression equation that was used as
the transfer .function. It summarizes descriptive
statistics associated with the 1070-year reconstruction (A.D. 901-1970) and the first 400year subset of this long-term reconstruction
(A.D. 901-13(0), the focus of this study.
-74-
Table 3.4. Descriptive Statistics Associated with 55 Final, Full-Period Calibrations and Long-Term
Reconstructions.
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.63
2.84
4.52
8.66
.95
3.91
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviatio n
Minimum Value
Maximum Value
Skewness
Kurtosis
.77
2.61
4.49
8.00
1.11
4.60
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.63
2.80
4.53
8.38
.91
3.74
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.65
2.83
4.52
8.65
.99
3.96
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
-.65
2.85
-4.48
8.88
1.06
4.18
Reconstruction 1: B 0.53 4.61
Full-Period
Calibration,
Reconstruction
A.D. 9011970
Recon PDSI
.63
2.22
5.29
4.17
.02
2.44
.55
2.45
7.87
6.65
.11
.13
Reconstruction 2: B 0.85 2.95
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.77
1.88
4.87
3.63
.15
3.05
.78
1.98
6.10
5.05
.07
.28
Reconstruction 3: B 1.02 3.65
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.63
2.18
5.24
4.11
.02
2.45
.55
2.40
7.74
6.46
.11
.13
Reconstruction 4: B 1.02 4.48
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.65
2.21
5.29
4.09
.02
2.41
.56
2.45
7.89
6.81
.11
.12
Reconstruction 5: B 1.02 5.40
FUll-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 901-1970
-.65
2.13
-4.29
3.66
.07
2.45
-75 -
-.48
2.45
-7.55
7.15
-.09
-.13
Reconstruction
A.D. 9011300
.57
2.42
7.20
6.65
.14
.01
Reconstruction
A.D. 9011300
.82
1.96
6.10
4.94
.16
.30
Reconstruction
A.D. 9011300
.57
2.38
7.06
6.46
.15
.00
Reconstruction
A.D. 9011300
.58
2.43
7.24
6.81
.13
.03 -
Reconstruction
A.D. 901-1300
-.48
2.49
-7.31
7.15
-.12
-.03
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.65
2.82
4.43
8.77
1.06
4.16
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.60
2.79
4.38
8.42
.97
3.89
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.60
2.78
4.35
8.25
.95
3.80
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.60
2.77
4.32
8.08
.94
3.73
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.59
2.77
4.28
7.96
.94
3.68
Reconstruction 6: B 1.14 6.44
Full-Period
Calibration,
Reconstruction
Recon POSt
A.D. 9011970
.65
2.10
4.21
3.60
.07
2.45
.48
2.42
7.46
7.13
.09
.12
Reconstruction 7: B 1.02 7.80
Full-Period
Calibration,
Reconstruction
A.D. 9011970
Recon PDSI
.60
2.05
4.06
3.54
.08
2.44
.43
2.37
7.25
7.06
.09
.11
Reconstruction 8: B 1.00 8.54
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.60
2.03
4.00
3.50
.08
2.44
.44
2.35
7.19
7.04
.08
.11
Reconstruction 9: B 1.05 9.45
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.60
2.00
3.93
3.45
.08
2.44
.44
2.33
7.11
7.01
.08
.10
Reconstruction 10: B 0.99 10.26
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.59
1.98
3.88
3.43
.08
2.43
-76-
.43
2.31
7.05
7.00
.08
.10
Reconstruction
A.D. 9011300
.48
2.46
7.22
7.13
.11
.02
Reconstruction
A.D. 9011300
.44
2.41
7.01
7.06
.10
.01
Reconstruction
A.D. 9011300
.44
2.39
6.95
7.04
.10
.00
Reconstruction
A.D. 9011300
.44
2.37
6.86
7.01
.09
.01
Reconstruction
A.D. 9011300
.43
2.35
6.80
7.00
.09
.01
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.59
2.76
4.17
7.58
.98
3.75
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.20
2.78
5.44
5.73
.43
2.53
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maxiroom Value
Skewness
Kurtosis
.39
2.54
5.35
5.84
.63
3.65
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maxiroom Value
Skewness
Kurtosis
.19
2.75
5.50
5.65
.40
2.56
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation·
Minimum Value
Maxiroom Value
Skewness
Kurtosis
.25
2.80
5.50
5.59
.46
2.47
Reconstruction 11: B 1.14 13.66
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.59
1.91
3.75
3.27
.09
2.42
.44
2.24
6.81
6.85
.07
.09
Reconstruction 12: C 0.53 4.61
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.20
2.04
4.49
4.49
.09
2.69
.16
2.07
5.90
5.83
.11
.31
Reconstruction 13: C 0.85 2.95
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.39
1.79
4.08
3.69
.11
2.76
.38
1.81
5.31
4.92
.10
.30
Reconstruction 14: C 1.02 3.65
.Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.19
2.03
4.45
4.46
.09
2.70
.16
2.06
5.85
5.80
.11
.31
Reconstruction 15: C 1.02 4.48
Full-Period
Calibration.
Reconstruction
Recon PDSI
A.D. 9011970
.25
2.03
4.55
4.40
.07
2.64
-77-
.20
2.05
5.94
5.65
.13
.32
Reconstruction
A.D. 9011300
.44
2.28
6.56
6.85
.08
.03
Reconstruction
A.D. 9011300
.20
2.06
5.59
5.26
.21
.35
Reconstruction
A.D. 9011300
.41
1.80
5.20
4.60
.19
.33
Reconstruction
A.D. 9011300
.20
2.05
5.56
5.25
.21
.35
Reconstruction
A.D. 9011300
.24
2.04
5.48
4.93
.23
.36
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.25
2.82
5.46
5.55
.49
2.43
Full-Period
Calibration,
Actual POSt
Mean
Standard Deviation.
Minimum Value
Maximum Value
Skewness
Kurtosis
.15
2.97
5.40
5.49
.50
2.22
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.16
2.99
5.39
5.77
.56
2.28
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
. Skewness
Kurtosis
.16
2.99
5.37
5.85
.57
2.29
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.19
2.98
5.35
5.91
.60
2.31
Reconstruction 16: C 1.02 5.40
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.25
2.02
4.53
4.38
.07
2.63
.19
2.04
5.92
5.62
.13
.32
Reconstruction 17: C 1.14 6.44
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.15
2.13
4.11
4.52
.05
2.45
.03
2.31
6.72
7.28
.11
.07
Reconstruction 18: C 1.02 7.80
Full-Period
Calibration,
Reconstruction
A.D. 9011970
Recon PDSI
.16
2.04
3.78
3.84
.08
2.36
.02
2.21
6.47
6.27
.13
.19
Reconstruction 19: C 1.00 8.54
Full-Period
Calibration,
Reconstruction
A.D. 9011970
Recon PDSI
.16
2.02
3.76
3.81
.07
2.37
.02
2.19
6.41
6.19
.14
.19
Reconstruction 20: C 1.05 9.45
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 901970
.19
2.01
3.78
3.75
.07
2.37
-78-
.01
2.17
6.39
6.05
.14
.19
Reconstruction
A.D. 9011300
.23
2.04
5.43
4.86
.24
.37
Reconstruction
A.D. 9011300
.03
2.37
6.57
7.28
.08
.02
Reconstruction
A.D. 9011300
.01
2.24
6.31
6.27
.19
.14
Reconstruction
A.D. 9011300
.01
2.22
6.25
6.19
.19
.14
Reconstruction
A.D. 9011300
.02
2.19
6.23
6.05
.19
.14
Table 3.4 (Continued).
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.20
2.97
-5.33
5.92
.61
2.31
Reconstruction 21: C 0.99 10.26
Full-Period
Calibration,
Reconstruction
A.D. 901-1970
Recon PDSI
-.05
-.20
1.88
1.88
-4.07
-5.29
4.64
3.71
-.18
-.03
-.37
2.3.6
Reconstruction
A.D. 901-1300
-.07
1.93
-5.29
4.64
-.25
-.40
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.19
2.93
-5.22
5.94
.64
2.32
Reconstruction 22: C1.14 13.66
Full-Period
Reconstruction
Calibration,
A.D. 901-1970
Recon PDSI
-.03
-.19
1.83
1.83
-5.14
-3.95
4.54
3.63
-.18
-.03
-.37
2.36
Reconstruction
A.D. 901-1300
-.06
1.88
-5.14
4.54
-.25
-.40
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.14
2.66
-6.57
6.23
.02
2.75
Reconstruction 23: 10.53 4.61
Full-Period
Calibration,
Reconstruction
A.D. 901-1970
Recon PDSI
-.14
-.54
2.11
2.08
-6.30
-4.66
3.75
6.03
.03
-.08
2.72
-.16
Reconstruction
A.D. 901-1300
-.58
2.09
-6.30
5.09
.03
-.19
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.19
2.55
-6.33
6.14
.13
2.80
Reconstruction 24: 10.852.95
Full-Period
Calibration,
Reconstruction
A.D. 901-1970
Recon PDSI
-.19
-.57
1.98
2.01
-4.51
-6.09
3.57
5.69
-.09
.04
-.15
2.73
Reconstruction
A.D. 901-1300
-.60
2.00
-6.09
4.73
.04
-.18
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.13
2.63
-6.59
6.14
-.02
2.75
Reconstruction 25: 11.02 3.65
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 901-1970
-.13
-.52
2.03
2.06
-4.52
-6.16
3.69
5.95
-.08
.04
2.73
-.15
Reconstruction
A.D. 901-1300
.55
2.05
-6.16
4.97
.04
-.18
-79-
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.12
2.67
6.71
6.12
.05
2.72
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.15
2.73
6.78
6.14
.00
2.62
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.22
2.77
6.80
6.14
.10
2.59
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.20
2.82
6.69
6.11
.14
2.48
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.19
2.82
6.59
6.07
.15
2.45
Reconstruction 26: I 1.02 4.48
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.12
2.05
4.57
3.72
.08
2.72
.51
2.07
6.18
5.94
.03
.16
Reconstruction 27: 11.02 5.40
Full-Period
Reconstruction
Calibration,
Recon PDSI
A.D. 9011970
.54
2.09
6.23
5.90
.03
.17
.15
2.07
4.68
3.72
.07
2.71
Reconstruction 28: 11.146.44
Full-Period
Reconstruction
Calibration,
A.D. 9011970
Recon PDSI
.65
1.91
5.79
4.98
.06
.27
.22
1.97
4.64
3.57
.08
2.46
Reconstruction 29: 11.027.80
Full-Period
Calibration,
Reconstruction
A.D. 9011970
Recon PDSI
.20
2.09
4.81
3.68
.03
2.68
.62
2.11
6.32
5.98
.02
.18
Reconstruction 30: 11.00 8.54
Full-Period
calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.19
2.17
5.38
4.11
.15
2.59
80-
.56
2.31
6.86
6.44
.01
.03
Reconstruction
A.D. 9011300
.55
2.06
6.18
5.02
.03
.19
Reconstruction
A.D. 9011300
.58
2.07
6.23
5.03
.02
.21
Reconstruction
A.D. 9011300
.68
1.89
5.79
4.98
.14
.27
Reconstruction
A.D. 9011300
.66
2.09
6.32
5.23
.01
.22
Reconstruction
A.D. 9011300
.58
2.29
6.21
6.44
.14
.14
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSJ
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.11
2.68
6.48
6.04
.17
2.34
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.07
2.96
6.41
6.03
.22
2.25
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.11
2.90
6.20
5.78
.28
2.27
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.28
2.29
5.11
5.34
.38
2.98
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.25
2.22
4.54
5.41
.57
3.03
Reconstruction 31: I 1.05 9.45
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.11
2.13
5.26
4.04
.07
2.43
.47
2.28
7.07
6.51
.07
.10
Reconstruction 32: 10.99 10.26
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.07
2.21
5.18
4.26
.16
2.52
.38
2.42
7.53
7.81
.02
.08
Reconstruction 33: 11.14 13.66
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.11
2.11
4.95
4.01
.15
2.52
.40
2.31
7.22
7.41
.02
.07
Reconstruction 34: M 0.53 4.61
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.28
1.71
4.19
3.44
.11
2.86
.45
1.77
5.22
4.85
.05
.26
Reconstruction 35: M 0.85 2.95
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.25
1.56
3.82
3.14
.11
2.86
81-
.40
1.62
4:76
4.43
.05
.26
Reconstruction
A.D. 9011300
.48
2.25
6.53
6.51
.05
.09
Reconstruction
A.D. 9011300
.39
2.43
6.35
7.81
.16
.31
Reconstruction
A.D. 9011300
.41
2.32
6.12
7.41
.15
.30
Reconstruction
A.D. 9011300
.48
1.75
5.22
4.85
.13
.25
Reconstruction
A.D. 9011300
.44
1.60
4.76
4.43
.13
.24
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.27
2.27
4.87
5.31
.41
2.94
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.26
2.30
5.22
5.31
.30
2.95
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation.
Minimum Value
Maximum Value
Skewness
Kurtosis
.27
2.30
5.43
5.34
.26
3.10
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.28
2.34
5.52
5.36
.24
2.92
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maxirrum Value
Skewness
Kurtosis
.30
2.37
5.59
5.39
.26
2.81
Reconstruction 36: M 1.02 3.65
Full-Period
Calibration,
Reconstruction
Recon POSt
A.D. 9011970
.27
1.68
4.13
3.38
.11
2.86
.44
1.75
5.14
4.79
.05
.25
Reconstruction 37: M 1.02 4.48
Full-Period
Reconstruction
Calibration.
A.D. 9011970
Recon PDSI
.43
1.80
5.26
4.95
.05
.26
.26
1.73
4.22
3.51
.11
2.86
Reconstruction 38: M 1.02 5.40
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.27
1.73
4.24
3.49
.11
2.86
.44
1.80
5.28
4.96
.05
.25
Reconstruction 39: M 1.14 6.44
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.28
1.83
4.46
3.68
.11
2.86
.46
1.89
5.56
5.21
.05
.26
Reconstruction 40: M 1.02 7.80
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.30
1.87
4.51
3.82
.10
2.82
82-
.47
1.04
5.68
5.23
.06
.27
Reconstruction
A.D. 9011300
.47
1.73
5.14
4.79
.12
.24
Reconstruction
A.D. 9011300
.46
1.78
5.26
4.95
.13
.24
Reconstruction
A.D. 9011300
.47
1.78
5.28
4.96
.12
.24
Reconstruction
A.D. 9011300
.50
1.87
5.56
5.21
.12
.24
Reconstruction
A.D. 9011300
0.51
1.92
5.68
5.21
.14
.28
Table 3.4 (Continued).
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.38
2.39
-5.61
5.37
.33
2.77
Reconstruction 41: M 1.00 8.54
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D.901-1970
-.38
-.54
1.86
1.93
-5.73
-4.56
5.13
3.76
-.06
.10
-.27
2.82
Reconstruction
A.D.901-1300
-.58
1.91
-5.73
5.12
-.14
-.28
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.41
2.39
-5.60
5.36
.38
2.78
Reconstruction 42: M 1.05 9.45
Full-Period
Reconstruction
Calibration,
Recon POSt
A.D. 901-1970
-.41
-.58
1.84
1.91
-4.59
-5.71
3.60
5.07
-.05
.10
-.26
2.84
Reconstruction
A.D. 901-1300
-.62
1.88
-5.71
5.07
-.13
-.26
Mean
Standard Deviation
Minimum Vatue
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actuat POSt
-.43
2.40
-5.59
5.37
.41
2.78
Reconstruction 43: M 0.99 10.26
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 901-1970
-.43
-.60
1.84
1.90
-4.60
-5.73
3.58
5.05
.10
-.05
2.85
-.26
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual POSt
-.36
2.52
-5.54
5.37
.37
2.48
Reconstruction 44: M 1.14 13.66
Full-Period
Calibration,
Reconstruction
Recon POSt
A.D. 901-1970
-.36
-.52
1.85
1.92
-4.54
-5.68
3.69
5.14
.10
-.06
2.84
-.27
Reconstruction
A.D. 901-1300
-.56
1.89
-5.68
5.14
-.14
-.26
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
Full-Period
Calibration,
Actual PDSI
-.11
2.07
-4.21
6.43
.57
3.82
Reconstruction 45: F 0.53 4.61
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 901-1970
-.11
-.14
1.26
1.30
-2.91
-3.79
2.78
3.59
.07
-.13
2.67
-.31
Reconstruction
A.D. 901-1300
-.17
1.30
-3.52
3.15
-.23
-.36
83 -
Reconstruction
A.D. 901-1300
.64
1.88
-5.73
5.05
-.13
-.26
Table 3.4 (Continued).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.10
2.00
3.57
6.16
.59
3.54
Full-Period
Calibration,
Actual POSt
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.11
2.03
3.95
6.34
.61
3.79
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.10
2.06
4.27
6.37
.54
3.82
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.12
2.09
4.55
6.41
.48
3.84
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.15
2.17
4.86
6.47
.41
3.66
Reconstruction 46: F 0.85 2.95
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.10
1.21
2.79
2.68
.08
2.69
.14
1.26
3.63
3.48
.12
.31
Reconstruction 47: F 1.02 3.65
Full-Period
Reconstruction
Calibration,
A.D. 9011970
Recon PDSI
.11
1.23
2.85
2.72
.07
2.67
.14
1.28
3.72
3.52
.13
.31
Reconstruction 48: F 1.02 4.48
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.10
1.25
2.88
2.78
.07
2.67
.13
1.29
3.76
3.59
.13
.31
Reconstruction 49: F 1.02 5.40
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.12
1.27
2.95
2.81
.07
2.67
.15
1.32
3.85
3.63
.13
.31
Reconstruction 50: F 1.14 6.44
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.15
1.38
3.20
3.03
.08
2.69
84-
.19
143
4.18
3.95
.12
.31
Reconstruction
A.D. 9011300
.17
1.25
3.42
3.11
.22
.35
Reconstruction
A.D. 9011300
.17
1.27
.3.45
3.08
.23
.36
Reconstruction
A.D. 9011300
.15
1.29
3.49
3.15
.23
.36
Reconstruction
A.D. 9011300
.17
1.32
3.57
3.18
.23
.36
Reconstruction
A.D. 9011300
.22
1.43
3.95
3.54
.22
.35
Table 3.4 (Concluded).
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.15
2.22
5.16
6.48
.34
3.63
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.16
2.23
5.31
6.44
.30
3.63
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.17
2.24
5.47
6.40
.27
3.63
Full-Period
Calibration,
Actual PDSI
Mean
Standard Deviation
Minimum Value
Maximum Value
Skewness
Kurtosis
.16
2.25
5.59
6.37
.21
3.61
Full-Period
Calibration,
Actual POSt
Mean
Standard Deviation
Minimum Vatue
Maximum Value
Skewness
Kurtosis
.21
2.24
5.99
6.22
.16
3.86
Reconstruction 51: F 1.02 7.80
Full-Period
Reconstruction
Calibration,
A.D. 9011970
Recon PDSI
.20
1.46
4.26
4.02
.12
.31
.15
1.41
3.26
3.08
.08
2.69
Reconstruction 52: F 1.008.54
Full-Period
Reconstruction
Calibration,
A.D. 9011970
Recon PDSI
.21
1.47
4.29
4.03
.12
.31
.16
1.42
3.30
3.09
.08
2.69
Reconstruction 53: F 1.05 9.45
Full-Period
Reconstruction
Calibration,
A.D. 9011970
Recon POSt
.17
1.40
3.28
3.05
.08
2.68
.22
1.46
4.27
3.98
.12
.31
Reconstruction 54: F 0.99 10.26
Full-Period
Calibration,
Reconstruction
Recon PDSI
A.D. 9011970
.16
1.40
3.25
3.05
.07
2.68
.20
1.45
4.24
3.97
.12
.31
Reconstruction 55: F 1.14 13.66
Full-Period
Calibration,
Reconstruction
Recon POSt
A.D. 9011970
.21
1.29
3.07
2.76
.07
2.68
85-
.25
1.34
3.98
3.60
.12
.31
Reconstruction
A.D. 9011300
.23
1.45
4.02
3.60
0.22
.35
Reconstruction
A.D. 9011300
.24
1.46
4.05
3.61
.22
.35
Reconstruction
A.D. 9011300
.25
1.45
4.02
3.53
.22
.36
Reconstruction
A.D. 9011300
.23
1.44
3.97
3.52
.22
.36
Reconstruction
A.D. 9011300
.27
1.33
3.73
3.18
.22
.36
An examination of these data reveal consis-
structed calibration data set has both minimum and maximum values that are less
extreme than their counterparts in the actual calibration period data set This is
also generally expected with regressionbased estimation given its tendency to
create relatively normal distributions with
predicted values closer to the mean.
tent patterning in these statistics from set to set
in each of the 55 reconstructions• The means of the reconstructed calibration
period data sets are always equivalent to
the means of the actual calibration period
data sets. This is expected since mean
values of the x and y variables will fall
directly on the regression line and thus
will be predicted by the regression equation, as well.
• In all cases, the reconstructed calibration
period data set produces distributions that
are less skewed (i.e.• closer to 0.00) than
those セッ the actual calibration period data
Lエセ
whIch tend to be characterized by positIVely skewed distributions.
• The calibration period mean is never
zero, although it often approaches zero. In
part this is due to the limited number of
years used to calibrate any reconstruction,
which results in a skewed sample that
does not fully capture the long-term climatic variation characteristic of a location. However, the greater part of this
phenomenon results from using a slightly
extended set of years to calculate the original PDSI values (i.e., all instrumented
climatic data were taken through 1983),
whereas the calibration period must end at
1970 because that is the final year of treering data. Because no reconstruction has
joint precipitation and temperature data
earlier than 1915, it is important to include as many years as possible in the
original PDSI calculations in order to derive more accurate long-term means used
in the PDSI equations. If, however, the
post-calibration years used to determine
the actual PDSI values were characterized
by precipitation and temperature conditions somewhat different than that of the
calibration period, the distribution of
those values and their mean will be different Although it varies for each of the
55 reconstructions for the 1971-1983
period, the post-1970 period was generally wetter with more years characterized
by extremes in wetness than extremes in
dryness (fable 3.5). Consequently, the
calibration period means differ from zero
or near-zero and are consistently on the
negative side.
• In all cases, the reconstructed calibration
period data set produces distributions that
better approximate normal kurtosis or
"peakedness" (normal=3.00) than the actual calibration period data sets.
• The means of both the full, long-term reconstruction (A.D. 901-1970) and the
first 400-year subset (A.D. 901-1300) are
somewhat different from the calibration
period means, reflecting the variation associated with a large number of years.
• The reconstruction period standard devia-
tions are similar to the calibration period
standard deviations and fall between those
generated for the actual full period calibration and the reconstructed full period calibration.
• The minimum and maximum values differ
from those of the calibration period,
reflecting more extreme conditions over
the 107Q-year reconstruction period than
in the reconstructed calibration period.
• Both reconstructed series tend to be normally distributed as judged by skewness,
and exhibit platykurtic distributions.
• The distribution of values within the 400
years used in this study exhibits characteristics quite similar to that of the full
long-term reconstruction. Small patterned
differences OCCW" from elevationaJ. zone to
elevational zone.
• The standard deviation of the reconstructed
calibration period is always less than the
actual calibration, and again this is an expected result of using a least-squares セ
gression line to predict or reconstruct
PDSI values.
The long-term June PDSI reconstructions
presented in this study compare favorable with
Rose's Zuni area (Martin Rose, personal communication, April 1989) and Southwestern
• In most cases (50/55 or 91 %) the recon-
-
86-
Table 3.5. Occurrence of Extreme Conditions estimated for the month of June (s -4.00 or 2+4.00 PDSI) in the Years Used to Calculate PDSI for all Combination of Weather Station and
Moisture Class.
Weather
Station
Bluff
AWca
Class
1
Extreme
Dry Years
(ore-1941 )
1934
(-4.52)
1934
(-4.49)
1934
(-4.53)
Extreme Dry Years
(1941-1970) and
r1971-19831 b
1951 (-4.40); 1956
(-4.06): 1959 (-4.02)
1951 (-4.28);
1959 (-4.08)
1947 (-4.05); 1951
(-4.32); 1959 (-4.03)
1934
(-4.52)
1934
(-4.48)
1934
(-4.43)
1934
(-4.38)
1947 (-4.07);
1951 (-4.31)
1947 (-4.07);
1951 (-4.30)
1947 (-4.05); 1951
(-4.25); 1956 (-400)
1947 (-4.05); 1951
(-4.24); 1956 (-4.01)
1934
(-4.35)
1934
(-4.32)
1934
(-4.28)
1934
(-4.17)
1947 (-4.05); 1951
(-4.23)' 1956 (-403)
1947 (-4.06); 1951
(-4.21)' 1956 (-4.05)
1947 (-4.04); 1951
(-4.21)' 1956 (-4.07)
1947 (-4.06); 1951
(-4.16)' 1956 (-4.15)
1951 (-4.93); 1959
(-4.13); [1977]
Bluff
2
Bluff
3
Bluff
4
Bluff
5
Bluff
6
Bluff
7
Bluff
8
Bluff
9
Bluff
10
Bluff
11
Cortez
1
Cortez
2
Cortez
3
Cortez
4
1951 (-5.50); 1959
(-4.84); [1977]
Cortez
5
1951 (-5.46); 1959
(-4.69); [1977]
Cortez
6
1951 (-5.40); 1959
(-4.47); [1972,1977]
Cortez
7
1951 (-5.39); 1959
(-4.27); [1972,1977]
1951 (-5.35); 1959
(-4.95): [19771
1951 (-5.50); 1959
(-4,93); [1977]
-
87-
Extreme
Extreme Wet Years (1941Wet Years 1970) and [1971-1983Jb
(pre-1941)
1941 (8.66); 1957 (4.25);
r19741979,1980,1983]
1941 (8.00); 1957 (5.00);
£1978 1979 1980 1983]
1941 (8.38); 1957 (4.36);
[1973,1978,1979,1980,
19831
1932
1941 (8.65);
(4.17)
r1973 1979 1980,19831
1932
1941 (8.88);
(4.53)
f1973 1979 1980 19831
1932
1941 (8.77);
(4.28)
£1973,1979 1980 19831
1928
1941 (8.42);
(4.68);
[1973,1979,1980,1983J
1932
(4.08)
1941 (8.25);
1928
(4.99)
£1973 1979 1980 19831
1941 (8.08);
1928
(5.36)
£197319781980,19831
1928
1941 (7.96);
(5.63)
f1973 1979 1980 1983];
1928
1941 (7.58);
(6.56)
f1973 1979 1980 19831
1932
1941 (5.16); 1942 (4.71);
(4.21 )
1957 (4.80);
£19731978 19791
1932
1941 (5.38); 1957 (5.84);
(4.01)
1965 (4.93); f1973 19791
1941 (5.19); 1942 (4.54);
1932
(4.15)
1957 (5.65); 1965 (4.73);
£1973.19791
1932
1941 (5.15); 1942 (4.87);
(4.28)
1957 (5.59); 1965 (4.61);
£1973 19791
1932
1941 (5.10); 1942 (5.18);
(4.49)
1957 (5.55); 1965 (4.51);
(1973 19791
1932
1941 (5.05); 1942 (5.48);
(4.75)
1957 (5.49); 1958 (5.04);
1965 (4.42);
£1973 1979 19801
1932
1941 (5.12); 1942 (5.n);
(4.92) .
1957 (5.48); 1958 (5.38);
1965 (4.18);
£1973 1979 19801
Table 3.5 (Continued).
Weather
Station
Cortez
AWCa
Class
8
Extreme Dry Years
Extreme
Dry Years (1941-1970) and
(pre-1941 ) r"1971-19831 b
1951 (-5.37); 1959
(-4.12); [1972,1977]
Extreme
Wet Years
(pre-1941)
1932
(5.03)
Cortez
9
1951 (-5.35);
[1972,1977]
1932
(5.19)
Cortez
10
1951 (-5.33);
[1972,1977]
1932
(5.31);
Cortez
11
1951 (-5.22);
[1972,1977]
1932
(5.94)
Ignacio
1
1934
(-6.57)
1946 (-4.40); 1950
(-4.54); 1959 (-4.53);
[1977]
Ignacio
2
1934
(-6.33)
1946 (-4.19); 1950
(-4.47); 1959 (-4.28);
[1977]
Ignacio
3
1934
(-6.59)
1946 (-4.37); 1950
(-4.56); 1959 (-4.41);
[1977]
Ignacio
4
1934
(-6.71 )
1946 (-4.44); 1950
(-4.54); 1959 (-4.47);
[1977)
Ignacio
5
1934
(-6.78)
1946 (-4.44); 1950
(-4.67); 1951 (-4.19);
1959 (-4.44); [1977]
Ignacio
6
1934
(-6.80)
1946 (-4.36); 1950
(-4.44); 1951 (-4.36);
1959 (-4.28); [1977]
Ignacio
7
1934
(-6.69)
1946 (-4.28); 1950
(-4.40); 1951 (-4.49);
1959 (-4.08); [1977]
1926
(5.04);
1927
(4.49)
1926
(4.85);
1927
(4.26)'
1926
(4.93);
1927
(4.45)
1926
(4.89);
1927
(4.56)
1917
(4.05);
1926
(4.81);
1927
(4.74)
1917
(4.29);
1926
(4.76);
1927
(5.04)
1917
(4.51);
1926
(4.n);
1927
(5.28)
88 -
Extreme Wet Years (19411970) and [1971-1983]b
1941 (5.06); 1942 (5.85);
1957 (5.41); 1958 (5.40);
1965 (4.09);
r1973 1979 19801
1941 (4.94); 1942 (5.91);
1957 (5.36); 1958 (5.26);
r1973 1989 1980)
1941 (4.87); 1942 (5.92);
1957 (5.30); 1958 (5.13);
r1973 1979 19801
1941 (4.69); 1942 (5.57);
1957 (5.27); 1958 (4.62);
r1973 1979 1980)
1941 (6.23); [1973,1979]
1941 (6.14); [1973]
1941 (6.14); [1973]
1941 (6.12); [1973,1979]
1941 (6.14); [1973,1979]
1941 (6.14); [1973,1979]
1941 (6.11); [1973,1979]
Table 3.5 (Continued).
Extreme
Dry Years
(Dre-1941 )
1934
(-6.59)
Extreme Dry Years
(1941-1970) and
r1971-19831 b
1946 (-4.25); 1950
(-4.37); 1951(-4.52);
[1977J
9
1934
(-6.48)
1946 (-4.22); 1950
(-4.34); 1951 (-4.54);
[1977]
Ignacio
10
1934
(-6.41 )
1946 (-4.18); 1950
(-4.31); 1951 (-4.56);
[1977]
Ignacio
11
1934
(-6.20)
1946 (-4.07); 1959
(-4.29); 1951 (-4.56)
Mesa
Verde
1
1951 (-5.11); [1977]
Mesa
Verde
Mesa
Verde
2
1951 (-4.54); [1977]
3
1951 (-4.87); [1977]
Weather
Station
Ignacio
AWCa
Class
8
Ignacio
Mesa
Verde
Mesa
Verde
Mesa
Verde
Mesa
Verde
Mesa
Verde
Mesa
Verde
Mesa
Verde
Mesa
Verde
Ft. Lewis
4
5
6
7
8
9
10
11
1
1934
(-4.04)
1934
(-4.14)
1934
(-4.18)
1934
(-4.15)
1934
(-4.14)
1934
(-4.14)
1934
(-4.13)
1934
(-4.20)
1951 (-5.22);
[1977198f1
1951 (-4.85);
[197719811
1951 (-5.52);
[197719811
1951 (-5.59);
(19771981)
1951 (-5.61);
f197719811
1951 (-5.60);
[1977 1981'1
1951 (-5.59); [1977]
1951 (-5.54); [1977]
1964 (-4.21): r19771
-89-
Extreme
Wet Years
(Dre-1941)
1917
(4.62) ;
1926
(4.77);
1927
(5.41)
1917
(4.71);
1926
(4.78);
1927
(5.54)
1917
(4.78);
1926
(4.81);
1927
(5.64)
1917
(5.10);
1926
(4.61 );
1927
(5.71)
1926
(4.03);
1938
(4.39)
1938
(4.32)
1926
(4.00);
1938
(4.43)
1938
(4.46)
1938
(4.45)
1938
(4.47)
1938
(4.49)
1938
(4.51 )
1938
(4.57)
1938 .
(4.61 )
1938
(5.09)
Extreme Wet Years (19411970) and [1971-1983Jb
1941 (6.07);
[1973,1979J
1941 (6.04); 1942 (4.56);
[1973.1979J
1941 (6.03); 1942 (4.65);
1958 (4.84); [1973.1979]
1941 (5.78); 1942 (4.72);
1958 (4.59);
[1973,1979,1980J
1941 (5.34)
1941 (5.41); 1957 (4.01);
[19791
1941 (5.31)
1941 (5.31); [1979]
1941 (5.34); [1973,1979]
1941 (5.36);
[1973.1979 19831
1941 (5.39);
[1973 1979 19831
1941 (5.37);
[1973 1979 19831
1941 (5.36);
f1973 1979 19831
1941 (5.37);
r19731979.19831
1941 (5.37);
r1973 1979 19801
1941 (6.43)
Table 3.5 (Concluded).
Weather
Station
Ft. Lewis
Ft. Lewis
Ft. Lewis
Ft. Lewis
AWCa
Class
2
3
4
5
Ft. Lewis
6
Ft. Lewis
7
Ft. Lewis
8
Ft. Lewis
9
Ft. Lewis
10
Ft. Lewis
11
Extreme
Dry Years
(pre-1941)
..
Extreme Dry Years
(1941-1970) and
1971-19831b
19771
19771
1964 (-4.27)· r19771
1959 (-4.06); 1964
(-4.55); r19n1
1959 (-41.2); 1964
(-4.86)· r19n1
1959 (-4.16); 1963
(-4.06); 1964 (-5.16);
r19771
1959 (-4.17); 1963
(-4.17); 1964 (-5.31);
(1977)
1959 (-4.13); 1963
(-4.29); 1964 (-5.47);
r19771
1959 (-410); 1963
(-438); 1964 (-559);
f19771
1963 (-472); 1964
(-599); f19771
Extreme
Extreme Wet Years (1941Wet Years 1970) and [1971-1983]b
(ore-1941)
1941 (6.16
1941 {6.34
1941 (6.37
1941 (6.41)
1941 (6.47); 1949 (4.07);
r19731
1941 (6.48); [1973]
1941 (6.44); 1949 (4.25);
[1973]
1941 (6.40); 1949 (4.32);
[1973]
1941 (637); 1949 (436);
[1973]
1941 (622); 1949 (453);
f19731
Inches
of SOIl and In the lower
aSee Table 3.3 for descnptlOn of the amount of water held In upper SIX
sod profile for each AWC class.
bYears between 1971 and 1983 that were used in calculating CAFEC values necessary for deriving
POSt, but not used in calibration procedures. Years 1941·1970 used in calibration. Pre-1941 years
used in verification.
Colorado July PDSI reconstructions (Rose et
al. 1982:385). Table 3.6 summarizes data from
these two studies for comparison with data in
Table 3.4.
Historic Commentary
The agreement between actual and reconstructed PDSI values may be assessed in a
qualitative manner, as well. While historic
commentary may be regarded as anecdotal, it
does provide a source of information that can
be used to evaluate the general robustness of
reconstructed PDSI values. This should be
particularly noticeable in years when more
extreme climatic conditions that affected local
populations and their perceptions of "abnormality" were recorded.
Historical remarks on weather conditions in
Montezuma and Dolores counties have been
assembled from-three sources: an oral history
of the Goodman Point area recently assembled
by MaJjorie Connolly of the Crow Canyon
Archaeological Center, Cortez (Connolly,
1990); printed commentary in Cortez's newspaper, the Montezuma Valley Journal (Ian
Thompson, personal communication, January
1990); and newspaper commentary gleaned
from the files of the Mancos Times, Mancos
Times-Tribune, and the Montezuma Valley
Journal as reported in a published history of
Montezuma County (Freeman 1958:207-317).
The earliest remarks begin in 1894 and are
contained in appendix B.
Table 3.7 lists the annual mean reconstructed PDSI values of the 34 reconstructions
associated with soils known to be used for agriculture in the Montezuma and Dolores County
area for the period of 1890-1970 (see section
on Historic Crop Yield and Total Regression,
this chapter, for more detail on the identification of these 34 reconstructions). It also notes
-90-
concurrence with available historic commentary, for those years where comments were
found. Figure 3.2 plots these annual mean
PDSI values. Figure 3.3 plots the standard
deviations for these mean values for the same
1890-1970 time period. Whereas the plot of
annual mean values tracks the prevailing condition noted across the region during a given
year, the plot of standard deviations provides
infonnation on the spatial variability present in
any year. The lower the standard deviation. the
more consistent and widespread the condition.
The higher the standard deviation, the more
localized and discontinuous the condition.
Table 3.6. Comparative Data from Long-Term
Reconstructions in the Zuni Area and in
Southwestern Colorado.
SW
coa
Mean
Standard
deviation
Minimum value
Maximum value
Skewness
Kurtosis
-1.07
1.85
-6.03
4.54
-0.04
-0.25
SW
Mountains b
-0.94
1.78
-6.29
4.16
-0.17
-0.30
Black
Rockc
-0.42
2.08
-6.58
5.31
-0.05
-0.21
aSouthwest Colorado reconstruction combines instrumented climatic data from Cortez, Mancos, Mesa
Verde, Ft. Lewis, Ignacio, and Pagosa Springs,
Colorado (Rose et al. 1982:237,385).
bSouthwest Mountains reconstruction combines
instrumented climatic data from Ft. Wingate, EI Morro,
Bluewater, Grants, San Fidel, and Marquez, New
Mexico (Martin Rose, personal communication, April
1989).
cBlack Rock reconstruction uses climatic data from
Black Rock, New Mexico (Martin Rose, personal
communication, April 1989).
Historic remarks (Appendix B) are associated with 41/81 (51%) years between 18901970. Of these, 37/41 (90%) unquestionably
support the reconstructed PDSI annual mean
values. 3/41 (7%) do not contradict the PDSI
values but do not strongly support it, and only
1/41(2%) seem not to support the PDSI value
at all. However, it is noted that in that year,
1947, there was much variability about the
11
Wet
10
9
セ
-=
セ
」N^セ
vjセ
cャセ
セ
8
セ
1
J
セ
7
6
j
5
セ
4
1
3
2
Dry
1
1890
1910
1930
1950
1970
Years
(A.D.)
Figure 3.2. Plot of annual mean PDSI values by class for 34 longterm reconstructions (REC34), A.D.
18901970. Class 1 represents conditions of extreme drought, Class 6 represents conditions that are near normal. and Class 11 represents conditions of extreme wetness' re
mainingclasses are intermediate.
•
-91-
Table 3.7. Comparison of Historic Comments with Annual PDSI Values for 34 Long-Tenn Reconstructions, A.D. 1890-1970.
Year
A.D.
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
Mean
REC34
125.32
136.85
111.15
94.15
94.53
125.15
78.91
134.53
119.82
73.24
100.24
124.65
73.41
137.38
72.06
145.21
141.71
146.08
133.41
139.71
126.21
146.88
133.32
98.38
150.24
152.32
149.41
139.65
97.32
138.15
133.06
123.06
129.47
103.15
135.21
105.59
147.21
127.91
119.24
113.71
116.97
102.41
137.44
103.77
93.12
132.24
110.06
134.15
127.15
111.50
114.24
156.09
128.4,1
S.D.
REC34
9.93
3.00
8.44
4.61
4.69
5.10
6.52
9.79
1.83
5.58
4.06
4.69
8.74
6.84
6.46
3.58
9.43
3.90
13.99
14.14
6.94
7.97
7.16
4.26
4.16
4.92
4.87
9.69
9.26
7.87
6.32
2.64
5.18
4.21
9.27
5.62
5.49
3.58
6.17
1.43
4.48
8.61
10.22
5.67
7.01
3.64
3.58
5.03
3.97
2.72
4.47
3.97
11.13
PDSI
Class
7
8
5
3
3
7
1
8
6
1
3
6
1
8
1
9
9
9
8
8
7
9
8
3
10
10
9
8
3
8
10
6
7
4
8
4
9
7
6
5
6
4
8
4
3
8
4
8
7
5
5
10
7
PDSI
Class Condition
incipient wet
slightly wet
incipient drought
moderate drought
moderate drought
incipient wet
extreme drought
slightly wet
near nannal
extreme drought
moderate drought
near nannal
extreme drought
slightly wet
extreme drought
moderately wet
moderately wet
moderately wet
slightly wet
slightly wet
incipient wet
moderately wet
slightly wet
moderate drought
unusually wet
unusually wet
moderately wet
slightly wet
moderate drought
slightly wet
unusually wet
near nonnal
incipient wet
slight drought
slightly wet
slight drought
moderately wet
incipient wet
near nonnal
incipient drought
nearnonnal
slight drought
slightly wet
slight drought
moderate drought
slightly wet
slight drought
slightly wet
incipient wet
incipient drought
incipient drought
unusually wet
incipient wet
-92-
Historic
Agreement
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
no, not in Cortez
ok
yes
ok
yes
yes
ok
yes
yes
yes
Table 3.7 (Concluded).
Year
A.D.
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
Mean
REC34
104.06
118.35
120.71
88.71
130.15
130.41
142.21
97.24
79.94
145.62
102.4'7'
108.68
110.35
98.50
133.03
122.09
80.27
126.27
108.62
120.88
108.56
110.68
145.18
119.77
106.62
120.15
112.53
115.59
S.D.
REC34
6.80
3.69
4.23
5.17
6.95
1.78
4 37
6.58
5.00
9.14
7.12
2.14
5.84
5.53
6.83
7.05
8.04
5.33
5.38
5.69
3.68
7.93
4.37
7.06
5.41
7.83
6.49
5.80
PDSI
Class
4
6
6
2
8
8
9
3
1
9
4
4
4
3
8
6
1
7
4
6
4
4
9
6
4
6
5
5
PDSI
Class Condition
slight drought
near normal
near normal
severe drought
slightly wet
slightly wet
moderately wet
moderate drought
extreme drought
moderately wet
slight drought
slight drought
slight drought
moderate drought
slightly wet
near normal
extreme drought
incipient wet
slight drought
near normal
slight drought
slight drought
moderately wet
near normal
slight drought
near normal
incipient drought
incipient drought
Historic
Agreement
ok
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
Note: For the entire 81 -year period, the overall PDSI class mean and standard deviation is 6.03 ± 2.50.
Historic remarks are listed in Appendix B by year. "Yes' indicates clear agreement with available historic
commentary; "ok" indicates no contradiction but no strong relationship; "no' indicates contradictory
conditions indicated from available remarks. The absence of a comment means no historic remarks were
found for that year.
mean, and meteorological drought varied considerably from place-to-place and soil-to-soil.
Figures 3.2 and 3.3 illustrate that the turnof-the-century years 1896-1904 were notably
dry, especially 1896, 1898, 1902, and 1904.
The period between 1905-1920 was notably
wet, especially 1914, 1915, and 1920, where
only 2/16 years (1913, 1918) were ever below
the long-term average (i.e., PDSI Class 6). The
years 1921-1930 were rather average as a
whole but exhibited much year-to-year variation. The decade between 1931 and 1940 was a
notably dry period with 6/10 years drier than
the long-term average, especially 1934. The
1941-1949 period was quite variable with
than was usual, especially
more extreme カセオ・ウ
the years 1942, a year of high spatial variability; 1941, a very wet year, and 1946, a very dry
year. The years 1950-1959 were notably dry,
particularly 1951 and 1959. The final years in
the series, 1960-1970, were moderate with little
inter-annual variation.
The 1905-1920 wet period and the 19311940 dry period were widely experienced in
the western United States and show up in numerous climatic reconstructions (Fritts 1965).
It is also interesting to note that the modem
practice of dry-farming in Montezuma and
Dolores Counties began in 1905 and 1906
near Mancos (Freeman 1958:215) and 19091910 near Yellowjacket (Freeman 1958:144)
in years with greater than average soil moisture. Experimentation continued through
-93-
16
14
12
'0
"'=
"
'0
e
=
10
tI)
fIl
file
U·セッ
MNセエiIセ
qセ
=.oQ
8
6
4
セ
セ
•
n
セ
l1 11
セ
セ
:II
3I!Il
II
2
X
セ
U
I!I
IrJ
•
イセ
iセ
o
1890
1930
1910
1950
1970
Years
(A.D.)
Figure 3.3. Plot of annual standard deviations for mean PDSI values for 34 long-term reconstructions
(REC34), A.D. 18901970. The mean standard deviation is 6.03 ± 2.50; ± one standard deviation ranges between 3.538.53, ± two standard deviations ranges between 1.0311.03.
Thus, three years, 1908, 1909, and 1942 are beyond two standard deviation units from the
normal amount of spatial variation exhibited in the Montezuma County and Dolores County area.
the wetterthanaverage years prior to and
during the First World War. It was found to be
very successful and was adopted by many
farmers in the area by 19261927 (e.g., within
Dove Creek, and
the communities of a」ャセュ・ョL
Cahone [Freeman 1958:148, 152]), again in a
period when soil moisture conditions were almost never unfavorable to this type of nonirrigated agriculture.
April 1989), using very similar treering data
as predictor variables. Further, it appears that
the reconstructed series of PDSI values closely
match the historic record available for Montezuma and Dolores Counties. This adds qualitative support for the general accuracy of the
reconstructed series and their usefulness for
paleoenvironmental reconstruction. Thus, all
reconstructions were judged to be acceptable
and were used in the final GIS analyses.
Conclusion
The 55 longterm reconstructions based on
the 55 fullperiod calibration transfer functions
appear to be quite reasonable and compare
well against other longterm reconstructions
for locales on the southeastern Colorado
Plateau. Reconstructions based on June PDSI
do not seem to be markedly different than the
July PDSI reconstructions made by Rose et al.
(1982) and Rose (personal communication,
GEOGRAPillC INFORMATION SYSTEM
PROCESSING
Elevation Data
Digital Elevation Models (OEMs) are not
instantly usable when they are obtained from
the USGS. There is usually some preprocessing
to read in, transform, mosaic, correct, and rescale images before they can be used in
94-
analysis. In this case, the ASCII tape containing
the 12 DEMs was mounted on a drive connected to the IBM 3090-200 mainframe at
WSU and processed by the VICAR command
VTOPA. VTOPA logs in the digitized data and
records selected header information about
each DEM file, including map identification
number, labeling information, and geographic
coordinates of the four comers of the image.
These were used thereafter as descriptive labels
for succeeding VICAR programs. VTOPA also
examines the data records of each DEM and
checks for possible errors in the elevation files.
Finally, it creates a binary 16-bit pixel rectilinear image with one DEM elevation sample
per pixel location in VICAR format.
Next, each image was rotated 90 degrees
counter-clockwise by VICAR command R..OT
so that the top of each image is north. Following rotation, the individual rectangular images
were pieced together to create a single uninterrupted image. This process is called "mosaicing" and was accomplished by VICAR
commands FASTMOS, which matches the
edges and joins the pieces, and by SAR, which
interpolates missing values.
Subsequent VICAR commands reexpressed
the data so that they could be processed by
EPPL7 programs. Command C converted the
16-bit pixels to 8-bit significance thereby compressing the original number of elevational
classes, representing altitudes from I,SOO3,012 m, to some number between セRUsL
the
maximum class range for EPPL7 (28= 2S6).
The 1512 different elevational classes or
"DNs" (Density Number, a term borrowed
from remote sensing) were reduced to 189 by
grouping elevational classes in groups of eight
meters. Thus, DN#1 equals elevations 1500IS07 m, DN#2 equals 1508-1515 m, etc.
Commands TIECONM, MGEOM, and
STRETCH resampled the 3D-m cell resolution
of the original DEM to match the 200-m cell
resolution of the soils data. Here, the original
image with 1,425 lines (rows) and 1,508 samples (columns) and a 30-m cell size was "degraded" by a factor of about 6.6 to produce
an output image with only 214 lines (rows) and
227 samples (columns) and a 2oo-m cell size.
Other programs created histograms of the DN
values and line printer maps, permitting visual
inspection of the data and facilitating error
detection.
The final mosaiced. corrected, and rescaled
output image, now formatted as a binary raster
file, was downloaded to an mM PC on which
the EPPL7 GIS program was loaded. This
EPPL7 file (CRVDEM.EPP) was the basic elevational data plane for this study. Later, a
"trimmed" or "windowed" version of this
image was created with EPPL command WINDOW. Here the 14 bottom-most rows were deleted from the analysis leaving an image with
200 rows and 227 columns (DEMWIND.EPP).
This was done largely because so little land in
Utat part of the study area had been mapped
for soils; much of this land is Ute Mountain
Ute Indian Reservation and had not been
examined by SCS personnel.
Finally, using EPPL's RECLASS command, this second "trimmed" elevational data
plane was used to create a third elevational data
plane (SELEV.EPP) that grouped the 189 elevation classes into five elevational classes. Here
each class was the range of elevations to be
modeled by each of the weather stations. Thus,
Class 1 included those cells on the image modeled by the set of Bluff reconstructions (DNs#
1-13) representing elevations 1,500-1,604 m,
which covered 2.71 % of the study area. Class 2
included those cells modeled by the Cortez reconstructions (DNs# 14-54) representing elevations 1,605-1,932 m, which covered 40.97%
of the study area. Class 3 included the Ignacio
reconstructions (DNs# 5S-71) representing
elevations 1,933-2,068 m, which covered
33.44% of the study area. Class 4 included
cells modeled by the Mesa Verde reconstructions (DNS# 72-93) representing elevations
2,069-2,244 m, which covered 20.45% of the
study area. Class 5 included cells modeled by
the Ft Lewis re-constructions (DNs# 94-189),
representing elevations 2,245-3,012, which
covered 2.32% of the study area.
Soil Data
As described in Chapter 2, information on
soil type and location was collected and recorded by hand for each 200 x 200 m cell in the
study area. Cells were grouped in blocks of
100 cells, 10 rows (west to east) by 10 columns
-95 -
(north to south). Each block was identified by
the UTM coordinate of the northwest comer of
the northwestenunost cell. This point corresponds to some readily visible tick mark on a
7.5-minute USGS topographic map or orthophoto quad for ease in location. In this manner, a total of 502 blocks representing 50,200
cells were recorded in the Soil Conservation
Service field office.
The first task, then, was to input these data
into a computer program that could be read by
VICAR and would allow each block to be inserted into a pre-specified "picture window"
in a fashion similar to placing puzzle pieces
into an image space. A mainframe editing program, WYLBUR, accepted the original soil data
input as numeric and alphanumeric codes.
Later, all alphanumeric codes were reexpressed
as 113 numeric codes, each representing a different soil map unit, to facilitate computer processing.
Next, an overall image or "picture window" was created with VICAR program GEN
to encompass the full set of soil<oded blocks
in their proper geographic positions. The image's overall dimension were taken from the
maximum dimensions of the three (north to
south) by four (west to east) grid of topographic maps equivalent to the 12 mosaiced
DEMs. The VICAR program QSAR pieced
each l00<ell block into its appropriate location within the larger picture window. After
extensive checking for code errors and block
mismatches, the finished 214 row x 227 column image was downloaded to the IBM microcomputer and assigned an EPPL7 filename
(CRVSOD....EPP) and was ready for EPPL7
processing.
The initial EPPL7 processing task was to
register the soil plane with the elevational plane
making sure they were georeferenced to the
same earth coordinates and expressed the same
extent of areal coverage. An auxiliary computer system, International Imaging Systems
(l2S), was used to check for basic registration.
An I2S command, FLICKER, alternately displayed both elevation and soils planes in such
quick succession that it was easy to see whether
or not map features fell into alignment A need
for a small adjustment to the position of the
soils plane was "detected by this method and
was easily made with EPPL7. EPPL command
WINDOW permitted the coregistration and
windowing of the soil plane that would match
the elevation plane and resulted in the creation
of a 200 row x 227 column soil plane (SOILWIND.EPP).
A third soils plane (lIAWC.EPP) was then
created with command RECLASS that grouped
all 113 potential soils into the study area into
II groups of soils based on their available
water holding capacity. This was the same
grouping of soils used in the calculation of
PDSI values and the long-term PDSI reconstructions.
New Data
After the creation of 5ELEV.EPP, which
identified the elevational zones to be modeled
by the five weather stations, and I IAWC.EPP,
which grouped all soils in the study area into
one of II soil groups based on soil moisture
characteristics, a new data plane was created
which combined these two. This data plane,
PDSIMAP.EPP, was created with EPPL command EVALUATE. It created 55 classes where
each class corresponds to a long-term PDSI
reconstruction and any given cell was associated with its assigned available water holding
class and its assigned elevational zone. In the
case of this study, only 44 of the 55 long-term
reconstructions were actually represented in the
study area, but their numbers ranged from 155, despite nonrepresented classes. The creation of PDSIMAP.EPP greatly facilitated the
processing of the final data planes and provided the link between the elevation and soils
data and the long-term PDSI reconstructions.
Arcbaeolo&ical Data
Block Survey Areas
The perimeters of the two survey areas selected to test the model's effectiveness, the
Mockingbird Mesa and Sand Canyon localities,
were digitized with the EPPL7 DIGITIZE program. These survey areas bad been plotted on
USGS 7.5-minute maps and were digitized with
standard digitizer table and mouse equipment
connected to the microcomputer containing
-96-
the digitizing program. The digitized files.
MOCK..DOT and SAND.DGT were used to
create survey areashaped "windows"
(MOCK..EPP and SAND.EPP) within the larger
study area. These windowed areas were irregular in shape. MOCK.EPP depicts the Mockingbird Mesa Survey Locality and incorporates
17.96 km 2 of land surface. SAND.EPP portrays the Sand Canyon Survey Locality and
encompasses 26.08 km 2 of land. They were
used in separate runs of the final model to
depict contrasting agricultural productivity in
two subregions of the study area and to compare the results with that of the entire study
area.
TreeRing Dated Sites
Nine sites were selected to test the model's
ability to discern spatial and temporal patterning in agricultural productivity on a local scale.
Two of these nine, 5MT6970 (Wallace Ruin)
and 5MT4126 (Ida Jean Ruin), are very close
to each other in time and in space and may, in
fact, be part of the same architectural complex.
Consequently, they were combined for the
purpose of this analysis, and a point halfway
between the two chosen to represent the center
of the site. The site number for Wallace Ruin
was retained and the site number for Ida Jean
dropped. Therefore, the UTM coordinates of
the eight sites were entered into an EPPL7 data
file via a microcomputer editing program,
KEDIT. The EPPL commands GRIDPOINT,
BUFFER, and EVALUATE were used to create
an eightcell (1.6 km) radius aroWld the cell
containing a given site. This created a roughly
circular, 3.4kmdiameter (7.88 km 2), "site
catchment" area for each of the eight sites
(5MT8371 5MT8839, 5MT2433, 5MT3834,
5MT6970, UセTQRVL
5MT2149, and
5Mf765). These were used later in eight separate iterations of the final model and generated annual data on productivity within the
area of the catchment.
LINKING SOn.. MOISTURE CONDmONS
AND LEVELS OF AGRICULTURAL
PRODUCTIVITY
As discussed in the section of this chapter
on data acquisition, information on historic
yields of drylaIid bean and maize was gathered
to address two methodological problems. The
first problem was to establish the nature and
the strength of the relationship between modem crop yield and modem soil moisture conditions as modeled by Palmer Drought Severity
techniques. If a relationship could be demonstrated, it would provide the shape of the function that must be fitted to the PDSI values generated by the longtenn PDSI reconstructions
and to the yield estimates of prehistoric agricultural production. The second problem was
to develop a method whereby specific yield
valuescurrently available for only about half
of the soils in the study area could be estimated for all soils. If these two problems
could be solved, crop yield estimates could be
associated with each soil type and each soil
linked to its appropriate 1,070year PDSI
reconstruction via GIS technology.
Historic Yield and Soil Mojsture Conditions:
Establishing a Relationship
Yield values for pinto bean and maize production grown by dry farming techniques have
been compiled by Bums (1983) for five COWlties in southwestern Colorado. Since the vast
majority of the study area is located in Montezuma County, only those records for Montezuma COWlty (Bums 1983:315) were used
for this particular study. These yield figures
are presented as poWlds of beans per harvested
acre or bushels of maize per harvested acre and
are presented as mean values (Table 3.8, columns 2 and 3). Bums' values for 19311960
were used in this analysis. The selection of
1931 as a begirming date was conditioned by
the availability of joint monthly precipitation
and temperature records for Cortez, Colorado,
which were required for the calculation of
Palmer Drought Severity Indices during the
historic period. The ending date of 1960 was
determined by the lack of maize data after that
time, even though bean data continued to be
recorded. As a commercial venture, dryland
beans grown in southwestern Colorado were
successful and quite competitive in regional,
national, and even international markets, but
maize grown in southwestern Colorado was
not Consequently, maize was decreasingly
grown as a commercial product and official
records end at 1960. Thus, only 30 years of
jointly available crop yield data, which were
97-
Table 3.8. Historic Crop Yields and PDSI Values. A.D. 1931-1960.
Bean Yield Maize Yield
Year
ACT ROC
ACT ROC
REC ROC
A.D.
(OR1G) PDSlc (GIS) PDSld
(lbs/ac)8
(bu/ac)b
PDSt e
1931
145
115
270
2.53
9
5.31
173
151
1932
350
12
-1.18
108
1933
12
113
240
1934
87
-3.34
4
99
170
101
-1.94
138
1935
240
10
88
-3.15
114
1936
280
9
128
143
1937
0.80
12
300
143
117
2.32
12
1938
360
106
-1.41
112
1939
260
9
106
125
-1.39
12
1940
510
169
157
4.87
1941
400
20
179
139
5.92
15
1942
350
144
2.41
120
1943
16
740
150
127
3.03
1944
12
380
118
125
-0.23
1945
16
430
92
-2.14
99
18
1946
420
123
-1.95
100
20
1947
600
133
0.63
126
1948
410
13
134
145
1.35
1949
480
19
97
83
-2.30
16
1950
350
67
-5.33
79
1951
15
170
146
131
2.61
1952
15
520
94
123
14
0.26
1953
480
99
108
1954
14
-2.08
330
103
96
14
-1.71
1955
530
101
-1.87
93
1956
304
18
173
115
1957
5.30
692
19
171
129
1958
16
5.13
350
96
81
-3.87
1959
220
15
100
132
16
-2.00
1960
280
aBean yield from Burns (1983:315).
bMaize yield from Bums (1983:310).
cACTROC, computed POSI values for ROC soil (reconstruction #21; C 0.9910.26 in).
dACTROC. in reexpressed POSI values in GIS units ranging from 1-240.
eRECROC, reconstructed POSI values (GIS units) for ROC soil.
fREC34, reconstructed mean POSI value for 34 soil group reconstructions (GIS units).
coterminous with the historic PDSI data, were
used in this particular study.
Historic Crop Yield and Total Regression
The Soil Conservation Service provides
nonirrigated bean yield data in the form of
pounds of beans per harvested acre for 46 dif
ferent soil types in Montezuma and Dolores
Counties (Table 2.3, column 9). Fortyfour of
these 46 occur in the study area. Of these 44
the soil type mapped as ROC, Witt Loam 36%
slope is by far the most abundant agricultural
soil in the study area. A total of 5,868 cells,
representing 15·.96% of the 36,759 analyzable
REC34
PDSlt
102
137
104
93
132
110
134
127
112
114
156
128
104
118
121
89
130
130
142
97
80
146
102
109
110
99
133
122
80
126
cells, was assigned this soil type.
For the purpose of this study, a combined
measure of the 44 soils currently evaluated for
modern dry bean farming, as well as two mea
sores of the ROC soil as it is modeled at the
Cortez elevation, were used to examine the re
lation between crop yield and soil moisture
conditions. In order to create a combined
measure that might better reflect the mean
yield values for Montezuma County compiled
by Burns (1983) (Table 3.8, columns 2 and 3),
nested frequency counts of the 55 PDSI re
constructions and all of the soil types in the
study area were examined. Often a soil occur-
-98 -
red in more than one elevational range. In
addition, different soil map units, because of
similarity in their waterholding capacities, fell
into the same POSI reconstruction (Table 3.9).
The 44 different soil types for which yield data
had been compiled were associated with 34 of
the 55 reconstructions. The POSls for each of
the 34 were summed annually from 19311960 and an annual mean was taken. This
mean index value was called "REC34" (Table
3.8, column 7). Similarly, but more simply, the
annual POSI value for reconstruction #21 (Le.,
those soils that fell into the soil class defmed
on its available water holding capacity of 0.99
inches of water in the upper six inches of soil
and 10.26 inches of water in the lower soil
profile using the Cortez climatic reconstruction), which includes ROC (Witt Loam, 3-6%
slope), was recorded for the 1931-1960 period
(Table 3.8, column 6), as was the original POSI
calculation for the historic period (Table 3.8,
columns 4 and 5).
First it was necessary to establish that a
strong relationship existed between the reconstructed POSI estimates and the actual POSI
calculations for the 30 years under consideration so that a case could be made for the usefulness of the reconstructed series for modeling soil moisture. A plot of the reconstructed
POSI values for the ROC soil and the actual
POSI calculations for the ROC soil during the
1931-1960 period produced a positive linear
regression line with an r = .65 (significant at
the .01 level). Values greater than .80 are considered high, around .50 considered moderate,
and below .30 considered low (Downie and
Heath 1970:101). While not perfect, regression
indicates that these two data sets are moderately
to highly correlated and that the reconstructed
values are reasonably accurate predictors of the
actual calculated values.
Second, the actual POSls and the reconstructed POSIs for 1931-1960 were regressed
against bean yield, first for the ROC soil and
then for the 34 soil groups (Table 3.10, middle
column, top half). The highest linear correlation was between actual POSI values for ROC
soil and bean yield, with an r = .41. Although
this is significant at the .05 level, the correlation is rather low. AU of these plots were better
represented by second-degree polynomial
equations in which an inverted U or a parabolic, curvilinear shape was fitted through the
points. Again, the actual POSI data for ROC
soil when plotted against the bean yield data
produced the best correlation of the test attempts (r .51).
=
Third, linear and curvilinear regression
lines were fitted to the maize data in the same
Table 3.9. Distribution of Soils for Which Nonirrigated Bean Yield Data Are Available by
PDSI Reconstruction Class.
Recon. Weather
Station
Bluff
Bluff
Bluff
Bluff
Bluff
Bluff
Bluff
Bluff
Bluff
10
Bluff
Bluff
11
Cortez
12
Cortez
13
Cortez
14
Cortez
15
Cortez
16
17
Cortez
#
1
2
3
4
5
6
7
8
9
AWCValues
(Su + SI)8
0.53 4.61
0.85 2.95
1.02 3.65
1.02 4.48
1.02 5.40
1.14 6.44
1.02 7.80
1.00 8.54
1.05 9.45
0.99 10.26
1.14 13.66
0.53 4.61
0.85 2.95
1.02 3.65
1.02 4.48
1.02 5.40
1.14 6.44
Soil Map Unit
M2DD, M2DD1
ROLC, ROLD, ROLD1
R7C,R7D,R7D1
H1CD, R4D, ROHe, ROHD
RHC
-99-
Table 3.9 (Concluded).
