Barrows, i.e. burial mounds, are amongst the most important of Europe’s prehistoric
monuments. Across the continent, barrows still figure as prominent elements in the
landscape. Many of these mounds have been excavated, revealing much about what
was buried inside these intriguing monuments. Surprisingly, little is known about
the landscape in which the barrows were situated and what role they played in their
environment. Palynological data, carrying important clues on the barrow environment,
are available for hundreds of excavated mounds in the Netherlands. However, while
local vegetation reconstructions from these barrows exist, a reconstruction of the
broader landscape around the barrows has yet to be made. This makes it difficult to
understand their role in the prehistoric cultural landscape.
It is argued in this book that barrows were built on existing heaths, which had been
and continued to be maintained for many generations by so-called heath communities.
These heaths, therefore, can be considered as ‘ancestral heaths’. The barrow landscape
was part of the economic zone of farming communities, while the heath areas were used
as grazing grounds. The ancestral heaths were very stable elements in the landscape and
were kept in existence for thousands of years. In fact, it is argued that these ancestral
heaths were the most important factor in structuring the barrow landscape.
Marieke Doorenbosch studied Biology at the Free University of Amsterdam and specialized in
paleoecology. From 2008-2013 she worked as a PhD student within the NWO-funded project
Ancestral Mounds at the Faculty of Archaeology at Leiden University of which this dissertation is
the result.
Sidestone Press
ISBN: 978-90-8890-192-8
9 789088 901928
Sidestone
ISBN 978-90-8890-192-8
ancestral heaths
reconstructing the barrow landscape in the
central and southern netherlands
Marieke Doorenbosch
ancestral heaths
In this book a detailed vegetation history of the landscape around burial mounds is
presented. Newly obtained and extant data derived from palynological analyses taken
from barrow sites are (re-)analysed. Methods in barrow palynology are discussed and
further developed when necessary. Newly developed techniques are applied in order to
get a better impression of the role barrows played in their environment.
Marieke Doorenbosch
ancestral heaths
ancestral heaths
Sidestone Press
ancestral heaths
reconstructing the barrow landscape in the
central and southern netherlands
Proefschrift
ter verkrijging van
de graad van Doctor aan de Universiteit Leiden,
op gezag van Rector Magnificus prof.mr. C.J.J.M. Stolker,
volgens besluit van het College voor Promoties
te verdedigen op donderdag 21 november 2013
klokke 15.00 uur
door
Marieke Doorenbosch
geboren te Amsterdam
in 1980
Promotiecommissie:
Promotor: Prof. Dr. C.C. Bakels
Co-promotor: Dr. D.R. Fontijn
Overige commissieleden:
Prof. Dr. J. Müller, Christian-Albrechts-Universität, Kiel
Dr. J.M. van Mourik, Universiteit van Amsterdam
Dr. A.B. Nielsen, Lunds Universitet
Prof. Dr. H. Fokkens, Universiteit Leiden
© 2013 M. Doorenbosch
Published by Sidestone Press, Leiden
www.sidestone.com
ISBN 978-90-8890-192-8
Lay-out & cover design: Sidestone Press
Photograph cover: K. Wentink
Contents
Part One
11
1 Introduction: why study the environment of barrows?
13
1.1 The academic significance of environmental barrow research
13
1.2 The societal significance of environmental barrow research
15
2 Environmental research on barrows, an overview so far
17
2.1 The vegetation history of the Netherlands in the Holocene
17
2.2 Environmental research on barrows
2.2.1 An overview
2.2.2 Pollen analyses for dating purposes
2.2.3 The reconstruction of local vegetation: regional and
cultural differences
19
19
20
21
2.3 Vegetation reconstructions of the barrow environment: open spaces in
the landscape
2.3.1 An overview of open spaces
2.3.2 Which open spaces were chosen for the building of barrows?
2.3.3 What was the size of the open spaces barrows were built in?
24
25
28
32
Conclusions
32
3 Barrow research, missing data
33
3.1 Research questions
33
3.2 Research area
34
3.3 Research methods
35
Part Two
Methodology
4 Sampling and treatment of soil samples
37
37
39
4.1 The sampling of barrows
4.1.1 The sampling of the old surface
4.1.2 The sampling of sods
4.1.3 The sampling of the soil profile underneath barrows
4.1.4 The sampling of ditch fills
4.1.5 The sampling of posthole fills
39
39
40
40
41
43
4.2 Chemical treatment and analysis of palynological soil samples
45
5 The palynology of mineral soil profiles
47
5.1 The theory behind the palynology of mineral soils
47
5.2 The time represented in a mineral soil pollen diagram
52
5.3. Absence of pollen grains in barrows
58
Conclusions
62
6 The pollen sum
63
6.1 Slabroek
64
6.2 Contemporaneous barrow pollen spectra
69
Conclusions
75
7 The size of an open place where a barrow was built
77
7.1 The size and the number of sods used in a barrow
7.1.1 An example:
77
78
7.2 The size of an open heathland area - examples from present Dutch
heathland areas
Sites and sampling methods
Methods of analysis
Results and discussion
Conclusions
78
79
81
81
84
7.3 The distance of a barrow to the forest edge - palynological modelling
Barrow landscape simulation
84
85
7.4 Discussion
92
Part Three
Case–studies
8 Northern and central Veluwe
93
93
95
8.1 Echoput
8.1.1 Site description
8.1.2 Pollen sampling and analysis
8.1.3 Results
8.1.4 Discussion
8.1.5 In conclusion: the history of the Echoput barrow landscape
96
96
101
102
109
114
8.2 Niersen-Vaassen
8.2.1 Site description and sample locations
8.2.2 Results
8.2.3 Discussion
115
115
120
123
8.3 Ermelo
8.3.1 Site description and sample locations
8.3.2 Results
8.3.3 Discussion
124
125
126
128
8.4 Putten
8.4.1 Site description and sample locations
8.4.2 Results and discussion
128
128
130
8.5 Vierhouten
8.5.1 Site description and sample locations
8.5.2 Results and discussion
130
130
130
8.6 Emst
8.6.1 Site description and sample locations
8.6.1 Results and discussion
132
132
132
8.7 Uddelermeer
8.7.1 Site description and sample locations
8.7.2 Results and discussion
132
132
132
8.8 Boeschoten
8.8.1 Site description and sample locations
8.8.1 Results and discussion
133
133
133
8.9 Ugchelen
8.9.1 Site description and sample locations
8.9.1 Results and discussion
133
133
133
8.10 Stroe
8.10.1 Results and discussion
135
135
8.11 Palynological results from peat and lake sediments
8.11.1 Site description and sample locations
8.11.2 Results and discussion
136
136
136
8.12 Summary: the barrow landscape of northern and central Veluwe
138
9 The Renkum stream valley
141
9.1 Site description and sample locations
Burial mounds belonging to the barrow alignment
Burial mounds outside the barrow alignment
141
141
143
9.2 Results and discussion
144
10 Gooi
149
10.1 Site description and sample locations
Baarn Group
Hilversum Group
Laren Group
Roosterbos
The Laarder Wasmeren area
149
149
149
150
151
151
10.2 Results and discussion
Gooi area
Laarder Wasmeren area
The (pre)barrow landscape of the Gooi
151
151
155
160
11 Toterfout-Halve Mijl and surroundings
163
11.1 Toterfout-Halve Mijl
11.1.1 Site description and sample locations
11.1.2 Results and discussion
163
163
170
11.2 Hoogeloon
11.2.1 Site description and sample locations
11.2.2 Results and discussion
172
172
172
11.3 Knegsel-Urnenweg
11.3.1 Site description and sample locations
11.3.2 Results and discussion
174
174
174
11.4 Knegsel-Moormanlaan
11.4.1 Site description and sample locations
11.4.2 Results and discussion
174
175
175
11.5 Steensel
11.5.1 Site description and sample locations
11.5.2 Results and discussion
175
175
175
11.6 Eersel
11.6.1 Site description and sample locations
11.6.2 Results and discussion
176
176
176
11.7 Bergeijk
11.7.1 Site description and sample locations
11.7.2 Results and discussion
176
176
177
11.8 Alphen
11.8.1 Site description and sample locations
11.8.2 Results and discussion
177
177
177
11.9 Goirle
11.9.1 Site description and sample locations
11.9.2 Results and discussion
178
179
179
11.10 Summary: the barrow landscape of Toterfout-Halve Mijl and
surroundings
179
12 Oss-Zevenbergen and surroundings
183
12.1 Oss-Vorstengraf area and Oss-Zevenbergen
12.1.1 Site description and sample locations
12.1.2 Results
12.1.3 Discussion
12.1.4 In conclusion: the history of the Oss-Zevenbergen landscape
183
183
194
207
212
12.2 Vorssel
12.2.1 Site description and sample locations
12.2.2 Results and discussion
213
213
213
12.3 Slabroek
12.3.1 Site description and sample locations
12.3.2 Results and discussion
213
213
215
12.4 Schaijk
12.4.1 Site description and sample locations
12.4.2 Results and discussion
218
218
218
12.5 Palynological results from palaeosoils, peat and lake sediments
12.5.1 Site description and sample locations
12.5.2 Results and discussion
218
218
219
12.6 Summary: the barrow landscape of Oss-Zevenbergen and surroundings 219
13 Ancestral heaths: understanding the barrow landscape
13.1 The barrow landscape
13.1.1 What did the barrow landscape look like in the central and
southern Netherlands during the 3rd to 1st millennium cal BC?
13.1.2 What was the history of the barrow landscape before
the barrows were built?
13.1.3 What does this mean?
13.1.4 What was the role of barrows in the landscape?
13.2 The heath open-forest passage landscape as part of the Dutch
prehistoric landscape
14 Conclusions: answers to the research questions
225
225
225
234
234
237
239
241
References
245
Appendix 1
261
Appendix 2
267
Acknowledgments
277
Curriculum Vitae
279
Part One
This thesis is divided into three parts. Part one (Chapters 1, 2 and 3) concerns
the background of this research, beginning (Chapter 1) with an overview of the
development of the palynological research of barrows. Following the overview
is an assessment of what data are available (Chapter 2) and what is still missing
(Chapter 3) from the palynological research of barrows.
Part two will go further into the methodology behind the palynology of barrows.
Chapter 4 gives an overview of sampling techniques used in this study. Chapter 5
discusses the theory of vegetation history reconstruction through the use of pollen
diagrams derived from mineral soils. In addition the relation between time and
depth in mineral soils will be discussed in this chapter. In Chapter 6 the so-called
pollen sum that is used in palynological analyses of barrows will be examined and
reconsidered. Chapter 7 concerns the determination of size of the open place a
barrow was built in. Three methods to determine the extent of an open space are
described and discussed.
In part three of this thesis (Chapters 8-14) the methodological theories described
and discussed in part two are applied to reconstruct the barrow landscapes of five
case study areas. Each case study area is dealt with in a separate chapter (chapters
8-12), including the presentation of palynological analyses of several individual
barrows and/or barrow complexes. In Chapter 13 the results of all case studies
are summarized and discussed, and the last chapter (14) submits answers to the
research questions put forward in this thesis.
Chapter 1
Introduction: why study the
environment of barrows?
1.1 The academic significance of environmental barrow
research
Barrows, i.e. burial mounds, are amongst the most important of Europe’s prehistoric
monuments. In the European landscape today hundreds of thousands of them are
still visible, and considering the large number of barrows that have disappeared
over time, it is not difficult to imagine the great importance barrows must have
had. Across Europe, barrows still figure as a prominent element in the landscape.
In Denmark alone, more than 80,000 barrows are known (Johansen et al. 2004).
Many barrows in Europe have been excavated, revealing much about what was
buried inside these monuments. Little is known, however, about the landscape in
which the barrows were situated. Palynological data, carrying important clues on
the barrow environment, are absent for most of the excavated barrows in Europe.
In the Netherlands however, the opposite is the case, with palynological data being
available for hundreds of excavated barrows, a fact which places the Netherlands
as a very important centre for the environmental research of barrows.
Some 3,000 barrows are presently known in the Netherlands (Bourgeois 2008).
Burial mounds were built from the 4th millennium BC until around 500 year BC,
with most being constructed during the 3rd and 2nd millennium BC. So many
barrows were built during this period that they must have visibly dominated the
landscape. Many of these barrows have been the subject of archaeological research
in the Netherlands. In 1906, Holwerda was the first to begin excavating barrows
near Hoog Soeren, the Veluwe (Holwerda 1907). Holwerda also did much to
popularize barrow archaeology, bringing it to the attention of the public. Van
Giffen, a contemporary of Holwerda, pioneered the quadrant method of barrow
excavation. With the quadrant method, the barrow is divided into four quadrants
and the opposing quarters are removed in order to identify internal features and
expose a continuous profile of the object through its centre along intersecting
axes (see figure 1.1). Van Giffen involved palynology, determination of bones and
seeds, geology and C14 dating in his archaeological research (Louwe Kooijmans
1979), in large part due to his training and background. After the Second World
War Glasbergen en Modderman continued to excavate numerous barrows. Around
1970 it was realised that burial mounds were valuable archaeological monuments
that needed protection, which led to the mounds being listed as cultural heritage
monuments protected by the state. Since then very few barrows have been
excavated and it was thought for a long time that there was more than enough
known about burial mounds.
Since 1906 around 800 barrows have been excavated. These excavations have
contributed not only to the knowledge we presently have on barrows, but also
to what we know of prehistoric man. However, this information has nowadays
become dated. In the past barrows were solely interpreted as burial places for
introduction
13
prestigious individuals or martial chiefs, but, based on the special and sometimes
exotic objects that are often found in barrows, especially barrows from the 3rd
and 2nd millennium cal BC, there is growing evidence pointing to barrows having
been highly important ritual places with a specific cultural value. The importance
of barrows in the past was emphasised by the fact that they were often re-used
again for burials and other ritual practices for hundreds of years and that barrows
formed in their entirety highly visible barrow landscapes. However, the specific
social and ideological significance of barrows is still unclear. What is further
lacking is information on the landscape surrounding the barrows. While local
vegetation reconstructions from many barrows in the Netherlands are available, a
reconstruction of the total landscape around the barrows has yet to made, without
which it would be difficult to understand their role in the prehistoric cultural
landscape. To improve our knowledge of barrows with respect to the problems
mentioned above, the project ‘Ancestral Mounds’ was started. The following
research questions were formulated (Fontijn 2007):
1. What was the social and ideological significance of barrow graves? In what
way do they differ, in terms of content, location, and landscape setting, from
contemporaneous other types of burials and ritual depositions? What does
this tell us about the social roles of the deceased buried in barrows?
2. What was the significance of barrows as landscape monuments? How were
they embedded in the by then emerging agrarian landscape and how did their
presence structure the landscape of later generations?
The ‘Ancestral Mounds’ project is divided into three PhD-projects, each focusing
on a different level of analysis:
Project one is pitched at the level of the grave(s) inside the burial mounds.
What was the social and ideological identity of the dead? This will be investigated
by analysing the life-cycles of all artefacts found in burial places (Wentink in
prep.).
Project two focuses on the barrow groups (Bourgeois 2013). How and why did
barrows come to form entire landscapes?
14
ancestral heaths
Figure 1.1. An example
of a barrow in which one
quadrant has been excavated
according to the quadrant
method pioneered by van
Giffen. The excavated barrow
in the picture is located at the
Echoput, near Apeldoorn (see
chapter 8.1). Photograph by
Q. Bourgeois.
Figure 1.2. Two barrows at the
Zuiderheide, near Hilversum.
Project three, which is the subject of this thesis, studies the barrow environment.
What did a barrow landscape look like and what was the role of barrows in
this landscape? In this thesis a detailed vegetation history around barrows is
reconstructed in order to get a better impression of what role barrows played in
their environment. The research questions and methods will be discussed more in
detail in Chapter 3.
1.2 The societal significance of environmental barrow
research
Besides the academic concern for doing research on barrows in the Netherlands,
there is also a societal concern. The Dutch public and landowners are very
interested in the barrows in their region. A tourist route in a nature reserve may
pass several burial mounds (see figure 1.2 for an example), with only a small sign
next to the barrow indicating the presence of a burial mound (see figure 1.3). It
is also often the case that very little information about the barrow is available.
Owners of areas with barrows have expressed a desire for more information about
the history of these barrows, and in some cases they want to show what the
barrow landscape looked like at the time of the barrow’s building. Nature reserves
such as the Staatsbosbeheer and Kroondomeinen are interested in reconstructing
barrow landscapes and including the burial mounds in their management and
development of the landscape. But in order to carry out this management, they
need to know what the barrow environment looked like.
The archaeological value of the barrows is not always clear to the public, as
evidenced by the disturbance of several barrows in recent years. For example in
Rhenen-Elsterberg a barrow had been dug into to presumably make a place for a
shelter (Arnoldussen et al. 2009). Greater awareness of the archaeological value of
barrows could prevent such unfortunate unwitting vandalism from occurring.
introduction
15
Many barrows in the Netherlands are protected. However, only the barrow itself
is considered a monument, although there are some exceptional cases where the
protected area around the barrow is extended to a maximum of 10 metres. Since
the role of the barrows in the landscape is not very clear at the moment, it might
be desirable to have the monumental area increased. Ceremonial post alignments
that are associated with the barrows for example may be situated outside the 10 m
zone (Fokkens et al. 2009b). In that case not only the barrow itself was important,
but also the area around it.
16
ancestral heaths
Figure 1.3. A standard
Dutch information sign at
barrow 2 at the Echoput, near
Apeldoorn. Photograph by
A. Louwen, taken during the
excavation campaign in 2007
(see chapter 8.1).
Chapter 2
Environmental research on barrows,
an overview so far
In this chapter an overview of previous environmental barrow research is given.
This chapter starts with a general overview of the Holocene vegetation history of
the Netherlands, followed by a more specific overview of environmental barrow
research.
2.1 The vegetation history of the Netherlands in the
Holocene
Before looking into detail at barrow landscapes it is useful to provide a sketch
of the regional vegetation development during the second part of the Holocene
(from the Subboreal period onwards), the period in which barrows were built.
This vegetation development is mostly derived from pollen records preserved in
peat and lake sediments. The following vegetation development will focus on
the central and southern Netherlands, this research’s area of interest (see section
3.2).
The Holocene is divided into periods based on artefact remains, the vegetation
history of the Holocene, however, is divided into climatic zones based on peat
stratigraphy (Blytt-Sernander) and on data from pollen cores. Three separate
pollen zone descriptive schemes (formulated individually by Firbas, Jessen/
Iversen, and the Rijksgeologische Dienst [RGD, the Dutch State Geological
Service]) are commonly used to describe the Holocene vegetation development in
the Netherlands (see table 2.1).
The first barrows were built during the Subboreal period. A deciduous forest
dominated the Netherlands during the preceding Atlantic period. Quercus, Tilia,
Ulmus and Corylus were the main forest species in the drier regions, with also
Fraxinus increasing its presence throughout this period. In the wetter areas Alnus
was the dominant species. Pinus, a coniferous tree that had been present in large
numbers in the preceding periods, rapidly decreased during the Atlantic and was
almost absent in the Netherlands.
During the Subboreal, which correlates with the Neolithic and the Bronze
Age for most of the Netherlands and pollen zone VIII (in the schemas of Firbas
and Iversen), several changes in vegetation occurred. At its start there is a decline
of Ulmus. In large parts of Northwest Europe this was a very rapid decline,
also referred to as the Ulmus fall. This decline was not as pronounced in the
Netherlands, but still a decrease of a few percentages that can be seen with respect
to the Atlantic. Tilia also decreased and almost disappeared at the end of the
Subboreal, a process that started in the north of the Netherlands and proceeded
to the middle and south of the Netherlands (Waterbolk 1954). This period is
also characterised by the appearance of Fagus. The Subboreal is also the period
where man seriously started to interfere with the landscape. The character of the
vegetation changed. Natural forests were cleared for agricultural activities. In
some places the forest could recover, but in others a Calluna-heath established
environmental research on barrows
17
years (cal BC)
Archaeological
period
Blytt
Firbas
Jessen/
Sernander
(1949)
Iversen
RGD
Vegetation development according
to general pollen diagrams
(1935-1941)
X
Modern History
Vb2
AD 1500
Anthropogenic indicators increase
Medieval Period
Vb1
AD 500
Roman Period
12
250
500
800
1100
Subatlantic
IX
1800
2000
Carpinus >1%
IX
Late Iron Age
Fagus >5%
Middle Iron Age
Va
Carpinus <1%
Early Iron Age
Late Bronze Age
Fagus >1%)
Tilia decreases
Middle BA B
1500
Fagus >5%
Middle BA A
IVb
Early Bronze Age
Late Neolithic B
2500
Late Neolithic A
Subboreal
VIII
VIII
2900
IVa
Middle
3750
Ulmus decreases (<5%)
Fagus increases
Increase anthropogenic indicators
Neolithic
4200
VII
Early Neolithic
4900
Querus, Ulmus, Tilia and Alnus dominant
Pinus decreases
VII
VI
Atlantic
7000
III
Mesolithic
Vb
VI
Va
V
IV
IV
Boreal
II
8000
Preboreal
10000
18
ancestral heaths
I
Corylus, Quercus and Ulmus dominant,
Alnus and Tilia increase, Pinus decreases
Pinus dominant, but decreasing
Quercus, Ulmus, Corylus increase
Pinus and Betula dominant
itself. Natural forests alternated with a cultivated landscape, such as fields, pasture
land and settlements.
The Subatlantic period that followed the Subboreal started around 800 cal
BC (when the Subboreal climate deteriorated) and continues to the present day.
Fagus and Carpinus expanded and Quercus declined. Tilia and Ulmus have almost
disappeared. Herbs became more prevalent, which seemed to be favoured by
human influence. Artemisia, Plantago, Cerealia and grasses gained importance. In
the early Middle Ages, also known as the Dark Ages, the vegetation changed. After
the fall of the Western Roman Empire, during the Migration Period (300-600
cal AD), human pressure on the vegetation seemed to lessen. Forests were able to
recover in the South and Southeast of the Netherlands, while a concurrent decline
of human influence was almost absent in the northern Netherlands (Janssen and
Ten Hove 1971, Renes 1988, Bunnik 1999). For example in the loess area in
the Netherlands between the Rhine and the Meuse Corylus and Quercus could
expand first, succeeded by Fagus and Carpinus. In the wetter areas Alnus was able
to expand enormously (Bunnik 1999). During the Merovingian and Carolingian
dynasty (ca. 600-900 cal AD) human cultivation activities increased again and in
the late Medieval Period most of the natural forest had disappeared due to forest
clearing. Due to (agri-) cultural activities, the soil impoverished and Callunaheath could establish itself at great scale. The heath was exploited (grazing, sod
cutting, etc.) and was therefore able to expand. From the 16th century onwards
the forest was able to regenerate, mostly due to the planting of trees. Pinus was
planted in enormous amounts in the 19th century, and at present Dutch forests
consist of about 20% deciduous forest, 20% coniferous forest and 50% of mixed
forest. (Waterbolk 1954, Janssen 1974, Berendsen 2004, Bastiaens and Deforce
2005).
2.2 Environmental research on barrows
2.2.1 An overview
Table 2.1. An overview of
commonly used pollen zones
for the Holocene period
and the general vegetation
development in the central
and southern Netherlands per
zone. RGD= Rijks Geologische
Dienst.
There are several ways to investigate the prehistoric landscape in which the barrows
were situated. The appearance of a landscape is for a great deal determined by the
vegetation that is in it. No understanding of a barrow landscape can be considered
complete without knowledge of its vegetation. Palynological analysis is a common
way of reconstructing a landscape’s vegetation in the past. In an ideal scenario
pollen analysis can be applied to a deposit that has accumulated over time, such as
peat or lake mud. The pollen rain that precipitated on the surface was embedded
in the deposit as it built up. In this way the peat or the sediment in a lake became
an archive of vegetation history for the surrounding area. When pollen precipitates
onto a soil surface there is no incorporation by layers built on top of the surface
and it is very likely that pollen grains on the surface will be corroded or washed
away. However, after construction of a barrow, the surface containing the pollen
precipitation was covered and protected from the air, reducing microbiological
activity and thus corrosion of the pollen grains. In addition, the tumulus will
prevent new pollen from precipitating on the old surface. The old surface under
the barrow is often still recognizable as a darker layer and can be sampled for pollen
analysis. Besides the old surface, the sods of which the barrow is constructed are
also suitable for pollen analysis, since they also contain the upper part of the soil
profile (Waterbolk 1954, van Zeist 1967b). This topic will treated more fully in
Chapter 4.
environmental research on barrows
19
Soils in Neolithic barrows were first investigated in Denmark by Müller and
Sarauw (Müller 1884, Sarauw 1898). As mentioned in Chapter 1, van Giffen
initiated investigations into the environmental aspects concerning barrows in the
Netherlands and the first pollen analysis of barrows dates from before World War
II. The ideas of van Giffen were carried out and improved upon by Waterbolk
(1954). The barrow database, and barrow interpretations, was later enlarged
through contributions made by van Zeist (1955), and Groenman-van Waateringe
and Casparie (1980). Barrow palynology was practiced mainly in the Netherlands,
although barrows in regions outside the Netherlands were also subjected to
palynological analyses. For example in Belgium barrows were palynologically
investigated by Groenman-van Waateringe (1977) and van Zeist (1963), and in
Denmark by Andersen (1988); Averdieck (1980) and Groenman-van Waateringe
(1979) investigated several barrows in Germany, Dimbleby and Evans (1974) in
England and Groenman-van Waateringe (1983) in Ireland. Knowledge and ideas
about barrows and their environment have evolved during the last century. These
developments will be discussed in the coming sections.
2.2.2 Pollen analyses for dating purposes
Palynology was at first primarily used to date peat and sediment sequences. A
general reference pollen diagram representing the vegetation history of the
Netherlands during the Holocene, based on pollen data from mainly peat and
lake deposits, can be divided into pollen zones (see table 2.1). Pollen spectra from
undated sediment layers can often be fitted into a certain pollen zone and thus
be linked to a certain time period. This method of dating is most reliable when
multiple pollen spectra or a local pollen diagram is provided instead of a single
pollen spectrum in order to create as much overlap with the reference pollen
diagram as possible. In addition, this technique was extended to the dating of
various archaeological objects and sites. An object that was found embedded in
a sediment can be linked to a certain depth in the pollen diagram obtained from
this sediment and therefore to a certain age, when the exact original location of
the object in the sediment is known.
Palynological dating has also been applied to barrow research. When grave goods
are absent, dating a barrow is difficult. When Waterbolk first derived palynological
data from the old surfaces beneath numerous barrows in the Netherlands, they
were used for dating purposes. Two barrows near Apeldoorn, extensively described
and discussed in Chapter 8 of this thesis, were first palynogically dated to a
pollen zone known to be contemporaneous to the Iron Age. This dating was later
confirmed by the radiocarbon dating of charcoal from the ring ditches surrounding
the barrows (see Chapter 8). However, dating barrows using palynological data
has not always been completely accurate. For example, a large group of barrows
at Toterfout-Halve Mijl has been chronologically ordered mostly based on their
pollen spectra by Waterbolk (1954). Bourgeois, however, has shown that the
chronological sequence of these barrows should probably be different, on the
basis of radiocarbon dates and the surrounding features (Bourgeois 2013; see also
Chapter 11).
Dating by palynological analysis is a form of relative dating, since chronologic
checkpoints from other sources are needed. Presently, absolute dating methods
like radiocarbon dating and OSL dating have displaced palynological dating to
all intents and purposes. However, the method still finds application on occasion,
when no datable material is available.
20
ancestral heaths
2.2.3 The reconstruction of local vegetation: regional and cultural
differences
Besides dating, pollen data have also been used to reconstruct vegetation in the
vicinity of the barrows at the time the barrows were built, providing information on
the agricultural systems used by prehistoric man. In addition, these palynological
analyses were used to show differences in land use (Waterbolk 1954, Van Zeist
1959, 1967a). Two different agricultural systems– the Iversen landnam and
Troels-Smith landnam- can be distinguished during the Neolithic. These types of
land use are named after the two Danish scientists who first described them. The
Iversen landnam is a Neolithic land occupation phase (in the Middle Neolithic B),
describing the clearance of the primeval forest by burning and cutting trees. The
Iversen landnam was first described by Iversen (1941, 1973), based on the results
of palynological analyses of Danish small lakes. The Iversen landnam consists of
three phases, which can be recognised in the pollen diagrams as follows:
Phase 1: The first phase represents the actual forest clearance by cutting and
burning: at first Ulmus declines, followed by the decrease of Tilia and Quercus.
The pioneer tree Betula shows an increase.
Phase 2: This phase corresponds with the agricultural phase, involving grazing
and crop cultivation. Anthropogenic and grazing indicators show a maximum;
particularly Plantago lanceolata, but also Cerealia, Poaceae and Rumex acetosella.
Phase 3: The third phase represents the abandonment of the pastures and fields,
allowing regeneration of the forest. This is shown by a maximum of Corylus,
the increase of mainly Quercus, Fraxinus and Tilia and the decrease of Betula.
Anthropogenic and grazing indicators decrease and disappear almost completely.
Troels-Smith introduced a second type of Neolithic occupation in Denmark,
prior to the Iversen landnam (Troels-Smith 1953). He found various agricultural
indicators, such as cereal pollen, contemporaneous with the Ulmus decline (around
3750 cal BC, see table 2.1). Troels-Smith suggested that the fall of the elm curve
reflected pollarding of the trees for the purpose of cattle fodder. Together with
the absence of pastures, deduced from very small numbers of Plantago lanceolata,
Troels-Smith concluded that a farmer culture existed preceding the Iversen
landnam, mainly based on small-scale arable farming with livestock kept within
enclosures throughout the year.
The investigations and interpretations offered by Iversen and Troels-Smith
triggered similar investigations in the Netherlands. In pollen diagrams derived
from peats in the province of Drenthe (in the north of the Netherlands) named
Bargeroosterveen, Emmen and Nieuw-Dordrecht, the two types of landnam
were shown to have occurred (Van Zeist 1959, 1967a). In the period between ca.
3700 cal BC and 2800 cal BC the Ulmus decline can be seen, together with low
percentages of Plantago lanceolata. The data reflect the type of land use described
by Troels-Smith, characterised by small forest clearances and cattle kept within
enclosures. In the period after ca. 2800 cal BC an increase of Plantago lanceolata
can be observed, signalling the Iversen-landnam, with rather large cleared forest
areas mostly used for grazing.
Van Zeist compared these pollen diagrams with spectra from Neolithic grave
monuments. Grave monuments from three Neolithic cultures were investigated:
megalithic tombs built by people from the oldest culture, the Funnel Beaker
Culture (FB) (ca. 3400-2900 cal BC, van den Broeke et al. 2005, 28) and barrows
built by people belonging to the later Protruding Foot Beaker Culture (PFB) (ca.
2900-2500 cal BC, van den Broeke et al. 2005, 28) and the Bell Beaker Culture
environmental research on barrows
21
(BB) (ca. 2500-2000 cal BC, van den Broeke et al. 2005, 28). Pollen spectra
from the FB grave monuments showed low values of Plantago lanceolata, Rumex
and Poaceae, indicating a Troels-Smith landnam. Pollen spectra from the PFB
burial mounds showed in general high percentages of Plantago lanceolata, Rumex
and Poaceae, corresponding to the Iversen landnam. Pollen spectra from the BB
mounds were very similar to those of the FB, characteristic of the Troels-Smith
landnam. Van Zeist ascribed the Troels-Smith phase in the Bargeroosterveld
diagram between ca. 3700 cal BC and ca. 2800 cal BC and the Funnel Beaker
Culture and the Iversen phase after ca. 2800 cal BC to the Protruding Foot Beaker
Culture (see figure 2.1). The Troels-Smith landnam used by the farmers of the
Bell Beaker Culture was not shown in the Bargeroosterveld diagram. Van Zeist
explained this by the dominating activities of the people of the Protruding Foot
Beaker Culture (Van Zeist 1959).
Waterbolk also noted in his thesis that the maxima of herbs (like Plantago,
Rumex, Poaceae, Dryopteris-type, Asteraceae and Caryophyllaceae) he found in
the pollen spectra from barrows must have been caused by activities of the Corded
Ware Culture, also known as the Protruding Foot Beaker Culture, who apparently
practised an Iversen landnam (Waterbolk 1954).
For the Early and Middle Bronze Age Period van Zeist suggested a difference
in farming practice between the north and the south of the Netherlands, based
on the pollen spectra from barrows (Van Zeist 1967a). In the northern part of
the Netherlands barrow pollen spectra showed high values for Plantago lanceolata,
Rumex and Poaceae, comparable to the spectra from the Protruding Foot Beaker
Culture (e.g. Iversen landnam). In barrows in the south of the Netherlands,
especially those belonging to a regional culture called Hilversum Culture, the
percentages of Plantago, Rumex and Poaceae were considerably lower than in the
north, suggesting that the farming practice more resembled that of the Funnel
and Bell Beaker Culture (e.g. Troels-Smith landnam). Van Zeist found that these
differences in agricultural practice interestingly coincide with differences in
culture between the north and the south of the Netherlands, namely the culture
of the Barbed-wire Beakers in the north and the Hilversum Culture in the south.
The theory that these differences in land use were culturally bound was criticised
by Casparie and Groenman- van Waateringe (1980). Their article (re)analysed
many pollen spectra from barrows north and south of the IJssel river (see figure
22
ancestral heaths
Figure 2.1. Cultural
differences shown in
barrow pollen spectra. I=
Funnel Beaker Culture, II=
Protruding Foot Beaker
Culture, III= Bell Beaker
Culture. Figure after van
Zeist (1959, figure 11).
Figure 2.2. The areas north
and south of the IJssel
examined by Casparie and
Groenman-van Waateringe
(1980).
2.2) and compared them to the peat pollen diagrams from Bargeroosterveen,
Emmen and Nieuw-Dordrecht (Van Zeist 1959, 1967a).
The differences between the Funnel Beaker and Protruding Foot Beaker period
ascertained by van Zeist in the peat diagrams were, according to Casparie and
Groenman-van Waateringe, not the result of differences in type of land use, but
of soil conditions and nature of the cleared forest. The FB people living near the
sampling sites apparently preferred to reclaim the Ulmus- and Tilia-rich forests
that were present on soils relatively rich in nutrients. These forests were mainly
situated on cover sand deposited on a weathered boulder-clay ridge. Because of
this boulder-clay in the subsoil, the sandy soils were loamy and moist to wet.