Recon. Weather
#
Station
18
Cortez
19
Cortez
20
Cortez
21
Cortez
22
Cortez
23
Ignacio
24
Ignacio
25
Ignacio
26
Ignacio
27
Ignacio
28
Ignacio
29
Ignacio
30
Ignacio
31
Ignacio
32
Ignacio
33
Ignacio
34
Mesa Verde
35
Mesa Verde
36
Mesa Verde
37
Mesa Verde
38
Mesa Verde
39
MesaVerde
40
Mesa Verde
41
Mesa Verde
42
Mesa Verde
43
Mesa Verde
44
Mesa Verde
Ft. Lewis
45
46
Ft. Lewis
4 7 F t . Lewis
Ft. Lewis
48
Ft. Lewis
49
50
Ft. Lewis
51
Ft. Lewis
52
Ft. Lewis
53
Ft. Lewis
54
Ft. Lewis
55
Ft.Lewis
AWC Values
(Su + Sl)a
1.02 7.80
1.00 8.54
1.05 9.45
0.99 10.26
1.14 13.66
0.53 4.61
0.85 2.95
1.02 3.65
1.02 4.48
1.02 5.40
1.14 6.44
1.02 7.80
1.00 8.54
1.05 9.45
0.99 10.26
1.14 13.66
0.53 4.61
0.85 2.95
1.02 3.65
1.02 4.48
1.02 5.40
1.14 6.44
1.02 7.80
1.00 8.54
1.05 9.45
0.99 10.26
1.14 13.66
0.53 4.61
0.85 2.95
1.02 3.65
1.02 4.48
1.02 5.40
1.14 6.44
1.02 7.80
1.00 8.54
1.05 9.45
0.99 10.26
1.1413.66
Soil Map Unit
R4C, R4C1
R3B, R3C, R3D, R3B1, R3C1, R3D1, WOBB
A5C, WOB
ROB, ROC, ROD, ROB1, ROC1
M2DD
ROLC, ROLD
R7C,R7D
H1CD,R4D,ROHC,ROHD
RHC
R4C,R4C1
R3B, R3C, R3D, R3C1, WOBB
ASC, A5D, WOB
ROB, ROC, ROD
M2DD
ROLC, ROLD
R7C,R7D
H1CD, R4D,ROHC, ROHD,S6C,S7
RHC
R4C, S5
R3C,R3D,S51,S5D,S8C,V4C,WOBB
ASC, ASD, V1 C, WOB, WOC
M5E, ROB, ROC, ROD, SSE
R1C,R1D
S6C, S7
S5
S5D, V4C, S8C
V1B, V1C
M5E, SSE
R1C,R1D
Note: This list associates 44 different soil types with 34 different PDSI reconstructions. Some
soil types occur with more than one reconstruction because they occur in different elevational
ranges. Two soil types listed in Table 3.3, C3CD and V1 D, are not present in the study area but
are located elsewhere within Montezuma and Dolores Counties and are associated with Reconstructions #52 and 53, respectively.
aAWC Values (Su +SI): Available waterholding capacity of upper six inches of soil and lower soil
profile.
way they had been to the bean data (Table
.3.10, middle column, bottom half). Again,
actual and reconstructed PDSI values for the
single reconstruction for ROC soil and the
combined reconstruction for 34 long-term reconstructions were used. Unlike the bean data,
the best linear equation was between the com-
bined 34 reconstructions instead of with the
single ROC. reconstruction. Despite this, none
of the linear equations are considered significant (at the .05 level), the best having only an r
= .25, which is considered to be low. As with
the bean data, all the maize plots were better
modeled by second-degree polynomial
100 -
Table 3.10. Results of Linear, Polynomial, and Partial Regression Studies of the Relationship Between
PDSI and Crop Yield Values. For the 30-year period, the mean bean yield = 380.53 ± 141.16
pounds/acre and the mean maize yield = 14.07 ± 3.69 bushels/acre.
Beans
Actual PDSI
for ROC
Reconstructed
PDSI for ROC
Mean
for REC34
Simple Linear Regression and
Polynomial Regression
y = 152.2844 + 1.8683x
f = .41 ,2 = .17 (significant at .05)
Partial Regression
y = 9782.55180 + 5.09523678x1 +
2.04829764x2 f= .52,2= .27 (significant at
.01)
y = -566.7831 + 13.8224x-0.0468x2
f= .51 ,2 = .26 (significant at .01)
y = 10.971.18229 + 5.71318142x1 +
y = 238.016 + 1.2064x
f= .17 ,2 = .03 (not significant at .05)
2.00384746x2 f= .38 ,2 = .14 (Significant at
.05)
y = -724.6224 + 18.1378x
-0.0723x 2 f=.29,2 = .08 (not
significant at .05)
y = 9667.04731 + 4.99434456x1 +
y = 82.1425 + 2.5672x
f= .35,2 = .12 (not significant at .05)
2.84843676x2 f = .47 ,2 = .22 (significant at
.01)
y = -633.0555 + 15.2679x 0.0549x 2
f= .38 ,2 =.14 (significant at .05)
Actual PDSI
for ROC
Simple Linear Regression and
Polynomial Regression
y = 10.8752 + 0.0261x
f= .22 ,2 = .05 (not significant at .05)
Reconstructed
PDSlforROC
y = 15.8618 0.0568x + 3.243e4x2
f= .23 ,2 =.05 (not significant at .05)
y 12.3279 + 0.0147x
f= .08 ,2 = .01 (not significant at .05)
Maize
=
y
Mean
for REC34
Partial Regression
y = 525.61342 + 0.27517129x1 +
0.03546922x2 f=.69,2 = .47 (significant at
.001)
y = 580.67051 + 0.30224351x1 +
0.05690542x2 f= .69 ,2 = .47 (significant at
.001)
=44.5n2 0.5525x + 0.0024x2
f= .30,2 = .09 (not significant at .05)
y = 8.5048 + 0.0479x
f= .25 ,2 .06 (not significant at .05)
=
=45.0841 0.6017x + 0.0028x2
f= .41 ,2 = .17 (significant at .05)
y = 532.58051 + 0.27718880X1 +
0.06346171x2
f =.70 ,2 =.49 (significant at .001)
y
equations. As with the linear model, the combined 34 reconstruction PDSI model had the
highest correlation coefficient (r .41) and
was significant at the .05 level, but is still
considered weak. However, this curvilinear .
equation, as were the other two, was fitted with
an upright Vshaped curve. The results did not
=
fulfill expectations about how the yield data
were related to soil moisture conditions.
Historic Crop Yield and Partial Regression
In general the results of the above analyses
were not very convincing. However, another
101-
course of action still remained to search for
pattern in the data that might provide the relationship needed to relate soil moisture conditions and crop yield. This was to use partial
linear regression on the data with, "time" also
considered as an independent variable. With
partial regression techniques, the contribution
to crop production by the first independent
variable, time, can be controlled for and the
unique contribution made by the second independent variable, soil moisture, assessed. This
was thought to be potentially very useful because time might also incorporate the influence
of modem technological changes on crop
yield. Burns (1983:70-73) refers to this as the
"technology trend" and includes the role that
technological components such as tractors,
fertilizer, pesticide, fungicide, seed selection,
and cultivation practices play in effecting crop
yield. He used the application of fertilizer to
crops as a proxy measure of the general influence of technology on crop production,
although he recognized that the use of chemical fertilizers on dry farming lands may be
questioned (Bums 1983:75-76).
The partial regression study was accomplished by specifying regression options in the
Statistical Analysis Systems (SAS) program,
REG. It was performed first on the bean data
and then on the maize data; summary statistics
and residual scatter plots were produced for
each. Three models were specified for each
crop type. The first specified that the dependent variable (crop yield) would be regressed
on independent variables year (time) and actual PDSI values for the single ROC soil group
for 1931-1960. The second specified that the
regressors were year and reconstructed PDSI
values for the ROC soil. The third specified
that the regressors were year and the mean
value for the 34 soil groups, which represented
the 44 yield-documented agricultural soils.
The data used in the study were those listed in
Table 3.8 and also used in the linear and polynomial regression studies. For the purpose of
comparability between the bean and maize
data, the 30-year period of 1931-1960 was
used for both crops.
The results of the analysis are summarized
in Table 3.10 (right column). The equations
and regression coefficients generated for the
same data for 1931-1960 period using simple
linear regression and the second-degree polynomial regression are included for comparison. The use of partial regression was successful in increasing the correlation coefficient (r)
and the index of determination (r 2 ) in every
case. Where only one simple linear equation
for bean yield and the actual PDSI calculations
was successful in producing a significant correlation, all six of the equations derived under
the partial regression method produced statistically significant correlations. Under both
simple linear and partial regression, the bean
yield was better explained by the actual PDSI
calculations for ROC soil than by the reconstructed PDSIs for ROC or the mean value of
the combined 34 soil groups. Also, under both
simple linear and partial regression, the combined 34 reconstructions model explained
more variation in maize yield than either of the
ROC data sets. In comparing the partial regressions on the bean data versus those on the
maize data, the correlation coefficients of the
maize models increased dramatically and moderately strong, positive, linear relationships
appeared between soil moisture and maize
yield. The "technology trend", as proxied by
the variable time, was much more strongly
reflected in the maize than in the bean data.
Once the contribution of this variable was removed, the relationship between increasing
amounts of soil moisture and increasing crop
production was visible graphically and numerically in the correlation coefficients. In
contrast, the bean data seemed to take on a
more curvilinear function with increasing soil
moisture correlated with increasing bean yields
up to a point and then decreasing yield with
the highest levels of soil moisture.
In sum, partial regression analysis succeeded in demonstrating a relationship between
PDSI values and crop yields. Of most importance was the partial regression for maize yield
and the 34 combined reconstructions, which is
thought to best reflect the soil types in the
study area from which the reported maize yield
was derived. This regression indicated that a
positive linear relationship exists between
reconstructed PDSI values and modem maize
production. The partial regression not only
demonstrated a significant relationship between
soil moisture and crop yield (r = .70, r 2 = .49,
significant at .(01), but it also correctly
Historic Crop Yield and Natural Plant.
Productiyity: Estimating Prehistoric Crop
Yields for IndiYidual Soils
The second methodological problem was
to develop a procedure whereby potential crop
yield values could be assigned to each soil type
in the study area and could vary under different climatic and soil moisture conditions. Such
a method was developed and it involved the use
of modem bean yield data for specific soils
compiled by the SCS (Table 2.3, colwnn 9)
and potential natural plant productivity data
for all soils estimated under favorable, unfavorable, and normal growing conditions (Table
2.3, columns 6-8). Ultimately, these estimates
of productivity expressed as bean units would
have to be reexpressed as maize units, for the
reconstruction of maize production is one of
the goals of this research.
The mean yield estimates for pinto beans
grown on nonirrigated lands in Montezuma
and Dolores Counties have been compiled by
the Soil Conservation Service. These estimates
have been made by SCS agents on the basis of
reported harvests by individual farmers who
commercially grew beans on these soils for a
number of years. No comparable maize data
exist. These values are associated with specific
soil types. Of the 113 different soil map units
in their two-county jurisdiction, the SCS has
compiled data on mean yield from 46 different
soil map units. Forty-four of these 46 are present in the study area and represent 45% of all
the 98 soil map units (by number) that actually
occur in the study area. The soils for which
yield data have been compiled represent the
majority of local soil types that modem dry
farmers use to grow pinto beans in commercial
quantities. These farmers use mechanized
equipment and regularly apply pesticides and
fungicides but less often use fertilizers to grow
beans on nonirrigated lands. Because their
capital investment is relatively great and the
price of harvested beans relatively low, they
must plant and harvest a significant area to
make a profit Consequently, their fields are
generally on interior mesatop locations where
the soil is deeper and can hold greater soil
moisture, where slope does not present major
problems for equipment and slope erosion, and
where large patches of arable land occur. All
this is to say that these 46 soils are not the only
ones that can successfully grow nonirrigated
crops, particularly at the scale and technological level believed to have been employed by
Anasazi farmers. Therefore, it is not unreasonable to suppose that additional soil types were
also used by prehistoric farmers of the area,
some not included on the list of 46.
Therefore, a general method for estimating
crop yield values on all soils in the study area
was devised. First, a linear regression equation
derived from the relationship between bean
yield and potential natural plant productivity
on the 44 known soils was used to estimate
crop yield values for the entire set of 98 soils
in the study area. The values were plotted, the
least-squares regression line calculated for the
scatter (y = -116.4617 + 3.1745x), and the
correlation coefficients determined to be significant (r .77, r 2 .59, significant at the .01
level) and valid for the purposes of linear estimation. Using this model, the natural productivity values for all soils are used to derive
predicted bean yields for all soils. This results
in a series of estimated crop yields for all the
soils in the study area grown under "normal
climatic conditions" (taken to mean average
precipitation and temperature conditions).
Next, the estimates of natural productivity
grown under favorable and unfavorable climatic conditions for all soils were used to estimate crop yields for all soils. Although this
could have been accomplished again through
linear estimation from the regression equations,
an alternative method was used. Here the percentage change from favorable to normal conditions and again from unfavorable to normal
conditions was calculated, and this percent used
as a multiplier to the normal value. For example, the natural productivity for one soil is
1,200 pounds per acre in favorable years, 900
pounds per acre in normal years, and 700
pounds per acre in unfavorable years. If the
crop yield for normal years was eStimated to be
320 pounds of beans per acre, then 320
pounds is multiplied by 1,200/900 or 1.33 to
derive 427 pounds of beans per acre for favorable years, and by 700/900 or .78 to derive
249 pounds of bean per acre for unfavorable
years. Subsequent to calculating the favorable
and unfavorable categories of yield, two intermediate sets of values representing a condition
between normal and favorable, and another
between normal and unfavorable, were created
-103 -
=
=
so that five different "states" or "choices"
were possible, and these five could be fit to the
PDSI values. This was accomplished simply by
halving the difference between normal and
favorable yield estimates to provide normal-tofavorable category and again by halving the
difference between normal and unfavorable
estimates creating a normal-to-unfavorable
category. Table 3.11 provides bean yield estimates for the five categories.
The Relationship Between Bean Yield and
Maize Yield
The values created by the above steps create a matrix of values where the crop yield reconstructed is dry bean production. This was
necessary because only values for dry bean
production per soil type were available from
the Soil Conservation Service for lands in
Montezuma and Dolores Counties. However,
the estimation of bean production for all soil
types under varying climatic and soil moisture
conditions is a necessary but intermediate step
toward the estimation of maize production under varying climatic and soil moisture conditions, for maize is really the crop of interest
relative to prehistoric Pueblo subsistence. Two
points must be emphasized with regard to these
bean yield estimates. First, as individual values
these numbers are only bestguess estimates
based on the leastsquares regression formula;
a real world value might be somewhat different. Second, these values are relative estimatesproduction values relative to the same soil
under different growing conditions, and relative to all other soil types present in the area.
Likewise, when the relationship is established
between bean growth and maize growth, and
when these relative production units are converted from beans to com, these qualities of
approximation and relativity must also be seen
to be associated with the resulting maize yield
estimates.
The relationship between bean and maize
yield was established by taking the mean value
of production of beans and for maize for the
combined Montezuma and Dolores County
area for the 1931-1960 period (Tables 3.12
and 3.13). The combined data were thought to
provide a larger, acreage-weighted sample that
might better reflect trends in the data. The crop
yield means were set equal to each other: 368
pounds per acre for nonirrigated beans and
13.9 bushels per acre (equivalent to 778
pounds per acre, where one bushel weighs 56
pounds) for nonirrigated maize. These numbers compare favorably with the long-term
Table 3.11. Table of Productivity Values For All Soil Types Within the Study Area Under Five Different
Climatic Conditions.
Soil
Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Map
Unit
A1B
A2B
A3B
A4
A41
A4A
A4B
A5C
A5D
A5W*
A6A
AOAB
AOB
AOC
AO*
BD
C2CE
C2D
Favorablea
FlNb
NormaJC
NlUd
Unfavorable 9
328
334
427
427
343
330
427
427
281
288
374
374
316
294
374
374
234
241
320
320
289
257
320
320
201
204
285
285
253
221
267
267
168
167
249
249
217
184
213
213
948
292
367
367
469
0
245
218
878
259
312
312
411
0
204
175
808
226
257
257
352
0
163
131
751
198
221
221
293
0
143
109
693
170
184
184
234
0
122
87
-
104-
Number
of Cellsf
296
108
55
315
134
56
201
293
35
0
36
230
168
103
93
42
377
44
Percent
of Area9
.77
.28
.14
.82
.35
.15
.53
.77
.09
.00
.09
.60
.44
.27
.24
.11
.99
.12
Table 3.11 (Continued).
Soil
Code
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Map
Unit
C2F
C2V
C3CD
C3E
C5C
C7D
CP
DOCE
GP
H1CD
H1DC
H6D
M1CE
M2C
M2CE
M2DD
M2DD1
M5E
M2C
M7E
M8D
OAB
P1C
R1C
R1D
R1HB
R1HC
R1HD
R3B
R3B1
R3C
R3C1
R3D
R3D1
R4C
R4C1
R4D
R7C
R7D
R7D1
R8B
RHC
RL
ROB
ROB1
ROC
ROC1
ROD
ROHC
ROHD
ROLC
ROLD
ROLD1
Favorable a
253
218
F/Nb
213
175
Normalc
173
131
NJUd
147
109
Unfavorabies
120
87
216
351
0
272
0
397
184
299
0
233
0
343
152
246
0
194
0
289
119
211
0
155
0
235
86
176
0
116
0
181
210
334
203
167
292
242
494
470
491
433
301
344
580
580
179
288
164
135
252
213
454
426
439
368
264
277
529
529
147
241
125
103
211
183
414
382
387
302
226
210
478
478
116
223
97
80
176
154
362
339
350
268
189
172
427
427
84
204
69
56
141
124
310
296
313
233
151
134
376
376
528
294
528
294
528
294
489
326
475
416
416
283
334
483
0
519
437
519
437
519
406
406
427
427
283
440
276
440
276
440
276
423
288
405
361
361
255
288
404
0
467
355
467
355
467
348
348
374
374
255
352
257
352
257
352
257
356
250
335
305
305
226
241
325
0
415
273
415
273
415
289
289
320
320
226
317
202
317
202
317
202
312
207
300
269
269
198
204
292
0
346
228
346
228
346
253
253
285
285
198
282
147
282
147
282
147
268
163
264
233
233
170
167
258
0
277
182
277
182
277
217
217
249
249
170
-105 -
Number
of Cellsf
104
49
0
0
387
43
9
10
36
200
0
67
36
3617
6322
3150
1093
369
25
15
21
59
45
116
29
0
0
0
3
60
137
199
462
75
1075
375
938
153
102
61
75
500
413
982
54
5868
152
1058
130
69
219
130
75
Percent
of Area9
.27
.13
.00
.00
1.01
.11
.02
.03
.09
.52
.00
.18
.09
9.46
16.54
8.24
2.86
.97
.07
.04
.05
.15
.12
.30
.08
.00
.00
.00
.01
.16
.36
.52
1.21
.20
2.81
.98
2.45
.40
.27
.16
.20
1.31
1.08
2.57
.14
15.35
.40
2.77
.34
.18
.57
.34
.20
Table 3.11 (Concluded).
Map
Soil
FavorUnfavorNumber Percent
NlUd
F/Nb
Unit
able a
Nonnalc
Code
able 9
of Cells f of Areag
72
S1C
294
257
276
221
184
365
.95
73
S5
611
560
509
458
407
8
.02
74
S51
609
541
507
676
473
40
.10
454
414
345
75
494
276
S53
250
.65
541
507
473
76
676
609
S50
64
.17
77
405
351
455
296
S6
504
61
.16
419
385
350
78
S6C
519
469
16
.04
289
244
199
343
79
397
278
S6E
.73
409
357
460
80
S7
544
502
14
.04
339
293
385
81
424
S8C
463
204
.53
82
T1B
0
.00
294
267
374
320
83
427
2
.01
T2C
116
147
179
84
14
84
210
T3E
.04
0
T4B
85
.00
0
86
T6
.00
87
0
.00
T6B
0
88
T6C
.00
0
89
T60
.00
470
604
722
336
4
839
90
V1B
.01
470
722
604
336
839
34
91
V1C
.09
92
V10
0
.00
281
234
201
328
168
93
V2B
552
1.44
234
201
94
V2C
281
168
123
328
.32
137
225
186
162
95
V2S
264
59
.15
V2V·
328
281
234
201
168
96
248
.65
137
V2W*
225
162
.97
264
186
146
.38
487
772
369
31
98
V4C
688
604
.08
257
221
184
VOC
330
294
5
.01
99
257
184
100
VOC
330
294
221
49
.13
101
VOCB
294
257
221
184
3
330
.01
VOCC
294
257
221
184
14
102
330
.04
103
WOB
361
289
217
253
325
173
.45
WOBB·
104
361
289
217
325
.42
253
159
105
WOC
361
325
289
253
217
7
.02
XC60
210
179
147
116
106
84
1083
2.83
107
XC6E
158
135
111
87
63
1572
4.11
119
XH20
206
181
156
4
108
82
.01
XH2E
170
145
120
94
109
68
275
.72
XH3E
217
186
155
124
110
93
23
.06
111
><M1F
0
.00
158
137
115
97
><M30
79
537
112
1.40
143
115
72
86
57
55
.14
113
Z
Note: All productivity values are Soil Produdivity Units Based on Beans (SPUBs) and are given in pounds per acre.
aFavorable Growing Season (assigned to PDSI Classes 10 and 11, Unusually Wet and Extremely Wet).
bFavorabletoNormai Growing Season (assigned to PDSI Classes 8 and 9, Slightly Wet and Moderately Wet).
cNormal Growing Season (assigned to PDSI Classes 5, 6, and 7).
dNormaltoUnfavorable Growing Season (assigned to PDSI Classes 3 and 4, Moderate and Slight Drought).
eUnfavorable Growing Season (assigned to PDSI Classes 1 and 2, Extreme Drought and Severe Drought).
fEach cell is 200 x 200 m (4 hal in area.
gThis is the proportion of a given soil type within the analyzable portion of the study area. The total study area
contains 45,400 cells. Of these, 7175 have not yet been mapped for soil associations. Therefore, only 38,225 cells
have been assigned a soil value.
•Artificially wet and possibly saline from irrigation water; consequently the natural potential productivity values
used to generate these yield estimates were reduced to better approximate the original, nonirrigated conditions
(SCS Soil Lエセゥョ・」s
Alan Price, personal communication, March 1989).
106 -
(A.D. 6521968) "adjusted" (Le., technology-free) mean estimates made by Bums
(1983) who used 374 pounds per harvest acre
for dry beans and 12.7 bushels per harvest acre
for nonirrigated maize (711 pounds per acre).
Additionally, the maize value compares favorably with Petersen's (1987) 1919-1960 estimate for mean maize production as 13.98
bushels (783 pounds per harvested acre). Finally, the bean value compares favorably with
the mean bean production value of 362
pounds per acre taken from the 46 different
soils as reported by the SCS.
Second, the bean yield values, which
ranged from 0-948 pounds per acre (Table
3.11), were reclassified as soil productivity values that must not exceed the 0-255 maximum
range defined by the GIS. This was accomplished simply by dividing each value by 10
and truncating the number to the right of the
decimal (e.g., 948 becomes 94.0) and assigning each to a new set of 94 classes. In this way,
a new range of soil productivity class values
ranging from 0-94 was created. Each class
preserves its relative measure of productivity
and is usable by the GIS. For example, for soil
#2 (A2B, Sili Silty Clay Loam, 0-3% slope),
the unfavorable through favorable growing
season yields expressed in bean units are 168,
201,234,281, and 328 pounds per acre. When
reexpressed these become 16, 20, 23, 28, and
Table 3.12. Nonirrigated Bean Data (in pounds) from Dolores and Montezuma Counties Used to
Generate a Mean Bean Yield Value for Years A.D. 1931-1960. Mean bean yield for this 30-year period is
368.10 ± 126.14 pounds per acre.
Dolores
Year
Co. Yield
1931
250
1932
300
1933
320
1934190
1935
230
1936
260
1937
300
1938
350
1939
320
1940
370
1941
400
1942
350
1943
770
1944
380
450
1945
1946
390
1947
590
1948
400
1949
450
1950
330
1951
250
1952
500
1953
530
1954
340
1955
470
1956
160
1957
700
1958
320
1959
180
1960
280
Montezuma
Co. Yield
270
350
240
170
240
280
300
360
260
510
400
350
740
380
430
420
600
410
480
350
170
520
480
330
530
304
692
350
220
280
Dolores Co.
Acres Harv.
1180
910
3,230
2,484
7,520
5,060
9,770
13,890
17,440
23,950
20,400
20,710
31,000
NA
NA
NA
NA
51,000
44,590
35,650
27,750
29,820
39,640
42,780
40,960
40,020
41,900
42,950
36,130
36,250
Montezuma Co. Dolores Co.
Acres Harv.
Total Production
8,190
295,000
5,090
273,000
6,300
1,033,600
6,774
471,960
14,010
1,729,600
12,510
1,315,600
20,110
2,931,000
30,250
4,861.500
42,670
5,580,800
35,920
8,861,500
28,150
8,160.000
30,100
7,248,000
39,440
23,870,000
NA
NA
NA
NA
NA
NA
NA
NA
49,000
20,400,000
40,520
20,065,000
41,680
11,764,500
25.100
6,937,500
29,200
14,910,000
33,670
21,009,200
51,230
14,545,200
39,660
19,251,200
54,560
6,403,200
58,420
29,330,000
59,620
13,744,000
51,800
6,503,400
52,950
10,150,000
Montezuma Co. Mean
Total Production Per Acre
2,211,300
267.5
1,781,500
342.4
1,512,000
267.1
1,151,580
175.4
3,362,400
236.5
3.502,800
274.2
6,033.000
300.0
10,890,000
356.9
11,094,200
277.4
18,319,200
454.0
11,260,000
400.0
10,535,000
350.0
29,185,600
753.2
NA
*380
NA
*440
NA
*405
NA
*595
20,090,000
404.9
19,449,600
464.3
14,588,000
340.8
4,267,000
212.0
15,184,000
509.9
16,161,600
507.1
16,905,900
334.6
21,019,800
499.5
16,586,240
243.1
40,426,640
432.2
20,867,000
337.4
11,396,000
203.6
14,826,000
280.0
Note: All of these dry bean values for Dolores and Montezuma Counties are derived from Bums (1983:313, 315).
*designates averaged but unweighted yields for years 1944-1947 when complete records were not kept.
-107 -
Table 3.13. Nonirrigated Maize Data (in bushels) from Dolores and Montezuma Counties Used to
Generate a Mean Maize Yield Value for Years A.D. 1931-1960. Mean maize yield for this 30-year period
is 13.9 ± 3.2 bushels per acre (or n8 ± 179 pounds per acre where one bushel equals 56 pounds).
Year
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1969
Dolores Montezuma Dolores Co. Montezuma
Dolores Co.
Co. Yield Co. Yield
Acres HaN. Co. Acres Harv. Total Prod.
9
9
1,030
4,100
9,270
13
12
1,700
5,540
22,100
11
12
1,700
4,690
18,700
5
4
1,950
3,440
9,750
27
10
2,613
3,073
70,551
10
9
2,380
3,010
23,800
11
12
2,830
3,260
31,130
12
12
2,510
3,230
30,120
12
9
2,070
3,050
24,840
14
12
2,610
2,350
36,540
17
20
2,540
1,970
43,180
10
15
1,230
1,480
12,300
15
16
1,470
1,230
22,050
17
12
NA
NA
NA
15
16
NA
NA
NA
17
18
NA
NA
NA
17
20
NA
NA
NA
14
13
960
910
13,440
19
19
1,040
810
19,760
15
16
260
380
3,900
14
15
100
260
1,400
12
15
90
320
1,080
14
14
60
350
840
13
14
360
350
4,680
13
14
310
110
4,030
12
18
250
70
3,000
17
19
320
60
5,440
15
16
230
30
3,450
13
15
240
150
3,120
17
16
30
30
510
Montezuma
Mean
Co.Total Prod. Per Acre
36,900
9.0
66,480
12.2
56,280
11.7
13,760
4.4
30,730
17.8
27,090
9.4
39,120
11.5
38,760
12.0
27,450
10.2
28,200
13.1
39,400
18.3
22,200
12.7
19,680
15.5
NA
*14.5
NA
*15.5
NA
*17.5
NA
*18.5
11,830
13.5
15,390
19.0
6,080
15.6
3,900
14.7
4,800
14.3
4,900
14.0
4,900
13.5
1,540
13.3
1,260
13.3
1,140
17.3
480
15.1
2,250
13.8
480
16.5
Note: All of these maize values for Dolores and Montezuma Counties are derived from Bums (1983:308, 310).
*designates averaged but unweighted yields for years 1944-1947 when complete records were not kep1.
32 soil productivity units (SPUs), respectively.
Next, the mean bean production (368
pounds per acre) is associated with soil productivity unit 36 (which includes any value
between 360 and 369 pounds of beans) as is
the mean maize production of 778. Thus, the
soil productivity unit 36 which is based on
bean production (SPUB) is equivalent to the
mean value of beans and the mean value of
maize.
Finally, an equation converting each.SPUB
into a SPUM (soil productivity unit based on
maize) is developed, so that each and every soil
productivity unit is associated with maize production. The SPU is multiplied by 10 and to
this the number four is added. This number
produces a value that exists at the midpoint of
its class and approximates the original bean
yield estimate. For example, SPU 36 represents
bean yields of 360-369 pounds for which the
range midpoint is 364; it is this number that is
now used to represent bean yield. 1bis number
is then divided by the ratio of 368 to 778 or
.473 (Le., the mean bean yield to the mean
maize yield) to produce the equivalent value in
pounds per acre of maize. As an equation, this
may be written as SPUM (pounds/acre) =
«SPUB *10)+ 4) / 0.473. Ultimately, this
SPUM value must be reexpressed as yield in
kilograms per hectare to be consistent with the
unit of analysis used in the GIS, which is a 4-ha
cell, 200 m on a side. To convert from pounds
108-
per acre to kilograms per hectare, the maize
yield is multiplied by 1.12. Thus, the fmal
equation is SPUM (kg! ha) = «(SPUB * 10) +
4) / O. 473) * 1.12. In order not to loose additional resolution in these maize estimates, it
was decided to apply the SPUB to SPUM conversion in the post-GIS processing of data with
SAS programs, after the GIS had analyzed and
summed all the SPUBs for each year in the
400-year study.
Calibratinl: Palmer Drought Severity Index
Values with Yield Values
To graphically portray contrasting areas of
agricultural production within the study cu:ea,
and to generate absolute yield values that m
tum could be used to estimate populations that
could be sustained by that yield, it was necessary to assign crop yield value.s to PDSI セオ・ウN
Depicting differential soil mOisture condltlOns
is not enough; this must be "translated" in1:O
equivalent measures of agricultural production.
Further, this had to be accomplished in such a
way that the GIS could manage and display the
PDSI and crop yield information.
The first step toward achieving this calibration was data reduction of the PDSI information into fewer, more interpretable categories.
The second was reexpression of original positive and negative PDSI values into GIS-compatible units that would preserve the meaning
and rank order of the original. The third was to
"fit" or "calibrate" the positive linear function created by the partial regression study to e
five categories created by the crop yield and
natural plant productivity study. Once this was
carried out, it was a relatively simple task for
the GIS to integrate these tabular data as
"look-up tables" with the spatial data already
resident in the system.
Data Reduction
In Palmer's (1965:28, Table 11) original
description of his method, he created a classification scheme to interpret the PDSI values that
resulted from his computations. Using Palmer's original classification system as a guide,
a number was assigned to each of the 11 interpretative classes that would retain their rank
order and meaning (Table 3.14). Here Class 1
Table 3.14. Eleven Nominal PDSI Classes.
Original PDSI
Values
PDSI Class
Extremelv Wet
3.00-3.99
Unusuallv Wet
2.00-2.99
Moderatelv Wet
1.00-1.99
Sliahtly Wet
0.50-0.99
Incipient Wet
(-0.49)-(0.49) Near Normal
(-0.50)-(-0.99) Incipient Drought
(-1.00)-(1.99) Slight Drouaht
(-2.00)-(2.99) Moderate Drought
(-3.00)-(3.99) Severe Drouaht
s;(-4.00)
Extreme Drouaht
P NTセ
PDSI
Nominal
Class
11
10
9
8
7
6
5
4
3
2
1
represents the category "Extreme Drought"
G;-4.00 PDSI), Class 11 represents the category
PDSI), Class 6 rep"Extremely Wet" セKTNP
resents the category "Near Normal" (-.49+.49 PDSI), with all other values falling into
intermediate positions (Figure 3.4, rows a and
b).
Data Reexpression
As explained previously, the original PDSI
values are expressed as both negative and positive real numbers. These values had to be reexpressed as positive whole numbers with a
maximum range from 0-255 so as to be manipulated by EPPL7. At the time this was done,
the exact range of reconstructed PDSI values
was not known, but it was known that values
below -6.00 PDSI and above +10.00 PDSI did
occur in the actual calculations for PDSI for
the 55 soil groups used in the study. To assure
that all PDSI values would fall within the true
range, a wide hypothetical range of -12.00 to
+12.00 was selected. These values were reexpressed in GIS Units ranging from 0-240.
This was accomplished by first rounding the
value one decimal place (Le., "rounding up"
where the value is equal or greater than five,
and "rounding down" where the value is less
than five); multiplying the number by 10 (to
produce a nondecimal number); and adding
120 to the number to insure that it be positive
and within the woIkable range of the GIS. For
example, an original PDSI value of 1.64
-109 -
PDSI = 0.00
GIS = 120 M M M M セ
WET
DROUGHT
A.
2
3
EJ.1teme
Severe
Maderale
SliQht
Incip- N.ar Incipient Narmal ienl SliQht
3.00
2.00
1.00
.50
PDSI Categories
8.
Values in
Original PDSI Units
C.
Reexpressed
Values in GIS Units
D.
I
4
Agricultural
Productivity
Categories
セ
4.00
セ
80
to
to
5
10
10
3.99
2.99
1.99
99
8190
91100
101110
111115
6
.49
to
.49
7
8
9
10
II
Madera'.
Severe
Ellrem.
3.00
.50
1.00
2.00
10
10
10
10
.99
1.99
2.99
3.99
116- 125124 129
130139
140149
155 159
4.00
セ
セ
160
PRODUCTIVITY
Unfovorable
Normol/
Unfavorable
Normal
Fa vorable /
Normal
Favorable
Figure 3.4. Schematic diagram illustrating the relationship among reconstructed POSI values, their
interpretive categories, and crop yield production categories.
becomes 1.6 when rounded, then 16 when
multiplied by 10, and finally is reexpressed in
GIS units as 136 when 120 is added to 16.
Similarly, an original PDSI value of 3.28 becomes, 3.3, then 33, and finally reexpressed
as 87 in GIS units (Figure 3.4, rows b and c).
Calibrating Yield Categories to PDSI
Categories
The goal of this step was to assign the 11
moisture categories (Le., extreme drought to
extremely wet) to the five crop yield categories
(i.e., unfavorable to favorable) using the knowledge that increasing crop yields are reason
ably well correlated with increasing amounts of
stored soil moisture. This was done subjectively, but guided by the knowledge gained
from the partial regression study. The central
PDSI categories 5, 6, and 7, representing the
incipient drought, near normal, and incipient
wet, were combined and assigned to the "normal growing conditions," the central category
of the crop yield estimates. PDSI categories 1
and 2, extreme drought and severe drought,
were assigned to the lowest crop yield estimate
category, "unfavorable growing conditions."
Correspondingly, the highest PDSI categories,
10 and II, representing very wet and extremely wet, were assigned to the highest crop yield
category, "favorable growing conditions."
PDSI categories 3 and 4, moderate and slight
drought, were assigned to low intermediate
"normaltounfavorable growing conditions"
category of yields. Finally, PDSI categories 8
and 9, slightly and moderately wet, were assigned to the high intermediate category of
"normaltofavorable growing conditions"
110-
category of estimated yields (Figure 3A, rows a
and d).
In sum. historic crop yield data. data on
natural plant productivity, and reconstructed
Palmer Drought Severity Indices. were used in
a variety of patternrecognition studies that allowed quantifiable estimates of prehistoric crop
production associated with climatically
controlled soil moisture conditions to be made.
These studies provided the data to link the treering based PDSI reconstructions with the data
manipulation and display capabilities of geographic information systems. Collectively, the
data and techniques made possible the visual
discrimination of agricultural variability as well
as the quantitative estimates of crop production
and human carrying capacity.
-111 -
Page Blank in Original
4
BUILDING THE MODEL:
FINAL DATA ANALYSES
This chapter completes the description of
the data, assumptions, and analytic techniques
used to create this model of prehistoric agricultural productivity. It details the use of the
EPPL7 geographic information system in integrating the environmental data and the results
of the long term reconstructions and agricultural productivity studies. Further, it identifies
the numeric output required from the GIS
analysis for subsequent analysis. It also explains the methods used to calculate total agricultural productivity and to estimate the number and density of human beings that could be
sustained on the available yield. The chapter
concludes with a tabular synopsis of the
model-building process described in chapters
2,3, and 4.
FINAL GEOGRAPlDC INFORMAnON
SYSTEM ANALYSES
Introduction
This final stage of GIS analysis can be
divided into several steps (Figure 4.1). First, a
PDSI value (PDSI values ranging from Qセ
240, or 0 if no data exists for that cell) is
assigned to each cell in the 45,400 cell image
for each year of the analysis. Second, this PDSI
value is reassigned to a PDSI Nominal Oass
(value ranging from 1-11). Third, the PDSI
Oass value is linked to one of five newly
created data planes called favorability planes.
These planes depict the specific soil productivity unit values (SPUs) that have been estimated for each soil type under different growing
season conditions as discussed in chapter 3.
Once linked, the soil productivity unit value
(SPU value ranging from 5-94) is assigned to
a final agricultural productivity data plane for
a given year. Fourth, and last, a tabular file is
produced that lists the range of SPU values
assigned to a yearly data plane and provides
counts and percentages for those values.
AssiWin2 a PDSI Value
First, EPPL command INT ABLE creates a
temporary data plane that depicts the reconstructed annual PDSI value for each cell in
PDSlMAP.EPP (the data plane that assigned
each cell to its appropriate long-term PDSI reconstruction). It uses the appropriate long-term
PDSI reconstruction table (see Van West
1990:448-555 for an example) and locates the
reconstruction number (a row value, ranging
from 1-55 for a given year and a column
value ranging from 1-20, where the maximum
number of years in a single table is 20). The
conjunction of row and column is the reconstructed June PDSI value for a given soil class
in a given year. A different version of this table
is used by EPPL, since it can only process positive intergers between セRU
(recall Fig. 3.4).
Table 4.1. PDSI Class Correspondence Table.
PDSI
Nominal
Class
Range of PDSI
Values (in GIS
Units)
1
QセP
2
3
4
5
6
7
8
9
81-90
91-100
101-110
111-115
116-124
. 125-129
130-139
140-149
150-159
160-240
10
11
PDSISoil
Moisture Class
Extreme Drought
Severe Drought
Moderate Drought
Slight Drought
Incipient Drought
NearNonnal
Incipient Wet
Slightly Wet
Moderately Wet
Unusually Wet
Extremely Wet
PRELIMINARY ANALYSES
EPPL
COMMANDS:
RECl.ASS
FINAL ANALYSES
EVAWA"ffi
Ir=r
EVALUATE
JNfARLE
PDSlyyyy.TBL
r+
20 tbls, 20 yrs
eh
DEMWIND.EPP.5ELEV.EPP .
(1189)
(15)
•
soilw ndNep セi awcNep Mャ
PDSIMAP.EPP --+- (TEMP)
(I55)
(0·240)
--+-
FAVNORM.EPP,
(POSI 8.9)
•
PDSI.CCT..TEMP.EPP+ NORM.EPP .... PRODyyyy.EPP セ
(0.240=
(III)
(POSt 5.6,7)
(5·94)
L.. エセpeGfnumro
1·11)
(1·113)
(1·11)
PDSI Rccon.Table
PDSt090I.TBL
(AD 901 920)
セ
DEMWIND.EPP.5ELEV.EPP セ
168
13
(2044m)
(IGNACIO)
.
POSIMAP.EPP
1#32
SOILWlND.EPP'IIAWC.EPP
1#62
1# 10
(ROB)
(.99 10.26")
t
-+-
!
(TEMP] セ
57
(POSt = .6.35)
Class CorY. Table
PROOyyyy.CNT
(594)
(POSI 3,4)
Mセ
--
I
FAV.EPP
(POSI 10.11)
UNFAV.EPP
(PDSt 1,2)
f。カッイ 「ゥャ エy A u ・セ
Productivity Planes
Prod. Cm. Files
FAV.EPP
(PDSt 10.11)
FAVNORM.EPP
(PDSI 8,9)
PDSI.CCT..TEMP.EPP
. NORM.EPP
0·80 =
post
(PDSt 5.6.7)
Class t
Class I
セ
LNORMUNF.EPP
(POSt 3.4)
PRODyyyy.EPP
SPU 27
セ
PROOyyyy.CNT
SPU 27,# ,%, HA
UNFAV.EPP
(post 1.2)
Figure 4.1. Diagram illustrating procedural steps used to process the final GIS-integrated analysis. The top half is a generalized representation.
Notations in parentheses denote the range of class values associated with each data plane or file. The bottom half illustrates a hypothetical
example for one cell of the 45,400 cell image containing Witt Loam Soil (ROB) at 2044 m elevation for the year A.D. 901.
aウゥャュoセ
a PDSI Class
Second, command INTABLE reassigns the
PDSI values into 11 PDSI Classes (ranging
from Class 1, which is equivalent to extreme
drought, to Class 11, which is equivalent to
extremely wet, see Table 3.14). Compression
into the 11 classes is guided by a "class correspondence table" in which the reclassification
rules are stored (pDSI. CCf). A reduced set of
PDSI classes is created by grouping the values
in Table 4.1. This not only makes the reconstructed values more interpretable, but it also
simplifies the data manipulation. Thus, a
named but temporary data file, TEMP.yyyy.EPP is created for every year, in which
every cell is assigned a soil moisture (PDSI)
class. The "yyyy" designates the year A.D.
Assilming a Soil Productivity Unit Value
Before EPPL command EVALUATE
could be used to assign agricultural productivity values to each cell of a final output image
for each year, five new data planes had to be
created using data in Table 3.11. These five
data planes-FAV.EPP, FAVNORM.EPP,
NORM.EPP, NORMUNF.EPP, and UNFAV.EPP-are called "favorability" planes.
They represent the varying levels of agricultural productivity (expressed as soil productivity units, SPUs) that are associated with
individual soil types as they are affected by
different growing season conditions. Each
favorability plane was created with EPPL command !NTABLE, which linked one of the five
levels of productivity from Table 3.11 (and
reexpressed in an EPPL7-compatible version
of this table called PROD.TBL, Table 4.2) with
the soil appropriate soil type in SOn.WIND.EPP. Class values (the SPUs) on the favorability data planes range from 5 to 94. As a set,
they represent the range of potential yields
from all soils in the study area. Favorability
layer FAV.EPP assigns the most favorable
productivity values for a given soil (pROD.TBL, Table 4.2, column 2) to the cells in its
image. These values range from 14 to 94, and
depict the agricultural situation in the study
area should every soil be producing at its
optimum level. This, of course, never occurs,
but it creates a data plane that can be used for
evaluation purposes. Likewise, FAVNORM.-
EPP depicts a study area where every soil uses
values from the third column of data in
PROD.TBL, with values ranging from 11 to 87.
NORM.EPP uses the fourth column of data in
PROD.TBL with values from 8-80. NORMUNF.EPP uses the fifth column of data in
PROD.TBL with values from 7-75. Finally,
UNFAV.EPP uses the sixth and final column
of data in PROD.TBL with values from 5 to 69,
and depicts the agricultural situation in the
study area should every soil be at its minimum
productive level.
Once these five data planes were created,
the third analytic step could be performed. Via
a series of logical "if-then" operations,
EVALUATE was used to examine each cell in
TEMPyyyy.EPP, which has been assigned a
PDSI Class, and match it with the appropriate
favorability plane. While it is rare that more
than eight or fewer than three different PDSI
Class values occur over the surface of the study
area during anyone year, it is quite usual for at
least four or five different PDSI classes to exist
simultaneously for different soil classes at different elevations. If the PDSI Class value assigned to a cell in TEMP0934.EPP for year
A.D. 934, for example, is either a Nominal
Class 1 or 2 (extreme or severe drought), then
command EVALUATE uses the UNFAV.EPP
favorahility plane, looks up the SPU value for
that cell in that plane, and as-signs that SPU
value to the same cell location in the final output data plane, PR0D0934.EPP. Similarly, if
the PDSI value associated with a cell in TEMP.0934.EPP for AD. 934 is Class 3 or 4 (moderate drought or slight drought) the program
uses the NORMUNF.EPP agricultural productivity plane and assigns those values to the
output data plane. PDSI Class Values 5, 6, and
7 (incipient drought, near normal, and incipient wet) are assigned to the NORM.UNF productivity plane; PDSI Class Values 8 and 9
(slightly wet and moderately wet) are assigned
to the FAVNORM.EPP productivity plane; and
PDSI Class Values 10 and 11 (unusually wet
and extremely wet) are assigned to the FAV.EPP productivity plane. Thus, the final data set
for year AD. 934 will have integrated the
values contained in selected cells of the five
different productivity planes. This evaluation
and assignment process is carried out for every
year from AD. 901-1300 and consequently
creates 400 PRODyyyy.EPP files.
115 -
Table 4.2. Productivity Table (PROD.TBL) used in GIS Analysis.
Soil Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
FAV
0
32
33
42
42
34
33
42
42
0
94
29
36
36
46
0
24
21
25
21
0
0
21
35
0
27
0
39
0
21
33
20
16
29
24
49
47
49
43
30
34
58
58
0
0
0
52
29
52
29
52
29
48
32
47
FAVNORM
0
28
28
37
37
31
29
37
37
0
87
25
31
31
41
0
20
17
21
17
0
0
18
29
0
23
0
34
0
17
28
. 16
13
25
21
45
42
43
36
26
27
52
52
0
0
0
44
27
44
27
44
27
42
28
40
NORM
0
23
24
32
32
28
25
32
32
0
80
22
25
25
35
0
16
13
17
13
0
0
15
24
0
19
0
28
0
14
24
12
10
21
18
41
38
38
30
22
21
47
47
0
0
0
35
25
35
25
35
25
35
25
33
116-
NORMUNF
0
20
20
28
28
25
22
26
26
0
75
19
22
22
29
0
14
10
14
10
0
0
11
21
0
15
0
23
0
11
22
9
8
17
15
36
33
35
26
18
17
42
42
0
0
0
31
20
31
20
31
20
31
20
30
UNFAV
0
16
16
24
24
21
18
21
21
0
69
17
18
18
23
0
12
8
12
8
0
0
8
17
0
11
0
18
0
8
20
6
5
14
12
31
29
31
23
15
13
37
37
0
0
0
28
14
28
14
28
14
26
16
26
Table 4.2 (Continued).
Soil Code
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
FAV
41
41
28
33
48
0
51
43
51
43
51
40
40
42
42
28
29
61
67
49
67
50
51
39
54
46
0
42
21
0
0
0
0
0
83
83
0
32
32
26
32
26
77
33
33
33
33
36
36
36
21
15
20
17
FAVNORM
NORM
30
30
22
24
32
0
41
27
41
27
41
28
28
32
32
22
25
50
54
41
54
40
41
28
46
38
0
32
14
0
0
0
0
0
60
60
0
23
23
18
23
18
60
25
25
25
25
28
28
28
14
11
15
12
36
36
25
28
40
0
46
35
46
35
46
34
34
37
37
25
27
56
60
45
60
45
46
34
50
42
0
37
17
0
0
0
0
0
72
72
0
28
28
22
28
22
68
29
29
29
29
32
32
32
17
13
18
14
117-
NORMUNF
26
26
19
20
29
0
34
22
34
22
34
25
25
28
28
19
22
45
50
34
50
35
38
24
40
33
0
29
11
0
0
0
0
0
47
47
0
20
20
16
20
16
48
22
22
22
22
25
25
25
11
8
11
9
UNFAV
23
23
17
16
25
0
27
18
27
18
27
21
21
24
24
17
18
40
47
27
47
29
35
19
35
29
0
26
8
0
0
0
0
0
33
33
0
16
16
13
16
13
36
18
18
18
18
21
21
21
8
6
8
6
Table 4.2 (Concluded).
Soil Code
110
111
112
113
FAV
FAVNORM
21
0
15
14
NORM
18
0
13
11
15
0
11
8
Creating Count files
The final step is the creation of a tabular
output file that contains the values used later to
calculate total annual maize productivity and
sustainable human population size. EPPL
command COUNT produces a file that lists
classes, cell counts, individual percentages, cumulative percentages. areas, and a legend that
briefly describes the classes. Here the classes
listed are the soil productivity classes (SPUBs)
present in year yyyy with class values ranging
from 5 to 94. The cell counts are the number
of 200 x 200 m cells included in each class
and the individual percentage is the proportion
of those cells to the total number of cells
counted. The cumulative percentage is the
running total of percentages calculated class by
class. The area is the total amount of land measured in hectares that is included in that particular class. The legend identifies the productivity class. Table 4.3 is an example of a GIS
tabular output "count file" for year A.D. 902.
Production of the count file, particularly
the values associated with class and area, is
essential to the model-building process for it
embodies the numeric values needed to estimate annual maize production and ultimately
the number of people that could be supported
on that yield. With creation of the count files
for each of the 400 years in the analysis, the
GIS portion of the analysis is largely concluded, although GIS technology continues to
be used for displaying final products (the annual "maps"). Hereafter, these data are transferred to a mainframe statistical analysis program for processing, since these particular
capabilities do not reside within EPPL7.
Mapping the Agricultural Niche
Once the integrated file for a given year
has been created by the above analysis, it can
-
NORMUNF
12
0
9
7
UNFAV
9
0
7
5
be viewed with EPPL7 DISPLAY programs. In
DISPLAY, the locations of cells representing
the various levels of agricultural productivity
are visible. The many soil productivity classes
(5-94) may be regrouped into fewer classes
(1-5) that are easier to interpret. To retain the
original detail of differential productivity reconstructed for each cell in the 45,400 cell image, simplification of productivity classes
(SPUBs) was performed via selective colorcoding during the display of the data only. In
the color-classification scheme that was devised, five colors are arranged in a rainbow-like
palette representing five general categories of
agricultural productivity: SPUB classes 5-18
represents "low productivity" locations
(purple), SPUB classes 19-27 represents "lowto-moderate productivity" locations (blue),
SPUB classes 28-44 represents "moderate
productivity" locations (green), SPUB classes
45-52 represents "moderate-to-high productivity" locations (yellow), and SPUB classes
53-94 represents "high productivity" locations (red). These divisions are based on the
distribution of bean yield values for the 46
documented soils in Montezuma and Dolores
Counties. Using this distribution of values, any
value within one standard deviation of the
mean is considered normal or "moderate"
yield. A value falling between one and two
standard deviations above the mean is considered "moderate-to-high" yield. A value greater than two standard deviations above the mean
is "high" yield. Conversely, a value falling between one and two standard deviations below
the mean is considered "moderate-to-Iow"
yield, and a value lower than two standard deviations below the mean is considered "low"
yield (Table 4.4, Figure 4.2).
Using this classification and color scheme
it is easy to visually identify which locations
represent the most productive lands in any
given year. Both large and small patches of
productive land can be discerned and precisely
118-
Table 4.3. Example EPPL7 Tabular Output: COUNT File of Agricultural Productivity in A.D. 902.
Classa
8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
27
28
29
30
31
32
33
34
35
36
37
38
40
41
42
43
44
45
46
47
50
52
54
56
60
68
80
87
Count b
17
2,713
521
2,273
5,047
205
122
1,863
1,045
315
10
70
3,677
276
1,020
152
2,252
225
651
32
123
54
1,193
219
313
1,118
150
185
201
925
8,494
452
2
284
41
18
40
8
103
153
7
144
10
20
16
Percent C
0.05
7.38
1.42
6.18
13.73
0.56
0.33
5.07
2.84
0.86
0.03
0.19
10.00
0.75
2.77
0.41
6.13
0.61
1.77
0.09
0.33
0.15
3.25
0.60
0.85
3.04
0.41
0.50
0.55
2.52
23.11
1.23
0.01
0.77
0.11
0.05
0.11
0.02
0.28
0.42
0.02
0.39
0.03
0.05
0.04
Cumulative d
0.05
7.43
8.84
15.03
28.76
29.32
29.65
34.72
37.56
38.42
38.44
38.63
48.64
49.39
52.16
52.57
58.70
59.31
61.08
61.17
61.51
61.65
64.90
65.49
66.35
69.39
69.80
70.30
70.85
73.36
96.47
97.70
97.70
98.48
98.59
98.64
98.75
98.77
99.05
99.46
99.48
99.87
99.90
99.96
100.00
Area (ha)
68
10,852
20,840
9,092
20,188
820
488
7,452
4,180
1,260
40
280
14,708
1,104
4,080
608
9,008
900
2,604
128
492
216
4,772
876
1,252
4,472
600
740
804
3,700
33,976
1.808
8
1,136
164
72
160
32
412
612
28
567
40
80
64
Legend (Prod. Level)e
8D-89 SPUs; Low
100-109 SPUs; Low
110-119 SPUs; Low
120-129 SPUs; Low
130-139 SPUs; Low
140-149 SPUs; Low
150-159 SPUs; Low
160-169 SPUs; Low
170-179SPUs;Low
180-189 SPUs: Low
190-199 SPUs; Low/Moderate
200-209 SPUs; Low/Moderate
210-219 SPUs; Low/ Moderate
220-229SPUs;Low/Moderate
230-239 SPUs; Low/Moderate
240-249 SPUs; Low/Moderate
250-259 SPUs; Low/Moderate
270-279 SPUs; Low/Moderate
280-289 SPUs, Moderate
290-299 SPUs; Moderate
300-309 SPUs; Moderate
310-319 SPUs; Moderate
320-329 SPUs; Moderate
330-339 SPUs; Moderate
340-349SPUs; Moderate
350-359SPUs;Moderate
360-369 SPUs; Moderate
370-379 SPUs; Moderate
380-389 SPUs; Moderate
400-409 SPUs; Moderate
410-419 SPUs; Moderate
420-429 SPUs; Moderate
430-439 SPUs; Moderate
440-449SPUs;Moderate
450-459 SPUs; Moderate/High
460-469 SPUs; Moderate/High
470-479 SPUs; Moderate/High
500-509 SPUs; Moderate/High
520-529 SPUs; Moderate/High
540-549 SPUs; High
560-569 SPUs; High
600-609 SPUs; High
680-689 SPUs; High
800-809 SPUs; High
870-879 SPUs; High
Notes: The variables used for subsequent analysis with SAS are class (SPUB) and area (HA). The others are not
used. 36,759 cells are onsite; 8,641 cells are offsite.
セィ・
SPUB value (soil productivity unit, initially expressed in bean units).
hthe number of cells exhibiting this class value. Each cell is 4 ha in area.
Cpercent of onsite cells or analyzable cells. Offsite refers to cells that cannot be analyzed. A cell may be offsite
because soils have not yet been mapped for that cell or because available water capacity or natural productivity
information were not available at the time of analysis. Alternatively, a cell may be near the margin of the image in a
p'lace where incomplete overlap between data planes exists.
Cicumulative percent.
ean optional but useful feature for identifying the meanings assigned to each class. This legend identifies the
agricultural productivity level associated with each class.
119-
25.0
-I S.D.
X
+1 S.D.
+2 S.D.
Ibs / acre
194
278
362
446
530
SPUB
19
28
36
45
53
I
-
I
YIELD
Low
Moderate
to Low
Moderate
Moderate
to High
High
-
Figure 4.2. Normal distribution of modern bean yield (Ibs/ac) from 46 agricultural soils in Montezuma and
Dolores Counties. Colorado.
located; associated elevations and soil types for
any location can be identified. These images,
therefore, depict the agricultural niche, however it may be defined, as it existed in any given year. Thereafter, these depictions may be
captured photographically, printed out in color
by a color printer, or depicted as shades of
gray or symbols by a standard computer printer (Figure 4.3).
Comments on GIS Data Processing
Given that the above operations were to be
repeated 400 times, once for every year in the
analysis, it was necessary to automate the processing of the data Here other computer soft. ware was enlisted to facilitate the processing of
data. Toward this end, a program was written in
interpretive BASIC (GENEPPL.BAS) that generated a PC file in less than five minutes containing 400 sets of EPPL instructions, one for
each year in the analysis. After some minor
editing, the file (now renamed PRODEPPL.IN
to signify that modifications had occurred) was
redirected into EPPL and the 400 years of data
planes were processed in approximately 12
hours on an IBM/PS2 Model 50z. The batch
job produced 400 agricultural productivity
data planes, a large disk file (pRODEPPL.OUT) containing statistics for each data plane
(a "COUNT' file), EPPL system messages,
and possible error messages (of which there
were none). GENEPPL.BAS is Appendix F in
Van West (1990:561-562).
SUPPLEMENTAL ANALYSES WITH SAS
An auxiliary system to the GIS was needed
to further manipulate the numeric output contained in the EPPL7 count files. A system was
needed to link files, read selected data, create
new data from various combinations and transformations of extant data, and generate numeric and graphic output of the results. While
-120 -
Table 4.4. Modem, Nonirrigated Bean Yield for 46 Agricultural Soils in Montezuma and Dolores
Counties, Colorado.
Number
1
2
3*
4
5**
6**
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Soil
Code
8
9
21
28
34
35
36
42
43
47
48
49
50
51
52
53
54
55
56
57
58
60
62
Soil Map
Unit
A5C
A5D
C3CD
H1CD
M2DD
M2DD1
M5E
R1C
R1D
R3B
R3B1
R3C
R3C1
R3D1
R3D1
R4C
R4C1
R4D
R7C
R7D
R7D1
RHC
ROB
Yield Per Soil
T e Ib/ac
275
250
400
250
300
250
450
450
400
350
275
350
275
300
250
375
273
300
350
300
250
350
450
Number
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42*
43
44
45
46
Soil
Code
63
64
65
66
67
68
69
70
71
73
74
75
76
78
80
81
90
91
92
98
103
104
105
Soil Map
Unit
ROB1
ROC
ROC1
ROD
ROHC
ROHD
ROLC
ROLD
ROLD1
S5
S51
S53
S5D
S6C
S7
S8C
V1B
V1C
V1D
V4C
WOB
WOBB
woe
Yield Per Soil
T e Ib/ac
300
450
300
400
350
300
350
300
250
450
450
400
400
423
424
433
500
500
450
500
500
250
500
Notes: Mean yield for 46 soils is 362.02 84.26 pounds/acre. With the exception of M2DD and M2DD1, (the only
non-SCS Capability Class 3 and 4 soils, and soils for which yield from only the Pulpit components has been
reported), nonirrigated Bean Yield Values per soil type are determined by multiplying the percent of that component
by the reported yield value per component, and summing these values, as listed in Table 2.3. For example, soil map
unit 16, R4C, is comprised of Cahona, Sharps, and Wrtt Ioams in a relationship 40:35:25 ((.40 x 350 Ibs/ac) + (.35 x
350 Ibslac) + (.25 x 450 Ibslac)) • 140 + 122.5 + 112.5.375 pounds/acre.
*Soils not present in study area.
"Inadvertently, these two soils were not weighted and the yield from the Pulpit component of each, which is the
only agricultural component, was used to represent the entire unit instead of only 45% of that map unit. Had this
not been overlooked, the yield for M2DD would have been given as 135 instead of 300, and the yield for M2DD1
would have been given as 112.5 instead of 250. Thus, the correct estimates would have produced soil productivity
values that were only 45% of the values used. The overall mean and standard deviation for the 46 soils would have
been 355.4076 ± 96.0086 pounds/acre.
a number of different programs exist to do this
work, including some on powerful microcomputers with large amounts of memory, mainframe systems have some advantages for large
datasets requiring complex analyses. Ultimately, the widely used Statistical Analysis System (SAS Institute, Inc. 1985a, 1985b) was
selected to do the job.
First, the redirected output file that contained the results of the 400 individual analyses
of soil moisture conditions and agricultural
productivity was uploaded to the WSU mainframe. Next, the EPPL7-redirected output file
was modified with a mainframe editing pro-
gram, CMSXEDIT, to remove U1Ulecessary
verbiage, add appropriate instructions, and
format it for use in SAS data analysis. Thereafter, methods were developed to calculate a
single value that represented the total potential
production of maize for each year in the 400year series. Further, methods were devised to
estimate the number and density of people that
could be supported by that annual yield and to
identify levels of carrying capacity that would
reflect the maximum, sustainable population
size over extended periods of time (Appendix
C). Subsequently, I wrote SAS programs to
concatenate all 400 yearly data sets, convert
soil productivity units based on bean yield
-121 -
AGRICULTURAL
PRODUCTIVITY
A.D. 902
_LOW
•
Low toModerate
.Moderate
dmッ、・イ。エ・MエセMh
i gh
DHigh
DNa
Data
Figure 4.3. EPPL7 dotplot of prehistoric agricultural productivity in the study area as reconstructed for
JUly 1, A.D. 902.
(SPUBs) into soil productivity units based on
maize yield (SPUMs), and calculate the annual
supply of maize (Appendix D, Program 1). I
also wrote SAS programs to estimate sustainable population given three different levels of
demand, calculate a maximum, longtenn population level and calculate a lower, more realistic, longterm population level for the study
area (Appendix D, Program 2). Finally, I wrote
SAS programs to graphically portray various
aspects of the yield and population information.
The procedure described above was conducted first for the entire study area. Thereafter the data for the areas encompassed by the
Sand Canyon Survey Locality and the Mockingbird Mesa Survey Locality were processed
in the same manner. Finally, data for the areas
contained within the radii surrounding the
eight treering dated sites, here referred to as
"site catchments," were processed. The results
of these analyses are described in chapter 5.
Estimating Annual Maize Production
A single value representing the potential
maximum maize production is calculated for
each yearly data set in the time series. The
value is the total amount of maize that could
have been produced from all arable soils in the
study area (minus the potential yield that could
be derived from the low yield classes for any
year). It represents the maximum yield that
could have been, but certainly never was, produced in a given year. However, it does provide
an upper value for the potential maize supply
that could have been produced by human populations, and as such, is the first value required
to estimate sustainable human populations. It is
computed by summing the individual yield
values associated with each productivity class
that is considered to produce at least lowtomoderate yields. This is done in the following
way (see Appendix D, Program 1).
First, each class value that represents the
122-
Soil Productivity Unit in bean units (SPUB) is
converted to a Soil Productivity Unit in Maize
Units (SPUM), as detailed in chapter 3. This is
done by means of an equation SPUM =
«(SPUB*1O)+4)/0.473)*1.12. Here the SPUB,
which represents crop yield in terms of pounds
of beans per acre, is converted to a SPUM,
which represents crop yield in terms of kg of
maize per hectare.
Second, each SPUM (kg/ha) is multiplied
by the area (or the total number of hectares)
included in that class to derive the number of
kg of maize produced by those particular soil
units or cells (kg of maize per SPUM =
SPUM*HA).
Finally, those soil productivity units
(SPUBs) that are equal to or greater than 19,
representing cells where bean yields are greater
than or equal to 190 pounds per acre or where
maize yields are greater than or equal to 459
kg/ha, are summed. The SPUB 19 was chosen
as the lower limit of "lowtomoderate yields"
based on the following data. The mean and
standard deviation for the 46 of 113 soils in
the Montezuma and Dolores County area for
which the Soil Conservation Service has assembled mean bean yield data is 362.02 ±
84.26 pounds per acre (Table 4.4). If the
yields associated with the 46 soils were distributed normally (Figure 4.2), then a nonnal
or "moderate yield" would be any value falling within one standard deviation of the mean.
A value falling between +1 and +2 standard
deviations would be considered a "moderateto-high yield," and between -1 and -2 standard
deviations would be considered a "moderateto-low yield." Any value greater than +2 standard deviations would be considered a "high
yield," and any value below -2 standard deviations would be considered a "low yield."
Assuming that soils that were capable of producing greater maize yields would have been
selected over poor producers, cells containing
soils that produced yields less than 19 SPUBs
were not allowed to contribute their yield subtotals to the final annual estimate. The use of
this minimum yield criterion produces a more
conservative estimate of total annual potential
maize production (fOTPROD) than would be
produced than if all cells were used. Further, it
partially compensates for the possibility of
having overestimated prehistoric maize pro-
ductivity per unit of land made from analogs
with modem maize production.
Estimatinl: Annual Maximum Human
Populations
Once an estimate of the potential annual supply of maize was obtained, a method for determining the maximum annual demand for
that potential yield had to be devised. Human
demand for the supply necessarily includes
consideration of land use and cultivation practices, predictable post-harvest crop losses, seed
retention rates, storage levels, and human consumption rates. Reasonable assumptions and
numeric estimates needed to be established to
model these parameters. Thereafter it would be
possible to estimate the number of people who
could have been supported on that potential
food supply, both on an annual basis, and for
an extended period of time. Furthennore, the
assembled population estimates for the full 400
years would embody a data set that would permit evaluation of the question of whether the
study area was capable of supporting large
numbers of people in certain periods of time,
particularly in the late thirteenth century.
The following section details the assumptions used to estimate the yearly maximum
population that could be supported by the total
annual maize yield (TOTPROD). A specific
example of how this is calculated, is found in
Appendix C. Equations are developed to estimate a yearly population requiring an annual
maize harvest equivalent to one (pOP1YR), two
(POP2YR), and three (pOP3YR) year's food
supply. They are used in the SAS program
presented in Appendix D, Program 2.
Assumptions Used in Calculating Population
Values
• Only a portion of the total land area potentially available to produce crops is used.
Some lands remain uncleared and are used
for activities and resources not connected
with agriculture (e.g., places for hunting,
wild plant gathering, or collecting of nonfood resources such as construction or
fuel wood). Some cleared places on the
landscape are essentially nonagricultural,
such as residences and public buildings.
123-
In addition, some agricultural lands are deliberately taken out of use or fallowed for
one or more seasons to allow moisture to
recharge soil profiles. to curtail soil nutrient depletion, and sometimes to reduce
loss of topsoil from wind erosion
(Beaglehole 1937:36). In this study it is
assumed that only 50% of the lands potentially usable for raising crops were
cultivated in any year. The remaining half
generates the potential gross yield. The
estimate of 50% was also used by Kohler
et al. (1986:528) in a recent attempt to
model agricultural productivity in the
vicinity of the Dolores Archaeological
Project, and it reflects ethnographic observations of land use by the Hopi (Forde
1931:370), the Tewa of San Juan Pueblo
(Ford 1968:157), and maize-growing
Mexican peasants (Sanders 1976:141143).
Hopi farmland was devoted to maize production.
• Not all the maize harvested will be available for consumption. Some will be lost
during transportation from the fields to
residences and storage facilities, and even
after being deposited in storage containers
or storage rooms, some additional amount
will be lost due to pests and spoilage. In
this study, the annual adjusted gross
maize yield is reduced by 10% to account
for these predictable losses. The resulting
value represents the potential net maize
yield available to prehistoric farmers. The
estimated loss rate of 10% was used by
Williams (1989:726) in her ethnohistoric
study of fanning populations in the Basin
of Mexico and by Hassan (1981:45) in
estimating agricultural supply and consumption in Pre--Dynastic Egypt
• Of what is harvested. stored. and potentially available for consumption, a certain
percent must be reserved for seed. This
seed will be used for planting the following season's crop and may also be used
for subsequent seasons if the following
season is particularly poor. In this study,
10% of the potential net yield must be reserved for planting subsequent crops. The
resulting balance represents the actual edible or adjusted net yield. All ethnographers refer to the seed that has been selected and stored with particular care for
use in later plantings, but rarely is a percent or weight of grain given. Williams
(1989:721) removed 10 kg/ha for her
study. Wetterstrom (1976:195) removed
10 kg/ha for each year in storage. Hassan
(1981:45) removed 25% in his Egyptian
example. While the value used in my
study may be high (although not as high
as Hassan's value), it may compensate for
underestimating or not taking into consideration losses in other parts of the
growing-harvesting-and-storing cycle.
• While it is not known precisely in what
proportions the major prehistoric cultigens of maize, beans, and squash were
grown, archaeologists have generally assumed that maize was the major crop of
Pueblo Southwest. Maize is more amenable to selection and local adaptation than
the other crops, and by virtue of its composition and its regular stOrage on a solid
cob, it preserves better than the other
two. Further, maize plays a major role in
the cultural systems of contemporary and
historically-known Pueblo people. Archaeologists have reasoned from various
evidence that these traditions have great
time depth and were likely established in
prehistoric Pueblo societies. In this study
it is assumed that approximately 80% of
the remaining lands available for agriculture were devoted to maize production
and the remainder used for other crops.
This value represents the adjusted gross
yield or the harvested maize production.
The estimate of 80% was used recently by
Williams (1989:726) in a study of a sixteenth century Aztec peasant community,
Santa Maria Asuncion, in the Basin of
Mexico, and has also been used by Sanders (1976:109, 145; 1979:376) in the
Basin of Mexico. This value is closely
ma1ched by the ethnographic reports of
Hack who said that modem Hopi at that
time devoted 72% ofall farmland to
growing maize (Hack 1942:19), and by
Judd (1981:39) who quoted Hough (1930)
and claimed that over three-quarters of
• While minimum daily caloric requirements
vary from person to person and depend on
factors such as age, gender, metabolism,
and level of activity, it is assumed here
that an average of 160 kg of maize per
year was required to sustain the average
person. This value is equal to .4384 kg
per day or 1534 calories per day where
3500 calories are produced from a kg of
maize (Cook and Borah 1979:164). It is
-124 -
accurate in predicting the actual number
of people that produced a crop of maize
than the calculation for the single year.
This may be true not because people
stored subsistence foods for this length of
time, but rather because they required
more yield per person to satisfy the various demands for that supply.
assumed that additional calories (and minerals and proteins) will be obtained from
foods other than maize. but it is not necessary here to specify a value representing
total daily caloric intake. The estimate of
160 kg of maize per year was used by
Sanders (1976:145) as an aggregate statistic for maize consumption and is based
on values from ethnographic data Williams used a like figure (1989:725). The
values used to estimate the rate of consumption vary in archaeological models.
The number of calories obtained from a
kg of maize varies (e.g., Ford [1968: 157]
uses 355 kcal/l00 g. Gross [1985:103]
uses 356 kca1/100 g, Hastorf [1980:99]
uses 360 kcal/100 g). The number of calories assumed to be necessary to sustain
an individual also varies (e.g., Schlanger
[1985:44] assumes 2,000 calories a day,
and Kohler et al. [1986: 528] assume
2,400 calories a day), and the amount of
maize consumed varies (e.g., Schlanger
[1985:44] assumes 50% of calories are
obtained from maize. and Kohler et al.
[1986:528] assume 60% of caloric requirements are obtained from maize).
• Population density is computed by dividing the number of people by land area. It
is desimble to calculate population density (rather than, or perhaps in addition
to, population size) because it represents
a standard unit of measure that can be
compared from place to place without
regard to size or total area. In this study,
the entire study area is comprised of
36,759 cells for which complete soils
data are available. Each cell is 200 by 200
m (4 hal in area. Thus, the total, analyzable area of the study reton is 147,036
hectares or 1470.36 km . The Sand
Canyon Survey Locality contains 652
cells and includes 26.08 km 2. The Mockingbird Mesa Survey Locality contains
449 cells and incorporates 17.96 km 2.
Six of the eight tree-ring dated sites
(5MT8371,5MT8839,5MT2433,
5MT3834, 5MT1566, and 5M1765) include 197 cells in their 1.6-krn radius,
representing 7.88 krn 2. 5MT6970 and
5MT2149 are located near the margins of
the swdy area and consequently a portion
of the 1.6-km radius fell outside the
boundary. Thus, 5MT6970 took in only
108 cells or4.32 km 2 and 5MT2149 included 168 cells or 6.72 krn 2. It is important to note that these area measurements
are used for every year in the analysis and
do not consider the distribution of individuals or settlements within that area.
Thus, the density values produced do not
reflect the degree of clustering or aggregation of settlement within the areas identified. Furthermore, the boundaries established for purposes of analysis are arbitrary and do not reflect the acwalbounding of human settlement or activity.
However, the estimated values do represent an average population density that
could be supported by an area so bounded
and are useful numbers for envisioning a
upper limit for supportable populations.
• In this study, the possibility of storing a
portion of edible harvest for one and even
two years beyond the needs of the current
year is considered (Bums 1983). Here an
estimate of population size is made on
the full consumption of the adjusted net
yield in a single year (POPIYR), on consumption of that yield if a portion is
stoced and predictably reduced and used for
a second year (POP2YR), and on consumption of that yield if a portion is
stored., reduced, and used over two additional years (POP3YR). The goal of attempting to annually grow two years of
maize is often cited by ethnographers as
the Pueblo ideal (Hough 1915:62; Forde
1931:393; Parsons 1936:954-955; Whiting 1939:13, 15; Titiev 1944:181; Ford
1968:162; Bradfield 1971:21) although it
is occasionally said that it is rarely
achieved. It seems clear, however, that
more maize was grown than was needed
to support the immediate producers for a
year. Maize was grown also to fulfill
reciprocity obligations among kin and
was used regularly in ceremonies, weddings (Kavena 1980:14), and trade and exchange transactions (Bradfield 1971:21).
Consequently, the calculations for two or
even three years of stomge may be more
It must be pointed out that the net effect of
the second-to-Iast point above, which concerns
itself with determining population size at three
-
125-
contrasting storage/demand levels, is that population estimates will be reduced rather than
inflated. In effect, I have assumed that while
populations grow and store more than one
year's demand for maize, they do this every
year without consideration of what "excess"
might remain in storage at the beginning of the
planting season. For example, if a population
in Year 1 managed to raise the equivalent of
two year's supply of maize and deposited it in
storage after the harvest of Year 1, it would still
have the better part of that yield in storage at
the beginning of the planting season of Year 2.
Given the ever-present risks associated with
crop production in the American Southwest,
this same population would attempt to plant
and harvest an equally large crop in Year 2.
Should the goals for the Year 2 maize crop be
realized at harvest time, the old grain in storage, which was marked for the second year, will
then be consumed (eaten, traded, used as barter
tender, or otherwise used) and the fresh grain
used to replace the old grain.
Should, however, a less than totally successful harvest be realized, some old grain will
be kept (i.e., the needed balance preserved)
and the amount in reserve from Year 1 and
Year 2 harvests will not exceed the storage level
modeled for that population. So, while I assume a population may carry over stored
reserves, in effect I do not build the carry over
into the model; these populations attempt to
grow as much as possible, but not more than
can be stored or maintained in their storage
facilities. Some might argue that this will result
in unrealistically low carrying capacity estimates for the storage scenarios. However,
because one conclusion of this study will be to
argue that short-term climate change is probably insufficient, by itself, to cause the total
depopulation of the Four Comers area, some
conservatism in population estimation seems
defensible. If, in fact, more people could have
been supported by their maize production than
I estimate, then that strengthens the argument
that the thirteenth-century depopulation is not
solely attributable to climate change.
It is important to emphasize that this model
attempts to estimate the maximum values for
productive acreage, maize yield, and population size possible at each level of storage or
demand for every year in the 4OO-year series.
Estimatio2 Lon2-Teun Sustainable Human
PQDuiatiQus
One Qf the goals of this study is to ascertain whether climatic conditions could have
negatively affected the supply Qf agricultural
resources to such an extent that they alone
could account fQr the major depopulation of
the Mesa Verde region in the late thirteenth
century. This entailed reconstructing relevant
climatic conditions, modeling agricultural
productivity, and estimating the size of human
populations that could be supported by the
available yield. Thanks to the availability of
high-resolution tree-ring data, this has been
done on an arumal basis. However, these yearly
values must be evaluated for patterning
through time to incorporate the effects of
annual and multiyear fluctuations in climate
and crop productivity on sustained population
levels.
In this study, the concepts of maximum,
critical, and optimal carrying capacity have
been adapted from Hassan (1981: 166-168) to
estimate sustainable pQpulation levels over
extended periods of time. These concepts are
depicted schematically in Figure 4.4. Maximum carrying capacity is a population size
estimate that is equivalent to the long-tenn
mean value of the estimated yearly maximum
population for the total period of 400 years.
This value is the upper limit on population size
and would represent a regional population that
would frequently experience yield shortfalls
when annual production fell noticeably below
the mean. Critical carrying capacity is a population size estimate that is equivalent to the
minimum annual population value in the 400year period. It is a value below the maximum
carrying capacity (the mean) and represents
the largest population that would exist
throughout the entire period without experiencing significant crop shortages. Thus, it is
detennined by the years when yield is least
abundant Optimum carrying capacity is a
population size estimate that falls within a
range of values. Its upper limit is roughly
equivalent to the critical carrying capacity
value (the minimum value). Its lower limit is
not specified, but it is the size of a population
operating well below the limit of fluctuating
productivity. It is almost always within the
range of values equal to 20--60% and is
126 -
Maximum
Car rying - 1--+---\-f---l-l---\-/--+---1'4-J-++++-++-+--I+--f'r--+1+-t-iY-+I--+Mean Val ue (100%)
Capacity
90%
60%
70 %
Cr itic al Carrying
Capacity
Zane of
Oplimal
Carrying
Capacity
Minimum
Va Iue (60%)
50%
hMZ[Nセイ⦅BtWエ
40%
30%
20%
10%
Time
Figure 4.4. Schematic diagram illustrating the relationship among three concepts of carrying capacity
(after Hassan 1981:167).
generally between 40-60% of the maximum
carrying capacity value (Hassan 1981: 175).
To obtain estimates of carrying capacity,
the mean value (the maximum carrying
capacity), the minimum value (the critical
carrying capacity), and other descriptive data
were calculated for each 400-year set of maize
demand levels (pOPIYR, POP2YR, POP3YR)
with SAS Programs (Appendix D, Program 2,
Proc Means). This was performed on the data
forthe study area as whole, for the two survey
Table 4.5. Summary of Model-Building Steps: Operationalizing the Model.
I. Paleoenvironmental Data
A. AcqUire appropriate tree-ring data (SWOLD7 for years A.D. 901-1970).
B. Reconstruct PDSI Values for the time period represented by the tree-ring data for all combinations
of weather stations (Bluff, Cortez, Ignacio, Mesa Verde, and Ft. Lewis) and soil moisture groups (11
classes based on available water-holding capacity in the upper and lower soil layers). Product is 55
sets of reconstructions.
1. Calculate PDSI for Instrumented Time Period.
2. Calibrate PDSI and Tree-Ring Data for Instrumented Time Period.
a. Initial Calibration.
b. Verification.
c. Final Calibration.
3. Retrodict PDSI Series, A.D. 901-1970.
C. Construct PDSI tables in GIS-compatible fonn (subset using A.D. 901-1300 organized as a set of
tables).
II. Elevational Data
A. AcqUire and process OEMs.
B. Produce a single, mosaiced, elevation data plane (DEMWIND.EPP).
C. Create new data plane depicting the zones of elevation to be modeled by the weather stations (in
this case, five zones modeled by five stations) (5ELEV.EPP).
-
127-
Table 4.5 (Concluded).
III. Soas Data
A. Acquire descriptive and locational data on soil map units.
B. Produce a single soils data plane depicting all soil map units in the study area (SOILWIND.EPP)
C. Create a new data plane depicting all soil map units reclassified into a reduced set of soil classes
based on potential water-holding capacity of soil layers (in this case, 11 soil classes subsuming 98
different soil map units) (11AWC.EPP).
IV. Soil Productivity Data
A. Acquire historic bean and maize production data for study area.
B. Acquire data on natural plant productivity for all soils in study area.
C. Perform regression-based analyses to determine the relationship between varying levels of
agricultural productivity and PDSI.
1. Determine the nature and strength of the relationship between PDSI and historic yield of bean
and maize.
2. Estimate bean and maize yields for all soils from known relationship between domestic plant
productivity (bean and maize yields) and natural plant productiVity per soil type on a sample of
soils.
D. Calibrate PDSI values with crop yield values.
E. Construct an agricultural productivity table in GIS-eompatible form (a table providing varying values
for agricultural yield per soil type given varying PDSI values and climatic conditions).
V. Ethnographic Data
A. Gather appropriate data useful for modeling prehistoric Pueblo agricultural practices, consumption
practices, and per capita demand for yield.
VI. GIS-Integrated Final Analysis
A. Create new data plane to facilitate the analysis.
1. Create an image that assigns each long-term PSDI reconstruction to its appropriate location
(PDSIMAP.EPP).
B. Run program that generates annual images of agricultural productivity.
1. Associate each cell on PDSIMAP with its POSt value for each year in the A.D. 901-1300 series,
via PDSI tables (product is the annual PDSI reconstruction, TEMPyyyy.EPP).
2. Associate each cell's PDSI value with a PDSI nominal class (1-11).
3. Associate each cell's PDSI nominal class value with a yield value, via agricultural productivity
table.
4. Depict the value for each cell's agricultural yield value on a new output image. Image may be
displayed and photographed (product is the annual productivity reconstruction,
PRODyyyy.EPP).
5. Produce tabular output files (COUNT files) that summarize the total number of cells and their
associated area in hectares for each productivity class (prodUct is annual productivity
reconstruction count file PRODyyyy.CNT).
6. Store tabular data for subsequent transfer to a statistical analysis program for further processing
and analysis.
VII. SAS Data Final Analysis
A. Calculate total and effective annual agricultural maize yields for study area.
B. Calculate the maximum annual population (number of people and density) that could be supported
on that effective yield for a population with varying requirements for maize storage for the study
area. Product is an annual estimate of the maximum population size, for three levels of storage.
C. Calculate the annual maximum, annual minimum, and the long-term mean population (number of
people and density) that could be sustained over the period of 400 years (A.D. 901-1300) for the
study area. Products are long-term estimates of population size, expressed as the maximum,
critical, and optimal carrying capacity, for three levels of storage.
128-
localities and for the eight site catchments.
THE COMPLETE MODEL
The products of the modeling procedure
described in chapters 2. 3. and 4 are the annual
images (or "maps") of agricultural productivity reconstructed for the A.D. 901-1300
period. the numeric estimates of annual maize
productivity. and the numeric estimates of annual and long-term population for the same
time period. An example of one of these annual images prepared as a black-and-white
EPPL7 DOTPLOT was presented as Figure 4.3.
The image is better viewed and interpreted as a
colored computer image seen directly on a
monitor. captured photographically as a color
slide. or observed as a series of colored images
prepared as an animated time series.
-
. The numeric estimates of annual total maize
productivity and annual and long-term population size for the study area are presented in
abbreviated form in Appendix E. The estimates
made for the two subareas. the Sand Canyon
Survey Locality and the Mockingbird Mesa
Survey Locality are presented in Appendices F
and G. The annual estimates for the eight treering dated site catchments are presented in
Appendices H through O.
The data and analytic tools used to reconstruct the prehistoric dryland farming habitat
of the Mesa Verde Region Anasazi may be
summarized in outline form. Table 4.5 shows
the steps taken to develop the final analytic
products. This table may also be seen as a
synopsis specifying the methods and sequential
products that allow the conceptual model
identified in chapter 1 to be operationalized.
129-
Page Blank in Original
5
RESULTS
This chapter presents and explores data
created by the model of prehistoric agricultural
productivity and sustainable human population
in southwestern Colorado for the tenth through
the thirteenth centuries A.D. The primary data
are estimated values of annual maize yield and
annual potential population for three different
levels of demand/storage. Appendix E supplies
the annual estimates of maize yield and annual
maximum population for the study area as a
whole. Appendices F and G present these estimates for two contrasting survey localities
within the study area, the Sand Canyon Locality and Mockingbird Mesa. Appendices H-O
supply these values for the catchments surrounding eight tree-ring dated sites within the
study area.
The purpose of this chapter is to examine
these data and suggest their usefulness for addressing problems of interest to archaeologists
working with horticultural societies. My exploratory studies, however, are neither prescriptive nor exhaustive; they are meant only to
demonstrate some of the uses for these data at
different scales of analysis.
The data generated for the study area as a
whole embody an assessment of agricultural
productivity and supportable population for a
1470.36 km 2 area. My analysis of these values
may be seen as representing one of the studies
that can be accomplished with such information on a broad geographic or regional scale.
Here I examine certain characteristics of the
dataset, determine human carrying capacity
levels that would have been sustainable over the
full 4OO-year period, and fmally arrive at some
tentative conclusions about the effect of known
climatic variation on agricultural production
and human population in the study area.
On a smaller geographic scale, the data
produced for the two block survey localities
within the study area may be seen as one of the
results that can be generated for a subregional
level of analysis. At this level it is possible to
compare archaeological estimates of population derived from architectural and artifactual
remains to the ecologically derived estimates
produced by the model. In this way, I compare
"observed" values to "expected" values and
draw conclusions as to local carrying capacity
limits and whether productive limits have been
exceeded. This exercise was particularly instructive because the environments of the two
localities contrast in some ways that potentially
affect agricultural productivity and the number
and density of people who may be supported
by agriculture. In addition to comparing the
two localities to each other, I compare the localities to the larger study area, to gain an appreciation for the degree to which each of
these spatial samples represents the whole.
The data produced for the eight tree-ring
dated sites may be seen as analysis on yet a
smaller geographic scale, the site catchment
level. Studies like those performed at the regional and subregional levels can, of course,
also be performed at this level. In addition, it is
possible to characterize the productivity and
population characteristics of the catchment
during the time that the site was occupied. I use
this information to evaluate whether the site
occupation period was better than, worse than,
or essentially the same as the long-term condition for that place. I also use the information
to predict the most favorable times for settlement of that catchment during the 400 years
examined. Further, it is possible to compare the
catchment of sites that are known to be contemporary and assess their relative productivity. In this way, these data may be used to
predict the more favorable places for productive agriculture during a given time. I do
this in a limited way for three partially contemporary "central place" sites in my sample
that are thought be cOimected to the Chacoan
occupation of the study area. I also do this with
two small village sites that are nearly identical
in their dating. I classify 5MT8371 (DCA site),
5MT8839 (Norton House), 5MT2433 (Aulston
Pueblo), and 5MT3834 (Mustoe Ruin) as small
village sites on the basis of their limited structural remains and non-public architecture. All
have been described as having two or fewer pit
structures and 10 or fewer surface rooms. Alternatively, I classify 5MT6970 (Wallace Ruin),
5MT1566 (Lowry Ruin), 5MT2149 (Escalante
Ruin), and 5MT765 (Sand Canyon Pueblo) as
"central place" sites on the basis of having 25
or more rooms and public architecture such as
large or special kivas and other unusual structures such as tri-walled, D-shaped buildings.
Finally, it is possible to compare classes of sites
to assess the role that differential agricultural
potential and sustainable population size may
play in site location or duration of occupation.
Here I compare potential productivity and
population values for the class of four "small
village" sites against the class of four "central
place" sites.
THE STUDY AREA
Annual values for total maize yield
(TOTPROD) and the maximum number and
maximum density of people who could be
supported by that yield for a population requiring either one year (pOPIYR), two years
(pOP2YR), or three years (pOP3YR) of maize
in storage at the end of harvest have been estimated (Appendix E). Total maize productivity
for the 400-year series is depicted graphically
in Figure 5.1 and trends in mean population
levels for the three storage/demand levels are
depicted graphically in Figure 5.2. Simultaneous inspection of these two graphs reveals that
population size varies (by defmition) directly
with maize production. In some years maximum potential yield and maximum potential
population are quite high and in some years
they are much lower; often this shift happens
quickly, one year following another. It is clear,
however, that the size of a real population
could not vary from year-to-year in this manner. Rather, longer-term trends in productivity
control the real size of a population that can be
sustained in a given place. Consequently, longterm trends were derived from the 400-year
series of values.
The 400-year mean value, the 400-year
minimum value, and a range of values equivalent to 20% through 60% of the 400-year
mean value were taken for POPI YR, POP2YR,
and POP3YR to represent the maximum carrying capacity, the critical carrying capacity, and
the optimal carrying capacity, respectively
(Table 5.l). These data better approximate a
sustainable population than do productivity
estimates from any individual year.
Hassan (1981) has suggested from a worldwide sample that the actual size of populations
that are able to maintain their numbers over an
extended period of time, usually exists within a
range of values between 20% to 60% (and
more commonly 40% to 60%) of the longterm mean estimate that can be calculated for
their sustaining area. This range is termed the
optimal carrying capacity. The maximum
number of people who could be supported
during the least productive years in the 400year series. including those infamous years of
the "Great Drought of 1276-1299" (Douglass 1929, see Burns 1976 for a discussion of
this concept), is the critical carrying capacity
and is represented by the minimum population
value of the 400-year sequence. Using the
POP2YR set of values, the low value of 21
persons/krn 2 occurs 15 times with values nearly
as low - 22 persons/krn2 and 23 persons/krn2occurring an additional 12 times. Occasionally
two (but never three) population lows occur
two years in a row (A.D. 906-907; 980-981).
Other lows occur at close intervals (e.g., AD.
901 and 906-907; A.D. 972 and 980-981;
A.D. 1062 and 1067; AD. 1146, 1150, 1156,
and 1161; and AD. 1254 and 1258). By
contrast, there are a few. fairly long intervals
when the three lowest population values do not
occur (A.D. 1020-1061, 1091-1130. 11871216. 1228-1253) and a number of periods of
15 or more years in length when the minimum
population supportable was. in fact. reasonably
high (Table 5.2). Generally, however, these
very low values occur once every 10 to 25
years, a time likely to be recalled by adults in
the population. The long-term minimum value.
the critical carrying capacity, then, represents
the long-term maximum number of people
whose demands for maize could be met without suffering, barring other factors that
132 -
1.IE+oa
1.0E+oa
9.0E+07
a.OE+07
..-.
w
w
セ
9
7.0E+07
セ
6.0E+07
5.0E+07
".OE+07
3 . Of +0 7
IIImn,nllllllfttlnIlRtllnlnHlnml.IIII11Il1I1I11MIIIIJlIIIIIIIAIIIIIIIIIIRlllfllllnlDllllflnMI1"nllA111RllllftIlHlmnnnUIIII"""mIIllPlU'KlIIIIIIIIIQlIUlllUlllllmUlIlnllltm"nnm"",,,,"IIUIIIOIIIIIIIIIIIIIII""""'IIII1I11""'"III""""I11"""'""III''"II'''I11'''''IIIII'III'"'''''''''""
900
1000
1100
"'
1200
YEARS (A.D.)
Figure 5.1. Maize production in the study area. A.D. 901-1300. The vertical needles represent annual yields. The undulating line represents
weighted mean yields.
.
1300
ISO
140
130
120
110
><
t::
100
CI)
セ
90
C
Z
....0
W
セ
セ
..J
::J
セ
0
セ
80
10
60
'0
40
30
20
10
0
,
,
900
1000
1100
1200
YEARS (A.D.)
Figure 5.2. Supportable population density in the study area, A.D. 901-1300. The vertical needles represent estimated maximum annual
population. The undulating lines represent the weighted means for POP1YR (top), POP2YR (middle), and POP3YR (bottom).
1300
Table 5.1. Carrying Capacity Estimates at Three Levels'ofStorage/Demand for A.D. 901-1300 for the
1470.36 km 2 Study Area.
.
Maximum Carrying
Storage Level
Capacnr
POP1YR
88± 19
(persons per km 2)
POP1YR
131,473 ± 28,222
(number of persons)
POP2YR
35±7
(persons per km 2)
POP2YR
53,246 ± 11 ,429
(number of persons)
POP3YR
19±4
(persons per km2)
POP3YR
...rumber of persons)
28,752 ± 6172
Mean Value ± Standard Deviation.
Critical Carrying
Capacityb
Optimal Carrying
Capacityc
52 (59%)
18 - 35 - 53
77,439
26,295 - 52,589 - 78,884
21 (60%)
7 - 14 - 21
31,363
10,649 - 21,298 - 31,948
11 (58%)
4-8-11
16,936
5,750 - 11,501 - 17,251
bMinimum Value.
ォBoセ
- 40%· 60% of Mean Value.
affect successful crop production. The maximum carrying capacity, the long-term mean, is
theoretically the highest long-term population
size that could be sustained most, but not all, of
the time. It is the least realistic of the three
levels of carrying capacity, but it does provide
an upper threshold for estimating a realistic
aggregate population for the study area as a
whole.
The population density values provided in
Table 5.1 are high; higher than some researchers might think possible, particularly towards
the end of the thirteenth century when the
Mesa Verde Area was permanently abandoned
by Anasazi populations. The estimates for
POPI YR are probably the least accurate; we
suspect that prehistoric Puebloans did attempt
to store at least two years of maize at the end of
harvest in the event that production was insufficient the following season. Consequently, the
population estimates for POP2YR are likely to
be better estimators of sustainable population,
although they suffer from the conservative bias
reported in the previous chapter. POP2YR
represents the population size that could be
supported for two years from the maize produced in the present year, or the population
that could be supported if the production in
the present year were followed by a year with
no production whatsoever. Since this is rarely
or never the case, POP2YR (and, all the more,
POP3YR estimates) have a conservative bias).
-
Thus, by placing the maximum inferential
weight on the use of the middle storage/demand level (pOP2YR) and the more likely
estimates of sustainable population size (the
critical and optional carrying capacity values),
I calculate that some 21,298-31,363 persons,
representing a density of some 14-21 persons/km2 for the 1,470.36 km 2 study area
could have been supported in any given year
within the A.D. 901-1300 time period. The
lower value represents the 40% of mean value
from the optimal carrying capacity range, and
the upper represents the critical carrying capacity value. This upper value is very similar to
the value recently suggested by Rohn (1989)
for aggregate population in the Montezuma
Valley, an area that includes the study area but
extends fully from the eastern slopes of the
Abajo Mountains in southeast Utah to the
valley bottom below the Mesa Verde in Southwest Colorado. Rohn asserts that 30,000 is a
conservative estimate of the number of people
who lived in the Montezuma Valley in the
thirteenth century (Rohn 1989:166), an estimate derived from his knowledge of archaeological data, particularly survey data in the
Montezuma and Dolores County area. Further,
this estimate of 30,000 does not include the
numbers of people he estimates for nearby
Mesa Verde or the Mancos Valley. Therefore,
it would seem that the POP2YR estimates are in
line with one of the few general estimates of
population that has been proposed for the
135-
Table 5.2. Periods of Greatest Occupational Attractiveness in the Study Area.
Rank by Rank by Minimum
Years a Pop. Densityb
4
6
7
7
8
8
1
4
2
2
5
1
3
3
6
5
Beginning
Date A.D.
931
955
1020
1091
1091
1100
1228
1259
Ending
Date A.D.
952
971
1034
1125
1120
1120
1250
1275
Number
of Years
22
17
15
35
30
21
23
17
Minimum Population
Density Value (km2)
26
26
26
27
29
30
29
27
aBased primarily on number of years and secondarily on the minimum population density that can be sustained.
bBased primarily on minimum population density and secondarily on the number of years that this condition can be
sustained.
heartland of the Mesa Verde region.
In sum, I believe that two conclusions may
be drawn from this preliminary analysis of the
data generated for the study area as a whole.
First, climatic variation alone as it affected
agricultural production cannot be cited as the
simple and sufficient cause for the major depopulation of the Mesa Verde Region of the
Anasazi in the late thirteenth century A.D.
Barring crop failures due to plant disease or
pests or loss of significant quantities of farmland due to soil erosion or nutrient depletion,
there was always enough productive land to
support thousands of people in the study area
(minimum population size of 31,363 person or
21 persons/km2), even during the difficult
times of the middle ll00s, which coincided
with the collapse of the Chacoan system, and
the socalled Great Drought of the late 1200s.
Second, the distribution of productive land
changes from year to year, although there are
locations in the study area that are more consistently productive and others that are more
consistently unproductive than other locations
over long periods of time. Inspection of the
annual images generated by the GIS makes this
point abundantly clear since it is possible to see
where the most and least productive lands are
located in any year. Examination of the full
400 years reveals those places that are reasonably constant, and therefore predictable, for
considerable intervals of time. Therefore, it
may be concluded that access to productive
resources (Le., the arable land or the products
produced from that land) was an issue of extreme importance (Adler 1990; Kohler 1989).
In this situation, either people must be allowed
to go to where they can grow the food, or the
food must be grown and distributed to where
the people are located. If mobility and access
to the productive lands are restricted, or if
redistribution systems are not in place to support dispersed populations or uneven production, then the potential aggregate population figures will overestimate the actual
population that could have been supported.
These issues fall as much in the sociopolitical
realm of human cultural systems as in the environmental realm, and they force us to seriously consider the complex interactions of
climatic variation. environment, and human
behavioral systems in issues of culture change
and stability.
LOCALITIES WITHIN THE STUDY AREA:
THE BLOCK SURVEY AREAS
Sand Canyon Survey Locality
The Sand Canyon survey locality, as digitized for this study, is a 26.08 k:m 2 area surrounding two large thirteenth century pueblos,
Sand Canyon Pueblo (5MTI65) and Goodman
Point Ruin (5MT604), located near the heads
of Sand and Goodman Canyons on the highland mesatop setting known as McElmo Dome
(Figure 2.4). The locality straddles portions of
two adjacent 7.5minute USGS quadrangle
maps, Woods Canyon to the west and Arriola to
the east, and is located within the southcentral
portion of the study area. Elevations in the
survey locality range from 1,8902,164 m
(6,2QO7,100 ft). Precipitation and temperature
136 -
patterns affecting local soils were therefore
modeled with PDSI reconstructions from the
Cortez, Ignacio, and Mesa Verde series.
Survey and testing within the locality was
undertaken by the Crow Canyon Archaeological Center between 1985 and 1987 (Adler
1988; Van West 1986; Van West et al. 1987);
429 sites were recorded. Of these, 420 single
and multiple component prehistoric sites were
attributed to the Anasazi culture. Of these 420,
18 habitation components with a mean room
count of six rooms per site date to the AD.
YSセYXP
period, 73 habitation components
with a mean room count of six rooms per site
date to the AD. Y X セ
1060 period, 89 habitation components with a mean room count of
eight rooms per site date to the AD. 1PVセ
1150 period, and 95 habitation components
with a mean room count of 20 rooms per site
date to the AD. Q UセQSP
period (Adler
1988:27, and revisions, personal communication, March 1989). Room counts and room
measurements on these habitation sites, in conjunction with assumptions about rooms per
household units, mean number of people per
room, and site longevity have permitted Adler
to make momentary population estimates
based on inferred habitation life spans of 20
years (Le., 20-year momentary population
estimates) and habitation life spans of 50 years
(Le., 50-year momentary population estimates)
for each archaeologically defmed time period
(Adler 1988:27). These archaeological data
represent a set of "observed" population
values that can be compared to the "expected" carrying capacity values derived by
the environmental model.
Annual values for total maize yield (TOTPROD) and population at three levels of
storage/demand (pOP1YR, POP2YR, and
POP3YR) for the Sand Canyon survey locality
are found in Appendix F. Total maize productivity for each year in the 4()()..year series
for the Sand Canyon survey locality is graphed
in Figure 5.3 and trends in mean population
density for the three levels of storage/demand
are displayed in Figure 5.4.
Mockingbird Mesa Survey Locality
The Mockingbird Mesa survey locality, as
digitized and measured in this study, is a 17.96
km 2 area cotenninous with the mesatop surface
of Mockingbird Mesa (Figure 2.4). Mockingbird Mesa is a north-northeast to south-southwest oriented mesa situated north and west of
Sandstone and Yellowjacket Canyons and
south and east of Hovenweep and Negro Canyons. It is depicted on the Negro Canyon 7.5minute USGS quadrangle map and is positioned in the west-central portion of the study
area. Elevations in the survey locality range
from 1,829-1,975 m HVLPセTX
ft). Consequently, reconstructions drawn from the
Cortez and Ignacio series were used to model
soil moisture conditions in local soils.
The Bureau of Land Management conducted archaeological survey in this locality
between 1981 and 1984; 684 single and multiple component sites was recorded. Of these,
700 components representing some 550 sites
were assigned to the Anasazi culture. Fettennan
and Honeycutt (1987:120), who summarize the
work, report that 47 habitations representing
66 households can be attributed to the early
Pueblo IT period (A.D. 925-10(0), 37 habitations representing 72 households to the late
Pueblo II periods (A.D. QッセPIL
47 habitations representing 118 households to the
early Pueblo III (AD.ll00--12oo), and 40
habitations representing 155 households to the
late Pueblo III (AD. 1200-1275). Household
counts, in conjunction with assumptions concerning the numbers of people per household,
human life expectancy, and site life expectancy, allowed them to make population estimates based on inferred habitation life spans of
12 years and 100 years (or in the case of late
Pueblo III, of 75 years) for each of the defined
archaeological periods (Fetterman and Honeycutt 1987:119, 120).
Sarah Schlanger (1985:194-195) used a
sample of these same archaeological data from
Mockingbird Mesa and ascribed sites and their
components to a different set of archaeological
periods based on ceramic assemblages resulting from the Dolores Archaeological Project.
She assigned two sites representing 20 people
to A.D. PXYセR
period, 30 sites representing
period, five
285 people to the A.D. URPQセXY
sites representing 50 people to the A.D. 10251100 period, 14 sites representing 155 people
to the A.D. 1100-1175 period, and 49 sites
137 -
2600000
2500000
2400000
2300000
2200000
2100000
.....
w
00
セ
......
セ
2000000
1900000
1800000
1700000
1600000
1500000
1400000
1300000
1200000
.,.
900
1000
1100
1200
1300
YEARS (A.D.)
Figure 5.3. Maize production in the Sand Canyon Survey Locality, A.D. 901-1300. The vertical needles represent annual yields. The undulating
line represents weighted mean yields.
200
190
180
170
UiO
15<>
140
b
130
(I)
120
ffi
110
Q
100
S
.....
w
セ
\0
セ
90
セ
80
70
60
5<>
40
30
20
10
0
.
900
,
1000
1100
1200
1300
YEARS (A.D.)
Figure 5.4. Supportable population density in the Sand Canyon Survey Locality, A.D. 901-1300. The vertical needles represent estimated
maximum annual population. The undulating lines represent the weighted means for POP1YR (top). POP2YR (middle), and POP3YR
(bottom).
representing 530 people to the A.D. 11751250 period. Household counts, in conjunction
with assumptions about the number of the
people per household, the frequency of structure reoccupation, and the life expectancy of a
living room, permitted her to make population
estimates based on an inferred habitation life
span of 20 years (i.e., 20-year momentary
population estimates) for the each of the archaeologically defined time periods (Schlanger
1985:188, 194, 195, 198, 199).
To explore the comparison of archaeological data with the data created by the agricultural productivity model, I have chosen to use
Schlanger's population estimates instead of
Fetterman and Honeycutt's estimates because
Schlanger's data are more comparable to those
generated by Adler for Sand Canyon. The archaeological time periods are more alike, both
use 20-year momentary population estimates,
and both provide density values for population
in their respective areas. Schlanger, Fetterman
and Honeycutt, and Adler use somewhat different values and assumptions for their estimates, however, so the results are not strictly
comparable. The interested reader should refer
to their publications for details of method.
Armual values for total maize yield (TOTPROD) and population at three levels of storage/demand (pOP1YR, POP2YR, and POP3YR)
for the Mockingbird Mesa survey locality are
reported in Appendix F. Total maize productivity for each year in the 400-year sequence
for the Mockingbird Mesa survey locality is
portrayed in Figure 5.5 and trends in mean
population density for the three levels of population are portrayed in Figure 5.6.
The Localities and Their Relationship to the
Study Area
Table 5.3 compares the 400-year mean
maize yield and population values for the two
localities with those of the larger study area.
While both the number and density of people
able to be supported by the 400-year mean
yield are provided, the more useful figure is
the population density value as it is standardized by area. The popUlation density values
predicted by the productivity model for the
-
Mockingbird Mesa survey locality are nearly
identical with those of the study area as a
whole, whereas the density values predicted by
the model for the Sand Canyon locality are
markedly higher than for the study area or for
Mockingbird Mesa. This would seem to indicate that the Mockingbird Mesa Locality is
generally representative of conditions in the
study area as a whole, although it does not
include the lowest and highest elevations or the
least and most productive soils in the study
area. It also demonstrates that there are places
within the study area that are better than others
insofar as productive land is concerned. The
Sand Canyon locality is a more productive and
predictable location in which to farm than is
the Mockingbird Mesa locality (The coefficient of variation-the ratio of the standard deviation to the mean multiplied by 100-indicates that there is more overall variation associated with the annual values for Mockingbird
Mesa than for the study area, and much less
variation associated with the Sand Canyon locality. The coefficient is an inverse measure of
overall predictability). Other places within the
study area, particularly lands in the north and
north-east portions of the study area, appear to
be to be equally if not more productive.
Table 5.4 provides the maximum, critical,
and optimal carrying capacity values for the
three areas expressed as persons/km2 for a
population requiring two years maize at the
end of harvest (pOP2YR). The Mockingbird
Mesa survey locality at any time within the
A.D. 901 - 1300 period could have supported
a population density of at least 18 persons/km2,
a value equal to 50% of the long tenn mean
value of 36 ± 9 persons/km 2. This minimum
value or critical carrying capacity occurs 16
times over the 4OO-year time span, in a similar
pattern to that of the study area. By contrast,
the data indicate that the Sand Canyon survey
locality at any time during the same 4OO-year
period could always have supported a density
of at least 37 persons/km2, a value equal to
67% of the long-tenn mean value of 55 ± 10
persons/km 2. Interestingly, this 4OD-year
minimum value occurs 29 times, or almost
twice as often as in the other locality. In absolute tenns, this long-tenn minimum value, or
critical carrying capacity, is twice as high as for
the Mockingbird Mesa survey locality and
140-
1300000
1200000
II 00000
1000000
....
セ
セ
セ
......
900000
800000
700000
600000
500000
400000
..
900
1000
1100
1200
1300
YEARS (A.D.)
Figure 5.5. Maize production in the Mockingbird Mesa Survey Locality, A.D. 901-1300. The vertical needles represent annual yields. The
undulating line represents weighted mean yields.
150
140
130
120
110
>-
t:
100
ffi
0
90
fI)
Z
10
j
70
9
セ
N
:J
ll.
0
ll.
60
50
40
30
20
10
0
,
,
900
1000
1100
1200
1300
YEARS (A.D.)
Figure 5.6. Supportable population density in the Mockingbird Mesa Survey Locality, A.D. 901-1300. The vertical needles represent estimated
maximum annual population. The undulating lines represent the weighted means for POP1YR (top), POP2YR (middle), and POP3YR
(bottom).
Table 5.3. Comparison of Population Values for the Study Area, Sand Canyon Survey Locality, and
Mockingbird Mesa Survey Locality A.D. 901-1300.
Study Area
POPNUM
1,470.36
Sand Canyon Locality
POPKM
POPNUM
26.08
26.08
Mockingbird Mesa Locality
POPKM
POPNUM
17.96
17.96
Area (km2)
POPKM
1,470.36
TOTPRODa
(kg)
C.V.b
64,925,217 ±
13,936,845
21.5
64,925,217 ±
13,936,845
21.5
1,779,087 ±
337,524
19.0
1,779,087 ±
337,524
19.0
POP1YRc
Mean, S.D.
Range
88±19
52....141
131,473 ± 28,222
77,439-207,382
137±26
93-198
3,602 ± 683 89±23
2,437-5,174 46-142
POP2YRd
Mean, S.D.
Range
35±7
21-57
53,246 ± 11,429
31,363-83,989
55± 10
37-80
1,458±276
987-2095
801,336 ±
204,994
25.6
36±9
18-57
801,336 ±
204,994
25.6
1,622 ± 415
834-2,553
656± 168
337-1,034
POP3YRe
Mean, S.D.
19±4
28,752±6,172
29±5
787±149
19±5
354±90
Range
11-30
16,936-45,354
20-43
532-1,131
10-31
182-558
aTOTPROO is the total mean productivity of maize rounded to the nearest whole number; all population values
(POPIYR, POP2YR, POP3YR) are truncated integers.
bCoefficient of Variation, the ratio of the standard deviation to the mean multiplied by 100, rounded to the nearest
tenth. This yields a percentage that is rounded to the nearest integer.
cPOP1YR .. «(TOTPROO*.324)/160)/area.
dpOP2YR .. «(TOTPROO *.324)*.81 )1320)/area.
epOP3YR .. «(TOTPROO*.324)*.6561 )/480)/area.
almost twice as high as for the study area (at
21persons/km2). It is a good indicator of the
productivity and predictability of the Sand
Canyon survey locality for fanning.
Table 5.4. Carrying Capacity Values (POPKM) for the Study Area, Sand Canyon Survey Locality, and
Mockingbird Mesa Survey Locality, based on POP2YRa Estimates for A.D. 901-1300.
Study Area
Sand Canyon Locality Mockingbird Mesa
1,470.36
26.08
17.96
Area (J<m2)
64,925,217
1,779,087
801,336
TOTPRODa (kg)
± 13,936,845
± 337,524
± 204,994
19.0
25.6
C.V.b
21.5
80
57
Maximum value
57
Maximum CC:c Mean value
35 ± 7
55± 10
36±9
37 (67%)
18 (50%)
Critical CC:
Minimum value 21 (60%)
Optimal CC: 60% of Mean
21
33
22
40% of Mean
14
22
14
20% of Mean
7
11
7
Note: All population values are truncated integers.
aTOTPROO is the total mean productivity of maize, rounded to the nearest whole number.
bCoefficient of variation (ratio of the standard deviation to the mean multiplied by 100) rounded to the nearest
tenth.
cMaximum, mean ± standard deviation, and minimum population density values (POPKM) for POP2YR are
found in Appendix E (entire study area) F (Sand Canyon locality) and G (Mockingbird Mesa locality).
-
143-
Comparing Archaeological Estimates with
Productivity Estimates of Population
Sand Canyon Survey Locality
Figure 5.7 plots the 20year and 50year
momentary population densities (based on inferred habitation life spans of 20 and 50 years,
respectively) for each of the archaeological
periods established by Adler (1988:27 and
revised values, personal communication, March
1989) for the Sand Canyon survey locality.
These are overlaid on a plot of the longterm
mean value (the maximum carrying capacity),
the maximum value, and the minimum value
(the critical carrying capacity) calculated for
the 400year period from the annual population density values associated with POP2YR.
Also depicted are those intervals when the minimum value (the critical carrying capacity)
"lifts" for a period of time from its low of 37
persons/km2 to some greater value. These elevated periods were determined and plotted in
the following way. Using population density
data contained in Appendix F for POP2YR, the
minimum population value for a consecutive
series of 10 years was recorded for each 10year period beginning at A.D. 901. For example, the minimum value for A.D. 901910
is 37 personjkm2. Similarly, the minimum value for each rurming set of lOyear periods
(A.D. 902911, 903912, 904913, and so
on) is recorded. When the minimum level
changed (e.g., from A.D. 908917 the minimum value is 46 rather than 37), that new
minimum value was plotted at the fmal year of
the 1Qyear period (e.g., 46 was plotted at A.D.
917) and was allowed to stay at that level until
a new 10year minimum value was recorded.
Here I chose to plot the minimum value at the
end of the lOyear period, rather than at the
beginning or middle, to simulate an effective
lOyear memory of past productivity that may
be drawn on to make decisions about current
or future agricultural behavior. In this short
term view of timea frame of reference more
appropriate to a human lifetime and human
recollectionthis "temporary" lifting of the
longterm minimum value may be perceived as
stability, depending on the length of time that
this condition persists. The actual or perceived
reality of a heightened and enduring productive landscape may have significant behavioral
ramifications, some of which may be detected
archaeologically. Evidence of population in.crease, agriCUltural intensification, or environmental degradation might, for example, be detected during or following an extended period
of heightened critical carrying capacity.
Examination of Adler's momentary population estimates (based on inferred habitation
life spans of 20 years) indicates that local, aggregate population in the Sand Canyon survey
locality never approached the critical carrying
capacity of 37 persons/km2, despite steady
population growth throughout the 400 years.
However, if momentary population estimates
are based on inferred habitation life spans of
50 years, then local population requirements
for maize (under the POP2YR assumptions)
would only have been met adequately during
the first three rriodS. The minimum value of
37 persons/km ,however. was exceeded at
times within the final A.D. 115a-13oo period,
an interval when archaeological estimates indicate a population density of 40 persons/km2
was present. This indicates that there would
have been years of shortage in meeting the
demands of a population requiring two years
of maize at the end of harvest if habitation sites
were occupied as long as 50 years. These years
of shortage in the fmal period are: A.D. 1150,
1156, 1176, 1185, 1191, 1215, 1227, 1254,
and 1295. Overall, however, it would appear
that the Sand Canyon survey locality was a
highly productive and predictable farming location and that potentially large local populations would have been living well below the
limits imposed by the productive environment.
Mockingbird Mesa Survey Locality
Figure 5.8 plots the 2Qyear mean momentary population density estimates (based on inferred habitation life spans of 20 years) for
each of the archaeological periods established
by Schlanger (1985:199) for the Mockingbird
Mesa survey locality. As in Figure 5.7, these
are overlaid on a plot of the longterm mean
value (the maximum carrying capacity), the
maximum value, and the minimum value (the
critical carrying capacity) calculated for the
4OQyear period using POP2YR values. Again,
periods of elevated carrying capacity are depicted.
144-
SAND CANYON SURVEY LOCALITY
POPULATION DENSITY
Mo'imum so
Value
so
70
70
セ
N
e
.¥
......
III
c
0
セ
60
60
Mean
Value (55)
-Moximum
50
50
t1>
a..
セ
>.
III
.f>o.
VI
Carrying
Capacity
c
セ
40
Minimum
Value (37)-
7
セ
t
7
A7 1
7
7R7
j;.
7
j.
X77-;=] 14 0
-
t1>
0
30
30
C
0
0
20
20
10
'a
Critical
Carrying
Capacity
Optimal
Carrying
Capacity
Zone (11-37)
セ
a.
0
a..
at
0
0
I'-"f "
aoセ o
i'
(C cic cr C <, cCf CC, C c,c C
950
I 900-950 I
950-9S0
セ
I()(X)
I
C coc
」セ
C <icc
1050
980-'060
セ
CC,t tfCC«<,cC( C <,t
1100
I
1060-1150
」セ
C tVCf CC,t t ( t c,c
1150
I
」セ
t < c,
セ
ttl' C
',' ' ( ' "
セGエサ\
QRセ
1200
1050-1300
'to
'('
iセo
I
Years (AD)
Figure 5.7. Comparison of estimates of carrying capacity for the Sand Canyon Survey Locality generated by the model with estimates derived
from archaeological survey in the locality, A.D. 901-1300.
N
e
セュオ ゥN。m
MOCKINGBIRD
MESA
SURVEY LOCALITY POPULATION DENSITY
60
Value(S7)
.>I-
"-
セ
III
セ
III
a..
セ
....
,J:..
0\
セッ
§
>-
III
40
40
Meon
Value (36)
30
30
C
III
Critical
-Carrying
20 Capacity
0
Mセ
c
C
:J
C-
Minimum 20
Value 08l10
o
Maximum
Corrying
Copocily
10
Optimol
Carrying
Capacity
Zone (7-18)
2 ".,tont/ .. J
a..
o
AD 900
.
Yセ
920-980
1000
1050
1150
'200
0
'300
1250
QRセPMGSP
YXPMQPRセ
Years (AD.)
Figure 5.8. Comparison of estimates of carrying capacity for the Mockingbird Mesa Survey Locality generated by the model with estimates
derived from archaeological survey in the locality, A.D. 901-1300.
A comparison of Schlanger's data for
Mockingbird Mesa with the longtenn estimates generated by the model reveals a different scenario than that reconstructed for the
Sand Canyon survey locality. While Schlanger
provides momentary estimates based only on
inferred life spans of 20 years for her five periods. these values reveal that two periods of major population growth are followed by two periods of major population decline. In both
cases. the productive capabilities of the Mockingbird Mesa survey locality were significantly
exceeded by a population estimated from archaeological data In the growth periods. A.D.
98G1025 and A.D. II 751250. archaeological population estimates are 28.5 pers/km2 and
53 persons/km 2• whereas the 4OOyear minimum value or critical carrying capacity was 18
persons/km2 and the lOyear minimum value is
always less than 35 persons/km 2. The earlier
period is followed by a period of low population (A.D. 10251100. archaeologically estimated to be 5 persons/km 2) and the later period is followed by a time of unknown population density (A.D. 125G1300). but a time
in which it is known that at some point the
Mesa is pennanently abandoned. These data
suggest a repeated history of population overshoot and collapse when human demand exceeded agricultural supply.
In concluding this section. it may be said
that localitybased data for which archaeological survey and population estimates are available are potentially very infonnative. They can
suggest patterns that would allow a variety of
specific hypotheses to be tested in a given location. In the case of the two localities examined here. the productivity data in conjunction
with the archaeological data were able to suggest two quite different histories of population
in the 400 years examined. TIlis is particularly
interesting since these two localities are relatively close to each other (about 15 km or 9.3
mi apart from center to center) and both possess excellent soils with relatively high available
water capacities. However. the Sand Canyon
survey locality is 124385 m (2()().620 ft)
higher overall in elevation than the Mockingbird Mesa survey locality. and it is not as
circumscribed by canyon topography. This
preliminary study of localitybased data is useful in that it clearly demonstrates that there are
places within the study area that are more pro-
ductive and predictable than others, and even
. among the better locations. there are some that
are consistently superior.
SITE CATCHMENTS WITffiN THE STUDY
AREA: TREERING DATED SITES
Small Village Sites
5MT8371, DCA Site
Description. This unnamed site that I have
chosen to call the "DCA site"(after the institution that excavated it. the Division of Conservation Archaeology at the San Juan County
Museum near Fannington. New Mexico) is
reported by Dykeman (1986). It is inferred to
be a single component. early Pueblo II habitation site consisting of a single household.
Excavations revealed a single, semisubterranean rectangular pit structure with multiple floor
features that appeared to be associated with a
number of extramural features including storage pits. postholes. shallow pits, an outdoor
activity area, and a midden (Dykeman 1986:i).
Location. The site is located in the SEI/4.
SEI/4. of Section 1. T37N. RI9W. N.M.P.M.
on the Negro Canyon 7.5minute quadrangle
map. with UTM coordinates 4.152.000 m N.
and 678.800 m E (Zone 12). It is situated on
the west edge of Mockingbird Mesa (although
not in the Mockingbird Mesa survey locality).
on a ridge near the head of Negro Canyon. at
an elevation of 1,914 m (6,280 ft). The PDSI
reconstructions used to represent soil moisture
in this 7.88 km 2 catchment were drawn from
the Cortez and Ignacio series.
Dating. The assignment of calendrical
dates to 5MT8371 by its excavator derived
from three sources: treering dates. ceramic
studies, and radiocarbon estimates. The finnest
dates are associated with the treering dated
samples (Table 5.5). All were charred wood
from the roof of Pit Structure 1. Five samples
were dated; two were cutting dates dating to
A.D. 935. The latest date on the site was 940 vv
(Dykeman 1986:68). A date of A. D. 952 for
the ceramic assemblage was calculated by
Dykeman on the basis of ceramic analysis and
the application of two modal ceramic date
fonnulae (Dykeman 1986:85). Four charcoal
147 -
Table 5.5. Tree-Ring Dates from Site 5MT8371, DCA site (TRL# A-666).
Structure and
Feature
Pit Structure 1-- Tr.2, Feat.2
Pit Structure 1--Tr.2, Feat.2
Pit Structure 1--Level 1
Pit Structure 1--Tr.2. Feat.2
Pit Structure 1--Roof fall
TRL Cat.
Number
CVM-114
CVM-115
CVM-121
CVM-113
CVM-119
Field
Number
32
33
126
32
59
Tree
Species
Juniper
Pinon
Pinon
Juniper
Juniper
Outside Date
Symbol
932 w
935 r
935r
936 w
940 w
Note: The symbols (Robinson and Harrill 1974:4-5) used with the outside date are as follows:
B Bark is present.
G Beetle galleries are present on the surface of the specimen.
L A characteristic surface patination and smoothness, which develops on beams stripped of bark, is present.
c
The outermost ring is continuous around the full circumference of the specimen. This symbol is used only if a
full section is present.
r
Less than a full section is present, but the outermost ring is continuous around the available circumference.
v
A subjective judgment that, although there is no direct evidence of the true outside on the specimen. the date
is within a very few years of being a cutting date.
vv There is no way of estimating how far the last ring is from the true outside.
+ One or more rings may be missing near the end of the ring series whose presence or absence cannot be
determined because the specimen does not extend far enough to provide an adequate check.
++ A ring count is necessary due to the fact that beyond a certain point the specimen could not be dated.
The symbols B. G, L, c, and r indicate cutting dates in order of decreasing confidence, unless a + or ++ is also
present.
The symbols L, G, and B may be used in any combination with other symbols except v and w. The rand c
symbols are mutually exclusive, but may be used with L, G, B, +, and ++. The v and vv are also mutually
exclusive and may be used with the + and ++. The + and ++ are mutually exclusive but may be used in
combination with all the other symbols.
samples from four different features were submitted for 14C analysis. Only one fell within
the range of the tree-ring dates and dates associated with ceramics. That uncalibrated date,
from Feature 1, was reported as 1000 ± 40 B.P.
(A.D. 1950, with a half-life of 5568 years) and
assigned an A.D. 950 date by the excavators
(Dykeman 1986:69). Given the beginning
cutting date of A.D. 935 and the final dates
from the structure, the structure was interpreted
as likely to have been constructed sometime
with the A.D. 935-940 period (Dykeman
1986:68). Since 5MT8371 is believed to have
been occupied for 15-20 years, it is assigned
an A.D. 935-950 date for this study.
Site Catchment and Agricultural Productivity. Table 5.6 summarizes population density data for 5MT8371 using POP2YR levels of
demand. Over the entire 4OQ-year period, it
appears that values generated for the maximum
carrying capacity for 5MT8371 are nearly
identical to those for the nearby Mockingbird
Mesa survey locality and to the study area as a
whole (35 ± 10 persons/km 2 versus 36 ± 9
persons/km2 and 35 ± 7 persons/km2, respectively). However, the critical carrying capacity,
-
the minimum value, is considerably less for
5MT8371 than either Mockingbird Mesa or
the study area (9 persons/km2 versus 18 persons/km2 and 21 persons/km2, respectively). In
every respect, 5MT8371 is markedly less productive than the figures estimated for the Sand
Canyon survey locality, which has a maximum
carrying capacity of 55 ± 10 persons/km2 and
a critical carrying capacity of 37 persons/km2
at all times.
During the 16-year period when 5MT8371
was occupied, the carrying capacity values for
5MT8371 were higher than they were for the
400-year period, indicating that the small village site was established and inhabited during a
time that was slightly better than the long-term
average conditions for that catchment. Again,
maximum and optimal carrying capacity values for 5MT8371 are very similar to those for
the Mockingbird Mesa survey locality and the
study area (38 ± 5 persons/km2 versus 38 ± 5
persons/km2 and 37 ± 4 persons/km2, respectively). The critical carrying capacity value for
5MT8371 during its period of occupation is
higher than its for either Mockingbird Mesa or
the study area as a whole (28 persons/km2
148-
Table 5.6. Comparison of Population Density Values (POPKM) for 5MT8371, DCA Site, for A.D.
901-1300 and A.D. 935-950.
5MT8371
Catchment
7.88
350,996
± 104,956
29.9
C.V.b
c
57
Maximum value
35 ± 10
Maximum CC: Mean value c
Minimum value c 9 (25%)
Critical CC:
OptimalCC: 60% of Mean 21
14
40% of Mean
7
20% of Mean
Mockingbird
Mesa
17.96
801,336
± 204,994
25.6
57
36±9
18 (50%)
22
14
7
A.D. 901-1300 (400 years)
Area (kfn2)
TOTPRODa(kg)
Sand Canyon
Locality
26.08
1,779,087
±337,524
19.0
80
55± 10
37 (67%)
33
22
11
Study Area
1,470.36
64,925,217
± 13,936,845
21.5
57
35±7
21 (60%)
21
14
7
A.D. 935-950 (16 years)
TOTPROD(kg)
372,413
851,962
1,855,196
67,853,516
± 57,456
±127,904
± 207,518
± 7,473,915
15.0
11.2
11.0
15.4
C.V.
48 (AD 938,
46 (AD 938) 65 (AD 942, 948) 41 (AD 938, 942,
Maximum value
944)
944, 946, 948)
Maximum CC: Mean value
38 ± 5
38±5
58±6
37±4
41 (71%)
24 (63%)
26 (70%)
Critical CC:
Minimum value 28 (74%)
(AD 937)
(AD 937,939) (AD 937, 944) (AD 937)
35
23
Optimal CC: 60% of Mean 23
22
15
23
40% of Mean
15
15
8
12
7
20% of Mean 8
Note: All population values are truncated integers.
aTOTPROD is the total mean productivity of maize, rounded to the nearest whole number ± standard
deviation
beoefficient of Variation, the ratio of the standard deviation to the mean, multiplied by 100, rounded to the
nearest tenth.
cMaximum, mean ± standard deviation, and minimum population density values (POPKM) for POP2YR
area found in Appendix H (OCA), G (Mockingbird Mesa), F (Sand Canyon locality), and E (entire study
area).
versus 24 persons/km2 and 26 persons/km2 , respectively), indicating its comparative attractiveness. Nevertheless, the catclunent is still less
attractive than other places in the region at exactly the same time, as indicated by the values
for the Sand Canyon survey locality, which are
much higher (maximum carrying capacity of
58 ± 6 persons/km 2 and a critical carrying capacity of 41 persons/km2 for the entire 400year period).
Site Catclunent and Prime Periods for Agricultural Prodyction. Given that the 16-year
occupation between A.D. 935-950 was more
productive than that over the long-term for the
site catclunent of 5MT8371, the data available
for 5MT8371 were searched for other equally
-
long periods that could support a minimum of
28 persons/km2 (Appendix H, Figure 5.9).
Table 5.7 summarizes these data for the catchment of 5MT8371. In this study, I assume that
the length of time a place can sustain a high
carrying capacity is a measure of temporal
predictability-the longer the time, the more
predictable a place. I also assume that the size
of the minimum population density a place can
sustain is a measure of spatial productivitythe greater the minimum population, the more
productive a place. By these two measures,
5MT8371, in comparison to the other time
periods of equal of greater occupational attractiveness, is characterized by moderate predictability and productivity, whereas the A.D.
1098-1120 period is the most productive
149-
Table 5.7. Periods of Equal or Greater Occupational Attractiveness in the Catchment of 5MT8371 (A.D.
935-950).
Rank by Rank by Minimum Beginning
Date A.D.
Years a Pop. Densityb
4
4 (5MT8371)
931
8
8
1069
1
2
1091
2
1
1098
5
5
1167
6
6
1187
3
3
1228
7
7
1259
Ending
Date A.D.