When the clearings were abandoned no great expansion of Plantago lanceolata
took place here. Rather than explain this by the type of landnam activity, Casparie
and Groenman- van Waateringe explained the absence of a Plantago lanceolata
resurgence was due to the compactness of the soil (Casparie and Groenman-van
Waateringe 1980, 59):
“It is conceivable that the loamy to very loamy soils of the boulder-clay ridge became
compacted very readily, a process that checked considerably the establishment of
Plantago lanceolata.”
The PFB people also cleared forest that had developed on cover-sand that was
generally considerably poorer in nutrients, more drought-susceptible and far less
loamy. Especially the latter was more in favour of Plantago lanceolata, which was
able to expand here. So, Casparie and Groenman-van Waateringe explained the
differences between the FB and PFB period as a result of which type of forest was
cleared (rich versus poorer) and the condition of the soil (loamy versus less loamy
and wet versus drier). In the period of the Bell Beaker (BB) Culture, Tilia shows a
definitive decline, with an expansion of Corylus, Pteridium, Poaceae and Plantago
lanceolata indicating the clearance of already degraded forest.
environmental research on barrows
23
In addition, pollen spectral differences between barrows belonging to the FB
and BB period on the one hand (e.g. low values for herbaceous plants, ascribed
to the Troels-Smith landnam) and PFB period on the other (e.g. relatively high
values for herbaceous plants, ascribed to the Iversen landnam) described by van
Zeist (1967a) were not as explicit in Casparie and Groenman-van Waateringe’s
results. Within each culture, pollen spectra showed considerable differences,
therefore the pollen spectra alone could not be used to culturally isolate a group.
The differences van Zeist found in barrow pollen spectra were more likely to
be due to dissimilarities in soil type, since all PFB barrows were located on the
Drents plateau (Drenthe, northern Netherlands) and the BB barrows were located
on the Veluwe (central Netherlands). Casparie and Groenman-van Waateringe
did in fact find some differences between barrow pollen spectra from the Drents
Plateau (north of the IJssel) and the Veluwe (south of the IJssel). The northern
barrow pollen spectra showed an earlier and more pronounced expansion of
heath than the southern spectra. These differences were ascribed to differences
in the hydrological situation. The northern barrows were nearly all situated on
the Drents Plateau, where soils were influenced by the presence of impervious
boulder-clay not far below the surface.
“It was therefore precisely here, that disturbance of the vegetation cover and
agricultural activities resulted in rapid exhaustion of the soil and a very droughtsensitive topsoil, that in many places facilitated rapid expansion of the heath.”
(Casparie and Groenman-van Waateringe 1980, 60)
In the area south of the IJssel the barrows analysed by Casparie and Groenmanvan Waateringe (1980) were situated on the Veluwe. The Veluwe is a landscape
consisting of pushed moraines, cover sands and fluvio-glacial material of porous
nature, where water seeps down more easily. As a result the soils are much drier
than on the Drents Plateau. According to Casparie and Groenman-van Waateringe
the forest was therefore probably more open at the Veluwe with a well-developed
undergrowth of herbaceous plants sufficient for grazing. Grazing pressure caused a
gradually opening up of the woodland, allowing grasses and heath to expand. The
research by Casparie and Groenman-van Waateringe showed that differences in
land use were not culturally bound. As Casparie and Groenman-van Waateringe
concluded:
“It seems more likely that prehistoric man adapted his methods of reclamation to
a great extent to the possibilities available, and in such a way that no culturallylinked pattern is evident.” (Casparie and Groenman-van Waateringe 1980, 62)
2.3 Vegetation reconstructions of the barrow environment:
open spaces in the landscape
It has become clear from research that most of barrows in the Netherlands were
built in open spaces. These open spaces might have been small or large. Waterbolk
(1954) mentioned that practically all barrows were built in an open space without
deliberate clearance of the area. Van Zeist (1967a) suggested after analysis of
pollen data from several Neolithic and Bronze Age barrows in the Netherlands
that they were constructed in either small clearings (Troels-Smith landnam) or
larger clearings (Iversen landnam). Casparie and Groenman- van Waateringe
concluded from their research (1980):
24
ancestral heaths
“The environment in the immediate vicinity of a barrow varied from only slightly
degraded forest to extremely degraded, heath-rich vegetations, with all possible
intermediate stages.”
De Kort palynologically investigated several barrows in the Netherlands.
For a cemetery complex in North Brabant called Oss-Zevenbergen (see for an
extensive description and discussion Chapter 12) he concluded that all barrows
he investigated were erected in an open place covered with heath vegetation. This
open place was probably already present before the oldest barrow was constructed
and continued to be present during the period the barrows were built (e.g. from the
late Neolithic until the Iron Age) (de Kort 2009, 166, 169). In another cemetery
complex near Slabroek in North Brabant, an urn field that also contained some
barrows that probably dated to the Bronze Age, de Kort found that the oldest
barrow was built in a small open place with heath vegetation (see also chapter 12).
The heath at this open place probably expanded during the Bronze Age when the
younger barrows were built (de Kort 2010, 64).
Barrows in regions besides the Netherlands were also found to have been built
in open places. Andersen found indications that in the Vroue area, West Jutland
(Denmark), Early Neolithic barrows were built in natural woodland with heath
patches (Andersen 1994-95). Later on trees became increasingly scarce and open
spaces became larger. For Early Bronze Age barrows in Thy, Denmark, Andersen
found indications that they were built in a rather treeless landscape, with remnants
of woodlands that probably had been in the area some time before the barrows were
built (Andersen 1996-97). It has been suggested that burial mounds in southern
Sweden were built in a rather open landscape, with forest cover estimated at 2040%, falling to 10% in the immediate surroundings of the barrow itself (Hannon
et al. 2008).
The open spaces barrows were built in have mostly been interpreted in terms
of prehistoric man’s land use. Let us now focus on the open space itself and its
relation to the barrow. First an overview of possible open spaces and their origin
will be given, than follows an overview of open spaces in which barrows were
built.
2.3.1 An overview of open spaces
Natural open spaces
The general view is, as has been described in section 2.1, that a closed canopy
forest developed in the beginning of the Holocene in Western and Central Europe.
When human interference with the landscape, the density of the forest decreased
and open spaces were created. There is, however, an alternative hypothesis: a halfopen park-like landscape, described by Vera (1997) as a landscape consisting of
a continuous grassland with clumps of shrubs and forest. Vera claims that the
initial Holocene vegetation of Western and Central Europe was not a closed
forest system, but a half-open-park-like landscape. He points out that Quercus
and Corylus would not be able to flower and regenerate in closed forests, while
these species were continuously present in considerable numbers in Central and
Western Europe since the last ice age. Vera’s suggested type of half-open-park-like
landscape was created and maintained by large herbivores, in a process he calls the
theory of cyclical vegetation turnover (Vera 1997).
In Vera’s cyclical vegetation turnover, thorny shrubs establish themselves in
the grassland. In these clumps of thorny shrubs trees could grow, protected from
grazers by the thorns. The trees developed into a forest, which would degenerate
environmental research on barrows
25
back into grassland again due to large herbivores and climatic events such as
drought and storms. The process could start over again, with the establishment of
thorny shrubs in the grassland (Vera 1997).
Mitchell (2005) tested the hypothesis of Vera that large grazers kept the forest
open. He compared palynological data of Quercus and Corylus from Ireland,
where only two large herbivores were present during the Early Holocene, to that
from other European countries with a greater assortment of large herbivores
(Mitchell 2005). He found no obvious differences in Quercus/Corylus regenerative
progression and concluded that large herbivores would have had little impact on
the abundances of Quercus and Corylus. Mitchell also argued that, based on data
from small forest hollows in Europe and eastern USA, opening up of the forest
canopy was mainly artificial and caused by human activities.
Nevertheless, other researchers join Vera in believing that the natural structure
of the northwest European forest in the Early Holocene was probably more open
than previously thought. Svenning stated that closed forest would be predominant
in ‘normal’ uplands, but with longer-lasting openings (Svenning 2002). These
openings would have mainly occurred on floodplains, on calcareous or poor,
sandy soil and in the continental interior of northwest Europe. At these locations
the appropriate conditions would have existed for the presence of open vegetation
like open woodland, scrub, heath and meadows. Fire would probably have been
an important agent involved in the maintenance of this vegetation. Bradshaw et
al. also argued that closed forest theory alone is not a perfect model for the Early
Holocene vegetation structure (Bradshaw et al. 2003). They agreed that closed
forest canopy is the dominant vegetation type, but they also argued that some
parts of the landscape were open. This openness might have been created and
maintained by events like floods, fires and wind throw. A combination of fire
and grazing pressure may have created proper circumstances for regeneration of
Quercus and Corylus. Whitehouse and Smith discussed that other proxy indicators
may provide useful information that contributes to this subject (Whitehouse
and Smith 2010). They showed, using beetle records from archaeological and
palaeoecological sites in Britain, that the early Holocene was characterized by
quite open woodland and that locally open areas may have played an important
role. They found little evidence that those open areas were maintained by grazing
activity of large herbivores, and proposed that other disturbance factors were
probably of more importance.
To conclude, there are numerous indications that the west European Holocene
landscape was probably more open than previously thought.
Fabricated open spaces: forest clearance
The landscape started to change rapidly with the onset of prehistoric man’s
interference. During the Neolithic, man switched from a hunter-gatherer strategy
to an agricultural strategy. Farmers started to plant their own food and began to
keep their own animals. This change to crop cultivation and animal husbandry had
great impact on the vegetation and consequently on the landscape. Agricultural
practise required open spaces for arable fields and livestock too, needed pasture to
graze in. Forests were cleared and from the period of around 4100 cal BC, human
influence becomes visible in palynological research in the form of cereal pollen
grains and weeds from both arable and pasture land (Louwe Kooijmans 1974,
Out 2009, Chapter 8 in this thesis). From 3000 cal BC there is a pronounced
human impact on the environment. Both agriculture and stock breeding were
practised on a large scale. For agriculture open space was needed on the most
26
ancestral heaths
fertile grounds. Forest areas were cleared, notable in palynological records by
the rapid decline in trees, the increase of herbs growing in open vegetation and
indicators for cultivation, such as cereal pollen (Louwe Kooijmans 1974).
The influence of human activity on the landscape was mainly notable in
pollen spectra by the presence of anthropogenic indicators (Behre 1986). Bakker
for example, reconstructed the emergence and expansion of agriculture on the
Drenthe Plateau (eastern Netherlands) by using the indicator-species approach
in combination with the use of modern pollen/land-use relationships (Bakker
2003). Bakker demonstrated that the first small-scale arable farming and livestock
foddering took place on the Drenthe Plateau in the Subboreal (4050-3450 cal BC,
according to Bakker 2003). An increase of Poaceae, Cyperaceae, Calluna, Plantago
lanceolata and Rumex acetosa-type indicate the presence of various types of grassrich vegetation, probably maintained by livestock. The appearance of Cerealia
indicates the presence of arable fields. In the following phase (3450-2600 cal BC,
Bakker 2003) more widespread clearances occurred, especially in the rich and
higher forest. The further increase of Cerealia indicates the increased importance
of arable fields.
After a period of decreased human influence on the vegetation during 26001770 cal BC (Bakker 2003), a more extensive clearance of the forest and their
replacement by agricultural fields can be seen in the later phase of the Subboreal
(1770-800 cal BC, Bakker 2003). Cleared forest areas could be used for crop
cultivation for several years until the soil was exhausted. On these fallow fields
grasses were able to expand and could be used as pasture (Groenman-van Waateringe
et al. 1968). Grazing animals prevented the forest from regenerating and besides
grasses heath was able to establish itself on the abandoned fields. Bakker (2003)
showed that in the Subatlantic (800 cal BC-1500 cal AD), the exhausted and
abandoned fields on the Drenthe Plateau were dominated by Calluna and extensive
heath fields dominated the landscape. Heath was also grazed and maintenance
and expansion of the heath was ascertained. The maintenance of these heath areas
will be further discussed extensively in Chapters 8-13. Forest clearance might also
have taken place for the sole purpose of providing pasture for grazing.
Forest clearings could have been accomplished by tree felling. Felled wood and
other vegetation from the forest clearances could have been used as raw material,
as fuel and served to cattle. As a raw material wood could serve as construction
material for several structures in a settlement, such as houses, sheds, fences and
palisades. Bakels for example showed that Linearbandkeramik settlements in the
southern Netherlands (ca. 5300-4900 cal BC) used large quantities of wood. For
a settlement of 200-250 houses that were built over a period of about 400 years,
a woodland area of 50-1000 ha was needed (Bakels 1978). Wooden structures
have also been found in association with barrows. Barrows were for example often
encircled by wooden posts in the form of palisaded ditches (Late Neolithic), widely
spaced post circles (1800-1400 cal BC) and closely spaced post circles (1700-1300
cal BC) (Bourgeois 2013, 34-36). Besides its use as fencing, wood was also used
for the pyre when a body was cremated and in some cases a body was buried in a
wooden coffin or a burial chamber constructed of wood (Bourgeois 2013).
Another method of forest clearing is burning. The deliberate use of fire to
manipulate the vegetation in prehistory has been suggested by several authors
(Mellars 1976, Simmons and Innes 1987, 1996a). Simmons and Innes suggest that
fires were a deliberate tactic for resource management as early as the Mesolithic
(Simmons and Innes 1996b). The resultant opening up of the landscape would
have facilitated hunting by improving the sight and/or making the landscape more
attractive for certain game species.
environmental research on barrows
27
A combination of cutting and burning wood is applied in the so-called slash and
burn agriculture. The forest is felled and the wood is left to dry, to be later burned.
With this technique the soil is mixed with ash, enhancing the soil’s fertility for
crop cultivation. It has been suggested that slash and burn agriculture was already
taking place in the Neolithic, as Iversen connected Neolithic forest clearance (e.g.
Iversen landnam, see section 2.2.3) with slash and burn (Iversen 1941). Large
amounts of charcoal in soil samples may be taken as indication of the use of fire.
Odgaard suggests that charcoal layers found in soil samples indicate the use of fire
in clearing woodland (Odgaard 1994). Andersen mentions deformed tree pollen
grains found in Neolithic barrow soil samples (Andersen 1994-95). The deformed
tree pollen grains were interpreted as an indication that trees had been felled
and burned, when lying on the ground (Andersen 1992, 1994-95). However,
deformed pollen grains were mixed with non-deformed herbaceous pollen grains,
indicating that regeneration of the burnt area had already started. Therefore, in
this case burning of the trees had already taken place sometime before the barrows
were built. In some barrows in the north of the Netherlands high concentrations
of charcoal particles were found, indicating that the local vegetation was burned
intentionally before the barrow was built (Casparie and Groenman-van Waateringe
1980). Hannon et al. (2008) also found charcoal particles in most of the barrows
they investigated on Bjäre Peninsula, southern Sweden. They concluded that slash
and burn agriculture was practised in the area.
2.3.2 Which open spaces were chosen for the building of barrows?
Open spaces, whether created by man or by nature, were present in the Neolithic
landscape. Since the Neolithic, man’s interference with the landscape grew in range
and magnitude. Forests were cleared and over time the vegetation became more
and more open. During the Neolithic period erecting barrows in open spaces was
already an established practise. Choice in open spaces was in all likelihood limited
at that time, although the landscape may have been more open than previously
thought (see section 2.3.1). In the Bronze and Iron Ages, the availability of open
space was certainly greater. However, what do we know about the open spaces in
which a barrow was set? Open spaces were created by man, but where these open
spaces also chosen as building site for a barrow?
Barrows in arable fields
Cleared forest areas, mostly used for agriculture (see section 2.3.1), may have been
chosen as sites for constructing barrows in. Some barrows were probably built on
arable land that had recently or since a longer period been abandoned. Casparie
and Groenman-van Waateringe (1980) found that, especially in the northeast of
the Netherlands (Drenthe; see figure 2.3), the open spaces where barrows are
placed were previously used as arable land, and that they were probably already
long abandoned before the barrows were built. The open spots might originally
have been cleared for agricultural purposes, but at the time the barrows were built
the agricultural fields were no longer in use. In the central Netherlands (Veluwe,
Gooi and Utrechtse Heuvelrug; see figure 2.3) indications for arable land are
scarce. Casparie and Groenman-van Waateringe (1980) noted the difficulty in
establishing with certainty whether an area had been used for crop cultivation, but
concluded that in general, barrows were seldom constructed on or in the vicinity
of arable land then in use.
Research outside the Netherlands has shown that barrows were not often built
on arable land. Andersen (1994-95) found indications that mounds in Denmark
were often built at sites that were less intensively exploited than areas in the near
28
ancestral heaths
Figure 2.3. The Drenthe,
Veluwe, Gooi, and Utrechtse
Heuvelrug regions.
vicinity. Pollen spectra from some barrows in the Vroue area, (West Jutland,
Denmark) dating to the Middle Neolithic, showed traces of agriculture, but on
the whole the pollen spectra from barrows in the Vroue area showed no evidence
of agricultural practice. In Thy (West Jutland, Denmark) Andersen concluded
that the Early Bronze Age barrows were built in pastureland and that only some
of the barrows were built in recently cleared coppice wood that had been used
for cereal cultivation prior to the barrow building. Lawson et al. suggests that
there seems to be a correlation between soils and the distribution of barrows in
Norfolk (Southeast England), where barrows were placed on agriculturally poor,
light soils (Lawson et al. 1981). Altogether there seems to be a preference for
building a barrow on a location that had not been used as arable land recently. It
has even been suggested that barrows were preferably built on marginal land, so
that no (economic) valuable land that could be used for cultivation was wasted
(Ovrevik 1990, Field 1998). However, as Downes mentioned, this marginal land
might have been very useful for other purposes (boggy ground could have served
as source for fuel for example) and not have been as insignificant as assumed
(Downes 1994)
Barrows in pastoral zones
The change in the Neolithic to a more agricultural way of living also included the
raising of livestock. Farming communities became more and more dependent on
livestock to provide meat, dairy products, manure and wool, leather or other raw
materials, as well as for pulling ploughs. Livestock needed pasture for grazing, at
least for part of the year. They might have been grazing in natural open places in
the forest. Groenman–van Waateringe found, however, that a Neolithic farmer
had to open up the forest, since woodland composed of less than 30% grasses was
not suitable for grazing (Groenman-van Waateringe 1993). Adams also mentioned
that forest cover needed to be less than 50% (Adams 1975). Forests were cleared
environmental research on barrows
29
Figure 2.4. It was often
assumed that barrows were
located close to settlements.
This figure shows a schematic
drawing of two Bronze Age
households, with barrows
located at the settlement site.
Figure after Fokkens (2005b,
figure 20.3A).
for crop cultivation, but possibly also to create grazing areas for cattle. In addition,
abandoned fields might have served as pasture (Groenman- van Waateringe et al.
1968).
Were spaces that were used as pasture also used to build barrows? Recent new
research on the Vorstengraf barrow in North Brabant (see also section 12.1)
shows that this barrow was probably built in an open space already present long
before the building took place. This open space was covered with heath vegetation
during that entire period, which might have lasted for several centuries. De Kort
concluded that this heath vegetation had been used as pasture, with probably
sheep grazing in the open spot. This might indicate that the barrow cemetery,
where besides the Vorstengraf several other barrows are located, was deliberately
kept open, while grazing prevented tree species from establishing and forest
gradually covering the open place (de Kort 2002). Casparie and Groenman-van
30
ancestral heaths
Waateringe (1980) found some indications that barrows were built on pasture
land, however, they conclude that it is extremely difficult to determine this with
certainty (Casparie and Groenman-van Waateringe 1980).
The practise of mounds construction on pasture land also finds support outside
the Netherlands, for example in Orkney (Bunting and Tipping 2001) and Thy,
Denmark (Andersen 1996-97). Odgaard reports that two barrows that were built
in Calluna heathland in Jutland (Denmark), where grazing had probably taken
place (Odgaard 1988). Karg concluded that the heathland where the barrows in
Skelhøj were built had been used as pasture as a form of heath management (Karg
2008).
Open spaces created for barrow building
There is also the possibility that open spaces were created for the purpose of
barrow construction. Some barrows had been constructed in an area where the
local vegetation was destroyed by fire shortly before the barrow was constructed.
Samples from these barrows consisted almost exclusively of charcoal particles,
which may indicate that the area was cleared intentionally before a barrow was
constructed. This intentionally burning of the area could have been some kind of
ritual activity. However, it may also represent a certain phase in the landnam and
have no direct connection to the burial. Casparie and Groenman-van Waateringe
(1980) could not find evidence of forest clearance for the purpose of burial of the
dead.
Barrows and settlements
It is often assumed that barrows were built close to settlements (see figure 2.4).
For the Middle Bronze Age Roymans and Fokkens argued that barrows were
constructed in the near vicinity of the houses (Roymans and Fokkens 1991).
Barrows were assumed to be family graves and families buried their deceased
relatives underneath a barrow close to their settlement. This theory is mainly based
on the settlement excavated in Elp, where a Bronze Age barrow and several flat
graves were situated close to several houses remains. This cemetery was assumed
to be in use by the inhabitants of the settlement (Waterbolk 1964). Bourgeois
and Fontijn tested the hypothesis of Roymans and Fokkens by re-analysing the
data from the only 15 sites where traces of both houses and barrows dating to the
Middle Bronze Age were found in close association (Bourgeois and Fontijn 2008).
The houses and barrow of Elp seem to be contemporaneous, which also applies
for three other sites Bourgeois and Fontijn analysed. They showed however, that
most barrows that were found in Middle Bronze Age settlements were much
older than the houses in question, with Elp forming an exception rather than
the rule. Middle Bronze Age barrows were not built close to houses, but Middle
Bronze Age houses were often built close to already existing barrows, which were
then re-used by the residents of the settlement. They emphasize however, that
the number of sites that could be used for such analyses is very low and that no
firm conclusions can be drawn yet. For the Late Neolithic and Early Bronze Age
barrows there is hardly any evidence that they were built close to houses. Casparie
and Groenman-van Waateringe (1980) mention some PFB barrows were built
on abandoned settlement areas, based on artefact finds. In fact there are very few
examples of Late Neolithic and Early Bronze Age settlements, making it difficult
to draw any conclusions on the relation between burial mounds and settlements
in the Late Neolithic and Early Bronze Age.
environmental research on barrows
31
On the one hand one could expect that barrows were placed close to a
settlement, where one lived close to one’s deceased ancestors. On the other hand,
the place a burial mound was located in could be seen as a ritual and/or sacred area
that would be kept separate from the world of the living.
2.3.3 What was the size of the open spaces barrows were built in?
Barrows were built in open spaces in the forest. It is likely that open spaces in
Neolithic times were smaller than in later periods, since prehistoric man created
more and more openness in the forest during the Holocene. However, there is
not much known about the size of the open places that were used to build burial
mounds. Sods were used to construct a burial mound. These sods were most likely
taken from the near surroundings of the place where the barrow was planned
(Waterbolk 1954, van Zeist 1967a). This suggests a larger open place was necessary
than just the size of the barrow. Jonassen concludes that in a forest non arboreal
pollen (NAP) shows values of approximately 10%, but that a few hundred metres
from the forest values rise up to about 100% NAP. Spectra with NAP of 100500% could indicate an open landscape in a forested area with forest at a distance
of about 1 km (Jonassen 1950, 71-72). Waterbolk (1954) estimated the size of
the open space around the Neolithic barrows at a few to tens of hectares. De Kort
estimated in his MA-thesis that the size of the open space that was needed to take
sods from to build the Vorstengraf barrow in Oss was about 1.5 ha (de Kort 1999;
see also Chapter 12).
Conclusions
A large amount of vegetation data of barrows is available, as has been described in
the previous paragraphs. The data that informs us on how the barrow landscape
looked like is still limited, however, and many questions about the barrow
environment remain. The next chapter will be on this subject.
32
ancestral heaths
Chapter 3
Barrow research, missing data
3.1 Research questions
The barrows of the Netherlands have been the source for many reconstructions of
prehistoric local vegetation. Barrows were built in open spaces, in areas that could
have been used for several purposes before the construction of the barrow (see
previous chapter). And yet, what the total landscape around the barrow looked
like during the barrow’s construction, and the history of the area prior to the
barrow’s erection, represents a great lacuna in the history of barrow research. This
lack prompts the first research question:
1. What did a barrow landscape look like and what was the vegetation (history)
around barrows?
Was the origin of the open space (e.g. how the open space originated and its
original function) influential, affecting the builder’s choice on the barrow’s
setting? Hardly any evidence supports the idea that the barrows were built
in areas that were cleared for burial rite activities. The open place that a
burial mound was raised in probably had a longer existence as an open space,
before becoming the site of a burial mound. It might have been used for crop
cultivation or as pasture, or the open space might have served as a settlement
location. It has been suggested that the barrow builders had a preference for
ancestral grounds, land that has been used by their ancestors. In several cases
indications have been found that barrows were built on a location with a
history of pasture (see section 2.3.2). This conscious decision, if true, suggests
there might be a relation between barrows and pastoral zones. The second
research question has been formulated as follows:
2. Were barrows built on ancestral grounds? What is the relationship with
pastoral zones?
In addition to our ignorance on the origin of open spaces, what also is unknown
is the size of the open spaces. The size of the open space is important for the
understanding of the role of barrows in the landscape, for knowing the size
of the open space tells us something about the visibility of the burial mound
and the barrow landscape: Were they built in small open spaces with a short
distance to the forest, where surrounding forest probably prevented the sight
from and towards the mound? Alternatively, were they built in large open
areas, so they were well visible from the environment and offered a good view
towards the surroundings? In addition, the size tells us about the method by
which it was cleared and the energy requirements in maintaining the open
space.
barrow research, missing data
33
3. What was the size of the open space barrows were constructed in and what
was the distance to the forest?
The previous research questions lead to the last research question, concerning
the role of barrows in the landscape.
4. What was the role of barrows in the landscape? How can the history of the
barrow environment be linked to that of the natural and cultural landscape
in the surroundings?
Since there is a public interest in knowing more about barrows (see Chapter
1), an additional research goal can be appended to the research questions
described above:
5. Supplying Staatsbosbeheer and other authorities with advice and suggestions,
to aide in reconstructing the original environment around barrows for
purposes of tourism.
3.2 Research area
The research area encompasses the southern and central Netherlands (see figure
3.1). This area was chosen for the numerous barrows found there and for the
time periods (from the late Neolithic to the Middle Bronze Age [2900-1100 cal
BC, see table 2.1]) that are represented by these barrows. Previous excavations in
these regions have yielded a lot of data, which will be reconsidered in this research
project (Waterbolk 1954, Casparie and Groenman-van Waateringe 1980). In
addition, the owners of nature reserves in this region are very interested in the
role that barrow research in the development of cultural tourism and adequate
heritage management.
Figure 3.1. An overview
of all case-study areas and
all known barrows in the
Netherlands. Boxes indicate
the areas presented in the
case-studies. Figure after
Bourgeois (2013), figure 1.4.
34
ancestral heaths
3.3 Research methods
Below a brief overview is given of the methods used to answer the research
questions. The methodology is further discussed in detail in part two of this thesis
(Chapters 4-7).
RQ1 and RQ2: What did the barrow landscape look like and were barrows built on
ancestral grounds?
Vegetation reconstructions (RQ1) were made using data derived from pollen
analyses taken from barrow sites. These environmental reconstructions provide
information about the prehistoric land-use that was in practice before and
at the time the barrows were built (RQ 2). Extant data sets were explored and
reconsidered in five case-studies (Chapters 8-12). To expand the original data sets
additional sampling of barrows was undertaken as well (Chapters 8 and 12). In
addition to single pollen spectra, pollen diagrams from the soils underneath the
barrows were made. From these diagrams vegetation development in the barrow
landscapes through time could be reconstructed. Despite possible factors of
disturbance (see Chapter 5), buried mineral soils appear to be suitable for pollen
analysis, as has been demonstrated by past researches. For example in Harreskov,
Jutland, where Odgaard and Rostholm obtained a pollen spectrum from a fossil
soil found under a barrow (Odgaard and Rostholm 1987). The diagram showed a
clear vegetation development, corresponding to the development shown by a peat
diagram. Calibration of these pollen diagrams is necessary to determine the timedepth relation. Until present a calibration value of 10 cm per 300 years was used,
defined by Dimbleby, based on a buried soil in Suffolk (East of England; Dimbleby
1985). A calibration based on pollen diagrams of Dutch Pleistocene sandy soils
with known age is necessary for this research. The necessity of this calibration
is further explained and discussed in Chapter 5. Besides pollen diagrams, single
pollen spectra were used to compare the ancient surface data from clusters of
barrows of differing ages belonging to one barrow group.
RQ3: What was the minimum size of the open spaces?
Barrows were constructed with sods, probably taken from the immediate vicinity
of the barrow. The number and size of these sods that were used to build the
barrow can provide information about the minimum size of the open area around
the barrow. Pollen data from sods were compared to pollen data from the old
surface, to ascertain whether the sods were taken in the immediate surroundings
of the barrow (Chapter 7).
The vegetation reconstructions undertaken provide information about the
size of the open spot. The ratio of arboreal to non-arboreal pollen was used to
estimate the distance of the barrow to the forest edge (Chapter 7). To refine these
reconstructions, a recent open area surrounded by forest with known vegetation
cover was sampled at increasing distances from the forest border. These pollen
spectra were used to calibrate the barrow pollen data.
RQ 4: What was the role of barrows in the landscape?
The answers to research questions 1, 2 and 3 provide the foundation from which
RQ 4 can be posed. To understand the role of the barrows in the landscape it is
necessary to know what the landscape looked like and what vegetation was present
at and around the barrow site prior to and at the time the barrows were built
(RQ1). To link the barrow landscape to the natural and cultural surroundings, the
origin of the open area, and what it was used for, should be reconstructed (RQ2).
barrow research, missing data
35
The reconstruction of the size of the open area (RQ3) gives valuable information
about the role of barrows in a wider landscape, while providing welcome insights
on the visibility and impact of a barrow on its surroundings (chapter 13).
RQ5: Cultural tourism
To reconstruct barrows and their original environments in nature reserves requires
a detailed vegetation history of the barrow landscape. The outcomes of this thesis
research will provide the owners of these areas with information that they may
use to include the barrows in their management and development of the nature
reserve areas (Chapter 14).
36
ancestral heaths
Part Two
Methodology
The methodology of palynological research can be rather complicated and requires
some exposition before palynological results can be interpreted appropriately. The
technique of sampling a barrow and its surroundings, and the chemical analysis
of the soil samples, is described in Chapter 4. Vegetation reconstruction of the
barrow’s locale does not follow as a matter of course from the soil samples taken
from those barrows. The theory underpinning the palynological research of
soil profiles is discussed in Chapter 5. The expression of palynological data in
percentages is common to palynology, enabling comparison of different sites and
time series with one another. These percentages are fractions of an arbitrarily
chosen pollen sum. Which pollen sum will be used in this research and the theory
behind this choice is explained in Chapter 6. One of the main research questions
concerns the size of the open place a barrow was built in. Chapter 7 discusses
three methods that can be used to determine the extent of the open area around
a barrow.
Chapter 4
Sampling and treatment of soil
samples
4.1 The sampling of barrows
As has been shortly explained in Chapter 2, pollen analysis has been proven to be a
good method for reconstructing past barrow landscapes. Pollen grains precipitate
onto the surface every year and are more or less evenly distributed in the top
soil. Pollen disappears due to corrosion and outwash, but normally there is an
equilibrium between the precipitation and disappearance of pollen. Therefore,
the pollen grains in de topsoil represent the surrounding present-day regional and
local vegetation. After a barrow was built, the surface underneath the barrow with
the pollen from the period the barrow was built in, including the previous years’
precipitation was sealed from the air (see figure 4.1). New pollen was prevented
from precipitating onto the old surface and the corrosion and outwash of the pollen
under the barrow was reduced. Analyses of the pollen grains in the old surface
underneath a barrow provide information about the vegetation of the barrow’s
locale before the barrow was built. This principle can been used to reconstruct the
landscape around barrows. In the following paragraphs a description will be given
of the methodology of the barrow sampling. A more detailed discussion about
the preservation of pollen grains in the soil underneath and in barrows is given in
Chapter 5.
4.1.1 The sampling of the old surface
The old surface underneath a barrow, i.e. the surface people lived on at the time
the barrow was built and consequently the surface the pollen grains precipitated
on in that period, is often still recognisable as a darker greyish layer in the soil
pollen rain
sod cutting
old surface
Figure 4.1. A schematic
illustration of pollen
precipitation and how pollen
grains are preserved in the old
surface underneath a barrow
and in its sods.
sampling and treatment of soil samples
39
b
a
profile. Sampling of the old surface can be accomplished by collecting about 10
cm3 of soil by cutting a piece of soil out of a clean section of the barrow of about
1 cm high, 5 cm broad and 2 cm deep. Care must be taken to sample from the old
surface itself and not from the building material above.
4.1.2 The sampling of sods
A barrow is usually constructed of sods (see figure 4.1). Strips of sod of an average
width of 10-25 cm were taken from the upper part of the soil and placed upside
down when building the barrow. The sod-structure of the barrow is in some cases
still visible in a barrow (see figure 4.2). The pollen grains in the old surface of
the sods represent the vegetation that was present at the sod location at the time
just before they were taken. It is tempting to assume that these sods were taken in
the close surroundings of the location where the barrow was built, but comparing
the sods’ pollen spectra with the old surface’s spectra should substantiate such
assumptions. Sampling of the sods is possible when they are clearly recognisable
in the soil section and should be carried out in the same way as the sampling of
the old surface.
4.1.3 The sampling of the soil profile underneath barrows
A new approach in the palynological research of barrows was applied in this
investigation: sampling the soil profile underneath the barrow. About 10 cm3
(height × width × depth ≈ 1×5×2 cm) of soil was collected every centimetre
downwards in the soil profile as exposed in a clean section, containing at least the
entire A (the old surface), B and as much of the C horizon as possible (see figure
4.3). The series of samples was used to make a pollen diagram representing the
vegetation development in the period before the barrow was built. The reliability
and value of pollen diagrams from mineral soils underneath a barrow will be
discussed in Chapter 5.