952
1084
1125
1120
1185
1204
1250
1275
Number of
Years
21
16
35
23
19
18
23
17
Minimum Population
Density Value (km2)
28
28
28
35
28
28
28
28
aBased primarily on number of years and secondarily on the minimum population density that can be sustained.
bBased primarily on minimum population density and secondarily on the number of years that this condition can be
sustained.
time, and the A.D. 10911125 period is the
most predictable for the catchment. If agricultural productivity and predictability were key
variables for settlement decisions in the catchment of 5Mf8371, then sites dating to this
time period should be present.
5Mf8839 Norton House
Description. This site is reported by Fuller
(1987) and by Kuckelman (1988). It is considered to be a single-component Pueblo II
habitation. The site was located in the right-ofway of one of the Dolores River Project canals,
the South Canal. Excavation revealed substantial architecture including "one pit structure
with an adjoining pit room, a separate pitroom
southwest of the pit structure and 11 exterior
features" (Kuckelman 1988:373). The main
rectangular pit structure was found to have had
two occupational surfaces and to have burned
on abandonment. No other pit structures or
surface structures were defmed during survey
and excavation, but it was suggested that modem plowing has likely destroyed surface rooms
(Kuckelman 1988:403).
Site Location. The site was located in the
SWI/4 of the NWI/4 of Section 3, T38N,
R18W, N.M.P.M., depicted on the Pleasant
View 7.5-minute quad map, at UTM 4161580
m N and 692100 m E (Zone 12) on deep
mesatop soil at an elevation of 2,093 m (6,868
ft). The site was in an area that is currently under cultivation, about three miles west of the
modem farming community of Pleasant View.
The PDSI reconstructions used to represent soil
-
moisture in the 7.88 km 2 catchment were
drawn from the Mesa Verde series.
Dating. Three sources of dates were explored to date this site: tree-rings, archaeomagnetism, and ceramics. All 19 tree-ring dated
samples came from the burned roof fall of Pit
Structure 1; eight of these were cutting dates
(Table 5.8). Of these eight, the earliest cutting
Table 5.8. Tree-Ring Dates from Site 5MT8839,
Norton House (TRL# A-699).
TRL Cat.
Number
CVM-174
CVM-184
CVM-189
CVM-176
CVM-178
CVM-170
CVM-171
CVM-172
CVM-175
CVM-177
CVM-186
CVM-179
CVM-181
CVM-185
CVM-188
CVM-180
CVM-187
CVM-190
CVM-182
Field
Number
8
24
34
10B
14
1A
4
5
9
12
27
15
18
26A
31A
17
290
35
20
Outside Date
Symbola
934 vv
971 vv
987 vv
1020 +H
1027 vv
1029 r
1029 H
1029 vv
1029 r
1029 v
1035 +rB
1037 r
1037 r
1037 r
1039 r
1040 +vv
1041 r
1041 HB
1048 H
Note: All specimens are from Pit structure 1, and all
are juniper.
aFor explanation of outside date symbols, see Table
5.5.
150-
1$0
140
130
120
llO
>-
t:CIl
ffi
0
-j
セ
....
....
Ul
セ
セ
0
セ
70
60
50
40
30
20
10
0
I
900
.
1000
lIOO
1200
YEARS (A.D.)
Figure 5.9. Population density supportable within the 1.6-km-radius catchment of 5MT8371, DCA Site, A.D. 901-1300.
1300
date is A.D. 1029 r. The fmal date from the
site is A.D. 1048 +T. A small cluster of dates at
A.D. 1029 and another at A.D. 1037 may represent the major construction episodes at the
site with the later dates representing minor
repairs. Alternatively, the primary construction
occurred in A.D. 1037 with significant reuse of
wood previously cut and/or used elsewhere, if
the second replastering of the floor was not accompanied by roof repair/construction.
Kuckelman concludes that it is likely that "the
pit structure was either constructed in A.D.
1029 or in A.D. 1037 and was abandoned after
A.D. 1048" (Kuckelman 1988:386). Archaeomagnetic dates from the hearth in Pit Structure 1 proved of little value (Kuckelman
1988:386). Ceramic associations, however,
corroborated a middle-to-Iate eleventh century
placement, ca A.D. 1000--1075 (Kuckelman
1988:386). For the purposes of this study, the
occupational dates assigned to this site are A.D.
1029-1048, a period of 20 years.
Site Catchment and Agricultural Productivity. The 400-year maximum carrying
capacity for the catchment is 67 ± 10 persons/km2 (Table 5.9). These long-tenn population density values are notably higher than
those sustainable in the Mockingbird Mesa
survey locality (36 ± 9 person/km2), in the
study area as a whole (35 ± 7 persons/km2), or
even in the Sand Canyon survey locality (55 ±
10 persons/km2). The critical carrying capacity
for 5MT8839 is also higher than the other
areas with a 4OO-year minimum value of 49
persons/km 2, whereas the others can only sustain a long-tenn minimum population of 18
persons/km 2, 21 persons/km 2, and 37 persons/km 2, respectively. These high population
levels reflect the generally greater potential for
Table 5.9. Comparison of Population Density Values (POPKM) for 5MT8839, Norton House, for A.D.
901-1300 and A.D. 1029-1048.
5MT8839
(Norton)
7.88
653,070
± 99,827
15.3
C.V.b
94
Maximum value c
Maxirn.am CC: Mean value c
67 ± 10
Minimum value c 49 (73%)
Critical CC:
40
Optimal CC: 60% of Mean
27
40% of Mean
20% of Mean
13
Mockingbird
Mesa
17.96
801,336
± 204,994
25.6
57
36±9
18 (50%)
22
14
7
A.D. 901-1300 (400 years)
Area (kfn2)
TOTPRODa(kg)
A.D. 1029-1048 (20 years)
TOTPROD(kg)
Sand Canyon
Locality
26.08
1,779,087
± 337,524
19.0
80
55± 10
37 (67%)
33
22
11
Study Area
1,470.36
64,925,217
± 13,936,845
21.5
57
35±7
21 (60%)
21
14
7
733,352
1,646,677
614,991
59,804,508
± 76,714
± 173,871
± 287,165
± 10,941,721
12.5
23.7
17.4
18.3
C.V.
77 (AD 1042)
47 (AD 1042) 65 (AD 1042)
44 (AD 1042)
Maxiroom value
Maximum CC:Mean value
33±6
64 ± 7
33 ± 7
51 ± 9
19 (58%)
37 (73%)
25 (76%)
Critical CC:
Minimum value 50 (78%)
(AD 1036, 1048) (AD 1035)
(AD 1032)
(AD 1035, 1041)
20
Optimal CC: 60% of Mean
38
20
31
13
40% of Mean
26
13
20
7
20% of Mean
13
7
10
Note: All population values are truncated integers; density km 2
aTOTPROD is the total mean productivity of maize, rounded to the nearest whole number ± standard deviation.
bCoefficient of variation, the ratio of the standard deviation to the mean, multiplied by 100, rounded to the nearest
tenth.
cMaximum, mean ± standard deviation, and minimum population density values (POPKM) for POP2YR are found in
Appendix I (Norton), G (Mockingbird Mesa), F (Sand Canyon) and E (entire study area).
-
152-
agricultural productivity of the nearly uniform,
deep, upland mesatop soils located in the northern portion of the study area.
During the 20 years when 5MT8839 is
thought to have been occupied, the maximum
carrying capacity is, in fact, not as good as the
average conditions for that catchment over the
400-year span, although the difference is
small. In those 20 years, the maximum carrying capacity is 64 ± 7 persons/km2. However,
the 20-year minimum value, or critical carrying capacity, is slightly higher than the longterm minimum value at 50 persons/km2. During its period of occupation, 5MT8839 still
bears the same superior productivity relationship to both sUIvey localities and the study area
as a whole.
Site Catelunent and Prime Periods for
Al:ricultural Production. The population density reconstructions for 5MT8839 were
searched for periods of at least 20 years when
the critical carrying capacity never fell below
50 people/km2 (Appendix I, Figure 5.10).
These periods are identified in Table 5.10.
Using measures of temporal predictability and
spatial productivity described for 5MT8371 ,
the most predictable period of high productivity in the catchment is A.D. 1010-1061,
whereas the most productive period of high
population density is A.D. 1227-1253. Therefore, it appears that 5MT8839 was occupied in
the period of highest temporal predictability.
5MT2433, Aulston Pueblo
Description. 5MT2433 is reported by Kane
. (1975a) and by Morris (1986a). It was located
within and adjacent to the right-of-way for one
of the Fairview laterals, a water pipeline associated with the Dolores River Project. Surface
evidence consisted of a large sherd and lithic
scatter that incorporated two concentrated areas
of rubble and eight areas with dark soil and
vitrified adobe. Analysis of surface ceramics
indicated that the site was multiple-component
with occupations dating to BM III or Pueblo I,
Pueblo II, and possibly Pueblo III (Morris
1986a:4). Subsurface investigations were restricted to two 5-6-m-wide project corridors
that passed thorough portions of the site. One
pipeline trench exposed three pits in one area
(Features 1-3), but these were not excavated
further. A second pipeline trench revealed the
profiles of two pit structures, a burned kiva
(Pit Structure 1) and a mealing room (Pit
Structure 2). These were fully excavated and
provide the dates for this analysis. Of the two,
only burned Pit Structure 1 provided wood
samples suitable for tree-ring dating, and on
the combined evidence derived from tree-rings
and ceramic and architectural characteristics, a
middle Pueblo II date is inferred for this
portion of the site.
Location. The site was located in the SEl/4
of the SEl/4 of Section 35, T38N, RI8W,
N.M.P.M. on the Pleasant View 7.5-minute
quadrangle map. Its UTM coordinates are
4,152,600 m N, 695,180 m E (Zone 12). The
elevation of the site is 2079 m (6,820 ft) on a
broad, relatively flat northeast-southwest trending ridge southeast of Sandstone Canyon and
northwest of Woods Canyon. Today this site is
totally within cultivated farmland. The PDSI
Table 5.10. Periods of Equal or Greater Occupational Attractiveness in the Catchment of 5MT8839
(A.D. 1029-1048).
Rank by
Years a
2
1
6
3 tie
5
4
3 tie
Rank by Minimum
Pop. Densityb
4
3 (5MT8839)
6
5 tie
2
1
5 tie
Beginning
Date A.D.
931
1010
1068
1091
1192
1227
1255
Ending
Date A.D.
971
1061
1089
1130
1214
1253
1294
Number of Minimum PopUlation
Years
Density Value (km2)
42
50
62
50
22
50
40
50
23
60
27
60
40
50
aBased primarily on number of years and secondarily on the minimum population density that can be sustained.
bBased primarily on minimum population density and secondarily on the number of years this can be sustained.
-
153-
300
>-
t:fIl
200
セ
0
,...
VI
セ
セ
Z
セ
セ
0
セ
\00
o ...セAiQG セG ZGuAiGZ GZ G[ BセG [G BGZ セBZ ゥ
900
••LAZュLセ LBZA セLB
\000
..ZL LZ L セL GHL A[L LZBG[ GB[ G ZG G[BG WG
1100
1200
YEARS (A.D.)
Figure 5.10. Population density supportable within the 1.6-km-radius catchment of 5MT8839, Norton House, A. D. 901-1300.
1300
Table 5.11. Tree-Ring Dates from Site 5MI2433,
Aulston Pueblo (TRL# A-689).
TRL Cat.
Number
CVM-143
CVM-145
CVM-155
CVM-136
CVM-191
CVM-126
CVM-194
CVM-124
CVM-144
CVM-156
CVM-140
CVM·141
CVM·153
CVM-123
CVM-152
CVM-193
CVM-122
CVM-128
CVM-131
CVM-154
CVM-151
CVM-157
CVM-125
CVM-146
CVM-134
Field
Number
42
44
54
33
39
13
57
6
43
55
37
38
52
5
51
56
3
14
23
53
50
58
7
45
29
Outside Date
Symbol a
966 w
973 w
989 w
1006 +w
1006 +w
1012 w
1025 w
1026 ++v
1027 w
1029 w
1030 v
1030 w
1030 r
1031 +w
1032 w
1034 ++r
1035 +v
1035 +v
1035 ++r
1035 w
1035 rB
1040 v
1041 +v
1041 r
1042 w
. tatiOns for the dates, with one assigning initial
roof construction in AD. 1035 followed by
remodeling shortly after A.D. 1041. The alternative interpretation places construction in
the early 1040s and regards the earlier timbers
as reused wood. In either case, the kiva was in
use during the decade of the 1040s and, according to the author, may have been use as
late as the early 1050s (Morris 1986a:29).
Based on the above evidence, Aulston
Pueblo is assigned an AD. 1030-1050 occupation for this study.
Note: All specimens are from Pit Structure 1 (a kiva),
and all are juniper.
aFor explanation of outside date symbols, see Table
5.5.
reconstructions used to represent soil moisture
in this 7.88 kIn 2 catchment were drawn from
the Mesa Verde and Ignacio series.
Dating. Architectural comparisons, ceramic
analysis, and treering samples provide the data
for assigning calendrical dates to this site. Excavation revealed two floors within the Pueblo
II-style kiva, but distinct dates could not be
assigned to each surface. Ceramic analysis
indicated a terminal occupation dating to approximately A.D. 1050 (Morris 1986a:29). All
25 tree-ring-dated samples came from the
burned roof fall of the kiva (Table 5.11). Five
of these are cutting dates. The earliest cutting
dates are A.D. 1030 v and 1030 r. The latest
sample from the site is AD. 1042 vv. Two
small clusters are discernible in the tree-ring
data, one ca. AD. 1035 and one in the early
10408. Morris suggests two alternative interpre-
-
Site Cate!unent and Agricultural Productivity. The 400-year maximwn carrying
capacity value for MT2433 is 52 ± 11 persons/km2 with a critical carrying capacity value
of 32 persons/km 2 (Table 5.12). It is markedly
higher than maximum and critical carrying capacity values for both the Mockingbird Mesa
survey locality and for the study area as a
whole (36 ± 9 persons/km2 with an 18 person/km2 minimum for Mockingbird Mesa, and
35 ± 7 セイウッョOォュR
with a minimum of 21 persons/km2 for the study area), but not as high as
values for the Sand Canyon survey locality (55
± 10 persons/km2 with a minimum of 37 persons/km2).
During the 21-year occupational period
assigned to this site, the maximum carrying capacity of the catchment drops to 48 ± 11 persons/km2 with a critical carrying capacity of
32persons/km 2. These estimates resemble those
for the Sand Canyon survey locality, which is
about 11 kIn to the south of the catchment for
Aulston Pueblo, however, the Sand Canyon locality exhibits somewhat greater productivity
and potential population densities for the same
period of time (52 ± 9 persons/km2 with a critical carrying capacity of 37 persons/km2).
Nevertheless, the catchment of 5MT2433 has
higher density values than either the Mockingbird Mesa survey locality or the study area as a
whole during this 21-year period (33 ± 7 persons/km2 with a critical carrying capacity of 19
persons/km2 for Mockingbird Mesa and 33 ± 6
persons/km2 with a critical carrying capacity of
25 persons/km2 for the study area).
5MT2433, Aulston Pueblo (ca. A.D.10301050) can also be compared to 5MT8839,
155-
Table 5.12. Comparison of Population Density Values (POPKM) for 5MT2433. Aulston Pueblo, for A.D.
901-1300 and A.D. 1030-1050.
.
5MT2433
(Aulston)
7.88
500,847
± 115,085
23.0
C.V.b
Maximum value
77
Maximum CC: cMean value
51 ± 12
Minimum value 32 (63%)
Critical CC:
OptimalCC: 60% of Mean 31
40% of Mean 20
20% of Mean 10
A.D. 901-1300 (400 years)
Area (krJl2)
TOTPRODa(kg)
A.D. 1030-1050 (21 years)
TOTPROD (kg)
462,421
± 106,346
23.0
C.V.
64 (AD 1037,
Maximum value
1042, 1049)
48 ± 11
Maximum CC:Mean value
Critical CC:
Minimum value 32 (67%)
(AD 1032)
Optimal CC: 60% of Mean 29
40% of Mean 19
20% of Mean 10
Mockingbird
Mesa
17.96
801,336
± 204,994
25.6
57
36±9
18 (50%)
22
14
7
Sand Canyon
Locality
26.08
1,779,087
± 337,524
19.0
80
55 ± 10
37 (67%)
33
22
11
735,616
± 165,411
22.5
47 (AD 1042)
1,667,214
± 295,324
17.7
65 (AD 1037,
1042, 1049)
52±9
37 (71%)
(AD 1032)
31
21
10
33± 7
19 (58%)
(AD 1035)
20
13
7
Study Area
1,470.36
64,925,217
± 13,936,845
21.5
57
35± 7
21 (60%)
21
14
7
60,277,686
± 10,888,254
18.1
44 (AD 1042)
33±6
25 (76%)
(AD 1035,1041)
20
13
7
Note: All population values are truncated integers.
aTOTPROD is the total mean productivity of maize, rounded to the nearest whole number.
bCoefficient of variation, the ratio of the standard deviation to the mean multiplied by 100, rounded to the nearest
tenth.
cMaximum, mean ± standard deviation, and minimum population density values (POPKM) for (POP2YR) are found
in Appendix J (Aulston Pueblo), G (Mockingbird Mesa), F (Sand Canyon) and E (entire study area).
Norton House, with which is likely contemporary (ca. A.D. 1029-1048). Aulston is located approximately 9.4 km (5.9 mi) southsouthwest of Norton House. Norton is only
slightly higher than Aulston (14.6 m or 48 ft),
but it is located in a larger area of deep mesatop soils than is Aulston. Both exhibit the same
trend of diminished agricultural productivity
and lower levels of human carrying capacity
during the occupational episode. Norton, however, notably outperforms Aulston and can
support a greater population density. Whereas
Aulston can support a maximum carrying
capacity of 48 ± 11 person/km2 and a critical
carrying capacity of 32 persons/km2 during
the occupational period, Norton could support
a maximum carrying capacity of 64 ± 7 persons/krn2 with a critical carrying capacity of 50
persons/km 2. All other things being equal,
-
Norton would have been a more desirable
location than Aulston. Aulston, however, was
still much superior to other locations in the
study area, and its catchment would have been
an attractive location for farming.
Site Catchment and Prime Periods for
Agricultural Production. It is interesting to
note that the 21 years assigned to Aulston's
occupation, A.D. LPUoQセS
include a year
that is as low as any year in the 4OO-year record. Thus, the minimum value for the 21-year
occupational episode is as low as that for the
400-year long-term record. Further, given that
the period of occupation is not as good as the
prevailing long-term conditions, identifying
time periods that were equally good or better
could not be done. Had it been done, anytime
within the 400 years would be equally as good.
156-
Instead, I reduced the length of time for my
search to 16 years beginning at AD. 1035, the
time of the first date cluster, and ending at
AD. 1050. In this 16year period the critical
c::?ing capacity is lifted by two people per
km , resulting in a sustainable population
minimum of 34 persons/km2 rather than 32
persons/km 2.
Using 34 persons/km2 as the minimum
population size and 16 years as the minimum
duration period, nine other time periods of
equal or greater potential for occupation can
be identified within the 400year series (Appendix J, Figure 5.11). Table 5.13 summarizes
these data for 7.88 krn 2 catchment of
5MT2433. A.D. 1091-1122 is the time of the
highest productivity and A.D. 1091-1130 ties
with A.D. 1255-1294 as the time of highest
temporal predictability within the 400 years
considered. It would seem from the ranking
that the component dated as A.D.1030-1050
at 5MT2433 existed in the catchment in an interval that was relatively predictable and productive, but not at either extreme. It would be
of interest to know if the other site components
at 5MT2433 that were not dated by absolute
methods do in fact date to these prime periods
for agricultural production.
5MT3834, Mustoe Ruin
Description. This site is reported by Gould
(1982) and its dating is discussed by Gould
and later by Ahlstrom (1985:499-501). It is a
multiple-component habitation site with an
earlier Pueblo II component dating to sometime between A.D. 900-1050 and a later
Pueblo III component dating to the AD.
Q WセQRSQ
period (Gould 1982:342-345).
Excavations recovered evidence of the earlier
component in the fonn of three surface rooms,
a kiva, a stockade partially enclosing the settlement, an extramural cist, and a trash midden.
The later component was represented by a
roomblock of at least five rooms, a kiva, a
tower, and a trash midden. The later kiva was
burned and provided all the samples used for
tree-ring dating. Consequently, this is the component used for this study.
Location. The site was located within the
Sand Canyon survey locality, approximately 1
km (0.6 mi) southwest of Goodman Point Ruin
(5MT604) and 4 km (2.5 mi) east of Sand
Canyon Pueblo (5MT765). It is located in the
NW 1/4 of the SWI/4 of Section 4, T36N,
R17W, N.M.P.M. on the Arriola 7.5-minute
quadrangle map. Its UTM coordinates are
4,142,050 m N and 700,550 m E (Zone 12). It
is situated on relatively level, deep mesatop
soils on the highland feature known as Goodman Point, west of Goodman Canyon, at an
elevation of 2,054 m (6,740 ft). The PDSI
reconstructions used to model soil moisture
conditions in this 7.88-km 2 catchment are
from the Ignacio and Mesa Verde series.
DatinK. The assignment of calendrical
dates to 5MT3834 derives primarily from
Table 5.13. Periods of Equal or Greater Occupational Attractiveness in the Catchment of 5MT2433
(A.D. 1030-1050).
Rank by
Years a
5
7
6
1 tie
2
9
8
4
3
1 tie
セb。ウ・、
Rank by Minimum
Pop. Densityb
3
4
7 (5MT2433)
5 tie
1
9
8
6
2
5 tie
Beginning
Date A.D.
931
954
1033
1091
1091
1134
1157
1192
1228
1255
Ending
Date A.D.
952
971
1054
1130
1122
1149
1174
1214
1253
1294
Number of Minimum PopUlation
Years
Density Value (km2)
22
39
18
36
22
34
40
34
32
40
34
16
18
34
23
34
26
40
40
34
ーイセュ。 ャケ
on イ・「セョ
of years 、ョセ
seco.ndarily on the ュオゥョセ
population density that can be sustained.
Based prlmanly on minimum population density and secondarily on the number of years this can be sustained.
-
157-
190
180
170
160
ISO
140
セ
f::en
セ
130
120
110
0
100
セ
til
00
E-<
90
j
80
::J
I:l..
0
70
I:l..
60
SO
40
30
20
10
0
I
I
900
1000
1100
1200
YEARS (A.D.)
Figure 5.11. Population density supportable within the 1.6-km-radius catchment of 5MT2433, Aulston Pueblo, A.D. 901-1300.
1300
treering dating and to a lesser degree on' .
artifactual associations. Twentynine wood
samples were dated (Table 5.14). Nine of the
29 were cutting dates. The earliest cutting date
is A.D. 1065 r and is apparently a reused
timber from an earlier structure. The next
seven cutting dates cluster from A.D. 1173 to
1175 and are interpreted as representing the
initial construction of the kiva (Ahlstrom
1985). A second cluster ca. A.D. 12271231
ends with the final cutting date and tenninal
date for the site. Ahlstrom speculates that this
was likely a repair episode due to its weaker
representation in the sample (Ahlstrom
1985:499). No evidence of occupational
discontinuity occurs in the A.D. 11731231
period. Thus, the full 59 years bracketing the
Table 5.14. Tree-Ring Dates from Site
5MT3834, Mustoe Ruin (TRL # A-452).
TRL Cat.
Number
CVM-58
CVM-68
CVM-49
CVM-45
CVM-48
CVM-55
CVM-23
CVM-36
CVM-21
CVM·51
CVM·46
CVM-30
CVM-27
CVM-43
CVM-40
CVM-67
CVM-24
CVM-26
CVM-63
CVM-28
CVM-29
CVM-32
CVM-66
CVM-25
CVM·50
CVM-54
CVM-60
CVM-33
CVM-62
Field
Number
43
56
34
30
33
40
5
20
2
36
31
13
9
28
24
55
6
8
50
10
11
16
54
7
35
39
46
17
49
Outside Date a
Symbol
1016 w
1027 w
1057 +w
1059 +w
1060 ++w
1065 r
1066 +B
1086 +w
1090 ++w
1093 +w
1139 ++B
1150 +v
1160 ++v
1173 r
1174 v
1174 v
1174 r
1174 r
1174 rB
1174 H
1174H
1174 H
1174 HB
1175 r
1229 +v
1229 HB
1230 w
1231 w
1231 rB
. A.D. 11731231 period is used to date this
site. with construction episodes inferred from
treering clusters at A.D. 1175 and 1231.
Site Catchment and Agricultural Productivity. 5MT3834 is within the Sand Canyon
survey locality. The long-tenn population
density values for the catchment of Mustoe are,
not unexpectedly, most like those for the Sand
Canyon survey locality. The 400-year maximum carrying capacity value for 5MT3834 is
56 ± 11 persons/km2 with a critical carrying
capacity of 39 persons/km 2, whereas the 400year maximum carrying capacity for the Sand
Canyon survey locality is 55 ± 10 persons/km2
with a critical carrying capacity 37 person/km2
(Table 5.15). Both 5MT3834 and the Sand
Canyon survey locality are noticeably more
productive than either the Mockingbird Mesa
survey locality or the study area as a whole (36
± 9 persons/km2 with a minimum value of 18
persons/km2 for Mockingbird Mesa survey
locality and 35 ± 7 persons/km2 with a minimum of 21 persons/km2 for the study area).
Given the long period of occupation assigned to this site, it is not surprising that the
maximum and critical carrying capacity values
for the 59-year occupation period at 5MT3834
are the same as those for the 400 years. Although there is a one person/km2 increase in
the maximum carrying capacity values for
both survey localities and the study area for the
A.D. 1173-1231 period, the critical carrying
capacity values for these three areas remain the
same as for the 4OO-year period.
Site Catchment and Prime Periods for
A2ricultural Production. Noting that population density values are not different for the occupation period and that this site is in a portion
of the study area that has already been cited as
being notably productive, a different method
was used to identify those time intervals of
high productivity. Figure 5.12 was visually examined for intervals of sustained higher population density. Data contained in Appendix K
were consulted and Table 5.16 summarizes the
results. The period that exhibits the greatest
temporal predictability is A.D. 1091-1122,
and the period with the highest productivity is
A.D. 1228-1257. The occupation dates assigned to 5MT3834 only partially overlap with
Note: All specimens are from the kiva, and all are
juniper.
aFor explanation of outside date symbols, see
Table 5.5.
-
159-
Table 5.15. Comparison of Population Values (POPKM) for 5MT3834, Mustoe Ruin, for A.D.
901-1300 and A.D. 1173-1231.
5MT3834
(Mustoe)
7.88
553,196
± 108,069
C.V.b
19.5
c
80
Maximum value
Maximum CC: Mean value c
56 ±11
Minimum value c 39 (70%)
Critical CC:
34
OptimalCC: 60% of Mean
40% of Mean
22
11
20% of Mean
Mockingbird
Mesa
17.96
801,336
± 204,994
25.6
57
36±9
18 (50%)
22
14
7
A.D. 901-1300 (400 years)
Area (kfn2)
TOTPRODa(kg)
Sand Canyon
Locality
26.08
1,779,087
± 337,524
19.0
80
55± 10
37 (67%)
33
22
11
Study Area
1,470.36
64,925,217
± 13,936,845
21.5
57
35±7
21 (60%)
21
14
7
A.D. 1173-1231 (59 years)
POP2YR
TOTPROD (kg)
C.V.
Maximum value
Maximum CC: Mean value
Critical CC:
Minimum value
567,693
1,823,532
817,051
66,314,895
± 112,686
± 195,801
± 339,201
± 13,901,244
19.9
24.0
18.6
21.0
80 (AD 1184)
57 (AD 1184) 80 (AD 1184)
57 (AD 1184)
37±8
57±10
36±7
59 ± 11
39 (66%) (AD
18 (49%) (AD 37 (65%) (AD
21 (58%) (AD
1175, 1186,
1186,1227) 1175,1186,
1186,1227)
1191,1215,
1191,1215,
1227)
1227)
34
35
22
Optimal CC: 60% of Mean
22
40% of Mean
24
15
23
14
7
11
20% of Mean
12
7
Note: All population values are truncated integers.
aTOTPROD is the total mean productivity of maize, rounded to the nearest whole number.
bCoefficient of variation, the ratio of the standard deviation to the mean multiplied by 100, rounded to the nearest
tenth.
cMaximum. mean ± standard deviation. and minimum population density values (POPKM) FOR POP2YR are found
in Appendix K (Mustoe Ruin), G (Mockingbird Mesa), F (Sand Canyon) and E (entire study area).
the last period of high population density.
However, the nonnal minimum population
density that can be supported in the catchment
area is already so high at 39 persons/km 2 that
this may not have been an effective issue here
for potential farmers and inhabitants.
Table 5.16. Periods of Greatest Occupational Attractiveness in the Catchment of 5MT3834 (A.D. 11731231).
Rank by
Rank by Minimum Beginning
Ending
Number of Minimum Population
Yearsa
Pop. Densityb
Date A.D.
Date A.D.
Years
Density Value (km 2)
952
22
46
3
2
931
1
3
1091
1122
32
45
2
1 (5MT3834)
1228
1257
30
47
aBased primarily on number of years and secondarily on the minimum population density that can be sustained.
bBased primarily on minimum population density and secondarily on the number of years that this condition can be
sustained.
160-
200
190
180
170
160
ISO
>-
l::CI.l
130
ffi
0
12:0
Z
9セ
110
100
j
90
セ
10
0
70
0'1
セ
セ
140
60
SO
40
30
2:0
10
0
I
,
900
1000
1100
1200
YEARS (A.D.)
Figure 5.12. Population density supportable within the 1.6-km-radius catchment of 5MT3834, Mustoe Site, A.D. 901-1300.
1300
Central Place Sites
5MT6970, Wallace Ruin and 5MT4126, Ida
Jean Ruin.
. masonry over the kiva walls and bench, and
built a Chaco style sub-floor ventilator shaft
under the original floor. Kiva B produced 24
cutting dates assigned to the year A.D. 1124.
Description. The Wallace Ruin is described
by Bradley (1974, 1984, 1988b) and Powers et
all (1983). The Ida Jean Ruin. sometimes
referred to as North McElmo #8, is described
by Brisbin and Brisbin (1973) and Powers et
al, (1983). These sites apparently are contemporary, are only 0.8 kIn (0.5 miles) apart, and,
along with the nearby Haney Ruin, are considered to be Chacoan outliers of the Lake
View Group in the Mesa Verde Area (powers et
a1.1983: 164-167).
Location. 5MT6970. Wallace Ruin. is located some 7 kIn (4.3 mi) northeast of the
modem town of Cortez. It is located within the
NW 1/4 of the SE1/4 of Section 16, T36N.
R15W, N.M.P.M. on the Dolores West 7.5minute quadrangle map. Its UTM coordinates
are 4139660 m N and 721000 mE (Zone 12).
It is situated on a valley bottom immediately
west of Simon Draw, an intermittent stream
flowing into McElmo Creek, at an elevation of
1,896 m (6,220 ft).
Site 5MT6970, Wallace Ruin, is a multistoried, V-shaped pueblo with an enclosed
plaza. It is fronted with gate-like, formalized
trash mounds and a road segment south of the
plaza wall. Although the entire site has not
been excavated, it is estimated to have almost
100 rooms, some of them built in three stories,
and five kivas. Four separate building episodes
have been identified by its excavator on the
basis of architectural characteristics, with a continuous occupation from approximately A.D.
1045-1125 (incorporating the first three
phases), followed by a hiatus, and then a final
reuse and modification of the structure in the
middle to late thirteenth century (Bradley
1988b:29).
5MT4126, Ida Jean Ruin, is 40 m north
and 810 m west of Wallace Ruin at VTM coordinates 4,139,700 m N and 720,190 m E
(Zone 12). It is located within the SW1I4 of the
NW1I4 of Section 16, T36N, R15W, N.M.P.M.
on the Dolores West 7.5-minute quadrangle
map. It is located on a ridge top west of Simon
Draw at an elevation of 1,917 m (6,290 ft).
Site 5Mf4126, Ida Jean Ruin, is less wellknown than the Wallace Ruin and has been
extensively damaged. It is a two-storied pueblo
with approximately 55 rooms built as a Vshaped or E-shaped roomblock with two enclosed kivas and a detached great kiva. Two kivas were adjacent in an upraised plaza enclosed
by a plaza wall. The western of these two kivas,
Kiva A, is described by Brisbin and Brisbin
(1973) who salvaged information from this
site. It originally had been connected to the
eastern kiva, Kiva B, by a tunnel. They note
that it had been originally built as a Mesa
Verde-style, keyhole-shaped kiva with six pilasters, a southern recess, and a floor level ventilator system. Later, renovations done in the
Chaco style had replaced the Mesa Verde-style
pillar pilasters with Chaco-style low pier pilasters, filled and sealed off the southern recess,
added a nine-foot-tall veneer of Chaco banded
For this study, a location positioned halfway between these two sites was chosen to represent the centerpoint of a single catchment
since they are believed to be contemporary and
components of the same cultural complex operating out of this location. This created a
catchment that was only 4.32 kIn 2 in area since
the two sites are adjacent to the eastern margin
of the study area. The name and number for
the better known 5MT6970, the Wallace Ruin,
was retained, and the name and number for the
lesser known 5MT4126, Ida Jean Ruin, was
dropped. PDSI reconstructions used to represent soil moisture conditions in the catchment
were drawn from the Cortez series.
Datin2. At the present time, a total of 44
wood samples provenienced to seven separate
features is reported for 5MT6970 (Table
5.17a). Feature 25, a three-story room associated with Building Period 1 (A.D. 1000-1075),
produced 15 datable samples. Three of the 15
were cutting dates, the earliest being A.D.1037
v, the next AD. 1045 rB, and the last AD.
1052 r. The terminal tree-ring date was A.D.
1071 +vv. Examination of the set of dates from
this feature reveals a strong cluster of 11
samples at ca. AD. 1045, which likely
162 -
Table 5.17a. Tree-Ring Dates from Site
5MT6970, Wallace Ruin (TRL# A-645).
Struct.,
Feature
Feat. 5
Feat. 5
Feat. 6
Feat. 6
Feat. 6
Feat. 6
Feat. 6
Feat. 6
Feat. 7
Feat. 7
Feat. 7
Feat. 7
"rRL Cat.
Numb.
WLR-13
WLR-14
WLR-11
WLR-4
WLR-5
WLR-10
WLR-9
WLR-6
WLR-19
WLR-18
WLR-31
WLR-34
Feat. 7
WLR-38
Feat. 7
WLR-36
Feat. 7
WLR-35
Feat. 7
WLR-32
Feat. 8
Feat. 8
Feat. 8
Feat. 8
Feat. 8
Feat. 8
Feat. 9
Kiva 3
Kiva 3
Kiva 3
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Feat. 25
Fill,
hearth 4
unknown
unknown
WLR-22
WLR-28
WLR-25
WLR-27
WLR·23
WLR-20
WLR-29
WLR-17
WLR-16
WLR-30
WLR-61
WLR-62
WLR-51
WLR-52
WLR-53
WLR-54
WLR-55
WLR-56
WLR-58
WLR-59
WLR·60
WLR-64
WLR-67
WLR-63
WLR-65
WLR-68
Field
Numb.
27
30
20
27
36
39
33
29
51
50
primary
secondary
secondary
secondary
secondary
secondary
138
139
134
129
137
132
37
B6
B2
112
113
95
96
97
98
101
103
110
111
109
117
116
119
27175
WLR·48 W2422
WLR-47 W2421
Tree
Spec.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Out. Date
Symbola
1064 vv
1101 vv
995 vv
1060 ++vv
1073 vv
1073 vv
1074 +vv
1096 vv
1017 vv
1061 vv
1073 vv
1037 vv
Jun.
1059 vv
Jun.
1061 vv
Jun.
1071 +vv
Jun.
1084 vv
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Pond.
Pond.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
1048 +vv
1055 vv
1076 vv
1077 vv
1080 vv
1084 vv
1040 +vv
1040 vv
1054 vv
1062 vv
1035 H
1037 v
1045 HB
1045 HB
1045 H
1045 rB
1045 H
1045 HB
1045 HB
1045 HB
1045 HB
1045 +rB
1045 HB
1052 r
1071 +vv
1039 ++B
Jun.
988 vv
Jun.
1021 vv
constrl!ction ッセ the room
represents the ゥセエ。ャ
and a major bUlldmg event m the ケイッエセゥィ
of the
multistoried pueblo. No dated treenng
samples are known to be associated with Build
ing Period 2 (A.D. 10751100). Features
(Rooms) 5. 6. 7. 8. 9. and Kiva 3 are associated
with Building Period 3 (A.D. 10751125) on
architectural evidence. but none of these
samples provided a 」オエゥョセ
セ。エ・N
N? treering
dates are available for BUlldmg Penod 4
(middle to late 1200s).
Thirtytwo wood samples taken from three
different kivas in Ida Jean Ruin proved datable
(Table 5.17b). Two noncutting date samples
Table 5.17b. Tree-Ring Dates from Site
5MT4126, Ida Jean Ruin (North McElmo #8,
TRL# A-384).
Struct., TRLCat.
Field Tree Out. Date
Feature Numb.
Numb. Spec. Symbol a
Kiva A MVC-1902 39
Jun. 1070 +vv
Kiva A MVC-1904 41
Jun. 1078 ++vv
KivaB
MVC-1876 13
Pond. 1118 vv
Kiva B MVC-1889 26
Pond. 1120vv
KivaB
MVC-1892 29
Pond. 1122 H
KivaB
MVC-1891 28
Pond. 1123 +v
KivaB
MVC-1894 31
Pond. 1123 +r
Kiva B MVC-1870
7
Pond. 1124 v
KivaB
MVC-1866
3
Pond. 1124 r
KivaB
MVC-1871
Pond. 1124 r
8
KivaB
MVC-1874 11
Pond. 1124 r
KivaB
MVC-1875 12
Pond. 1124 r
Kiva B MVC-1877 14
Pond. 1124 r
KivaB
MVC-1878 15
Pond. 1124r
KivaB
MVC-1879 16
Pond. 1124r
Kiva B MVC-1880 17
Pond. 1124r
KivaB
MVC-1881 18
Pond. 1124 r
Kiva B MVC-1882 19
Pond. 1124 r
KivaB
MVC-1884 21
Pond. 1124 r
KIvaB
MVC-1885 22
Pond. 1124 r
Kiva B MVC-1886 23
Pond. 1124 r
KivaB
MVC-1887 24
Pond. 1124r
KivaB
MVC-1890 27
Pond. 1124 r
KivaB
MVC-1893 30
Pond. 1124 r
Kiva B MVC-1895 32
Pond. 1124 r
KivaB
MVC-1897 34
Pond. 1124 r
KivaB
MVC-1898 35
Pond. 1124 r
KivaB
MVC-1899 36
Pond. 1124r
KivaB
MVC-1900 37
Pond. 1124 r
KivaB
MVC-1868
5
Pond. 1124c
Kiva B MVC-1873 10
Pond. 1124 c
KivaC
MVC-1901 38
Jun. 998 vv
aFor explanation of outside date symbols, see Table
5.5.
163-
Table 5.18. Comparison of Population Values (POPKM) for 5MT6970, Wallace Ruin, for A.D. 901-1300 and A.D. 1045-1125.
A.D. 901-1300
400 years
Area(kJn2)
TOTPRODa (kg)
C.V.b
Maximum value
MaximumCC:MeanvahJe
Critical CC: Min. value
Optimal CC:60%C of Mean
40% of Mean
20% of Mean
セ
A.D. 1045-1125
81 years
TOTPROD (kg)
C.V.
Maximum value
Maximum CC: Mean value
Critical CC: Min. value
Optimal CC: 60% of Mean
40% of Mean
20% of Mean
5MT6970
(Wallace)
4.32
297,798 ±
53,732
18.0
78
56±10
37 (66%)
34
22
11
Mockingbird
Mesa
17.96
801,336 ±
204,994
25.6
57
36±9
18 (50%)
22
14
7
Sand Canyon
Locality
26.08
1,779,087 ±
337,524
19.0
80
55±10
37 (67%)
33
22
11
5MT1566
(Lowry)
7.88
409,875 ±
102,360
25.0
68
42±10
26 (62%)
25
17
8
305,680 ±
52,243
17.1
78 (A.D.
1065)
58±9
37 (64%)
(A.D. 1067,
1090)
828,844 ±
202,383
24.4
57 (A.D. 1065)
1,831,160 ±
369,007
20.1
79 (A.D. 1116)
433,100
37±9
18 (49%)
(A.D. 1067,
1090)
35
23
12
22
15
57 ± 11
37 (65%)
(A.D. 1062,
1067, 1081,
1082, 1090)
34
23
11
7
± 112,453
5MT2149
(Escalante)
6.72
291,075 ±
53,857
18.5
58
34±6
24 (70%)
20
14
7
295,823
± 60,601
26.0
68
(AD 1065, 1116)
45± 11
26 (58%) (A.D.
1055, 1062, 1067,
1068, 1081, 1082,
1090)
27
18
20.5
58 (AD 1079, 1087,
1116)
36± 37
24 (67%) (A.D. 1048,
1053, 1055, 1062,
1067, 1068, 1081,
1082, 1092, 1104)
22
14
9
7
f'Jofe: Air population values are fruncated integers.
atotal mean productivny of maize, rounded to the nearest whole number.
bCoefficient of Variation, the ratio of the standard deviation to the mean multiplied by 100, rounded to the nearest tenth.
cpercent of mean value rounded to the nearest whole number.
Study Area
1470.36
64,925,217 ±
13,936,845
21.5
57
35±7
21 (60%)
21
14
7
67,377,205 ±
14,683,637
21.8
56 (A.D. 1065)
37±8
21 (57%)
(A.D.1067, 1090)
22
15
7
came from Kiva A, and one noncutting date
sample came from Kiva C. Neither structure
has been described in print. The remaining 29
samples came from Kiva B and 24 of these
were cutting dates, all dating to A.D. 1124.
Ahlstrom (1985:499500) describes this as a
"strong terminal cluster" and states that it
likely represents a major construction event at
the site. This large cluster provides strong evidence of kiva roof building activities in A.D.
1124 or 1125.
Thus, the Wallace Ruin contributes the
initial occupation date of A.D. 1045 and the
Ida Jean Ruin provides the terminal date of
A.D. 1124. Architectural and artifactual data
provide strong support for the initial construction and use of the Wallace Ruin and Lake
View Group in general. In this study, the period of A.D. 10451125 is used to represent the
major occupational period of the combined
WallaceIda Jean "central place" location.
セN
Site Catchment and Agricultural ProducTable 5.18 summarizes population values for 5MT6970. The 400year maximum
carrying capacity is 56 ± 10 persons/km2 with
a critical carrying capacity of 37 persons/km 2 .
These very high density values are almost identical to the 400year values for the Sand Canyon survey locality, which has a maximum
carrying capacity of 55 ± 10 persons/km2 and
a critical carrying capacity of 37 persons/km2.
They are both markedly higher than those
400year values for the Mockingbird Mesa
survey locality and for the study area (maximum carrying capacity of 36 ± 9 persons/km2
with a critical carrying capacity value of 18
persons/km2, and a maximum carrying capac. ity of 35 ± 7 persons/km2 with a critical carrying capacity of 21 persons/km2, respectively).
Potential mean population values for the
81year occupation period at 5MT6970 are
even higher than the longterm values. The
maximum carrying capacity for the A.D.
10451125 :r;riOd in this catchment is 58 ±9
persons/km with a critical carrying capacity of
37 persons/km 2. The Sand Canyon survey
locality mean values for this period are also a
little higher with a maximum carrying capacity
of 57 ± 11 persons/km 2 and a critical carrying
capacity value of 37 persons/km2. Again, both
are much higher than those for the 81year
period for the Mockingbird Mesa survey locality and for the study area (maximum
carrying capacity of 37 ± 9 persons/km2 with a
critical carrying capacity of 18 persons/km2,
and a maximum carrying capacity of 37 ± 8
persons/km2 with a critical carrying capacity of
21persons/km2), but they, too, are slightly
higher than the longterm conditions for these
places (Figure 5.13).
Site Catchment and Prime Periods for
Agricultural Production. As noted for several
other sites, the minimum population density
that can be sustained over the 81year period
between A.D. 10451125 is 37 persons/km 2,
the same critical carrying capacity value characteristic of the longterm period. Inspection
of values contained in Appendix L and Figure
5.13 revealed that there are, however, a number
shorter time intervals that support noticeably
higher pgpulation densities than 37 persons/km 2. Table 5.19 summaries the results.
Table 5.19. Periods of Greatest OCcupational Attractiveness in the Catchment of 5MT6970 (A.D.
1045-1125).
Ranka by
Yrs.
4
3
6
1
2
5
Rankbby Min.
Pop. Density
3
6
5 (5MT6970)
2 (5MT6970)
1
4
Beginning
Date A.D.
931
982
1042
1091
1228
1281
Ending
Date A.D.
952
1004
1061
1122
1250
1300
Number of
Years
22
23
20
31
23
20
Minimum Population
Density (km2)
45
41
43
45
46
44
aBased primarily on number of years and secondarily on the minimum population density that can be
セオウエ。ゥョ・、N
Based primarily on minimum population density and secondarily on the number of years that this condition
can be sustained.
165 -
200
190
180
110
160
ISO
140
セ
t::
130
CI)
セ
120
0
0\
0\
セ
110
100
90
::J
c..
0
c..
80
10
60
SO
40
30
20
10
0
,
,
900
1000
1100
1200
YEARS (A.D.)
Figure 5.13. Population density supportable within the 1.6-km-radius catchment of 5MT6970, Wallace Ruin, A.D. 901-1300.
1300
A.D. 10911122 is the period of highest temporal predictability at a minimum population
density, while A.D. 1228-1250 is the period of
greatest production that is sustained for a considerable interval of time. It would appear that
the main occupation and use of the combined
Wallace - Ida Jean Ruin occurs during the
times of highest predictability and noticeably
high productivity.
5MT1566, Lowry Ruin
Description. 5MT1566, Lowry Ruin, is a
multiple component, multi-story habitation
often assigned a role in the Chacoan presence
in the Mesa Verde area, although earlier Basketmaker and later, non-Chacoan occupation is
present as well. Paul Martin excavated major
portions of this site in the 1930s and provides
the major report on this site (Martin 1936).
Stabilization work. in the 1970s and reanalysis
of wood samples in the 1980s has resulted in
additional information on Lowry Ruin (Ahlstrom 1985; White and Bretemitz 1976). Excavation revealed a rectangular structure of 40
or more rooms, some up to three stories tall,
eight kivas, and a detached great kiva 61 m
(200 ft) southeast of the major roomblock.
Portions of a road segment have also been
identified at the site (powers et al. 1983: 170).
Location. 5MT1566 is located in the
SWl/4 of the NWl/4 of Section 2, T38N,
RI9W, N.M.P.M. on the Ruin Canyon 7.5minute quadrangle map. Its UTM coordinates
are 4,161,640 m N and 683,620 m E (Zone
12). It is situated on a high point of land on a
mesa top approximately 1 km (.6 mi) west of
Cow Canyon at an elevation of 2,049 m (6,720
ft) and southwest of the modem farming community of Cahone. The PDSI reconstructions
used to model soil moisture conditions in this
7.88-km 2 catchment are drawn from the
Ignacio series.
Dating. Contrasting architectural styles and
stratigraphic relationships indicated to Martin
that Lowry was not a single pre-planned unit
built at one point in time. Rather, he identified
at least six construction episodes (Martin
1936:37, 194) associated with the masonry
pueblo, some occurring after occupational
hiatuses. Based on what tree-ring dates he had
-
av·ailable to. him for portions of the pueblo,
. Martin assigned an A.D. 1086-1106+ date for
the site (Martin 1936:204). Reassessment of
the 53 wood samples from the entire site
(Table 5.20) permitted Ahlstrom to identify 33
cutting dates associated with the first three
building stages (Ahlstrom 1985:339). He concluded that the first episode associated with the
original nucleus of the building (Rooms 10,
15, 19,21, and the great kiva) occurred at ca
A.D. 1085-1090, the second episode (Rooms
8, 11,26, Kiva B, and Great Kiva) between
A.D. 1106-1120, and the third episode (Kiva
A) at ca. A.D. 1120 (Ahlstrom 1985:340).
Thus, for the purpose of this study, a main
occupation period from A.D. 1086-1120 is
assigned to 5MT1566. The first cutting date
for the first building episode associated with
the construction of the Great Kiva is A.D.
1086. The final cutting date associated with the
third period and the remodeling of Kiva A is
A.D. 1120.
Site Catchment and Agricultural Produc-
tiY.i.tt. Table 5.21 summarizes population
values for 5MT1566. The 400-year maximum
for 5MT1566 is 42 ± 10
carrying 」。セゥエケ
persons/km with a critical carrying capacity of
26 persons/km 2. These values are not as high
as those for the Sand Canyon survey locality
(maximum carrying capacity of 55 ± 10
persons/km 2 and 37 person/km2), but they are
higher than for both the Mockingbird Mesa
survey locality and the study area as a whole
(maximum carrying capacity of 36 ± 9 persons/km2 with critical carrying capacity of 18
persons/km2, and maximum carrying capacity
of 35 ± 7 persons/km2 with a critical carrying
capacity of 21 persons/km2, respectively).
The 35-year occupation period for
5MT1566 has a higher maximum carrying capacity (46 ± 10 persons/km2), but the same
critical carrying capaci!y as for the 4OO-year
period (26 persons/km 2)for this catchment.
This is true for the other three areas during this
period, as well. The maximum carrying capacity lifts to 58 ± 10 persons/km2 for the Sand
Canyon survey locality, to 38 ± 8 persons/km2
for the Mockingbird Mesa survey locality and
to 38 ± 7 persons/km2 for the study area
167-
Table 5.20. Tree-Ring Dates from Site 5MT1566, Lowry Ruin (TRL#
A-323)
TRL Cat.
Number
Great Kiva
GP-586 a,b
GP:580-2
Great Kiva
GP-588
Great Kiva
Great Kiva
GP-569 (567,
568,570,571 )
GP-574 (575)
Great Kiva
GP-587
Great Kiva
Great Kiva
GP-580
Great Kiva
GP-580-1 a,b
Great Kiva
GP-584 a,b,c
Great Kiva
GP-577
Great Kiva
GP-576
(578,585)
Room 10
GP-566 (565)
LOW-49
Room 15
Room 15
LOW-21a
Room 15
LOW-25
FML 33-2
Room 19
FML 33-10
Room 19
FML 33-15
Room 21
LOW-41
Kiva A
LOW-37
Kiva A
Kiva A
LOW-38
LOW-39
Kiva A
Kiva A
LOW·40
Kiva A
LOW·42
LOW-36
Kiva A
FML 33-3
Kiva A
LOW-45
KivaB
LOW-47
Kiva B
LOW-46
Kiva B
LOW-32
Kiva 1
GP-559 a,b
Kiva 1
LOW-31
Kiva 1
GP-562
Kiva 1
GP-560
Kiva 1
LOW-27
Kiva II orB
LOW-26
Kiva II or B
LOW-3
Room 8
GP-563
Room 8
LOW-30
Room 8
LOW-28
Room 8
LOW-22
Room 11
LOW-29
Room 11
GP-591
Room 25
FML-33-14
Room 26
FML-33-1
Room 27
FML-33-20
Room 27
FML-33-21
Room 27
GP-582-1 a,b
Room 35
Room on east GP-561
side of ruin
Field
Num.
26
21
6
2
3
4
5
7
1
10
12
11
Tree
Outside
Species Date/Symbol
Jun.
910 w a
Jun.
932 ++w
Jun.
1043 ++B
1086 rB
Jun.
Jun.
Jun.
Pili.
Pond.
Jun.
Jun.
Jun.
1086 +v
1086 v
1106 v
1106v
1109 ++v
1119 w
1172 r
Pond.
Pond.
Pond.
Pitt
Jun.
Jun.
Pitt
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Pond.
Pond.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Pin.
Pin.
Jun.
Jun.
Pin.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
968w
988 +w
989w
1088 v
1085 HL
1090 +r
1089 r
985+w
1027 +w
1040 ++w
1053 +w
1066 vv
1098 +B
1108 ++v
1120 rL
1105 vv
1105 vv
1109 +v
988 vv
991 vv
1105 vv
1105 vv
1108 rB
1106 v
1106 r
1016 vv
1055 ++vv
1106 r
1110 r
839 vv
1120 vv
990 vv
1106 cL
1067 w
1103 rL
1103 rL
1097 vv
1011 +r
-
168-
Site Catchment and
Prime Periods for Agricultural Production. The
minimum population value
for the 35year period is
equivalent to the minimum
value for the entire 400year
series for that catchment, 26
persons/km2. Thus, the value
is useless for purposes of
identifying periods of equal
or greater occupational
attractiveness. If instead a
slightly shorter period of
occupation is identified,
A.D. 10911120, representing 30 years of continuous occupation, then a
minimum value of 32 person/km2 can be used to
search for like sequences.
Inspection of the data contained in Appendix M and
Figure 5.14 for 5MT1566
reveals that no other consecutive sequence of years in
the 400year time series
produces a minimum of 30
years with a minimum population value as high as 32
persons/km2. The two closest
time periods are A.D. 12281253 for 26 years when the
minimum population value
is never lower than 32 persons/km2 and A.D. 931952
for 22 years when the minimum population value is 31
persons/km2 (fable 5.22).
Therefore, it would seem that
interval used to date the
main occupation of Lowry
Ruin contains within it the
period of highest temporal
predictability and productivity for the entire 400 years
under consideration for this
catchment.
5Mf2149, Escalante Ruin
Description. This site is
most thoroughly described
by Hallasi (1979) and is
Table 5.20 (Concluded).
Struet .•
Feature
No
Provenience
No
Provenience
No
Provenience
No
Provenience
TRL Cat.
Number
FML-33-24
Field
Num.
Tree
Species
Jun.
Outside
Date/Symbol
925 +v
FML-33-9
Jun.
FML-33-13
Jun.
1084 w
LOW-34
Pirt.
1089 w
946+w
Note: For explanation of outside date symbols, see Table 5.5.
also discussed by Powers et al. (1983) and
White and Bretemitz (1979). Escalante is a
multiple-component, roughly square, single
story pueblo of some 25 rooms enclosing a
Chaco-style kiva. A second, smaller, Mesa
Verde-style kiva is located immediately south
of the structure. Construction and/or use of this
site dates to two periods, an earlier Chacoan
period and a later Mesa Verdean period, with at
least three occupational surfaces being recognized.
Location. 5MT2149 is located at the apex
of a prominent hill on the west side of the
Dolores River and its floodplain, at a place
where the west-flowing river takes a great bend
to the north. The location is the highest point
in the area at 2,195 m (7,200 ft) and takes in a
particularly panoramic view of the Mesa Verde
region. It is located in the NW1/4 of the SE1/4
of Section 7, T37N R15W, N.M.P.M. on the
Dolores West 7.5-minute quadrangle map. Its
UfM coordinates are 4,150,500 m N and
717,080 m E (Zone 12). The PDSI reconstructions used to model soil moisture in this 6.72
km 2 catchment are drawn from the Mesa Verde series. (The catchment is a little smaller than
the normal 7.88 km 2 formed by creating a
1.6-km radius around the site. This is due to its
position close to the eastern margin of the
study area where the eastern edge of the
catchment is truncated).
Dating. Occupation dates have been assigned to this site on the basis of architectural
styles, stratigraphic evidence, ceramic associations, and limited tree-ring data. Calendrical
assignments are made primarily from ceramics
and tree-ring dates. However, limited excava-
-
tions and few cutting dates
result in a tentative and incomplete chronology of site
use. Ceramic dates for the
earlier Chacoan period place
initial construction sometime
ca. A.D. 1075-1100. A second occupation at or about
A.D. 1150 after a short abandonment is suggested, and a
third and final post-A.D. 1200
reoccupation after a second
abandonment is inferred
(Hallasi 1979:396-397).
Thirty-five wood samples were dated for
Escalante Ruin (Table 5.23). Of these, nine
were cutting dates. Ahlstrom (1985:340--343)
reviews dating arguments for this site. He
acknowledges the interpretive problems posed
by ceramics, stratigraphy, and the tree-ring
data. He tentatively suggests that the weak
cluster of dates in the 1130s may be remodeling events associated with Kiva A, and the
strong cluster at A.D. 1129 for Room 20 may
represent a construction episode at the site.
Whether major construction and use occurred
prior to the late 1120s is not clear from the
current list of tree-ring dates from the site.
No dated samples are available for the
stratigraphically earliest excavated room. The
earliest cutting date from 5MT2149 is A.D.
1124 r from Room 9, a room built during the
second occupation. The latest cutting date is
A.D. 1138 v from Kiva A. The terminal date
on the site is A.D. 1138 vv from Kiva A. Thus,
the period from A.D. 1124-1138 is used to
represent a likely period of occupation at
Escalante Ruin for this study. These dates are
assigned with full acknowledgment that the
initial construction and occupation of the site
may have preceded the beginning date by two
or more decades and persisted for some years
later than the final date.
Site Catchment and Airicultural ProducTable 5.24 summarizes population
values for 5MT2149. The 400-year maximum
carrying capacity for the site is 34 ± 6 persons/km2 with a critical carrying capacity of 24
persons/km 2. These values are most similar to
the values for the study area as a whole and the
セN
169-
170
160
1.50
140
130
>
120
t::
110
セ
100
C/)
0
j
Z
0
....
-.J
0
::J
0..
0
90
80
70
60
0..
.50
40
30
20
10
0
I
I
900
1000
1100
1200
YEARS (A.D.)
Figure 5.14. Population density supportable within the 1.6-km-radius catchment of 5MT1566. Lowry Ruin, A.D. 901-1300.
1300
Table 5.21. Comparison of Population Values (POPKM) for 5MT1566, Lowry Ruin, for A.D. 901-1300 and A.D. 1086-1120.
A.D. 901-1300
400 years
Area(kffi2)
TOTPROoa (kg)
C.V.b
Maximum value
MaxirTlJm CC: Maximum Vaule
Critical CC: Min Value
Optimal CC: 60%C of Mean
40% of Mean
20% of Mean
....
....
A.D. 1086-1120
35 years
TOTPROD(kg)
.....:J
C.V.
Maximum value
Maximum CC: Mean value
Critical CC: Min value
Optimal CC: 60% of Mean
40% of Mean
20% of Mean
5MT1566
(Lowry)
7.88
409,875 ±
102,360
25.0
68
42 ± 10
26 (62%)
25
17
8
Mockingbird
Mesa
17.96
801,336 ±
204,994
25.6
57
36 ± 9
18 (50%)
22
14
7
Sand Canyon
Locality
26.08
1,779,087 ±
337,524
19.0
80
55 ± 10
37 (67%)
33
22
11
5MT6970
(Wallace)
4.32
297,798 ±
53,732
18.0
78
56 ± 16
37 (66%)
34
22
11
5MT2149
(Escalante)
6.72
291,075 ±
53,857
18.5
58
34 ± 6
24 (70%)
20
14
7
442,524 ±
97,810
22.1
68 (AD 1116)
46± 10
26 (57%)
(AD 1090)
28
18
9
843,185 ±
184,260
21.9
54 (AD 1116)
38±8
18 (47%)
(AD 1090)
23
15
8
1,861,444 ±
335,490
18.0
79 (AD 1116)
58± 10
37 (64%)
(AD 1090)
35
23
12
308,549 ±
46,326
15.0
74 (A.D.1116)
58 ±8
37 (64%)
(A.D.1090)
35
23
12
301,478 ±
62,644
20.8
58 (AD 1116)
36± 7
24 (67%)
(A.D.1090, 1104)
22
14
7
Note: All population values are truncated intergers.
a is the total mean productivity of maize rounded to the nearest whole number.
bCoefficient of Variation, the ratio of the standard deviation to the mean multiplied by 100, rounded to the nearest tenth.
cpercent of mean value rounded to the nearest whole number.
Study Area
1470.36
64,925,217 ±
13,936,845
21.5
57
35± 7
21 (60%)
21
14
7
68,542,192 ±
13,413,144
19.6
55 (AD 1116)
38± 7
21 (55%)
(AD 1090)
23
15
8
Table 5.22. Periods of Greatest Occupational Attractiveness in the Catchment of 5MT1566 (A.D.
1086-1120).
Rank by
Yrs. a
3
1
2
Rank by Min. Pop.
Densityb
3
1 (5MT1566)
2
Beginning
Date A.D.
931
1091
1228
Ending
Date A.D.
952
1120
1253
Number of
Years
22
30
26
Minimum Population
Density Value (km2)
31
32
32
aBased primarily on number of years and secondarily on the minimum population density that can be
sustained.
bBased primarily on minimum population density and secondarily on the number of years that this condition
can be consecutively sustained.
Table 5.23. Tree-Ring Dates from Site 5MT2149, Escalante Ruin
(TRL# A-354).
Struct.,
Feature
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Kiva A
Room 6
Room 6
Room 6
Room 9
Room 9
Room 9
Room 9
Room 20
Room 20
Room 20
Room 20
Room 20
Room 20
Room 20
Room 20
Room 20
TRL Cat.
Number
ESC-23
ESC-19
ESC-21
ESC-6
ESC-18
ESC-16
ESC-65
ESC-66
ESC-22
ESC-11
ESC-26
ESC-69
ESC-70
ESC-10
ESC-5
ESC-7
ESC-25
ESC-9
ESC-8
ESC-44
ESC-45
ESC-48
ESC-52
ESC-64
ESC-56
ESC-53
ESC-123
ESC-130
ESC-122
ESC-127
ESC-124
ESC-125
ESC-126
ESC-129
ESC-131
Field
Num.
64
46
51
5
45
43
72
73
52
17
69
79
80
13
4
7
68
10
9
32
33
40
53
62
57
54
56
63
54
60
57
58
59
62
64
Tree
Species
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Pond.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Pond.
Jun.
Pond.
Jun.
Pond.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
Outside
Date/Symbol
988 w
1002 w
1027 w
1057 w
1073 +w
1088 w
1100 w
1105 w
1110 w
1116w
1118 w
1127 w
1128 +w
1128 w
1134w
1136 +v
1137 w
1138 w
1138 v
1084 w
1084 w
1098 w
1112 w
1118 ++rB
1124 r
1134 +B
1112w
1128 +r
1129 r
1129 r
1129 rB
1129rB
1129 rB
1129 rB
1129 rB
Note: For explanation of outside date symbols, see Table 5.5.
-
172-
Mockingbird Mesa survey
locality (maximum carrying
capacity of 35 ± 7 persons/km2 with a critical
carrying capacity of 21 persons/km 2, and maximum
carrying capacity 36 ± 9 persons/km2 with a critical carrying ケエゥ」。セ
of 18 persons/km , respectively). The
400-year maximum and
critical carrying capacity
values for the Sand Canyon
survey locality are much
higher at 55 persons/km2 and
37 persons/km 2.
The 15 years associated
with the occupation for
5MT2149 have only a slightly
higher maximum carrying
capacity at 35 ± 6 persons/km2, but the critical
carrying capacity remains the
same at 24 persons/km 2. Maximum carrying capacity values
for the study area and the
Mockingbird Mesa survey
locality remain the same
during this period although
their critical carrying capacity
is lifted a slight amount (22
persons/km2 for both areas).
Both maximum and critical
carrying capacity values for
the Sand Canyon survey
locality are elevated slightly
from the long-term conditions
during the A.D. 1124--1138
period (56 ± 12 per-sons/km2
Table 5.24. Comparison of Population Values (POPKM) for 5MT2149, Escalante Ruin, for A.D. 901-1300 and A.D. 1124-1138.
A.D. 901-1300
400 years
Area (kffi2)
TOTPRODa (kg)
C.V.b
Maximum value
Maximum CC: Mean value
Critical CC: Min. value
Optimal CC: 60%C of Mean
40% of Mean
20% of Mean
.....
....,J
V.l
A.D. 1124-1138
15 years
TOTPROD (kg)
5MT2149
(Escalantei
6.72
291,075 ±
53,857
18.5
58
34 ± 6
24 (70%)
20
14
7
Mockingbird
Mesa
17.96
801,336 ±
204,994
25.6
57
36 ± 9
18 (50%)
22
14
7
Sand Canyon
locality
26.08
1,779,087 ±
337,524
19.0
80
55 ± 10
37 (67%)
33
22
11
5MT6970
(Wallace)
4.32
297,798 ±
53,732
18.0
78
56 ± 10
37 (66%)
34
22
11
5MT1566
(lowry)
7.88
409,875 ±
102,360
25.0
68
42 ± 10
26 (62%)
25
17
8
405,238 ±
293,201 ±
1,782,290 ±
808,937 ±
291,715 ±
125,280
400,000
203,427
58,062
51,911
30.9
19.8
22.4
25.2
17.8
C.V.
66 (A.D. 1124)
76 (A.D. 1129)
42 (AD 1124, 55 (AD 1129)
74 (AD 1124)
Maximum value
1127,1129)
42 ± 13
55 ± 11
36±9
56 ± 12
35±6
Maximum CC: Mean value
26 (62%)
40 (73%)
22 (61%)
37 (66%)
24 (69%)
Critical CC: Min. value
(AD 1131, 1133) (A.D. 1126, 1131, (A.D. 1131,
(AD 1131,
(AD 1131)
1133)
1137)
1133)
25
33
34
21
22
Optimal CC: 60% of Mean
17
22
14
14
22
40% of Mean
8
11
11
20% of Mean
7
7
Note: All population values are truncated intergers.
atotal mean productivity of maize rounded to the nearest whole number.
bCoefficient of Variation, the ratio of the standard deviation to the mean multiplied by 100, rounded to the nearest tenth.
cpercent of mean value rounded to the nearest whole number.
Study Area
1470.36
64,925,217 ±
13,936,845
21.5
57
35±7
21 (60%)
21
14
7
64,471,631 ±
15,725,883
24.4
49 (AD 1129)
35±8
22 (63%)
(AD 1131)
21
14
7
ISO
140
130
120
110
>t::
100
ffi
0
90
CI)
Z
10
j
70
9
.....
J
+>-
I
セ
60
セ
0
セ
SO
40
30
20
10
0
900
,
,
1000
1100
1200
YEARS (A.D.)
Figure 5.15. Population density supportable within the 1.6kmradius catchment of 5MT2149, Escalante Ruin, A.D. 9011300.
1300
Table 5.25. Periods of Greatest Occupational Attractiveness in the Catchment of 5MT2149
(A.D.11241138).
Rank by
Yrs. a
5
2
4
3
1
Rank by Min.
Pop. Densityb
5
2 (5MT2149)
4
3
1
Beginning
Date A.D.
938
11 05
1192
1228
1259
Ending
Date A.D.
954
1130
1214
1253
1286
Number of
Years
16
26
23
25
28
Minimum Population
Density Value (km2)
30
30
30
30
30
aBased primarily on number of years and secondarily on the minimum population density that can be
liustained.
bBased primarily on minimum population density and secondarily on the number of years that this condition
can be sustained.
sons/km 2. This value is no different than the
longterm minimum population value that can
be sustained over the 4OOyear time series.
Site Catchment and Prime Periods for Ag
Instead, Figure 5.15 was inspected for intervals
of at least 15 years that sustained higher popuricultural Production. The main period of
lation densities. Data contained in Appendix N
occupation for 5MT2149 is estimated to be
A.D. 11241138 with a minimum population
were also reviewed and the results are listed in
value for the 15year period of 24 per
Table 5.25. Since there is no difference in the
minimum population density
for any of these intervals,
Table 5.26. TreeRing Dates From Site 5MT765, Sand Canyon Pueblo they can only be ranked by
(TRL# A856).
the length of time that this
high level endures. The data
Struct.,
TRL Cat.
Field
Tree
Outside
indicate that the A.D. 1259Feature
Number
Num. Species Date/Symbol
1286 period is the most
CCC210
Jun.
1250 w
Struet. 101
53
predictable period of at least
87
927w
Struet. 102
CCC362 a,b
Jun.
15 years in length in the
Struet. 102
CCC234
39
Jun.
933 +w
entire
4ooyear sequence.
44
Struet. 102
CCC239
Jun.
958w
However,
the interval to
Struet. 102
CCC217
22
Jun.
988w
which
the
Escalante Ruin is
Struet. 102
CCC241 a,b
46
Jun.
1005 +w
currently
dated
is almost as
Struet. 102
CCC360
85
Jun.
1029 ++r
long. Therefore, it would
Struet. 102
CCC155
40
Jun.
1039 +w
Struet. 102
CCC339
48
Jun.
1075 w
appear that during the earlier
51
Struet. 102
CCC166
Jun.
1076 w
Chacoan episode at the
Struet. 102
CCC161
46E
Jun.
1090 ++w
Escalante Ruin, the site was
Struet. 102
CCC224 a,b,c
29
Jun.
1112 w
used and grew during a time
Struet. 102
CCC346a,b
56
Jun.
1115 w
of high temporal predict
Struet. 102
CCC356
1124 w
66
Jun.
ability
and high spatial pro
Struet. 102
CCC215
20
Jun.
1130 w
ductivity.
Struet. 102
CCC355 a,b
Jun.
1130 w
65
Struet. 102
CCC132
Jun.
1137 w
8
5Mf765, Sand Canyon
Struet. 102
CCC131
7
Jun.
1146 w
Pueblo
Struet. 102
CCC147
30
Jun.
1160 w
Struet. 102
CCC127
1
Jun.
1165 w
29
Struet. 102
CCC145
Jun.
1167 +w
Description. Sand Can
CCC186
Struet. 102
78
Jun.
1172 w
yon Pueblo is an extensive,
Struet. 102
CCC370
Jun.
1175 ++w
singlecomponent, late
CCC172
Struet. 102
50
Jun.
1180 +w
Pueblo III site with some
CCC235 a,b,c
Struet. 102
40
Jun.
1181 w
300400 rooms, 89 kivas, 14
Struet. 102
CCC348 a,b
58
Jun.
1190 ++w
towers, aDshaped, bi or triwith a critical carrying capacity value at 37
persons/km2).
175-
Table 5.26 (Continued).
Struet.,
Feature
Struct. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struct. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
TRL Cat.
Number
CCC-144
CCC-231 a,b
CCC-207
CCC-350 a,b
CCC-367
CCC-185
CCC-154
CCC-179
CCC-174
CCC-160
CCC-176
CCC-133
CCC-136
CCC-342 a,b
CCC-232
CCC-208
CCC-345
CCC-162
CCC-163
CCC-135
CCC-204
CCC-363
CCC-219
CCC-148
CCC-178
CCC-343
CCC-347
CCC-184
CCC-364
CCC-202
CCC-369 a,b
CCC-349
CCC-341
CCC-343
CCC-222
CCC-225
CCC-227
CCC-230
CCC-236 a,b
CCC-238
CCC-228 a,b,c
CCC-151
CCC-366
eeC-167
CCC-226
eCC-152
CCC-169
CeC-158
CeC-168
Cee-221
CCC-223
CCC-153
CCC-159
CCC-177
Field
Num.
28
36
FS4
60
92
76
38
67
60
46B
61
9
12
52
37
FS 7
55
47
48
10
FS 7
88
24
31
66
53
57
72
89
FS 7
59
51
53
27
30
32
35
41
43
33
34
91
54
31
35
56
44
55
26
28
37
45
62
Tree
Outside
Species Date/Symbol
Jun.
1195 w
Jun.
1198 w
Jun.
1200 w
Jun.
1200 w
Jun.
1213 w
Jun.
1213 w
1214 w
Jun.
Jun.
1217 ++w
Jun.
1220 w
Jun.
1223 w
Jun.
1223 w
Jun.
1226 ++w
Jun.
1226 ++w
Jun.
1228 w
Jun.
1228w
1229 w
Jun.
Jun.
1230 ++v
1232 r
Jun.
1232 w
Jun.
Jun.
1233 +v
Jun.
1233 ++w
Jun.
1234 +v
Jun.
1234 +v
Jun.
1235 r
1235 r
Jun.
Jun.
1235 +v
Jun.
1235 r
1235 rB
Jun.
Jun.
1235 r
Jun.
1235 r
Jun.
1235 r
Jun.
1235 r
Jun.
1235 v
Jun.
1235 +v
Jun.
1235 r B
Jun.
1235 r
Jun.
1235 r
Jun.
1235 w
Jun.
1235 r
1235 r
Jun.
Jun.
1236 v
Jun.
1242 w
Jun.
1242 r
Jun.
1242 r
Jun.
1242 r
Jun.
1247 +w
Jun.
1248 +w
Jun.
1249 +v
Jun.
1249 r
Jun.
1249 H
Jun.
1249 +w
Jun.
1250 HB
Jun.
1250 +w
Jun.
1250 H
-
176-
walled structure (containing
two kivas), a Great Kiva. and
at least one prehistoric
reservoir. Its presence was
noted by Prudden (1903)
and Fewkes (1919), it was
recorded in 1967 (Lister
1967), and portions were
stabilized in 1978 (Metzger
and BreternitzGoulding
1981). Systematic and
intense study began in 1983
by the Crow Canyon Archaeological Center of Cortez,
Colorado. The most recent
overview of this work is in
Bradley (1992). Annual
reports of investigations are
also available (e.g., Adams
1985; Bradley 1986, 1987,
1988a; Kleidon and Bradley
1989).
Location. 5M1765 is
situated on the mesa rim and
along the upper canyon
slopes of Sand Canyon, one
of the major tributaries to
McElmo Creek, directly
north of Sleeping Ute
Mountain. It is a large,
walled, U-shaped community
conforming to the topography of the canyon headlike rincon, incorporating a
significant spring within its
immediate territory. The site
is located in the N1(1 of the
NE1/4 of Section 12, T36N,
R18W, N.M.P.M. on the
Woods Canyon 7.5-minute
quadrangle map. Its geographic center is situated at
an elevation of 2,073 m
(6,800 ft) at UTM coordinates 4,141,200 m N and
696,800 m E (Zone 12). The .
PDSI reconstructions used to
model soil moisture in soils
of this 7.88-km 2 catchment
are drawn from the Mesa
Verde and Ignacio series.
Datin!:. 5M1765 is one
Table 5.26 (Continued).
Struet.,
Feature
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 102
Struet. 107
Struet. 107
Struet. 108
Struet. 108
Struet.204
Struet.204
Struet.208
Struet.208
Struet.501
TRL Cat.
Number
CCC-206
CCC-218 a,b
CCC-233
CCC-351 a,b
CCC-353
CCC-170
CCC-344
CCC-340 a,b
CCC-352 a,b
CCC-359
CCC-201
CCC-138
CCC-150
CCC-200
CCC-214
CCC-354
CCC-165
CCC-209
CCC-171
CCC-205
CCC-194
CCC-192
CCC-371
CCC-372 a,b
CCC-375
CCC-374 a,b
CCC-850
CCC-409
CCC-787
Struet.501
Struet.501
CCC-843
CCC-831
Struet.501
Struet.501
CCC-842
CCC-824
Struet.501
Struet. 501
CCC-758
CCC-784
Struet.501
CCC-795
Struet.501
CCC-792
Struet. 501
CCC-834
Struet.501
CCC-836
Struet.501
CCC-788
Struet. 501
Struet.501
Struet.501
Struet.501
Struet.501
CCC-841
CCC-760
CCC-741
CCC-n2
CCC-791
Field
Num.
FS 7
23
38
61
63
57
54
49
62
70
FS7
15
34
FS7
19
64
52
FS2
58
FS7
16
11
4
331/ 1
216/2
388/
114
693/ 10
3881
166
693/9
3881
162
388/83
3881
224
3881
131
3881
113
3881
219
3881
221
3881
115
693/5
388/86
388/2
388/90
3881
119
Tree
Outside
Species Date/ Symbol
Jun.
1250 +v
Jun.
1250 HB
Jun.
1250 +B
Jun.
1250 w
Jun.
1250 w
1251 H
Jun.
Jun.
1251 +w
Jun.
1251 +v
Jun.
1251 +v
1251 H
Jun.
1257w
Pitt
Pit'!.
1261 w
1264 r
Pitt
Pit'!.
1264w
1267 rB
Pin.
Jun.
1270 H
1271 rB
Jun.
1271 rB
Jun.
1274 r
Pitt
1274 r
Pili.
Jun.
1160 +w
Jun.
1166 +w
Jun.
1036 ++w
Jun.
1229 w
Pitt
1235 w
Pin.
1267 +w
702w
Jun.
Pitt
1244 rB
Jun.
904+w
Jun.
Jun.
925w
928w
Jun.
Jun.
9nw
983 ++w
Jun.
Jun.
987w
994 ++w
Jun.
994+w
Jun.
998 +w
Jun.
1007w
Jun.
1019 +w
Jun.
1038 ++w
Jun.
Jun.
Jun.
Jun.
Jun.
1061 w
1063 w
1084 +w
1088 w
1125 ++w
177-
of the best-dated sites in this
portion of the Mesa Verde
Region. Architectural and
artifactual associations place
the entire site finnly within
the thirteenth century. As of
the spring of 1990, a total of
329 wood samples had been
dated for 5MT765 (Table
5.26). Of these, 124 were
cutting dates. A total of 12
masonry structures, either a
room or a kiva, had produced datable wood samples.
Five of these structures produced samples with cutting
dates and architectural evidence sufficiently strong to
suggest a specific year of
construction (Bruce Bradley,
personal communication,
May 1990). Table 5.27
summarizes these data.
The earliest cutting date
for excavated portions of
5MT765 is A.D. 1200 rB
and the latest is A.D. 1274 r,
which is also the [mal date
for the site. Major clusters of
treering dates occur at A.D.
1203, 1225, 1235, 1250,
1251, 1252, 1260, 1262,
1266, and 1274. Based on
the [mdings of current
investigations in the greater
Sand Canyon survey locality
(Adler 1988; Huber 1989;
Huber and Bloomer 1988;
Van West 1986; Van West et
at. 1987; Varien 1990)
Bradley suspects wholesale
movement of kiva roof lumber in the thirteenth century
to 5MT765 from previous
village locations (Bruce
Bradley, personal communication, March 1990). He
tentatively concludes that
construction and use dates
for 5MT765 do not begin as
early as the first cutting dates
in the 1200s might suggest.
Rather, he places the first
Table 5.26 (Continued).
Struet.,
Feature
Struet.501
TRL Cat.
Number
CCC-793
Struet.501
CCC-825
Struet.501
Struet.501
CCC-757
CCC-806
Struet.501
CCC-813
Struet.501
Struet. 501
CCC-743
CCC-822
Struet.501
CCC-782
Struet.501
Struet.501
Struet.501
CCC-769
CCC-742
CCC-808
Struet.501
CCC-807
Struet.501
Struet.501
Struet. 501
Struet.501
CCC-838
CCC-754
CCC-766
CCC-780
Struet.501
CCC-802
Struet.501
Struet.501
Struet.501
Struet.501
Struet.501
Struet.501
Struet. 501
Struet.501
CCC-846
CCC-748
CCC-761
CCC·764
CCC-765
ccc-no
CCC-n3
CCC-776
Struet.501
Struet.501
CCC-n8
Struet.501
CCC-797
Struet.501
CCC-798
Struet.501
CCC-800
Struet.501
CCC-801
Struet.501
CCC-803
Struet.501
CCC-809
Stnlet.501
CCC-810
Struet.501
CCC-816
CCC-796
Field
Tree
Outside
Num.
Species Date/ Symbol
1141 w
388/
Jun.
127
1145 w
388/
Jun.
163
1146 ++w
388/82 Jun.
1149 w
388/
Jun.
124
1152 w
388/
Jun.
149
1170 w
Jun.
388/4
1188 w
Jun.
388/
158
1189 w
Jun.
388/
104
1190 w
388/77 Jun.
1194 w
388/ 1
Jun.
1195 ++8
388/
Jun.
138
Douglas 1196 +r
388/
fir
126
1196w
693/6
Jun.
1200 w
388/22 Jun.
1200 r8
388/74 Jun.
1201 v
388/
Jun.
102
1201 w
388/
Jun.
121
1202 r
552/1
Jun.
388/ 14 Jun.
1203 r
1203 r
388/87 Jun.
1203 r
388/92 Jun.
1203 r
388/73 Jun.
1203 r
388/78 Jun.
1203 r
388/93 Jun.
388/
Jun.
1203 r
120
1203 r
388/99 Jun.
1203 r
388/
Jun.
143
1203 r
Jun.
388/
132
Jun.
1203 r
388/
135
1203 r
388/
Jun.
137
1203 r
388/
Jun.
139
Jun.
1203 r
388/
142
Jun.
1203 v
388/
140
1203 r
Jun.
388/
144
1203 r
Jun.
388/
152
-
178-
solid evidence for con
struction ca. A.D. 1252 on
the basis of architectural
properties in conjunction
with treering dates, and he
has evidence for more-orless continuous construction
and site use until at least
A.D. 1274.
For the purposes of this
study, an occupation of A.D.
1252-1274, a period of 23
years, is assigned to
5MT765, with the knowledge that the initial construction and use of this site
may be several decades earlier and it may have been
used for a decade later than
the last date.
Site Catchment and Agricultural Productivity. Table
5.28 summarizes population
values for 5MT765. The
400-year maximum c。イケゥョセ
capacity within the 7.88 kIn
radius of the site is 47 ± 9
persons/km2 with a critical
carrying 」。セゥエケ
of 10
persons/km . This extremely
low population density value
happens only once in the
4OO-year sequence, at A.D.
1277. Otherwise, the low is
31 per-sons/km 2, which
happens 31 times in the four
centuries. Although the
maximum carrying capacity
for this catchment is higher
than the maximum carrying
capacity values for either the
Mockingbird Mesa survey
± 9 perlocality セSV
sons/km ) or the study area
as a whole (35 ± 8 persons/km2), the critical carrying capacity value of 10
persons/km2 for the catchment is significantly lower
than either of these two respectively). Furthermore, the
maximum and critical
Table 5.26 (Continued).
Struet..
Feature
Struet.501
TRL Cat.
Number
CCC-817
Struet.501
CCC-819
Struet.501
CCC-823
Struet. 501
CCC-830
Struet. 501
CCC-835
Struet.501
Struet.501
Struet.501
CCC-840
CCC-845
CCC-815
Struet.501
Struet.501
CCC-774
CCC-789
Struet.501
Struet.501
Struet.501
CCC-740
CCC-744
CCC-805
Struet.501
Struet.501
CCC-775
CCC-786
Struet.501
CCC-804
Struet.501
CCC-818
Struet.501
CCC-832
Struet. 501
Struet.501
Struet.501
Struet.501
CCC-745
CCC-762
CCC-759
CCC-820
Struet.501
CCC-829
Struet.501
Struet.501
Struet.501
Struet. 501
Struet.501
Struet.501
CCC-749
CCC-753
CCC-756
CCC-767
CCC-771
CCC-799
Struet.501
CCC-828
Struet.501
CCC-833
Struet.501
CCC-811
Struet.501
CCC-794
Struet.501
CCC-821
Outside
Tree
Field
Species Date/ Symbol
Num.
1203 r
Jun.
388/
225
1203 r
Jun.
388/
155
1203 r
Jun.
388/
161
1203 r
Jun.
388/
164
1203 r
388/
Jun.
220
1203 r
Jun.
693/8
1203 r
Jun.
548/7
1204 w
Jun.
388/
151
1215 +r
388/94 Jun.
Jun.
388/
1215 ++v
117
Jun.
1220 +r
388/3
Jun.
388/8
1220 +v
Jun.
388/
1220 +w
123
388/96 Jun.
1221 r
Jun.
388/
1221 r
110
Jun.
388/
1221 w
122
Jun.
1221 r
388/
154
388/
Jun.
1221 r
171
Jun.
388/6
1222 r
388/88 Jun.
1222 r
388/85 Jun.
1223 w
388/
Jun.
1223 w
159
Jun.
388/
1223 v
170
388/ 15 Jun.
1224 r
388/21 Jun.
1224 w
388/81 Jun.
1224 r
388/75 Jun.
1224 r
388/79 Jun.
1224 r
388/
Jun.
1224 r
136
388/
Jun.
1224 r
169
388/
Jun.
1224 r
172
Jun.
388/
1225 r
146
388/
Jun.
1238 r
128
388/
Jun.
1238 r
160
179-
carrying capacity values for
the Sand Canyon survey
locality, the area in which
UVWtセU
L、・セ」ッャウゥ
are far
higher than those for the site
(55 ± 10 persons/km2 and
37 persons/km2). This discrepancy can be partially
accounted for by noting that
UセtWV
is located on the
mesatop edge and sloping
canyon walls of Sand Canarea
yon. The ・セゥ、ュ
around the site is characterized by shallow or steeply
would
sloping sediments セィエ
result in low yields given
dryfarming techniques.
Therefore, exploitation of
arable lands higher on the
Lーッセウ・ュ
and thus, at a
ァイ・セ イ
distance from the site
proper, would have been
essential to increasing the
minimum population supportable at the site itself.
The maximum and
critical carrying capacity
values for the 23year
ョッゥセーオ」 ッ
period assigned
to UVセ
indicates that
this A.D. 12521274 period
would have been able to
support a mean population
density of 48 ± 8 persons/km2 with a minimum
population density of 31
persons/km2 セ all times.
While the 23year maximum
carrying capacity value is
only slightly higher than the
400year average of at 47 ±
9 persons/km2, the critical
carrying capacity value is
over three times as high セ
31 persons/km 2. During the
same 23year interval, the
is
catchment for U セ t W V U
again markedly better than
rd 。ウ・セ
the ゥ「ァョォ」ッセ
survey locality or the study
area as whole (37 ± 9 persons/km2 with a minimum of
Table 5.26 (Continued).
Struet.,
Feature
Struet.501
Struet.501
TRL Cat.
Number
CCC-755
CCC-812
Struet.501
Struet. 501
Struet.501
Struet.501
Struet.501
CCC-746
CCC-751
CCC-763
CCC-768
CCC-n9
Struet.501
CCC-837
Struet.501
CCC-790
Struet.501
Struet.501
CCC-747
CCC-814
Struet.501
Struet. 504
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struct. 1004
Struct. 1004
Struet. 1004
CCC-839
CCC-851
CCC-737
CCC-736
CCC-717
CCC-703
CCC-718
CCC-731
CCC-699
CCC-712
CCC-700
CCC-723
CCC-716
CCC-725
CCC-726
CCC-701
CCC-730
CCC-739
CCC-729
CCC-732
CCC-713
CCC-719
CCC-702
CCC-714
CCC-706
CCC-696
CCC-708
CCC-721
CCC-735
CCC-738
CCC-695
CCC-707
CCC-734
CCC-728
CCC-727
CCC-710
CCC-698
CCC-704
Field
Tree
Outside
Num.
Species Date/ Symbol
388/80 Jun.
1246 ++v
388/
Jun.
1249 +r
145
388/7
Jun.
1250 r
723/1
Jun.
1250 r
388/91 Jun.
1250 r
388/76 Jun.
1250 r
388/
Jun.
1250 r
100
388/
Jun.
1250 r
223
388/
Jun.
1251 +r
116
388/ 11 Jun.
1252 +vv
388/
Jun.
1252 +r
150
693/7
Jun.
1252 v
345/1
Jun.
1065 vv
530/1
Jun.
911 VV
584/1
Jun.
924 +vv
530/33 Jun.
994vv
530/ 13 Jun.
1155 vv
530/34 Jun.
1156 ++vv
530/29 Jun.
1170 vv
530/8
Jun.
1183 +vv
530/ 14 Jun.
1184 vv
530/9
Jun.
1189 vv
530/41 Jun.
1190 vv
530/25 Jun.
1200 ++B
530/26 Jun.
1201 ++vv
530/43 Jun.
1205 vv
530/ 10 Jun.
1209 vv
530/27 Jun.
1223 vv
530/3
Jun.
1223 vv
530/28 Jun.
1238 +r
530/30 Jun.
1242 vv
530/ 18 Jun.
1243 r
530/35 Jun.
1244 r
530/ 11 Jun.
1246 vv
530/ 15 Jun.
1247 vv
530/22 Jun.
1248 vv
530/5
Jun.
1249 r
530/ 17 Jun.
1249 +r
530/36 Jun.
1249 r
1249 r
530/2
Jun.
1249 r
588/1
Jun.
586/1
Jun.
1250 r
530/ 19 Jun.
1250 v
530/32 Jun.
1250 r
530/40 Jun.
1253 +vv
530/44 Jun.
1262 vv
530/24 Jun.
1263 +vv
1264 +r
530/7
Jun.
1264 +r
530/20 Jun.
180-
18 persons/km2, and 36 ± 7
persons/km2 with a minimum of 21 persons/km2,
respectively). Nevertheless,
the catclunent for the site
during this interval is not as
good as other potential
catclunents within the Sand
Canyon survey locality,
which has a population density potential that is virtually
unchanged from longterm
conditions (maximum carrying capacity of 55 ± 14 persons/km2 and a critical
carrying capacity of 37 persons/km2).
Site Catclunent and
Prime Periods for AgriculturaJ Production. With
the exception of year A.D.
1277 when the maximum
arumal population density
that could be supported on
the potential productivity
from the 7.88 km 2 catchment was only 10 persons/km2, the minimum
population density value for
the 23year occupation
period is the same as it is for
the 400year period at 31
persons/km 2. Consequently,
Figure 5.16 and values contained in Appendix 0 were
examined for intervals of
sustained higher population
density. The results of this
search are contained in
Table 5.29. The period of
highest temporal predictability is the A.D. 931971
period and the period of
highest productivity is the
A.D. 12281253 period. It
would seem likely that
5MT765 was established in
this period of greatest productivity for the catchment,
and continued through
another period of relatively
high and predictable productivity (A. D. 1255-
Table 5.26 (Continued).
Struet.,
Feature
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1004
Struet. 1203
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1205
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet. 1206
Struet.
Struet.
Struet.
Struet.
Struet.
Struet.
1206
1206
1206
1206
1206
1206
TRL Cat.
Number
CCC722
CCC724
CCC711
CCC705
CCC410
CCC416
CCC414
CCC423
CCC412
CCC422
CCC425
CCC421
CCC418
CCC424
CCC413
CCC420
CCC411
CCC415
CCC417
CCC377
CCC198
CCC397
CCC395
CCC408
CCC400
CCC399
CCC381
CCC405
CCC338 a,b
CCC197
CCC404
CCC396
CCC391
CCC398
CCC337 a,b
CCC299,300
CCC388
CCC379
CCC383
CCC386
CCC387
CCC389
CCC390
CCC332
CCC328
CCC302
CCC306,320
a,b
CCC318 a,b
CCC295
CCC296
CCC301,307
CCC311 a,b
CCC312
1276).
Field
Tree
Outside
Num. Species Date/ Symbol
1264 H
530/39 Jun.
1264 H
530/42 Jun.
1265 H
530/ 16 Jun.
1266 r
530/21 Jun.
Jun.
242/1
1169 +w
Jun.
246/8
1026 +w
1031 w
Jun.
246/6
1061 w
Jun.
182/8
1181 w
182/7
Jun.
1208 w
Jun.
246/4
Jun.
182/9
1229 +w
1231 H
246/ 13 Jun.
1236w
246/ 10 Jun.
Jun.
182/9
1248 w
Jun.
1249 HB
246/5
246/ 12 Jun.
1251 +rB
246/3
Jun.
1252 ++r
246/7
Jun.
1260 rB
Jun.
246/9
1260 rB
Jun.
166/2
934w
74
Jun.
972+w
Jun.
1031 +w
231/6
231/ 14 Jun.
1036 ++w
224/3
Jun.
1036 +w
231/7
Jun.
1049 ++w
231/ 21 Jun.
1058 +w
Jun.
166/3
1059 w
174/ 1
Jun.
1090 +w
Jun.
78
1138 +w
82
Jun.
1171 ++w
143/27 Jun.
1176 ++w
231/ 12 Jun.
1181 +w
Jun.
143/1
1193 w
231/4
Jun.
1202 ++w
77
Jun.
1224w
27
Jun.
1241 HB
167/2
Jun.
1241 w
166/3
Jun.
1242 rB
167/ 1
Jun.
1242 r
167/2
Jun.
1242 v
167/2
Jun.
1242 r
167/2
Jun.
1242 r
126/22 Jun.
1242 r
Jun.
71
1242 r
67
Jun.
1242 r
32
Jun.
1242 r
36
Jun.
1242 r
52
23
24
31
41
43
Jun.
Jun.
Jun.
Jun.
Jun.
Jun.
1242
1242
1242
1242
1242
1242
181 -
r
rB
rB
rB
rB
rB
Concluding Remarks
Table 5.30 presents the
longterm population
density values at the
POP2YR level of demand for
the catchments of the eight
treering dated sites. While it
is unwise to suggest general
trends in behavior on the
basis of only eight sites, it is
instructive to conclude this
examination of catchment
data with a ranking of these
different places on the productive landscape.
Of the four small village
sites, 5MT8839 (Norton
House) is clearly located in
the potentially most productive and most predictable
catchment. This is indicated
by the relatively high long
term mean value, the maximum carrying capacity, and
relatively low coefficient of
variation. Based on the mean
population density value for
each catchment, the ranking
for small village sites is
5MT8839 (Norton) is highest, followed by 5MT3834
(Mustoe), 5MT2433 (Aulston), and finally 5MT8371
(DCA Site). Of the four
central place sites, 5MT6970
(Wallace) has the highest
longterm mean population
density, followed by
5MT765 (Sand Canyon),
5MTl566 (Lowry), and
5MT2149 (Escalante).
It is interesting to note
that of the two contemporary
small village sites, 5MT8839
(Norton) and 5MT2433
(Aulston), the former is by
far the more attractive as a
locale for high potential
Table 5.26 (Concluded).
Struct.,
TRL Cat.
Feature
Number
Struct. 1206
CCC-317
Struct. 1206
CCC-323,322
Struct. 1206
CCC-326a,b
Struct. 1206
CCC-331
Struct. 1206
CCC-335 a,b
Struct. 1206
CCC-336
Struct. 1206
CCC-315 a,b
Struct. 1206
CCC-327a,b,c
Struct. 1206
CCC-324
Struct. 1206
CCC-393
Struct. 1206
CCC-333
Struct. 1206
CCC-385
Struct. 1206
CCC-378
Struct. 1206
CCC-334
Struct. 1206
CCC-382
Struct. 1206
CCC-380
Struct. 1206
CCC-406
CCC-407
Struct. 1206
Struct. 1206
CCC-305
CCC-308
Struct. 1206
Struct. 1206
CCC-314 a,b
Struct. 1206
CCC-321
Struct. 1206
CCC-325 a,b
Struct. 1206
CCC-329
Struct. 1206
CCC-297,310
Struct. 1206
CCC-303
Struct. 1206
CCC-294
Struct. 1206
CCC-196
Struct. 1206
CCC-313 a,b
Struct. 1206
CCC-330
Struct. 1206
CCC-319 a,b
Struct. 1206
CCC-394
Struct. 1206
CCC-373 a,b
Feature 2
Note: For explanation of outside date
Field
Num.
51
59
65
70
75
76
48
66
62
231/ 1
72
167/ 1
166/3
73
166/3
166/3
174/ 1
174/ 1
35
38
45
55
64
68
25
79
22
32
44
69
53
231/2
Tree
Outside
Species Date/ Symbol
Jun.
1242 rB
Jun.
1242 rB
Jun.
1242 rB
Jun.
1242 rB
Jun.
1242 rB
1242 rB
Jun.
Jun.
1242 v
1242 v
Jun.
1242 w
Jun.
Jun.
1245 ++rB
Jun.
1259 w
1260 w
Jun.
Jun.
1261 rB
Jun.
1261 rB
Jun.
1261 B
Jun.
1262 r
Jun.
1262 r
Jun.
1262 r
Jun.
1262 r
Jun.
1262 r
Jun.
1262 r
Jun.
1262 +r
Jun.
1262 w
Jun.
1262+r
Jun.
1262 +r
Jun.
1262 +r
Jun.
1262 +rB
Jun.
1262 +rB
Jun.
1262 +rB
Jun.
1262 +rB
Jun.
1262 w
Jun.
1265 +rB
Jun.
1242 rB
2
symbols, see Table 5.5.
maize yields and high population densities. Similarly,
of the three partially con
temporary central place sites,
5MT6970 (Wallace) is po
tentially more productive
than 5MT1566 (Lowry) and
that both are markedly more
capable of sustaining higher
population densities than
5MT2149 (Escalante). If the
dates used to bracket the
major Chacoan components
present at these sites are
reasonably correct, then
Wallace's productive pri
macy may be a factor in its
early establishment and
enduring presence in the
region.
The mean population
value (the maximum carry
ing capacity) for the four
small villages sites as a group
is 53.0 ± 12.9 persons/km2
(C.Y. = 24.3) whereas the
mean for the four central
place sites is 45.0 ± 8.8
persons/km 2 (C.Y.= 19.6).
Likewise, the mean mini
mum popUlation density (the
critical carrying capacity)
for all the small village sites
is 32.35 ± 17.0 (C.Y.= 52.7)
persons/km2 and the mean
Table 5.27. Suggested Construction Dates (A.D.) for Tree-Ring Dated Structures at 5MT765.
Block
No.
100
200
500
1000
1200
Struct.
No.
101
102
107
108
204
208
501
504
1004
1203
1205
1206
Function
room
kiva
roomlkiva
kiva
room
kiva
kiva
room
kiva
room
room
kiva
Last Date or
Range of Dates
1250w
1232 r-1274 r
1166 +w
1229w
1267 +w
1244 rB
1200 rB-1252 +w
1065w
1243 r-1266 r
1169 +w
1260 rB
1242rB-1265+rB
182-
Total
Samples
1
99
2
2
2
2
98
1
40
1
14
67
Cutting
Dates
0
26
0
0
0
1
51
0
10
0
2
34
Suggested
Constr. Date
?
1274
?
?
?
?
1252
?
1266
?
1260
1262
Table 5.28. Comparison of Population Values (POPKM) for 5MT765, Sand Canyon Pueblo, for A.D. 901-1300 and A.D. 1252-1274.
A.D. 901-1300
400 セ・。イAI⦅
Area (km2)
TOTPRODa(kg)
C.V.b
Maximum value
Maximum CC: Mean value
CriticalCC: Minimum value
Optimal CC: 60%C of Mean
40% of Mean
20% of Mean
....
00
u.J
5MT765
(Sand Canyon)
7.88
459,114 ± 92,902
20.2
71
47 ± 9
10 (21%)
28
19
9
Mockingbird
Mesa
17.96
801,336 ± 204,994
25.6
57
36 ± 9
18 (50%)
22
14
7
Sand Canyon
Locality
26.08
1,779,087 ± 337,524
19.0
80
55 ± 10
37 (67%)
33
22
11
Study Area
1470.36
64,925,217 ± 13,936,845
21.5
57
35±8
21 (60%)
21
14
7
A.D. 1252-1274
23 years
TOTPROD (kg)
466,172 ± 81,809
65,843,368 ± 12,921,943
815,258 ± 202,795
1,750,494 ± 468,955
C.V.
17.6
24.9
19.6
26.8
46 (AD 1259, 1271)
Maximum value
59 (AD 1255, 1271) 54 (AD 1259)
72 (AD 1271)
Maximum CC: Mean value
48 ± 8
36± 7
37±9
55± 14
21 (58%) (AD 1254)
Critical CC: Minimum value 31 (65%) (AD 1254) 18 (49%) (AD 1254) 37 (67%) (AD 1254)
22
22
33
29
Optimal CC:60% of Mean
14
22
19
15
40% of Mean
7
7
10
11
20% of Mean
Note: All population values are truncated integers.
atotal mean productivity of maize rounded to the nearest whole number.
bCoefficient of Variation, the ratio of the standard deviation to the mean multiplied by 100. This value is rounded to the nearest tenth.
cpercent of mean value rounded to the nearest whole number.
180
170
160
ISO
140
130
>-
l::
120
セ
110
tI)
0
Z
0
....
90
-(
..J
110
I--
..-
00
セ
100
:J
00
0-
70
60
SO
40
30
20
10
0
I
900
1000
1100
1200
YEARS (A.D.)
Figure 5.16. Population density supportable within the 1.6-km-radius catchment of 5MT765, Sand Canyon Pueblo, A.D. 901-1300.
1300
Table 5.29. Periods of Greatest Occupational Attractiveness in the Catchment of 5MT765 (A.D.
1252-1274).
.
Rank by
Yrs. a
1
6 tie
2
4
5
3
6 tie
Rank by Min. Pop.
Densityb
5
6tie
4
2
3
1 (5MT765)
6 tie(SMT765)
Beginning
Date A.D.
931
1033
1091
1105
1192
1228
1255
Ending
Date A.D.
971
1054
1130
1130
1214
1253
1276
Number
of Years
42
22
40
26
23
26
22
Min. Population
Density (km2)
32
32
33
38
38
39
32
aBased primarily on number of years and secondarily on the minimum population density that can be
セオウエ。ゥョ・、N
Based primarily on minimum population density and secondarily on the number of years that this condition
can be sustained.
minimum population density for the central
place sites is 24.25 ± 11.1 persons/km2 (C.V.=
45.8) These statistics would seem to indicate
that the small villages were located in catchments that were more productive and more
capable of sustaining higher minimum population levels than those for the central place
sites. Although this is an admittedly small
sample, these data seem to indicate that the
agricultural potential of a residential location
and its immediate environs was a stronger
factor in site selection for inhabitants of small
villages than it was for inhabitants of central
place sites.
Only two of the eight sites were occupied
in time intervals that could support a higher
minimum population density (the critical car
-
rying capacity) than in the 4OOyear interval.
Seven of eight (the exception being 5MT3834,
Mustoe), however, were dated to a time period
considered among the best for their particular
catchments, seemingly indicating an awareness
of and selection for locations and intervals of
high agricultural productivity.
Finally, in closing, it must be said that it is
not possible to know how accurate are the absolute population density values presented in
this study. Nevertheless, they provide at least a
relative ordering of the productive capabilities
for the eight site catchments, as well as the two
survey localities and the study area. As such,
they are useful for making comparisons and
suggesting research questions.
185-
Table 5.30. Summary of Population Values (POPKM) Using POP2YR Estimates for Eight Tree-Ring Dated Sites,
A.D. 901-1300.
00
0\
Occupation Dates
Area (km2)
TOTPRODa (kg)
C.V.b
Maximum value
Maximum CC: Mean Value
Critical CC: Minimum Value
Optimal CC: 60%C of Mean
40% of Mean
20% of Mean
5MT8371
(DCA)
A.D. 935-950
7.88
350,669 ± 104,956
29.9
57
35 ± 10
9 (26%)
21
14
7
5MT8839
(Norton)
A.D. 1029-1048
7.88
653,070 ± 99,827
15.3
94
67 ± 10
49 (73%)
40
26
13
5MT2433
(Aulston)
A.D. 1030-1050
7.88
500,847 ± 115,085
23.0
77
51 ± 12
32 (63%)
31
20
10
5MT3834
(Mustoe)
A.D. 1173-1231
7.88
553,196 ± 108,069
19.5
80
56 ± 11
39 (70%)
34
22
11
Occupation Dates
Area (km2)
TOTPROD (kg)
C.V.
Maximum value
Maximum CC: Mean Value
Critical CC: Minimum Value
Optimal CC: 60% of Mean
40% of Mean
20% of Mean
5MT6970
(Wallace)
A. D. 1045-1125
4.32
297,798 ± 53,732
18.0
78
56 ± 10
37 (66%)
34
22
11
5MT1566
(Lowry)
A.D. 1086-1120
7.88
409,875 ± 102,360
25.0
68
42 ± 10
26 (62%)
25
17
8
5MT2149
(Escalante)
A. D. 1124-11 38
6.72
291,075 ± 53,857
18.5
58
34±6
24 (70%)
20
14
5MT765 (Sand
Canyon Pueblo)
A.D. 1252-1274
7.88
459,114 ± 92,902
20.2
71
47±9
10 (21%)
28
19
9
7
Note: All population values are truncated integers.
atotal mean productivity of maize rounded to the nearest whole number.
bCoefficient of Variation, the ratio of the standard deviation to the mean multiplied by 100 and rounded to the nearest tenth.
cpercent given is percent of mean value rounded to the nearest whole number.
6
SUMMARY AND EVALUATIONS
This chapter briefly summarizes the
methods and basic conclusions of the study. It
also describes the strengths and weaknesses of
the methods and data used to reconstruct
paleoclimate and agricultural productivity. I
conclude with suggestions for future research
that would improve the model and with suggestions for additional studies that might be
accomplished with the resulting databases.
POP2YR, POP3YR). From these annual estimates, longer time period estimates of a sustainable population-carrying capacity estimates-are made. Three different approximations of carrying capacity are defmed: a
maximum carrying capacity equal to the longterm (400-year) mean; a critical carrying capacity equal to the long-term (400-year) minimum annual value; and an optimal carrying
capacity range equal to some value between
20% and 60% of the long-term mean value.
SUMMARY
The preceding chapters describe a method
of reconstructing paleoclimate for the late
Anasazi occupation period of southwestern
Colorado (A.D. 901-13(0) and the predicted
influence of climatic variation on dryland
agricultural potential. Tree-ring data, strongly
reflecting the regional climate of the southeastern portion of the Colorado Plateaus are
used to retrodict 1070 years (A.D. 901-1970)
of Palmer Drought Severity Indices (PDSI) for
the month of June for local soils and local elevational settings. Reconstructed PDSI values associated with each soil and each 4-ha unit of
space in the 1470.36-k.m2 study area are reexpressed as potential maize yields. The integration, quantification, and visual display of
these productivity values are coordinated
through geographic information system (GIS)
technology. The method results in the production of 1) annual maps depicting the variable character of the potential agricultural
environment and 2) annual values for total
maize productivity (TOTPROD) that can be
translated into the number (POPNUM) and
density (pOPKM) of people that can be supported on that yield. Three different estimates
for population are provided, representing a
population that annually demands the equivalent of either one, two, or three years of maize
in storage at the end of harvest (pOP1YR,
The conservative POP2YR value, representing a population requiring two years of maize
to be placed in storage at the end of harvest,
and the critical carrying capacity value, representing the maximum population size/density
that can be sustained during the lowest producing year(s) of the 400-year record, are used
explore archaeological questions at three spatiallevels. First, they are used to question
whether maize production in the study area was
sufficient to sustain a large number of people
during the 400 years examined, and to examine the proposition that climatic fluctuation resulted in the collapse of Anasazi agricultural
systems and directly promoted the thirteenth
century abandonment. Second, they are used
to assess the productivity and predictability of
maize agriculture in two localities within the
study area, and to compare these model-derived estimates of carrying capacity with
archaeological estimates of population in order
to determine whether the productive limits of
more localized areas had been reached at different times and different places. Third, they
are used to evaluate and compare the productive catchments of eight tree-ring dated sites
and to suggest the climatic and productive
conditions under which habitation sites are
established and maintained.
While each of these explorations of the re-
suiting database was used to address different
archaeological questions, it is clear from all of
them that productivity varied considerably
from placetoplace and from yeartoyear
within the study area, but also that there was
always enough productive land to produce
enough maize to support a very large population (for example, 31,363 persons at a density of 21 persons/km2 in the 1470.36 km 2
study area over the 400year period), even in
the relatively dry times of the middle twelfth
and late thirteenth centuries. If mobility and
access to productive land were not restricted, or
if redistribution systems were in place to support dispersed populations or uneven production, then the productive environment
always could have sustained many people, even
during the socalled Great Drought of A.D.
12761299. If, however, mobility and access to
productive resources were severely restricted,
and extensive extracommunity food sharing
was not regularly practiced, then there would
have been times when some populations who
were confined to living in certain places were,
by virtue of their size, characterized by a demand for maize that was not met by local
annual supply. Nevertheless, it is important to
emphasize that there were always locations
somewhere within the study area that could
produce adequate maize crops, and at no time
was the "potential dry farming belt" (petersen
1988) completely pinched out by climatic
fluctuations. In other words, climatic fluctuations as they affected crop production in and
of themselves cannot be used as the sole and
sufficient cause for the total abandonment of
Northern San Juan Region at the end of the
thirteenth century.
. tern of annual growth discerned in treering
data. Further examination of the patterning is
necessary before this correlation can be ascertained, but if this were true it may imply that
the prehistoric Pueblo farmers of the Mesa
Verde area were not willing to work hard for
only very minimal returns, even though they
could have met their annual requirements by
working more land of lower productive potential. Instead of turning to less productive soils,
it is possible that they moved to places where
high productivity was more predictable, which
occasioned both local abandonments and relocations within the region, and ultimately the
regional abandonment of the Northern San
Juan District. Should this type of cultural selection for only the best lands have occurred,
then estimates of total maize productivity
should only be drawn from the highest yield
categories. This would, of course, result in lower population estimates and carrying capacity
values than the preceding study has indicated.
An alternative to this possibility is that
environmental resources other than agricultural
productivity were the limiting factors in sustaining a large population in the study area
Critical environmental resources that have been
suggested elsewhere as potentially limiting
include potable water shortages (Herold 1961),
woodresource depletion (Kohler and
Matthews 1988), soil nutrient depletion in
pinyonjuniper woodland zones (Matson et al.
1988, Kohler 1992), and animal protein
deficiency (Speth and Scott 1989). Whereas
the shortage of drinking water can be linked to
meteorological conditions, shortages of lumber
and fuel, shortages of animal protein, and
depletion of soil nutrients are probably more
closely linked to human overuse and poor
management practices than limits imposed by
the natural environment. Should one or several
of these factors be the limiting conditions for
sustaining large number of people, then the
population estimates currently generated by
the model will exceed the true maximum value.
The implications of the results are several,
given the validity of the model for reconstructing prehistoric agricultural productivity. If
Anasazi populations were aware of the differential productivity of places on the landscape
and selected for those locations that would
consistently produce good yields of maize, as
the site catcmnent studies show, then it might
indicate that populations only considered the
most arable soils (Le., soils that would be classified as moderatetohigh and high yield by
the model) as worth farming. Tantalizing hints
of this possibility exist in the data generated by
the model; the annual contribution of moderatetohigh yield soils seems to mimic the pat
Yet another possibility is that nonenvironmental factors (i.e., social or cultural factors)
are responsible for the ultimate abandonment
of the region. It is beyond the scope of this
study to explore what some of these nonenvironmental factors might be (see, e.g., Adams
1991). It is likely, however, that some
188
combination of factors is responsible for the
depopulation of the region in the late thirteenth century. Although this study does not
identify what factor or combination of factors
led to the final abandonment, it does suggest
that some previous assumptions about maize
production cannot be used to account for this
event.
Second, the model is quantifiable, which
allows other potential users to judge the appropriateness of the variables and values used and
permits modification where these values are
found to be inappropriate or inaccurate. In this
regard, the model gains strength from using
the most current data on the distribution and
quality of soils in the study area, newly available digital elevation data, and newly compiled
information on soil productivity.
STRENGTHS OF THE MODEL
The tree-ring-based model of agricultural
productivity and sustainable population described in the preceding chapters derives
strength from at least four sources. First and
foremost, it is a high resolution model. Treering data provide annual information about
regional patterns of precipitation and temperature that are made locally relevant by application of Palmer Drought Severity Indices
(PDSI), which have been calculated for local
soils and local weather stations. Further, it is
recognized that tree-ring data provide the
highest temporal resolution reconstructions of
past environmental conditions (Dean 1988b),
that PDSI are integrative measures of precipitation and temperature, which incorporate the
effects of previous values on current soil moisture, and that retrodictions of PDSI produce
dendroclimatological reconstructions with the
highest explained variance of all reconstructed
variables (Rose et al. 1982). My use of multiple local weather stations to model the local
influences of elevation on precipitation and
temperature patterns, and my use of multiple
classes of soil moisture rather than just one as
suggested in the original method (palmer
1965), is new. As such, it helps to translate
regional climate into local terms. A cell size of
4 ha (200 x 200 m) contributes to the model's
high spatial resolution, particularly given the
size of the study area (1,816 km 2 or 701 mi 2).
which required recording values for a minimum of 45,400 cells (200 rows x 227 columns). While other researchers in the Mesa
Verde Region have conducted productivity
studies at this scale of spatial resolution before.
albeit not for such an extensive study area
(e.g., Cordell 1975; Darsie 1982; Kohler et al.
1986), no one before has created such a high
resolution. locally pertinent record of climatic
variability as it affects agricultural production.
Third, the implementation of the model by
means of geographic information system technology permits the manipulation of large data
sets and the visual display of results in a form
that can be easily grasped and also numerically
expressed. Thus. GIS serves as a management
device, as well as an analytic tool, and is ideally
suited to this kind of research. Conceptually,
the maps produced by the GIS are similar to
those produced by Dean and Robinson (1977,
1979) by simpler computer cartographic techniques, in that they both depict changing dendroclimatic values across a study area for a
lengthy period of time. The maps produced in
this study, however, are more useful for local
analysis than the contour maps of the northern
Southwest produced by Dean and Robinson
because they are annual, not decadal, and because they translate regional climatic patterns
.into a measure of local agricultural productivity that is more interpretable from a human behavioral point of view.
Fourth, the population estimates are deliberately conservative; I consciously tried to underestimate rather than overestimate the number of people and their average per km density
that could be supported in any place and at
any time during the 400 years. In part this is
due to my attempt to represent the behavior of
a population that annually grows and stores
two years of maize. This level of agricultural
demand is repeated ever year, regardless of
how much food may be left in storage facilities. Thus, the demand is maintained constantly and results in lower population estimates than if carryover in storage from year to
year had been allOWed. Furthermore, the population density estimates are based on the total
number of persons divided by the total area of
land (e.g., the full 1,470.36 km 2 study area or
the full area included in the localities or site
189 -
catchments) after the yields from the lowest
producing lands have been removed from
consideration. This results in lower population
density values than would have been derived if
only the actual but varying area of arable land
(some value less than 1470.36 km 2 in the
study area, for example) had been used to calculate total arumal productivity.
LIMITATIONS OF THE MODEL
At least three sources of potential weakness
can be identified. One pertains to the method
of paleoclimatic reconstruction selected. The
second pertains to data quality (inherent error
and operator error), and the final pertains to
potentially important variables not selected for
inclusion in the model.
Just as the model gains strength by using
tree-ring data to reconstruct high frequency
paleoclimatic variation, it is limited by the same
factors that limit all tree-ring-based reconstructions. Dean (l988b:132-135) reviews
these advantages and shortcomings. Perhaps
the most unfortunate and unavoidable of these
is that there are years when growing conditions
are so limiting than very narrow rings or even
missing rings (indicating no growth) are produced. This can result in the underestimation
of the true severity of the climate (especially
the accurate estimate of precipitation) for that
year. This limitation is partially mitigated by
using chronologies where the specimen depth
(the number of tree-ring samples) is adequate
and by using multiple chronologies, as in this
study (Le., SWOLD7), where unevenness in the
frequency of samples associated with each individual chronology is partially evened out by
using indices from other chronologies. Similarly, tree-ring records are better at predicting
low precipitation values than high precipitation
values because more growth is attributed to
non-climatic variables when the tree is not
water-stressed. This means that reconstructions
at lower elevations, which generally receive less
precipitation and are warmer, are generally
better than reconstructions for the highest
elevations where more precipitation and cooler
temperatures are typical. This dendroclimatological fact was observed in the varying r 2
values associated with the final full-period
calibration equations for the 55 different re-
constructions (Table 3.3). Fortunately, however, only 2.4% of the land in the study area
fell into the range of elevations modeled by the
highest weather station (Ft Lewis), minimizing
the impact of this inherent problem. Nevertheless, it is true that no dendroclimatic reconstruction, whether from relatively low elevations on the Colorado Plateau or not, ever retrodicts 100% of the variance observed in the
reconstructed variable, because a perfect correlation between tree-ring growth and climate
does not exist. Finally, it must be said, that treering chronologies are poor estimators of low
frequency climatic variation due to the methods used to standardize ring-width values,
which is the kind of longperiodicity variation
that alluvial geomorphology and palynology
can detect. This method-dependent deficiency
to monitor low frequency variability is not
particularly problematic for this study, which
instead focuses on the high frequency variation
in potential annual dry-farming productivity.
However, it would be a greater problem if the
focus were on floodplain agriculture, for
example.
Second, the accuracy of the model and its
results are limited by data quality. The accuracy of the reconstructions of PDSI, the estimates of bean and com yield, and the estimates
.of population are dependent on the quality of
the primary data described in chapter 2. If the
raw data amassed by other agencies or individuals are inherently inaccurate (e.g., if soils were
mapped incorrectly by Soil Conservation Service), estimates for bean yields were markedly
underestimated or overestimated, the weather
stations selected to represent the local expression of regional climate were not truly representative, or the Digital Elevation Models (the
base maps used for this study) were inaccurate,
then the resulting manipulations of these data
would also be inaccurate. Similarly, if I recorded these data improperly (e.g., errors in judgment, transcription errors, encoding errors, or
miscalculations) or overgeneralized certain
variables by virtue of the 200 x 200 m cell size
resolution (e.g., dominant soil type), then the
resulting estimates will be equally compromised. While care to record and calculate
values was exercised throughout this study
(and on occasion monitored, as with the recording of soil map units from aerial imagery),
it was not possible to control for inherent error
190 -
in the primary data sources, other than acquiring official and presumably valid data kept by
federal agencies. For the present, these data,
particularly the soil distribution data, are taken
as given, with the acknowledgment that more
refined mapping may be necessary, especially
for canyon topography within the study area
While this is not seen as a major drawback in
the search for patterning at the larger geographic scales of the study area and survey
localities, it potentially is a limitation for the
smallerscale site catchment investigations. This
is so because small patches of arable land may
have been missed by soil scientists or ignored
by me in the process of determining the dominant soil map unit per 4ha cell if they covered less than half that area.
Third, the ability of the model to accurately depict past envirorunental conditions and
their influence on agricultural production may
be reduced by noninclusion of all the important variables. Certain envirorunental phenomenon that are known or have been suspected by
some to be relevant to successful prehistoric
maize harvests have not been incorporated in
this model. The infestation of crops by insects,
molds, fungi, and other parasites that can decrease yields or even totally destroy crops has
not been included in the model, per se, although a percentage of crop loss is withdrawn
from the potential annual yield to partially
simulate this problem. Regular depletion of
soil nutrients and largescale erosion of sediments that remove soils and farmland have not
been built into the model, although these are
variables that might be considered in future
versions of the model. However, at this time
there is no evidence to indicate that nutrient
depletion or major soil loss in the study area
was a critical factor in diminishing production.
Further, while degree of slope has been incorporated in the soil data, aspect, or the direction
of slope, has not. This is a variable that could
easily be incorporated into future versions of
the model, particularly because of the GIS capability to calculate aspect from elevational
data, but only if it was found to significantly
affect yield figures. Given the dominant south
and southwest aspect of the terrain in the study
area as a whole, it was not considered as a major discriminating variable.
Finally, the pooling effects of cold air
. drainage in canyons and within the troughs of
low hills and small upland basins (Adams
1979) has only partially been considered in the
model. The significance of cold air drainage
into topographic lows is that it reduces temperature and shortens the length of the growing season for plants. While this phenomenon
was emphasized in the studies by the Dolores
Archaeological Project, given the centrality of
the Dolores River and its tributaries, the present
study is primarily upland and not canyon or
lowland topography. In addition, the reduced
productivity of shallow soils, soils on steep
slopes, and soils within canyons, is reflected in
reduced natural potential productivity values,
reduced crop yield values or low available
waterholding capacity values associated with
such places. These locations ultimately were included in the lowest yield category and effectively were removed from contributing (or only
occasionally contributed) to the estimate for
total annual productivity (TOTPROD). Therefore, the effect of cold air drainage is indirectly
taken into consideration in the model.
RECOMMENDATIONS FOR RJTURE
RESEARCH
Future promising work should include tests
. for the sensitivity of the current model to
changes in values, thereby assessing its robustness; refinement of the model through improved methods of estimation and more detailed (or additional) data; and continued exploration of the archaeological record with the
extant database.
Perhaps the most obvious study is a thorough examination of the values used in the
current model to assess the strength or robustness of the results. Through repeated runs of
the model, where values used to represent
critical variables are changed one at a time, it
will be possible to determine how great or how
small a change is required to substantially alter
the basic conclusion of the study. Given that
the agricultural productivity of the study area
is always high enough to support a minimum
of 21 persons/km2 and that this is the more
conservative POP2YR value, I believe that the
model is quite robust In other words, the
model produces results that are solid and will
not wildly fluctuate given small changes in
191 -
critical variables. Nevertheless, sensitivity
analysis should be done to add confidence to
the conclusions reached in this research.
Refinements to the basic model, once tested
and confirmed acceptable, could result from
experimentation with the use of different values, which might lead to higher levels of variance explained in the reconstructed environmental variables. For example, instead of using
the instrumented June PDSI values, which depict the state of soil moisture conditions as they
exist on July 1, July PDSI values, which depict
conditions on August 1, might produce higher
correlations between tree-ring data and PDSI
values and better reflect water stress to maize
plants during a critical stage of their development. My reconstructions are based on June
PDSI values. The variance (r2 ) in June PDSI
values explained by the tree-ring data is very
similar to variance explained for July PDSI
values derived by Rose et al. 1982 for the
southwestern Colorado and by Rose for the
Zuni area (personal communication, April
1989). Further, my June PDSI reconstructions
closely mirror qualitative commentary drawn
from historic sources for the study area Nevertheless, it would be useful to quantify the different results attributed to the use of one
month or the other, although I suspect the variance explained will be quite similar given high
autocorrelation from one month to the next.
Experimentation with different values
related to agricultural demand, as well as to
agricultural supply, could lead to an appreciation for the tolerances of the resource-topopulation relationship. For example, it would
be instructive to reduce the portion of the total
land area potentially available to produce
crops, perhaps incrementally through time, to
simulate what might be the carrying capacity
ramifications of an increasingly deteriorating
environment for agriculture. Alternatively,
crop yield values could be modified to simulate lower prehistoric yields than those currently
estimated, particularly if experimental garden
research using native seed could demonstrate
markedly different production values.
Improvements to the database are always
warranted, and that is particularly true for the
mapping of soils. A more detailed survey to
identify and chart the location of distinct soil
. types and measure natural and agriCUltural
productivity as a check on the values used in
this study is recommended. This would be useful in locations known to be characterized by
significant archaeological remains so as to
compare the distribution of soils of variable
quality to the distribution of permanent habitations and seasonal field houses.
The recalculation of PDSI values should be
accomplished at some future date when treering chronologies have been extended to incorporate more years beyond the current upper limit of A.D. 1970. It would be ideal if
extremely dry years in southwestern Colorado
(such as 1989, a year drier than recorded historically by any of the five weather stations
used in this study) could be incorporated into
the instrumented PDSI calculations. If more
extremes in climatic fluctuation could be incorporated into the calibration data set, better
long-term reconstructions would result.
Finally, additional explorations of the
archaeological record using these data can be
envisioned. If archaeological sites dated by
surface ceramic assemblages can be better
dated than is currently the practice and assigned calendrical dates rather than portions of
the Pecos Classification system (e.g., A.D.
. 1050-1075 versus late "Pueblo II"), then it
should be possible to closely examine the
distribution of settlements of various time periods with respect to the distribution of productive land. This would be done in much the same
way that I examined the tree-ring-dated archaeological sites. Should the pattern of site establishment, site use, and site abandonment be
found to reflect the temporal and spatial characteristics of productive agricultural land, as
suggested by the locality and site catchment
studies reported herein, then it will be possible
to predict similar patterns for other areas and
other places in the Anasazi world. Further,
when clusters of settlements can be associated
as components of prehistoric communities and
the spatial distribution of various prehistoric
communities can be identified, then it may be
possible to infer which communities coalesced
when and where on the basis of access to and
control of productive lands. Kohler (1989) and
Adler (1990) have suggested recently that
competition for and control of productive land
may be the key in the formation of
192 -
communities and in aggregation. The data
contained in this model may provide the backgroWld infonnation to test this idea. In this
way, it may be possible to reconstruct the territories of contemporary commWlities and gain
insight into the dynamics of intra- and intercommunity cooperation or strife.
The maps of annual productivity should be
examined closely to identify the best patches
of agricultural land in those years or periods
when productivity is low or below the longterm conditions for the study area. By this
method, the most optimal areas for settlement
in stressful times can be established and archaeological survey can be conducted to test the
proposal that productive and predictable land
was indeed a significant draw in stressful times.
In order to further explore the utility of the
data, the model could be applied to the well
studied and nearby Wetherill Mesa, for which
extensive survey has been accomplished
-
. (Hayes 1964) and where simulations of site
establishment and abandonment have already
been conducted (Cordell 1975). Comparison
of the estimates of population and patterns of
settlement should be interesting and informative.
It may also be possible to refine the dating
of poorly dated sites by suggesting the most
likely occupational intervals based on the reconstruction of the temporal variability in agricultural productivity in the vicinity of those
sites.
In conclusion, it is probably fair to say that
this study provides a defensible approach to
modeling prehistoric agricultural productivity,
one that can be tested, refined, and augmented
by future research. As such, it offers a solid
base upon which subsequent work may be built
and may bring a sense of methodological order to studies of prehistoric settlements and
human communities.
193-
Page Blank in Original
BIBLIOGRAPHY
Adams,E. C.
1979 Cold Air Drainage and Length of Growing
Season in the Hopi Mesas Area In The Kiva
44:285296.
1984 Preliminary Report of Work Accomplished
During 1983 Under State Permit Nos. 83-8
and 839. Crow Canyon Archaeological
Center. Submitted to the Office of the State
Archaeologist. Colorado Historical Society.
Denver.
1985 Annual Report of Test Excavations at
5M1765, Sand Canyon Pueblo, and
Archaeological Survey in T36N, R18W,
Sections 12, and 24, and T36N, R16W,
Sections 29 and 30. Crow Canyon
Archaeological Center. Submitted to the Office
of the State Archaeologist. Colorado Historical
Society, Denver.
1991 The Origin and Development of the Pueblo
Katsina Cult. The University of Arizona
Press, Tucson.
Adams, K. R.
1980 Relative Numbers of Native Macrofossils
in Strata of Poor Preservation, with Emphasis
on Flotation. In Pan 7: Salmon Ruin
Ethnobotanical Report, by V. L. Bohrer. In
investigations at the Salmon Site: the
Structure of Chacoan Society in the Northern
Southwest ill, edited by C. IrwinWilliams and
P. H. Shelley, pp. 251301. Eastern New
Mexico University, Portales.
Adler, M. A.
1988 Archaeological Survey and Testing in the
Sand Canyon PueblolGoodmon Point Ruin
Locality, Montezuma County, Colorado, 1987
Field Season. Crow Canyon Archaeological
Center. Submitted to the Bureau of Land
Management. San Juan Resource Area,
Durango and Office of the State Archaeologist,
Colorado Historical Society, Denver.
1990 Population Aggregation and the Anasazi
Social Landscape: The View From the Four
Comers. Paper presented at the 1990
Southwest Symposium, Albuquerque.
Adler, M. A., and M. Metcalf
1991 Draft Report on Archaeological Survey of
Lower East Rock and Sand Canyons,
Montezuma County, Colorado. Crow Canyon
Archaeological Center, Cortez, Colorado.