40
ancestral heaths
Figure 4.2. Two examples
of a sod-built barrow with
visible sods. The barrow in
figure 4.2a is barrow 7 at
Oss-Zevenbergen (see chapter
12.1.1). Photograph by Q.
Bourgeois. The barrow shown
in photograph 4.2b is barrow
2 at the Echoput (see chapter
8.1). Figure by Q. Bourgeois.
removed soil
for safety
removed soil
for samples
Figure 4.3. The sampling
of a soil profile of mound
1 at the Echoput. The top
10 cm is removed to allow
for clean pollen sampling.
Figure by J.W. de Kort/M.
Doorenbosch.
4.1.4 The sampling of ditch fills
Sampling of the old surface underneath a barrow is not always possible as is the
case when dealing with an urnfield. In urnfields the cremated body was buried in
an urn under a much smaller barrow, usually with a diameter of 4-6 m. A ditch was
dug surrounding the urn and the soil material that came from the ditch was put
on top of the urn, creating a small barrow. Most of the barrows in urnfields have
disappeared, but the ditch is often still recognisable as a darker discolouration in
the soil. In this case, given that the old surface is gone, the best option for pollen
analysis is to sample the ditch fill.
A similar case is presented by larger barrows levelled in historical times, where
circular structures such as ditches, may be the only features left. However, what
can be deduced from the pollen spectrum of a ditch fill? This is highly dependent
on what happened to the ditch after it was dug. When was the ditch filled and
how deep was it? If the ditch was open, pollen could precipitate on the bottom of
the ditch. When the ditch was filled, i.e. when the bottom of the ditch was buried,
the latest pollen precipitation was archived. When the ditch was filled slowly, new
pollen could infiltrate again and reach the bottom of the fill easier than when the
ditch was filled fast. In addition, the material that filled the ditch contained both
older and younger pollen. In all cases the pollen grains at the bottom of the ditch
would probably provide the most reliable information about the period that is
closest to the period the urn was buried or the barrow was built.
sampling and treatment of soil samples
41
42
ancestral heaths
20
40
60
80 100
P
NA
20 40
s
nu
Al
60
20
s
lu
ry
Co
40
1
1
1
5
20
us la
m tu
Ul Be
5 10 5 10 1
s
s
s inu
cu
s
gu ax cea nu er
ia
Fa Fr Pi Pi Qu
Til
Trees and shrubs
Figure 4.4. Pollen spectra of the ‘Op de Kiek’ barrow. A spectrum of
the old surface and a spectrum of the surrounding ditch are shown.
Pollen spectra of the two samples are quite similar. Spectra are given
in % based on a tree pollen sum minus Betula pollen. In the total AP
(=arboreal pollen) Betula is included. In the total NAP (= non arboreal
pollen) spores are included, non pollen palynomorphs are excluded.
Different scales have been used, indicated with different colours.
Alphen_Kiek_os_per1
Alphen_Kiek_ditch
AP
Alphen Op de Kiek
Old surface (per1) and ditch
40
60
20
ae
ce
ica
r
E
40
60
80
Grazing indicators
Anthropogenic indicators
Upland herbs Ferns and aq. mosses
1
1
1
5
1
1
1
1
1
5
1
5
519 1124
ae
a)
ul
or
a
e
ifl
et
ar
at
ul
- B um
ol
lg
P
b
e
u
s
v
(A
tu
nc
n
e
la
m um
ae
lle
ea e ri s ium
a
isi race alia eae tago i sa rac cea pte od dium gnu n s l po
p
m
e
i
a
c
a
e
c
e
o
a
e
r
l
y
n
a a
l
r
p b y
l
t
c
h
te t
e
Ar As Ce Po Pl Su Cy Fa Dr Po Pt Sp Po
To
587 1224
Heath
The reliability of the pollen grains in the ditch fills representing the ‘urn/
barrow period’ has been confirmed by several investigations. Bakels for example
compared the pollen composition of ditch fills with control samples, taken from
the undisturbed subsoil next to the ditches and taken from the soil on top of
the ditches (Bakels 1975). The control samples from the undisturbed subsoil did
not contain any pollen, making it unlikely that older pollen that was already
present in the subsoil (the material the ditch was filled with) influenced the pollen
composition of the ditch fill. The control samples from the soil above the ditch
showed a different and younger pollen composition than the ditch fill, indicating
that infiltration of younger pollen from above is negligible. In addition, 14C
dates from the ditches were in agreement with the age indicated by the pollen
composition of the ditch fill.
Another example is given by Casparie and Groenman-van Waateringe (1980).
This example concerns the ditch around barrows from the Middle Bronze Age
period (“Alphen Op De Kiek”). Samples were taken from both the old surface and
the ditch fill. Pollen spectra of the two samples are quite similar (see figure 4.4),
indicating that they indeed represent the same period, the period the barrow was
built.
4.1.5 The sampling of posthole fills
Another rather new approach was applied by taking samples from the filling of
postholes found in the neighbourhood of barrows. In figure 4.5 a hypothesis is
described to explain the pollen spectrum derived from such a posthole fill and how
it would be interpreted. Before the post was placed a surface was present where
pollen grains could precipitate on. Pollen could infiltrate into the soil and pollen
stratification, as has been described in section 4.1.3 and Chapter 5, could be the
result. When the post was going to be placed a hole was dug into this soil. The soil
material coming out of the hole contains the pollen grains that were previously
present in these stratigraphic layers, now mixed up, representing different times
of periods before the hole was dug. The post was placed into the soil and it is very
likely that the remaining hole was filled with the soil material that came out of
the hole in the first place. This soil material contains a mixture of pollen grains.
The soil next to the post continued to develop, with new pollen precipitating on
and infiltrating into the soil. In time, the length of which is usually unknown,
the post will decay or be pulled out. A post could have decayed due to the attack
of soil fauna and fungi. When a post has decayed the part of the post that was
below surface would have been slowly filled up with mostly material from above.
Sediment from above would probably have filled the spaces that emerged due to
the decay since the soil will most likely collapse a little. Younger pollen could
infiltrate into the soil with this incoming sediment. In addition younger pollen
probably has infiltrated with the micro-organisms that were responsible for the
decay. The postpipe will still be visible as a darker coloration in the soil. The
sediments of which this postpipe consists will most likely now contain a mixture
of pollen that mostly represents the period when the post was subject of decay.
With the post decayed, soil development can now also take place from the surface
downwards at the location the post was placed. New pollen precipitates also on
this location and will be transported downwards. However, when the postpipe
itself is still clearly visible in the soil profile, most likely some distance below the
surface as a darker coloration of the soil, it is probable that the soil development
has not reached this depth yet. Assuming that the transportation of pollen grains
downwards into the soil is correlated to the soil development (as will be discussed
in Chapter 5), it is also likely that newly precipitated pollen grains (i.e. pollen
sampling and treatment of soil samples
43
pollen rain
mixture A & B
mixture A & B
A
A
B
B
A
postpipe
decaying post
mixture A & B pollen infiltration mixture A & B mixture of pollen mixture A & B mixture of pollen
infiltrated during
infiltrated during
decay of post
decay of post
mixture A & B
B
mixture A & B and mixture A & B and infiltrated pollen
infiltrated pollen Bottom of posthole:
mixture A & B and pollen infiltrated closest
to period of pull-out
grains that precipitated on the surface since the post had disappeared) did not
reach this depth yet. A sample that is taken in the centre at the bottom of this
clearly visible postpipe contains then pollen grains that were present during the
decay of the post. The pollen spectrum would then represent a mixture of time
periods, but only (or at least mostly) from the time after the post was placed, when
it was subject to decay, until the post had completely decayed. A possible dating
based on the pollen spectrum would give a terminus post quem date.
When a post had been pulled out of the soil a hole was left behind. It is likely
that the hole collapsed and that sediment from the sides and from above filled
up at least part of the hole. The filling of this hole now contains pollen that is
mixture of pollen that was originally present in the posthole-fill (older than the
pulling out of the post) and younger pollen that precipitated on the soil after
the post was placed. The situation is now comparable to when a ditch was dug
(see section 4.1.4). When the post was immediately backfilled, the latest pollen
precipitation was archived underneath. When the posthole was filled slowly, new
pollen could infiltrate again and rejuvenate the pollen spectrum more easily than
when the posthole was filled fast. When a post has been pulled out, the place
where the post was present can often not be distinguished from the original post
44
ancestral heaths
Figure 4.5. An illustration
that shows the theory of pollen
distribution in postpipes and
posthole fills. A: decaying of
the post. B: pulling out of the
post.
hole in which the post was placed. Pollen grains at the centre of the bottom of the
post hole will probably provide information about the period that is closest to the
period that the post was pulled out. An example of sampling posthole fills and the
interpretation of their pollen spectra will be discussed in section 8.1.
Based on the hypothesis above it is best to take samples from the centre of
the bottom of the postpipe or from the centre of the bottom of the posthole fill,
providing a terminus post quem date for the placing of the post.
4.2 Chemical treatment and analysis of palynological soil
samples
Pollen was extracted by adding potassium hydroxide (KOH) to 1 cm3 of the
sediments to remove humic acids. To every sample one to five Lycopodium tablets
were added as a marker, in case pollen concentrations need to be calculated. Heavy
liquid separation, using a mixture of bromoform (CHBr3) and alcohol with a
specific gravity of 2.0, was performed to separate the inorganic material from the
organic material. Finally the samples were acetolysed with a mixture of sulphuric
acid (H2SO4) and acetic anhydride, to remove the large plant remains. Grains were
identified with the aid of the keys of Beug (2004), Faegri and Iversen (Faegri and
Iversen 1989), Moore et al. (1991), Punt et al. (1976, 1980, 1981, 1984, 1988,
1991, 1995, 2003, 2007, 2009) supplemented by Reille (1992, 1995, 1998),
several lists set up by van Geel (van Hoeve and Hendrikse 1998) and the reference
collection of the Faculty of Archaeology of Leiden University. The spectra were
calculated using a pollen sum of ∑AP–Betula (Van Zeist 1967a). A minimum of
300 arboreal pollen grains (excluding Betula) per sample were counted. For more
information about the pollen sum see section 4.3. Pollen spectra and diagrams
have all been plotted with the Tilia software, version 1.7.16 (Grimm 1992).
sampling and treatment of soil samples
45
Chapter 5
The palynology of mineral soil profiles
5.1 The theory behind the palynology of mineral soils
Pollen grains are very resistant to decay and often well preserved across a range
of circumstances. They can, however, be subject to degradation. Pollen is best
preserved under waterlogged (anaerobic) conditions. In aerobic conditions pollen
grains oxidize, causing thinning of the sporopollenin wall of the grains (Havinga
1964). Besides oxidation, the degradation of pollen by biological activity such as
bacterial attack is probably at issue under aerobic conditions (Havinga 1967, 1984).
In addition, pollen can be mechanically damaged during transport (Holloway
1989). When pollen grains have precipitated on the surface of a mineral soil,
and hence under aerobic circumstances, they will be subject to corrosion and
they will wash away (outwash). However, there will be an equilibrium between
the disappearance of pollen grains due to corrosion or outwash and the supply of
pollen to the surface. In addition pollen grains are incorporated into the faeces of
the soil fauna that is responsible for the decomposition of the litter layer on the soil
(van Mourik 2003; this process of incorporation will be discussed in detail in the
following paragraphs). Faeces provide good conditions for preservation of pollen
grains. As a consequence the top soil will contain an assemblage of pollen grains
that represents the surrounding vegetation. After construction of a barrow the
surface containing this pollen assemblage has been covered and protected from the
air, reducing microbiological activity and thus corrosion of the pollen grains. In
addition, the outwash of pollen has been diminished. Therefore, the construction
of a barrow provides good circumstances for preservation of the pollen grains in
the top soil underneath a barrow that had been precipitated on the surface shortly
before the barrow was built. This allows for reconstructing the vegetation of the
barrow building period by sampling the old surface underneath and the sods from
the barrow as has been described in sections 4.1.1 and 4.1.2.
As explained in section 4.1.3, a pollen sequence can be extracted from the soil
profile underneath a barrow providing a pollen diagram that shows a vegetation
development from the period before the barrow was built. Ideally pollen diagrams
are derived from samples taken from peat or lake sediments. The formation
of peat and lake sediments is well known. Both peat and lake sediments are
formed by accumulation processes. Peat is formed by the accumulation of
partially decayed vegetation matter. Organic materials can accumulate when the
production of biomass is greater than its chemical breakdown. Lake sediments
consist of accumulated organic and inorganic material, forming layers containing
an environmental archive. In both peat and lake sediment pollen was caught in
each layer. There is hardly any vertical movement of material and therefore pollen
from the lower layers represents the oldest vegetation. The anaerobic condition
found in both peat and lake sediment enable good preservation of pollen grains,
in contrast to mineral soils. Mineral soils do, however, often show a pollen
stratigraphy. Several investigations have shown that mineral soil pollen grains
can provide a vegetation history (Havinga 1963, Munaut 1967, Dijkstra and van
the palynology of mineral soil profiles
47
Mourik 1995, van Mourik 2003). Van Zeist (1967) published a mineral soil pollen
diagram with a clear vegetation development that generally corresponded to the
known vegetation history of the Netherlands, which was reconstructed from peat
pollen analysis. The value of mineral soil diagrams has been the subject of much
discussion. This discussion has mainly revolved around two issues: the conservation
of pollen grains in a mineral soil, and the distribution of pollen grains in a mineral
soil. Does pollen show a real stratigraphic organization and can they be used
to reconstruct a vegetation development? Several theories have been suggested
about the processes taking place in a mineral soil that influence the distribution of
pollen in the soil and their possible stratigraphy. It was thought for some time that
a similar process of accumulation like in peat and lakes also took place in mineral
soils (Beijerinck 1933, Benrath and Jonas 1937, Florschütz 1941). However,
Dimbleby (1952, 1957, 1961) and Havinga (1962) concluded that processes of
sand accumulation by drifting or by soil fauna are of minimal significance for
the development of pollen stratigraphy, since the pollen concentration decreased
significantly with depth. Munaut similarly disagreed with the theory, showing
in his thesis that most Pleistocene cover sands do not contain contemporary
pollen from this period and pollen grains from other periods must have infiltrated
into these layers (Munaut 1967, 136-137). There is indeed an influx of organic
material in mineral soils, but it are soil forming processes and not accumulation
processes that cause decomposition and transportation of pollen material deeper
into the soil. Other theories involved the infiltration of pollen into the soil as the
underlying process that causes pollen distribution in the soil.
Mothes, Arnoldt and Redman thought percolating water to be the cause of
pollen grain infiltration into the soil (Mothes et al. 1937). Their experiments
showed a selective penetration of pollen grains, with large pollen grains such as
Pinus being transported much more slowly than smaller pollen grains such as
Quercus. Mothes et al.’s conclusions are discussable, since their laboratory situation
was not adequately representative of natural conditions. According to Munaut
(1967, 138) they ignore the influence of organic material in the soil. He states
that pollen grains are incorporated into aggregates of organic material and very
fine mineral particles, causing pollen grains not being able to move around freely
in between the soil particles. In addition, Munaut showed that the infiltration
speed differed between sites with comparable soil types, which should not be the
case when percolating water had been responsible for this (Munaut 1967, 138139). Firbas et al. and Trautman considered percolating water as cause for selective
infiltration of pollen into the soil as well (Firbas et al. 1939, Trautmann 1952).
However, Munaut found no examples of the expected high concentration of small
pollen grains at the lowest parts of the soil and has mentioned that the differences
in the diagrams these authors based their conclusions on, could very well be the
result of differences in local vegetation (Munaut 1967, 144-145).
Havinga, like Munaut, disagreed with the theories of percolating water being
the main cause of pollen distribution into the soil (Havinga 1962). In his thesis
he explained the distribution of pollen grains into the soil by intense biological
activity during the homogenization phase preceding soil formation, especially
podsolization. Pollen are incorporated in the faeces of burrowing animals such as
earthworms and transported into the soil by these animals. A mixture of older and
recently precipitated pollen grains is the result. During this phase pollen grains
disappear due to corrosion and the pollen assemblage is constantly rejuvenated.
During the podsolization phase the homogenization depth decreases due to
decrease of biological activity and pollen below this homogenization depth was
preserved (see figure 5.1). At the top the process of rejuvenation of the pollen
assemblage continues and a pollen profile with at the bottom older and at the top
48
ancestral heaths
Figure 5.1. The change
of a pollen profile
under the influence of
homogenisation. On the left:
the homogenisation depth
is the same during period
B as during the preceding
period A. On the right: the
homogenisation depth is
less during period B than
during period A. The pollen
composition a represents the
vegetation during the older
period A, while the pollen
composition b represents the
vegetation during the younger
period B. Figure after Havinga
(1962, figure 4).
younger pollen assemblages evolves. This means that during the homogenization
phase and cases of incipient soil formation a homogenous pollen assemblage is
present, showing a similar vegetation pattern in the top as well as deeper in the
soil.
Havinga also discussed selective corrosion of pollen grains in mineral soils. Selective
corrosion could be responsible for changes in a pollen profile, mistakenly interpreted
as changes in vegetation. Selective corrosion would more easily take place in sandy
soils than in peat. Based on differences between pollen diagrams of a mineral soil
and a peat bog Havinga concluded that Quercus pollen is largely destroyed in sand
under dry conditions (Havinga 1962, 70-76), but these differences could also
have been the result of local vegetation differences, caused by edaphic differences
between the soil types (Munaut 1967, 145). Havinga showed that pollen grains
that have been oxidized are more easily destroyed by subsequent microbial attack
(Havinga 1964). Later on, Havinga tested selective oxidation in a laboratory
situation, showing a relation between corrosion by oxidation and the amount of
sporopollenin in a pollen grain (Havinga 1967, 1984). This implies that some
pollen grains are more susceptible to oxidation hence corrosion than other pollen
grains, causing selective corrosion. Havinga emphasises that his investigations
were not carried out under perfectly natural conditions. However, differences
in susceptibility for corrosion should be accounted for when interpreting pollen
spectra from mineral soils.
Munaut (1967) agreed with Havinga that the depth of infiltration of pollen grains
into the soil is related to the depth of homogenization by biological activity.
However, he found no homogenous pollen profiles as described by the theory of
Havinga, not even in little developed soils. He also found sharp transitions from
one to another pollen association (Munaut 1967, 141). The research of Munaut
also showed that in less developed soils, those with high biological activity, the
infiltration speed of pollen is higher and the disappearance of older pollen spectra
by microbial attacks is more pronounced. Despite this, Munaut concluded that
biological activity is not the only driving mechanism behind the pollen distribution.
He assumed that the most likely explanation is a combination of both percolating
water and biological activity being responsible for the distribution of pollen in
the soil (Munaut 1967, 141-142), as was suggested by Erdtman (1943). Munaut
concluded that percolating water could be primarily responsible for the depth
of pollen distribution, but that the biological activity is probably responsible for
the activation, delay and stop of this process. Pollen grains are incorporated into
organic aggregates by soil fauna and thereby fixed at a certain level in the soil.
When a pollen grain is freed from its organic aggregate by microbial attack it can
be transported deeper into the soil by percolating water and decomposed or reincorporated again. According to Munaut this explains the higher infiltration speed
and shorter vegetation history in less developed soils, where humic complexes are
less stable and easier to decompose by microbial attack. However, Guillet states
the palynology of mineral soil profiles
49
that soil infiltrating water could not be responsible for pollen transport because
pollen grains have hydrophobic properties and their mean grain size does not allow
vertical transport in single grain conditions through soil pores (Guillet 1970).
Van Mourik (1985, 1986) continued the discussion about pollen infiltration and
conservation in mineral soils (van Mourik 1985, 1986). He studied pollen and
spores micromorphologically in thin sections from several mineral soils. Like
Havinga, he concluded that the distribution of pollen in various mineral soils
is directly correlated with the distribution of soil fauna activity. Pollen grains
incorporated into faunal excrements, were protected from decay. Van Mourik
did not find free pollen grains in the pores of the soil, as would be expected
if transport by percolating water, as suggested by Munaut, had taken place. In
addition van Mourik differentiated syn-sedimentary and post-sedimentary pollen;
syn-sedimentary pollen being present in the sediment when deposited and postsedimentary pollen being that which is brought into the sediment by soil fauna
during soil formation, both being present in excrement. Syn-sedimentary pollen
would be present in a constant concentration throughout the sediment and give
information about the vegetation present at the time of sediment deposition,
while post-sedimentary pollen would decrease in concentration with depth and
give information about the vegetation present at the time of soil formation. This
could be much later then the time of deposition. They can be differentiated from
each other because excrement containing syn-sedimentary pollen is randomly
distributed in the matrix, while the excrement containing post-sedimentary pollen
is mainly concentrated in burrow channels. A good interpretation of mineral
soil pollen diagrams makes the distinction between syn- and post-sedimentary
pollen.
The topic of distribution and conservation of pollen grains in a mineral soil has
also been the focus of recent studies. Davidson concluded that the activity of soil
fauna, mainly earthworms, is an important factor in the redistribution of pollen
(Davidson et al. 1999). Pollen that has been precipitated on the soil surface is
consumed, digested and excreted by soil fauna. He claimed that the depth of
incorporation of the pollen grains is dependent on the depth of the soil fauna
activity. However, he also stated that the result is a mix-age pollen assemblage
and that age-stratification of pollen assemblages is only possible in the top surface
organic horizon of a podzol or soils with accumulating organic horizons like peaty
soils. Van Mourik showed that the vegetation development from heathland to
closed (planted pine) forest was recorded in pollen assemblages in undisturbed
acid soil profiles that had developed underneath the forest at several locations
in the Netherlands (Dijkstra and van Mourik 1995, Dijkstra and van Mourik
1996, van Mourik 2003). These soils could develop after plantation of a pine
forest on a former heath area. Pollen zonation was already visible in the organic
top layer (F, H and A horizons, see figure 5.2) of the soil. The pollen content of
terrestrial humic soils like these is post-sedimentary (van Mourik 2001), meaning
that the pollen assemblage represents the vegetation that was present during the
development of these soils, since they infiltrated into the soil during this process.
Micromorphological observations of thin sections from these soils revealed
that when pollen grains precipitated on the surface, they were incorporated in
excrements from soil fauna in the upper part of the F horizon. The pollen grains
were then released again in the lower part of the F horizon and the H and A
horizons. Then they were reincorporated again in small excrements of soil fauna.
These excrements, which are very stable and are only slowly decomposed by
Soil fauna burrows channels while moving through the soil leaving their excrements behind in these
channels.
50
ancestral heaths
Figure 5.2. Pollen diagram
from a micropodzol that
had developed underneath a
Larix forest. Incipient pollen
zonation is visible in the top
organic layer (F, H, AE and
AB horizons). Figure after van
Mourik (2001).
fungal attack, preserved good conservation conditions for pollen grains. Pollen
grains that were not reincorporated were destroyed by microbial consumption
(van Mourik 2003). When soil-mixing animals were absent a pollen stratification
representing the vegetation development was present. In layers where soil-mixing
animals were present pollen were easier oxidized and the pollen distribution was
more even throughout the soil layer (Dijkstra and van Mourik 1996). Pollen grains
are transported deeper into the soil, into the B-horizon or even the C-horizon by
the activity of soil fauna like earthworms. Since they show a retrogressive activity
during the soil formation the oldest pollen assemblages will be found at the lowest
parts (van Mourik 1999).
How can this principle of mineral soil pollen palynology be used in the palynological
research of barrows? During pedogenesis pollen grains are transported deeper into
the soil. However, this process is interrupted when the soil was covered by a burial
mound and the soil was well preserved until excavation. The soil profile that had
developed before the barrow was built is often still recognizable. This indicates
that after the construction of the barrow the soil profile had not or hardly been
disturbed. This furthermore indicates that the pollen profile that was present in
the soil before the barrow was built was also preserved. When a barrow was built
the soil was sealed away from outside influences. As a consequence, biological
activity decreased, creating a more stable environment for pollen conservation
and preventing homogenization of the soil that would consequently disturb the
soil profile. In addition, the barrow also prevented pollen from precipitating on
the soil. Podzols found underneath barrows have often developed in the top of
Pleistocene cover sands. These sediments were originally free from pollen (Koster
1978), so the pollen content of the soil underneath barrows is mainly postsedimentary (van Mourik 2001). This means that pollen infiltrated into the soil
during pedogenesis. Infiltration of younger pollen grains into the soil can alter
the composition of the pollen assemblage and as a consequence the interpretation
on which the vegetation reconstruction is based. It is likely that there is a mix of
pollen grains of different ages in each zone, but it is also likely that the majority of
the youngest pollen grains will be in the top the soil and the deeper into the soil
the higher the average age of the pollen grains will be.
In conclusion, based on the results of previous investigations described above it
seems to be possible to read a rough vegetation history from a mineral soil pollen
diagram from underneath a barrow, however, with the usual caveats.
the palynology of mineral soil profiles
51
5.2 The time represented in a mineral soil pollen diagram
What stretch of time is represented in pollen diagrams derived from mineral soils is
important for dating and linking a vegetation development to a certain period. The
duration of the downward movement of pollen in the soil indicates the period that
is represented in a mineral soil pollen diagram. For the most part, it is not possible
to date the soil using dating techniques like 14C or OSL (optically stimulated
luminescence). It has been suggested, however, that the infiltration speed of
pollen grains into the soil can be generalised. Dimbleby (1985) suggested that the
average rate of downward movement of pollen in a mineral soil is about 10 cm
in 300 years. Although he stated that this rate could vary according to prevailing
pedological conditions, this average rate is still often used in the interpretations of
mineral soil pollen diagrams (Groenman- van Waateringe 1986, de Kort 2002).
As has been explained in the previous paragraph, the downward movement is
dependent on the activity of the soil fauna, which is active during pedogenesis.
The speed at which soil fauna moves through the soil distributing the pollen grains
incorporated in their excrements is highly dependent on several factors such as the
hydrology, the acidity, and compaction of the soil. As a consequence it is highly
unlikely that the speed at which soil fauna distributes pollen grains into the soil
is similar across different locations. The 300 years in 10 cm Dimbleby found
may very well have been true in that particular situation, this cannot however be
applied to every mineral soil. A few examples now follow of cases that contradict
10 cm/300 years downward rate of pollen in mineral soils.
Example 1: The Laarder Wasmeren area
That the formation of a podzol is a complex process and can differentiate even
in a small area can be seen in the Laarder Wasmeren (LWM) area. The Laarder
Wasmeren area is a nature reserve in the Netherlands (see figure 5.3). The area
had been used to discharge waste water in the 20th century, polluting the area
with heavy metals and toxic organic compounds. In 2003, remediation of the area
Figure 5.3. Location of the
nature reserve area Laarder
Wasmeren, Weerterbergen
and Gieten.
52
ancestral heaths
started by ending the discharge and thus draining the area, and subsequent removal
of polluted sludge and soil. Underneath this soil a Holocene drift sand landscape
was discovered. The complex stratigraphy and genesis of this landscape, with four
drift sand phases, two lacustrine phases and five phases of soil formation, was
studied in detail by Sevink et al. (in press) who investigated several representative
soil profiles in the area. The soil profiles showed three or four podzols on top
of each other separated by layers of drift sand. Every time, during a period with
stable conditions, soil formation led to the development of a podzol, which was
buried under drift sand during the next phase of landscape instability. During the
stable phase, the local vegetation caused a constant precipitation of pollen grains
on the soil surface. Over time these pollen grains were transported deeper into the
soil by soil fauna as has been explained above. Due to unstable conditions that
probably resulted from land use impacts, vegetation would become scarce and
under the influence of wind the topsoil of bare surfaces was blown away. When
the surface of the LWM area was being covered by sand, pedogenesis and pollen
distribution in the soil stopped. When circumstances were stable again vegetation
could establish itself on the newly deposited sand and pedogenesis and pollen
transportation could take place again. The pollen spectra that are recorded in
a buried soil thus represent the vegetation history of the stable period until the
surface was being covered. Likewise, pollen spectra from the soil underneath a
barrow represent the vegetation history of the landscape in the period before the
barrow was built. The duration of this period is dependent on the time the soil
had to develop. The buried soils in the LWM area, unlike the soil underneath
barrows, could be dated. This was accomplished by taking OSL samples. The
various phases that formed this landscape could be dated providing information
about the length of the period that is represented by the pollen record in which a
certain vegetation development has taken place. In addition all major soil profiles
have been sampled for pollen analysis. Monoliths were taken from the profiles
and from these monoliths every second centimetre a sample of 1 cm was taken
for analysis. For an exact overview of the site, sample locations and methods of
preparation see Sevink et al. (in press). Prepared slides were provided by van Geel
to the author of this thesis for pollen analysis. For this research two profiles have
been selected to analyse. Profile II and Profile V consist of respectively four and
three podzol soils on top of each other.
Based on the OSL dates a reconstruction of soil formation and drift sand phases
in time could be made. For a detailed discussion see Sevink et al. (in press). A
summary of these results is shown in figure 5.4. Profile II consists of four podzols.
The first phase of soil formation (S1) has taken place in Pleistocene cover sand,
deposited around 11500 years BP. A drift sand phase took place broadly between
6500-8500 years BP, which means that the development of S1 could have taken
3000-5000 years. In the drift sand layer D1 a second soil (S2) could develop. This
soil was covered by a new drift sand layer approximately around 5800-6400 years
BP, after which the development of S2 stopped. This indicates that the time span
S2 represents has a length of approximately 100-2400 years. According to Sevink
et al. (in press) the soil phase S2 was probably rather short (a few hundred years),
based on the poor development of the podzol.
The second drift sand period started around 5800-6400 years BP. In this
sand layer, S3 developed until it was covered by a third sand layer (D3). D3
was deposited between approximately 4800-5300 years ago. This leaves 500-1600
years for soil phase S3. A fourth podzol (S4) could develop in D3 until it was
covered by another, more regional phase of aeolian activity (D4), which dates
from the Late Middle Ages or even more recent. However, D4 is missing in LWM
the palynology of mineral soil profiles
53
LWM II
340
OSL
D-3 S-4
Age (ka)
E
Bh1
330
LWM V
D-2 S-3
310
Ahb1
5,8 ± 0,3
320
BC1
Eb1
5,8 ± 0,3
310
BC2
300
Bhb1
Ahb2
6,5 ± 0,4
Eb2
270
CS
S-1
Ahb3
260
8,8 ± 0,4
Age (ka)
5,3 ± 0,2
D-2
S-3
280 CS S-1/2
270
Ahb1
Eb1
6,4 ± 0,3
Ahb2
Eb2
Bhb2
Eb3
260
250
Bh1b3
Bh2b3
250
240
BCb3
240
230
OSL
Bh2
290
280 D-1 S-2
E
Bh1
330
300
290
S-2
4,8 ± 0,2
BC
320
340 D-3
230
II and hence information about the exact time span of S4 in LWM II is lacking.
Profile IV consists of 3 podzols. In this profile S1 and S2 have merged together,
representing a period of about 3100-7400 years. The results of the pollen analysis
are shown in figure 5.5a and 5b. The vegetation development shown by the pollen
diagram will be discussed in detail in section 10.2. Since both profiles are situated
close together and both represent the same soil development phases they would be
expected to be identical. The vegetation development shown by the pollen diagrams
derived from the two profiles is indeed similar. However, the time represented per
centimetre in each profile is not alike (see figure 5.6 and table 5.1). S3 in profile
II and V represent the same period of time (500-1600 years), but the thickness of
S3 in profile II is 27.5 cm, while only 14 cm in profile V. The Dimbleby factor of
30 years per centimetre could be applicable to profile II (this would indicate that
soil phase S3 would have taken around 825 years), but not to profile V. When
applying the Dimbleby-factor to profile V, the middle podzol would have been
estimated to represent about 400 years, while according to the reconstruction by
Sevink et al. this podzol represents about 500-1200 years (Sevink et al. in press.).
Koster (2005) has argued that the rate of pedogenesis in drift sands is highly
dependent on the origin of the drift sand. Drift sand can consist of former A and
E horizon material (like S2), in which a new podzol can form relatively fast. When
the deposited sand originally was C material, development of a podzol is a much
slower process (like S3 and S4).
Profile LWM,
soil phases
Estimated time per soil Thickness of deposit
fase based on OSL (yr) (cm)
Figure 5.4. Cross sections of
the Laarder Wasmeren II and
V profiles with the according
soil formation and drift sand
periods. The location of the
samples for OSL dating
have been indicated with the
corresponding OSL dates (see
also figure 5.6). Figure after
Sevink et al., in press., figure
5 and 7.
Estimated year/cm
LWM-II-4
S4 in D3
?
25-0 cm
LWM-II-3
S3 in D2
500-1600
52,5-25 cm
18-58
LWM-II-2
S2 in D1
100-2400
66-52,5 cm
7.4-177
S1 in coversand
3000-5000
x-66 cm
LWM-V-4
S4 in D3
?
43-0 cm
LWM-V-2
S3 in D2
500-1600
57-43 cm
LWM-V-1/2
S1/2 in coversand/D1
3100-8000
x-57 cm
LWM-II-1
54
ancestral heaths
36-114
Table 5.1. The estimated
time that is represented per
centimetre in every soil phase
of Laarder Wasmeren II and V.
the palynology of mineral soil profiles
55
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
h
pt
De
L
)
ol
ith
m
(c
y lu
s
Ah
20 40 60
s s
g u nu
Fa P i
20 40 60 80 100 1
r
Co
d
an
r
he
bs
20 5
5
20 1
20
1268
100 200
20 5
20
100 200 300 400 500 600
rs
to
ca
di rs
s
n
i
se
ic a to
ts os
s
en dic
rb lan m
g
e
n
d
h
po i
cp n
ro ing
nd a ti ns a ae
la
th az
u r
g
Up Aq Fe
Al
An Gr
100 200 300 400 500 5
es
s
tre
c u ix
us
er tula a th
er
a
l
i
h
m
l
u
l
t
e
i
Q
Sa T U
O B
He
Trees and shrubs
Figure 5.5a-b (this page and next page). Simplified pollen diagrams of the
Laarder Wasmeren. Spectra are given in % based on a tree pollen sum minus
Betula pollen. In the total AP (=arboreal pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are included, non pollen palynomorphs are
excluded. Different scales have been used, indicated with different colours.
Bh
BC
us
Eh
n
Al
20 40 60
P
NA
20 40 60 80 100
y
og
AP
Figure 5.5a: LWM II.
8690 ± 430
6040 ± 330
5580 ± 290
5410 ± 320
4710 ± 250
L
OS
s
te
da
ar
(ye
P)
sB
Laarder Wasmeren
LWM II, simplified diagram
i
ng
Fu
20
740
90
173
120
68
130
82
450
310
356 685
317 750
209 647
328 598
314 591
409 725
350 634
632 1057
406 612
585 799
418 529
918 1093
797 923
701 823
475 622
437 876
404 2297
339 769
385 933
303 759
302 740
343
S1
S2
S3
a)
ul
et
-B m
e
P
l
su
(A
ab
m le n
in
m
su pol
er
n
l
t
e
P de
ll
ta
Po To
NP I n
Soil phase
341 1951
374 1728
359 1171
323 1115
313 995
300 1162
S4
308 1105
319 1248
308 1856
326 4566
366 925
340 860
390 1282
437 1826
455 1507
372 916
380 880
56
ancestral heaths
e
at
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
h
pt
De
0
P
NA
Bh
20 40 60 80 100
Eh
y
og
ol
th AP
Li
)
m
(c
Figure 5.5b: LWM V.
5790 ± 300
4760 ± 220
O
d
SL
ar
ye
s(
P)
sB
Laarder Wasmeren
LWM V, simplified diagram
s
BC
r
Co
Ah
20 40 60 80
nu
Al
s
20
s s
g u nu
F a Pi
20 40 60 80 5
ylu
d
an
bs
ru
sh
20 1
5
5
5
100 200 300
s
ee
s
s r tr la h
cu x
er
li ilia lmu the etu eat
u
a
Q
S T U O B H
Trees and shrubs
t
An
20
rs
to
ica
nd
i
ic
en
og
p
o
hr
40
100 200
g
in
az
Gr
rs
to
ca
di
in
20 40 5
5
20
es
s ss
s
nt o
rb
la d m
e
p
h
n
d
ti c a
i
lan
ua rns gae n g
Fu
Aq Fe Al
Up
20
334
941
230
833
704
435
533
722
452
508
682
788
580
737
648
907
715
652
609
509
475
630
441
380
384
426
340
263
356
314
342
177
315
337
202
259
274
412
337
311
391
318
306
520
364
130
S1/2
S3
)
la
tu
Be m
e APl
u
s
(
ab
in um llen
s
o
rm
te llen al p
e
t
d Po
To Soil phase
In
315 1186
313 1182
349 854
43 127
143 566
154 531
300 825
69 174
331 643
153 250
311 1153
S4
36
91
120 280
cm
above
NAP
340
II
OSL
OSL
4.8
320
D-3
5.8
5.8
D-2
300
6.5
280
5.3
6.4
D-1
8.8
CS
260
Figure 5.6. Cross sections of
LWM II and V with profiles,
phases and OSL datings.
Figure after Sevink et al., in
press., figure 2A.
V
240
220
200
Example 2: Gieten
A 70 year old forest soil in the forestry of Gieten (the Netherlands, see figure
5.3) was investigated by van Mourik (van Mourik 2003). The age of the soil
is known because a former heath area (originally formed on Pleistocene cover
sand) which had been used for sod taking, had been deeply ploughed to prepare
‘fresh’ parent material after which the area had been reforested around 1930. After
plantation of Larix and Fagus trees a forest soil started to develop. This forest soil
is described by van Mourik as ‘micropodzol’ with well-developed humus forms
(mormoders). At the time of the investigation the soil formation had reached a
depth of 10 centimetres. During the formation of this soil pollen grains had been
distributed in this soil by the soil fauna by processes of incorporation, release
and reincorporation into faeces as has been described in the previous paragraph.
The soil profile was palynologically investigated and showed the vegetation
development since the reforestation. With an age of 70 years and a decimetre in
depth it is implied that every cm of soil represents an average of 7 years.
Example 3: Weerterbergen
In another study van Mourik et al. (2010, 2012a) investigated a polycyclic
Holocene soil-drift sand sequence near Weerterbergen (the Netherlands, see
figure 5.3). The investigated profile shows a sequence of four phases with drift
sand deposits in which podzols had developed. Two nearly identical profiles are
involved in this research. In 2002 a profile was sampled for OSL dating. The OSL
ages were used to compare the different soil phases and the time they represent
(see table 5.2). The youngest soil formation phase shows an average of 18 yr/cm,
while the next two phases respectively show 16.7, 75 and 266 yr/cm. Once again,
these data differ from the Dimbleby-factor as well.
Table 5.2. The OSL dates of
the Weerterbergen profile
and the according estimated
time that is represented per
centimetre in every soil phase
(based on van Mourik et al.
2010, 2012a and van Mourik
pers.comm., September 2013)
Profile Weerterbergen,
soil formation phases
Estimated time per soil
formation phase based
on OSL
Thickness of soil phase
(cm)
Estimated year/cm
based on OSL
4
90
5
18
3
250
15
16.7
2
3370
45
75
1
5320
20
266
the palynology of mineral soil profiles
57
Conclusions
These examples show that the ‘Dimbleby-factor’ of 1 cm/30 years cannot be
used as standard. The thickness of a soil and soil pedogenesis is probably highly
dependent on local circumstances. These circumstances are variable through time
and place, even very locally. Hence, the vegetation developments from soil profiles
underneath barrows below cannot be placed in time without additional dating. To
estimate the age of a soil underneath a barrow dating techniques are necessary.
The best technique for dating phases of soil formation is probably OSL (van
Mourik et al. 2010, van Mourik et al. 2011, Sevink et al. in press). However, in
order to do this at least two podzols should be present on top of each other. This is
usually not the case underneath barrows. Radiocarbon dating of organic soil layers
might also be possible, although precaution should be taken when interpreting
these dates. This is clearly shown by the study of van Mourik described in example
3, where in addition to OSL radiocarbon dates have been determined based on
samples from different fractions (humin and humic acid, see van Mourik et al.
2010) of the soil organic matter taken from the buried A horizons in this profile.
The OSL samples provided ages of the sedimentation and soil formation phases.
The radiocarbon dating however did not correspond with the OSL dating, which
is probably due to the presence of older charcoal particles in organic aggregates,
causing an overestimation of the 14C age. Underestimation of the 14C age is
possible when younger organic particles have infiltrated. Van Mourik concludes
that due to the complexity organic matter of the soil radiocarbon ages of buried
horizons cannot be used to date drift sand and soil formation phases (van Mourik
et al. 2010).
5.3. Absence of pollen grains in barrows
As has been described in section 5.1 the soil underneath barrows (and the soil
the barrow has been constructed of ) often provides good pollen preservation
conditions and consequently contains fossil pollen. Professor Waterbolk (University
of Groningen) for example, who has palynologically investigated a great number
of barrows, has never encountered barrow sediments that did not contain pollen
grains (H.T. Waterbolk pers.comm., August 2011). However, pollen is not present
under/in every barrow. In this study we have encountered the problem of a total
absence of pollen grains even under comparable conditions. In Chapter 8 (Case
studies) the palynological results will be discussed of several barrows that did
contain pollen. One of these case studies concerns two barrows in the region of
Apeldoorn at a location called the Echoput, excavated in 2007. The Echoput
barrows did contain reasonably preserved pollen, sufficient for a vegetation
reconstruction that will be discussed in section 8.1. Close to the Echoput barrows,
about four kilometres to the northeast, three more barrows situated at a location
called the Wieselse Weg (WW) were excavated in 2008 and 2009 (Fontijn and
Louwen in prep.). Given that the WW barrows are situated in the same geographic
region as the Echoput barrows, it was expected that they would contain pollen.
However, in contrast to the Echoput, the WW barrows had little or no pollen.
An explanation for the absence of pollen in the WW barrows could possibly be
found in the differences in soil texture between the two locations. Although in
general soil textures were very similar (the soil at both locations was classified as
In order get an even more exact image of the time represented in a mineral soil pollen diagram, one
should also account for syn- and post-sedimentary pollen (see section 5.1, p.50-51).
58
ancestral heaths
Figure 5.7. The locations
of the barrows that have
been used for grain size
analysis.
an Umbric Podzol (ISRIC-FAO 2006) (Dutch classification: Holtpodzol, gY30
[see Bodemkaart van Nederland], according to soil scientist J. Boerma the soil of
the Echoput barrows was loamier than the soil at the Wieselse Weg. In addition,
the podzol underneath the WW barrows was much harder to recognize. Contrary
to the barrows of the Echoput sods were not recognizable and also the old surface
was hard to detect. Possibly the soil on which the WW barrows were constructed
consisted of a somewhat coarser sediment than the Echoput barrows, favouring
a better aeration of the soil which caused the pollen grains of the WW barrows
to be subject of oxidation (Havinga 1984). Besides oxidation, pollen grains were
more easily outwashed as a consequence of a higher susceptibility of the soil.
To test this hypothesis, soil samples of the Echoput and the WW barrows were
selected and analysed for grain size. In addition, soil samples from a barrow with a
well-preserved soil profile and well-preserved pollen from another region, Barrow
7 from the barrow group of Oss-Zevenbergen (section 12.1), were analysed for
comparison (see figure 5.7). Can the results of sediment observations be used to
determine in advance the utility of conducting pollen analysis? Eight soil samples
from the WW barrows, eight samples from the Echoput barrows and four samples
from the Oss-Zevenbergen barrow were selected and analysed for grain size by
the Sediment Analysis Laboratory of the Free University Amsterdam with a Laser
Particle Sizer Helos KR Sympatec. An overview of the selected samples is given
in table 5.3.
A summary of the results is shown in figure 5.8. This figure shows the
distribution frequency q3 of all samples plotted against particle size. To discuss
the results in detail, percentages per classification of the three sites have been
compared with each other (see table 5.4a-b). Figure 5.8 and table 5.4a show
that there are hardly any differences between the Wieselse Weg and the Echoput
All soil types have been classified according to the World Reference Base (ISRIC-FAO 2006), unless
indicated otherwise.
Bodemkaart van Nederland 1:50.000 toelichting kaartblad 33 west Apeldoorn, p. 27, 67-8.
the palynology of mineral soil profiles
59
barrows. This is also demonstrated in table 5.4b, which shows the statistical
results. No significant differences can be detected between WW and Echoput,
with exception of the Middle Coarse Sand fraction. However, almost all fractions
Nr.
Location
Barrow
Sample location
Sample name
1
Echoput
barrow 1
sod 1
MT 266
2
sod 2
MT 267
3
old surface 1
MT 268
4
old surface 2
MT 269
5
sod 1
VNR 99
6
barrow 2
sod 2
VNR 100
7
old surface 1
A2.1 old surface 2
old surface 2
A2.1 old surface 1
profile west
sample 1
8
9
Wieselse Weg
barrow 1 (p101)
10
sample 5
11
barrow 2 (p201)
profile west
12
sample 1
sample 5
13
barrow 3 (p301)
MT 801
14
MT 802
15
MT 803
16
MT 804
sod 1
VNR 275
18
sod 2
VNR 276
19
sod 3
VNR 277
20
sod 4
VNR 279
Density distribution q3*
17
Oss-Zevenbergen
2.15
2.10
2.05
2.00
1.95
1.90
1.85
1.80
1.75
1.70
1.65
1.60
1.55
1.50
1.45
1.40
1.35
1.30
1.25
1.20
1.15
1.10
1.05
1.00
0.95
0.90
0.85
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
barrow 7
Location
Echoput
Echoput
Echoput
Echoput
Echoput
Echoput
Echoput
Echoput
WW
WW
WW
WW
WW
WW
WW
WW
Oss-Z
Oss-Z
Oss-Z
Oss-Z
0.1
60
0.2
0.4
Barrow
barrow 1
barrow 1
barrow 1
barrow 1
barrow 2
barrow 2
barrow 2
barrow 2
barrow 1
barrow 1
barrow 2
barrow 2
barrow 3
barrow 3
barrow 3
barrow 3
barrow 7
barrow 7
barrow 7
barrow 7
0.6 0.81.0
ancestral heaths
Table 5.3. Overview of samples
that have been analysed for
grain size.
Figure 5.8. Results of the
grain size analyses, showing
the density distribution q3
versus the particle size (μm).
Sample location Sample name
sod 1
MT 266
sod 2
MT 267
old surface 1
MT 268
old surface 2
MT 269
sod 1
VNR 99
sod 2
VNR 100
old surface 1
old surface 2
profile west
sample 1
profile west
sample 5
profile west
sample 1
profile west
sample 5
MT 801
MT 802
MT 803
MT 804
VNR 275
sod 2
VNR 276
sod 3
VNR 277
sod 4
VNR 279
2
4
6
8 10
20
40
particle size / µm
60 80100
200
400 6008001000
2000
4000
% Clay
% Very fine Silt
% Fine Silt
% Coarse Silt
(<8 µm)
(8-16 µm)
(<16-32 µm)
(32-63 µm)
WW
Echoput Oss-Z
WW
Echoput Oss-Z
WW
Echoput Oss-Z
WW
Echoput Oss-Z
7,82
6,69
2,56
2,6
2,67
1,23
4,07
4,4
1,48
8,82
8,37
1,9
6,71
5,1
2,67
2,28
1,82
1,33
3,5
2,81
1,84
7,65
5,73
2,52
5,59
6,72
1,75
1,96
2,71
0,81
3,06
4,26
1,02
6,16
7,32
1,33
6,45
6,83
1,72
2,18
2,69
1,36
3,28
4,14
2,06
6,65
8,01
2,08
8,12
5,11
2,55
1,93
3,36
2,88
6,48
5,58
7,93
4,79
2,47
1,84
3,2
2,73
6,57
5,11
6,57
5,98
2,15
2,42
3,08
3,65
6,48
6,81
1,94
4,05
0,78
1,59
1,03
2,38
1,37
4,43
% Very fine sand
% Fine sand
(63-125 µm)
% Middle coarse sand
(125-250 µm)
% Coarse sand
(250-500 µm)
(500-1000 µm)
WW
Echoput Oss-Z
WW
Echoput Oss-Z
WW
Echoput Oss-Z
WW
Echoput Oss-Z
10,76
8,59
6,73
13,35
8,9
43,4
28,93
23,16
38,86
22,89
33,11
3,83
9,05
6,5
6,95
9,9
8,31
35,67
24,18
25,85
38,35
30,56
36,44
10,56
7,56
7,61
6,05
11,45
10,67
46,32
30,61
28,51
39
31,85
30,32
3,71
8,21
9,22
5,82
12,61
12,77
42,44
30,63
25,13
40,86
28,97
27,73
3,67
8,23
6,22
18,38
7,43
34,18
19,13
18,23
34,74
8,16
5,77
13,86
7,63
33,51
20,4
23,59
36,18
8,42
7,62
14,06
8,79
32,66
20,94
24,57
34,49
1,81
4,84
4,95
5,24
19,05
15,14
51,14
38,17
% Very coarse sand
(1000-2000 µm)
WW
Echoput Oss-Z
0,75
4,12
0
6,16
7,45
0,11
1,77
1,88
0
1,02
3,49
0
0,48
16,97
0,72
15,55
2,03
9,31
17,93
24,16
Table 5.4a. Results of the grain
size analyses in percentages
per grain size.
of Oss-Zevenbergen show significant differences compared to both the Wieselse
Weg and Echoput. The soil underneath the Oss-Zevenbergen barrow 7 consists
mostly of fine sand, while the sediments of the other two sites mainly consist of
middle coarse and coarse sand.
The finer composition of the Oss-Zevenbergen sediment could be part of the
explanation why pollen grains have been well preserved. However, this does not
count for the difference in conservation between the Wieselse Weg and Echoput.
Based on these results it is unlikely that differences in particle size of the sediment
are the main causes for differences in pollen conservation. Another possible
the palynology of mineral soil profiles
61
C
Echo
Oss
VFSi
Echo
Oss
FSi
Echo
Oss
WW
x
***
WW
x
***
WW
x
*
Echo
***
Echo
***
Echo
CSi
Echo
Oss
VFSa
Echo
Oss
FSa
Echo
Oss
WW
x
***
WW
x
x
WW
x
***
***
Echo
x
Echo
Echo
**
***
MCSa
Echo
Oss
CSa
Echo
Oss
VCSa
Echo
Oss
WW
*
***
WW
x
**
WW
x
x
****
Echo
****
Echo
Echo
**
explanation for the relatively good pollen conservation of the Echoput barrows
is the locally wet conditions of the Echoput area (Fontijn 2011a, 30) compared
to the much drier conditions of the Wieselse Weg. The soil at the location where
the Echoput mounds are situated contains some loam which is practically absent
at the Wieselse Weg. Although loam has not been shown by grain size analysis,
loam might have been present somewhat deeper in the subsoil, causing moist
conditions at the Echoput. These moist conditions at the Echoput site could have
provided favourable conditions for pollen conservation, reducing the availability
of oxygen and the (micro)biological activity (Havinga 1962, 1984), where at the
Wieselse Weg the drier conditions favoured the degradation of pollen grains. The
soil in Oss-Zevenbergen is also dry, but much poorer in nutrients, which also
reduces microbial activity.
Conclusions
The purpose of the soil texture measurements described in the previous section
was to determine whether the presence or absence of pollen grains in a soil could
be predicted with these relatively simple and quick measurements. However, based
on the results it is not possible to differentiate a pollen containing sediment from
a non pollen containing sediment only by judging the texture of the sediment.
Further research is recommended.
62
ancestral heaths
Table 5.4b. Results of the
statistic analyses (tested with
unpaired t-test) after grain
size analysis. x means not
statistically different, * means
statistically different p<0,05),
** means statistically
different (p<0,01), ***
means statistically different
(p<0,001), **** means
statistically different
(p<0,0001).
Chapter 6
The pollen sum
Figure 6.1. Young Betula trees
appearing as pioneer trees 6
years after the excavation of
barrows at Oss-Zevenbergen.
Photograph by R. Jansen.
The absolute number of pollen grains found per sample can differ significantly per
sediment. To be able to compare pollen spectra, pollen types are usually expressed
as percentages of a pollen sum. The pollen sum used can be a total pollen sum, so
with all pollen types included, or it can be based on a selection of pollen types.
The pollen sum should be chosen in a way to get the most representative reflection
of the vegetation composition that produced the pollen. To quote Faegri (1966,
136): “Pollen sums must be adapted to the problem they are supposed to elucidate, and
then the basic rule is extremely simple: the pollen sum should contain the pollen taxa
from those plants that are of interest in elucidating the actual question.” For example,
when tree abundance in a forest area is of main interest, the pollen sum should
only include arboreal pollen taxa and exclude herbal taxa, because the variation
in herbal abundance would influence the percentages of the arboreal taxa that are
unrelated to differences in tree abundance. However, in areas with little forest,
herbal vegetation is of much more importance and when interested in the ratio
between arboreal and non-arboreal vegetation herbal taxa should be included in
the pollen sum. It would therefore be expected that in barrow palynology the
pollen sum would include both arboreal and non-arboreal taxa. However, the
pollen sum used in the palynological studies is a tree pollen sum minus Betula
∑AP-B. According to van Zeist this is the most appropriate sum to use. Betula
and herbs are excluded because they grow locally at a barrow site and when they
are included pollen percentages strongly fluctuate between barrow pollen spectra,
even between samples from the same barrow (van Zeist 1967a). Betula is a pioneer
tree (see figure 6.1); it settles and flowers easily in an open space and can therefore,
like herbs, vary significantly at short distances. The consequence of using relative
the pollen sum
63
numbers is that the change in one species affects the percentage of a species that
does not change at all. Therefore, to reflect the regional vegetation best species
that can vary locally should be left out of the pollen sum. Van Zeist tested several
pollen sums for two tumuli originating from the same period and situated in the
same area. It is expected that these spectra look very similar at least concerning the
regional vegetation. He concluded that ∑AP-B was the most suitable pollen sum
indeed. Since then ∑AP-B was commonly used in barrow palynology. However,
this pollen sum has only been tested once and needs to be reconsidered.
6.1 Slabroek
As explained above, to get a reliable image of the regional vegetation, species
that grow locally on a barrow site should be left out of the pollen sum, since
their frequency can differ greatly at short distances. In a barrow pollen spectrum
non-arboreal species are in general species that grew close to the sample site.
Arboreal species however do not necessarily solely reflect regional vegetation.
Van Zeist concluded that besides herbal species, Betula is also a local species and
should therefore be excluded from the pollen sum (van Zeist 1967a). However,
how can regional vegetation be best ascertained? Pollen diagrams derived from
peat are assumed to give a reliable image of regional vegetation. As has also been
explained in section 4.1.3, peat is an accumulation of organic material and in
each layer pollen grains are caught. Peat provides conditions for good preservation
of pollen grains, and since there is no vertical movement of pollen, each layer of
peat reflects the vegetation from the period the pollen was caught. By comparing
barrow pollen spectra with a contemporaneous pollen spectrum from peat from
the same area it should be possible to determine the local vegetation of the barrow
site, which can then be excluded from the pollen sum. The peat spectrum gives
information about the regional vegetation of the barrow landscape. Species in the
Figure 6.2. Location of
Slabroek, the Venloop. OssZevenbergen, Echoput and
Hijken.
64
ancestral heaths
Urnfield Slabroek
0
250
500
1000 m
Sample
location
Venloop
m NAP
63 m
0m
Nistelrode
Urnfield Slabroek
0
250
500
1000 m
Sample location
m NAP
63 m
Venloop
0m
Figure 6.3. Map with the
location of the urnfield at
Slabroek and the Venloop.
The map is based on digital
elevation model of the AHN
(copyright www.ahn.nl).
0
250
500
1000 m
Sample location
barrow pollen spectrum thatm NAP
differ greatly from the peat spectrum probably are
m
part of the local vegetation at the 63
barrow
site. These species can then be excluded
from the barrow pollen sum.
0m
To examine this theory a case study was conducted with data from a nature
reserve area called the Maashorst, situated in the province of North Brabant, the
Netherlands. A prehistoric urnfield at ‘Uden-Slabroekseheide’ (see figure 6.2 and
6.3) was investigated in 1923 by Remouchamps. He discovered 38 barrows from
the Iron Age and Early Roman Period (Remouchamps 1924). Due to plans to
change the area into a nature reserve and to reconstruct the urnfield, the area
was reinvestigated by Archol (the excavation unit of the Faculty of Archaeology,
University of Leiden, The Netherlands) in 2005 (van Wijk and Jansen 2005b).
In 2010 it was decided to excavate the remainder of the urnfield area (Jansen and
Louwen in prep.). During this last excavation several ring ditches were found.
Two of these ring ditches (43 and 12, see figure 12.17 for the exact locations of
the ditches), originally encircling the urnfield barrows, were sampled for pollen
analysis. Their history as a pollen trap is the same as that of a ditch around a
barrow. Pollen samples taken from the bottom of the ditch fill were analysed (as
representing the vegetation of the area at the time of the digging of the ditch,
see section 4.1.4). Samples were prepared as described in section 4.2. The pollen
spectra from the ditch fills are shown in figure 6.4, percentages are based on a tree
pollen sum minus Betula.
About one kilometre from the urnfield at Uden-Slabroekseheide, van Mourik
(2011) analysed a peat core taken from the Venloop (van Mourik et al. 2012b).
The pollen diagram from the Venloop as it has been published by van Mourik is
shown in figure 6.5; percentages are based on a total tree pollen sum. It shows
the development from a wetland to a peat to a deforested agricultural landscape.
The start of the peat accumulation was 14C dated 750-410 cal BC, which is the
start of zone 2 in the diagram and is contemporaneous with the urnfield period
of Slabroek, which can be placed in the Early Iron Age (800-500 cal BC, see table
2.1) (Jansen and Louwen in prep.). Since vegetation in the pollen diagram does
the pollen sum
65
66
ancestral heaths
60
20 40
60
1
20
40
5 10
1
5
20
5
1
20
100
s
s
us
le
in s rcu a
us la
i
ica
m tu
ax n u e
F r Pi Q u
Til Ul Be
Er
Figure 6.4. Pollen spectra of the samples taken from the ditches 12 and
43 of the urnfield at Slabroek. Spectra are given in % based on a tree
pollen sum minus Betula pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (= non arboreal pollen) spores are
included, non pollen palynomorphs are excluded. Different scales have
been used, indicated with different colours.
80 100
s
gu
200
300
1
5
5
1
1
5 10 1
5
1
5
5
1
1
707
327 1349
20 40
Fa
Ditch12_base49-48cm
us
in lus
rp ory
a
C C
407 1081
s
Ditch12_base45-44cm
nu
Al
e
ra
flo
412
P
NA
Anthr. indicators Grazing indicators Upland herbs Ferns and mosses
es
or
sp
e
e
a)
p
n
a
r
ul
e
ty
or ta
fe are
et
li
yp
aifl ola
B
e
t
l
e
u
g
t
l
u
a
P
a
a
um
ub e ig ce
(A n s
sil vu
op
os
e t typ ae l lan
e
ur eae te p ium m um
et
l
a
l
c
e
a
o
e
e
u
c e od n
a
o
s
a
isi c lia c g ae
g n
lp
ex isa ut ra ol
m ra a ra ta ce
m cc sc pe on lyp ha lle ota
te te re te an a
Ru Su Cu Cy M Po Sp Po
Ar As Ce As Pl Po
T
361 741
Heath
Ditch43_base35-32cm
Ditch43_base32-31cm
AP
Trees and shrubs
not seem to vary greatly around this period (see figure 6.4), a more exact dating
of the ditches is not necessary and their pollen spectra can be compared with the
pollen diagram.
The pollen spectra from the two ditch fills and the pollen diagram from the
Venloop peat core provide information about the local and regional vegetation
in the area in the Early Iron Age. By comparing the Venloop pollen diagram with
the ditch pollen spectra it is possible to determine regional and local vegetation
and therefore which pollen sum should be used to best display this in the ditch
pollen spectra. To be able to compare the peat diagram with the ditch pollen
spectra percentages should be based on a similar pollen sum. In table 6.1 average
percentages of zone 2 of the peat diagram (VL) and percentages of both ditch
pollen spectra (43 and 12) are shown based on similar pollen sums and tested
with several pollen sums. The first pollen sum tested is ∑AP-B that, according
to van Zeist, should show similar percentages of tree pollen between both the
ditch spectra as well as the peat diagram. What can be seen is that there are clear
differences in percentages of Quercus and Corylus between the ditch spectra and
the peat spectrum. When Betula is included in the pollen sum the differences
do not disappear. The assumption that Alnus is overrepresented in the Venloop
diagram is easily made, as this is a dominant species in the wetland shown in the
diagram. When Alnus is left out of the pollen sum however, differences in Corylus
and Quercus become greater. Several pollen sums have been tested, shown in table
6.1, but no pollen sum could be found to match the ditch spectra to zone 2 of
the peat diagram. Based on these results it is not possible to come to a conclusion
about which pollen sum is most suitable for barrow pollen analysis and in section
6.3 another case-study is carried out.
Furthermore, the origin of the differences between Venloop and the ditch spectra
is not clear. This could indicate that the ditch spectra and zone 2 of the pollen
diagram are not contemporaneous and that they reflect two periods with different
vegetation composition. However, based on the 14C-dating of the diagram and
Table 6.1. Pollen percentages
of the samples taken from
the ditches 12 and 43 at the
urnfield at Slaboek and the
average percentages of zone
two of the Venloop pollen
diagram.
Pollen
∑AP-Betula
sum
Alnus
Betula
Corylus
Pinus
Quercus
A
B
C
P
Q
Ditch 12
55.8
11.8
25.6
4.4
9.1
49.9
10.5
22.9
4.0
8.1
Ditch 43
52.4
3.0
34.1
5.0
4.2
50.8
3.0
33.1
4.3
4.0
Venloop
66.2
25.0
7.4
1.5
19.1
55.1
20.8
6.1
1.2
15.9
Pollen
% total pollensum
sum
A
B
C
P
Q
%AP
%NAP
A
B
C
P
Q
D12
21.0
4.4
9.6
1.7
3.4
42.1
57.9
45.7
9.7
20.9
3.6
7.4
D43
25.5
1.5
16.6
2.4
2.0
50.2
49.8
46.8
2.7
30.4
4.5
3.7
VL
43.3
16.3
4.8
1.0
12.5
81.7
18.3
44.6
16.8
5.0
1.0
12.9
Pollen
%total AP
%total sum-heath
%AP-Alnus
%AP-B-A
sum
A
B
C
P
Q
A
D12
99.6
21.1
55.7
0.4
16.2
D43
103.3 6.0
67.2
9.8
VL
122.5 46.3
13.6
2.7
B
%AP-B-C
C
P
Q
A
B
126.1 26.7
57.8
0.6
20.6
74.9 15.8 34.3 0.3
8.2
109.9 6.4
71.5
10.5
8.7
79.4 6.4
35.4
195.7 73.9
21.7
4.3
56.5
71.4 27.0 7.9
the pollen sum
C
P
Q
12.2
71.5 10.5 8.7
1.6
67
20.6
68
ancestral heaths
Figure 6.5. Pollen diagram of
the Venloop. Figure after van
Mourik (2012, figure 5).
the minimal change in vegetation around this 14C-date and the rather narrow
archaeological dating of the ditches, this is not very likely.
Another possibility is that the distance between the Venloop and the urnfield
is too great. This would imply that the Venloop diagram and the ditch spectra
both show local vegetation, since they are only one kilometre apart from one
another. This is surprising in regard to the peat diagram, since the assumption
has always been that pollen diagrams derived from peat reflect regional vegetation
composition. This would have implications for reconstructions of regional
vegetation history in the Netherlands, which are mainly based on peat and lake
sediment analyses. To (dis)prove this further research is necessary.
6.2 Contemporaneous barrow pollen spectra
In this paragraph another approach to reconsider the pollen sum ∑AP-Betula will
be discussed. First the data used by van Zeist will be re-inspected in detail. Van Zeist
based his conclusions on the results of pollen analysis of two contemporaneous
barrows at Hijken. Since these two tumuli originate from the same period and
are situated in the same area it is expected that their pollen spectra are similar.
The original complete spectra of the barrow cemetery of Hijken were published
in van Zeist (1955). Pollen percentages based on several pollen sums are shown
in table 6.2. In addition the standard deviation is calculated based on the pollen
percentages from both barrows to show the variation in the data. The lower the
standard deviation, the less variation in pollen percentages is present. Van Zeist
only looked at two species, Alnus and Corylus. His conclusion that least variability
is shown when a tree pollen sum minus Betula is used is true concerning Alnus.
Other arboreal species, including Corylus, however are least variable when a total
pollen sum is used. In addition, percentages of the non-arboreal species Calluna
and Poaceae are least variable with a total pollen sum. This still does not exclude
n thBetula as the best pollen sum to use, since it is very unlikely that Alnus was
growing on the barrow site, being more likely to have grown at the lake side about
five kilometres from the barrows (van Zeist 1955). Differences in the other species
should then be caused by locally varying appearance, meaning that especially
Betula trees were growing on the barrow site. This is very likely, since Betula is
a very common tree in heathland areas. When Betula is included in the arboreal
pollen sum the variability of Alnus indeed increases. The lower variability of other
arboreal species with a total pollen sum can be explained by the high numbers of
Calluna, suppressing the percentages of species present in a lower abundance. The
higher variability of percentages of Calluna and Poaceae with ∑AP-Betula can, as
has been explained by van Zeist, be caused by local differences (van Zeist 1955).
Another case study used to investigate the pollen sum is located at the Echoput
near Apeldoorn, the Netherlands. Two barrows were excavated and samples were
taken from the old surface and some sods of both barrows for pollen analysis. The
barrows both dated to the 4th or 3rd century cal BC, and were probably built at the
same time or one relatively quickly after the other one (Fontijn 2011b, 153). For
a full description of the site, the barrows, sampling for pollen analysis and a more
detailed discussion of the results, see Chapter 8, case-study Echoput. The results
of the pollen analysis are shown in table 6.3. Percentages are based on several
pollen sums. It is expected that both barrows, since they were built in the same
period, show a comparable regional vegetation pattern and even locally they are
not expected to differ greatly since they were built less than 20 m apart from each
The total pollen sum includes all arboreal pollen taxa, all herbal pollen taxa and all spores from ferns
and mosses.
the pollen sum
69
% AP-Betula
Tum 5
Tum 6
Alnus
Betula
Corylus
Quercus
Tilia
Calluna
Poaceae
o.s.
64.5
14.1
22.4
6.8
3.4
39.5
51.8
sod 1
61.8
31.9
16.1
19.0
0.8
53.9
21.7
sod 2
65.9
13.5
23.6
6.6
2.4
107.1
26.4
o.s.
65.2
25.1
20.0
10.4
3.0
50.4
66.0
sod 1
61.3
38.3
20.2
15.2
0.8
98.8
27.4
sod 2
70.4
82.5
18.5
7.9
0.6
50.3
41.7
sod 3
66.4
56.5
22.1
7.5
1.3
65.7
48.3
S.D.
3.1
24.8
2.6
4.8
1.2
26.1
16.1
% total AP
Tum 5
Tum 6
A
B
C
Q
T
Ca
P
o.s.
56.5
12.4
19.6
6.0
3.0
34.6
45.4
sod 1
46.9
24.2
12.2
14.4
0.6
40.9
16.4
sod 2
58.0
11.9
20.8
5.8
2.1
94.4
23.3
o.s.
52.1
20.1
16.0
8.3
2.4
40.3
52.8
sod 1
44.3
27.7
14.6
11.0
0.6
71.4
19.8
sod 2
38.6
45.2
10.1
4.3
0.3
27.6
22.8
sod 3
42.4
36.1
14.1
4.8
0.8
42.0
30.8
S.D.
7.3
12.2
3.8
3.7
1.1
23.8
13.8
% total pollen sum
Tum 5
Tum 6
A
B
C
Q
T
Ca
P
o.s.
8.9
1.9
3.1
0.9
0.5
5.4
7.1
sod 1
26.8
13.8
7.0
8.2
0.3
23.4
9.4
sod 2
16.6
3.4
6.0
1.7
0.6
27.0
6.7
o.s.