Report submitted to the Bureau of Land
Management, San Juan Resource Area Office,
Durango, Colorado.
Ahlstrom. R. V. N.
1985 The Interpretation ofArchaeological TreeRing Dates. Ph.D. dissertation, University of
Arizona, University Microfllms, Ann Arbor.
Allen, K. M. S., S. W. Green, and E. B. W. Zubrow
1990 Interpreting Space: GIS in Archeology.
Taylor and Francis, London.
Arrhenius, G., and E. Bonatti
1965 The Mesa Verde Loess. In Contributions of
the Wetherill Mesa Archaeological Project. assembled by D. Osborne, pp. 92-101. Memoirs
of the Society for American Archaeology No.
19. Washington, D.C.
Aronoff, S.
1989 Geographic Information Systems: A
Management Perspective. WDL Publications,
Ottawa
Avery, T. E., and G. L. Berlin
1985 Interpretation ofAerial Photographs. 4th
ed. Burgess, Mirmeapolis.
Bailey, R., S. Hackenberger, D. Howes, and D.
Wherry
1985 Geographic Information Processing in Land
Use Modeling and Testing in the Columbia
River Basin. Paper presented at the 50th
Annual Meeting of the Society for American
Archaeology, Denver.
Beaglehole, E.
1937 Notes on Hopi Economic Life. Yale
University Publications in Anthropology 15.
New Haven.
Berry,J. K.
1989a Beyond Mapping. GIS World 2(3):8, 3839.
1989b Beyond Mapping. GiS World 2(4):34,
47-48.
Betancourt, J. L., and O. K. Davis
1984 Packrat Middens from Canyon de Chelly,
Northeastern Arizona: Paleoecological and
Archaeological Implications. Quaternary
Research 21:56-64.
Ones. 2d ed., edited by J. H. Berger and E. F.
Berger, pp. 122-127. Interdisciplinary
Supplemental Educational Programs, Cortez,
Colorado.
Betancourt, J. L., and T. R. Van Devender
1981 Holocene Vegetation in Chaco Canyon,
New Mexico. Science 214:656-658.
1986 1985 Annual Report of Test Excavations al
Sand Canyon Pueblo (5M1765). Crow
Canyon Archaeological Center, Cortez,
Colorado. Submitted to Bureau of Land
Management, Durango, Colorado.
Blasing, T. J.
1975 Methods for Analyzing Climatic Variations
in the Nonh Pacific Sector and Western North
America for the Last Few Centuries. Ph. D.
dissertation, University of Wisconsin,
University Microfilms, Ann Arbor.
1987 Annual Report of Excavations at Sand
Canyon Pueblo (5M1765). Montezuma
County. Colorado. 1986 Field Season. Crow
Canyon Archaeological Center, Cortez
Colorado. Submitted to Bureau of Land
Management, Durango, Colorado.
Bohrer, V. L.
1980 Part 7: Salmon Ruin Ethnobotanical
Report. In Investigations at the Salmon Site:
The Structure of Chacoan Society in the
Northern Southwest Jl/. edited by C. IrwinWilliams and P. H. Shelley, pp. 163-535.
Eastern New Mexico University, Portales.
1988a Annual Report on the Excavations at Sand
Canyon Pueblo, 1987 Field Season. Crow
Canyon Archaeological Center, Cortez,
Colorado. Submitted to Bureau of Land
Management, Durango, Colorado.
1981 Methods of Recognizing Cultural Activity
From Pollen in Archaeological sites. The Kiva
46:135-142.
1988b Wallace Ruin Interim Report. Southwestern Lore 54(2):8-33.
Bohrer, V. L., and K. R. Adams
1977 Ethnobotanical Techniques and Approaches
at Salmon Ruin. New Mexico. San Juan
Valley Archaeological Project, Technical
Series No.2. Contributions in Anthropology,
Vol. 8, No.1. Eastern New Mexico
University, Portales.
Bolton, H. E.
1972 Pageant in the Wilderness: The Story of
the Escalante Expedition to the Interior Basin.
1776. Utah Slate Historical Society, Salt Lake
City.
Bracken, I., G. Higgs, D. Martin, and C. Webster
1989 A Classification of Geographical Systems
Literature and Applications. Environmental
Publications, University of East Anglia,
Norwich.
Bradfield, M
1971 The Changing Pattern ofHopi AgriCUlture.
Royal Anthropological Institute Occasional
Paper 30. London.
Bradley, B. A.
1974 Preliminary Report of Excavations at
Wallace Ruin, 1969-1974. Southwestern Lore
40(3 & 4):63-71.
1984 The Wallace Ruin and the Chaco
Phenomenon. In 1nsighJs 1nto the Ancient
-
1992 Excavations at Sand Canyon Pueblo. In
The Sand Canyon Archaeological Project: A
Progress Report, edited by W. D. Lipe, pp.
79-99. Occasional Paper No.2. Crow Canyon
Archaeological Cemer, Cortez, Colorado.
. Brandegee, T. S.
1976 The Flora of Southwestern Colorado. The
United States Geological and Geographical
Survey Bulletin 3(3):227-248.
Breternitz, D. A.
1986 Notes on Early Basketmaker ill sites in
Mancos Canyon, Colorado. The Kiva 51:263264.
Breternitz, D. A., and D. W. Martin
1973 Report of the Dolores River Project
Archaeological Reconnaissance, 1972-1973.
Ms. on file, Midwest Archaeological Center,
National Pm Service, Lincoln, Nebraska
Bretemitz, D. A., C. K. Robinson, and E. T. Gross
(compilers)
1986 Dolores Archaeological Program: Final
Synthetic Report. Bureau of Reclamation,
Engineering and Research Center, Denver.
Brew,J. O.
1946 The Archaeology of Alkali Ridge,
Southeastern Utah. Papers of the Peabody
196-
Museum of American Archaeology and
Ethnology No. 21. Harvard University,
Cambridge, Massachusetts.
Brisbin, J. M., and C. J. Brisbin
1973 Contribution to the Sites Being Excavated
in the North McElmo Drainage Area of
Southwestern Colorado. Ms. on file, Crow
Canyon Archaeological Center, Cortez,
Colorado.
Briuer, F. L.
1988 Geographic Information Systems and
Automated Archaeological Information:
Investments and Dividends for Archaeological
Resource Management Paper presented at the
53rd Annual Meeting of the Society for
American Archaeology, Phoenix.
Brown, J.
1975 Preliminary Report on the 1975
Excavations at Site 5MT3 Near Yellowjacket,
Colorado. Southwestern Lore 41(4):47-50.
Brown,L. A.
1938 A Basisfor Rating the Productivity of
Soils on the Plains ofEastern Colorado.
Colorado Experimental Station Technical
Bulletin 25. Colorado State College, Ft.
Collins.
Bullard, W. R.
1962 The Cerro Colorado Site and Pithouse
Architecture in the Southwestern United States
Prior to A.D. 900. Papers of the Peabody
Museum of Archaeology and Ethnology No.
44. Harvard University, Cambridge,
Massachusetts.
Bums,B. T.
1976 The "Abandonment Controversy" and a
Proposal for its Partial Resolution. Preliminary Exam Paper, Department of Anthropology, University of Arizona. Ms. on file,
Arizona State Museum Library, University of
Arizona, Tucson.
1983 Simulated Anasazi Storage Behavior Using
Crop Yields Reconstructed From Tree-Rings:
A.D. 652-1968. 2 vols. Ph.D. dissertation,
University of Arizona, University Microfilms,
Ann Arbor.
Burrough, P. A.
1986 Principles of Geographic Information
Systems for Land Resource Assessment.
Oxford University Press, Oxford.
Calamia, M. A.
1986 Geographic Information System Applicationsfor Cultural Resource Management.
Bureau of Land Management, Denver.
Brown, P. E., and B. H. Rubin
1982 Patterns of Desert Resource Use: An
Integrated Approach to Settlement Analysis. In
Granite Reef" A Study in Desert Archaeology,
edited by P. E. Brown and C. L. Stone, pp.
267305. Arizona State University
Anthropological Research Papers No. 28.
Tempe.
Bryan,K.
1925 Dates of Channel Trenching (ArroyoCutting) in the Arid Southwest Science
62:338-344.
1941 Pre-Columbian Agriculture in the
Southwest, as Conditioned by Periods of
Alluviation. Annals of the Association of
American Geographers 31:219--242.
1954 The Geology of Chaco Canyon. New
Mexico. in Relation to the Life and Remains
of the Prehistoric Peoples ofPueblo Bonito.
Smithsonian Miscellaneous Collections, Vol.
122, No.7. Smithsonian Institution,
Washington, D.C.
-
Carmichael, D. L., and D. Christenson
1988 Predictive Modeling of Prehistoric Site
Distribution in Central Montana. Paper presented at the 53rd Annual Meeting of the
Society for American Archaeology, Denver.
CASA
1981 Cultural Resource Inventory of the Shell
Mobile C02 Wet Gathering System,
Montezuma and Dolores Counties. Colorado.
Complete Archaeological Services Associates,
Cortez, Colorado. Submitted to Bureau of
Land Management, San Juan Resource Area.
Durango.
Cattanach, G. S., Jr.
1980 Long House. Mesa Verde National Park.
Colorado. Publications in Archeology NO.7-H.
National Park Service, Washington, D.C.
Chandler, S. M., A. D. Reed, and P. R. Nickens
1980 Ecological Variability and Archaeological
Site Location in Southwestern Colorado: The
Class II Cultural Resource Inventory of the
Bureau ofLand Management's Sacred
Mountain Planning Unit. Centuries Research,
197-
Montrose, Colorado. Submitted to the Bureau
of Land Management, Montrose District
Office, Montrose, Colorado.
Clarke, K. C.
1986 Advances in Geographic Information
Systems. Computers, Environment and Urban
Systems 10:175-184.
Collins, S. H., and G. C. Moon
1981 Algorithms for Dense Digital Terrain
Models. Photogrammetric Engineering and
Remote Sensing 47:71-76.
Connolly, M.
1990 A Historic Land Use Study of Goodman
Point. Ms. on file, Crow Canyon Archaeological Center, Cortez, Colorado.
Creamer, W.
1985 The Upper Klethla Valley: Computer
Generated Maps of Site Location. Paper presented at the 50th Annual Meeting of the
Society for American Archaeology, Denver.
Darsie, R F., III
1982 Prehistoric Agricultural Potential and Site
Location in the Dolores River Valley, Southwest Colorado. Unpublished M.A. thesis,
Department of Anthropology, Washington
State University, Pullman.
Dean, J. S.
1988a A Model of Anasazi Behavioral Adaptation. In The Anasazi in a Chonging Environment, edited by G. J. Gumerman, pp. 25-44.
Cambridge University Press, Cambridge.
1988b Dendrochronological and Paleoenvironmental Reconstruction on the Colorado
Plateaus. In The Anasazi in a Chonging
Environment, edited by G. J. Gumerman, pp.
119-167. Cambridge University Press,
Cambridge.
Cook, E. R
1982 A Long-Term Drought Sequence for the
Hudson Valley, New York. In Climate from
Tree Rings, edited by M. K. Hughes, P. M.
Kelly, J. R Pilcher, and V. C. LaMarche, Jr.,
pp. 163-165. Cambridge University Press,
Cambridge.
Cook, E. R, and G. C. Jacoby
1977 Tree Ring Drought Relationships in the
Hudson Valley, New York. Science 198:399401.
Cook, S. F., and W. Borah
1979 Essays in Population History: Mexico and
California. University of California Press,
Berkeley.
Dean, J. S., and W. J. Robinson
1977 Dendroclimatic Variability in the American
Southwest A.D. 680 to 1970. Laboratory of
Tree-Ring Research, University of Arizona
Submitted to the National Park Service, Interagency Archeological Services, Denver. U.S.
Department of Commerce, National Technical
Information Service PB-266 340.
1978 Expanded Tree Ring Chronologiesfor the
Southwestern United States. Chronology
Series 3. Laboratory of Tree-Ring Research,
The University of Arizona, Tucson.
Cordell. L. S.
1975 Predicting Site Abandonment at Wetherill
Mesa. The Kiva 40:189-202.
1979 Computer Cartography and the Reconstruction of Dendroclimatic Variability in the
American Southwest, A.D. 680 to 1970. In
Statistical Cartographic Applications to
Spatial Analyses in Archoeological Contexts,
edited by S. Upham, pp. 79-94. Anthropological Research Papers No. 15. Arizona State
University, Tempe.
1984 Prehistory of the Southwest. Academic
Press, New York.
Cole, S. J.
1989 Iconography and Symbolism in Basketmaker Rock Art. In Rock Art of the Western
Canyons, edited by J. S. Day, P. D. Friedman,
and M J. Tate, pp. 59--85. Colorado Archaeological Society Memoir No.3. Denver.
Cowen, D. J.
1988 GIS versus CAD versus DBMS: What Are
the Differences? Photogrammetric Engineering
and Remote Sensing 54:1551-1555.
Dean, J. S., R C. Euler, G. J., Gumerman, F. Plog,
R H. Hevly and T. N. V. Karlstrom
1985 Human Behavior, Demography, and Paleoenvironment on the Colorado Plateaus. American Antiquity 50:537-554.
Debloois, E. I., and D. F. Green
1978 SARG Research on the Elk Ridge Project,
Manti-La Sal National Forest, Utah. In
-
198-
Investigations of the Southwestern Anthropological Research Group: The Proceedings of
the 1976 Conference, edited by R. C. Euler
and G. J. Gumennan, pp. 1324. Museum of
Northern Arizona, Flagstaff.
Douglass, A. E.
1929 The Secret of the Southwest Solved by
Talkative TreeRings. National Geographic
Magazine 54:737-770.
Douthit, M. L.
1984 Excavations at Site 5MT3777 and Site
5MT3778, Casa de Suenos. Technical Report,
August 21,1984. Bureau of Land Management, Anasazi Heritage Center, Dolores,
Colorado.
Dueker, K. J.
1987 Geographic Infonnation Systems and
Computer-Aided Mapping. Journal of the
American Planning Association 53:383-390.
Duff, U. Francis
1904 Some Exploded Theories Concerning
Southwestern Archeology. American Anthropologist [n.s.] 6:303-306.
U.S.G.S. National Cartographic Infonnation
Center, Reston, Virginia.
Erdman, J. A., C. L. Douglas, and J. W. Marr
1969 Environment of the Mesa Verde, Colorado.
Archeological Research Series No. 7-B.
National Park Service, Washington, D. C.
Euler, R. C., G. J. Gumennan, T. N. V. Karlstrom,
J. S. Dean, and R. H. Hevly
1979 The Colorado Plateaus: Cultural Dynamics
and Paleoenvironment. Science 205:10891101.
Farley, J., W. F. Limp, and J. Lockhart
1990 Theoretician's Workbench: Integrating
GIS, Remote Sensing, EDA, and Database
Management In Interpreting Space: GIS and
Archaeology, pp. 141-164. Taylor and
Francis, London.
Fettennan, J., and L. Honeycutt
1980 Cultural Resource Inventory, MAPCO's
Liquid Hydrocarbons Pipeline. Southwestern
Colorado. Bmeau of Anthropological Research, Dove Creek, Colorado. Submitted to
Woodward-Clyde Consultants, San Francisco.
1982a Cultural Resources Inventory, Shell C02
Laterals and Shell Oil Flowlines. Bureau of
Anthropological Research. Submitted to
Woodward-Clyde Consultants, San Francisco.
Dykeman, D. D.
1986 Excavations at 5MT8371 , An Isolated
Pueblo II Pithouse Structure in Montezuma
County, Colorado. Studies in Archaeology
NO.2. Division of Conservation Archaeology,
San Juan County Museum Association,
Fannington, New Mexico.
1982b Testing and Excavation Report,
MAPCO's Rocky Mountain Liquid Hydrocarbons Pipeline, Southwest Colorado. Bureau
of Anthropological Research, Dove Creek,
Colorado. Submitted to Woodward-Clyde
Consultants, San Francisco.
EddY,F. W.
1966 Prehistory of the Navajo Reservoir District, Northwestern New Mexico. Museum of
New Mexico Papers in Anthropology Vo1.15,
Nos. 1 and 2. Santa Fe.
Eddy, F. W., A. E. Kane, and P. R. Nickens
1984 Southwest Colorado Prehistoric Context:
Archaeological Background and Research
Directions. Office of Archaeology and Historic
Preservation, Colorado Historical Society,
Denver.
Ekren, E. B., and F. N. Houser
1965 Geology and Petrology of the Ute Mountains Area Colorado. Geological Survey
Professional Paper 481. U.S. Government
Printing Office, Washington, D.C.
Elassal, A. A., and V. M. Caruso
1983 Digital Elevation Models. U.S. Geological
Survey Circular 895-B. User Services Section,
-
1987 The Mockingbird Mesa Survey. Cultural
Resource Series No. 22. Bureau of Land Management, Denver, Colorado.
Fewkes, J. W.
1919 Prehistoric Villages, Castles and Towers of
Southwestern Colorado. Bmeau of American
Ethnology Bulletin No. 70. Washington, D.C.
Flint, P. R., and S. W. Neusius
1987 Cottontail Procurement among Dolores
Anasazi. In Dolores Archaeological Program:
Supporting Studies: Settlement and
Environment, compiled by K. L. Petersen and
J. D. Orcutt, pp. 257-273. Bureau of
Reclamation, Engineering and Research
Center, Denver.
199-
Ford, R. I.
1968 An Ecological Analysis Involving the
Population of San Juan Pueblo. New Mexico.
Ph.D. dissertation, University of Michigan.
University Microfilms, Ann Arbor.
1987 Cultural Resource Inventoriesfor the
Dolores Project: The Dove Creek Canal
Distribution System and Dawson Draw
Reservoir. Four Comers Archaeological
Project Report NO.7. Complete Archaeological Service Associates, Cortez, Colorado.
Submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City.
Forde, C. D.
1931 Hopi Agriculture and Land Ownership.
Journal of Royal Anthropological Institute
61:357406.
1988 Cultural Resource Inventoriesfor the
Dolores Project: The Ute Irrigated Lands
Survey. Four Comers Archaeological Project
Report No. 13. Complete Archaeological
Service Associates, Cortez, Colorado. Submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City.
Forrest, E., G. E. Montgomery, and G. M. Juhl
1990 Intelligent Infrastructure Workbook: A
Management-Level Primer on GIS. AEC
Automation Newsletter, Fountain Hills,
Arizona.
Freeman, I. S.
1958 A History of Montezuma County Colorado. Johnson Publishing, Boulder, Colorado.
Fritts, H. C.
1965 TreeRing Evidence for Climatic Changes
in Western North America. Monthly Weather
Review 93:421-443.
1976 Tree Rings and Climate. Academic Press,
London.
Fritts, H. C., T. J. Blasing, B. P. Hayden, and J. E.
KUlZbach
1971 Multivariate Techniques for Specifying
Tree Growth and Climatic Relations and for
Reconstructing Anomalies in Paleoclimate.
Journal of Applied Meteorology 10:84--864.
Gill, S. J., and D. Howes
1985 A GIS Approach to the Use of Surface
Samples in Intra-Site Distributional Analysis.
Paper presented at the International Symposiwn on Data Management and Mathematical
Methods in Archaeology and International
Union of Pre- and Protohistoric Sciences,
Denver.
Gillespie, W.
1976 Culture Change at the Ute Canyon Site: A
Study of the Pithouse-Kiva Transition in the
Mesa Verde Region. Unpublished M.A. thesis,
Department of Anthropology, University of
Colorado, Boulder.
GIS World
1990 The 1990 GIS Sourcebook. 1st edition,
GIS World, Fort Collins, Colorado.
Fritts, H. C., G. R. Lofgren, and G. A. Gordon
1981 Past Climate Reconstructed from Tree
Rings. In Climate and History: Studies in
Interdisciplinary History. edited by R. I.
Rotberg and T. K. Rabb. pp. 193213.
Princeton University Press, Princeton, New
Jersey.
Gleichrnan, C., and P. Gleichman
1992 The Lower Sand Canyon Survey. In The
Sand Canyon Archaeological Project: A
Progress Report, edited by W. D. Lipe, pp.
25-31. Occasional Paper No.2. Crow Canyon
Archaeological Center, Cortez.
Fuller, S. L.
1985a Dove Creek Canal, Reach 2 Research
Design. Four Comers Archaeological Project
Report No.1. Complete Archaeological
Service Associates, Cortez, Colorado.
Submitted to Bureau of Reclamation, Upper
Colorado Region, Salt Lake City.
Gould, R. R.
1982 The Mustoe Site: The Application of
Neutron Activation Analysis in the Interpretation of a Multi-Component Archaeological Site. Unpublished Ph. D. dissertation,
Department of Anthropology, University of
Texas, Austin.
1985b South Canal Research Design. Four
Comers Archaeological Project Report No.3.
Complete Archaeological Service Associates.
Cortez, Colorado. Submitted to Bureau of
Reclamation, Upper Colorado Region, Salt
Lake City.
-
Graybill, D. A.
1989 The Reconstruction of Prehistoric Stream
Flow. In The 1982-1984 Excavations at Las
Colinas. Environment and Subsistence. by
D.A. Graybill, D. A. Gregory, F. L. Nials, S.
K. Fish, C. H. Miksicek, R. E. Gasser, and
200-
C. R. Szuter, pp. 2538. Archaeological
Series 162, Vol. 5, Pt. 1. Cultural Resource
Management Division, Arizona State Museum, University of Arizona, Tucson.
Hallisy, S. J.
1974 Salvage Excavations at Sites
5MTUMR2344 and 5MTIJMR.2347 in
Mancos Canyon, Southwest Colorado.
Submitted to Bureau of Indian Affairs,
Albuquerque Area Office, Albuquerque.
Gross, G. T.
1985 Anasazi Storage Facilities in the Dolores
Region. Colorado. A.D. 600-920. Ph.D. dissertation, Deparunent of Anthropology, Washington State University, Pullman.
Hammack, L. C.
1983 Cultural Resource Inventory: Shell Mobil
C02 Project 25 KV Electric Distribution
System. Yellowjacket and Hovenweep Areas.
Complete Archaeological Services Associates,
Cortez, Colorado. Submitted to Bureau of
Land Management, San Juan Area, Durango,
Colorado.
Gross, G. T., and D. Howes
1986 Cougar Springs Cave (Site 5MT4797). In
Dolores Archaeological Program: Anasazi
Communities at Dolores: Early Small
Settlements in the Dolores River Canyon and
Western Sagehen Flats Area, compiled by T.
A. Kohler, W. D. Lipe, and A. E. Kane, pp.
80--1 DO. Bureau of Reclamation, Engineering
and Research Center, Denver.
Hammack, N. S.
1984 Archaeological Testing and Monitoring.
ShelllMobile C02 Yellow Jacket and
Hovenweep Systems. Montezuma County,
Colorado. Complete Archaeological Service
Associates, Cortez, Colorado. Submitted to
Bureau of Land Management, San Juan Area,
Durango, Colorado.
Gumennan, G. J. (editor)
1988 The Anasazi in a Changing Environment.
Cambridge University Press, Cambridge.
Hammett, J. E., and N. H. Olsen (editors)
1984 Hovenweep 1977. Archeological Report
No.4. Office of Sponsored Programs, San
Jose State University, San Jose, California.
Haase, W. R., IV
1983 Pueblo II and Pueblo JJJ Settlement
Patterns on Cedar Mesa, Southwestern Utah.
Unpublished MA. thesis, Department of
Anthropology, Washington State University,
Pullman.
Hamon, C. E.
1983 Detennining Capability Class of Soils.
SCS Technical Notes. Agronomy Note No.
56, 3rd rev., July.
Haase, W. R. IV
1985 Domestic Water Conservation Among the
Northern San Jan Anasazi. Southwestern Lore
51(2):15-27.
Harris, A. H.
1970 Past Climate of the Navajo Reservoir
District American Antiquity 35:374-377.
Hack, J. T.
1942 The Changing Physical Environment of
the Hopi Indians ofArizona. Papers of the
Peabody Museum of American Archaeology
and Ethnology, Vol. 35, No.1. Cambridge,
Massachusetts.
Hart, J. A., and D. B. Wherry
1984 VICAR/IBIS User Reference Manual.
Washington State University Digital Image
Analysis Laboratory, Computing Service
Center, Pullman.
Hall, S. A.
1977 Late Quaternary Sedimentation and
Paleoecologic History of Chaco Canyon, New
Mexico. Geologic Society of America Bulletin
88: 1593-1618.
Hal1asi, J. A.
1979 Part II: Archaeological Excavation at the
Escalante Site, Dolores, Colorado, 1975 and
1976. In The Archaeology and Stabilization of
the Dominguez and Escalante Ruins, pp. 197425. Cultural Resource Series No.7. Colorado
State Office, Bureau of Land Management,
Denver.
-
Hasenstab, R. J.
1983a A Preliminary Cultural Resource Sensitivity Analysis for the Proposed Flood
Control Facilities Construction in the Passaic
River Basin ofNew Jersey. Report submitted
for the Passaic River Basin Special Studies
Branch, Department of the Anny, New York
District, Anny Corps of Engineers.
201-
1983b The Application of Geographic Information Systems to the Analysis of Archaeological Site Distribution. Paper presented at
the 48th Annual Meeting of the Society for
American Archaeology, Pittsburgh.
Summers of 1875 and 1876. In Tenth Annual
Report of the U.S. Geological and Geographical Survey of the Territories, Embracing
Colorado and Parts ofAdjacent Territories,
Being a Report of Progress of the Exploration
for the Year 1876 (compiled) by F.V. Hayden,
pp. 383-407. U. S. Government Printing
Office, Washington, D.C.
Hassan, F. A.
1981 Demographic Archaeology. Academic
Press, New York.
Hastorf, C.
1980 Changing Resource Use in Subsistence
Agricultural Groups in the Prehistoric
Mimbres River Valley, New Mexico. In
Modeling Changes in Prehistoric Subsistence
Economies, edited by T. Earle and A.
Christensen, pp. 79120. Academic Press,
New York.
Honeycutt, L., and J. E. Fetterman
1982 Cultural Resource Inventory of the Empire
Electric 115 KV Transmission Line. Bureau of
Anthropological Research, Dove Creek,
Colorado. Submitted to Empire Electric
Association, Cortez, Colorado.
1985 Letter Report on the Cultural Resource
Inventory ofParcels Associated with the Dove
Creek Canal-Reach I. Bureau of
Anthropological Research, Dove Creek,
Colorado. Submitted to Bureau of
Reclamation, Cortez Projects Office, Cortez.
Hayes, A. C.
1964 The Archeological Survey of Wetherill
Mesa, Mesa Verde National Park, Colorado.
Archaeological Research Series No.7A.
National Park Service, Washington, D. C.
Hayes, A. C., and J. A. Lancaster
1975 Badger House Community, Mesa Verde
National Park, Colorado. Publications in
Archeology 7E. National Park Service,
Washington, D.C.
Haynes, D. D., J. D. Vogel, and D. G. Wyant
(compilers)
1972 Geology, Structure, and Uranium Deposits
of the Cortez Quadrangle, Colorado and Utah.
Miscellaneous Investigations Series, Map 1629. U.S. Geological Survey, Washington,
D.C.
Hevly, R. H.
1988 Prehistoric Vegetation and Paleoclimates
on the Colorado Plateaus. In The Anasazi in a
Changing Environment, edited by G. J. Gumerman, pp. 92-118. Cambridge University
Press, Cambridge.
Herold,J.
1961 Prehistoric Settlement and Physical
Environment in the Mesa Verde Area. University of Utah Anthropological Papers No.
53. Salt Lake City.
Hogan, P. F.
1987 Prehistoric Agricultural Strategies in
West-Central New Mexico. Ph. D. dissertation, Washington State University, University
Microf1lrns, Ann Arbor.
Holmes, W. H.
1878 Report of the Ancient Ruins of Southwestern Colorado, Examined During the
-
Hough, W.
1915 The Hopi Indians. The Torch Press, Cedar
Rapids, Iowa.
1930 Ancient Pueblo Subsistence. Proceedings
of the 23rd International Congress ofAmericanists, 1928, pp. 67-69. New York.
Huber, E. K.
1989 Preliminary Report of Excavations, 1987
and 1988 At Green Lizard (5MT390l), Montezuma County, Colorado. Crow Canyon
Archaeological Center, Cortez, Colorado, and
Washington State University, Pullman. Submitted to Bureau of Land Management, San
Juan Resource Area. Durango.
Huber, E. K., and W. W. Bloomer
1988 Annual Report of Investigations at Green
Lizard (5MT390l), Montezuma County,
Colorado. Crow Canyon Archaeological
Center, Cortez, Colorado, and Washington
State University, Pullman. Submitted to the
Bureau of Land Management, San Juan
Resomce Area, Durango, Colorado.
Hughes, M K., P. M. Kelly, J. R. Pilcher, and V. C.
LaMarche, Jr. (editors)
1982 Climate from Tree Rings. Cambridge University Press, Cambridge.
Hunt, C. B.
1956 Cenozoic Geology ofthe Colorado Plateau.
Geological Survey Professional Paper 279.
202-
U.S. Government Printing Office,
Washington, D.C.
Management, Anasazi Heritage Center,
Dolores, Colorado.
Hurst, C. T., and V. F. Lotrich
1932 An Unusual Mug from Yellow Jacket
Canyon. El Palacio 33(2122): 195198.
Jackson, W. H.
1878 Report on the Ancient Ruins Examined in
1875 and 1877. In Tenth Annual Report of the
U.S. Geological and Geographical Survey of
the Territoriesfor 1876, pp. 411-450. Washington, D.C.
1933 Another Unusual Bowl from Yellow Jacket
Canyon. El Palacio 34(15-16):111-115.
Huxhold, W. E.
1990 An Introduction to Urban Geographic
Information Systems. Oxford University
Press, Oxford.
Ives, G. A., and M. Errickson
1983 Dove Creek Canal Reach II, Fairview
Pumping Plant, Cahone Pumping Plant, and a
Portion of Dove Creek CanalReach I. Ms.
on rue, Bureau of Land Management, Anasazi
Heritage Center, Dolores, Colorado.
Ives, G. A., and J. D. Orcutt
1981 The Dove Creek Canal, Reach I, Survey
1981. Ms. on file, Bureau of Land Management, Anasazi Heritage Center, Dolores,
Colorado.
Ives, J. C.
1971a Preliminary Report of Archaeological
Excavations by Fort Lewis College Under
Pennit #7O-C0--022 on Public Lands Under
the Supervision of the Bureau of Land Management, 1970-1971. Ms. on file, Bureau of
Land Management, Anasazi Heritage Center,
Dolores, Colorado.
1971b Preliminary Report of Archaeological
Excavations by Fort Lewis College Under
Pennit #71--C0--021 on Public Lands Under
the Supervision of the Bureau of Land Management, 1971. Ms. on rue, Bureau of Land
Management, Anasazi Heritage Center,
Dolores, Colorado.
1972 Preliminary Report of Archaeological
Excavations by Fort Lewis College Under
Pennit #72--C0-022 and a Survey Under
Pennit #7O-C0--021 on Public Lands Under
the Supervision of the Bureau of Land Management, 1972. Ms. on rue, Bureau of Land
Management, Anasazi Heritage Center,
Dolores, Colorado.
1973 Preliminary Report of Excavations Conducted by Fort Lewis College Under Pennit
#73--C0-034. Ms. on rue, Bureau of Land
-
Judd, N. M.
1981 The Material Culture of Pueblo Bonito.
Reprinted. Reprints in Anthropology, vol. 23.
J.L. Reprints, Lincoln, Nebraska. Originally
published 1954, Smithsonian Miscellaneous
Collections, vol. 24. Smithsonian Institution,
Washington, D.C.
Kane, A. E.
1975a Archaeological Resources in the Great
Cut-Dove Creek Area. Dolores Project Report
of the 1974 Season. Ms. on rue, Bureau of
Land Management, Anasazi Heritage Center,
Dolores, Colorado.
1975b Archaeological Resources of the Dolores
River Project: Report of the 1975 Field
Season. Ms. on rue, Bureau of Land Management, Anasazi Heritage Center, Dolores,
Colorado.
Karlsb'om, T. N. V.
1988 Alluvial Chronology and Hydrologic
Change of Black Mesa and Nearby Regions. In
The Anasazi in a Changing Environment,
edited by G. J. Gumerman, pp. 45-91. Cambridge University Press, Cambridge.
Kavena, J. T.
1980 Hopi Cookery. The University of Arizona
Press, Tucson.
Kidder, A. V.
1927 Southwestern Archaeological Conference.
Science 66:489-491.
Kleidon, J., and B. Bradley
1989 Annual Report of the 1988 Excavations at
Sand Canyon Pueblo (5MT765). Crow
Canyon Archaeological Center, Cortez,
Colorado.
KoWer, T. A.
1989 Chapter 1: Introduction. In Bandelier
Archaeological Excavation Project: Research
Design and Summer 1988 Sampling, edited by
T. A. KoWer, pp. 1-12. Reports of Investiga-
203-
tions 61. Department of Anthropology,
Washington Slate University, Pullman.
1992 Prehistoric Human Impact on the
Environment in the Upland North American
Southwest Population and Environment: A
Journal ofInterdisciplinary Studies 13:255268.
Kohler, T. A., and M. H. Matthews
1988 Longtenn Anasazi Land Use and Forest
Reduction: A Case Study from Southwest
Colorado. American Antiquity 53:537-564.
Project Report No. 11. Complete Archaeological Service Associates. Submitted to
Bureau of Reclamation, Upper Colorado
Region, Salt Lake City.
Kvamme, K. L.
1983a Computer Processing Techniques for
Regional Modeling of Archaeological Site
Locations. Advances in Computer Archaeology QZRセU N
1983b New Methods of Investigating the Environmental Basis of Prehistoric Site Location. Unpublished Ph.D. dissertation, Department of Anthropology, University of California, Santa Barbara
Kohler, T. A., J. D. Orcutt. E. Blinman, and K. L.
Petersen
1986 Anasazi Spreadsheets: The Cost of Doing
Agricultural Business in Prehistoric Dolores.
In The Dolores Archaeological Program: Final
Synthetic Report. compiled by D. A Bretemitz, C. K. Robinson, and G. T. Gross, pp.
525-538. Bureau of Reclamation, Engineering
and Research Center, Denver.
1984 Models of Prehistoric Site Location near
Pinyon Canyon, Colorado. In Papers ofthe
Philmont Conference on the Archaeology of
Northeastern New Mexico. edited by C. J.
Condie, pp. 34 -370. New Mexico Archaeological Council Proceedings No.6, Albu-
querque.
Kohler, T. A, and C. R. Van West
1994 The Calculus of Self Interest in the
Development of Cooperation: Sociopolitical
Development and Risk among the Northern
Anasazi. To appear in Resources Stress,
Economic Uncertainty, and Human Response
in the Prehistoric Southwest, edited by J. A
Tainter. Santa Fe Institute, Studies in the
Science of Complexity, Addison-Wesley,
Reading MA, 1994. In press.
1985 Detennining Empirical Relationships
Between Natural Environmental and Prehistoric Site Locations: A Hunter-Gatherer
Example. In For Concordance in Archaeological Analysis: Bridging Data Structure,
Quantitative Technique. and Theory, edited by
Christopher Carr, pp. 208-238. Westport
Publishing, Kansas City.
1986a An Overview of Geographic Infonnation
Systems for Archaeological Research and Data
Management In Proceedings ofthe National
Workshop on Microcomputers in Archaeology. Society for American Archaeology and
Society for Professional Archaeologists,
Washington, D. C.
Kuckelman, K. A.
1986 Cultural Resource Inventory for the
Towaoc Canal Project: Highline Ditch, Lower
Hermana Lateral. Rocky Ford Laterals, and
Portions of the Upper Hermana and Lone Pine
Laterals. Four Comers Archaeological Project
Report No.8. Complete Archaeological
Service Associates. Submitted to Bureau of
Reclamation, Upper Colorado Region, Salt
Lake City.
1986b The Use of Geographic Infonnation
Systems for Modeling Archaeological Site
Distributions. In Geographic Information
Systems in Government, vol. 1, edited by B.
K. Opitz, pp. 345-362. Deepak Publishing,
Hampton, Virginia.
1988 Excavations at Norton House (Site
5MT8839), A Pueblo IT Habitation. In Archaeological Investigations on Douth Canal,
Vol. 2, pp. 373-403. Four Comers Archaeological Project Report No. 11. Complete
Archaeological Service Associates. Submitted
to Bureau of Reclamation, Upper Colorado
Region, Salt Lake City
1988 Development and Testing of Quantitative
Models. In Quantifying the Present and
Predicting the Past: Theory, Method, and
Application of Archaeological Predictive
Modeling, edited by W. J. Judge and L.
Sebastian. pp. 325-428. U.S. Government
Printing Office, Washington, D.C.
Kuckelrnan, K. A, and J. N. Morris
1988 Archaeological Investigations on South
Canal, 2 vols. Four Comers Archaeological
-
204-
1989 Geographic Infonnation Systems in Regional Archaeological Research and Data
Management In Archaeological Method and
Theory, Vol. 1. edited by M. B. Schiffer, pp.
139-203. The University of Arizona Press,
Tucson.
Department of Anthropology, Washington
State University, Pullman.
Lightfoot, R. R., and C. R. Van West
1986 Excavation and Testing Conducted on
Private Land Under State Permit 85-22. Crow
Canyon Archaeological Center, Cortez,
Colorado. Submitted to the Office of the State
Archaeologist, Colorado Historical Society,
Denver.
Kvamme, K. L., and M. A. Jochim
1989 The Environmental Basis of Mesolithic
Settlement. In The Mesolithic in Europe:
Papers Presented at the Third International
Symposium, edited by C. Bonsall, pp. 1-12.
John Donald Publishers, Edinburgh, Scotland.
Lightfoot, R. R., and M. D. Varien
1988 Report of 1987 Investigations at the
Duckfoot Site (5Mf3868), Montezuma
County, Colorado. Crow Canyon Archaeological Center, Cortez, Colorado. Submitted
to the Office of the State Archaeologist,
Colorado Historical Society, Denver.
Kvamme, K. L., and T. A. Kohler
1988 Geographic Infonnation Systems: Technical Aids for Data Collection, Analysis, and
Display. In Quantifying the Present and
Predicting the Past: Theory, Method, and
Application ofArchaeological Predictive
Modeling, edited by W. J. Judge and L. Sebastian, pp. 493-547. U.S. Government Printing
Office, Washington, D.C.
Lipe, W. D.
1989 Social Scale of Mesa Verde Anasazi Kivas.
In The Architecture of Social Integration in
Prehistoric Pueblos, edited by W. D. Lipe and
M. Hegmon, pp. 53-71. Occasional Papers
No.1. Crow Canyon Archaeological Center,
Cortez, Colorado.
Larson, D.O., and J. Michaelsen
1990 Impacts of Climatic Variability and Population Growth on Virgin Branch Anasazi Cultural Developments. American Antiquity
55:227-249.
Lipe, W. D., and R. G. Matson
1971 Human Settlement and Resources in the
Cedar Mesa Area, Southeast Utah. In The
Distribution ofPrehistoric Population
Aggregates, edited by G. J. Gumerman, pp.
126-151. Prescott College Anthropological
Reports No.1. Prescott, Arizona
Lange, F., N. Mahaney, J. B. Wheat, and M.
Chenault
1986 Yellowjacket Jacket, a Four Corners
Anasazi Ceremonial Center. Johnson Books,
Boulder.
Lipe, W. D. (editor)
1992 The Sand Canyon Archaeological Project:
A Progress Report. Occasional Paper No.2.
Crow Canyon Archaeological Center, Cortez,
Colorado.
Lightfoot, R. R.
1985 Report of 1984 Investigations Conducted
Under State Permit 84-15. Crow Canyon
Archaeological Center, Cortez, Colorado.
Submitted to the Office of the State Archaeologist, Colorado Historical Society, Denver.
1987 Annual Report of Investigations at the
Duclifoot Site (5MT3868J, Montezuma
County, Colorado. Crow Canyon Archaeological Center, Cortez, Colorado. Submitted to
the Office of the State Archaeologist, Colorado
Historical Society, Denver.
1988 Roofing an Early Anasazi Great Kiva:
Analysis of an Architectural Model. The Kiva
53:253-272.
1992 Archaeology of the House and Household:
A Case Study ofAssemblage Formation and
Household Organization in the American
Southwest. Unpublished Ph.D. dissertation,
-
Lister, R. H.
1967 Progress Report on the 1967 Inventory of
Archaeological Remains in Lands Controlled
by the Bureau of Land Management in the
Cortez - Dove Creek Area of Southwestern
Colorado. Ms. on ftle, Bureau of Land
Management, Denver.
Lowe, C. H.
1964 Arizona's Natural Environment: Landscapes and Habitats. The University of Arizona
Press, Tucson.
Mackey, J. c., and S. J. Holbrook
1978 Environmental Reconstruction and the
Abandonment of the Largo-Gallina Area, New
205-
Martin, D. W., and J. D. Scurry
1988 The Role of MOSS of Archaeological Predictive Modeling. Paper presented at the 5th
Annual MOSS User's Workshop, New
Orleans.
Mexico. Journal ofField Archaeology 5:2949.
McDonald, C. L., and 1. K. Cain
1985 Applied Computer Graphics in a Geographic Information System: Problems and
Successes. IEEE Computer Graphics and
Applications 5:34-39.
Martin, P. S.
1929 The 1928 Archaeological Expedition of the
State Historical Society of Colorado. Colorado
Magazine 6(1):1-35.
McEwen, R. B., H. W. Calkins, and B. S. Ramey
1983 Overview and USGS Activities. U.S.
Geological SlDVey Circular 895-A. User
Services Section, U.S.G.S. National Cartographic Information Center, Reston, Virginia
1930 Archaeological Expedition of the State
Historical Society of Colorado in Cooperation
with the Smithsonian Institution. Colorado
Magazine 7(1):1-40.
Maffini, G.
1987 Raster versus Vector Data Encoding and
Handling: A Commentary. Photogrammetric
Engineering and Remote Sensing 53: 13971398.
1936 Lowry Ruin in Southwestern Colorado.
Anthropological Series Vol. 23, No.1. Field
Museum of Natural History, Chicago.
1938 Archaeological Work in the AckmenLowry Area, Southwestern Colorado, 1937.
Anthropological Series, Vol. 23, No.2. Field
Museum of Natural History, Chicago.
Macomb, J. N.
1876 Report of the Exploring Expedition from
Santa Fe. New Mexico to the Junction of the
Grand and Green Rivers of the West in 1859.
Government Printing Office, Washington,
D.C.
1939 Modified Basket Maker Sites: AckmenLowry Area, Southwestern Colorado. 1938.
Anthropological Series Vol. 23, No.3. Field
Museum of Natural History, Chicago.
Marble, D. F.
1984 Geographic Information Systems: An
Overview. PECORA 9 Proceedings, Spatial
Information Technologies For Remote Sensing Today and Tomorrow. pp. 1&-24. Sioux
Falls, South Dakota.
Martin, P. S., and W. Byers
1965 Pollen and Archaeology at Wetherill Mesa.
Memoirs of the Society for American Archaeology 19:122-135.
Marozas, B. A., and J. A. Zack
1990 GIS and Archaeological Site Location. In
Interpreting Space: GIS and Archaeology,
edited by K. M. S. Allen, S. W. Green, and
E. B. W. Zubrow, pp. 165-172, Taylor and
Francis, London.
Mather, J. R.
1977 Workbook in Applied Climatology. Rev.
ed. Publications in Climatology 30(1). Laboratory of Climatology, Elmer, New Jersey.
Matson, R. G.
1991 The Origins of Southwestern Agriculture.
The University of Ariwna Press, Tucson.
Martin, C. W.
1976 Archaeological Inventory of the Sand
Canyon Cliff Dwelling Area. Montezuma
County. Colorado. Mesa Verde Archaeological
Research Center. Submitted to Bureau of Land
Management. Montrose District Office, Montrose, Colorado.
Martin, D. W.
1971 Evaluation of the Archaeological Resources
on Public Lands of the Bureau of Land Management in the Cortez-Dove Creek (McElmo
District) of Southern Colorado. Ms. on fUe,
Bureau of Land Management. Anasazi Heritage
Center, Dolores, Colorado.
-
Matson, R. G., and W. D. Lipe
1978 Settlement Patterns on Cedar Mesa: Boom
and Bust on the Northern Periphery. In Investigations of the Southwestern Anthropological Research Group: The Proceedings of
the 1976 Conference. edited by R. C. Euler
and G. J. Gumerman, pp. 1-12. Museum of
Northern Ariwna., Flagstaff.
Matson, R. G., W. D. Lipe, and W. R. Haase
1988 Adaptational Continuities and Occupational
Discontinuities: The Cedar Mesa Anasazi.
Journal ofField Archaeology 15:245-264.
206-
1986b Archaeological Excavations on Reach II
of the Dove Creek Canal. Four Comers
Archaeological Project No.9. Complete
Archaeological Service Associates. Submined
to Bureau of Reclamation, Upper Colorado
Region, Salt Lake City.
Meko, D. M., C. W. Stockton, and W. R. Boggess
1980 A TreeRing Reconstruction of Drought in
Southern California. Water Resources Bulletin
QVZUYTセN
Metzger, T. R., and S. L. BreternitzGoulding
1981 Ruin Stabilization Activities at Sand
Canyon Pueblo Cliff Dwelling Area.
Montezuma County, Colorado. Mesa Verde
Research Center. Ms on file, Bureau of Land
Management, Anasazi Heritage Center,
Dolores, Colorado.
Neily, R. B.
1983 The Prehistoric Community on the
Colorado Plateau: An Approach to the Study
of Change and Survival in the Northern San
Juan Area of the American Southwest. Ph.D.
dissertation, Southern lllinois University.
University Microfilms, Ann Arbor.
Mitchell, J. M., Jr., C. W. Stockton, and D. M.
Meko
1979 Evidence of a 22Year Rhythm of Drought
in the Western United States Related to the
Hale Solar Cycle Since the 17th Century. In
Solar Terrestrial Influences in Weather and
Climate, edited by B. M. McConnac and T. A.
Seliga. pp. 125143. Reidel Publishing,
Dordrecht, Holland.
Newberry, J. S.
1876 Geological Report. In Report of the
Exploring Expeditionfrom Santa Fe, New
Mexico to the Junction of the Grand and Green
Rivers of the West in 1859, by J. N.
Macomb. U.S. Government Printing Office.
Nickens, P. R.
1977 Cultural Resources of the Dolores Project
Area, Southwestern Colorado. Submitted to
the Bureau of Reclamation, Upper Colorado
Region, Salt Lake City.
Monmonier, M. S.
1982 Computer Assisted Cartography: Principles
and Prospects. PrenticeHall, Englewood
Cliffs, New Jersey.
1981 Pueblo III Communities in Transition:
Environment and Adaptation in Johnson
Canyon. Memoirs of the Colorado Historical
Society NO.2. Boulder.
Morley, S. G.
1908 The Excavation of the Cannonball Ruins
in Southwestern Colorado. American
Anthropologist, n.s., 10:596610.
Nickens, P. R., and D. A. Hull
1982 Archaeological Resources of Southwestern
Colorado: An Overview of the Bureau of Land
Management's San Juan Resource Area. In
Archaeological Resources of Southwestern
Colorado, edited by S. Eininger, P. J. Gleichman, D. A. Hull, P. R. Nickens, A. R. Reed,
and D. D. Scott, pp. 1-308. Cultural Resource
Series No. 22. Bureau of Land Management,
Denver.
Morley, S. G., and A. V. Kidder
1917 The Archaeology of McElmo Canyon
Colorado. EI Palacio 4:41-70.
Morris, E. H.
1939 Archaeological Studies in the La Plata
District, Southwestern Colorado and
Northwestern New Mexico. Carnegie Institute
of Washington Publication No. 519.
Washington, D.C.
Morris, E. H., and R. F. Burgh
1954 Basketmaker II Sites Near Durango,
Colorado. Carnegie Institute of Washington
Publication No. 604. Washington, D.C.
Morris, J. N.
19800 Monitoring and Excavation at Aulston
Pueblo (Site 5MT2433), a Pueblo II Habitation Site. Four Corners Archaeological Project
Report No.6. Complete Archaeological
Service Associates. Submitted to Bureau of
Reclamation, Upper Colorado Region, Salt
Lake City.
-
Orcutt, J. D., and D. A. Goulding
1986 Archaeological Survey of the McPhee
Reservoir. In Dolores Archaeological Program:
Research Designs and Initial Survey Results,
compiled by A. E. Kane, W. D. Lipe, T. A.
Kohler, and C. K. Robinson, pp. 185-375.
U.S. Department of the Interior, Bureau of
Reclamation, Research and Engineering
Center, Denver.
Palmer, W. C.
1965 Meteorological DroughJ. Research Paper
No. 45. U.S. Department of Commerce,
207-
Office of Climatology, U.S. Weather Bureau,
Washington, D. C.
Parker, H. D.
1988 The Unique Qualities of a Geographic
Information System: A Commentary.
Photogrammetric and Engineering and Remote
Sensing 54:1547-1549.
Parker, S.
1985 Predictive Modeling of Site Settlement
Systems Using Multivariate Logistics. In For
Concordance in Archaeological Analysis:
Bridging Data Structure, Quantitative Techniques, and Theory, edited by C. Carr, pp.
173207. Westport, Kansas City.
1986 The Role of Geographic Information
Systems in Cultural Resource Management. In
Geographic Information Systems in Government, Vol. I, edited by B. K. Opitz, pp.
133140. Deepak Publishing, Hampton,
Virginia.
Parker, S., W. F. Limp, J. A. Farley, and I. Johnson
1986 Database Management Geographic Information Systems and Predictive Modeling in
the Forest Service. Arkansas Archaeological
Survey. Draft report submined to the U.S.
Department of Agriculture, Forest Service,
Region 8, Atlanta.
Parsons, E. C.
1936 Hopi Journal of Alexander M. Stephen.
Columbia University Contributions to Anthropology 23 and 24. Columbia University
Press, New York.
Petersen, K. L.
1981 Ten Thousand Years of Climate Change
ReconstructedfromFossil Pollen. Ph.D. dissertation, Washington State University. University Microfilms, Ann Arbor.
1987 Tree-Ring Transfer Functions For Estimating Com Production. In Dolores Archaeological Program: Supporting Studies: Settlement and Environment, compiled by K. L.
Petersen and J. D. Orcutt, pp. 217-231.
Bureau of Reclamation, Engineering and
Research Center, Denver.
Petersen. K. L., and V. L. Clay
1987 Characteristics and Archaeological Implications of Cold Air Drainage in the Dolores
Project Area In Dolores Archaeological
Program: Supporting Studies: Settlement and
Environment, compiled by K. L. Petersen and
J. D. Orcutt, pp. 187-216. Bureau of Reclarnation, Engineering and Research Center,
Denver.
Petersen, K. L., and P. J. Mehringer, Jr.
1976 Postglacial Timberline Fluctuations, La
Plata Mountains, Southwestern Colorado.
Arctic and Alpine Research 8:275-288.
Petersen. K. L., and J. D. Orcutt (compilers)
1987 Dolores Archaeological Program: Supporting Studies: Settlement and Environment.
Bureau of Reclamation, Engineering and Research Center, Denver.
Petersen, K. L., V. L. Clay, M. H. Matthews, and S.
W. Neusius (compilers)
1985 Dolores Archaeological Program: Studies
in Environmental Archaeology. Bureau of
Reclamation, Engineering and Research
Center, Denver.
Powers, R. P., W. B. Gillespie, and S. H. Lekson
1983 The Outlier Survey: A Regional View of
Settlement in the San Juan Basin. Reports of
the Chaco Center No.3. Division of Cultural
Research, National Park Service, Albuquerque.
Price, A. B., W. D. Nettleton, G. A. Bowman, and V.
L. Clay
1988 Selected Properties, Distribution, Source,
and Age of Eolian Deposits and Soils of
Southwestern Colorado. Soil Science Society
ofAmerica Journal 52:450-455.
Prudden, T. M.
1903 The Prehistoric Ruins of the San Juan
Watershed in Utah, Arizona., Colorado and
New Mexico. American Anthropologist, n.S.,
5:224-288.
1914 The Circular Kivas of Small Ruins in the
San Juan Watershed. American Anthropologist, n.s., 16:33-58.
1918 A Further Study of the Prehistoric Small
House Ruins in the San Juan Watershed.
Memoirs of the American Anthropological
Association, vol. 5, No.1.
1988 Climate and the Dolores River Anasazi.
University of Utah Anthropological Papers
No. 113. Salt Lake City.
-
208-
1972 Social Implications of Pueblo Water
Management in the Northern San Juan.
Sonderdruck: Zeitschrift Fur Ethnologie, Band
97, Heft 2, Braurschweig, pp. 212-219.
Reed, A. D.
1979 Part I: The Dominguez Ruin: A McElmo
Phase Pueblo in Southwestern Colorado. In
The Archeology and Stabilization of the
Dominguez and Escalante Ruins, pp. 1196.
Cultural Resource Series NO.7. Colorado
State Office, Bureau of Land Management,
Denver.
1974 Payne Site Investigations. Southwestern
Lore 40(3 & 4):40--52.
1975 A Stockaded Basketmaker III Village at
Yellowjacket, Colorado. The Kiva 40:113120.
Riley, C. L.
1948 An Archeological Survey at Hovenweep
National Monument. Ms. on fLle, Mesa Verde
Research Center, Mesa Verde National Park,
Colorado.
1977 Culture Change and Continuity on Chapin
Mesa. Regents Press of Kansas, Lawrence.
1989 Northern San Juan Prehistory. In Dynamics of Southwest Prehistory, edited by L. S.
Cordell and G. J. Gumerman, pp. 149-177.
Smithsonian Institution Press, Washington,
D.C.
1950 "Defensive" Structures in the Hovenweep
Monument. El Palacio 57: 339344.
Rinaldo, J.
1950 An Analysis of Culture Change in the
AckmenLowry Area. Fieldiana 36(5):93-106.
Rose, M. R., J. S. Dean, and W. J. Robinson
1981 The Past Climate ofArroyo Hondo, New
Mexico, Reconstructed From Tree-Rings.
Arroyo Hondo Archaeological Series Vol. 4.
School of American Research Press, Santa Fe.
Ripple, W. J.
1987 GIS for Resource Management: A Compendium. American Society of Photogrammetry and Remote Sensing and American
Congress on Surveying and Mapping, Bethesda, Maryland.
Rose, M. R., W. J. Robinson, and J. S. Dean
1982 Dendroclimatic Reconstructionfor the
Southeastern Colorado Plateau. Final Report
to Dolores Archaeological Project and Division of Chaco Research. Laboratory of TreeRing Research, The University of Ariwna,
Tucson.
1989 Fundamental of GIS: A CompendiUm.
American Society of Photogrammetry and
Remote Sensing and American Congress on
Surveying and Mapping, Bethesda, Maryland.
Roberts, F. H. H., Jr.
1925 Report on Archaeological Reconnaissance
in Southwestern Colorado in the Summer of
1923. Colorado Magazine 2(2):1-80.
Roys, L.
1936 Lowry Ruin as an Introduction to the
Study of Southwestern Masonry. In Lowry
Ruin in Southwestern Colorado, by P. S.
Maron, pp. 115-142. Anthropological Series
Vol. 23, No.1. Field Museum of Natural
History, Chicago.
1930 Early Pueblo Ruins in the Piedra District,
Southwestern Colorado. Bureau of American
Ethnology Bulletin 96. Washington, D. C.
Robinson, W. J., and B. G. Harrill
1974 Tree-Ring Datesfrom Colorado V, Mesa
Verde Area. Laboratory of Tree-Ring Research,
The University of Ariwna, Tucson.
Rohn, A. H.
1963 Prehistoric Soil and Water Conservation on
Chapin Mesa, Southwestern Colorado. American Antiquity 28:441-455.
1971 Mug House, Mesa Verde National Park,
Colorado. Archeological Series No.7-D.
National Park Service, Washington, D. C.
-
Sanders, W. T.
1976 The Agricultural History of the Basin of
Mexico. In The Valley of Mexico: Studies in
Pre-Hispanic Ecology and Society, edited by
E. R. Wolf, pp. 101-159. University of New
Mexico Press, Albuquerque.
1979 The Basin ofMexico: Ecological Processes
in the Evolution of a Civilization. Academic
Press, New York.
SAS Institute
1985a SAS User's Guide: Basics. 5th ed. SAS
Institute, Cary, North Carolina.
209-
1985b SAS User's Guide: Statistics. 5th ed.
SAS Institute, Cary, North Carolina
R. D. Jones. pp. 207-222. Arkansas Archaeological Survey Research Series No. 14.
Fayetteville.
Schaafsma, P.
1980 Indian Rock Art of the Southwest. School
of American Research and University of New
Mexico Press. Santa Fe and Albuquerque.
Schulman. A.
1950 Pre-Columbian Towers in the Southwest.
American Antiquity 15:288-297.
Schlanger. S. H.
1985 Prehistoric Population Dynamics in the
Dolores Area, Southwestern Colorado. Ph.D.
dissertation, Washington State University.
University Microfilms. Ann Arbor.
Sebastian, L.
1988 Leadership, Power, and Productive Potential: A Political Model of the Chaco System.
Ph.D. dissertation, University of New Mexico.
University Microfilms, Ann Arbor.
1986a 1982 Probabilistic Sampling Survey of
Windy Ruin and Yellowjacket Crest Locality.
In Dolores Archaeological Program: Research
Designs and Initial Survey Results. compiled
by A. E. Kane, W. D. Lipe, T. A. Kohler. and
Bureau of RecC. K. Robinson, pp. TWセPN
lamation. Research and Engineering Center,
Denver.
1992 The Chaco Anasazi: Sociopolitical development in the Prehistoric Southwest. Cambridge University Press, Cambridge.
Soil Conservation Service
1951 Soil Survey Manual. U.S. Department of
Agriculture Handbook 18. U.S. Government
Printing Office, Washington. D. C.
1986b Population Studies. In Dolores Archaeological Program: Final Synthetic Report.
compiled by D. A. Breternitz. C. K. Robinson, and G. T. Gross, pp. 493-524. Bureau
of Reclamation, Research and Engineering
Center, Denver.
Speth, J. D., and S. L. Scott
1989 Horticulture and Large-Mammal Hunting:
The Role of Resource Depletion and Constraints of Time and Labor. In Farmers as
Hunters: the Implications of Sedentism, edited
by S. Kent, pp. 71-79. Cambridge University
Press, Cambridge.
Schlanger, S. H., and P. L. Harden
1986 1979 and 1980 Probability Survey of Cline
Crest, Grass Mesa, Beaver Point, Trimble
Point, and Hoppe Point Localities. In Dolores
Archaeological Program: Research Designs and
Initial Survey Results, compiled by A. E.
Kane, W. D. Lipe, T. A. Kohler, and C. K.
Robinson, pp. 379-446. Bureau of Reclamation, Research and Engineering Center,
Denver.
Schoenwetter. J.
1970 Archaeological Pollen Studies of the Colorado Plateau. American Antiquity SUZセXN
Stiger. M. A.
1979 Mesa Verde Subsistence Patterns from
Basketrnaker to Pueblo m. The Kiva 44: 133144.
Stockton, C. W.
1975 Long-Term Streamflow Records Reconstructed from Tree Rings. Papers of the .
Laboratory of Tree Ring Research 5. University of Arizona, Tucson.
Schoenwetter, J.• and F. W. Eddy
1964 Alluvial and Palynological Reconstruction
of Environments, Navajo Reservoir District.
Museum of New Mexico Papers in Anthropology, No. 13. Santa Fe.
Scholtz, S. C.
1981 Location Choice Model in Sparta. In
Settlement Predictions in Sparta: A Locational
Analysis and Cultural Resource Assessment
on the Uplands of Calhoun County. Arkansas.
by R H. Lafferty m, J. L. Otinger, S. C.
Scholtz, W. F. Limp, B. Watkins and
-
Star, J., and J. E. Estes
1989 Geographic Information Systems: An
Introduction. Prentice-Hall, Englewood Cliffs,
New Jersey.
Stockton, C. W., and W. R. Boggess
1980 Augmentation of Hydrological Records
Using Tree Rings. In Proceedings of the
Engineering Foundation Conference "Improved
Hydrological Forecasting, Why and How."
Asilomar Conference Grounds, Pacific Grove.
California. March 25-30, 1979, pp. 239-265.
American Society of Civil Engineers and
American Meteorological Society, New York.
210-
Stokes, M. A. and T. L. Smiley .
.
1968 An Introduction to Tree-Ring Dating. The
U.S. Bureau of the Budget
1979 U.S. National Map Accuracy Standards. In
University of Chicago Press, Chicago.
Mapsfor America: Cartographic Products of
the U.S. Geological Survey and Others, by M.
Thompson, p. 104. U.S. Geological Survey,
Reston, Virginia.
Swannack, J. D., Jr.
1%9 Big Juniper House, Mesa Verde National
Park, Colorado. Archeological Research Series
Van Devender, T. R., and W. G. Spaulding
1979 Development of Vegetation and Climate in
the Southwestern United States. Science
204:701-710.
No. 7C. National Park Service, Washington,
D.C.
Swedlund, A. C.
1969 Human Skeletal material from the Yellow
Jacket Canyon Area, Southwestern Colorado.
Van West, C. R.
1986 Cultural Resource Inventory for the 1985
Field Season in the Crow Canyon and Sand
Canyon Areas, Montezuma County, Colorado.
Crow Canyon Archaeological Center, Cortez.
Submitted to the Office of the State Archaeologist, Colorado Historical Society, Denver.
Unpublished M.A. thesis, Department of
Anthropology, University of Colorado,
Boulder.
Thompson, C., S. A. Baker, J. R. Allison, and K. L.
Winteh
1986 The Nancy Patterson Village Archaeological Research Project, Field Year 1985.
Preliminary Report No.3. Brigham Young
1990 Modeling Prehistoric Climatic Variability
and Agricultural Productiopn in Southwestern
Colorado: A GIS Approach. Unpublished
University, Museum of Peoples and Cultures,
Technical Series No. 8627. Provo, Utah.
Ph.D. dissertation, Deparnnent of Anthropology, Washington State University,
Pullman.
Tipps, B. L.
1978 Stabilization at Three Prehistoric Ruins in
the San Juan Basin, 1977. Mesa Verde Research Center. Ms. on file, Bureau of Land
Management, Anasazi Heritage Center,
Dolores, Colorado.
1994a Reconstructing Prehistoric Climatic
Variability and Agricultural Production in
Southwestern Colorado, A.D. 901-1300: A
GIS Approach. Paper presented at the Anasazi
Symposium, Mesa Verde National Park. Oct.
3-4, 1991, in press. Ms. 1991.
Titiev, M.
1944 Old Oraibi, A Study of the Hopi Indians of
Third Mesa. Papers of the Peabody Museum of
1994b The Heuristic Value of Estimates of
Prehistoric Agricultural Production: A Case
Study from Southwestern Colorado. Paper presented at the Third Annual Southwestern
Symposium, January 17-18, 1992, Tucson,
Ariwna. To be published in Interpreting
Archaeology and Ethnology Vol. 22, No.1.
Cambridge, Massachusetts.
Toll, H. W., III
1977 Dolores River Archeology: Canyon
Adaptations as Seen Through Survey. Cultural
Southwestern Diversity: Underlying Principles
and Overarching Patterns. Arizona State Uni-
Resource Series No.4. Bureau of Land
Management, Denver.
versity Anthropological Papers, Tempe, in
press. Ms. 1992.
Tomlin, C. D.
1995 Agricultural Potential and Carrying Capacity in Southwestern Colorado, AD 901-1300.
Paper presented at the conference, "Pueblo
Cultures in Transition: AD 1150-1350 in the
American Southwest." March 25-28, 1990,
Crow Canyon Archaeological Center, Cortez,
Colorado, in press. Ms. 1992.
1990 Geographic Information Systems and
Cartographic Modeling. Prentice-Hall, Engle-
wood Cliffs, New Jersey.
Tomlinson, R. F., H. W. Calkins, and D. F. MMble
1976 Computer Handling of Geographic Data:
An Examination ofSelected Geographic
Information Systems. Natural Resources Re-
search xm. UNESCO, Paris.
Trewartha. G. T.
1954 An Introduction to Climate. McGraw-Hill,
New Yorlc.
-
Van West, C. R., and T. A. Kohler
1994 A Time to Rend, A Time to Sew: A New
Perspective on Northern Anasazi Social
Development in Later Prehistory. Paper presented at the Conference, "The Anthropology
211-
ofHwnan Behavior Through Geographic Information and Analysis," February 1-2, 1992,
University of California, Santa Barbara Oxford University Press, Oxford, in press. Ms.
1992.
Wansleeben, M.
1988 Geographic Information Systems in
Archaeological Research. In Computer and
Quantitative Methods in Archaeology 1988,
edited by S. P. Q. RaIltz, pp. 435-451. BAR
International Series 446 (ii), British
Archaeological Reports, Oxford
Van West, C.R., and W. D. Lipe
1992 Modeling Prehistoric Climate and Agriculture in Southwestern Colorado. In The Sand
Canyon Archaeological Project: A Progress
Report, edited by W.D. Lipe, pp. 105-119.
Occasional Paper No.2. Crow Canyon Archaeological Center, Cortez, Colorado.
Warren, R. E.
1988 Predictive Model of Archaeological Site
Location: Two Case Studies from the Midwest. Paper presented at the 53rd Annual
Meeting of the Society for American Archaeology, Phoenix.
Van West, C. R., M. A. Adler, and E. K. A. Huber
1987 Archaeological Survey and Testing in the
Vicinity of Sand Canyon Pueblo, Montezuma
County, Colorado, 1986 Field Season. Crow
Canyon Archaeological Center. Submitted to
the Office of the State Archaeologist, Colorado
Historical Society, Denver, and the Bureau of
Land Management, San Juan Resource Area,
Durango, Colorado.
Warren, R. E., S. G. Oliver, J. A. Ferguson, and R.
E. Druhot
1987 A Predictive Model of Archaeological Site
Location in the Western Shawnee National
Forest. Ms. on fIle. Illinois State Museum
Contract Archaeological Program, Springfield.
Wehde,M.
1982 Grid Cell Size in Relation to Errors in
Maps and Inventories Produced by Computerized Map Processing. Phatogrammetric
Engineering and Remote Sensing 48:12891298.
Varien, M. D.
1990 Preliminary Descriptive Report on the
Excavations at Lillian's Site (5MT3936),
Roy's Ruin (5MT3930) , Sharlene's Site
(5MT3918), and Troy's Tower (5MT3951).
Crow Canyon Archaeological Center, Cortez,
Colorado.
Wetterstrom, W. E.
1976 The Effect ofNutrition on Population Size
at Pueblo Arroyo Hondo, New Mexico. 2
vols. Ph.D. dissertation, University of Michigan, University Microfilms, Ann Arbor.
Varien, M. D., and R. R. Lightfoot
1989 Ritual and Nonritual Activities in Mesa
Verde Region Pit Struetmes. In The Architecture of Social Integration in Prehistoric
Pueblos, edited by W. D. Lipe and M. Hegmon, pp. 73-87. Occasional Papers No.1.
Crow Canyon Archaeological Center, Cortez,
Colorado.
Wheat, J. B.
1955 MT-1 Basketmaker ill Site Near Yellowjacket, Colorado. Southwestern Lore 21(2):1826.
1980 Yellowjacket Canyon Archaeology. In
Insights Into the Ancient Ones, 2d ed., edited
by J. H. Berger and E. F. Berger, pp. 60-66.
Interdisciplinary Supplemental Educational
Programs, Cortez, Colorado.
Varien, M., W. Lipe, B. Bradley, M Adler, and I.
Thompson
1990 Mesa Verde Region Settlement, A. D.
1100 to 1300. Ms. on file, Crow Canyon
Archaeological Center, Cortez., Colorado.
Walkenhorst, R. N.
1983 Cahone Laterals Letter Report. Ms. on file,
Bureau of Land Management, Anasazi Heritage
Center, Dolores, Colorado.
White, A. S.
1976 Emergency Stabilization of 5MT263, East
Rock Canyon, Colorado. Mesa Verde Research
Center. Ms. on file, Bureau of Land Management, Anasazi Heritage Center, Dolores, Colo-
mOO.
Walkenhorst, R. N., D. Goulding, M. Errickson, and
G.lves
1983 Fairview Lateral Letter Report. Ms. on
file, Bureau of Land Management, Anasazi
Heritage Center, Dolores, Colorado.
-
White, A. S., and D. A. Bretemitz
1976 Stabilization of Lowry Ruins. Cultural
Resources Series No.1. Bureau of Land Management, Denver.
212-
1977 Hovenweep 1976. Department of
Anthropology, San Jose State University
Archaeological Report No.3. San Jose,
California.
1979 Stabilization of Escalante Ruin 5MT2149
and Dominguez Ruin 5MT2148, Dolores,
Colorado. In The Archeology and Stabilization
of the Dominguez and Escalante Ruins. Cultural Resource Series NO.7. Bureau of Land
Management, Denver.
Whitten, P., T. Keams, and M. Swift
1986 A Report on the Archaeological Survey and
Testing of a C02 Pipeline Right-of-Way on
Cajon Mesa, San Juan County, Utah and
Montezuma County, Colorado. Studies in
Archaeology No. 1. Division of Conservation
Archaeology, San Juan County Museum
Association, Fannington, New Mexico.
Whiting, A. F.
1939 Ethnobotany of the Hopi. Museum of
Northern Arizona Bulletin 15. Flagstaff.
Wilshusen, R. H.
1989 Unstuffmg the Estufa: Ritual Floor
Features in Anasazi Pit Sttuctures and Pueblo
Kivas. In The Architecture of Sociallntegration in Prehistoric Pueblos, edited by W.
D. Lipe and M. Hegmon, pp. 89-111.
Occasional Papers No. 1. Crow Canyon
Archaeological Center, Cortez, Colorado.
Williams, B. J.
1989 Contact Period Rural Overpopulation in
the Basin of Mexico: Carrying Capacity
Models Tested with Documentary Data.
American Antiquity 54:715-732.
Williamson, R. A.
1984 Living the Sky: The Cosmos of the
American Indian. University of Oklahoma
Press, Norman.
Winter, J. C.
1975 Hovenweep 1974. Department of Anthropology, San Jose Stale University, Archaeological Report No. 1. San Jose, California.
1976 Hovenweep 1975. Department of Anthropology, San Jose State University, Archaeological Report No.2. San Jose, California.
-
1984 The Towers of Hovenweep: Theories of
Use and A History of Tower Research. In
Insights Into the Ancient Ones, 2d ed., edited
by J. H. Berger and E. F. Berger, pp. 73-80.
Interdisciplinary Supplemental Educational
Programs, Cortez, Colorado.
Woods Canyon Archaeological Consultants
1985 Archaeological Testing Program of Four
Sites in the Goodman Point Area. Montezuma
County, Colorado. Woods Canyon Archaeological Consultants. Submitted to Shell
Pipeline Corporation. Ms. on file, Bureau of
Land Management, Anasazi Heritage Center,
Dolores, Colorado.
Woosley, A. 1.
1977 Hovenweep 1977, Preliminary Report. San
Jose State University. Ms. on fIle, Bureau of
Land Management, Anasazi Heritage Center,
Dolores, Colorado.
Wycoff, D. G.
1977 Secondary Forest Succession Following
Abandonment of Mesa Verde. The Kiva
42:215-231.
Zahner, R., J. R. Saucier, and R. K. Myers
1989 Tree-Ring Model Interprets Growth Decline in Natural Stands of Loblolly Pine in the
Southeastern United States. Canadian Journal
for Resources 19:612-621.
Zubrow, E. B. W.
1990 Modelling and Prediction with Geographic
Information Systems: A Demographic Example from Prehistoric and Historic New
York. In Interpreting Space: GIS and Archaeology, edited by K. M. S. Allen,
S. W. Green, and E. B. W. Zubrow, pp.
307-318. Taylor and Francis, London.
213-
Page Blank in Original
ApPENDIX A
EIGENVECTOR AMPLITUDES GENERATED BY A
PRINCIPAL COMPONENTS ANALYSIS ON SEVEN
EXPANDED TREERING CHRONOLOGIES
(SWOLD7)
Year
AD
Axis
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920.
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
.394
.540
.427
.259
.954
.220
.144
.361
.176
.023
.080
.136
.042
.201
.233
.299
.181
.165
.178
.051
.252
.132
.029
.133
.110
.279
.059
.616
.024
.410
.033
.378
.140
.002
.148
.184
.843
.377
.115
.315
.545
.044
.219
.039
.204
.329
7
AX1S
6
.128
1.114
.299
.290
.352
.163
.167
.510
.636
.715
.223
.854
.350
.109
.804
.414
.533
.074
.161
.456
1.156
.237
.683
.679
.325
.378
.075
.631
.482
.055
.433
.386
.550
.719
.436
.648
.110
.224
.725
.383
.077
.280
.246
.556
.257
.007
Axis
AX1S
Axis
AX1S
Axis
4
5
3
2
1
.117
.308 1 .510
.545 4.179
.095
.153 .481 1 . 142
.288
.092 1 .651 .444
.670 1 .910
.486
.237 .487
.913 1 .135
.402
.261 .016
.124 2.920
.479 .376
.845
.208 5.392
.280 .118
.376 .089 4.438
.106
.571
.202
.600 3.303
.368 .947 1.313 .737
.278
.639 1 .382 .424 .256
.055
.173
.879 .779 1 .396 3.029
.214 1 .505 .109 1.411
1 .987
.125
.239
.299 .765 .343
.166
.486 1 .077 .910
.013
.285 .743 .681
1 .566 .153
.276
.445 1.409 1.096
.299
.826
.855 .277 .491 2.646
.857
.773
.489 .175 2.193
.675
.206
.663 .493 ""2.625
.447 .318 .469
.769 1.022
.476
.842
.606 .756 .289
.021 .367 .055
.631 4.273
.492 1 .713 1 .411 .104
.283
.135 .459 1 .193 .851
3.235
.558
.908 .031
.113 2.157
.087 .864
.387
.221 .227
.225 .199
.024 .180 1.695
.054
.957
.298 .391 3.109
.645 .890
.183 .708 .104
.069 .868 1 .359
.811 2.296
.163
.712 .842 .502 .186
.064 .129
.583 .224 .039
.615 .817
.005 .046
.586
.129 .350 2.177 1 .810
.398
.636 1.147
.993
.276
.168
.557
.216 .214
.267 1.050
.044
.734 2.889
.052 .546
.444
.074 2.127
.819
.090
.942 .163 1 .417
1.376 .275
.346 1.047
.636 .098 1 .079
.182
.935 1.699
.662 .492
.826 .298 .636 .615 .287
1.104
.720 1 .201
.201 .128
.291 .044
.416 .392 1.325
1 .120
.218 .201
.340 .189
.343 ;660 .059
.337 2.117
Year
AD
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
Axis
Axis
Axis
Axis
Axis
Axis
Axis
7
5
4
6
3
1
2
.725
.165
.351
.066 .895 .395
· 111
.035
.387 .516 .724 .212 .160 .850
.333
.690
.021
.244
.867
.704 2.843
.175
.225 .496
1. 214 1 .628
.981
1 .009
.622 .219 .981
.290 .735 .368 2.972
.643 .180 .185 .051
.462 .464
· 198
.647
.098 .668 .066 1 .025 .439 2.094
.245 .003
.024 .480 .546
.013 4.221
.547
.688
1 .077
.847
.044 .282
· 101
.383 .530 .047 1 . 109
.037
.393 3.366
.481
.691
.242 .241
.768
.908 2.260
.012
.310 .325
.736 1 .012 1. 538 2.533
.208 .834 .169 .068 .159 1 .080 2.459
.689 .064
.025 .747
1 .597 .586 3.202
.369 .267 .253
.886 . 182 1.370
.844
.190
.547 1.129 1 .465
.863
.523 3.067
.101
.281
.482
.167
.240 .331
.826
.098
.099
.797 1.117 .733
.674 1 .087
.056 1 .086 .696
.770 2.274 1 .382 1 .047
.484
.100
.304 .594
.298
.251 4.081
.293 .531
.549 .246
. 191
.740 2.343
1 .014 .008
.498 .272 .167
1.236 .212
.044
. 1 71
.585
1 .296 .721 .672 2.885
.269 .331
.567 1 .075
.422
.269 3.683
.744
.217 .216 .714 1 .004 1 .088
.592
.338
.557 .144 .173
.096
.958 4.724
.229
.396
.468
.401
.980 2.315 3.298
.243 .095
.122 .597 1 .498 1. 530
.817
.212
.604 .525 .588 .514 .392 5.149
.006 1 .025 1 . 141 .182
.092 1 .407
.301
.030
.175 .156
.233 .279 .835 2.384
.534
.419 .609 .429 1 .055 .379 2.588
.222
.199
• 1 19 .325 1 .072
.280 .507
.835 .215
.011
.206
.180
.639 セN 143
.636
.700
.175
.441
.034
.312 3.542
.499
.428 .501
1.038 1.275 .825
· 141
.330 .014 .142
.923 .586
.022 1 .750
'.480 .304 .070
.232
.448 .521
1 .822
.489
.068 .309
.415 .152 .075 1. 924
.561
.293
.822
.791
.359
. 118 2.678
.212
.176 .855 .783
.035
.297 4.004
.985 .868
.024 .419
.037 .222 4.761
.593
.029 .102 1.992 .004
.183 4.990
.086 .548 1 .361
.439 .457
.024
.221
.515 .637 .017
.803 .559 1 .410 3.643
.043
.335
.005
.412 .370 .508 1 .994
.004
.857
.636
.292 .796 .329 2.388
.080 .315
.073
.792
.909 2.552
.085
.456 .306
.170 1. 556 1.690 1 .461 1. 957
216-
Year
AD
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
101 1
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
Axis
Axis
Axis
Axis
AX1S
Axis
Axis
7
6
5
4
3
2
1
.818
.397 2.433 1 .223 1. 984
.398 1 .752
1 .834 1 .044 1 .995
.029 .845 1.010 .507
.738
.470 1 .520
.228 1 .033
.832 .800
.508 .986 , .029 3.986
.197
.049 .148
. 176 1 .300
.286 .431
.136 2.353
.538
.054 .188
.292
.079 .435 2.926
.333
.413
. 107
.535
.336
.193 1 .277
.039
.245 .295 1 .067
.389 .840
.864
.298
.017
.226 .053
.818
.493
.282 .063
.476 1 .402 4.516
.070 .076 1 .091
.211
.235
.638 2.022 .956
. 713 .202
.332
.763
.773
.681 2.671
.745
.023
· 181
. 171
.709 .561 1 .843
.448 .594 .677
.886 .640
.354 3.595
.096 .547
· 161
.261
.142 1 .369
.669
..829
.627 .380
.411
.454
1
.239
.709
.016
1.509
· 151
.498 1.418
.402
.397 .589
.195 .306
.113 1.311
1. 719
.301
1 .694
.242 .129
.548
.290 .491 1.411
.561
3.620
. 101
.131
.674
.321 1. 719
.688 .610
.474
.301
.092
.329 .464
.217
.309 2.107
.415
.960 .456 1 .380 .512
.366
.125
.143 .543 .187
.469 .727
1 .295
.363
.302 .601
.106
.728
.185
.234 3.320
.026 .906
.275 1 .016 1.359 2.519
· 140
.208 .125 .314 .505 .354
.122 .242
.212 .379 .468 .027 1 .260
.390 1 .041
.173
.434
.146 2.151
.877 1. 329 1 .683
.422 .373 .784 .018
.299 .049 3.947
.477 .224 .956
.685 .932
.653 2.371
.215 . 141
1 . 187
.790 1.070 2.294 1 .969
.476 .581
.157
1 .504 1.419 .946
.145
.052
.529
.042
.394 .474
.966 1.276
.192
.087 .255
.519
.172
.699 3.146
.477
.305
.885 1 .958 .203 .660 1 .496
.052
.409
.396 .142 .631
1 .106
.986
.282
.306 .069
.266
.139
.325 1 .609
.258 .058 .267
.944 .334 1 .007 .649
.002 .227
.129
.568 1 .376 .790
.708
.340
.147
.062 .474 .328 .820 4.816
.455
. 181
.787 .461
.300 1 .018 . 1 .882
.490
.665 .362 .291 .393 .454 1 .098
.114
.484 .166
.363
.402
.509 .618
.106
.099 .653
.275
.261 .469
.075
.264
.551 .393 .336 .728
.229
.566
.337
.335 .296 .139
.874 .390 3.274
.487 .103 .599 .146
.653
.054 2.410
.650
.207 1.580
.154 .033
.473 .153
.096 .103
.285
.386 .143 .786 2.941
217-
Year
AD
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
Axis
7
.946
.035
.028
.416
.052
.475
.334
1 . 180
.221
.046
.079
.367
.000
.003
.762
.214
.537
.059
.533
.008
.397
.256
.035
.432
.048
.359
.266
.133
.231
.295
.225
.175
.411
.104
.205
.334
.137
.210
. 161
.039
.319
.193
.037
.456
.430
.145
.295
.471
.107
Axis
6
.258
.415
.157
.397
.584
.075
.138
. 165·
.697
.679
.242
.710
.225
.245
.144
.480
.323
.216
.490
.199
.045
.425
.315
.446
.282
.350
.002
.254
.858
.355
.050
.023
.409
. 151
.424
.436
.721
.175
.021
.537
.224
.371
.264
.169
.213
.251
.324
.333
.039
Axis
Axis
5
4
.003 .239
1 .225 .154
. 115 .773
.140 .340
.600 .961
.218 .760
.903
.094
.196 .861
.078 .594
.675
1.244
.492
.380
.270 .094
.063
.343
.830 .484
.837 .125
.332
.61 7
.252
.906
.105 .039
.360 .912
.087
.151
.865 .550
.474 .707
.167 .256
.493 .145
.017 .151
.123 .921
. 169
.676
.267 .121
1.073 .601
.606 .118
.114
.425
.545
.207
.613 .099
.499 .041
.287
.833
.362 1.697
.496
.389
.054
.631
.381 .352
.039 .252
.344 1 .011
.551 1 .330
.368 .067
.346 .742
.491 1 .043
.073
.462
.076
.200
.195
.085
.931
.186
218-
Axis
3
1 .510
.400
1 .038
.824
.626
.308
.880
.430
.557
.166
.265
·1 .237
.152
.396
.384
.788
. 121
.803
1.310
1.273
.098
.868
.375
.137
.285
.031
.137
.267
.445
.421
1.226
.183
.230
1.078
. 147
1.176
.324
.633
.649
.153
1 .124
.077
1 .469
1 .309
.098
.444
.045
1 .268
.243
Axis
2
.686
.863
MNUQセ
.160
.647
.331
.461
1 .066
1 .480
.720
1 . 165
.233
.052
1 .583
.545
.766
.856
.685
.271
.257
.193
1 .035
.666
.273
.651
.470
.316
.087
.212
.095
1.513
. 168
.163
.614
1.703
.949
.577
2.590
.002
.600
.308
.818
2.220
1 .290
.967
.367
.500
.691
.267
Axis
1
.075
1 .413
.832
4.295
.792
1.376
2.067
4. 133
.623
.998
1 . 192
.332
.990
.041
1.123
2.188
1 . 160
2.626
3.964
2.430
4.924
.744
4.172
2.965
.519
.457
.546
1.265
.676
.274
2.003
2.272
1 .184
.315
2.105
1.729
2.677
1 .887
3.066
.857
2.780
.885
2.047
1 .438
.099
5.669
2.877
2.310
2.596
Year
AD
1094
1095
1096
1097
1098
1099
1 100
1101
1102
1103
1104
1105
1 106
1 107
1108
1109
1 110
1 111
1 1 12
1113
1114
1 1 15
1 1 16
111 7
1118
1 1 19
1120
1 121
1122
1123
1124
1125
1126
1127
1128
1129
1130
11 31
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
Axis
7
· 107
.403
1 .756
.483
.949
.094
.155
.083
· 141
.177
.817
.140
.020
.098
.449
.123
.147
.144
.782
.368
.349
.267
.136
.354
.444
.073
· 1 13
.381
.043
.185
.382
.599
.609
.182
.199
.181
.467
.583
.242
.291
.242
.196
.333
.456
. 117
1.101
.088
.182
.172
AX1S
Axis
Axis
4
5
6
3
.172 .939 . 175
.631
.109
.261
1 .081
1 .220
.300 1 .076 .262 . 1 10
.415 .485
.217 1.117
.448
.608 .299 .210
.574 .081
.360 . 511
.182
.406
.631
.108
.019
.888 2.135
.651
.005 .356
.724 .192
.346
.248
· 1 19 .350
.413 .424 1 .085 .769
.582
.339
.076 .214
· 1 1 7 .324 1 .438 .512
.250
.854 .686 .639
.594 .562
.505 .844
.798 .058
.942
.729
.136 .663 .213
.183
.095
.317 . 161
.335
.546
.043
.272 1. 398
.416 .018
.147 .422
. 141
.449 .857
· 157
.473
.816 .265 .522
.959
.426 .300 .007
.437
.304 .395 1 .421
.514
.861
.090 .776
.224
.249
.489 .859
. 118
.237 .867 .485
.353 .486 .403 .019
.919
.215 1. 025 .292
.575
.345
.176 1 .035
.361
.320 .637 .508
.154 .219 .065 .183
.001
.201
.305 1 .641
.090 .072 .636 .340
.194
.005 .428
.096
.850 . 781
.627 .214
.344 .248
.489
.040
.007
.580
.120
.816
.208
.154 .026 .548
セNPSQ
.129
.886 .106
.354 1 .056 .300 .800
.017 .470
.436
.154
.357 .246
.374 .042
.052 .367
. 114 1 .140
.856 1. 196 .309
.063
.296
.349
.829 .344
.159
. 117
.754 .133
.929 .801
.256
.897
.648 .113 .090 1 .516
2
.780
.874
1 .004
.551
. '90
.900
.684
.836
.619
1 .687
2.134
1 .279
1 .222
.486
.542
.501
.193
.930
.323
.474
1.628
.649
1.227
.456
.474
.169
.779
. 119
. 161
.938
.909
.251
.627
1 .241
.807
.010
.234
.956
.786
1 .399
1 .528
.675
.485
.342
.418
.329
.103
.082
.092
1
.756
2.675
1. 622
3.138
.599
3.019
.707
.744
2.252
.993
.082
.459
.317
.739
1 . 160
3.119
1 .260
.045
2.246
.587
.510
.628
2.957
2.121
1.479
1 . 185
.160
3.197
3.240
1.160
1.917
.256
1 .313
.071
1.817
4.266
.508
3.101
.028
.617
2.797
1 .601
.346
1 . 769
.670
1 .134
2.970
2.656
2.224
Axis
AX1S
AX1S
219-
Year
AD
1143
1144
1 145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1 161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
Axis
AX1S
7
6
.596
.041
.442 .372
.210
.399
.368 .:230
.434
.71 2
.439 1 .149
. 1 13
.306
. 119
.302
.338 .028
.337 1 .345
.273
.968
.645
.379
.209
.361
.158
.445
.304
.566
.164
.172
.539 .615
.479 .466
1 .359
.157
.125 1.010
.611
.407
.836
.258
.309
.481
.241
.365
.053 .533
.080
.302
.308 .037
.069 .444
.322 .230
.109 .799
.218 .924
.161
.114
.144
.050
.211
.400
.390
.122
.008
.350
.150 .278
.639
.350
.786
.293
.528
.404
. 114
.281
.727 .477
.185 .092
.046
.266
.439
.132
.089 .138
.076
.094
.083 .748
.072
.596
Axis
Axis
Axis
AX1S
AX1S
5
4
1
3
2
.185 .051
.015 .103
1 .546
.:220 .384 .102
.488 .122
.741
.074 1.201
.995 .716
.430 .001
.420 .185 4.160
.412
.794
.385
.01 1
.538
.041
.733
.880
.31 2 1 .768
.193 . 157 .526
.457
.325
.024
.042 .097
.360 4.472
.107
.029 .437
.261
3.381
.494
.107
.217 4.588
.203
.550 .238 .422
.927 .562
.511
.061
1 .037 .426
.277
.037 .357
.653 .497 1.118
.442 .330 .346
.755 3.914
.004
.628 .797 1 . 133 1 .620
.495
.177
.813 . 114 2.239
.366 .624 .192 .061 3.180
.160 .179
.465
.290 1. 193
.047
.684
.698 .259 3.411
.438
.009
.485
1 .949 5.430
.529 .112
.756
.020 2.244
.493
.787 ":".008 1 .427 .681
.186 .265
.624
.527 .461
.144 1 .502 1.089 .623 1 .950
1 .381 1.967
.447
.590 .605
1 .058 2.461
.303 .565 .615
.132 1 .086
2.217
.335 .461
. 131
.474 .714
.087
.034
·.904 1 .071 3.389
.637
.230
.416 1 .014
.017 .741
.318
.100
.733 1. 110
.195 2.290
.270
.405 .071
.556 2.511
.676 1.210
.490
1 . 792
1.298
. 177
.572
.709 .444
.095
.146 .462 .549
.069 3.068
.548
.100
.272
.036 .989
.008
.441
.510
.648
.022
.403
.364 .685
. 191
.703
.554 1.294 1 .399 1.099 1 .426
.192 .666 .345 .385 2.861
.317 .296 .103 .190
.202
.221
.612 .659 .984 3.863
.214 .093 .400
.237 1 . 186
. 121
.006
.738
.120 4.696
.510 .252
.456 1 .493 .689
.134
.450 .222 .456 .235
.169 .262 .731 1 .068 1.589
.066
.042 .918 .032 1 .109
.612 2.618 .255
.075 .062
220-
Year
AQ
1192
1193
1 194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
Axis
Axis
7
6
.346
.327
.914
.342
.341
.219
.641
.379
.042
.232
.683 .035
.269 .257
.154 .354
.022
.516
.461
.825
.316 .531
.303 .263
.056 .170
.699
.150
.410
· 158
.370 .843
.340
.051
.010 .060
.217
· 111
.005 .907
.081
.346
.659 .923
.101
.142
.174 .112
.107
.034
.354
.412
.260
.624
.255 .611
.426
.580
.307 .069
.240 .430
.215
· 110
.059 .048
.108
.971
.195 .206
.157
.108
.254 .327
.248 1 .317
.627
.149
.068
.157
.286
.434
.511 .525
.662 .583
1 .233 .769
.769
.525
.281
.886
.318
.563
.421 1.278
.091
.217
Axis
Axis
Axis
Axis
AXis
4
1
3
5
2
2.791
.098 .445 .854 .621
.032 .640 .657
.004 .624
.454 .646 .329 .592 1.215
.248 .497 .416 2.436
.551
.586 .482 .415 .789 2.171
.315 .750 .703 1.271 2.095
.593 1. 713
.061
.005 .090
.616
.040
.252 1 .334
.922
.134 .031 3.549
.596 1 . 128
.444 2.597
.226 .766 .084
.210 1.145
.798 1 .029 2.249
.405 1 .528
.430 .698 .523
.376 .098 .066 .925
.104
.304
. 119 1 .017 .298 2.901
.357 .296 1 .233
.288 .421
.822 2.687
.143
.656 .215
.077 .128
.526
. 114
.046
.323 .772 2.488 .504 3.005
.029 .194 .323 .094 .727
.737
.473 1 .568
.661
.035
.750
.010
.362
.292 .211
.810 1 .039
.506 1. 1 52
.059
.083 .464 .294 .239 2.525
.992 3.044
.375
.537 .258
.033
.898
.066 2.578
.356
.086 .371
.343 .453 4.795
.135 .160
.728
.220 1 .331
.205
. 121
.647 .170 2.334
.328 .023
.336 .467 1 .607
.502
.097 .323 .625 2.237
.360 .678 .261
.143 .058
.865 .040
.225 .536 .782
.147
.350
.095 1 .465 1 .224
.713 .850
.878
.026 .535
.352 1 .074 .011
.168 .617
.347
.348
.406
.252 4.392
.177
.491 2.383
.532
.660
.220 .684
.709
.497 1.778
.772
.066
.280
.681 1 .489
.417
.537 .127
.066 .660
.243
.079
.757
.642 1.925
.197 .096 .523 .262
.930
.065
.945
1 .064
.067 .567
.196
.107
.477 1.169 2.155
.224
.386
.013
.233 1.503
.215
.644
.882
.149 2.034
.795 .546
.084 .221 2. 137
.450
.323 1 .138 1.029 1.056
.434 .119 .331
.116 2.473
221-
Year
AD
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
Axis
7
.251
.484
.149
.438
.055
.308
.902
.662
.758
.037
.104
.635
.521
.335
.069
.753
.643
.243
1.066
.291
.339
.534
.305
.418
.362
.339
.039
.133
.503
1.542
. 161
.613
.350
.466
.465
.210
.577
.108
.606
.768
.357
.198
.284
.308
.502
.089
.487
.720
.311
Axis
6
.178
.177
.370
.034
.505
.445
.902
.358
.234
.804
.312
.189
.231
.180
.073
.493
.170
.567
.412
.532
1 .493
1 .180
.555
.795
.074
1 .158
.796
.736
.941
.762
.556
.280
.389
.434
.385
.492
.637
.278
.169
.658
. 116
.430
.128
.552
.148
.078
.177
.718
.001
Axis
Axis
Axis
5
4
3
.319
.630 .320
.089 .053 .140
.891
.182
.095
.125
.430
.859
.476 .718
.136
. 161 .457
.065
.004 .572
.379
.328 .274 .981
.021
.518
.568
.216 .037 .967
.159 .561
.352
.677
.007 .050
.088 .191 1.232
.156
.189
.549
.922
.555 . 151
.338 .094 .551
.280
.285 .325
.310 1 .050
.162
.265
.975 1 .352
.222
.268 .807
1 .353
.317
.941
.035
.043 .967
.143 1 .062
.521
.390 .702 .284
.469 .367 .644
.594
.972
.658
.012
. 117 .378
. 135 .140
.373
.515 1. 209 セ Q N V R W
.576
.018 .588
.242 1 .042 .535
.480 .323 .893
.466 1 .269
.415
.335
.574
.237
.841
1.129 1 .255
.376 .405
.195
.617 1.106 .092
.708 .175
.700
.747
.216 1 .183
.468 .412 .181
1 .078
.382 .248
.411
. 166 .218
.828 .155
.697
.265
.130 .750
.373
.102 .805
.331 .197 .354
.067 .162 1.370
.193
.245 1.168
.419 .613 .707
222-
Axis
2
1 .364
.356
.998
.588
1 . 176
.107
.001
2.634
.393
.500
.872
1.399
.035
.053
.786
.741
.574
.310
. 141
.775
1 .500
.481
.666
.460
.518
.641
.227
1 .514
.545
.653
.825
.765
.147
.300
.297
.507
.683
.429
.389
.590
.298
.818
1 .575
.317
.894
1.425
2.323
.517
.506
Axis
1
3.654
.871
.040
2.043
.717
3.137
.342
.362
1 .830
1 .584
4. 141
1 .747
1.339
6.078
1 .946
.875
1 .954
4.493
4.559
1 .044
.380
2.636
2.257
.992
.655
1. 398
1 .551
3.014
.086
.126
1 .184
.015
1 .969
.899
3.821
4.534
1.052
3.490
2.051
4.473
1.221
1.352
.066
1.680
.864
2.244
.706
3.600
.612
Year
AD
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1 311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
Axis
Axis
6
7
.470 .009
.543 .414
.215 .876
.275
.409
.486
.106
.109
.060
.412
.319
.192 1.106
1 .074
.219
.358 .308
.417
.707
.655
.792
.734
.154
.596
.160
.323 .783
.098 1 .230
.210 .503
.096 .263
.542
.288
.262 1 .087
.192
1 .029
.342 .091
.040
.266
.531
.453
.008 .640
.048 1. 939
.077 .277
.330
.189
. 151
.641
.592 .396
.072 1 .261
.107 .182
.346 . 11 1
.001
.220
.342 .493
.526
.323
.416
.302
.285 .038
.633
. 128
.173 .330
.107
.421
.209 .094
.537 .371
.063
.308
.201
.287
.243
.045
.142
.305
.093
.324
.326 .173
Axis
5
.803
.435
.755
.461
.012
.200
.268
.373
.339
.150
.335
.275
.149
.002
. 175
.553
. 211
. 119
.075
.087
.174
. 185
.414
.039
.210
.762
.545
.549
.340
.033
.310
.416
.505
.203
.230
.323
.169
.086
.024
.386
.441
.353
.686
.739
.060
.059
.506
.373
.327
Axis
4
.053
.017
.380
.715
.244
.134
.421
.681
.108
.674
.017
.504
.867
.035
.018
.675
.388
.426
.656
.927
.012
.572
.857
1 .031
.549
.859
1.136
.050
.425
.467
.096
.380
. 113
.248
.179
.535
.389
.351
.257
.601
.125
.407
.248
1 .468
1 . 764
.180
.519
.079
セXUY
223-
Axis
Axis
Axis
1
2
3
2.153 1. 118 2.810
.687 1 .924 1 .607
.249
.955
1 .308
.223
1 .293 2.566
.241 1 .252 1 .415
.695 2.195 2.427
1 .300 1 .441
.301
.548 1 .467 .863
.258
.564 3.113
1 .657
2.100 2.314
1 . 108 1 .381
.533
セNWU
.093 1.463
.020
.266 2.039
1 .334 1.810 1. 137
.250 1.008
1 .697
.189 .781 3.775
.003 .262
1 .388
.446
1 .619 1 .413
.993 . 111
1 .038
.120 .598 3.514
1 .580 1.059 3.003
.515 .474
.244
.734 1 .964 .518
.422
.150 4.432
.545 1 .061 1.013
1.212 .978 3.368
1 .548
.476 3.229
.743
.826 1.596
.432
.414 2.352
.698 .331 1 .261
.681
.655 .642
.263 .412 2.125
1 .375 .385 .443
.695 .640
.625
.582 .059 1 .921
1 .068 .316 4.279
.827 1 .343 .281
.721
.208 .327
.109 .260 2.154
.933
.959 .180
1.190 .933 3.199
.894 1 .060 .326
.372 .024 2.891
.845 .623 2.910
.659 1 .044 .504
.440
.385 2.855
3.133 2.636 2. 159
.005 1.634 1 .029
.150
.842 3.885
Year
AD
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
Axis
7
.158
.289
.425
.065
.096
.330
. 11 7
.256
.205
.260
.109
.505
.027
.074
.139
.376
.536
.256
.014
.005
. 182
.382
.052
.041
.090
.468
.126
.060
.236
.305
.745
.254
.164
.112
.096
.072
.346
.446
.062
.267
.104
.089
.016
.609
.069
.496
.004
.226
.292
Axis
Axis
Axis
AX1S
6
4
5
3
.013 .506
.869
.077
.106 .272
.202 .426
.446
.280 .186 .905
.245 .103
.212
.240
.637
.341
.816
· 135
.033 .155
.082
.395
.036
.066
.629
· 167 .
.473 .953
.409 .366
.076
.312 .091
.255
.663 .360
.098 .294
.166 .399
.895 1.127
.702
.595
.204 .078
.494 .312
.885 .015
.289
.323 .039 .262
.262 . 196
.285 1 .743
.793
.127 .455
.301
.437 .323 .270 .988
.423 .158
.362 .578
.722 .355
.865
.936
.933
.041
.468
.558
.577 .252 .272
.740
.650 .122 .178
.015
.767 .675 .154
.445
.466 1.007
.484
.226
.380 .181
.516
.466
.028 .023
.416
.540
.221
.602 .514 .346
.247 .363 .442
.126
.400 .209
.015
.519
.730 .674
.751
.583
.311
.808 .018
· 196
.365 .479 .744 .774
.823
.138 .083
.288
.725 .800 .004
.015
.590
.013
.821
.289
.350 .387
.479 .836
.565
.080 1 .232
.062
.378
.954 .128 .373
.340 .751
1 .252
.098
.175 .239 .253 1 .063
.038 .499
.494 .773
.329 .167 .773 1.205
.244
.069
.618 .143
.583
.375 1.147 1.514
.453 .734 .554 1 .839
.155
.228 .379
· 11 3
.612
.376
.241
.310
.374 .349
.135
.206
.327
.635
.052
.370
Axis
2
.494
.208
.193
.616
.477
.911
.333
1 .297
1 .078
.055
.227
.741
1 .603
1 .522
.370
.375
.503
.21 7
1.375
.074
.661
.079
1 . 147
.190
.202
.505
1.048
1 .098
. 125
.902
.510
.284
.095
1 .018
.281
.108
.256
.055
.600
.898
.017
1 .887
.044
2.481
.013
.450
.930
.019
.184
1
1 .834
2.740
1 .919
.925
.893
.846
.392
.375
4.198
.258
1 .567
1 .446
.458
2.940
3.461
1. 895
1 .621
2. 114
1 .055
3.610
2.475
1 .821
.397
2.123
2.658
1.158
.386
.065
.045
2.577
3.051
2.897
.581
1.632
.298
2.048
2.089
2.805
. 113
1 .060
.749
2.148
.207
.484
3.207
.187
1.072
1 .069
.349
Axis
224-
Year
AD
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
. 1433
1434
1435
1436
AX1S
7
.280
.280
.172
.757
. 111
.029
.261
.379
.292
.352
.328
.200
.460
.219
.203
.117
1.043
.051
.594
.253
.038
.148
.203
.013
.471
.767
.307
.404
.058
.136
.561
.787
.344
.402
.401
. 199
.086
.040
.007
.623
.424
.562
.011
.308
.092
.516
.064
.424
.078
Axis
Axis
5
6
.904
.812
.001
.596
.823
.470
.194
.886
.468
.627
. 113 1 .168
.102
.839
.790
.51 6
. 151 .756
.095 .456
.027 .494
.892
.309
.525 .172
.364 .206
.189 .243
.028
.109
.434
.185
.380
.466
.339 .408
.074
.235
.011
.468
.554 .750
.588 1 .330
.384
.467
.703 .632
.437 . 151
.407 .957
.020 .020
.084 .259
.441 .744
.392 .427
.164 .130
.088 .144
.713
.198
.389 .380
. 71 1 .425
.139
.065
.040 .242
.348
.406
.028 .520
.990 .248
.247
.575
.063 .015
.334
.863
.510
. 113
.356
.719
.008 .576
.720 .031
.009
.521
AX1S
Axis
Axis
Axis
1
4
3
...
.438
.136
.025
.318
.417
.339
.292 1 .358
.421
1 .051
.034 3.597
.019
.315 .006 .045
.420
.737
.646 1.479
.369
.223
.176 .887
.102 .167 .200 .365
.231
.167 1 .241 1 .446
.284
.574 .036 1 . 164
.079 .664 .297
1 .909
.874
.175 1 .691
.937
.178 .170 .710 3.814
.;.. . 179
.378
.211
2.527
.139
.981
.152
.122
.083
.770
.294
.997
.159
.834 1 .519
.168
.858 .33"3 1 .020 2.619
.071
1 .008
.065 .196
.307 .475
.379 1.876
.532 .461
.667 2.906
.829 1 .546
.098 .393
.155 .239 .907 3.097
.585 .683
.898 1.215
.357 2.850
.578
.392
.274 .183
.537 3.298
.174
.077 1 .419 4.222
.073 1 .426 1 . 105 3.770
.004
.886 .839 3.460
.061
.645
.557 .381
.547 .762 .280 .829
.621
.593 .296 2.608
.141
.990
1 .053 .762
.529 1 .454 1 .367 .160
.175 1 .052 1 .981
1 .745
.447
.704
.803 1 . 701
.141
.255 .478 3.417
.246 1 .272 1 .924 1 .051
.514
.291 1.378
.867
.490
.991 1.171 2.860
.223
.051
.358 1 .824
.915
.677 4.557
.322
.546 .375 .960
.383
.015
.234 .157 .244
.519 .138
.242 1 .459
.207
.282 1 .598
.235
.487 セ N P Q X
.319 3.363
.663
.381
1 .076 1 . 186
.947
.574 1 .047
.337
.090 1.272 .011
.956
')
225-
Year
AD
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
Axis
7
.581
.495
.467
. 118
.178
.018
.201
.045
.327
.392
.168
.176
.227
.105
.459
.224
.597
.072
.176
.571
.395
.521
.246
.293
.020
.060
.494
.449
.134
.446
.147
.013
. 127
.041
.246
.143
.215
.437
.289
.724
.030
.833
.259
.054
.615
.421
.669
.531
.360
Axis
6
.125
.169
.537
.056
.145
.326
.125
.891
.376
.045
.227
.703
.572
.205
. 113
.407
.302
.071
.434
.338
.308
.452
.356
.583
.343
.989
.499
.152
.675
.353
.127
.404
.332
.147
.186
.870
.616
.136
.504
.067
.007
.899
.059
.178
.408
.568
.642
.690
.490
Axis
5
.652
.680
.199
.134
.550
.556
.749
.148
.582
.324
.951
.327
.311
.023
.329
.007
.476
.344
.366
.364
.631
.425
.050
.827
.452
.236
.457
.302
.044
. 161
.763
1 . 147
.826
. 711
.328
.120
.345
.412
.194
.335
.318
.143
.171
.396
.519
.469
.437
.004
.589
Axis
4
.581
.089
.821
.186
.236
.348
.050
.032
.247
.431
.119
.283
.333
.266
.345
. 151
.245
.719
.282
.221
.613
.040
.436
.522
.344
.499
.558
.005
.430
.127
.454
.187
.418
.432
.395
.280
.285
.279
.303
.725
.058
.182
.002
.048
1 .334
.423
.006
.452
.107
226-
Axis
AX1S
Axis
3
1
2
.339
.21 6 .004
.109
.572 3.282
.051
.712 2.285
.345
.160 .502
.200 .269 2.536
.297
1 .881
1. 736
1 .064 1 .535 .924
.445 .683
.281
.122 .609 2.998
.586 2.159 .980
.176
.923 .408
.843
.728
.996
.444
.039 2.844
.406 .084 2.558
.753 .483 1 .843
.129 .129
.593
.463 .408 .782
.337
.165
.307
.354
.973 3.961
.695 .524
.404
.272 1.199 3.278
.923 .945 3.337
.248 .612
1.086
1 .268 .617
. 186
1 .406 1. 179 2.022
.034 1. 379 2.244
.092
.085 1 .997
.820 1 .052 3.183
.. 369 1.150
.449
.387
. 161 3.081
.016 .691 2.181
.643
.832
.611
.306 .322 1.723
.013 .536
.848
.198 . 197 4.219
.484
.878 .528
.597
.529
1 .895
.397 .763 1 .255
.315 .542 3.479
.378
.325 .542
.176 1.760
.602
.045
.707 1.599
.343 1 .256 1.699
.259 .276 4.609
.061
.105 1. 929
1.189 .340 .982
.193 .431
.762
.625 .370 5.364
1 .083
.564 2.560
Year
AD
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
Axis
Axis
Axis
Axis
AX1S
Axis
Axis
7
4
6
2
1
5
3
.570 1 .459 1.304
.524 .786
.543 2.439
.402 .441
.185 .253 .573
.282 2.807
.920 1 .091
.971
.276
.059 .771 1. 782
.415 .440 .261
.377
.429 .051 1 .522
.358 .324
.518
.815 .070
.103 .948
.289 .075 .153 .147 .269 2.002 2.830
.382
.178 .234 . 137
.189
.867 .954
.017
.174
.616 .688 1 .007
.642
.033
.551 .792
.579 .125
.634
.537 1 .071
.703
.145
.543 6.018
.256
· 110 .425
.298
1 .003
.868 .044
.866 .427 .226
.108 .278
.076 .089 .269 .165 1 .362
.542
.259 .439 3.829
1 .028 .383 .657
.406
.143 .773
.334 1 .072 .976
.695
.471
1 .327
.070
.751
3.997
.070
.688
.349 .509
.241
.062
.405 .286
.165
. 161
.018
.052 .964 .720
.254 .136
.168 .217
.145 .750
.104 .062
1 .082
.535 .138
.904 .746 .606
.724 .022
.310 1.007 .204
.056
.807 1 .080 1 . 1 70
.127
.632
.109
.125
.439 4.541
.169
.054
.039 1 .467
.208 1 .214
· 112 .031
.688 .083 1 .483 .624
.714
.654 .133
.164 .221
.853
.993 1 .006
.202 .241
.082
.157
.514 1 .350 1 .650
.076
.599
.778 .830 .866 1 .047 1 .660 .206 3. 110
.387
.572 .791
.037
.610 .699 1.078
.247
.581 1 .460
.084
.614 3.292
.618
.335
.573 .143
.104 1 .094
.399 1 .606
.061
.078 .695 1 .475
.941
.474 2.556
.050
.011
.439
.317
.930 .828 5.238
.084 .667 .402
.080 .593
.211
2.299
.032
.870
.988 1 .344 .773
· 113 .807
.807 .466
.700 .626
.157 .194 1.663
.333
.536 .337 .072 .265
.337 1 .704
.095 .615 .178
.073
.908 1 .921 1.671
1 .088
.041
.456
.791 .034
.282 3.674
.534 .858 .844
.348 .566 .300 1.259
.316
.209
.389
.562 .250 .904 2.115
.584 .633
. 115
.274
.057 .043
. 11 5
.436 .366
.252
.690 .321
.853 '1.261
.098
.300 .418 .795
.231 .274
.096
1 .032 .407 .119 1 .043 .098
.208
.228
.667 .706 .911
.731 .319
.972 1.751
.326
.498
.252 .231
.123 .024 .058
.139
.372 .160
.705
.134
.348 1 .106
.245
.124 .075
.109 1 .367
.768 2.862
.498
.043
.762 .109 1 . 133
.361
.273
.252
.124 .051
.218
.245
.544 .557
227-
Year
AD
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
Axis
7
.259
.202
.103
.387
1 . 103
.927
.877
.192
.956
.384
.364
.626
.145
.443
.125
. 110
.097
.338
.692
.725
.517
.199
.251
.314
.406
.355
.232
.569
.636
.270
.436
.569
.363
.299
1 .047
.606
.143
1 .029
.084
.534
.224
1 • 149
1 .736
.071
.381
.051
.083
.073
.402
Axis
6
.207
.319
.006
.087
.291
.501
. 199
. .180
. 141
1.113
.094
.429
.682
.748
.230
.357
.325
.220
.059
.487
1 .099
.140
.082
.097
1 .288
.584
.323
.429
.201
.521
.252
.249
.261
.395
.330
.098
.720
.267
.005
.567
.220
.666
.419
.087
.436
.159
.131
.506
.564
Axis
5
.207
.998
.185
.466
.001
.364
.409
.099
.225
.840
.738
.379
.384
.169
.265
.736
.446
.088
.011
· 114
.926
· 122
.032
.679
.037
.414
.093
.580
.074
.117
.429
· 115
.781
.515
.223
.960
.675
.802
· 166
..320
.422
.135
.590
.284
.002
.233
.072
.027
.168
Axis
Axis
Axis
AX1S
4
1
3
2
.472
.656 .832 1 .748
.455 .150
.318 1.714
.603
.523 .351 1.761
.552
.479
.776 2.335
.019
.496
.769 1. 961
.557 .444 1 .463 2.752
.090 .243 1 . 109 .897
.044 1 .31 1 .124 4.586
.169 .172
.879 2.909
.747
.583 .246
.480
1 .028
.583 .801
1 .880
.693 .,...304 .813
.840
1 . 136 .499
.679
.431
.298 . 161 .559
.767
1 .007 .000 .787
.382
.522 1.010
.190 2.041
.207
.154 .692
1 .699
.626 .992
. 111
.249
.344
.440 1.800
1 .301
.143 . 131
1 .394 1 .318
.801
.242
.708 1 .495
1 . 103 .823 1 .244 1 .931
.504 .025 .362 1 .780
.407
.633
.299 .331
.403
.710
.358
.127
.066 .825 1.667 2.315
.376 .681
.997 1 . 701
.251
.355 .750 2.812
.369
.052 .989 .306
.309 1.164 .496 .360
.007
.764 .395 2.133
.146
.668
.521
1 .318
.055
.196 2.094
.220
1 .083 .261
.280 .129
.766 1.150 1.417
.282
.391
.372 .245 1 .831
.218 1 .071
.836
.290
.496 1 .525
.442
.874
.254 .201
.288 4.757
.487 1.100
.449
.254
.347 .420
.400
.739
.417 .909 .014 2.065
1 .597 2.252
.034
.241
. 11 1 1.278
.509
.483
.166
.427 .613 4.307
.774 3.362
.358 .901
.334 .074
.061
1 .324
.412 2.154
.038
.056
.335 .134 .385 3.664
228-
Year
AD
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
.1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
Axis
Axis
7
6
.314 .389
.424
.421
.027 .156
.219
.521
.302 .076
.779
.358
1.012
.120
.243
.339
1 .253
.862
.923 .135
.514 .041
.347
.202
1 .349 .057
.274
.398
.064
.305
.075
1 .905
.345
.982
.651
.298
.372
.239
.245 .712
.618 .361
.043 .265
1 .066 .304
.242
.245
.511
.134
.436 .326
1 .040
.422
.588
.274
.463
.145
.066
.142
.656
.555
.075 .363
.012
.644
.369
.153
.648 .023
.342
.115
.651
.726
.014 .303
. 119
.532
.540
.346
.081
.859
.311
.024
.469
.507
1 .062 1. 192
.169
. 180
.028 .013
.124 .259
.278
.235
1 .138 .246
Axis
Axis
Axis
Axis
Axis
4
5
3
2
.370 . 177
.265
.930
.184 .095
.041
.343
.292 .806 .902
.005
.057
.086
.583
.422
.514 .142 .393 .330
.666
".312 .740 1 . 155
.91 1 .052
.273
.229
.213
.027
.522
.655
1.077
.073
1 .021
.145
.368 .350
.352 1.176
1 .088
.069 .436 .683
.087
.064 .372
.899
.264 2.213
.453
.009
.755 .950 .381
2. 135
.104 .328
.208 .496
.224 .274
.536
.434
.734
.819 .007
.158
.155
.001
.518 1 .495
.346 .439 .391
.588
.240
.022
.800
.628
.783
.058
.508
.298
.220 .077 .104 .465
.278 .304 .033
.062
.575 .009
.284
.667
.707
.260 .782 .899
.140
.256 .245 .198
.404 .966
.238
.836
.041
1 .015 .477
1 .267
.055
.495
.298
.882
.581
.904 1.911
.511
.287
.382
.051
.214
1 .393 .629 .124 .363
.024 .224 .599 1. 243
.703
.782 .743 .778
.528
. 196 .909 .200
. 151
.266 .391
.969
.644 .134 .740
.650
1 .512
.105 .020 .058
.494 .060 .269
.474
.774
.317 .012 1 .855
.013
.828 .445 .381
.304 .330
.506 .534
.345
.471
.231
.580
.238
.318
.082 .193
.148
.299 .137 .780
.008 .345 .284
.691
.019
.722 .555 .075
.325
.006 1. 135
.073
.158
.328 .042
.937
1
2.551
4.020
.662
1 .903
.892
. 110
3.484
229-
セNVWX
1 .994
3.143
3.804
1 .060
2.637
1.137
2.566
2.327
3.394
.846
.026
1 .091
.043
1.871
.136
.844
1.627
1.788
4.584
2.507
.690
1.679
.400
.602
.730
2.964
2.825
1 .520
2.908
3.191
.543
.989
2.379
2.430
1.423
2.472
1 .542
2.369
1.208
1 . 181
2.448
Year
AD
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
Axis
7
.636
.312
.942
.005
.946
.006
.393
.350
.609
.200
.401
.342
.294
.166
· 1 19
.776
.214
.247
.462
.795
. 151
.454
.636
.053
.184
· 174
.318
· 171
.506
.822
.285
.073
.420
.350
.008
.068
.021
.301
.213
· 162
· 169
· 192
.068
.246
.159
.705
.317
.109
.843
Axis
6
.849
.281
.405
.668
.713
.003
.096
.222
.310
.448
.371
.216
.685
.110
.995
.280
.477
.220
.514
1 .391
.720
.662
.440
.780
.471
.166
.676
.538
.050
.430
.400
.610
.087
.095
.205
.404
.589
.027
.638
.525
.205
.864
.772
.274
.096
.749
. 101
.859
.089
Axis
Axis
Axi s
Axis
Axis
5
3
4
1
2
1 . 1 14
.169 .510 .541 1 .709
.066 .064
.049
.372 1 . 144
.258 .626 1 .056 .345 1 .360
.872
.159 1 .898
.153
.492
. 761
.122 .176
1 .006
.674
.580 .295 .357 .823 2.322
.459
.993
.050
.747 1 .522
.390 .162 1 .676
.443 .247
.012 1 .481
.478 .341
.617
.634 .268 .164 .186 1 .025
.195
.103
.053
.453 .774
.015 1 .616 .218 .327
.263
.517 .609 2.025
.613 .388
. ·148
1 .954 .074
.204 2.677
1 .402
.687
.243
.100 1. 839
1 .294
.787
.507 .094 2.179
.104 .204
1 . 105
.314 1 .407
.417
.246
.216
.814 .419
.202
.996
.086
.010 3.629
.547
.288 1 .273 1.204 .700
1.578 .717 .025 .810 .378
.527
.678 1 .280 2.390
.258
.871
.012 .291
.346 3.206
.398 .402
.243 .895 .314
.105 .619 .053
.610 2.556
.143 .168
.221
.449 1 .846
.303
.153 .000 .719
.848
.308
.464 .531
.048 1. 660
'.048 1 .537 .556
.450
.530
.290 .437
.780
.448
.991
.210
.510
.297
.271
.322
.557 .384
.350 .330 3.619
.420 .478
.027 .727 1.749
.299
.189
.629 .170 1 .929
. 175
. 114
.059
.218 2.860
.058 .248 .583
.332 2.384
.378 .274
.886 .474 1 .239
.609
.251 1 .073
.086 1 . 779
.480 .142 .352 .363 .029
.190
. 116
.107 .369 . .258
.230
.089 .166 .461 1.747
.541 .407 .700 .209 .760
.543 .651 .267
.230 .978
.229
.130
.427 .044 3.046
. 101 .231
.454
.294
.106
.434 1 .441 .366 .608
.233
.463
.218
.927
.553
.534
.541 .285
.334
. 181 2.740
.572
.073 .098 1 .271
.613
230-
Year
AD
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
Axis
7
.425
.349
.304
.162
.440
. 11 7
.114
.119
.010
.086
.104
.061
.158
.077
.055
.106
.059
.235
.151
.217
.187
.221
.431
.562
.088
.144
.378
.136
.079
.583
.359
.305
.046
.467
.203
.201
. 110
.001
.144
.103
.570
.029
.124
.035
.013
.118
.012
.545
.076
Axis
Axis
Axis
4
6
5
3
.209
.246 .474 .304
.092
.132
.963 .236
.712 .540
.080
.021
.162 .045
.743
. 191
.257 .424 1 .018 .640
.015
.472
.193 .227
.412
.493
.955 .357
.754 .264 .013 .242
.438 .005 .379 .928
.036 .255 .107 .756
.684 1 .127 .175 .147
.496
.015
.375
.104
.738 .428
.660
.195
.782
.105
.143
.045
.344
.049 .004
.075
.341
.039
.823 .095
.376 .002 .556 .415
.179
.651 .308 1 .170
.712 .228 1 .304
.405
.545
.501
.720 .313
.347 .400
.612
.172
.885 .160 .680 .032
.682
.188
.396 .232
.186
.055 .186 .414
.860
.275 .934
.139
.511
.237 .059 .066
.061
.323 .555 .323
.104 1 .014 .170
.510
1 .304 .575
.378 .526
.401
.012
.129 .469
.463 .299
.054
.237
.132
.060
.208 .698
.312
.997
.353 1 .021
.412 .624
.257 .522
.377 .127 .651
.737
.094 .092
.237 .124
.077
.590 .448
. 111
.350
.100 1.296 .593
.046
.811
.142 .601
.215 1.675 .112 .687
.062 .911 .252
.422
.565
.249
.258 .750
.786 .393
.407 .214
.468
.269
.060 . 181
.008
.368
.401
.243
.489 .190 .477 1 .442
.814
. 116 .000 .635
.436 .106
.579
.666
.048 .381
.549 .319
2
.423
.361
.616
.037
.192
.148
.579
.743
.725
.235
.003
.391
.527
.251
.327
.320
1 .003
1.178
.396
.815
.007
.413
.020
1. 717
.673
.936
.807
.994
.155
.970
.025
1 .341
.956
.774
.164
.669
.616
1 .435
1 .449
1 .364
.899
.121
1 .682
1 .960
.046
. 126
.515
.319
.739
.680
1.811
3.198
5.023
1 .335
.377
.262
3.007
1 .067
.366
3.461
.672
.728
.742
3.349
1 . 190
.782
3.265
.863
3.134
1.488
.363
2.196
.192
3. 161
2.588
.649
.650
3.524
.565
1.825
.544
1 .835
.519
2.890
.180
1 .357
1.529
5. 114
.259
.516
1.985
1.688
2.471
2.985
.247
1 .760
4.854
.045
Axis
Axis
Axis
231-
1
Year
AD
17 31
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
17541755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
Axis
7
. 152
.012
.828
.332
.433
.325
.269
.209
.359
.609
.001
.037
.055
.047
.095
.491
1.033
.251
.265
.414
.407
.685
.004
-.818
-.754
-.222
.103
.562
-.016
.447
-. 131
-.334
.258
-.388
.054
. 1 12
.377
-.438
-.640
-.259
-.740
-.883
-.054
-.013
-.460
-.762
-.202
-.012
-.399
Axis
6
.157
.500
.759
.412
.099
.360
.025
.780
.785
.083
.450
.373
.653
. 116
.504
.500
1 .745
.283
.325
.195
.493
.771
.758
-.587
.009
-.093
-.447
.139
.072
-.104
-.427
.053
-.625
.160
.087
.637
-.637
1 .217
.914
.050
1.438
.917
-.650
-.129
.046
.183
-.194
-.112
-.192
Axis
5
.061
.920
.479
.299
1. 057
1.113
.397
.665
1.131
.417
.449
.162
.527
.569
.206
.045
.360
.274
.742
.524
1.024
.153
.440
.776
.169
.395
-.504
-.106
-.418
.517
-.843
-.007
-.364
.445
-1 .998
1 .065
.265
.361
.984
-.834
-.627
.584
-.696
.814
.332
-.299
.103
-.850
.545
-
Axis
4
.036
.564
.233
1 . 134
.669
.325
.039
.074
.066
.366
.412
.21 7
.170
1 .454
.203
.132
.338
.164
.083
.292
1 .607
.388
.784
1 .003
-.274
-1.243
.366
-.068
.008
1 .311
-.267
.347
-.692
.737
.513
.045
-.548
.242
-.433
.705
.889
1 .248
.557
.428
-.032
-.020
-.712
.167
-.260
232-
Axis
AX1S
3
2
.081
.258
.912 .355
. 161
. 1 14
.497 .222
.919
.491
.081
.196
.006 1. 189
.097 .937
.399 .897
.142
.025
.537 .250
.654
.062
.150 .505
.312
.133
.252 .487
.505
.006
.887
.292
.558
.494
.201
.425
.777 .163
1 .441
1 .453
.299 .144
.557
1 .445
-.364 1 .058
-.580 1 .293
-.227
.148
1 .471
-.856
.666
.232
.869 -.379
-.747 -.373
-.069 1 .311
-.307
.701
-.395 -1.276
-.401
.400
-.258
.565
-.178 -.552
1 .073 1.575
-.207
.125
-.808
.069
-.023 1 .025
-.371
.023
-.851 -.269
.516 -.047
-.264 -.378
.531 -.052
-.486 -.167
-.837
.448
1 .362
.499
-.896 -.502
Axis
1
.092
1.032
2.461
1 .408
3.506
.609
2.603
.829
2.662
.736
.201
.851
2.629
.921
3.227
4.501
4.275
5.993
4.676
1 .654
.875
2.933
.192
-2.074
1 .549
1 .913
1 .986
-.832
-1 .096
.869
-1.181
-1. 694
2.665
-2.436
.863
-2.544
.663
-2.065
-.630
-1 .640
-4.675
-.693
4.609
-.131
.574
.044
1 .946
-.102
1 .870
Year
AD
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
Axis
7
. 161
.713
.038
.025
.691
.842
.303
.532
1 .013
.320
1 . 164
.035
.204
.398
.402
1 .214
. 111
.344
.897
.732
.090
.084
.010
.409
.055
.334
.774
.549
.060
.016
.824
.646
.214
.264
.493
.149
.015
1 .076
.334
1 .019
.761
.407
.622
.531
.253
.215
.513
1 . 190
.396
Axis
6
.083
.139
.422
.516
.057
.510
.671
.526
.173
.309
.479
.025
.425
.333
.290
.237
.871
.098
.225
.710
.553
.380
.376
.098
.486
.242
.280
.203
.255
.061
.850
. 131
. 110
.019
.250
.406
.108
.428
.356
.112
.114
.323
.456
.089
.800
.165
.296
.075
.209
Axis
Axis
Axis
Axis
4
5
3
2
.644
.01 1 .712
.176
.169 .018 .068 .584
.008 .092
.602
.927
.359 .998
.521
.728
.487 .021
.286
.466
1.068 1 . 129 .830
.134
.237
.428
.031 .413
.574
.056
.014
.400
.581
.523 .013
.785
.145 1 .226 .270 .555
.368
.567 1 . 511
.862
.276 .609 .126 .489
.586 .287 .536
.022
.543
.776 . 118
1 .027
.519 .284
.425
.209
.704
.354
.424
.440
.147
.578 .119 .346
1. 293 .277 .459 .607
.318 .583
.301
.801
.51 1 .245
.076 .021
1.693 .604
. 191
.553
.601
.204
.321 .622
.201
.489 .919 . 311
.770 .405
.027
.886
.653 .314 .162 .180
.109 .157
.044
.907
.048 .589 .086 .543
.455
.013
.233 .048
.414
.326
.800
.329
.479 .385 1 .034 .639
.641
. 172 .821
.509
.180 .605
.342 .173
.143 .440 .750 .500
.063
.537
.626
.623
.241
.145
.787 1 . 152
.156
.168
.898 .679
.144 .137 .070 1 . 128
.489 .563 1.173 .634
.090
.701
.280
.873
.087
.236
.724
.341
.653
.962
.017 1 .412
.212 .370 1.549 .463
.732 .468 .507 .921
1.111
.866
.222
.261
.208
.342 .441
.256
.390 .154 .597
.215
.762
.128 1.226 .295
.146 .537 .096
.889
.010
.806
.104
.021
233-
Axis
1
3.350
.050
2.447
2.483
3.099
.848
1 .476
2.288
2.044
1 .267
.011
2.341
1.913
4.809
.845
.099
1 .094
.81 1
.643
.482
.337
1 .527
.430
.729
1 .039
2.058
3.710
2. 152
.912
.511
.517
.681
1.146
1 .001
.020
2.732
5.305
1.770
3.886
3.492
.736
2.110
4.350
2.374
1. 117
.535
.408
1 .091
3.423
Year
AD
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
セXTV
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
Axis
7
.604
.927
.542
.446
.573
.424
1 .222
1 .481
.130
.208
.503
.022
.980
.613
.213
-.027
.155
.276
-.320
.777
.034
.367
.196
1 .207
.153
.582
-.079
.749
.766
.848
.233
.819
.127
.773
.782
1 .420
.269
.241
.571
.262
.067
.368
1 .303
.603
.952
.353
1 .010
.282
-.110
Axis
Axis
Axis
6
4
5
.432
.934 .778
. 611
.516 .043
.769 .464
.682
.747 .472 .329
.227
.182 .102
.016 .108 .520
.81 1
.209
.232
.265
.353
.597
.046
.480
.156
.580 1 .092 1 .008
.401
1.004
.396
.108
.240
.935
.746 .135
.364
.369 .194
.846
.403
.304 1.277
.549
.696 1.984
-.706 1 .450
.488
-.170
.911 1 .485
.597 -.270
.820
.145 .313
.202
.537 1 .911 .482
-.065
.143
.165
.682 -.085
.265
.389
.066
.549
.397
.692 .637
-.890 .135
.173
.620 -.301
.643
.226 .261
.559
.192 .192
.190
.646
.716
.447
.357 . 181
.517
-.307 -.007
.461
.468 -.031
.193
.773
.208
.807
.163 .198 .419
-.180 .511
.075
.093
.128 -.205
.392 1.189 -.304
.196 .228 -.406
1 .075
.953
.653
.762 .997
.079
.935
.354
.388
.163 .453 .187
-.047 -.081
.112
1 .319 .419 -.400
.679
.396
.024
.398
.566 .284
.789
.572 -.800
.802 .499 1 .840
234-
Axis
Axis
Axis
3
2
1
1 .082
.906 1 . 71 2
1 .126 .178 1. 107
.621
.244
.445
1 .489 .424 .593
.232
. 171 , .995
.936 1 .263 1.127
1 .044 .674 .889
.368 1.411
.037
.400
.689 1 . 779
.304 .517 2.901
.359
.338 4.655
.308
.018 4.454
.885
.958 2.256
.461
.028 2.613
.356
.266 .331
.587 .398 1. 862
.586
.720 2.003
.146
.086 .447
.339
.842 4.698
-.506 1.118 1 .321
.477 1.083 -3.034
.745 .839 .795
.532 . 149 3.830
.688 .991 2.965
. 511
.333
.162
.432
.421
-.076
. 191
. 131 1 .859
1.255
.620 1.314
';".510 1 .382 -.079
-.082 .329 1 .833
.558 -.310 2.589
1 .124 -.520 1 .094
.280
.725 5. 177
1.787 1.716 .874
.646
.308
.632
.459
.655 3.247
.495
.145 1 .085
.600
.614 -2.250
.759 .118 -2.550
1 . 191
.195 -3.650
-.208
.341 -2.880
.688
.870 2.232
.653 1 .597
1 .687
.418 1 .641
.431
-.501
.137 1 .457
.078 -.089 -.410
-.220
.353
.702
.382 1 .055 2.299
.744 -.503 2.753
Year
Ap
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
191 1
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
Axis
Axis
7
.388
· 116
.393
.213
· 1 19
.214
.124
.372
.381
.202
.571
1 .416
.390
.675
.306
1 .993
1.922
.445
.843
.261
.639
1 .555
1 .080
.532
.485
1 .503
2.040
.710
2.099
.208
.307
.989
.547
.565
.741
.805
· 110
.278
.750
.028
1 .086
1 .186
1.050
1.215
1 .467
.938
.425
. 113
1.017
6
.724
.126
.743
.107
1 . 132
1 .230
1 .391
1 .001
.458
.206
.822
.970
.662
.075
.705
.692
.433
.419
.432
.686
.315
.524
.918
.310
.018
1. 169
.813
2.021
1.705
.701
.368
1 .245
.681
1.373
.190
.788
.806
.665
.180
.246
.342
.516
.509
.371
.252
.796
.471
.608
.196
Axis
Axis
Axis
Axis
5
4
3
. 113
.057
.316
.246 .671
.795
.200 .377 .413
.034 .661
.064
1 .255
.093 .041
.304 .122
.108
.478
.146 .423
.364 .396 .002
.084 .069 .967
.358 .883 1 .654
.810 .954
.323
.560 .073 .397
.366
.387
.301
.677
.352
.313
.433
.231 1 .427
.559 .124
.178
.046
.658
.516
.461
.915
.339
.713
.550 .345
.827 .678 .853
.544 .293
.015
.501
.210 .001
.027 .639 .307
.347" .945 .008
.557 .681
.877
.645
.315 1. 225
.085 .244 .020
.012 .204
.595
.495
.334
.703
.395 .376
.142
1 .083 .177 2.154
.087 .652 .061
.007
.276 .396
1 .287
.492
1 .079
1.436 1.276 .150
1 .074 .762
.470
.069
. 161
.288
.21 1
.025 .388
1 .824
.630
.540
.526 .332
.274
1 .247
1 .642
.094
.550
.685
.662
1 .066 .154 .206
.708 .193
.200
1.016
.294 .923
.580
.401
.168
.617
.849 1 .035
.718 1 .048 1.279
.697 .916
.895,
2
.600
1 .019
.957
.500
.780
.339
.109
.072
.230
1 .802
1.016
.646
1 .609
.760
.026
.279
.295
.362
.756
.264
.398
.838
.284
.320
1. 713
1 .661
1 . 1 10
.646
1.763
.367
1 .029
2.526
1 .689
1 .229
. 112
.489
.873
1 .326
.875
1 .847
.167
.293
.300
.086
1.326
.266
1 .254
1.151
.492
1
1 .800
1 .537
3.664
1 .104
.824
1 .966
1 .061
1 .482
.166
.250
2.463
.445
1.208
1 .968
1 .325
2.579
2.438
1 . 145
3.968
2.347
.268
4.918
2.233
.307
3.340
.059
4.679
3.244
1 .255
3.542
1. 658
.882
1 .291
3.034
2.013
1 .903
3.357
2.859
3.762
.418
1 .819
4.058
4.550
1 .068
.503
1 .410
1 .457
3.499
3.617
Axis
235-
Yea,..