11.5
4.4
3.5
1.8
0.5
8.9
11.6
sod 1
20.4
12.7
6.7
5.1
0.3
32.8
9.1
sod 2
22.9
26.9
6.0
2.6
0.2
16.4
13.6
sod 3
18.9
16.1
6.3
2.1
0.4
18.7
13.7
S.D.
6.3
8.9
1.6
2.6
0.1
9.7
2.9
% total pollen sum-Heath
Tum 5
Tum 6
70
A
B
C
Q
T
Ca
P
o.s.
9.4
2.1
3.3
1.0
0.5
5.7
7.5
sod 1
35.0
18.0
9.1
10.7
0.5
30.5
12.3
sod 2
22.8
4.7
8.2
2.3
0.8
37.0
9.1
o.s.
12.6
4.8
3.9
2.0
0.6
9.7
12.7
sod 1
30.3
18.9
10.0
7.5
0.4
48.8
13.5
sod 2
27.4
32.1
7.2
3.1
0.2
19.6
16.2
sod 3
23.3
19.8
7.7
2.6
0.5
23.0
16.9
S.D.
9.2
10.9
2.5
3.6
0.2
15.2
3.4
ancestral heaths
% AP-Alnus
Tum 5
Tum 6
A
B
C
Q
T
Ca
P
o.s.
129.8
28.4
45.2
13.7
6.9
79.6
104.2
sod 1
88.3
45.5
23.0
27.1
1.1
77.0
31.0
sod 2
138.2
28.4
49.6
13.9
5.0
224.8
55.5
o.s.
108.8
41.9
33.4
17.4
5.0
84.1
110.2
sod 1
79.6
49.7
26.2
19.8
1.0
128.3
35.5
sod 2
62.8
73.6
16.5
7.0
0.5
44.9
37.2
sod 3
73.7
62.7
24.5
8.3
1.4
72.9
53.5
S.D.
28.9
16.8
12.2
6.9
2.5
59.6
32.9
% AP-Betula-Alnus
Tum 5
Tum 6
A
B
C
Q
T
Ca
P
o.s.
181.4
39.7
63.1
19.2
9.6
111.3
145.6
sod 1
162.0
83.5
42.1
49.7
2.1
141.4
56.8
sod 2
193.0
39.6
69.2
19.4
7.0
313.8
77.4
o.s.
187.4
72.1
57.5
29.9
8.6
144.8
189.7
sod 1
158.3
99.0
52.1
39.4
2.1
255.2
70.7
sod 2
238.2
279.1
62.5
26.7
2.0
170.3
140.9
sod 3
197.9
168.5
65.8
22.3
3.9
195.8
143.8
S.D.
26.6
85.8
9.3
11.4
3.3
71.4
49.5
% AP-Betula-Corylus
Tum 5
Tum 6
Table 6.2. Pollen percentages
of Hijken.
A
B
C
Q
T
Ca
P
o.s.
83.1
18.2
28.9
8.8
4.4
51.0
66.7
sod 1
73.7
38.0
19.2
22.6
1.0
64.3
25.8
sod 2
86.2
17.7
30.9
8.7
3.1
140.2
34.6
o.s.
81.5
31.4
25.0
13.0
3.8
63.0
82.5
sod 1
76.8
48.0
25.3
19.1
1.0
123.7
34.3
sod 2
86.4
101.2
22.7
9.7
0.7
61.8
51.1
sod 3
85.3
72.6
28.3
9.6
1.7
84.4
61.9
S.D.
4.9
30.6
4.0
5.6
1.5
34.5
20.6
other. Samples originate from the old surface and from several sods of which the
barrows were constructed. Pollen spectra from the sods could possibly show some
dissimilarity in local vegetation, since they could originate from a wider area.
Pollen spectra from the old surface samples however, should be identical, since the
barrows were built very close together.
As can be seen in table 6.3 differences between arboreal pollen spectra are
smallest when a total pollen sum is used. This also counts for Alnus, in contrary
to the pollen spectrum of Alnus in the Hijken barrows (see 6.2). Since pollen
spectra are expressed in percentages, species that occur in large numbers have great
influence on percentages of other species when this species is included in the pollen
sum. In this case, when Calluna is included in the pollen sum, the percentages of
other species automatically decrease, since very high numbers of Calluna pollen
have been found in the Echoput samples. With these large numbers of Calluna
included in the pollen sum the variance of other species automatically decreases.
To avoid this effect Calluna should be left out of the pollen sum and an arboreal
the pollen sum
71
pollen sum should be applied. In addition, when expressed as a percentage based
on a total pollen sum the variation in Calluna is not as obvious as when expressed
as percentage based on a tree pollen sum since it is then bound to a maximum of
100%. Therefore, to reflect local variability in herbal vegetation it is best to use
an arboreal pollen sum. Since there are only a few Betula pollen grains present
in the samples it is not possible to judge the exclusion of this species as has been
suggested by van Zeist (1967).
Pollen
% AP-Betula
sum
Alnus
Corylus
Quercus
Fagus
Calluna
Poaceae
Barrow 1 sod1
65.5
14.5
19.1
0.2
246.8
36.6
Barrow 1 sod3
66.3
17.5
12.9
2.6
135.3
52.1
Barrow 1 sod4
58.6
15.9
22.6
1.6
159.9
61.5
Barrow 1 os1
64.7
15.0
19.3
0.7
157.8
46.7
Barrow 1 os2
57.1
15.0
23.7
2.8
174.1
34.3
Barrow 1 os3
74.5
16.1
8.1
1.3
195.2
62.6
Barrow 1 os4
57.5
11.8
28.1
2.2
125.6
63.9
Barrow 2 sod1
44.9
12.0
39.6
1.3
322.2
47.8
Barrow 2 sod2
58.7
19.6
20.8
0.6
140.1
71.5
Barrow 2 sod3
69.2
15.7
12.6
1.3
190.6
101.6
Barrow 2 os1
66.1
14.8
18.4
0.3
261.6
29.4
Barrow 2 os2
63.3
16.9
18.9
0.9
124.3
80.2
Barrow 2 os3
37.7
22.7
33.6
1.8
140.5
75.0
S.D.
9.9
2.9
8.6
0.8
60.4
20.5
Pollen
% total AP
% total pollen sum
sum
A
C
Q
F
Ca
B1 sod1
65.2
14.4
19.0
0.2
B1 sod3
66.3
17.5
12.9
B1 sod4
58.4
15.9
22.5
B1 os1
64.7
15.0
B1 os2
57.1
B1 os3
A
C
Q
F
Ca
P
245.8 36.5
16.3
3.6
4.8
0.0
61.6
9.1
2.6
135.3 52.1
21.5
5.7
4.2
0.8
43.9
16.9
1.6
159.4 61.3
17.2
4.7
6.6
0.5
47.0
18.1
19.3
0.7
157.8 46.7
20.1
4.7
6.0
0.2
49.1
14.5
15.0
23.7
2.8
174.1 34.3
17.8
4.7
7.4
0.9
54.3
10.7
73.8
16.0
8.0
1.3
193.3 62.0
19.8
4.3
2.1
0.3
52.0
16.7
B1 os4
57.0
11.7
27.8
2.2
124.4 63.3
19.2
3.9
9.4
0.7
41.9
21.3
B2 sod1
44.2
11.8
38.9
1.2
317.1 47.0
9.2
2.5
8.1
0.3
66.2
9.8
B2 sod2
58.5
19.5
20.8
0.6
139.6 71.2
18.0
6.0
6.4
0.2
43.1
22.0
B2 sod3
69.2
15.7
12.6
1.3
190.6 101.6
16.6
3.8
3.0
0.3
45.7
24.3
B2 os1
65.7
14.7
18.3
0.3
259.9 29.2
16.3
3.7
4.5
0.1
64.6
7.3
B2 os2
63.1
16.8
18.9
0.9
123.9 79.9
20.0
5.3
6.0
0.3
39.3
25.3
B2 os3
37.4
22.5
33.3
1.8
139.2 74.3
11.3
6.8
10.1
0.5
42.1
22.5
S.D.
10.0
2.9
8.4
0.8
59.4
3.5
1.2
2.3
0.3
9.1
6.1
72
ancestral heaths
P
20.4
Pollen
sum
% AP-Alnus
A
C
Q
F
Ca
Q
F
Ca
B1 sod1
42.5
9.4
12.4
0.1
160.2 23.8
187.4 41.5
54.6
0.5
706.6 104.9
B1 sod3
38.4
10.1
7.5
1.5
78.3
30.1
197.1 51.9
38.5
7.7
401.9 154.8
B1 sod4
32.5
8.8
12.5
0.9
B1 os1
39.5
9.2
11.8
0.4
88.5
34.0
140.5 38.2
54.2
3.8
383.2 147.3
96.4
28.5
183.3 42.6
54.6
1.9
447.2 132.4
B1 os2
38.9
10.2
16.1
1.9
118.6 23.3
133.1 35.1
55.2
6.5
405.8 79.9
B1 os3
41.3
8.9
4.5
0.7
108.2 34.7
281.7 61.0
30.5
4.9
737.8 236.6
B1 os4
33.0
6.8
16.1
1.3
72.1
132.4 27.2
64.7
5.1
289.0 147.1
B2 sod1
B2 sod2
27.4
7.3
24.1
0.8
196.1 29.1
79.3
21.2
69.8
2.2
568.7 84.4
31.7
10.6
11.2
0.3
75.6
38.6
140.8 46.9
50.0
1.5
336.2 171.5
B2 sod3
30.5
6.9
5.5
0.6
84.0
44.8
224.5 51.0
40.8
4.1
618.4 329.6
B2 os1
46.2
10.4
12.8
0.2
182.7 20.5
191.6 43.0
53.3
0.9
757.9 85.0
B2 os2
32.9
8.8
9.8
0.5
64.6
41.7
171.2 45.6
51.2
2.4
336.0 216.8
B2 os3
19.5
11.8
17.4
0.9
72.7
38.8
59.7
36.0
53.2
2.9
222.3 118.7
S.D.
7.1
1.5
5.3
0.5
44.4
7.5
58.8
10.5
10.3
2.2
179.9 71.4
Pollen
sum
Table 6.3. Pollen percentages
of the Echoput barrows.
% total pollen sum-heath
P
36.7
% AP-Betula-Alnus
A
C
A
C
P
% AP-Betula-Corylus
Q
F
Ca
A
C
Q
F
Ca
B1 sod1
189.5 42.0
55.2
0.6
714.4 106.1
76.6
17.0
22.3
0.2
288.6 42.9
B1 sod3
197.1 51.9
38.5
7.7
401.9 154.8
80.4
21.2
15.7
3.1
163.9 63.1
B1 sod4
141.5 38.5
54.6
3.8
386.2 148.5
69.7
18.9
26.9
1.9
190.2 73.1
B1 os1
183.3 42.6
54.6
1.9
447.2 132.4
76.2
17.7
22.7
0.8
185.8 55.0
B1 os2
133.1 35.1
55.2
6.5
405.8 79.9
67.2
17.7
27.9
3.3
204.9 40.3
B1 os3
292.4 63.3
31.6
5.1
765.8 245.6
88.8
19.2
9.6
1.5
232.7 74.6
B1 os4
135.3 27.8
66.2
5.3
295.5 150.4
65.2
13.4
31.9
2.5
142.4 72.5
B2 sod1
81.6
71.8
2.3
585.1 86.8
51.1
13.7
45.0
1.4
366.2 54.3
B2 sod2
141.9 47.3
50.4
1.6
338.8 172.9
72.9
24.3
25.9
0.8
174.1 88.8
B2 sod3
224.5 51.0
40.8
4.1
618.4 329.6
82.1
18.7
14.9
1.5
226.1 120.5
B2 os1
195.2 43.8
54.3
1.0
772.4 86.7
77.7
17.4
21.6
0.4
307.2 34.5
B2 os2
172.6 46.0
51.6
2.4
338.7 218.5
76.2
20.3
22.8
1.1
149.5 96.4
B2 os3
60.6
36.5
54.0
2.9
225.5 120.4
48.8
29.4
43.5
2.4
181.8 97.1
S.D.
60.2
10.7
10.7
2.2
185.3 72.0
11.5
4.2
10.2
1.0
66.9
21.8
P
P
25.5
A third case-study that is valuable in the question about the pollen sum is a barrow
complex near Oss in the province of North Brabant in the Netherlands (see fig
6.1). Several barrows are situated here, which have been excavated during several
campaigns. During these excavations most barrows were sampled for pollen
analysis. Two barrows of this barrows complex are similar in age, dating to the
Hallstatt C period (Fokkens et al. 2009b, Fokkens et al. 2012). Hence, pollen
spectra should display similar results in at least the regional vegetation pattern. In
chapter 12, case study Oss-Zevenbergen, the site will be discussed more in detail,
along with an extended overview and discussion of palynological results.
As can be seen in table 6.4 all tested species showed the least variance when a
total pollen sum is used. The variance is slightly higher when an arboreal pollen
sum is used. This also accounts for the herbal species, probably meaning that there
is not much local variability in herbal vegetation. As in the case of the Echoput, an
the pollen sum
73
% AP-Betula
Pollen
sum
Alnus
Corylus
Quercus
Fagus
Calluna
Poaceae
Barrow 3 o.s.
50.8
43.5
3.6
0.0
56.5
3.3
Barrow 3 sod1
55.3
21.9
13.7
3.6
79.3
4.6
Barrow 3 sod2
55.9
24.5
10.6
5.4
75.2
3.6
Barrow 3 sod3
52.8
39.0
3.6
0.5
48.9
41.7
Barrow 7 sod1
56.8
27.8
11.2
0.0
76.4
5.1
Barrow 7 sod2
52.0
28.1
16.3
0.0
65.7
4.9
S.D.
2.4
8.5
5.2
2.4
12.2
15.3
% total AP
Pollen
sum
% total pollensum
A
C
Q
F
Ca
P
A
C
Q
F
Ca
P
B3 o.s.
48.8
41.8
3.5
0.0
54.4
3.2
30.7
26.3
2.2
0.0
34.2
2.0
B3 sod1
52.3
20.7
12.9
3.4
75.0
4.3
28.8
11.4
7.1
1.9
41.3
2.4
B3 sod2
53.8
23.5
10.2
5.2
72.4
3.5
30.0
13.1
5.7
2.9
40.4
1.9
B3 sod3
50.6
37.4
3.5
0.5
46.9
40.0
26.4
19.5
1.8
0.2
24.4
20.8
B7 sod1
55.1
27.0
10.9
0.0
74.2
5.0
29.2
14.3
5.7
0.0
39.3
2.6
B7 sod2
49.4
26.7
15.5
0.0
62.4
4.7
29.2
15.8
9.2
0.0
36.9
2.8
S.D.
2.5
8.3
5.0
2.3
11.7
14.6
1.5
5.4
2.8
1.3
6.3
7.5
Q
F
Ca
P
% total sum-heath
Pollen
sum
% AP-Alnus
A
C
Q
F
Ca
P
A
B3 o.s.
46.6
39.9
3.4
0.0
52.0
3.1
103.1 81.7
6.9
0.0
106.3 6.3
B3 sod1
49.1
19.4
12.1
3.2
70.4
4.0
109.6 43.4
27.1
7.2
157.2 9.0
B3 sod2
50.3
22.0
9.5
4.9
67.7
3.3
116.4 50.9
22.0
11.3
156.6 7.5
B3 sod3
34.9
25.8
2.4
0.3
32.3
27.5
102.3 75.7
7.0
0.9
94.9
B7 sod1
48.1
23.5
9.5
0.0
64.7
4.3
122.9 60.1
24.2
0.0
165.4 11.1
B7 sod2
46.2
25.0
14.5
0.0
58.4
4.4
97.5
52.8
30.7
0.0
123.3 9.2
S.D.
5.6
7.2
4.8
2.1
14.0
9.7
9.6
15.0
10.3
4.9
29.8
29.5
P
% AP-Betula-Alnus
Pollen
sum
A
C
C
80.8
% AP-Betula-Corylus
Q
F
Ca
P
A
C
Q
F
Ca
B3 o.s.
103.1 88.3
7.4
0.0
114.8 6.8
89.8
76.9
6.5
0.0
100.0 5.9
B3 sod1
123.8 49.0
30.6
8.2
1.4
10.2
70.8
28.0
17.5
4.7
101.6 5.8
B3 sod2
126.7 55.5
24.0
12.3
0.0
8.2
74.0
32.4
14.0
7.2
99.6
4.8
B3 sod3
111.7 82.7
7.7
1.0
3.1
88.3
86.6
64.0
5.9
0.8
80.2
68.4
B7 sod1
13.1
64.3
25.9
0.0
176.9 11.9
78.7
38.5
15.5
0.0
105.9 7.1
B7 sod2
10.8
58.5
34.0
0.0
136.7 10.2
72.3
39.1
22.7
0.0
91.4
6.8
S.D.
54.6
15.7
11.4
5.3
79.9
7.9
19.4
6.5
3.1
9.2
25.4
74
ancestral heaths
32.2
Table 6.4. Pollen percentages
of barrows 3 and 7 of
Oss-Zevenbergen.
arboreal pollen sum can be used to show (lack of ) variations in both regional and
local vegetation. Since Betula pollen is rare in these samples it is hard to verify van
Zeist’s assumption that Betula should be excluded from the arboreal pollen sum.
Determining the pollen sum is determining the way to look at the landscape.
In the reconstruction of barrow landscapes several approaches are of interest: what
did the immediate surroundings of the barrow look like, what was the vegetation
character of the open place in which the barrow has been built? Knowing the local
vegetation is indispensible. Besides characterising the immediate surroundings,
knowing what the further surroundings, the regional vegetation, looked like is
necessary to be able to say something about the significance of barrows in the
landscape. What did the wider landscape look like? When trying to determine
the vegetation composition in a wider area it is best to leave the local vegetation
species out of the pollen sum, since they can vary at short distances. An arboreal
pollen sum would then be appropriate to answer these questions. Whether to inor exclude Betula from the pollen sum seems to be site dependent. There is no
reason to exclude Betula from the tree pollen sum according to the case-studies of
Echoput and Oss-Zevenbergen. However, not many pollen grains of Betula have
been found in the samples from these sites. Based on the results of Hijken, Betula
is indeed very variable and of a probably local origin.
When having several contemporaneous samples from one site, information
is provided about the heterogeneity of the herbal vegetation in the immediate
surroundings of the site. In the case of Oss-Zevenbergen the variability of all
species is least when a total pollen sum is used, but when an arboreal pollen
sum is employed the variability is not much higher. Assuming that, besides the
regional vegetation, the local herbal vegetation, which is most likely Calluna,
probably is also very similar, indicates that the area around Oss-Zevenbergen was
a quite homogenous heathland. In the case of the Echoput barrows the situation
is different. Although also showing least variability of all species when a total
pollen sum is used, the variability of especially Calluna is much higher when an
arboreal pollen sum is used. Arboreal species show also higher variability, but
the differences compared with a total pollen sum are small. It is likely that the
variability in Calluna is caused by local variance of the species at the barrow site,
indicating a more heterogeneous heathland.
Conclusions
In conclusion, since herbal vegetation can vary significantly even at short distances,
the best pollen sum to use in barrow palynology is an arboreal pollen sum. Whether
to in- or exclude Betula in this pollen sum seems to be site dependant, but one has
to take into account that beside herbs also arboreal species can vary locally.
Although ∑AP seems to be the most useful pollen sum to apply in
reconstructing the barrow landscape, percentages based on a total pollen sum
do give valuable additional information about the barrow landscape. In barrow
landscape reconstructions it is also of interest to estimate the size of the open place
where the barrow has been built. The ratio between forest and herbal vegetation is
a first indication and a total pollen sum should be applied to get this information.
However, one has to take into account that the non-arboreal pollen percentage
can fluctuate locally and preferably such an estimate should not be based on one
sample. The ratio arboreal versus non arboreal pollen in relation to the size of the
open space will be elaborated in the following chapter.
the pollen sum
75
Chapter 7
The size of an open place where a
barrow was built
Previous research has indicated that barrows were built in open places (see
Chapters 2 and 3), mostly with heath vegetation, surrounded by forest. Knowing
the size of these open spaces would give a more detailed vision of what a barrow
landscape looked like, giving valuable information about the visibility of barrows
in the landscape and about interaction of prehistoric man with the landscape (see
section 3.3). As has been pointed out in Chapter 2, not much is known about
the dimensions of the heathland area a burial mound was constructed in. In this
research it has been attempted to reconstruct the open space the barrows were
built in. Not only by determining the vegetation that was present, but also the
distance of the open vegetation to the forest edge.
In this investigation three steps were taken to ascertain the size of the open
spaces surrounding the barrows. The first step was to determine the minimum size
of the open area, by analysing the construction of the barrow itself (section 7.1).
The second step involved the comparison of pollen spectra of barrow soil samples
with pollen spectra of present Dutch heathland areas (section 7.2). The third step
expands on step two by involving palynological models into the reconstruction of
the barrow landscapes (section 7.3).
7.1 The size and the number of sods used in a barrow
A first indication about the size of the open space can be obtained from the barrow
itself. Barrows were built from sods and these sods were generally taken from
treeless vegetation areas. When the original size of the barrow is still preserved or
can be reconstructed and the size of the sods can be determined, these data can
be used to calculate the treeless area that had been used for sod taking to build
the barrow. Barrows were built in an open area. It is generally assumed that sods
were taken in the direct surroundings of the place the barrow was going to be
built, which can be tested by comparing pollen spectra from the sods and from
the old surface beneath the barrow. In addition the sediment of the sods and the
barrows should be similar when sods were cut close to the barrow location. When
sods were indeed taken in the vicinity of the barrow, the size of the barrow and
the sods can be used to reconstruct the minimum size of the open place where the
barrow was built in as has been suggested by de Kort (1999). The assumption has
to be made that the barrow was a smoothly shaped spherical segment (see figure
7.1). The volume of this spherical segment can be calculated by the following
formula:
1
Vss = _ • π • h • (3r² + 1h²)
6
Vss = Volume spherical segment
h = height of the barrow
r = radius of the barrow
Knowing the thickness of the sods, the area that needed to be stripped for 1 m3 of
barrow can be calculated.
the size of an open place where a barrow was built
77
sod
h
r
7.1.1 An example:
Two barrows are situated at a location called the Echoput near Apeldoorn (see also
8.1). The measurements of the barrows are (van der Linde and Fontijn 2011, 33;
Bourgeois and Fontijn 2011, 65):
Barrow 1: r=9.5 m (d=19 m), h=1.08 m
Barrow 2: r=7.25 m (d=14.5 m), h=1.0 m
Sods: average h=0.25 m
These measurements can be used to calculate the area to be stripped for both
barrows with the formula discussed above, which represents the minimum size of
the open space. For barrow 1 an area of 615 m2 was stripped and for barrow 2 an
area of 332 m2. These results will be further discussed in section 8.1.
7.2 The size of an open heathland area - examples from
present Dutch heathland areas
A second indication about the extent of the open area is provided by the
palynological analyses. Palynological analysis gives insight in the type of landscape
that has been present at the time the pollen precipitated. The quality of the
landscape can be determined by the achieved pollen spectra of an area. In addition,
the quantitative reconstruction of past landscape has been an important goal in
many palynological researches. To achieve this goal it is important to understand
the relation between pollen and vegetation. A pollen spectrum cannot be directly
translated into a vegetation composition, with other words; there is not a simple
linear relationship between pollen and vegetation. Since (most) barrows were built
in heath vegetation, it is the relation between pollen from a heathland area and
its vegetation that is of interest in this study; or, to be more precise, the size of
the heathland area and the position of the barrow inside. A comparative study has
been conducted in Dutch heathland areas to investigate the relation between the
pollen spectra from barrows and the distance from the barrow to the forest. Heath
areas have been selected based on a few criteria. Since the purpose is a comparison
of barrow landscapes, the vegetation composition from the recent heathland
areas should be similar to the heathland of the barrow period. This implies that
the main vegetation of the area should be Calluna vulgaris and that the heath
should be surrounded by forest. Forest in the barrow periods consisted mainly
of deciduous trees, coniferous trees were not present in large numbers yet, given
that most were planted in the Netherlands from the 19th century onwards (Janssen
1974, 57). At present most forests consist for a significant part of coniferous trees.
To make the comparison as realistic as possible heathland areas was selected with
a minimal amount of coniferous trees in and around them. Surface samples were
taken in the heath area with several distances from the forest edge and analysed
for pollen. Pollen spectra and the distance from the sample location to the forest
edge were tested for correlation.
78
ancestral heaths
Figure 7.1. A schematic
drawing of a barrow. To
calculate the minimum area
that has been used for sod
cutting to build a barrow,
a barrow can be seen as a
smoothly shaped spherical
segment, which has been
built with uniform sized sods.
Figure after Doorenbosch
2011, figure 5.6 by J. Porck.
Sites and sampling methods
Herikhuizerveld I (HHV1)
Herikhuizerveld is an extensive heathland area in the National Park Veluwezoom
in the east of the Netherlands (see figure 7.2). The heath vegetation is dominated
by Calluna vulgaris. The forest surrounding the open heathland area is a mixture
of deciduous and coniferous trees. At the edge of the forest one single common
hazel (Corylus avellana) was present. Corylus avellana was a very common shrub in
the barrow periods and pollen of hazel were often found in considerable numbers
in barrow pollen spectra. Nowadays, the shrub is much scarcer and finding a
heathland area with common hazel in the close surroundings was very difficult.
The presence of the species in this heathland area, even if it was just one single
shrub, was an important selection criterion for this area. The area is being grazed
by sheep, horses and Highland cattle. A transect of twelve moss surface samples
was taken (see figure 7.3), or when moss was not present some litter was collected
from the surface. Samples were prepared by the method described in section 4.2.
Herikhuizerveld II (HHV2)
Another transect of eleven moss/litter samples was taken in the Herikhuizerveld,
about 1 km east from the first transect (see figure 7.3). The vegetation criteria
were also applicable here, except for Corylus avellana, which was not present close
to this transect. Samples were prepared by the method described in section 4.2.
Zuiderheide (ZH)
The Zuiderheide is an area in a nature reserve called the Goois Natuurreservaat in
the middle of the Netherlands (see figure 7.2). It is an area of about 300 ha with
forest, heath, drift sand areas and some small lakes. The area is grazed by sheep
and cattle (mostly Highland Cattle). Pollen samples were taken along a transect
Figure 7.2. The locations of the
heath areas at Herikhuizerveld
(HHV), Zuiderheide (ZH) and
St Anthonisbos (StA).
the size of an open place where a barrow was built
79
Sample location HHV1
Sample location HHV2
Bare ground
Calluna heath
Calluna heath with Vaccinium myrtillus (blue berry)
Erica heath
with HHV1
Vaccinium myrtillus
Sample
location
Mixed heath
(Calluna
Sample
location
HHV2and Erica)
Ericaground
heath
Bare
Poaceaeheath
(grass)
Calluna
Ulex europaeus
(common
gorse)
Calluna
heath with
Vaccinium
myrtillus (blue berry)
Pteridium
bracken)
Erica
heath(common
with Vaccinium
myrtillus
Cytisusheath
scoparius
(common
broom)
Mixed
(Calluna
and Erica)
0 125 250
500m
Mixed
forest (deciduous and coniferous)
Erica
heath
Poaceae (grass)
Ulex europaeus (common gorse)
Pteridium (common bracken)
Cytisus scoparius (common broom)
Mixed forest (deciduous and coniferous)
0 125 250
500m
Figure 7.3. Map of the
transect of pollen samples
taken at Herikhuizerveld I
and II.
Sample location ZH
Drift sand
Mixed forest (deciduous and coniferous)
0
75 150
300 m
Mixed heath (Calluna and Poaceae)
Sample location ZH
Drift sand
Mixed forest (deciduous and coniferous)
0
75 150
300 m
Mixed heath (Calluna and Poaceae)
Figure 7.4. Map of the
transect of pollen samples
taken at Zuiderheide.
Sample location StA
Bare ground
Aquatic vegetation
Alder carr
Betula-Quercus forest
Mixed forest (deciuous and coniferous)
Quercus-Fagus forest
Shrubs
Calluna heath
Mixed heath (Calluna and Poaceae)
Calluna heath, moderately grassy
Grassy Calluna heath
Erica heath
Grassland
Wet grassland
0
250 500
1000 m
Water
through heath vegetation, from one forest edge to another (see figure 7.4). Heath
mainly consists of Calluna vulgaris, the forest is a mixture of deciduous (Fagus
sylvatica, Betula pendula, Quercus robur, Amelanchier lamarckii) and coniferous
trees (Pinus sylvestris). Sixteen moss surface samples were taken, or when moss
was not present some litter was collected from the surface. Samples were prepared
according to the method described in section 4.2.
80
ancestral heaths
Figure 7.5. Map of the sample
locations at St Anthonisbos.
Sint Anthonisbos
In 1999 de Kort analyzed the Sint Anthonisbos (forestry of Sint Anthonis) as part
of his MA-thesis. The forestry of Sint Anthonis is a roughly 800 ha nature reserve
area in the province of North Brabant in the Netherlands (see figure 7.2). It is
a varied landscape of production forest for the most part, coniferous alternating
with deciduous trees, and also having heath, fields and pasture areas. In the forest
a heath-drift sand area of approximately 150 ha is situated, dominated by Calluna
vulgaris. The area is currently being grazed by sheep and Highland Cattle. North
of this area an alder brook forest is situated. Eight samples were taken in different
characteristic parts of this landscape (see figure 7.5). Surface samples were taken
from the upper litter layer. Sample preparation and analysis were performed by
J.W. de Kort (for methods see de Kort 1999).
Methods of analysis
Pollen samples were analysed as described in section 4.2. A pollen sum of total
pollen minus the coniferous trees has been used. Pinaceae have been left out of
the pollen sum since they were rarely present at the time the barrows were built.
In addition pollen from Pinaceae are known to be transported over long distances
and can therefore influence the arboreal pollen percentage, while not coming
from within the nearest forest (Pidek et al. 2010). The percentages of arboreal
(AP) and non-arboreal pollen (NAP) have been calculated based on this total
pollen sum. From every sample point the distances to the surrounding forest edges
were measured and the average distance to the forest edge (ADF) was determined.
Using SPSS 19 Pearson product-moment correlation was carried out to identify
significant positive relationships between the percentage of NAP and the ADF
(obvious outliers have been removed from the data), which was the case in all of
the sites. Then regression analysis was applied to the data to show the correlation
between the percentage of NAP and the ADF. It was tested with Graph Pad Prism
5 whether the lines of best fit were significantly different or whether one line
could fit all data sets.
Results and discussion
There has been discussion whether pollen percentages can be correlated with
vegetation openness. Sugita et al. state that pollen percentages are insufficient
to quantify the percentage of land cover in open to semi-open land (Sugita et
al. 1999). Svenning however, found good correlation of NAP percentage with
vegetation openness in interglacial sites (Svenning 2002). This was independently
confirmed by data from beetle, molluscs and plant macro fossils. In this
investigation arboreal percentages and the ADF were significantly positively
correlated per site and these data will be used to estimate the size of an open place.
In table 7.1 arboreal pollen percentages for each sample per site are presented
based on a total pollen sum minus Pinaceae. In addition the ADF per sample is
shown. The AP was plotted against ADF and the line of best fit, which is a loglinear function, shows the relationship between the AP and the ADF (see figure
7.6). The lines of best fit of Sint Anthonisbos and Goois Natuurreservaat did not
differ significantly from each other (p=0.26), the lines that best fit the data of
Herikhuizerveld 1 and Herikhuizerveld 2 were significantly different from each
other and from the other two data sets. However, the lines of best fit of HHV1
and ZH were almost significantly similar (p=0.0495). The differences between
the lines become clear when applied to high percentages of arboreal pollen. Since
most barrow pollen spectra show AP percentages between 30% and 60% (see
the size of an open place where a barrow was built
81
HHV 1
HHV2
ADF (m)
AP (%)
93.5
146.3
StA
ADF (m)
AP (%)
58.4
10
40.1
392.5
142.1
43
140.4
53.5
139.5
144.8
ZH
ADF (m)
AP (%)
ADF (m)
AP (%)
64.4
10
42.8
123
81.1
10.0
93.4
38.2
260.0
52.7
391.25
26.4
372.5
45.5
318.75
25.5
266.7
34
325
16
258.3
26.7
57.4
360
44.9
350
15.6
370
11.2
271.7
19.2
41
313.75
22.7
270.0
18
149.7
62.2
156.0
57.5
337.5
28.2
316.25
23.5
280.0
19.8
320
39.7
318.75
26
285.0
38.6
163.4
270.7
42.6
295.0
34.3
24.4
298.3
27
272.7
18.6
303.3
20.1
263.3
17.1
16.7
78.3
100
31.7
76.6
221.7
38.5
232.3
31.9
y = -38,68ln(x) + 241,58
90
y = -8.4095Ln(x) + 83.675
80
y = -17,43ln(x) + 121,7
70
y = -19,51ln(x) + 138,25
60
AP (%)
Table 7.1. Average distance to
the forest (ADF) and arboreal
pollen (AP) percentages for
each sample per site, based on
a pollen sum minus Pinaceae.
HHV1
50
HHV2
40
StA
30
ZH
20
10
0
0
50
100
150
200
250
300
350
400
450
ADF (m )
Chapter 8) these differences do not have great influence on the interpretation of
the barrow pollen spectra in relation to the size of the open space which line of
best fit will be used.
Hypothetical arboreal percentages from 0 to 100% and the according average
distance to the forest edge based on the best fit lines shown in figure 7.6 are
presented in table 7.2. Most barrow pollen spectra show AP percentages between
30% and 60%. This corresponds with small open spaces with a radius of 25-100
m for 60% AP and rather large open spaces with a radius up to 500 m for 30%
AP.
Other studies have investigated the relation between pollen percentages and
land cover. These studies gave comparable results as described above. Contemporary
moss polster pollen data indicate that closed canopy forests produce arboreal
82
ancestral heaths
Figure 7.6. AP’s plotted
against ADF, showing
the lines of best-fit.
HHV=Herikhuizerveld,
StA=St Anthonisbos,
ZH=Zuiderheide
Site
HHV1
AP (%)
ADF (m)
AP (%)
ADF (m)
AP (%)
ADF (m)
AP (%)
ADF (m)
0
516
30
237
60
109
90
50
HHV2
118845
552
26
1
StA
1077
193
35
6
ZH
1195
257
55
12
HHV1
Table 7.2. Hypothetical non
arboreal percentages from 0
to 100% and the according
average distance to the forest
edge (ADF) based on the best
fit models (see figure 7.6).
10
398
40
183
HHV2
4272
198
StA
607
ZH
716
HHV1
HHV2
20
307
1536
50
70
84
100
39
9
0.4
109
19
3.5
154
33
7
142
71
80
65
3
StA
342
61
11
ZH
429
92
20
pollen percentages of 60-90%. Completely open sites showed arboreal pollen
percentages of less than 50% (Mitchell 2005). Tinsley and Smith took surface
samples across a woodland/heath transition in northern England (Nidderdale,
Yorkshire; Tinsley and Smith 1974). The results showed that close to the woodland
(oak) edge arboreal pollen percentages exceeded 50%. They also showed a rapid
decline of arboreal pollen percentage within 100 m from the forest edge. The
results in this investigation also show a fast decline of arboreal pollen when the
distance to the forest increases, although not as extreme as in Nidderdale. Arboreal
pollen percentages of less than 20% have not been found in the Dutch heathland
areas. This can be explained by the fact that average distances to the forest edge do
not exceed 400 m in the investigated Dutch heathland areas. The woodland may
be further away in one direction, but then in another direction other woodland
would be nearer. Another research by Tinsley in southwest England again showed
a rapid decline in tree pollen with increasing distance from the woodland edge
(Tinsley 2001) and also Lanner showed the main decrease of arboreal pollen to lie
within 160 m from the woodland (Lanner 1966).
When interpreting a pollen spectrum one has to take into account that
individual trees present in an open space have a great influence on the percentage
of arboreal pollen. For example HHV2 sample 10 was taken very close to an oak
tree (Quercus) and the according arboreal pollen percentage is 74.1%. This is a
solitary tree in the heathland area, with the closest forest edge at approximately
250 m, so the distance to the forest edge and the size of the open space would be
clearly underestimated in this case. This is a problem that is difficult to avoid; in
this case study this sample was excluded. In all of the Dutch heathland areas of
this investigation solitary trees were present and it is likely that these trees have
increased the percentage of arboreal pollen. As a consequence the distance to
the forest edge would be overestimated when applying these lines of best fit to
data where no individual trees are present. However, when assuming prehistoric
heathland areas also contained solitary trees, this effect would be compensated.
On the other hand, in the case of an individual tree in close proximity to a barrow,
the arboreal pollen percentage would be too high and the distance to the woodland
edge would be underestimated.
This investigation has focussed on pollen spectra from heathland areas, since
most barrow pollen spectra show considerable percentages of heath pollen.
Precaution has to be taken when the non-arboreal component of a pollen
spectrum is dominated by grasses (Poaceae) instead of heath in combination
with high arboreal percentages. Groenman-van Waateringe has investigated the
the size of an open place where a barrow was built
83
effect of heavy grazing of sites dominated by Poaceae on the pollen production
(Groenman-van Waateringe 1993). Heavy grazing prevents grasses from flowering
and as a consequence from dispersing pollen grains. Pollen spectra from these sites
can display very high percentage of arboreal pollen, since the percentage of the
main herbal vegetation (Poaceae) is kept low by grazing. Consequently, a pollen
spectrum with a high arboreal pollen percentage in combination with Poaceae
being the main component of the non-arboreal pollen can indicate a site in or
close to the forest, but it can also indicate a larger open grassland area that is
heavily grazed.
Conclusions
This research has shown a positive correlation between the percentage of arboreal
pollen and the distance to the forest edge in Dutch heathland areas. Although the
relation seemed to be complicated and precautions must be taken, based on these
results it seems that most barrows, showing arboreal percentages between 30%
and 60%, have been built in open places varying in size from rather extended to
narrow. The larger open spaces used for barrow building would have had a radius
of 200-500 m, the smaller open places a radius of 25-100 m. The ratio of AP
versus NAP should only be used as a rough approximation of the size of an open
place. The pollen spectra from the two barrows at the Echoput (see also 8.1) will
serve as an example. In table 7.2, the ratios of AP versus NAP for samples taken
from barrow 1 and barrow 2 are shown. The average arboreal percentages for each
barrow (30% and 27%) imply an open space of 200-500 m (see table 7.1).
7.3 The distance of a barrow to the forest edge palynological modelling
Palynological analysis gives insight in the type of landscape that has been present
at the time the pollen precipitated. As has also been demonstrated in section 7.2
the relation between pollen and vegetation is quite complicated. The relation is
highly dependent on the dispersal of pollen grains from the pollen source into the
surroundings. An important factor that determines the dispersal of pollen is the
pollen productivity of a taxon. Other factors include pollen-specific characteristics
(size, weight, shape), wind speed, height of the vegetation in the surroundings,
etc. In the last few decades models of pollen dispersal and deposition have been
developed, improving the understanding of this pollen-vegetation relationship.
These models could be of great value when reconstructing barrow landscapes,
possibly offering the opportunity to show a more detailed view of the vegetation
in the surroundings of a barrow and the size of the open space.
Extended R value (ERV) models have been developed to convert pollen
percentages into relative plant abundances. These models have been based on the
R-value model developed by Davis, Ri=pi,k/vi,k, describing the linear relationship
between the pollen percentage (pi,k) and the vegetation cover percentage (vi,k) of
taxon i at site k (Davis 1963). This model has been adjusted in the following
decades to account for background pollen. This resulted in the ERV-model. The
basic assumption of the ERV-model is that the pollen loading (number of pollen
grains) of taxon i at site k (γi,k) is linearly related to the distance-weighted plant
abundance (in kg/m2 or m2/m2) of taxon i around site k (xi,k), γi,k = αixi,k + ωi. The
pollen productivity of taxon i (α) and the background pollen loading of taxon i
(ωi) are constant for every taxon (Prentice and Parsons 1983). Since pollen loadings
can vary greatly within sites and are often not available (for example in fossil
samples) percentage data have to be used. However, interdependence of pollen
84
ancestral heaths
percentages can cause non-linearity of the relationship between pollen and
vegetation percentages (Fagerlind-effect; Fagerlind 1952). The ERV-model has
taken the Fagerlind-effect into account by introducing so-called site-factors to be
able to relate pollen percentages to vegetation percentages. There are three
submodels of the ERV-model that have different assumptions about the background
pollen loading. In submodel 1 (Parsons and Prentice 1981), the background
pollen loading (ωi) is assumed a constant proportion of the pollen loading of all
pollen taxa γk (ωi = zi • γk, where zi is the background pollen percentage). Submodel
2 (Prentice and Parsons 1983) assumes constant background pollen percentage
relative to the total plant abundance of all taxa φk (ωi = zi • φk). Submodel 3 (Sugita
1994) relates the pollen percentage to the absolute vegetation abundance, where
submodels 1 and 2 used vegetation percentages. This model assumes constant
background pollen loading ωi for each taxon. Distance weighing of the vegetation
data is necessary, since plants close to the sampling point contribute more pollen
than plants further away. The simplest way to weigh the vegetation is by dividing
the plant abundance by the distance d between the plant and the sampling point,
1_
(Prentice and Webb 1986), or by the square of the distance, d1_ (Webb et
d
al. 1981). Two other models use a weighting method that is based on Sutton’s
(1953) equations for dispersal of small particles in the atmosphere (Sutton 1953).
These models are referred to as the Prentice/Sugita models (Prentice 1985, Sugita
1993). In these models the pollen loading is dependent on the distance, atmospheric
conditions and taxon-specific properties.
The parameters α and ω can be estimated using maximum likelihood methods
(Parsons and Prentice 1981), meaning that the values of these parameters are the
most likely of having produced the observed values. The lower the maximum
likelihood function score, the better the fit of the observed data to the model
estimated data. Because the pollen assemblage is a distance-weighted function
of the plant abundance, the maximum likelihood function score should decrease
and approach an asymptote as the vegetation area increases. The relevant source
area of pollen (RSAP), the area beyond which goodness of fit between pollen
and vegetation data does not improve (Sugita 1994), can be determined. The
pollen coming from beyond the RSAP can be estimated as the background pollen
loading. The pollen productivities should be estimated at or beyond the RSAP
(Broström et al. 2008). When the vegetation data are properly distance weighted
the slope of the ERV-models represents an estimate of the pollen productivity
(PPE=pollen productivity estimate), the y-intercept represents the background
pollen loading.
2
Barrow landscape simulation
Software, called HUMPOL (Middleton and Bunting 2004), has been developed in
which vegetation cover maps can be used to generate modelled pollen assemblages.
This allows for comparing multiple landscape scenarios to fossil pollen spectra
(Nielsen 2004, Gaillard et al. 2008, Soepboer and Lotter 2009). The openness of
the landscape around a barrow based on the ratio NAP versus AP (see section 7.2)
can be tested when PPEs are known. PPEs have been calculated for several sites
throughout Europe based on the theory described above. An overview has been
published by Broström et al. (2008). Differences in PPEs between sites are not
unusual. Several explanations for these variances are given. Environmental factors
such as climate, vegetation structure and vegetation composition have influence
on the pollen production. In addition, different vegetation survey methods
the size of an open place where a barrow was built
85
(visual estimate of cover, modified circle walking and rooted frequency) result
in differences in PPEs, so PPEs should only be compared when the vegetation
survey method is similar (Bunting and Hjelle 2010). Also the method of pollen
collection (moss polsters, lake sediments) can cause differences in PPE (Broström
et al. 2008). A third factor that can cause differences is the reference taxon. PPEs
are calculated relative to a reference taxon (the pollen productivity of the reference
taxon is set to 1). Poaceae has become a standard reference taxon (Broström et al.
2008), since it is a widespread taxon and present in most vegetation communities.
However, it is very likely that the pollen production of Poaceae differs between
sites, causing differences in PPEs relative to the PPE of Poaceae. When comparing
PPEs between sites, the reference taxon should be similar, but the variability of
the reference PPE should be borne in mind. PPEs have not yet been determined in
the Netherlands. Detailed vegetation and pollen data are necessary to determine
PPEs for a certain region. After intensive search, it would appear that such a
combination of data is not possible for Dutch heathland areas at the moment.
As an alternative, PPEs from a comparable region should be used for testing the
Dutch barrow landscapes. PPEs derived from Southern Sweden and Norway are
probably best comparable to Dutch PPEs. Although southern Sweden has longer
days in summer and temperatures in western Norway are slightly lower than in the
Netherlands, both regions have a moderate maritime climate like the Netherlands.
Since in southern Sweden PPEs have been calculated for herbal as well as arboreal
taxa, these data will be used to simulate a Dutch heathland area to test whether
these PPEs can be applied in simulating barrow landscape scenarios. However,
PPEs for tree taxa were originally based on Juniperus and later recalculated for
Poaceae as reference taxon (Sugita et al. 1999, Broström et al. 2008). This should
be kept in mind when interpreting the results. The software that has been used in
this simulation is HUMPOL (Middleton and Bunting 2004).
St Anthonisbos
The St Anthonisbos (see section 7.2) seemed to be appropriate for simulation. A
digital vegetation cover map from the area was provided by Staatsbosbeheer. This
digital vegetation map was converted into ASCII for the HUMPOL software to
process it (see figure 7.7). Vegetation communities have been simplified based on
the available PPEs. Fall speeds of the included taxa have been based on Sugita et
al. (1999) and Broström et al. (2004). See for an overview of used PPEs and fall
speeds table 7.3. Wind speed was set to 4.05 m s-1 according to the daily mean
wind speed measured at the Bilt from 1904-2012. The wind rose was set in 16
directions according to the frequency distribution of daily mean wind direction
at the Bilt (The Netherlands) from 1904-2012. Both wind speed and wind rose
are based on data provided by the Royal Netherlands Meteorological Institute.
It is often assumed that the prevailing wind direction in the Netherlands is
south-southwest. The mean frequency distribution of the wind direction for the
last century shows that this is not entirely correct (see figure 7.8). November,
December and January have been excluded for the calculation of the mean wind
speed and wind rose, since the majority of plants do not produce pollen during
these months.
Within a series of concentric rings around the sample point a visual estimate of cover for each taxon
within each full ring is recorded.
(http://climexp.knmi.nl)
Corylus (hazel) is one of the earliest to flower in February, while Ericaceae (heath) can still flower up
to October (Weeda et al. 1988, 37).
86
ancestral heaths
Figure 7.7. Vegetation cover
map of St Anthonisbos as
used in the simulations.
Betula-Quercus=15% Betula,
30% Quercus, 10% Poaceae
; Bare= no vegetation; Alder
carr= 70% Alnus, 5% Salix,
10% Poaceae , Dry heath=
90% Calluna; Dry heath,
slightly grassy= 70% Calluna,
10% Poaceae ;Dry heath,
moderately grassy= 60%
Calluna, 30% Poaceae; Dry
heath, grassy= 40% Calluna,
30% Poaceae; Wet heath=
30% Calluna, 20% Poaceae ;
Grass= 90% Poaceae.
Bare
Alder carr
Betula-Quercus
Dry heath
Dry heath, slightly grassy
Dry heath, moderately grassy
Dry heath, grassy
Wet heath
0
250
500
1000 m
Grass
Figure 7.8. Wind rose
according to the frequency
distribution of daily mean
wind direction at the Bilt (The
Netherlands) from 1904-2012.
November, December and
January have been excluded.
Table 7.3. Pollen productivity
estimates (PPE) and fall
speeds of the pollen from the
taxa used in the simulations.
PPE
Fall speed (m/s)
Alnus
4.2
0.021
Betula
8.9
0.024
Corylus
1.4
0.025
Fagus
6.7
0.057
Quercus
7.6
0.035
Salix
1.3
0.022
Tilia
1.3
0.032
Calluna
4.7
0.038
Poaceae
1
0.035
Pollen spectra have been determined from eight sample locations in the area by
de Kort (see also 7.2). Simulated pollen percentages of these eight locations have
been calculated by the simulation program based on PPEs from south Sweden.
The simulated and the percentage data observed in the real samples for the taxa
used in the simulation are shown in table 7.4 (simulation 1 versus observed 1). As
can be seen the simulated data fit the observed data considerably well. Although
the ratio between Calluna and Poaceae (see table 7.4) appears to fall short of
accurately portraying the situation, the ratio between arboreal and non arboreal
pollen percentages seems to be appropriate, with exception of samples 5, 6 and 7.
This is probably due to pine trees close to these sample locations (de Kort 1999).
When Pinus is left out of the simulation, the simulated data fit the observed
pollen data very well (see table 7.4). The ratio Calluna versus Poaceae not being
the size of an open place where a barrow was built
87
sample 1
sample 2
Sim. 1
real 1
sim. 2
Alnus
69.82
77.09
74.73
Betula
5.07
6.48
5.43
Pinus
6.60
3.63
0
Quercus
8.19
3.48
8.76
Salix
1.59
1.42
1.71
Calluna
1.82
0.79
2.06
real 2
sim. 1
real 1
sim. 2
real 2
80
2.85
6.72
10.80
18.29
3.32
20.04
4.47
12.55
4.90
0
14.10
8.75
0
0
3.61
17.44
3.31
20.28
3.62
1.48
0.08
0.00
0.09
0
0.82
6.62
7.98
8.08
8.74
Poaceae
6.91
7.11
7.32
7.38
48.10
57.20
55.68
62.69
AP
91.27
92.10
90.63
91.80
45.27
34.82
36.24
28.57
sample 3
sample 4
sim. 1
real 1
sim. 2
real 2
sim. 1
real 1
sim. 2
real 2
Alnus
1.12
2.57
1.26
2.89
0.38
2.73
0.40
3.11
Betula
8.37
8.28
9.41
9.30
4.68
4.54
4.95
5.18
Pinus
11.11
10.93
0
0
5.90
12.26
0
0
Quercus
13.65
9.32
15.28
10.47
7.38
4.54
7.82
5.18
Salix
0.03
0.42
0.04
0.47
0.01
0.23
0.01
0.26
Calluna
50.81
19.07
57.74
21.41
67.33
52.76
72.26
60.14
Poaceae
14.90
49.41
16.27
55.47
14.31
22.94
14.55
26.14
AP
34.28
31.52
25.99
23.13
18.35
24.30
13.18
13.72
sim. 1
real 1
sim. 2
real 2
sim. 1
real 1
sim. 2
real 2
Alnus
0.23
1.14
0.23
1.28
0.09
2.46
0.09
3.13
Betula
5.95
2.93
6.12
3.29
5.11
7.58
5.45
9.64
Pinus
7.45
10.78
0
0
6.63
21.31
0
0
Quercus
9.35
4.61
9.62
5.17
8.21
5.74
8.76
7.29
Salix
0.01
0.00
0.01
0.00
0
0.20
0.00
0.26
Calluna
51.40
44.25
65.83
49.60
75.90
53.48
82.07
67.97
Poaceae
25.62
36.29
18.19
40.67
4.05
9.22
3.62
11.72
AP
22.99
19.46
15.98
9.73
20.04
37.30
14.30
20.31
sample 5
sample 6
sample 7
sample 8
sim. 1
real 1
sim. 2
real 2
sim. 1
real 1
sim. 2
real 2
Alnus
0.12
2.25
0.13
3.06
0.17
2.68
0.18
3.44
Betula
6.77
6.29
7.36
8.53
7.71
7.79
8.49
9.97
Pinus
8.49
26.33
0
0
9.58
21.88
0
0
Quercus
10.64
7.35
11.58
9.98
12.05
9.26
13.28
11.86
0
0.47
0.00
0.64
0.01
0.40
0.01
0.52
Calluna
37.20
28.00
41.73
38.00
12.72
2.42
15.12
3.09
Poaceae
36.77
29.30
39.19
39.77
57.77
55.57
62.92
71.13
AP
26.02
42.70
19.07
22.22
29.52
42.01
21.96
25.77
Salix
88
ancestral heaths
Table 7.4. Simulated and
real pollen percentages for St
Anthonisbos. Simulation 1 is
including Pinus, simulation 2
excluding Pinus.
Table 7.5. Average arboreal
percentages (AP) and non
arboreal percentages (NAP)
for the two Echoput barrows.
AP (%)
NAP (%)
Barrow 1
31.25
68.75
Barrow 2
28.37
71.63
accurate can be explained by the vegetation data of the vegetation cover map not
being detailed enough.
Echoput
As a pilot study the PPEs from south Sweden will be applied to a barrow
landscape. The assumption has to be made that PPEs did not change through
time. Two barrows are situated near Apeldoorn at a location called the Echoput.
These barrows have been excavated and palynologically investigated. The case
study will be extensively discussed in chapter 8 (see 8.1; see also 5.3 and 7.2). In
MOSAIC (part of the HUMPOL software) a barrow landscape of 1.5 by 1.5 km
has been designed based on the palynological results of the Echoput barrows. The
taxa considered in this simulation are Calluna, Poaceae, Alnus, Corylus, Fagus and
Quercus. The average percentage of AP in the pollen spectra based on the taxa
used in the simulation is 30% (see table 7.5). Based on the results this open space
consisting of a mixture of Calluna heath and grasses was created with a radius of
300 m (see table 7.2).
The Echoput samples showed rather high percentages of Corylus. Corylus is a tree
that requires light conditions to grow and it will not be able to survive in the
reduced light conditions in a closed forest. This indicates that the open space was
probably surrounded by a mantle vegetation mainly consisting of Corylus, shown
by a white ring in the simulation landscape (see figure 9a). Based on the observed
pollen percentages in the Echoput samples, the surrounding forest consisted
mainly of Quercus and some Fagus. In the simulation landscape a hypothetical
mixture based on an educated guess (e.g. observed pollen percentages and fall
speed of the taxa) of 75% Quercus and 15% Fagus was created, surrounding the
Corylus vegetation mantle. The last species to add to the simulation landscape
is Alnus. Alnus is a tree that requires moist conditions to grow and it is most
likely that alder carr was present in the lower and wetter brook valleys in the
environment. Since these were located south and west of the Echoput barrows
this is where alder carr was placed in the simulation landscape. The simulation
landscape with the according vegetation communities is shown in figure 7.9a.
Fall speed and PPEs from southern Sweden have been applied to simulate pollen
assemblages at the locations of the two Echoput barrows. The wind rose has been
set as it was in the St Anthonis simulation.
Table 7.6 shows the simulated (e.g. the percentages that were obtained by the
simulation) and the observed (e.g. in the ‘real’ Echoput samples) pollen percentages,
based on a total pollen sum of taxa used in the simulation. The total arboreal (AP)
and non-arboreal pollen (NAP) percentages have been calculated as well. Note
how the simulated AP and observed AP are very different from one another. This
indicates that the size of the open space was overestimated in the first simulation.
To improve the simulation, the open space in the simulation landscape was
decreased to a radius of 200 m, all other components being left unchanged (figure
7.9b). In table 7.5 can be seen that the percentage of AP now fits the observed AP
percentage. However, the composition of the vegetation communities seems not to
be accurate. Several simulation landscapes with varying vegetation compositions
were tested to come to a possible landscape scenario where the simulated data fit
the observed data. Such a possible landscape scenario is shown in figure 7.10.
the size of an open place where a barrow was built
89
Bare
Alder carr
Corylus forest edge
Fagus
Quercus
0 100
300m
0 100
300m
Dry heath, moderately grassy
This landscape shows an open place with a grassy heathland. The alder forest is
adjacent to this heathland open space. A wide forest edge consisting of Corylus
surrounds the open space. The surrounding forest is quite open, with only 50%
Quercus and 15% Fagus. However, this is not a very realistic landscape. The
geology of the environment is also of great influence on the landscape. To create a
more realistic landscape a digital elevation map of the environment (Dutch: AHN
Actueel Hoogtebestand Nederland) was combined with the simulation landscape in
figure 7.10. The result is shown in figure 7.11. In this scenario the heath area is
not circular, but stretched along the top of the plateau the Echoput barrows were
built on. This grassy heathland, 40% Calluna and 50% Poaceae, is approximately
700 m long and 300 m wide. The barrows are not located in the centre, but in
the southwest; the ADF is approximately 250 m. Alder carr is situated in the
lowest areas, the brook valleys. The heath is surrounded by a zone of Corylus of
approximately 150 m in width. This zone gradually shades into the forest, which
is very open with 30% Quercus and 5% Fagus. Simulated pollen percentages show
that the alder carr is greatly underestimated and that the abundance of Quercus is
slightly overestimated. It is however not very likely that the alder carr was situated
closer to the barrows. It is possible that the PPE of Alnus used in this simulation
was too low. A study performed in England has for example calculated a PPE for
Alnus of 11.4 (Broström et al. 2005) and in Estonia an Alnus PPE of 13.92 was
found (Poska et al. 2011). When the PPE of Alnus in this simulation is set to 11.4
the simulated pollen percentages are indeed much more similar to the real pollen
percentages (see table 7.5).
The investigation gives promising results for the use of models in the reconstruction
of past barrow landscapes. Quantitative reconstructions of barrow landscapes
would be very useful to enhance our knowledge about the environments that
the barrows were built in. These simulations were based on pollen productivity
estimates (PPEs) that have been determined in southern Sweden. As has been
shown by Broström et al. (2008) PPEs can differ between sites. Although the PPEs
from southern Sweden seem to be fairly applicable in Dutch simulations it is very
likely that the actual PPEs in the Netherlands were not exactly alike, as has also
been been shown in the last simulation. Even within the Netherlands PPEs could
show variations between different types of landscapes. Barrows were mainly been
built in heath vegetation. It is therefore recommended to have detailed vegetation
surveys and collection of pollen data in Dutch heathland areas to be able to achieve
PPEs of the major Dutch heathland taxa. The vegetation survey area should be
large enough to cover the RSAP of the pollen. Bunting and Hjelle showed that
the relevant source area of non-arboreal pollen in heathland areas is less than 4
m (Bunting and Hjelle 2010). This would make the achievement of PPEs of the
herbal heathland taxa relatively simple. For the determination of arboreal PPEs
90
ancestral heaths
Figure 7.9a-b. Simulated
landscape for the Echoput
barrows with different sizes
of heathland (a=300m and
b=200m). Alder carr=90%
Alnus; Quercus=100%
Quercus; Fagus=100% Fagus;
Corylus forest edge= 70%
Corylus, 10% Calluna, 10%
Poaceae; Dry heath= 50%
Calluna, 40% Poaceae.
Figure 7.10. Example for an
Echoput landscape scenario.
Alder carr= 90% Alnus;
Quercus= 100% Quercus;
Fagus= 100% Fagus, Bare=
no vegetation, Corylus
heath edge= 70% Corylus,
10% Calluna, 10% Poaceae;
Corylus forest edge= 70%
Corylus, 5% Quercus, 10%
Poaceae; Dry heath= 30%
Calluna, 40% Poaceae.
Bare
Alder carr
Corylus forest edge
Corylus heath edge
Fagus
Quercus
0 100
Figure 7.11. Example for an
Echoput landscape based on
the AHN. Alder carr= 90%
Alnus; Background forest=
30% Quercus, 5% Fagus,
20% Poaceae; Corylus heath
edge= 70% Corylus, 10%
Calluna, 10% Poaceae;
Corylus forest edge= 70%
Corylus, 5% Quercus, 10%
Poaceae; Dry heath= 40%
Calluna, 50% Poaceae.
300m
Dry heath, moderately grassy
Alder carr
Background forest
Corylus forest edge
Corylus heath edge
0 100
300m
Dry heath, moderately grassy
Alnus
barrow 1
barrow 2
real
sim. 1
sim. 2
sim. 3
sim. 4
sim. 5
19.88
3.00
2.86
12.17
6.05
14.87
16.30
2.98
2.84
12.92
5.66
14.01
4.76
0.39
0.54
2.13
3.93
3.56
4.91
0.39
0.54
2.02
4.35
3.97
Corylus
barrow 1
barrow 2
Quercus
barrow 1
barrow 2
real
sim. 1
sim. 2
sim. 3
sim. 4
sim. 5
16.59
23.63
11.4
10.89
9.87
16.59
16.56
23.59
11.01
11.37
10.36
16.56
Poaceae
Table 7.6. Simulated and real
(average) pollen percentages
for the two Echoput barrows.
Sim1=landscape shown in
fig.7.10a (ADF=300 m);
sim2= landscape shown in
fig.7.10bx (ADF=200 m);
sim3= landscape shown in
fig. 7.11; sim4=landscape
shown in fig. 7.12 (based
on AHN) ; sim5=landscape
shown in fig.12, with PPE
for Alnus adjusted to 11.4.
Wind speed was set to 4.05
and wind rose according to the
Dutch growing season for all
simulations.
real
sim. 1
sim. 2
sim. 3
sim. 4
sim. 5
barrow 1
16.15
11.30
10.14
19.01
17.08
15.47
barrow 2
19.26
11.31
10.15
18.95
16.99
15.48
real
sim. 1
sim. 2
sim. 3
sim. 4
sim. 5
barrow 1
0.53
1.97
2.86
1.94
1.03
0.93
barrow 2
0.31
1.96
2.85
1.87
1.07
0.98
real
sim. 1
sim. 2
sim. 3
sim. 4
sim. 5
barrow 1
52.60
66.75
59.98
53.35
61.03
55.3
barrow 2
52.59
66.79
60.03
53.22
60.55
55.2
real
sim. 1
sim. 2
sim. 3
sim. 4
sim. 5
barrow 1
31.25
21.95
29.89
27.64
21.9
29.23
barrow 2
28.15
21.89
29.82
27.82
22.45
29.32
Fagus
Calluna
AP
the size of an open place where a barrow was built
91
the survey area should be much wider, approximately 1500-2000 metre radius
(Broström et al. 2008). A relatively simple approach of modelling landscape has
been used here as a pilot study to test PPEs on a barrow landscape scenario. When
Dutch PPEs are known this approach might be extended by using the ‘Landscape
Reconstruction Algorithm’ (LRA) developed by Sugita (2007a, b). The inverse
forms of the ERV-models can be used to reconstruct past vegetation abundance.
With an estimate of the pollen productivity and known pollen proportions,
vegetation proportions can be calculated. However, the background pollen
component is not constant and changes over time. This change in background
pollen can be estimated when regional plant abundance and the pollen source area
are known. For that reason the LRA has been developed, estimating regional and
local vegetation abundance. To obtain a reliable estimate of the regional vegetation
it is advisable to use pollen assemblages from large sites (>100-500 ha) or when
such large sites are not available, from as many smaller sites as possible. It would
be of value to investigate whether this approach can be applied to Dutch barrow
landscapes.
7.4 Discussion
In this chapter an attempt was made to give an estimate of the size of the open
space barrows were built in. The number of sods that were used to construct
a barrow can provide a first indication of the minimum size of this open spot
(section 7.1). The ratio between AP and NAP has been used to get a second
approximation of the size of the heath area a barrow was built in, by estimating the
Average Distance to the Forest (ADF; section7.2). A positive correlation between
the size of an open place and the percentage of NAP was found, but there are
differences present between the sites and the relation was therefore complicated.
Also Broström et al. showed the ratio AP versus NAP cannot be simply translated
in vegetation openness, and that differences in background pollen appear to play
an important role in the relative representation of NAP (Broström et al. 1998).
Sugita et al. too underline the importance of background pollen coming from
the regional vegetation (Sugita et al. 1999). The landscape simulation models
and software that have been developed could give better insight in the relation
between Non Arboreal Pollen percentage and landscape openness and enhance
our understanding on what a barrow landscape looked like (7.3). To apply these
models properly to Dutch barrow landscapes further research is recommended.
This study will therefore not test further barrow landscape scenarios with the
palynological modelling methods, excepting the Echoput.
The three approaches have been applied to the Echoput case study. The first
approach suggested a minimum size of the open space of about 615 m2 for barrow
1 and 332 m2 for barrow 2. This could indicate a circular open spot with an
ADF of about 14 m and 10 m respectively, which seems rather small. The second
approach yielded an ADF of approximately 300 m, which is indeed much greater.
The third approach indicated that an ADF of 300 m might be an overestimation.
The ADF was corrected to approximately 250 m. The following chapters of this
thesis will discuss several case studies and for each case study the size of the open
space a barrow was built in will be estimated. Since the palynological modelling
approach still needs further research, the second approach (e.g. AP:NAP) will be
used as a standard in the following chapters. It should be kept in mind though
that a slight overestimation of the size of the open space could occur.
92
ancestral heaths
Part Three
Case–studies
Part 3 will discuss several case studies, gathering information in answer to the
questions that were put forward in the beginning of this thesis (Chapter 3): what
did the barrow landscape look like? An answer to this question is needed in order
to understand the function of barrows in the landscape and how barrows relate
with the natural and cultural landscape surrounding them. What was the original
impetus behind the creation of the open space a barrow was built in, and what was
that open space used for? Human activity played an important role in the history
of an open space. An open space could for example have been used as a grazing
area, for the cultivation of crops, or it could have served as settlement location.
In part 2, Chapters 4-7, several methodological aspects of pollen sampling in
barrow research have been described and discussed. In addition the uncertainties,
assumptions and consequences for the results were discussed as well as how to
interpret the results. All have a bearing on part three of this thesis. In Chapter 6
it was concluded that the best pollen sum to use in barrow research is an Arboreal
Pollen (AP) sum. Inclusion of Betula (birch) should be decided per site.. To be
able to compare all sites with one another it has been decided to apply an arboreal
pollen sum minus Betula to all sites. The percentages of arboreal and non arboreal
pollen have been calculated based on a total pollen sum (see sections 6.2 and
7.2).
In Chapters 8-12 several case-studies are discussed. Chapters 8-10 consider
three research areas that are all situated on the push moraines that were formed
during the Early and Middle Pleistocene in the northern half of the Netherlands.
Chapters 11 and 12 will study two research areas that are situated in the
southern part of the Netherlands, where cover sand was deposited during the
Late Pleistocene. Most of the palynological data discussed in Chapters 8-12 were
originally obtained by other researchers (for references see the corresponding casestudies). For the case-studies most of the data were re-analysed and pollen spectra
and/or pollen diagrams were re-plotted. In some cases pollen percentage data had
to be recalculated based on the appropriate pollen sum for the present barrows
study (e.g. tree pollen sum minus Betula, see Chapter 6). Subsequently all (reanalysed) data have been reinterpreted by the author of the present work.
Not all methods described in Chapters 4-7 could be applied to all casestudies. As has been mentioned above, much of the palynological data used in the
following chapters were obtained by other researchers many years ago. It is often
the case that documentation about the sampling method and the exact sample
locations is not available, especially when dealing with older excavations. Many
burial mounds were excavated in the 1920s-1940s. Knowledge about barrows has
grown enormously since then and excavation methods are much more detailed
now. The older excavations sometimes lack proper documentation, and when
this is available, is sadly incomplete. Dating of the excavated barrows was usually
based on grave goods, but the exact location of these finds was not always well
documented. Multi-period barrows were often not recognized, while stratigraphic
differences were not distinguished. Hence, it is difficult to relate grave goods to
proper dating of the barrow. Many of these barrows were re-excavated during
the 1970s. These re-excavations were mostly based on the documentation of the
older excavations, which we now know was not always correct or complete. Many
samples for pollen analysis were taken during these re-excavations. Since (the
original) documentation was not always accurate it is not in all cases clear what
the exact sample location was in a barrow. It is therefore hard to establish the
relation of the samples to the dating of the barrow. The old surface of the second
period of barrow at Stroe (section 8.10), for example, was sampled following the
documentation, while with the present knowledge it cannot be confirmed that
there was indeed a second period in that barrow. In such a case it is hard to
specify the exact sample location and to say more on the dating of the according
pollen spectrum. However, since it is not always possible to retrieve the necessary
information to clarify this relation, one has to rely on data that are available. As
a consequence one has to assume that not all pollen spectra are correctly dated.
However, only about 5% of the barrows discussed in this thesis seem to have
encountered this problem and this will be accounted for in the discussion of
the corresponding case-studies (e.g. 8.3 Ermelo, 8.10 Stroe, 8.11 Uddelermeer,
9.1 Warnsborn and 9.1 Wolfheze). For a more extensive discussion about the
reliability of older excavations and consequences for interpretation in present
research see Bourgeois (2013, p.47-48).
As a final introductory note to the case-study chapters, it should be mentioned
that many barrows are known by several names. Barrows were often re-excavated
and across several publications the same barrows were assigned different names.