AD
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
Axis
Axis
Axis
AX1S
AX1S
4
7
6
.5
3
.466 1. 226 .081
.474 .048
.932 .348
.270 .560 .803
1 .430
. 150 1 .488 .168
.134
.493
. 131
. 131
.310
.340
.340
.354
.257 1. 636
.205
.852
.047 .770 .353
.271
. 715 .290 .294
.040
.893
1 .256
.189 .406 1 .029
.158
.071
.052 1 .228 .640
.290
.448
1 .384 .013 .373
.021
.538 .130 .645
.551
.273
.700 1.131
.589 .839
.132
.226
.469 .008 .567 .009
.790 . 911
.231
.216
.433
.034
.060
.502 .695
. 1 16
. 11 7 1 .050 .183
.487 .929
. 114 .807
.417
.130
.529
.331
.575 1.191
.086
.602
.580
.235
.539
.700
.212
1 .944
.377
.472
.498 .321
.807
.914
. 145 .022
.639
.042 1 .051
.393 .155
.220
.367
1.123 1 .342
.170 .248
.181
.431
.363
.061 .154
.064
.011
.837
.465 .322
.561
.251
.782
.415
.152
.389
.245 .645
.170
.401
.301
.081
.577
.170
.351
';".717
.452 .637 .086 .471
.974 .042
.065 .778
.185
.460 .168
.656
.287 .558
.576 .229
.095 1.274 .106
.467 .482
.482 .168
.534
.521
.010 1 .164 .760 .118
1 • 164 .421
.337 .154 .836
.616 .229
.746 1 . 101
.202
.288 .274
.437
.284 .419
.346 .244 .157 .978
.831
.830 .622
.797
.842
.102
. 196 .088 .005 .103 1.386
.976
.182 1.062
.510
.093
. 711
1.133 1 .038 .247
.005
.781
.274
.489 .246
.055
.050
.833 .092 .887
.531
236-
AX1S
2
1. 067
.136
.096
.295
1 .275
.715
.398
.312
.031
.072
.075
1.101
.041
.555
.842
.496
1 .200
.264
. 171
.434
1 . 138
.251
.433
.907
.318
1.939
.803
.183
.849
.235
1 . 794
.506
1 .455
.245
.060
.891
.055
.451
1 .041
.736
.845
.501
.807
.678
,
Axis
.109
.386
.555
.774
.563
3.635
.801
2.382
2.154
.647
2.686
.324
.871
.633
4.352
2.221
.005
.891
.687
3.209
.391
1.621
2.921
4.176
4.651
1. 363
2.996
1 .325
2.666
3.121
.573
1 .053
2.723
1.410
1.444
.260
1 .138
1 .535
2.807
1 .315
2.481
1.555
.615
1 .148
ApPENDIX
B
HISTORIC COMMENTARY ON CLIMATE AND
AGRICULTURE IN MONTEZUMA COUNTY, A.D.
1894-1970
1894
1896
1899
1902
1904
1905
1906
1908
"The season of 1894 was one of water shortage allover the county and there is much talk of doing something about it" (Freeman 1958:248).
"The season of 1896 was another very dry season. Grain crop is short and there is only one crop
of hay. This set the people to thinking about a reserve water supply" (Freeman 1958:249).
"The summer of 1899 [in Mancos] was very dry, no feed on winter range. Severe winter follows
and there is heavy loss of stock from starving" (Freeman 1958:211).
"The Cortez town board has a well drilled for artesian water for a town water supply. They drill
to 594 ft., 8 in., but get not water. The board pays the bill and quits.... Cortez is still having
water trouble.... This was in the spring of 1902.... The dry season brought about a movement
in Cortez to incorporate the town so they can bond and build a water system as the only solution
of the water problem. Sentiment is divided. The town is small and valuation low for such an undertaking" (Freeman 1958:252).
Montezuma Valley Journal, September 5, 1902 refers to the drought of the past three months
and says it is "probably the severest year known in this section of the county ... [yet] one tract
produced 33 bushels of wheat per acre...[and a] ranch a few mile southwest of Cortez averaged 25
bushels weighing 72 pounds per bushel...[and] a good cutting of alfalfa has been harvested south
of Cortez this season.... The fact of the case is many farmers of this valley have been so accustomed to growing 4050 bushels of wheat and 610 tons of alfalfa per acre that they think there
is almost a failure of crops when these figures were lowered by the drought of the present season
(Thompson, personal communication, January 1990).
''The summer of 1902 very dry. Mancos river ceases to flow in August Town wells go dry.
Montezuma Valley and Indian reservation very dry. Water for stock very scarce. Navajo Indians
hard hit They are leaving the reservation and begging for food" (Freeman 1958: 213).
"Winter of 19031904 was very dry and there was almost no snow, even on the mountains.
Some farmers planted no crops in the spring because they figured there would be not water. Most
farmers planted their land and raised pretty good crops. It was cold and ice accumulated in the
mountains, melted and made lots of water for a time. There were some good summer showers"
(Freeman 1958:253).
"As a result of the dry winter and spring of 1904 the spring at Navajo Springs, always freely
flowing, all but went. Plans were made to move the [Ute Mountain Ute] agency if the water
failed entirely" (Freeman 1958:254).
Montezuma Journal, July 29, 1904 records that a statewide weekly crop bulletin reported that
"rainfall has been light and more is needed, especially in the extreme southern counties where severe drought prevails (Thompson, personal communication, January 1990).
"On May 15, 1905, the Mancos river was the highest it had been for many years. It floods the
lower street of town and the water was within six inches of the bridge. The Dolores river was
also on a rampage" (Freeman 1958:214).
"In 1906 the county produced the largest grain crop in its history up to that time" (Freeman
1956:255).
"The years of 1906 and 1907 were years of excessive rain in Montezuma Co. The water commissioner at Mancos was not called out these two years" (Freeman 1958:217).
"There were big crop yields in 1908. Wheat made from 25 to 50 bushels per acre, oats 40 to 80
bushels threshermen report There was also a good hay crop. Range cattle got very fat and an unusual amount of good cattle were shipped" (Freeman 1958:257).
1909
1910
1911
1913
1917
1918
1920
1922
1924
'There was more rain than in any year in recent history, and there was a big crop of everything
grown on the farms: a big fruit and potato crop, good grain and hay crops, thousands of fat cattle
and sheep were marketed, and the market for every product was the best in recent years" (Freeman
1958:257258).
''Rains of early September, 1909, wiped out bridges, roads and railroads; crop damage is
widespread. Never in the history of this part of the state has the country undergone such a persistent bombardment form the elements, and continuous rains and recurring floods wrought disaster
to every industry. Farms, mine, mills, the railroads and highways-all have had their share of
loss. Since July 18 it has frequently rained two days and nights without ceasing. All streams
were higher than they had ever been known" (Freeman 1958:218).
"During the rainy seasons of 1909 and 1911 flood waters came down the [McElmo] canon in
such volume and with such terrific force that it washed away houses and large parts of orchards
and farms causing damage that ran into the thousands of dollars. One farmer lost 40 acres of
grain and several acres of his farm went with it. Two and three acres were a common loss. It was
at this time that the channel widened and deepened to what it is today" (Freeman 1956: 132).
"In 1910 Montezuma County farmers are harvesting a bountiful grain crop, 60 bushes of wheat
and 80 to 90 bushels of oats per acre is not an uncommon yield" (Freeman 1958:261).
"A rain historic in intensity and in its disastrous consequences fell the week of July 10, 1911, in
Montezuma Valley, and well over the county. A number of important flumes, in the irrigating
system, were broken in a dozen or more places, and it was reported that McElmo Creek was the
highest it had ever before been known. The flood worked havoc and the repair work was a big
expense to the district, already burdened with financial difficulties. The rain was already heavy in
the mountains and all streams were on a rampage and much damage was done to roads, bridges
and crops" (Freeman 1958:264).
'The year 1911, with its abundance of rain, produced good crops and many record yields of
wheat, oats, barley and hay were reported from every part of the county. There was an unusually
good crop of corn in McElmo, and in other localities where com is grown. There are report of
yields of 100 bushels of oats per acre, 75 of barley, 40 to 60 bushels of wheat and com up to
100 bushels per acre" (Freeman 1958:265).
"After a season of unprecedented rainfall, a number of minor floods and one major one, the wet
season climaxes by a general downpour on the evening and night of Oct. 6, 1911; resulting in
the greatest flood in this southwestern section since it has been known to the white man
(Freeman 1958:221)." In this wet year, Freeman notes that historic sources report that the rivers
were raging. These rivers included the San Miguel, Dolores, Mancos, La Plata, Animas, Pine,
Piedra, San Juan, Navajo, Chama, and Rio Grande (Freeman 1958:221-222).
'The summer of 1913 very dry. No winter range in the lower country. Some canlemen bought
hay. Others shipped, and shipping was heavy. Plans to restock in the spring" (Freeman
1958:267).
"The year 1917 brought a dry summer. There are good crops, but feed on the winter range is very
poor and many more cattle must be fed through the winter than usual" (Freeman 1958:271).
"On June 21 and June 23 the county had two good rains, the second assuming flood proportions.
Very unusual rains for June, but they were welcome as the previous winter and spring had been
dry" (Freeman 1958:273).
"The spring floods of 1920 were the most destructive in several years. The railroad along the
Dolores river was washed out in a dozen places, and there was heavy damage along several other
stretches of the road in the Basin. Water was so high no repairs could even be started for days. It
was weeks before through service could be resumed" (Freeman 1958:313).
Winter of 1921-1922 was a winter of heavy snowfall at Mesa Verde National Park (Freeman
1958:173).
"There is a good grain crop in the county and dealers report much more binding twine than ever
before is being used. Albert Roessler reports 60 bushes of wheat per acre. He has been on his
ranch nine years, has had a better crop every year, and this year the best ever" (Freeman
1958:233).
"Another good farm, the W.A. Bay place, reported good yields: 98.5 bushels of oats per acre off
of 8 acres, 40 lbs. per bushel; 6 acres of Marquis wheat yields 50 bushels per acre, the result of
good farming for some years. Owner says he is just getting the land in shape again to produce
good crops" (Freeman 1958:233).
-
238-
1926
1927
1934
1935
"It was a rainy season and the growers estimate that 50 per cent of the seed was lost in harvesting" (Freeman 1958:283).
.
"Alfalfa seed growers made a little history in November, 1926, when they shipped 100,000
pounds of alfalfa seedtwo full car loadsfrom Mancos. The seed brought 14 cents a pound or
$14,000 for the shipment. Thomas Coppinger and Talcott and Carpenter were the main producers, with many small producers. In some cases, the seed grown on an acre of land was worth
nearly as much as the land was worth" (Freeman 1958:237).
"In the month of April, 1926, the rain gauge at Spruce Tree Camp registered a rainfall of 7.31
inches, and in May following 1.36 inches was recorded" (Freeman 1958:175).
"Heavy rains of flood proportions, struck the county last days of June. Washed out roads, railroad
and bridges, two railroad bridges on Dolores river above town and badly damaged one below
town. This rain was notable in that is was the heaviest rain ever known to fall in this county in
the month of June, uniformly a dry month with almost no rain" (Freeman 1958:284).
"In the fall of 1927 beans definitely came to the county's farming industry as a main crop. Many
yields of 800 pounds per acre reported and the contract price to the growers was $5 per hundred.
Many fields of 60 to 120 acres were planted to beans. Highest yields were reported from the dry
farming territory near Lewis. Beans were shipped by the car load and many cars were shipped"
(Freeman 1958:306).
"The Ackmen section of the dry lands has more wheat than can be harvested" (Freeman
1958:306).
"A record drought struck Montezuma County in 1934 and was probably the driest year the
county has experienced since it occupancy by the white man. With a very light winter snowfall,
ahnost no spring storms and very scant summer rains, crop production was at the lowest level
the county has ever known and distressing conditions were bound to ensue. These conditions being generally prevalent in the southwest, the government took measures to relieve the situation.
Funds were made available for buying cattle off the rancher thus reducing the nwnber to what
could be taken care of. Buyers came to every ranch and bought cattle paying $12.50 a head for
average cattle of every kind and young stock in proportion. The best of these were shipped into
southern areas where cattle were less plentiful and there was plenty of feed. Hundreds of head of
the poorest stuff in the county were simply driven off to some isolated spot and shot In cases
where stock were mortgaged half of the proceeds were given to the mortgage holder and half to
the owners. There was no flood water from snow and little water in the reservoirs. A Thompson
Park reporter said they had the only drouth the Park had ever experienced" (Freeman 1958:286287).
"Grasshoppers, prairie dogs, bean beetles, potato psyllids added to the dry weather calamities in
Montezuma County. The government furnished tons of grasshopper poison and prairie dog poison and bean growers and potato growers fought with sprays, and with the constant help and advise of County Agent Barr, any great calamity from these plagues was averted" (Freeman
1958:267).
"Rain fell during the week of July 20, 1934, some rain in every part of the county and a good
rain in the vicinity of Ackmen and Yellow Jacket. Bean and corn had survived the dry weather
and the rain brought on a fair crop in may places in the dry farming area It was something to
have a good crop in 1934" (Freeman 1958:287).
1934 is mentioned frequently by present and former Goodman Point residents as a drought year
that resulted in low crop yields (Connolly, personal communication, January 1990).
"The dry year of 1934 was followed by one of plenty of water and good crops were reported from
every part of the county. Thirty to forty bushels of wheat per acre, 60 to 90 bushes of oats, 40
to 70 bushels of barley were common yields where a good job of farming was done.
Considerable certified pure seed was used this year and plots planted with this seed invariably
outyielded crops from other seed.... Almost all seed grain and potatoes planted this year were
treated before planting to ward off disease and grain smut" (Freeman 1958:288).
"The long drouth of 1934, affecting the entire county, was effectively broken in the winter of
19341935. Although the soil was dry to a great depth the fall of rain and snow was sufficient to
restore moisture all the way down" (Freeman 1958:239).
239-
1937
1946
1947
1951
1954
1956
1959
1960s1970s
"The snowfall of the winter of 19361937 on Mesa Verde National Park was the greatest on
record according to the report made public in April, 1937. The total snowfall, according to the
record, as reported by Paul R. Frank, Acting Superintendent, was 145.5 inches. It was a winter
of deep snows allover the San Juan Basin and deep mud the following spring" (Freeman
1958:240).
Montezuma Valley Journal, August I, 1946 headline states "wheat harvest now exceeds previous
dry year estimates" and records that "some of the nonirrigated wheatfields are now yielding 15
bushels to the acre and less fortunate growers who have as low as 35 bushel returns are nevertheless showing a profit due to the fact that the local price is very favorable" (Thompson, personal communication, January 1990).
1947 recalled as being an example of a year with good crops by Goodman Point area residents
(Connolly, personal communication, January 1990).
1951 is recalled as a year of severe drought resulting in almost total crop failure by former and
present Goodman Point area residents (Connolly, personal communication, January 1990).
Montezuma Valley Journal, July 12, 1951 reports that "precipitation records compiled by
Montezuma Valley Irrigation show only 0.02 inches of rainfall in June with a sparse 3.26 inches
recorded since January.... Few vigorous stands of wheat are reported left unparched on proximity
farms with pinto beans planting destined to suffer the same fate if no rainfall is forthcoming in
the next two or three weeks" (Thompson, personal communication, January 1990).
Montezuma Valley Journal, October 25, 1951, records the following: "With inspection reports
coming in on the insured bean acreages the 1951 county yield of bean crops is showing up as the
lightest per acre the county has ever known, according to EL. Murphy, the County FMA
chairman. Mr. Murphy has been raising beans in the Pleasant View area for 30 years and has
seen some light crops and some extra heavy ones, but never a complete failure. Even this year,
with moisture for the last two years at only 30% of the one year average, the crop is not a
complete failure. The yield on a county average, will be about 20% of normal" (Thompson, personal communication, January 1990).
"Only 65 per cent of normal snowfall was recorded in the winter of 19531954, and a shortage of
irrigation water ensued" (Freeman 1958:299).
"This is one of the shortest crop years the [Yellowjacket] locality has ever experienced.... A
700,000 bushel wheat crop is the largest handled to date. There is probably a 200,000 bushel
bean crop this season, one of the lightest ever produced and a 300,000 bushel wheat crop"
(Freeman 1958: 146).
"In the summer of 1956, an extremely dry period of weather, and on account of SO much water
being taken out above [of the Dolores river], the river quit running at the pump station and the
town [of Dove Creek] was again desperately short of water for a while" (Freeman 1956:153).
'The past few years [19531956] have been rather drier than the average season, but there has
been a fair crop right along, nothing approaching a complete failure, and during the summer of
1956, good bean and wheat land was exchanging right along at $100 to $125 an acre [in the
Dove Creek area]" (Freeman 1958:155).
Montezuma Valley Journal, September 24, 1959 records that "About 20% of the bean crop has
been harvest but the yield has been reported as the worst in years. Etter said the yield is running
from one half to five bags with an average of two sacks per acre" (Thompson, personal communication, January 1990).
These two decades remembered by Goodman Point area residents as being generally good and relatively predictable years for crop production (Connolly, personal communication, January 1990).
240-
ApPENDIX
C
CALCULATING ANNUAL POPULATION SIZE:
AN EXAMPLE FOR A.D. 902
The criteria specified in chapter 4 were used to detennine the total potential productivity (TOTPROD)
for the region as a whole during a single year, A. D. 902. This was a very moderate year which exhibited
soils that were typified by PDSI classes 6, 7, and 8nearly nonnal, incipient wet, and slightly wet In
A.D. 902, a total of 73,033,490 kg of maize was estimated to be the maximum potential yield that could
be produced from the 1470.36km 2 study area. The following example details how estimates for POP1YR,
POP2YR. and POP3YR were calculated.
Eセ。オエャoョ
1. TOTPROD * .5
Prod uct
Potential gross yield
=
2. Potential gross yield*.8
=
Adjusted gross yield
3. Adjusted gross yield*.9
=
Potential net yield (yrl)
4. Potential net yield*.9
=
Adjusted net yield (yrl)
5. Adj. net yield (yrl) 1
POP1YR consumption rate =
6. POPIYR 1 total area =
POPIYR
POP1YR (density)
7. Adj. net yield (yrl)*.9 =
Potential net yield (yr2)
Pot. net yield (yr2) *.9
=
Adj. net yield (yr2)
8. Adj. net yield (yr2) 1 POP2YR
consumption rate =
9. POP2YR 1 total area =
POP2YR
POP2YR (density)
10. Adj. net yield (yr2) *.9
=
Potential net yield (yr3)
11. Pot. net yield (yr3) *.9
=
Adj. net yield yield (yr3)
12. Adj. net yield (yr3) 1 POP3YR
POP3YR
consumotion rate =
POP3YR (density)
13. POP3YR 1 total area =
ExampJe
73,033,490 kg *.5
=36 516,745 ォセ
36,516,745 kg *.8
=29,213,396 kg
29,213,396 kg *.9
=26,292 056 kl!
26,292,056 kg *.9
=23 662,851 kl!
23,662,851 kg 1160 kg per person =
147,893 oeoDle
147,893 people 11470.36 km 2 =
100.6 people/km2
23,662,851 kg*.9
=21,296,566 ォセ
21,296,566 kg *.9
= 19,166,909 ォセ
19,166,909 kg 1320 kg per person =
59 897 people
59,897 people 1 1470.36 k:m 2
=40.7 people/km2
19,166,909 kg.*.9
= 17,250,218 kl:!
17,250,218 kg*.9
15,525,196 kl:!
15,525,196 kg 1480 kg per person =
32 344 peoDle
32,344 people 1 1470.36 km 2
21.9 oeoDIe/km2
=
=
The above can be simplified as:
POPIYR = ««TOTPROD*.5)*.8)*.9)*.9)/160 or
a POPIYR =(TOTPROD*.324)/I60 and
POP1YR (density) = ««(TOTPROD*.5)*.8)*.9)* .9)/160)/1470.36 or
a POP1YR (density) =«TOTPROD* .324)/160)/1470.36
a POP2YR =«TOTPROD*.324)*.81)/320 and
a POP2YR (density) =«(TOTPROD*.324)*.8l)/320)/1470.36
a POP3YR = «TOlPROD*.324)*.6561)/480 and
a POP3YR (density) = (TOlPROD*.324)* .6561)/480)/1470.36
«
aThese are the equations that were used in SAS programs to calculate population size for the entire study
area.
242-
SAS
ApPENDIX D
PROGRAMS FOR CALCULATING
TOTPROD AND POPKM
The frrst program takes the edited set of count fJ1es generated for each year in the study and does the
following. It concatenates all 400 data sets, frrst as sets of 50 years and then as a single set Next, it creates
a new variable SPUM by an equation that converts SPURs to SPUMs. It also creates a new variable called
PRODMKG (productivity of maize in kilograms). The data set CONCAT is sorted by year and a listing file
is created that reports the results for each variable (SPUR HA YEAR SPUM and PRODMKG) in the form
of 5 columns and as many rows as there are SPURs for each of the 400 years. Finally, it calculates a
TOTPROD (total maize productivity) value for each year by summing the SPUM values when the SPUR
values are greater than or equal to 19. A disk file contains the results of this step and is a list with one
column (TOTPROD) and 400 rows, one for each year. This disk me is renamed and used as data for
programs designed to calculate population size and population density.
Program 1 generates TOTPROD:
1
2
3
4
38
39
40
41
85
86
87
88
OPTIONS TMSG=NOTES;
DATA Y901;
INPUT SPUB 5-6 HA 34-39; YEAR=901; CARDS4;
(EDITED COUNT FILE DATA FOR A.D.901 HERE)
16486
16487
16488
16530
16531
16532
16533
16534
16535
16536
16537
16538
16539
16540
16541
16542
16543
16544
16545
16546
16547
16548
16549
16550
16551
;;;; DATA Y1300;
INPUT SPUB 5·6 HA 34-39; YEAR=902; CARDS4;
(EDITED COUNT FILE DATA FOR A.D. 1300 HERE)
;;;; DATA CONCAT1; SET Y901 Y902 Y903 Y904 Y905 Y906 Y907 Y908 Y909
Y910 Y911 Y912 Y913 Y914 Y915 Y916 Y917Y918 Y919 Y920 Y921 Y922 Y923
Y924 Y925 Y926 Y927 Y928 Y929 Y930 Y931 Y932 Y933 Y934 Y935 Y936 Y937
Y938 Y939 Y940 Y941 Y942 Y943 Y944 Y945 Y946 Y947 Y948 Y949 Y950;
DATA CONCAT2; SET Y951 Y952 Y953 Y954 Y955 Y956 Y957 Y958 Y959 Y960
Y961 Y962 Y963 Y964 Y965 Y966 Y967 Y968 Y969 Y970 Y971 Y972 Y973 Y974
Y975 Y976 Y9n Y978 Y979 Y980 Y981 Y982 Y983 Y984 Y985 Y986 Y987 Y988
Y989 Y990 Y991 Y992 Y993 Y994 Y995 Y996 Y997 Y998 Y999 Y1 000;
DATA CONCAT3; SET Y1 001 Y1002 Y1003 Y1 004 Y1 005 Y1 006 Y1 007 Y1 008
Y1009 Y1010 Y1011 Y1012 Y1013 Y1014 Y1015 Y1016 Y1017 Y1018 Y1019
Y1020 Y1 021 Y1 022 Y1 023 Y1024 Y1 025 Y1 026 Y1027 Y1 028 Y1 029 Y1 030
Y1031 Y1032 Y1033 Y1034 Y1035 Y1036 Y1037 Y1038 Y1039 Y1040 Y1041
Y1042 Y1043 Y1044 Y1 045 Y1046 Y1047 Y1 048 Y1 049 Y1 050;
DATA CONCAT4; SETY1051 Y1052 Y1053 Y1054 Y1055 Y1056 Y1057 Y1058
Y1059 Y1 060 Y1 061 Y1062 Y1063 Y1 064 Y1 065 Y1 066 Y1 067 Y1 068 Y1 069
Y1070 Y1071 Y1072 Y1073 Y1074 Y1075 Y1076 Y01n Y0178 Y0179 Y1080
Y1081 Y1082 Y1083 Y1084 Y0185 Y1086 Y1087 Y1088 Y0189 Y1090 Y1091
Y1092 Y1093 Y1094 Y0195 Y1096 Y1097 Y1098 Y1099 Y1100;
DATA CONCAT5; SET Y1101 Y1102 Y1103 Y1104 Y1105 Y1106 Y1107 Y1108
Y1109 Y111 0 Y1111 Y1112 Y1113 Y1114 Y1115 Y1116 Y1117 Y11118 Y1119
Y1120 Y1121 Y1122 Y1123 Y1124 Y1125 Y1126 Y1127 Y1128 Y1129 Y1130
Y1131 Y1132 Y1133 Y1134 Y1135 Y1136 Y1137 Y1138 Y1139 Y1140 Y1141
;;;; DATA Y902;
INPUT SPUB 5-6 HA 34-39; YEAR=902; CARDS4;
(EDITED COUNT FILE DATA FOR A.D. 902 HERE)
;;; DATA Y903;
INPUT SPUB 5-6 HA 34-39; YEAR=902; CARDS4;
(EDITED COUNT FILE DATA FOR A.D. 903 HERE)
16552
16553
16554
16555
16556
16557
16558
16559
16560
16561
16562
16563
16564
16565
16566
16567
16568
16569
16570
16571
16572
16573
16574
16575
16576
16577
16578
16579
16580
16581
16582
16583
16584
Y1142 Y1143 Y1144 Y1145 Y1146 Y1147 Y1148 Y1149 Y1150;
DATACONCAT6; SETY1151 Y1152Y1153Y1154Y1155Y1156Y1157Y1158
Y1159 Y1160 Y1161 Y1162 Y1163 Y1164 Y1165 Y1166 Y1167 Y1168 Y1169
Y1170 Y1171 Y1172 Y1173 Y1174 Y1175 Y1176 Y1177 Y1178 Y1179 Y1180
Y1181 Y1182 Y1183 Y1184 Y1185 Y1186 Y1187 Y1188 Y1189 Y1190 Y1191
Y1192 Y1193 Y1194 Y1195 Y1196 Y1197 Y1198 Y1199 Y1200;
DATA CONCAT7; SET Y1201 Y1202 Y1203 Y1204 Y1205 Y1206 Y1207 Y1208
Y1209 Y1210 Y1211 Y1212 Y1213 Y1214 Y1215 Y1216 Y1217 Y1218 Y1219
Y1220 Y1221 Y1222 Y1223 Y1224 Y1225 Y1226 Y1227 Y1228 Y1229 Y1230
Y1231 Y1232 Y1233 Y1234 Y1235 Y1236 Y1237 Y1238 Y1239 Y1240 Y1241
Y1242 Y1243 Y1244 Y1245 Y1246 Y1247 Y1248 Y1249 Y1250;
DATA CONCAT8; SET Y1251 Y1252 Y1253 Y1254 Y1255 Y1256 Y1257 Y1258
Y1259 Y1260 Y1261 Y1262 Y1263 Y1264 Y1265 Y1266 Y1267 Y1268 Y1269
Y1270 Y1271 Y1272 Y1273 Y1274 Y1275 Y1276 Y1277 Y1278 Y1279 Y1280
Y1281 Y1282 Y1283 Y1284 Y1285 Y1286 Y1287 Y1288 Y1289 Y1290 Y1291
Y1292 Y1293 Y1294 Y1295 Y1296 Y1297 Y1298 Y1299 Y1300;
DATA CONCAT; SET CONCAr1 CONCAT2 CONCAT3 CONCAT4 CONCAT5
CONCAT6 CONCAT7 CONCAT8;
SPUM=(((SPUB *10)+4)/.473r1.12;
PRODMKG=HA*SPUM;
PROC SORT; BY YEAR;
PROC PRINT;
DATA_NULL_; SET CONCAT; BY YEAR; FILE PUNCH;
IF FIRST.YEAR THEN DO;
TOTPROD=O;
PUT YEAR @;
IF SPUB GE 19 THEN TOTPROD+PRODMKG;
RETURN;
END;
IF SPUB GE 19 THEN TOTPROD +PRODMKG;
IF LAST.YEAR ; PUT TOTPROD;
RETURN;
The second program calculates population density for a population intending to place one year, two
years, and three years of maize in storage. A variant of this program produces the maximum number of
people who could be sustained on TOlPROD rather than producing the density of people per km 2. tィセ
only difference between these two is the deletion of the division by the total area in the equation that
calculates population size for each storage level.
The program does the following. First it specifies the input two variables, YEAR and TOlPROD. The
data set that contains this information is contained within the renamed punch file generated by the first
program. Next, it creates three new variables, POPlYR, POP2YR, and POP3YR by the equations given.
Next, it sorts the data by year and creates an output file with five colums (YEAR TOTPROD POPl YR
POP2YR POP3YR) and 400 rows, one for each year in the study. Finally, it calculates the mean, standard
deviation, minumum value, and the maximum value, among other useful descriptive statistics, for the
variables TOlPROD, POPl YR, POPl YR, and POP3YR, and then appends this to the output file.
Program 2 generates POPKM:
1
2
3
4
5
404
405
OPTIONS TMSG=NOTES;
DATA POP;
INPUT YEAR TOTPROD;
CARDS;
(DATA FROM RENAMED PUNCH FILE THAT REPORTS ANNUAL TOTPROD HERE)
244-
406
407
408
409
410
411
412
413
414
DATA POP1; SET POP;
POP1YR = INT(((TOTPROD*.324)/160)/1470.36);
POP2YR= INT((((TOTPROD*.324)*.81)/320)/1470.36);
POP3YR= INT ((((TOTPROD*.324)*.6561)/480)/1470.36);
PROC SORT DATA= POP1; BY YEAR;
TITLE1 'REGIONAL POPULATIONAL DENSITY (PERSONS/KM2)';
PROC PRINTDATA=POP1;
PROCMEANS DATA = POP1;
VAR TOTPROD POP1YR POP2YR POP3YR;
-
245-
ApPENDIX
E
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the
Study Area, A.D. 9011300a
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
38351244
73033490
49252912
52740208
80344869
38241754
38647208
74453111
76407925
72693730
97327555
70538178
68766106
72502842
59249961
55794185
80986788
77865620
80097380
55078313
68494597
38332301
49468899
49738722
74199938
68448963
53010315
78762226
Number of People in 1470.36 kfTl2
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
16984
31452
52
77661
21
11
59896
32344
100
147892
21
40
40393
21812
67
99737
27
14
43253
23356
72
106798
29
15
65892
35582
162698
110
44
24
16936
31363
52
77439
21
11
17115
31695
53
21
11
78260
61060
32972
102
41
150767
22
154726
62664
33838
105
42
23
32193
21
147204
59617
100
40
79820
43103
134
197088
54
29
57850
31239
97
142839
39
21
30454
94
139251
56396
38
20
59461
32109
146818
99
40
21
48592
26239
81
119981
33
17
24709
31
16
112983
45758
76
111
66419
35866
45
163998
24
157677
63859
34484
107
43
23
162197
35472
110
24
65689
44
111533
45171
24392
75
30
16
56174
94
138701
30334
38
20
31437
77622
16976
52
21
11
100174
40570
21908
68
27
14
100720
40791
22027
27
14
68
150254
32860
102
41
60853
22
138609
56136
30313
94
38
20
107345
43475
23476
73
29
15
159493
64594
34881
108
43
23
121133
59819157
49059
26491
1300
82
33
Mean (n400)
131473
53246
28752
64925217
88
35
19
Standard
7
28222
11429
13936845
6172
Deviation
(N400)
Minimum Value
21
52
38241754
77439
31363
16936
(N=400)
Maxiroom Value 102410868
57
30
83989
141
45354
N=400
ate: All
Dvalues are rounded to the nearest whole number. All population values
rOP1 YR. POP2YR. セrySpo
are truncated Nセエョゥ
For a fuller presentation, see an West 1990:5
79.
18
19
4
11
30
F
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Sand Canyon
Survey Locality, A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
1203519
1927934
1341613
1471751
2174287
1203519
1203519
1917705
2185747
2067733
2483208
2067733
1916569
2039129
1579631
1472888
2186221
2051726
2127782
1472888
1916569
1203519
1340477
1465670
1917705
1915716
1472888
2064797
Number of People in 26.08 km 2
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
2437
987
532
93
37
20
3904
1581
853
149
60
32
2716
1100
594
104
42
22
2980
1207
651
114
46
24
4402
1783
962
168
68
36
2437
987
532
93
37
20
2437
987
532
93
37
20
3883
1572
849
148
60
32
4426
1792
967
169
68
37
4187
1695
915
160
65
35
5028
2036
1099
192
78
42
4187
1695
915
160
65
35
3881
1571
848
148
60
32
4129
1672
903
158
64
34
3198
1295
699
122
49
26
114
46
25
2982
1207
652
4427
1792
968
169
68
37
4154
1682
908
159
64
34
4308
1745
942
165
66
36
2982
1207
652
114
46
25
3881
1571
848
148
60
32
2437
987
532
93
37
20
2714
1099
593
104
42
22
2967
1202
649
113
46
24
148
60
32
3883
1572
849
148
60
32
3879
1571
848
2982
1207
652
114
46
25
914
160
64
35
4181
1693
1300
3244
1313
1601984
709
Mean (n=400)
3602
1458
787
1779087
Standard
276
149
337524
683
Deviation
(N=400)
Minimum Value
2437
1203519
987
532
(N=400)
MaximJm Value
5174
2555172
2095
1131
N=400
D values are round to the nearest whole number. All
Note: IT TP
rOP1YR, POP2YR, POP3YR) are truncated integ:rs.
For a fuller presentation, see Van West 1990:58 591.
124
137
26
50
55
10
27
29
5
93
37
20
198
80
43
population values
G
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Mockingbird Mesa
Survey Locality, A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
412084
958569
573175
590413
1036234
412084
412084
1023922
893405
872947
1196681
792818
860728
875788
695925
686075
1036234
1023922
1029226
670541
855330
412084
648568
586170
1023922
855330
590413
1023922
Population Density per km2
Number of People in 17.96 km 2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
834
337
182
46
18
10
1941
786
424
108
43
23
1160
470
253
64
26
14
1195
484
261
66
26
14
2098
849
458
116
47
25
834
337
182
46
18
10
834
337
182
46
18
10
2073
839
453
115
46
25
1809
732
395
100
40
22
386
98
39
21
1767
715
134
54
29
2423
981
529
89
36
19
1605
650
351
1742
705
381
97
39
21
1773
718
387
98
39
21
1409
570
308
78
31
17
303
77
31
16
1389
562
2098
849
458
116
47
25
115
46
25
2073
839
453
2084
844
455
116
46
25
1357
549
296
75
30
16
1732
701
378
96
39
21
834
337
182
46
18
10
1313
531
287
73
29
15
1186
480
259
66
26
14
453
115
46
25
2073
839
96
39
21
1732
701
378
484
261
66
26
14
1195
2073
839
453
115
46
25
1300
695925
1409
570
308
78
31
Mean (n=400)
1622
801336
656
354
46
18
9
23
Standard
204994
415
168
90
Deviation
(N=400)
46
Minimum Value
18
834
412084
337
182
(N=400)
57
Maxirrum Value
1034
1261125
2553
142
558
N=400
Note: All T TPR D values are rounded to the nearest whole number. All population values
rOP1 YR, POP2YR, POP3YR) are truncated Nウイ・セエョゥ
For a fuller presentation. see Van West 1990:59 -603.
17
10
5
10
31
H
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT8371 (DCA Site), A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
93426
406041
278234
278234
468553
93426
93426
468553
365883
361336
498407
300529
361336
365883
343587
339040
468553
468553
468553
339040
361336
93426
327675
231861
468553
361336
278234
468553
Number of People in 7.88 km2
Population Density per kfTl2
POP1 YR POP2YR POP3YR POPIYR POP2YR POP3YR
189
76
41
24
9
5
822
333
179
104
42
22
71
28
15
563
228
123
71
28
15
563
228
123
48
26
948
384
207
120
189
72
41
24
9
5
76
41
24
9
5
189
48
26
948
384
207
120
740
300
162
94
38
20
731
296
160
92
37
20
51
28
1009
408
220
128
608
246
133
77
31
16
37
20
731
296
160
92
94
38
20
740
300
162
695
281
152
88
35
19
686
278
150
87
35
19
48
26
948
384
207
120
948
384
207
120
48
26
948
384
207
120
48
26
686
278
150
87
35
19
731
296
160
92
37
20
189
76
41
24
9
5
663
268
145
84
34
18
469
190
102
59
24
13
120
48
26
948
384
207
731
296
160
92
37
20
71
28
15
563
228
123
948
384
207
120
48
26
1300
343587
695
281
152
Mean (n=400)
350669
709
287
154
Standard
104956
212
46
86
Deviation
(N=400)
Minimum Value
189
93426
76
41
(N=400)
Maximum Value
556372
1126
456
246
N=400
to the nearest whole number.
TPR D values are rou
Note: All
rOP1YR, POP2YR, POP3YR) are truncated integers.
For a fuller presentation, see Van West 1990:604-615.
88
89
26
35
35
10
19
19
5
24
9
5
142
57
31
I population values
I
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT8839 (Norton House), A.D. 9011300 a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
478574
737789
576793
579445
789503
478574
478574
695641
684843
682949
903350
682949
681055
682949
582097
579445
789503
787609
787609
579445
681055
478574
576793
576793
681055
681055
579445
787609
Number of People in 7.88 km2
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
969
392
211
122
49
26
1494
605
326
189
76
41
1168
473
255
148
60
32
1173
475
256
148
60
32
1598
647
349
202
82
44
969
392
211
122
49
26
969
392
211
122
49
26
1408
570
308
178
72
39
71
38
1386
561
303
175
1382
560
302
175
71
38
94
50
1829
740
400
232
1382
560
302
175
71
38
1379
558
301
175
70
38
71
38
1382
560
302
175
1178
477
257
149
60
32
1173
475
256
148
60
32
82
44
1598
647
349
202
81
44
1594
645
348
202
202
81
44
1594
645
348
1173
475
256
148
60
32
70
38
1379
558
301
175
392
211
122
49
26
969
32
1168
473
255
148
60
1168
473
255
148
60
32
70
38
1379
558
301
175
1379
558
301
175
70
38
60
32
1173
475
256
148
81
44
1594
645
348
202
582097
653070
99827
1178
1321
202
477
535
81
257
288
44
149
167
25
60
67
10
32
36
5
478574
969
392
211
122
49
26
909146
1841
745
402
233
94
51
population values
ApPENDIX
J
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT2433 (Aulston Publo), A.D. 9011300a
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
314812
556164
328830
393274
612425
314812
314812
554175
661581
617539
689806
617539
554175
598312
453607
393274
615740
568288
588556
393274
554175
314812
328830
419491
554175
554175
393274
571603
Number of People in 7.88 kn12
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
637
258
139
80
32
17
1126
456
246
142
57
31
665
269
145
84
34
18
796
322
174
101
40
22
34
1240
502
271
157
63
17
637
258
139
80
32
17
637
258
139
80
32
1122
454
245
142
57
31
1339
542
292
170
68
37
64
34
1250
506
273
158
71
38
1396
565
305
177
1250
506
273
158
64
34
1122
454
245
142
57
31
1211
490
264
153
62
33
918
372
200
116
47
25
796
322
174
101
40
22
64
34
1246
504
272
158
146
59
31
1150
466
251
61
33
1191
482
260
151
796
322
174
101
40
22
1122
454
245
142
57
31
258
139
80
32
17
637
84
34
18
665
269
145
107
43
23
849
344
185
454
245
142
57
31
1122
1122
454
245
142
57
31
796
322
174
101
40
22
1157
468
253
146
59
32
1300
454460
920
372
201
116
47
Mean (n=400)
500847
1013
410
221
128
51
Standard
115085
233
94
50
29
12
Deviation
(N=400)
Minimum Value
314812
637
258
139
80
32
(N=400)
Maximum Value
739948
1498
606
327
190
77
_CN=400)
Note: All TOTPROD values are rounded to the nearest whole number. All population values
(POP1YR, POP2YR, POP3YR) are truncated Nセ・エョゥ
aFor a fuller presentation, see Van West 1990:62
39.
25
27
6
17
41
K
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT3834 (Mustoe Ruin), A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
377040
592250
388311
460010
672000
377040
377040
593387
732902
667359
760407
667359
592250
659782
526121
461146
672000
604469
634304
461146
592250
377040
387174
441654
593387
591398
461146
605605
Number of People in 7.88 kl1l2
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
763
309
166
96
39
21
61
33
1199
485
262
152
786
318
171
99
40
21
203
118
47
25
931
377
1360
551
297
172
69
37
763
309
166
96
39
21
763
309
166
96
39
21
1201
486
262
152
61
33
1484
601
324
188
76
41
37
1351
547
295
171
69
1539
623
336
195
79
42
1351
547
295
171
69
37
1199
485
262
152
61
33
1336
541
292
169
68
37
1065
431
233
135
54
29
204
118
47
25
933
378
1360
551
297
172
69
37
1224
495
267
155
62
33
1284
520
280
163
66
35
933
378
204
118
47
25
1199
485
262
152
61
33
763
309
166
96
39
21
784
317
171
99
40
21
894
362
195
113
45
24
1201
486
262
152
61
33
1197
485
261
151
61
33
204
118
47
25
933
378
1226
496
268
155
63
34
1300
536539
1086
440
237
137
55
Mean (n=400)
1119
553196
453
244
141
56
Standard
27
11
108069
218
88
47
Deviation
(N=400)
Minimum Value
96
377040
763
39
309
166
(N-400)
Maxirrom Value
80
776849
1573
637
344
199
N=400
Note: All T TPR D values are rounded to the nearest whole number. All population values
NryQpoセ
POP2YR, POP3YR) are truncated integers.
For a fuller presentation, see Van West 1990:64Q-651.
30
30
6
21
43
L
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT6970 (Wallace Ruin), A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
198673
329891
236351
243549
361620
198673
198673
357832
330611
310152
402935
280886
302196
323981
300094
278309
364462
355559
360673
275847
295945
198673
248948
216953
357832
301249
243549
360673
Number of People in 4.32 km2
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
402
162
87
93
37
20
154
62
33
668
270
146
478
193
104
11 0
44
24
493
199
107
114
46
24
732
296
160
169
68
37
402
162
87
93
37
20
402
162
87
93
37
20
724
293
158
167
67
36
669
271
146
154
62
33
628
254
137
145
58
31
815
330
178
188
76
41
124
131
53
28
568
230
611
247
133
141
57
30
656
265
143
151
61
33
607
246
132
140
56
30
563
228
123
130
52
28
738
298
161
170
69
37
720
291
157
166
67
36
730
295
159
169
68
36
558
226
122
129
52
28
599
242
131
138
56
30
402
162
87
93
37
20
116
47
25
504
204
110
439
177
96
101
41
22
724
293
158
167
67
36
57
30
610
247
133
141
493
199
107
114
46
24
730
295
159
169
68
36
132
140
56
1300
300094
607
246
Mean (n=400)
297798
602
243
131
139
56
Standard
53732
108
44
23
25
10
Deviation
(N=400)
87
93
37
Minimum Value
198673
402
162
(N=400)
183
194
78
MaxiroomVaJue
414774
839
340
(N=400)
Note: All TOTPROb values are rounded to the nearest whole number. All population values
(POP1YR, POP2YR. POP3YR) are truncated integers.
aFar a fuller presentation, see Van West 1990:652-663.
30
30
5
20
42
M
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT1566 (Lowry Ruin), A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
256677
450955
261696
314642
501817
256677
256677
451713
555520
511478
568212
511478
450955
495850
363799
315400
503806
456922
483632
315400
450955
256677
260939
344212
451713
450387
315400
459669
Number of People in 7.88 knl2
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
519
210
113
65
26
14
199
115
46
25
913
369
529
214
115
67
27
14
637
258
139
80
32
17
1016
411
222
128
52
28
210
113
65
26
14
519
113
65
26
14
519
210
914
370
200
116
47
25
1124
455
246
142
57
31
419
226
131
53
28
1035
1150
466
251
146
59
31
1035
419
226
131
53
28
913
369
199
115
46
25
1004
406
219
127
51
27
736
298
161
93
37
20
638
258
139
81
32
17
1020
413
223
129
52
28
925
374
202
117
47
25
979
. 396
214
124
50
27
139
81
32
17
638
258
913
369
199
115
46
25
519
210
113
65
26
14
528
214
115
67
27
14
697
282
152
88
35
19
914
370
200
116
47
25
912
369
199
115
46
25
638
258
139
81
32
17
930
376
203
118
47
25
1300
369292
747
302
163
94
38
Mean (n=400)
409875
829
335
180
104
42
Standard
26
10
102360
207
83
45
Deviation
(N=400)
Minimum Value
65
26
256677
519
210
113
(N=400)
MaxilTXJm Value
657925
1332
68
539
291
169
N=400
Note: All T TPR D values are rounded to the nearest whole nu er. All population values
rOP1YR, POP2YR, POP3YR) are truncated integers.
For a fuller presentation, see Van West 1990:664-675.
20
22
5
14
36
ApPENDIX
N
Total Annual Maize Productivity and
Maximum Potential Population for Three
Levels of Storage in the Catchment of
5MT2149 (Escalante Ruin), A.D. 9011300 a
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
202822
316176
249706
249706
351031
202822
202822
310588
310588
310588
477949
310588
310588
310588
249706
249706
351031
351031
351031
249706
310588
202822
249706
249706
310588
310588
249706
351031
Number of People in 6.72 km2
Population Density per kl1l2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
410
166
89
61
24
13
640
259
140
95
38
20
505
204
110
75
30
16
505
204
110
75
30
16
710
287
155
105
42
23
410
166
89
61
24
13
410
166
89
61
24
13
628
254
137
93
37
20
628
254
137
93
37
20
628
254
137
93
37
20
967
391
211
144
58
31
628
254
137
93
37
20
628
254
137
93
37
20
628
254
137
93
37
20
505
204
110
75
30
16
204
110
75
30
16
505
710
287
155
105
42
23
710
287
155
105
42
23
155
105
42
23
710
. 287
110
75
30
16
505
204
628
254
137
93
37
20
410
166
89
61
24
13
505
204
110
75
30
16
505
204
110
75
30
16
628
254
137
93
37
20
628
254
137
93
37
20
505
204
110
75
30
16
710
287
155
105
42
23
1300
204
249706
505
110
75
30
Mean (0=400)
291075
588
238
128
87
34
Standard
16
6
108
44
53857
23
Deviation
(N=400)
Minimum Value
61
24
410
202822
166
89
(N=400)
Maximum Value
967
58
477949
391
144
211
N=400
Note: All
TPR D values are rounded to the nearest whole number. All popUlation va ues
rOP1YR, POP2YR, POP3YR) are truncated integers.
For a fuller presentation. see Van West 1990:676-687.
16
18
3
13
31
0
Total Annual Maize Productivity and Maximum
Potential Population for Three Levels of Storage
in the Catchment of 5MT765 (Sand Canyon
Pueblo), A.D. 9011300a
ApPENDIX
Year
A.D.
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
Maize Yield
TOTPROD (kg)
303731
507235
370220
378176
566526
303731
303731
501646
516895
510455
652432
510455
501646
508182
385469
378176
566526
559991
563590
378176
501646
303731
370220
383045
501646
501646
378176
559991
Number of People in 7.88 krJ12
Population Density per km2
POP1YR POP2YR POP3YR POPIYR POP2YR POP3YR
615
249
134
78
31
17
1027
415
224
130
52
28
749
303
163
95
38
20
765
310
167
97
39
21
1147
464
250
145
58
31
615
249
134
78
31
17
615
249
134
78
31
17
1015
411
222
128
52
28
132
53
29
1046
423
228
1033
418
226
131
53
28
167
67
36
1321
535
288
1033
418
226
131
53
28
128
52
28
1015
411
222
416
225
130
52
28
1029
780
316
170
99
40
21
167
97
39
21
765
310
1147
464
250
145
58
31
143
58
31
1133
459
248
.462
249
144
58
31
1141
167
97
39
21
765
310
1015
411
222
128
52
28
134
78
31
17
615
249
163
95
38
20
749
303
314
169
98
39
21
775
1015
411
222
128
52
28
1015
411
222
128
52
28
167
97
39
21
765
310
1133
459
248
143
58
31
1300
386227
782
316
171
99
40
Mean (0=400)
459114
929
376
202
117
47
Standard
92902
188
76
41
23
9
Deviation
(N-400)
204
82
44
25
10
Minimum Value
101079
(N.400)
Maxiroom Value
683688
1384
560
302
175
71
-.1N=400,
Note: A110TPROD values are rounded to the nearest whole number. All population values
(POP1YR, POP2YR, POP3YR) are truncated integers.
aFor a fuller presentation, see Van West 1990:688-699.
21
25
5
5·
38
Washington State University Department of Anthropology
REPORTS OF INVESTIGATIONS
1957 Stallard, Bruce, An Archaeological Survey in the Wells Reservoir in the State of Washington. (Out of
Print).
2
1958 Stallard, Bruce, Preliminary Surveys for Highway Salvage Archaeology in the State of Washington: A
3
1959 Daugherty, Richard D., Archaeological Excavations in the Ice Harbor Reservoir, 1959 Progress Report.
4
1959 Newman, Thomas Stell, Toleak Point-An Archaeological Site on the North Central Washington
5
6
1959 Osborne, Douglas, Archaeological Tests in the Lower Grand Coulee, Washington. (Out of Print).
1960 Sprague, Roderick, Archaeology in the Sun Lakes Area of Central Washington. Appendix by Roald
7
1961 Clinehens, Stephen S., Further Archaeological Excavations in the Lower Grand Coulee of Central
9
1%1 Fryxell, Roald, Geologic Field Examination of the Park Lake Housepit Site (45-GR-90), Lower Grand
10
1961 Daugherty. Richard D., Excavations in the lee Harbor Reservoir, 1957-1960: A Preliminary Report.
12
14
1961 Mallory, Oscar L., An Archaeological Survey of Pacific Gas Transmission Company's Alberta to
California Pipeline System: MP 108.0 to MP 722.0. (Out of Print).
1961 Combes, John D., An Archaeological Survey of Pacific Gas Transmission Company's Alberta to
California Pipeline System: MP 108.0 to MP 722.0. Phase JI. (Out of Print).
1962 Mallory, Oscar L., Continued Archaeological Appraisal of the Lower Coulee, Central Washington.
15
16
1962 Ice, Dannie M., Archaeology of the Lava Bulle Site, Deschutes County, Oregon. (Out of Print).
1962 Guinn, Stanley J., White Rock Village Archaeological Site: A Preliminary Report of Investigations.
181
1963 Combes, John D., A Report on the Excavations of a Late Indian Burial Site in the Ice Harbor Reservoir
1811
1962 Fryxell, Roald, Geologic Examination of the Ford Island Archaeological Site (45-FR-47), Washington.
19
1962 Daugherty, Richard D., Archaeological Research in the Boundary Dam Reservoir Area. Final Report.
20
21
1962 Combes, John D., Excavations at Fort Spokane, Summer of 1962: A Preliminary Report. (Out of Print).
1962 Fryxell, Roald (in cooperation with Richard D. Daugherty), Interim Report: Archaeological Salvage in
22
1963 Guinn, Stanley J., A Maritime Village on the Olympic Peninsula of Washington. (Out of Print).
1963 Fryxell, Roald (in cooperation with Richard D. Daugherty), Late Glacial and Post Glacial Geological and
Final Report. (Out of Print).
(Out of Print).
Coast. (Out of Print).
Fryxell. (Out of Print).
Washington. (Out of Print).
Coulee, Washington. (Out of Print).
(Out of Print).
13
(Out of Print).
(Out of Print).
Region, Washington. Section 1. (Out of Print).
Section JI. (Out of Print).
(Out of Print).
the Lower Monumental Reservoir, Washington. (Out of Print).
23
24
1963
25
26
1964
1964
27
1964
28
1964
Archaeological Chronology of the Columbia Plateau, Washington: An Interim Report to the National
Science Foundation, 1962-1963. (Out of Print).
Steen, Virginia, A Study of the Chemical and Physical Characteristics of Pumice from Glacier Peak,
Washington and Mt . Mazama (Crater Lake), Oregon: Report of Progress, June 16 to September 7,1963.
(Out of Print).
Barnes, Paul L., Archaeology of the Dean Site: Twin Falls County, Idaho. (Out of Print).
Ackerman. Robert E., Prehistory in the Kuskokwim-Bristol Bay Region, Southwestern Alaska. (Out of
Print).
Fryxell, Roald, and Earl F. Cook, editors, A Field Guide to the Loess Deposits and Channeled Scablands
of the Palouse Area, Eastern Washington. (Out of Print).
Ackerman, Robert E., Archaeological Survey, Glacier Bay National Monument, Southeastern Alaska,
Part 1. (Out of Print).
Continued inside back cover . . .
Continued from back cover . . .
29
1964 Combes, John D., Excavations at Spokane House-Fort Spokane Historic Site, 1962-1963. (Out of
Print).
30
1964 Combes, John D., A Preliminary Investigation at Old Military Fort Spokane, Washington. (Out of
Print).
3I
1964 Fryxell, Roald, and Richard D. Daugherty, Demonstration of Techniques for Preserving Archaeological
Stratigraphy: A Report to the Wenner-Gren Foundation. (Out of Print).
32
1965 Sprague, Roderick, The Descriptive Archaeology of the Palus Burial Site, Lyons Ferry, Washington.
(Out of Print).
1965 Rice, David G., Archaeological Test Excavation in Fryingpan Rockshelter, Mount Rainier National
33
Park.
34
1965 Nelson, Charles M., Archaeological Reconnaissance in the Lower Monumental and Little Goose Dam
Reservoir Areas, 1964. (Out of Print).
1966 Kenaston, Monte Ray, The Archaeology of the Harder Site, Franklin County, Washington. (Out of
35
Print).
1965 Ackerman, Robert E., Archaeological Survey: Glacier Bay National Monument, Southeastern Alaska.
36
Part 11. (Out of Print).
37
1966 Sprague, Roderick, and John D. Combes. Excavations in the Little Goose and Lower Granite Dam
Reservoir, 1965. (Out of Print).
1966 Larabee, Edward McM., and Susan Kardas, Archaeological Survey of Grand Coulee Dam National
38
Recreation Area. Part 1: Lincoln County Above Normal Pool.
391 1966 Nelson, Charles M., A Preliminary Report on 45COI, A Stratified Open Site in the Southern Columbia
Plateau. (Out of Print).
39II 1966 Rice, David G., An Archaeological Reconnaissance of the Wynochee Dam Reservoir, Greys Harbor
County, Washington. (Out of Print).
1967 Daugherty, Richard, Barbara A. Purdy, and Roald Fryxell, The Descriptive Archaeology and Geo40
chronology of the Three Springs Bar Archaeological Site, Washington. (Out of Print).
41
1967 Walker, Deward E., Jr., Mutual Cross-Utilization of Economic Resources in the Plateau: An Example
from Aboriginal Nez Perce Fishing Practices. (Out of Print).
42
1967 Chance, David H., Archaeological Survey of Coulee Dam National Recreation Area. Part 11: Spring
Draw-Down of 1967. (Out of Print).
43
1967 Sprague, Roderick, A Preliminary Bibliography of Washington Archaeology. (Out of Print).
44
1968 Ackerman, Robert E., The Archaeology of the Glacier Bay Region, Southeastern Alaska, Part 111. (Out
of Print).
45
1968 Rice, David G., Archaeological Investigations in the Coulee Dam National Recreation Area, Spring
1968.
1969 Bernard, Russell, editor, Los Otomies: Papers from the lxmiquilpan Field School.
46
47
1969 Nelson, Charles M., The Sunset Creek Site (45-KI-28) and Its Place in Plateau Prehistory. (Out of Print).
481 1971 Bernard, H. Russell, and Michael Kenny, editors, The Inland Empire: Papers from the Spokane Area
Field School in Cultural Anthropology and Linguistics. (Out of Print).
48II 1971 Irwin, H. T., D.J. Hurd, and R. M. LaJeunesse, Description and Measurement in Anthropology.
Computer Series 1. (Out of Print).
49
1971 Leonhardy, Frank C., Gerald Schroedl, Judith Bense, and Seth Beckerman, Wexpusnime (45-GA-61):
Preliminary Report.
50
1972 Rice, David G., The Windust Phase in Lower Snake River Region Prehistory. (Out of print).
5I
1973 Schroedl, Gerald F., The Archaeological Occurrence of Bison in the Southern Plateau.
52
1974 Croes, Dale. Jonathan Davis, and H. T. Irwin, The Use of Computer Graphics in Archaeology: A case
Study from the Ozette Site, Washington.
1975 Adams, William H., Linda Gaw, and Frank C. Leonhardy, Archaeological Excavations at Silcott,
53
Washington: The Data Inventory. (Out of Print).
54
1977 Adams, William H.• Silcott, Washington: Ethnoarchaeology of a Rural American Community.
1978 Bodley, John H., Preliminary Ethnobotany of the Peruvian Amazon.
55
1979 Bodley. John H., and Foley C. Benson, Cultural Ecology of Amazonian Palms.
56
1979 Brown, Christopher L., and Carl E. Gustafson, A Key to Postcranial Skeletal Ref/wins of Cattle/Bison,
57
Elk and Horse (reprinted 1990).
Continued inside front cover . . .