For this thesis all the barrows that are discussed have been given a new name. In
Appendix I an overview can be found of the other names and numbers a given
barrow was assigned in the several publications from which data were extracted for
use in the following chapters.
Chapter 8
Northern and central Veluwe
In the northern and central part of the Veluwe (the Netherlands), palynological
data was obtained from several barrows that exist in an area of approximately
20 by 20 km (see figure 8.1). In the following sections the palynological results
of these barrows will be described and discussed, based on the theory set out in
part two of this thesis (Chapter 4–7). This chapter will start with two barrows
at the Echoput. All the data from these barrows were collected by the author.
Most of the methods described in Chapters 4-7 have been applied to the barrows
of the Echoput and therefore these barrows will feature first. The second group
of barrows that will be discussed in this chapter is located at Niersen-Vaassen.
The data from two barrows of Niersen were collected by the author. The data of
all other barrows in this chapter were obtained from other researchers and they
will be discussed after the discussion of the Echoput barrows and the barrows of
Niersen-Vaassen. At the end of this chapter a pollen diagram derived from a lake
sediment (Uddelermeer, see section 8.11) will be presented after all the barrows
have been discussed. This pollen diagram will provide more information about the
vegetation in the wider surroundings of the barrows.
Vierhouten
Emst
Niersen
Celtic Field
Ermelo
Vaassen
Putten
Uddelermeer
Echoput
Boeschoten
Stroe
0
1500
3000
6000 m
Ugchelen
Sampled barrow
Figure 8.1. Detailed map of
the Echoput and surroundings
with the location of all
discussed barrows. The map
is based on digital elevation
model of the AHN (copyright
www.ahn.nl).
Celtic field
Other barrows
m NAP
150 m
0m
northern and central veluwe
95
organic layer
Figure 8.2. The Echoput
barrows one year after they
were excavated.
A horizon on top of barrow
barrow, with sods visible
old surface
A horizon under barrow
B horizon under barrow
(B)/C horizon under barrow
8.1 Echoput
Close to Apeldoorn two barrows are situated on a small hilltop. The site that these
barrows are located at is known as the Echoput. Excavation of these barrows took
place in the summer of 2007 (see figure 8.2). For an extensive description of the
excavation results see Fontijn et al. (Fontijn et al. 2011)
8.1.1 Site description
Both barrows showed similarities in construction and soil properties. They were
both built on a surface in which an Umbric Podzol (Dutch soils classification:
Holtpodzol gY30 [see Bodemkaart van Nederland]) had developed. The barrows
were constructed of sods that were still clearly visible, which were taken from a
Holtpodzol identical to the one they were placed on top of. The old surface was
well recognizable in the soil profile (see figure 8.3). The barrows were dated to
Bodemkaart van Nederland 1:50.000 toelichting kaartblad 33 west Apeldoorn, p. 27, 67-8.
96
ancestral heaths
Figure 8.3. A section of
Echoput barrow 2 (section 2.1
of barrow 2 in trench 2) with
the old surface clearly visible.
Photograph by Bourgeois
2012, figure 3.2).
15
P27
9
17
8
P10
16
P7
P1
14
3
22
6
7
Barrow 2
5
23
2
10
24
19
18
4
P2
Figure 8.4. Plan of all
trenches of the excavation of
the Echoput barrows. Trench
numbers are indicated. The
P-numbers indicate the
post features of which the
fills have been sampled for
pollen analysis. Figure by P.
Valentijn/ M. Doorenbosch.
20
P5
P12
21
1
Barrow 1
11
13
12
0
10 m
the Middle or Late Iron Age, based on 14C of charcoal from both ring ditches:
2225±30 BP (GrA-44706; 331-203 cal BC, calibrated with Oxcal 4.2; mound
1) and 2240±35 BP (GrA-44879; 326-204 cal BC, calibrated with Oxcal 4.2;
mound 2) as post quem dates. In addition, a terminus ante quem date for mound
1 of 2190±35 BP (GrN-32158; 376-171 cal BC, calibrated with Oxcal 4.2)
was derived from charcoal from a pit (S1) that was dug into the mound. The
combination of post and ante quem dates and the similarity of both mounds
make it likely that both mounds were constructed in the 4th or 3rd century cal BC
(Fontijn 2011, 152-153).
Excavation of the surroundings revealed a large amount of features including
a round post structure and two other post structures (see figure 8.4). Traces
dating to the Late Mesolithic and the Late Neolithic B period have been found
underneath both mounds (van der Linde and Fontijn 2011, 60-61; Bourgeois and
Fontijn 2011, 85).
The Echoput is a somewhat aberrant place in the local environment. It is
one of the highest places in this part of the Veluwe (95 m above Amsterdam
Ordnance Datum). The Veluwe exhibits an average yearly precipitation sum that
is considerably higher than in most parts of the Netherlands, since orographic
precipitation occurs on the elevated parts, like at the Echoput. The moist air is
forced to ascend where the landscape is elevated, causing the air to cool down,
form clouds and rain out. The local (loamy) soil conditions prevents the water
from draining off immediately, which makes the Echoput hill a rather wet place,
with pools of water forming regularly (see for a more detailed description Fontijn
2011a, 29-31). The surrounding area is covered with mixed forest (deciduous and
coniferous forest). The modern deciduous forest consists mainly of oak coppice
(Quercus sp.), with an undergrowth of blueberries (Vaccinium myrtillus) and
grasses, but also birches (Betula sp.) and beeches (Fagus sylvatica) are present. The
northern and central veluwe
97
0
ancestral heaths
pollen samples
profile pollen samples
restauration sand
natural sub-soil
charcoal
sod (vegetation side is darker)
feature
tree-trunk
soil on top of barrow
disturbance
5m
1.9
{
{
98
le
Figure 8.5a
NW
0
1m
Figure 8.5a-c. Profile sections of Echoput
barrow 1 (a-b) and barrow 2 (c) with the
location of the pollen samples indicated.
Figure 4a-b after van der Linde and
Fontijn 2011, figure 2.17 and 2.18
with changes; figure after Bourgeois
and Fontijn (2011, figure 3.8 A with
changes).
95.367 m NAP
ofi
pr
profile 1.9
o.s. 1
profile 1
sod 2 sod 1
o.s. 2
SE
95.367 m NAP
northern and central veluwe
99
0
5m
le
ofi
pr
0
1.1
NE
Figure 8.5b
sod 3 sod 4
o.s. 3
o.s. 4
profile 1.10
0
1m
SW
95.366 m NAP
95.366 m NAP
ofi
pr
100
le
5m
3.1
ancestral heaths
94.779 m NAP
NW
0
le
ofi
pr
2.1
profile 3.1
Figure 8.5c
profile 2
o.s. 3
profile 2.1
o.s. 2
o.s. 1
sod 3 sod 2 sod 1
0
1m
SE
94.821 m NAP
Sample location
Echoput1
Sample name
Profile 1.9
Soil profile series
1-19
20-35
Profile 1.9
Sod samples
Echoput1_sod1
Echoput1_sod2
Old surface samples
Echoput1_os1
Echoput1_os2
Profile 1.10
Ditch samples
Echoput1_ ditch
Sod samples
Echoput1_sod 3
Echoput1_sod 4
Old surface samples
Echput1_os 3
Echoput1_os 4
Echoput 2
Table 8.1. Overview of the
samples taken from the
Echoput barrows and their
surroundings. The samples
that have been analysed are
indicated by a shade. Those
with a darker shade did not
contain any or not enough
pollen. For the exact location
of the analysed samples, see
figure 8.4 and 8.5. os = old
surface underneath mound.
Level 10
Structure 17
Pit 1
Profile 2.1
Soil profile series
1-24
Profile 2.1
Soil profile series
25-29
Profile 2.1
Sod samples
Echoput2_ sod 1
Echoput2_ sod 2
Echoput2_ sod 3
Old surface samples
Echoput2_ os 1
Echoput2_ os 2
Echoput2_ os 3
Trench 9
Level 1
Post 10
Post 27
Trench 16
Level 1
Post 1
Trench 18
Level 1
Post 2
Trench 21
Level 1
Post 5
Post 12
coniferous forest consists mostly of pines (Pinus sp.), together with some Douglasfirs (Pseudotsuga menziesii) and Larches (Larix sp.). The barrows were located in
the forest, overgrown with trees and other vegetation, making them difficult to
spot. In 1999 both barrows were consolidated. The above ground parts of the trees
found on and around the barrows were removed, and the barrows were covered
with white sand to regain their presumed original shape.
8.1.2 Pollen sampling and analysis
During the excavation samples were collected for pollen analysis. For each mound,
individual samples were taken from different locations in and under the barrows
by Bakels and Achterkamp (University of Leiden, the Netherlands). From each
mound several samples were taken from the old surface underneath the mounds,
where the old surface was clearly visible. In addition several samples from the
top (e.g. the old surface) of different good recognizable sods of both mounds
were taken. The bottom of the ditch around mound 1 and the fill of a small pit
(structure 17) that was found underneath mound 1 were also sampled. Sampling
was done using methods described in Chapter 4. From these samples a selection
was made for analysis, based on the quality (colour and texture) of the soil. An
overview of the samples that were taken and analysed is shown in table 8.1. The
location of the analysed samples in the mounds is given in figure 8.5. In addition
samples were taken from the soil profile underneath both mounds. Samples were
collected as has been described in 4.1.3 over a length of 30 cm, containing the A
and most of the B horizon (see figure 8.5). In addition many samples were taken
from the fill of post features that were found in the surroundings of which four
northern and central veluwe
101
were analysed. These four post features belonged to four different structures (see
figure 8.4). A description of the sampling method and discussion can be found in
4.1.5. Chemical treatment and analysis of the samples took place as described in
4.2. For all pollen spectra a pollen sum of ∑AP–Betula (chapter 6) has been used,
except for the AP and NAP. These percentages have been based on a total pollen
sum. A minimum of 300 arboreal pollen grains (excluding Betula) per sample
have been counted by the author of the present work.
8.1.3 Results
For mound 1 four samples of the old surface, four sod samples, a sample taken
from the ditch (profile 1.10) and a sample from a small pit (level 10, structure
17) underneath mound 1 have been analysed (see table 8.1). Sample 2 from the
old surface did not contain enough pollen to count, as did the ditch sample and
the sample from the pit. The remaining samples contained sufficiently preserved
pollen. In addition the soil profile underneath mound 1 was sampled, from which
pollen could be obtained from 1 to 19 cm below the old surface. From mound 2
the three samples from the old surface and the three sod samples gave good results,
although pollen preservation was relatively poor. From the pollen series that was
taken the soil profile underneath mound 2 results could be obtained from 1 until
25 cm below the old surface. The samples derived from 25-29 cm were very poor
in pollen numbers. Below the results will be described.
Pollen from the old surface underneath the mounds and from the
sods
The pollen spectra from the two barrows show no clear differences and therefore
they will be discussed together. In addition, no differences could be noted between
the pollen spectra from the old surface and the sods of both mounds, so the
result description below counts for both the old surface and the sod spectra. The
percentage of non arboreal pollen (NAP) exceeds the percentage of arboreal pollen
(AP) in all samples (see figure 8.6). Especially heather (Calluna vulgaris) and less
but still in considerable amounts Poaceae (grasses) show high percentages. The
most abundant tree pollen types are Alnus (alder, 35-70%), Quercus (oak, 15-40%)
and Corylus (hazel, 15-25%). The presence of Carpinus (hornbeam) in some of the
spectra should be noted. Some Pinus (pine) pollen is present, but it is unlikely that
this tree was present in the surrounding forest. Pinus is not a common native tree
in the Netherlands in the time period after the Boreal but before the large scale
Pinus plantation starting in the 19th century AD (Janssen 1974, 57) and therefore
the Pinus pollen in the pollen spectra most likely came from long-distance. This
accounts for all pollen spectra that will be discussed. Anthropogenic indicators (cf.
Behre 1986) and grazing indicators (cf. Hjelle 1999) are present in all the samples.
One pollen grain of Secale (rye) was found in one of the sods (2) of mound 2.
Non-pollen palynomorphs were mostly represented by Sphagnum (peat moss) and
moss spores, but also algae like Debarya glyptosperma and Zygnema type 314 (van
Geel in: van Hoeve and Hendrikse 1998) are notable.
102
ancestral heaths
northern and central veluwe
103
Echoput2_os3
Echoput2_os2
Echoput2_os1
Echoput2_sod3
Echoput2_sod2
Echoput2_sod1
Echoput1_os4
Echoput1_os3
AP
20
40
60 80 100
20
s
nu
Al
40
Figure 8.6. Pollen spectra from the sod and
old surface samples taken from Echoput
barrow 1 and 2. Spectra are given in %
based on a tree pollen sum minus Betula
pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (= non
arboreal pollen) spores are included, non
pollen palynomorphs are excluded. Different
scales have been used, indicated with
different colours. MIA= Middle Iron Age,
LIA= Late Iron Age.
MIA/LIA
Echoput1_os2
Echoput1_os1
Echoput1_sod4
Echoput1_sod3
Echoput1_sod1
P
NA
Echoput
Barrow 1 and 2, old surface and sods
60
80
1
40
5
5
1
5
20 40
l ix
s
he
s
s inu era s rcu
u
x
u
e
d
g a
n
F a F r He Pi Qu
1
1
5
5
200
Upland herbs
100
a
lg
pe
vu
a
- ty
us lix ia tula llu n
b
R u Sa Til Be Ca
300
400
5
1
5
1
1
1
5
50
100
Ferns and aquatic mosses
20
ae
ae
or
a
or
at
l ifl e
ifl
u
ol
e
ul
b cea
p
g
ce
u
i
t ia
l
n
e
-ty
a
e
e
p l
d
a
a
m
-ty o e
ce
i u isi a ace op o e
m ag ea
in
l
r
ra
liu an t oac
c c rtem ste hen eca ste
a
a
l
V A A C S
G P P
A
Anthropogenic indicators
Algae
150
1
1
5
-ty
pe
Non pollen palynomorphs
o
t
pe ce
-ty . a
ris a/ R
c
s
a o
us et
u l ac
nc ex c isa
u
m
c
n
R a R u Su
lla
se
Grazing indicators
1
1
1
1
e
yp
1
1
1
1
5
1
1
1
1
1
p
-ty
um
1
e
1
1
1
1
20
1
5
5
20
5
1
5
1
1
5
1
1
5
20
952
987
938
220
734
338 1076
315 1259
318 1327
312 1016
328 1549
313
310 1165
362 1152
309
316 1069
309
es
s
or
re
sp
o
n
t
sp
er
a)
pe
arn te f e
ul
e
t
l ic is
-ty
r
fe
et
ns
ra
ge estr
e
ca lga or es
e
-B
ea
e
s
e
t
e
m
e
o
e
n
u
a
a
f
e
a
P
a
a
p
m
l
v
p
r
u
a lv
r
e
t
v
p
e
su
e ar pe
sil
(A
er v sp
ab
- ty ti ac iu o
-ty pe
ch sy
ys
n
pe p
ac
y
in
m
ar us e lu s i s sa yll plen eur eae ul a um
e v i um f ern um
ep
ric a
ty a-ty laria m-t
at
lle
u
t
t
t
9
1
3
e
a
m
h
a
c
c
a
i
o
s
s
3
8
d
r
5
d
a
3
b
m
a
c
e
e
n
en ry e t1 t1 t3 t5 t7 tA te
li c ris ce ga a op so u t ra n ri ne ea th ell gu liu A rz ol
n
ol p o te ag
lp
n c ba gn G G G G G G d e
ge th ia tra nn r y ry sc p e lip e pe sio i ac en un er ifo pe p e on
ll e ota
on ly ile h
An An Ap As Ca C a Ch Cu C y F i Hy Ja Lil M Pr Sp Tr Ty Hu M
Po
Co De Zy Bv Bv Bv Bv Bv Bv In
T
M Po Tr Sp
544 2103
20
us
in us
rp oryl
a
C C
ris
Heath
Pollen from the soil profile underneath mound 1
The zones described below are biostratigraphical units, based on palynological
changes in the diagram (see figure 8.7). This means they are not automatically
equivalent to geochronological zones.
Zone 1
A slight decrease in the arboreal pollen component from approximately 40 to 25%
can be seen. This is mainly due to a decrease in Tilia pollen, which starts at 20%
and decreases to around 2%. The forest cover in the surroundings of the Echoput
was dominated by Alnus. In addition the forest consisted mainly of Quercus and
Corylus. A high percentage of heather is present which even starts to expand further
at the end of this zone. Besides Calluna vulgaris grasses (Poaceae) were present in
considerable amounts as well as Polypodium vulgare (common polypody). Pollen
of anthropogenic and grazing indicators such as Artemisia (mugwort) and Plantago
lanceolata (ribwort plantain) are present in low amounts.
Zone 2
The expansion of Calluna vulgaris, which started in Zone 1, continues followed
by an expansion of Poaceae. The forest cover does not appear to be subject to
extreme changes in total, there is however an increase in Quercus and a decrease in
Corylus pollen percentage. In addition an increase in Fagus pollen is shown. The
anthropogenic and grazing indicators have expanded to some extent.
Zone 3
In zone 3 an increase in Tilia pollen percentage can be seen, together with a
decrease in Calluna vulgaris. Poaceae shows an increase as well as most other herbs
and ferns.
Pollen from the soil profile underneath mound 2
The zones described below are biostratigraphical zones, based on palynological
changes in the diagram (see figure 8.8). This means they are not automatically
equivalent to geochronological zones.
Zone 1
In this oldest part of the diagram, a decrease in AP can be seen, from 40% to
20%. The forest at the beginning of this period consisted mainly of Tilia (lime),
Quercus and Alnus. A decline of Tilia pollen is notable in this zone, as well as
the appearance of Fagus (beech) pollen. The percentage of Alnus pollen shows an
increase as well. Heather shows an expansion, as well as Poaceae. Anthropogenic
indicators, like Artemisia and Asteraceae tubuliflorae are present in low amounts,
grazing indicators like Poaceae, Asteraceae liguliflorae and Plantago lanceolata are
present in higher amounts.
Zone 2
In Zone 2 Tilia decreases further until almost no Tilia pollen is found. Corylus
shows an increase and the other tree species remain quite stable. Calluna vulgaris
fluctuates between 100 and 200%, Poaceae between 50 and 100%. Anthropogenic
and grazing indicators are present in higher amounts than in Zone 1. The
percentages of ferns and mosses have decreased, as well as Sphagnum.
104
ancestral heaths
Zone 3
Zone 3 shows a peak in Tilia pollen numbers and a decrease of Calluna vulgaris.
This is also shown in Zone 3 of Diagram 1. Zone 3 of Diagrams 1 and 2 is based
on the top samples taken from the soil profile and it is very well possible that
part of the sod above the old surface has been included in these samples. This
sod also contains a soil profile, similar to the soil profile underneath the barrow.
As a consequence it is likely that these samples do not represent the youngest
vegetation composition in this diagram, but older, comparable to part of Zone 2
in the diagram.
In all samples from both soil profiles particles of charcoal have been found.
Pollen from the post features
Trench 9
A very low percentage of arboreal pollen grains, 15-20%, can be seen (see figure
8.9). The absence of Tilia is notable in comparison to the pollen spectra obtained
from the barrows, as well as fairly high percentages of Fagus pollen and the
presence of Carpinus. The herb pollen types are dominated by Calluna vulgaris,
with percentages over 500%. Grasses show high percentages as well, around 70%.
Anthropogenic indicators are present in low amounts; however, the percentage of
Secale is relatively high.
Trench 16
This spectrum also shows a low percentage of arboreal pollen, around 15%.
Tilia is absent, Fagus and Carpinus are present in considerable amounts. Calluna
vulgaris is the dominating species, together with a high percentage of Poaceae. The
presence of Fagopyrum and Centaurea cyanus (cornflower) should be noted.
Trench 18
This spectrum is similar to the spectrum from Trench 16, except for a lower
percentage of Poaceae.
Trench 21
These spectra looks very much like the spectrum of Trenches 16 and 18 as well,
including the presence of Fagopyrum and Centaurea cyanus. Remarkable is the very
high percentage of Calluna vulgaris found in one of the spectra.
The size of the open space
The minimum size of the open space can be estimated by the amount of sods that
was used to build the barrows as has been explained in section 7.1. Knowing the
height and the diameter of the mounds and the thickness of the sods the minimum
size of the open area that was stripped can be calculated (see also table 8.2).
northern and central veluwe
105
106
ancestral heaths
18
19
13
14
15
10
11
1
2
3
4
5
6
7
8
9
AP
20
l
ta
To
40
60
l
ta
To
80 100
P
NA
Echoput
Profile barrow 1
20
s
nu
Al
40
60
Figure 8.7. Pollen diagram derived
from the series of samples taken from
underneath Echoput barrow 1. A
percentage diagram is shown, with %
based on a tree pollen sum minus Betula.
In the AP (= arboreal pollen) Betula
is included. In the total NAP (= non
arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours.
Depth (cm)
20
us
in lus
rp or y
C
1
Ca
40
20
5
5
20
s
us
s
in s rcu
gu
ax nu e
Fr Pi Qu
Fa
40
20
lix ilia
T
1
Sa
Trees and shrubs
40
20
100
200
300
400
5
Upland herbs
5
20
5
5
50
1
1
5
5
1
1
1
pe
1
1
150
1
20
5
20
40
60
1
5
20
80
Algae
20
1
5
5
1324
1333
1443
1031
302 1098
312 784
325 1095
320 1003
303 1087
316
333
311
311
Zone 1
)
a
la
es r m
tu
or sp e
Be
p
m
P
s o
su
(A
te p t
ila gly 4 u m
m
len
l
s
u
a
s
o
p y
1
gn
lp
te ar t3 en
ha
ta
i le eb G ll
Sp
Tr D Bv Po
To
298 1354
320 1539 Zone 3
331 1035
311 869
304 1344
303 1288
64 328 Zone 2
Ferns and aquatic mosses
100
pe
ty pe
is- ty
cr saa
to
s
lu ce
cu a a
un mex ccis
n
R a R u Su
e
ae
ra
or
ta
lifl ae liflo
la
u
u
b e
eo
t u iac e lig pe anc
e
l
d
a o
a
a
-ty o e
isi ce op ace
m ag ea
r
m ra
liu ant ac
te t e en t e
Ga Pl Po
Ar As Ch As
1
Grazing indicators
Anthropogenic indicators
es
s
or
re
sp
n
po
r
s
fe
pe um
rn
e e
ty
fe
-ty rag
at ar
ae
na ssif
ae pae
uc ulg
at
r
l
e
a
a
i
r
t
v
i
o
c
l
a fo
ae la ro e o n m
ps
ve m
e c e yl eu ea m iu e o s ti te
te iu
le o d
ea ca ph t a ac e ec ea s/ R la le
ac assi r yo scu per sion rt h sac bu pha ono ono ly p
i
Ap Br Ca Cu Cy Ja Na Ro Ru Ty M
M Po
1
na
us la
llu
m tu
Ca
Ul Be
is
ar
lg
vu
Heath
northern and central veluwe
107
Depth (cm)
20
AP
40
60
P
NA
80 100
l
ta
To
20 40
s
nu
Al
60 80
5
1
20 40
us s
la in l u
tu rp ry
Be Ca Co
Figure 8.8. Pollen diagram derived
from the series of samples taken from
underneath Echoput barrow 2. A
percentage diagram is shown, with %
based on a tree pollen sum minus Betula.
In the AP (= arboreal pollen) Betula
is included. In the total NAP (= non
arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
l
ta
To
Echoput
Profile barrow 2
20
1
20
20
1
20
s
cu
er
lix ia
Qu
Sa Til
40
Ca
vu
is
pe
Other herbs
100 200 300
na
llu
ar
lg
Anthropogenic indicators
1
20
1
5
1
20
1
1
1
5
1
5
1
1
t
a-
1
e
yp
5
1
-ty
m
tu
1
1
1
20 40
5
20 40
20
50
100
150
60
1
1
1
Algae
20
pe
ty pe
is- ty
cr saa
o
pe
lus c et
-ty
cu x a sa
m
n
u
i
e
c
n u m cc
iti
R a R u Su
Tr
Grazing indicators
NPP
a
a)
m
es
ul
er
or
et
sp
-B
sp
to
um
P
e
p
A
t
y
ns
ila 4
m
gl
m
le
l
s
u
4
9
u
a
o
gn
e p t31 ry t2 t28 t12 t22 A ns
lp
h a r ilet vG eba vG vG vG vG vg t oll e ota
Sp
T
T B D B B B B B P
319 1108
343 971 Zone 3
315 1037
322 1052
306 1227
320 1250
304 1154
304 1099
316 1216
402 1259
306 1090 Zone 2
219 891
305 943
299 1291
186 738
80 316
34 218
95 451
168 583
0
0
37 153 Zone 1
56 206
179 544
281 680
80 20 5 1 1 5 5 1 1 5
Ferns and aquatic mosses
5
ae
ae
or
ta
or
la
ifl
li fl
ul
eo
ae
bi
g
e
u
i
l
t
e lanc
a
i
p
e
e
d
p
a
a
a
-ty go
e
ce
e s op o le
isi ce
ea
al
ra
m ra
ium ta
re hen eca ste
te te
ac
al Pl an
e
o
S
Ar As
G
P
A
C C
Heath
s
re
s
po
re
e s re
e
po
t
s
s
a
a
a
n
r
e
c g
ae m e
ve
at
rfo
r u ul
ar pe
c e iu op e
sil
er v
pe ype
ae
ia ty
lla en ur
e v di um
ep
e ac e bi s hy spl ta e c ea cum l a-t ae lar mt
t
a
i
a
p so u
ra ri el ce gu liu A rz ole
ole p o
ce ic a
ia as s ann r yo hry usc p e y pe un sa er ifo p e up e on
on ly
M Po
Ap Br C C a C C C y H Pr Ro Sp Tr ty H M
1
us
s
s
in a
gu
ax ce nu
Fa
F r Pi Pi
Trees and shrubs
108
ancestral heaths
80 100
AP
lN
ta
To
20 40 60
P
20 40 60
s
nu
Al
5
20
20
20
us
in l us us
a
rp ry
g
ce
Pi
Fa
Ca Co
Figure 8.9. Pollen spectra derived from
the post hole fill samples. A percentage
diagram is shown, with % based on a tree
pollen sum minus Betula. In the AP (=
arboreal pollen) Betula is included. In the
total NAP (= non arboreal pollen) spores
are included, non pollen palynomorphs
are excluded. Different scales have been
used, indicated with different colours.
T21F12
T21F5
T18F2
T16F1
T9F27
T9F10
lA
ta
To
Echoput
Post fills
20
20 40 60
s
cu
s
er
nu
Pi
Qu
5
1
1
1
1
20
is
ar
lg
vu
Heath
Anthropogenic indicators
5
1
5
5
20
20 5
Upland herbs
20 40
60
5
5
5
Ferns and mosses
80 100
Algae
5
e
ta yp e
l a s-t typ
e o acri sac
o
n
l a lus ce t
go cu a
ta un mex ci sa
n
n
c
a
Pl Ra Ru Su
Grazing indicators
NPP
1
1
1
1
5
1
1
pe
1
1
1
1
5
1
5
1
5
5
5
5
20
1
5
1
2039 Trench 18
310
618
3980
Trench 21
3137 Trench 16
365
322 10700
3018
320
Trench 9
pe
-ty
es
a)
m
or
ul
ty
e
sp are
a
et
e ens
r
e
B
e ea
g
o
t
p
l
e
f
a
e
v
r
P
p
e p
um
i la vu
e yp
-ty r
(A
pe
ns
-ty l ac ro e
e m aa
ps m
yp -t
a
le
m
e l us yl e u ea ia um l a -t us e yp i u a ri
te iu um 4
ol
5 A 2 26 nsu
ea aga o ph uta rac o rb ric ntil e ll a an t cea us-t gan gul sp A o le pod gn t31
p
5
1
c
t
t
t
l
ia str
G G G l le ota
ry s c pe p h ype te un in sa b a r e r ola pe o n ly ha G
Bv Bv Bv Po
T
Ap A Ca Cu Cy Eu H Po Pr Rh Ro Ru Sp Sp Vi ty M Po Sp Bv
333 2485
-ty
5
ae
ae
or
or
i fl
l ifl us
ul
bi an
ae
pe
g
e
u
i
y
y
t c
l
a
e
-t
p di m
ae a
ae -typ e
m
iu is ia ce ure le s po yru
ce
ra li um cea
cin tem tera nt a rea eno gop ca le
e
c
a
t
Ga Po
Va Ar As Ce Ce Ch Fa Se
As
1000 2000 3000 4000 10 5
us
pe
uc
ty s a
na
i x mb il ia le x- lmu etul a ll u
l
C
Sa Sa T U U B
a
gr
ni
Trees and shrubs
98 m
N
NAP
91 m
00
old surface
underneath
Old surface
underneath barrow
barrow
meters
50 m
25
50
area used
cutting
Area for
usedsod
for sod
cutting
Figure 8.10. View of the
Echoput hill with the two
excavated barrows, based on
digital elevation model of the
AHN (copyright www.ahn.
nl), with around each barrow
an indication of the area that
had to be used for sod cutting.
The measurements of the barrows are (van der Linde and Fontijn 2011, 33;
Bourgeois and Fontijn 2011, 65):
Barrow 1: r=9.5 m (d=19 m), h=1.08 m
Barrow 2: r=7.25 m (d=14.5 m), h=1.0 m
Sods: average h=0.25 m
The calculated area to be stripped for Mound 2 is 332 m2. For Mound 1 a
correction should be made, because this barrow was not completely spherical, but
had a flattened top. Taking this into account, the stripped area for Mound 1 was
902 m2. A total area of 1234 m2 was used for sod cutting (see figure 8.10).
The size of the open space can also be estimated by the percentage of arboreal
pollen as has been described and discussed in section 7.2. The arboreal pollen
percentage of the Echoput barrows is on average only 29%. This implies an open
space with an average distance to the forest (ADF) of approximately 300 m.
8.1.4 Discussion
Dating the barrows
The first thing to point to in the palynological results is the resemblance between
the two barrows. Pollen spectra from the old surfaces indicate a similar vegetation
pattern at the time the barrows were built, which makes it likely that they were
built in the same period. This is in line with what was expected on the basis of the
14
C-datings and the general similarities between the mounds. The occurrence of
Carpinus suggests that this period can be placed in the Iron Age (Janssen 1974).
Both their contemporaneity as well as their Iron Age dating are in agreement with
the excavation results, on the basis of which the dating could be further specified
to the late Middle or earlier Late Iron Age (van der Linde and Fontijn 2011, 62;
Bourgeois and Fontijn 2011, 87).
northern and central veluwe
109
The barrow landscape
The similarity of the pollen composition of the old surface and the sods indicates
that the sods were cut in the close surroundings of the barrows, where vegetation
composition was similar to the spot where the barrows were built. The following
discussion about the barrow landscape is based on the results of the samples of both
the old surface and the sods of the two mounds, which represent the vegetation
composition at the time just before the barrows were built.
Figure 8.6 shows the pollen spectra of the mentioned samples. They indicate
that herbs are much more abundant than trees. Especially heather (e.g. Calluna
vulgaris) and less, but still in considerable amounts, grasses (e.g. Poaceae) dominate
the herb species. Heather pollen tends not to spread outside the heathland where
the pollen is produced (de Kort 2002). This implies that the Echoput barrows were
built in an open spot, where heather was the most dominant species. Non-pollen
palynomorphs such as Debarya glyptosperma and Zygnema type 314 (van Hoeve
and Hendrikse 1998) suggest the presence of some water at the site, at least part
of the year, conditions which nowadays still exist (the pools of water that remain
after rain for some time). Anthropogenic indicators are present amongst the herbal
pollen. These are dominated by Plantago lanceolata and Asteraceae tubuliflorae.
Remarkable is the find of one pollen grain of Secale in the pollen spectrum from
sod 2 of Mound 2. This cereal species (rye) had not been commonly introduced
in the Netherlands during the Iron Age yet, however, some early Iron Age finds in
northern and western Europe have been reported (van Zeist 1976, Behre 1992).
The anthropogenic indicators suggest the presence of human activity at the site,
which is consistent with the find of pottery sherds and flint fragments in the
sods and the old surface (van der Linde and Fontijn 2011, 59-60; Bourgeois and
Fontijn 2011, 87). However, the pollen percentages of anthropogenic indicators
are too low to conclude the site was a settlement area or with (former) arable
fields nearby. This is consistent with the data from the excavations in the close
surroundings of the barrows (Valentijn and Fontijn 2011).
The tree pollen that is present in the pollen spectra is mainly Alnus, Quercus
and Corylus. Alnus is likely to have grown on the lower sites in the surroundings
of the heathland, where hydromorphic soils occurred like Gleyic Podzols, Umbric
and Histic Gleysols. This indicates that alder carr was probably present in the
stream valleys in the surroundings of the Echoput hill. The dominance of Alnus
pollen within the total arboreal pollen content could imply an open landscape
where the alder pollen was free to travel in from out of the alder carr, since no other
sizeable forest blocked their way. In addition, Alnus blooms before Quercus and
Corylus get their leaves, making it easier for Alnus pollen to travel freely. Corylus
is a tree that requires light conditions to grow; it will not be able to survive in the
reduced light conditions in a closed forest. The tree requires moist soil, but not
wet conditions. It is very likely that Corylus grew on the slopes around the Echoput
hill, together with Quercus, a tree that has also has a preference for soil that is not
very wet (Weeda et al. 1985, 113). The presence of alder carr in the valleys and
the more open vegetation in the surroundings of the barrows indicates that forest
clearing had only taken place in the higher and drier places around the Echoput
hill. The forest was not cleared recently before the barrows were built, indicated
by the presence and the diversity of the herb vegetation. The herb vegetation
had already had some time to establish and to develop and the open place must
have existed some time before the mounds were constructed. Heath vegetation is
not a natural vegetation type in the Netherlands (with exception of the coastal
area). This implies that the barrow landscape was already managed to maintain
the heathland. The amount of grasses (Poaceae) together with Plantago lanceolata,
110
ancestral heaths
Asteraceae liguliflorae, Succisa ( and Galium-type could be an indication that the
heathland was kept open by grazing (Hjelle 1999) and as such was part of the
economic zone of settlements.
The size of the open space
It has already been mentioned that the barrows were built at the same time, or one
relatively quickly after the other. The similarity of the pollen spectra from the old
surface and the sods indicates that the sods were taken in the near surroundings
of the place where the barrows were built. In addition, the similarities between
pollen from sods and the old surface underneath the mound and in lithology of
sods and the Echoput hilltop all imply that the sods were cut from the Echoput
hilltop and not from the hill flanks. Regeneration of heath after sod-cutting
takes a period of 5-40 years, depending on the thickness of the sods. Thin sods,
preferably containing only the F horizon of the soil, were traditionally used as
fuel or as bedding in stables. Regeneration after cutting thin sods takes only 5-8
years (Pape 1970). When thicker sods were cut, containing the A- and E-horizon,
regeneration takes up to 40 years. Such sods were for example used as construction
material (Stoutjesdijk 1953, cf. Bakels and Achterkamp 2013). Assuming that the
period between the construction of the first and the second burial mound had
been too short for the heath vegetation to regenerate the open place had to be
large enough to cut sods for building two barrows. The soil profile shows that the
surface beneath both barrows was not used for sod cutting (Fontijn 2011b, 154),
which also implies that the barrows were built at the same time or that at least
part of the area had already been kept free from sod-cutting as a reservation for
the construction of the second burial mound.
As has been shown in the results, the area to be stripped for Barrow 1 is 902
m2 and for barrow 2 332 m2, so a total area of 1234 m2 was used for sod-cutting.
This implies that a minimum area of 1683 m2, the surface beneath the barrows
included, consisted of open vegetation. Based on the arboreal percentage the open
space had an ADF of approximately 300 metre, implying an open area of about 28
ha (πr2 = π . 3002 = 282743 m2). Although this size could have been overestimated
(see section 7.3, according to palynological modelling the ADF was probably
about 200 m) the open space is considerably larger than based on the amount
of sods that had been used to construct the burial mounds. The combination
of these two methods builds an image of how the burial mounds were situated
in the landscape. The barrows were located in an area that was dominated by a
heath and grass vegetation. Trees could probably not be found in the first 200 to
300 metre around the mounds. The barrows, already located on a relatively high
place in the environment, were probably even more prominent in the landscape,
knowing that the direct surroundings were cleared from both vegetation and the
topsoil, creating a bare environment (see figure 8.10). This will have increased
their visibility in the surrounding landscape.
The pre-barrow landscape
Based on the theory presented in Chapter 5, the pollen diagrams derived from
the soil under Barrows 1 and 2 represent the vegetation development of a certain
period before the barrows were built. Since the soil profiles have not been dated
the duration of the period represented is not clear (see Chapter 5). The pollen
diagrams show that heath was already present at the place where later on the
barrows were built since at least the time span that is represented by the diagrams.
The presence of an Umbric Podzol (Dutch classification: Moderpodzol) suggests
that heath vegetation could not have been present for a very long time, since
northern and central veluwe
111
underneath heath vegetation a Carbic Podzol (Dutch classification: Humuspodzol)
soil would develop. This could however take several centuries (Andersen 1979).
During the oldest zone represented in the diagram the AP is higher than at the time
the barrows were built, 40% compared to 20%. The forest was mainly dominated
by Tilia and Quercus at the drier sites and Alnus at the wetter sites. Despite the
low pollen counts in some of the lower samples of diagram 2 clear trends can be
seen in both diagrams. A decline of Tilia is indicated by decreasing Tilia pollen
percentages. Such developments in forest cover is presumed to have taken place
generally in the Netherlands as has been shown by several pollen analyses of lake
and peat sediments (Janssen 1974, van Geel 1978). At the same time an increase
of Fagus is visible in the diagram, comparable to the general increase of Fagus in
several parts of the Netherlands, since its arrival between ca. 3700 cal BC and ca.
500 cal BC (Fanta 1995). An increase of Alnus pollen that can be noticed might be
primarily related to the decrease of Tilia or could indicate an expansion of the wet
forest. The decrease of forest cover seems to go hand in hand with an expansion
of the heath vegetation.
At the time the barrows were built vegetation was dominated by heather, at
least locally. It is not entirely clear how the open place was created nor what
it was used for in the period before the barrows were built. Indications of the
presence of human activities at the site in several periods before the barrows were
built are evidenced by finds from below and beyond the mounds, although they
certainly do not indicate a very intensive use of this site in the Bronze Age or early
Iron Age (Louwen et al. 2011, 141). The absence of cereal pollen grains and low
amounts of arable weeds like Artemisia vulgaris in the diagram demonstrate that
the location had not been used for crop cultivation. The size of the heathland can
be estimated. Based on the ratio of arboreal pollen versus non arboreal pollen, the
size of the open space is estimated to have been from approximately 200 metre
ADF to approximately 300 metre ADF at the moment the barrows were built. To
maintain the heath, the landscape must have been managed. Methods of heath
management can involve sod-cutting, grazing, mowing and burning (Stortelder
et al. 1996, 287).
Sods were cut in the area, at least with the purpose of building barrows.
With sod cutting the soil is stripped from all vegetation. For heath to recover
it is dependent on re-establishment by seeds that were present in the deeper soil
layers or by expansion of surrounding heath vegetation. Recovering of the heath
vegetation after sod cutting will take 5-40 years, depending on the thickness of
the sods that were removed (see above). The area needed for building the barrows
was most likely much smaller than the total heath area in which the barrows were
built (see above, r=200 to 300 m). Consequently, sod cutting for the purpose of
building the barrows would not be sufficient to maintain the entire heath area.
Large scale sod cutting in heathland areas is mainly known from the Medieval
Period into the 19th century, when the sods were laid in stables to catch animal
dung and subsequently were used on arable fields as fertilizer. Small scale practise
of this way of farming may have taken place at the time the Echoput barrows were
built. There are however no indications of such arable fields in the environment.
In addition, manual sod cutting is quite labour-intensive and it is not likely that
this heath area was managed by sod cutting alone.
The amount of grasses (Poaceae) together with Plantago lanceolata, Asteraceae
liguliflorae, Succisa and Galium type could be an indication that the heathland
has been grazed (Hjelle 1999). Mowing and grazing are comparable since they
both keep the plants down. Grazing is more selective than mowing, with animals
having a preference for certain species. Sheep prefer young Calluna heath and
grass and herb vegetation in between the heath vegetation. They are not very
112
ancestral heaths
fond of older Calluna plants (Elbersen et al. 2003). Cattle eat mainly grasses,
although some landraces also eat young Calluna plants (cf. Lake et al. 2001, 31).
Archaeozoological evidence from several excavations suggests that prehistoric
farming communities kept mainly sheep and cattle (Brinkkemper and van
Wijngaarden-Bakker 2005, 493). Both sheep and cows are used in present times
to maintain heathland areas by grazing. Historical data show that in Medieval
Period grazing using only sheep was sufficient to maintain heathland vegetation.
A stocking rate of 1 sheep/ha is assumed (Piek 2000). Also in present heathlands
several studies mention that an average of 1 sheep/ha/yr should be sufficient to
manage the heathland (Elbersen et al. 2003, Verbeek et al. 2006). The size of the
stocking rate of cattle in the past is not clear, although it is clear that cattle grazing
in Dutch heathlands occurred on large scale before the 18th century (Bieleman
1987). Bokdam en Gleichman investigated the influence of grazing cattle on the
development of Calluna heath (Bokdam and Gleichman 2000). A stocking rate
of 0.2 livestock unit per hectare per year appeared not to be adequate against
invasion by grasses and tree growth. Natuurmonumenten, a Dutch organization
that protects and manages nature reserves in the Netherlands, has over 30 years
of experience with grazing in heathland areas. They experienced that in dry
heathland areas 1 head of cattle per 5-6 ha is sufficient to prevent grasses from
getting dominant in heathland areas (Siebel and Piek 2001). This is however in
the present environmental circumstances with higher deposition of nutrients,
and it is likely that in the past less cattle would have been adequate enough for
maintaining heathland vegetation. When an indication of the minimum size of
livestock from a prehistoric farming community should be calculated that was
responsible for managing the heathland area where the barrows are being built
in, an average of 1 sheep per hectare and/or 1 head of cattle per 6 hectare will be
used. At the Echoput, based on the ratio of arboreal versus non arboreal pollen
grains the area that was covered with heath vegetation at the time the barrows
were built is estimated to have been 28 hectare (π . 3002), implying a livestock
size of approximately 28 sheep and/or 4-5 head of cattle. Mowing can be seen as
a kind of grazing, although grazing is more selective.
Regular burning is also a traditional way of heath management. When the
heath is being burnt every 10-20 years the heath vegetation can be maintained
by rejuvenating the heath (Mallik and FitzPatrick 1996, Yallop et al. 2006). A
combination of burning and grazing is nowadays often applied, which seems to
be very effective. Small scale burning provides young vegetation, which is more
nutritious to the grazing stock. The remains of charcoal found in all the pollen
samples from the Echoput barrows may be an indication that human burnt the
heath vegetation. Particles of charcoal have been found elsewhere as well during
excavations of barrows and in soil samples that were taken for palynological
analyses (Karg 2008). A combination of grazing and burning and perhaps some
sod cutting seems a plausible explanation of how the heath was managed at the
Echoput.
Posts at the barrow site
The pollen spectra from the four possible structures that have been sampled have
a different composition than the barrow spectra. As was discussed in section 4.1.5
the posts might be dated based on their pollen spectra. The pollen spectra from the
four posthole structures (see figure 8.4) show a vegetation composition that can be
dated to a much younger period than the period the barrows were mainly built in.
This is implied by the presence of Secale, which is known as a common crop in the
Netherlands only after being introduced during the Roman Period (Behre 1992,
northern and central veluwe
113
RADAR 2006). In addition, the relatively high percentages of Carpinus and Fagus
indicate a rather young pollen composition. Both species show an increase during
the Holocene vegetation development in the Netherlands since the Subatlantic
period up to the Medieval Period (Janssen 1974). In addition, all posthole fillings,
with exception of the postholes from Trench 9, contained pollen from Fagopyrum
and Centaurea cyanus, which are only present in the Dutch pollen spectra from
the Late Medieval period (Bakels 2000, 2012). However, can the dating of the
pollen spectra be coupled to the dating of the posts? In other words, can the posts
also be dated in the Late Medieval Period? As has been discussed in section 4.1.5,
the Medieval pollen could have come from the vegetation that was present at the
Echoput hill at the time the posts were placed or they could have infiltrated in the
soil from some time before the posts were placed. The posts could then be dated
in the Late Medieval Period or later (as a terminus post quem date). The pollen
spectrum from the posthole from Trench 9 lacked pollen that indicates the Late
Medieval Period and consequently the Roman Period can be assessed as a terminus
post quem date for this posthole filling.
What did the landscape look like at the time the posts were placed? The
posthole fill pollen spectra indicate a landscape that was more open than during
the time the barrows were built. The amount of Alnus had decreased. This implies
deforestation of the lower sites as well, or a change in soil water content. The
barrow site was at this time an open spot as well, but the character of the place had
slightly changed compared to the barrow landscape. Calluna had expanded at the
cost of the forest. The diversity and quantity of other herbs increased. At Trench
21 a very high percentage of Calluna pollen can be seen, which is not visible in
any of the other samples. This could indicate a local abundance of heather, for
example the covering of the roof of the structure could have been made of it.
8.1.5 In conclusion: the history of the Echoput barrow landscape
It is generally assumed that most barrows were built in open spaces in a forest area.
However, the origin of these open spaces is little known. The pollen analyses of
two barrows at the Echoput show the vegetation history of the open space from a
period before the barrows were built. This showed that the clearing in the forest
was indeed much older than the barrow, as has been suggested in section 2.3.
When and how the open space was created is not known.
From the beginning of the period that our data represent, the open spot was
mainly covered by heath vegetation mixed with grasses and several other herbs.
The open space, surrounded by a forest of Tilia and Quercus, had been used for
at least a few centuries by prehistoric man. This is indicated by several features
dating to the Middle Bronze Age period. The presence of anthropogenic indicators
confirms the influence of prehistoric man in the environment. Mesolithic and
Bell Beaker features were also present (Louwen et al. 2011), though it is not
known if the forest was already cleared by then. Although we did not uncover
any evidence for a settlement near the mounds, it is clear that the area has been
used by prehistoric man. However, what did they use the open place for since
the Bronze Age? It is very likely that it was included in the economic zone of
farming communities as grazing grounds, keeping the vegetation open. Based on
the high percentage of pollen from Poaceae, in combination with the presence of
Plantago lanceolata, Asteraceae liguliflorae, Succisa and Galium type, the use of
this open spot as pasture is very plausible (following Hjelle 1999). Furthermore,
regular burning of heath could have occurred, indicating that a form of heath
management was used to keep the area open. The use of fire is indicated by the
amounts of charcoal found in the pollen records.
114
ancestral heaths
Before the barrows were built the open area seems to have been used solely
as a place for the living, since no indications have been found that people were
buried there. This changed when the burial mounds were constructed in the later
Middle Iron Age or early Late Iron Age. At this time the vegetation surrounding
the Echoput hill had changed. The Tilia dominated forest had shrunk and forest
with a more open character mainly consisting of Quercus and Corylus had taken
its place. The heath vegetation at the open place at the top of the Echoput hill had
expanded. This change in vegetation was probably due to human activities, such
as burning and cattle grazing. The upper surface of a large part of the heathland
at the Echoput hill was stripped in order to get sods for the construction of
the barrows. The surface where the barrows were going to be located was left
untouched. Whether the barrows were built at exactly the same time or with a
short period in between does not change the fact that both places had already
been designated as barrow location, based on the observation that the surface
underneath both barrows were not used for sod-cutting. The two barrows must
have been quite pronounced features in the landscape; placed on one of the highest
locations in the area, cleared from surrounding vegetation. It is unknown whether
the surrounding landscape was kept open after the barrows were built. However,
one of the mounds had been re-used as a burial location (van der Linde and
Fontijn 2011, 64). In addition, during the Roman Period and the Late Medieval
period (based on palynological dating of the post hole fillings) there was a very
large open spot covered with heath vegetation. It is likely that the place had been
kept open all this time.
8.2 Niersen-Vaassen
In the north-eastern part of the Veluwe several barrow alignments are situated.
Several of these barrows were excavated over a series of campaigns. An extensive
description and analysis of the barrow alignments have been made by Bourgeois
(2013). Barrows not part of alignments are also present in this area. Dating and
palynological data are available for five barrows in the area, of which two were part
of a larger alignment. In addition palynological data are available from samples
taken from a Celtic field present in the same area (see figure 8.11). Combining
these data makes it possible to reconstruct the vegetation development in this area
from the Neolithic until the Iron Age.
8.2.1 Site description and sample locations
Niersen, barrow 4 and 6
The two investigated barrows of Niersen form part of a 6 km long alignment
containing at least 46 barrows (Bourgeois 2013). The original excavation of Niersen
4 and 6 took place in 1907 by Holwerda (Holwerda 1908). Holwerda described
Niersen 4 as a Bell Beaker tumulus with a height of 1.65 m and a diameter of 36
m. He noticed that this barrow was situated approximately 2.25 m higher than
the other barrows in this area. In the barrow a grave was found in which skeletal
remains of more than one individual were present. Holwerda decided to take
out the entire grave-area after plastering to be able to examine the remains later.
This plaster box has recently been rediscovered in the collection of the National
Museum of Antiquities in Leiden and has been subject of research by the museum
in cooperation with the University of Leiden (Bourgeois et al. 2009). They dated
the grave, on the basis of stylistic parallels, to the late Neolithic period (2600-2200
cal BC). Samples for pollen analysis were taken from the sediment in between
northern and central veluwe
115
Niersen 4
Niersen 6
Celtic Field
Vaassen III
Vaassen IIVaassen I
Vaassen 1
Vaassen 3
Vaassen 2
0
250
500
1000 m
Sampled barrow
Other barrows
Celtic Field
m NAP
100 m
0m
the skeletal remains, but unfortunately pollen could not be obtained from these
samples. In 1984 the Niersen 4 barrow was consolidated by the ROB (presently
known as the Cultural Heritage Agency of the Netherlands). They described the
tumulus as a bank-and-ditch barrow with a diameter of 28 m. Niersen 6 was a
barrow with a height of 1.50 m and a diameter of 16 m. The tumulus probably
dates to the early Bronze Age (Bourgeois 2013). The ROB report corrects the size
of the barrow to a diameter of 19 m. During the conservation carried out on the
barrow, pollen samples were taken from the old surface underneath both mounds
and from the mounds themselves by Groenman-van Waateringe10. One sample of
the old surface per mound was prepared and analysed by the author. Methods of
sample preparation have been described in Chapter 4.
10
Due to poor documentation it is not completely certain that during reconstruction of the barrows
by the ROB the barrows were identified correctly as barrow 4 and 6 (Bourgeois et al. 2009). Samples
for pollen analysis were taken during this reconstruction.
116
ancestral heaths
Figure 8.11. Detailed map of
the Niersen-Vaassen area with
the locations of the barrows
of Vaassen and Niersen and
the Celtic Field of Vaassen.
The map is based on digital
elevation model of the AHN
(copyright www.ahn.nl).
Figure 8.12. Sample locations
in the sections Vaassen I, II
and III at the Celtic Field at
Vaassen. Figure redrawn after
Brongers (1976), plate 13.
Vaassen, barrows 1-3
Three barrows at Vaassen were excavated by Bursch and Tromp in 1941; reexcavation took place in 1970-1971 by Lanting and van der Waals (1971). During
that last excavation samples were taken and analysed for pollen by Casparie and
Groenman- van Waateringe (1980, 28, 35).
Vaassen 1 (V1) is a single period barrow radiocarbon dated to 2850-2600 cal
BC (Bourgeois 2013, 53). Underneath V1 a sherd of a PF beaker and some flint
was found. The original dimensions of the barrow were probably a diameter of 13
meter with a height of approximately 1 metre. Samples for pollen analysis were
taken from the old surface. Vaassen 2 (V2) is a two-period barrow of which the
first period can be dated to the Bell Beaker Period based on the find of a Veluvian
Bell Beaker (Lanting and van der Waals 1971a). The second period is dated to the
Middle Bronze Age. The primary barrow was approximately 8 m in diameter and
approximately 30 cm high. For the secondary period the barrow was expanded to
a diameter of approximately 15 m and a height of 1.40 m. The thickness of the
sods used for the second period is approximately 25 cm. Samples were taken from
the old surface of the primary mound and from sods belonging to the second
period. Vaassen 3 (V3) is also a two-period barrow of which the first period has
been radiocarbon dated to 2885-2625 cal BC (Bourgeois 2013, 53). The second
period has been dated to the Bell Beaker period. The diameter of the barrow is not
known; its surrounding feature measured approximately 7.5 m across. The height
northern and central veluwe
117
of the barrow was about 0.3 m. Samples for pollen analysis have been taken from
the first period from the intermediate ditch and the outermost palisaded ditch
(Casparie and Groenman-van Waateringe 1980, 28).
Vaassen, Celtic Field
In the woods west of Vaassen (municipality of Epe), a Celtic field is situated on
a 15 hectare heathland and continuing over a surface of almost 100ha. Three
parts of the Celtic field were excavated: Vaassen I, II and III (Brongers 1976).
Vaassen I was situated at the south boundary of the Celtic field, Vaassen II was
situated west of Vaassen I and Vaassen III could be found at the east side of the
heathland (see figure 8.12). Sections were made at these locations, revealing a
sequence of several layers. These layers, an old surface and three agricultural layers,
represent several phases. Local agricultural activities started on an old surface that
became partly denuded. Part of the A-horizon of the podzol belonging to this
old surface was homogenized and changed into an arable layer. The remaining
part of the A-horizon of this podzol is called the denuded old surface (DOS). The
arable layer, which does not belong to the bank system of the Celtic field (CF),
is called the pre-Celtic field (PCF) layer. On top of the PCF-layer a bank system
was constructed, forming a Celtic Field. At Vaassen III an older arable layer was
present (OAL) on top of the DOS and underneath the PCF layer. This OAL layer
was not present at the other two locations. Underneath the banks the DOS and/or
OAL, PCF and CF layers were clearly visible. In between the banks the DOS was
seriously disturbed and the PCF and CF layer could not be differentiated from
each other. Soil samples were taken for pollen analysis from all layers (see figure
8.12). At Vaassen I samples were taken from or underneath a bank. Three samples
were taken from the DOS, one sample from the PCF and one sample from the
CF. At Vaassen II a sample was taken from the DOS, underneath a PCF layer
that was covered by a bank. At Vaassen III two samples were taken. One sample
was derived from the OAL layer that was overlain by the PCF layer. The second
sample was taken from the CF layer in the bank that covered this PCF layer. The
samples were analysed by Casparie. The pollen data that were published in 1976
(Casparie 1976) were re-used in this research, in addition to the barrow data in
the Epe area.
Dating the Celtic Field
Several locations in the Celtic field and the layers underneath were sampled for
14
C. Remains of a farmhouse (Haps type) were discovered at Vaassen I. The house
plan was covered by the CF-layer and possibly also the PCF layer (CF and PCF
could not be differentiated here). The house was dated by fragments of charcoal
to 2420 ± 65 BP (GrN-5498; 671-396 cal BC, calibrated with Oxcal 4.2), dating
the part of the Celtic field that was situated on top of the house to 671-396 cal
BC terminus post quem. Such farming houses were often found associated with
Celtic fields and it well is possible that part of the Celtic field had already been
developed when the farmhouse was still in use (Brongers 1976). The DOS is
difficult to date and the dating of the samples from the DOS depends on the
depth at which they are taken. At Vaassen I the dating of the house plan can be
interpreted as a terminus ante quem date for the DOS layer (e.g. 671-396 cal BC);
samples were taken approximately 25m west of the house. The pollen spectra,
which will be discussed in more detail below, show the presence of Fagus and
Carpinus. Carpinus appears in the Netherlands around 1500 cal BC and both
118
ancestral heaths
northern and central veluwe
119
CF8
CF5
PCF 7
OAL 4
DOS 6
DOS 2
DOS 3
DOS 1
Vaassen2 sod1_per2
Vaassen2 sod2_per2
Niersen6_os
Vaassen2_os_per1
Vaassen2_sod_per1
Niersen4_os
Vaassen3_outert
Vaassen3_intert1
Vaassen3_intert2
Vaassen1_os1
Vaassen1_os2
20 40 60 80 1 5
s
e r nu
Ac Al
20 40 60 80 100 1
Figure 8.13. Pollen spectra from samples
taken from the barrows at Vaassen, the
barrows at Niersen barrows and the
Celtic Field at Vaassen. Spectra are given
in % based on a tree pollen sum minus
Betula pollen. In the total AP (=arboreal
pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are
included, non pollen palynomorphs are
excluded. Different scales have been used,
indicated with different colours.
2800-2600 BC
2500-2200 BC
2000-18000 BC
1600-1400 BC
1000-400BC
800-150 BC
AP
P
NA
1
1
1
1 5
20 40 5
5
20 5
20 40 60 80
e
yp
s-t
s a
ix rbu lia lmu t ul
l
Sa So T i
U Be
e
ra
fl o
Anthropogenic indicators
1 5
5
20
20
20 1
20 1
20
Aquatic herbs
1
20
.a
/R
t
ce
a
ell
os
-ty
pe
Grazing indicators
sa
eto
ac a
x
is
e
m
cc
Ru Su
Ferns and mosses
100 200
ae
ta
or
la
lifl
uli
e
a e ia
u
eo
b
e
p
g
d
c
u
li pe anc
t
e
a
-ty
i
e
m
l
ae
m
od o
ea -ty o a e
a
iu sia e
op tag a ra c um tag
in mi ra c a li
e
ica acc rte st e ere hen lan rtic st e al i lan oac
r
P
G
A
P
V
U
E
C
A
C
A
P
Upland herbs
100 200
na
llu
Ca
is
ar
lg
vu
Heath
es
or
pe
)
sp
-ty
n e
ia
us
r
ula
r
r
e
r
a
at
et
a
e f lga
is nct
sic e
e ae
ino
-B
pe ifoli
t
v
m
e
r
e
y
l
a
P
a
e
u
e p
a m
u
il
v
a
su
e e
-t t
(A
ab
e
n
pe p ty ce ba
sp
sl
m us
ae lac lac
ps m
ae
in u m
s ae ace us
lle
e a ce nu hyl cea
-ty num illa- c ula isor la trum m niu a ng t eris et e diu um cero cero num 14 a 8
9
a
m
u
a
s
o
a
e
r
5
l
c
e ic a p
ra
ac ce ni u n eri th go nt n u gu lic liu rga a p
ol po idi ho
lp
ho ag G t3 G t1 G t5 G t1 ete llen
iac a ss m p r yo pe
ta
ps ba ra m sio n en l y t e nu ng er ha rifo pa yph r yo
on Pol y Pter Ant Ant Sph
d
T T S T D
Di Fa G e Hu Ja Lo M Po Po Ra Sa Sp
Po
M
Bv
Bv Bv Bv
In
A p Br Ca Ca Cy
To
329 971
385 1839
539 1289
535 1512
1140 1593
399 654
517 928
416 595
1132 1998
1140 2238
315 570
685 1399
519 1117
301 959
685 1379
520 959
399 769
284 818
291 695
1 1 1 20 20 40 1 1 1 1 1 5 1 5 1 1 1 20 1 5 1 1 20 20 5 5
20 5
20 1 1 20 20 20
1
s
s
s inu ra
s rcu
ca
g u a x de x yri cea nu e
Fa Fr He Ile M Pi Pi Qu
20 40 5
pe
ty
aed us s
om in lu
dr rp ry
An Ca Co
Trees and shrubs
species are known to expand in the Netherlands since the Iron Age (Janssen
1974). Since the percentages of Fagus and Carpinus are still low (<1.5%) a dating
of around 1000 – 400 cal BC is suggested.
Traces of post holes have also been found at Vaassen III, covered by the OALlayer. Charcoal from one of the post holes was dated to 3020 ± 55 BP (GrN5895, 1418-1114 cal BC, calibrated with Oxcal 4.2). This implies that the first
agricultural activities started after 1418-1114 cal BC. The presence of Fagus and
Carpinus (respectively 1.5 and 1.3%) suggests a date around 1000 cal BC.
The third date is provided by charcoal found in a pit underneath the CF layer.
The pit was dug into the CF layer, since part of the arable layer (PCF and/or CF)
had sunk down into the pit. The 14C-date of 1800 ± 55 BP (GrN-5495, 82-352
cal AD, calibrated with Oxcal 4.2) can be considered as a terminus post quem date
for the end of the agricultural activities at the Celtic Field. Brongers (1976, 64)
argued that this date coincides with the period the Celtic Field came to an end,
since this disturbance of the arable layer is probably the result of unstable times
during the Roman occupation.
8.2.2 Results
Figure 8.13 shows the results of the pollen analyses of all barrows and the
Celtic field. The pollen spectra were placed in chronological order to see the
vegetation development in the area. It should be noted that the different phases
show some gaps or overlap in time, so the spectra do not show a continuous
vegetation development. Secondly, the spectra belonging to the Celtic field could
only very roughly be dated (see above). Although the spectra have been derived
from different types of samples (barrow versus agricultural layers) it has been
decided by the author to all compare them with each other. All Celtic Field spectra
probably represent a longer period of time, since the soil has been mixed up due to
agricultural activities. The herbal vegetation composition shown by the spectra is
very local and cannot be expanded to the barrow sites nearby, but the extra-local
and regional forest vegetation probably can.
Phase 1: 2800-2600 cal BC
Vaassen 1
The amount of forest pollen represents approximately 57% of all pollen (including
spores). The herbal vegetation consisted mainly of grasses and Calluna heath.
The surrounding forest mainly consisted of Betula with some Quercus and Tilia,
although Betula might also have been present locally on the heathland. Corylus
was present in high amounts. In the wetter areas Alnus was the dominating tree.
Vaassen 3, period 1
The ratio between arboreal and non-arboreal pollen is the same as that of barrow
of Vaassen I. There seem to be some differences in the forest composition: Quercus
decreased, while Tilia increased. There is a considerable decrease in Betula pollen.
Grasses have decreased, while heather was able to expand a little. Together with
the decline of Betula this could indicate some heath management, for example
by grazing activity. This prevented new Betula trees from establishing and grasses
from flowering.
120
ancestral heaths
Phase 2: 2500-2200 cal BC
Niersen 4
Compared to the barrows at Vaassen, which are at almost 2 km of distance apart,
there is a great difference in the vegetation composition at Niersen. At Niersen
there seems to have been a larger open space, dominated by heather (Calluna
vulgaris), in which the barrow has been built (AP=32%). There were hardly
any Betula trees present and the amount of grass was considerable, indicated
by pollen percentages of 25-50%. This species-poor heathland could have been
maintained by heath management, preventing Betula to re-establish and Calluna
to expand. The surrounding forest consisted mainly of Quercus, Tilia and Corylus.
In addition, some peaks can be seen in Succisa pollen and fern spores, indicating
moist conditions. The Alnus forest in the stream valleys seems not to show any
differences with that of Vaassen.
Vaassen 2, period 1
The vegetation character derived from the pollen analysis of Vaassen 2 is
comparable to Vaassen 1 and 3. The percentage of Betula pollen is comparable
to Vaassen 3. This means that the percentage of Betula is higher than in Niersen,
but considerably lower than at Vaassen 1. There seems to be a slight increase in
Quercus pollen compared to the other barrows of Vaassen.
Phase 3: 2000-1800 cal BC
Niersen 6
Compared to the other barrow at Niersen, Niersen 4, there has been an increase in
trees (AP=57%). The percentage of tree pollen is comparable to Vaassen II. This
increase of trees is probably mainly caused by a decrease in heather pollen. An
increase of Betula can be seen, although the amount of Betula pollen is still very
low compared to Vaassen. Re-establishment of Betula might have been possible
because heath management has been less intensive, also causing the heathland to
decrease in size.
Phase 4: 1600-1400 cal BC
Vaassen 2, period 2
There has been an increase of tree pollen, compared to all previous phases (both
Niersen and Vaassen). All arboreal pollen has increased, except Quercus and
Corylus. Heath seems to remain unchanged. Some Cerealia pollen is present, but
only in very low amounts and other anthropogenic indicators are also not very
numerous.
Phase 5: 1000-400 cal BC
DOS, Celtic field
The percentage of tree pollen is high, accompanied by a low percentage of herbal
pollen. The percentage of anthropogenic indicators is very low as well. This
suggests that forest was present at this site before the start of agricultural activities.
This forest, with mainly Quercus and Corylus, might have been present when the
barrows were constructed, although at that time Carpinus and Fagus were not part
northern and central veluwe
121
of it. Tilia pollen is present in very low amounts and might have been replaced by
Fagus and Carpinus, confirming a younger dating than the barrows. Alder carr is
present in the stream valleys, as in the barrow period.
Phase 6: 1000 cal BC- 150 cal AD
OAL, Celtic field
The pollen spectrum of the OAL might represent a period that is older than the
period represented by the DOS samples, since this arable layer was present at
another location. However, the higher percentage of Carpinus and Fagus suggests
that this spectrum represents a slightly younger period (see also 2.1 and 8.1.4).
The percentage of arboreal pollen is considerably lower than in the DOS-spectra,
while the amount of cereal pollen and other anthropogenic indicators is much
higher. Calluna is also present in considerable amounts.
PCF, Celtic Field
The sample from the Pre Celtic Field layer is taken from the layer above one
of the DOS-samples at Vaassen I described above (DOS1). Compared to this
spectrum the percentage of arboreal pollen has decreased, while the percentage of
anthropogenic indicators and Poaceae has increased. The amount of cereal pollen
is in contrast to the OAL-spectrum very low.
CF, Celtic Field
One CF-sample is taken from the layer above the PCF-layer, the spectrum of which
is described in the previous paragraph; the other sample is coming from Vaassen
III, from the layer covering the PCF layer above the OAL. At both locations the
amount of tree pollen has further declined. The Alnus forest had not changed
or increased some, but the dry forest had decreased in size. Cerealia and other
anthropogenic indicators are present, but there is a difference between the CF at
Vaassen I and the CF at Vaassen 3: at Vaassen 3 the percentage of cereals is much
higher than at Vaassen 2.
The size of the open spaces
The minimum size of the open spaces can be estimated by the measurements of
the barrows and the height of the sods that had been used in the construction of
the mounds (see 7.1).
The height of the sods is only known for the second period of the Vaassen 2
barrow (0.25 m). Fontijn et al. (2013, 99-100, figure 4.25) have measured the
length and thickness of many sods at a barrow site called Oss-Zevenbergen (see
also 12.1) and concluded that the average thickness of sods used at that site was
on average 20-35 cm. In addition, the thickness of the sods of the Echoput was
approximately 0.25 m as well and apparently this is a suitable thickness to build
barrows. For the calculations of the other barrows a height of 0.25 m will be
assumed as well. This leads to the following minimum areas to be stripped per
barrow (see also table 8.2):
Niersen4: 2041 m2, ropenarea≈25.5 m, based on a circular open spot
Niersen6: 858 m2, ropenarea≈16.5 m
V1 540 m2, ropenarea≈13 m
V2 period 1: 30 m2, ropenarea≈3 m
V2, period 2: 268 m2, ropenarea≈9 m
V3: 30m2, ropenarea≈3 m
122
ancestral heaths
These numbers indicate minimum areas. The dating of the barrows is not detailed
enough to determine whether some of the barrows were built at the same time as
was probably the case with the Echoput barrows (see 8.1). Hence, the calculated
areas of the Vaassen 1 and Vaassen 3 barrows cannot be added together. Based
on the ratio of arboreal versus non arboreal pollen percentages (see 7.2) the open
spaces were larger than the stripped area. The ADF of the Vaassen open spot is
estimated at 25-100 m. The ADF of the open area at Niersen was at the oldest
phase (Niersen 4) 100- 200 m and was somewhat smaller (around 50-100m) when
Niersen 6 was built.
8.2.3 Discussion
The pollen spectra show that the barrows at both Vaassen and Niersen were built
in open places with heath vegetation. The barrows of Vaassen were built in an
open spot with an ADF of approximately 100 m based on the ratio AP versus
NAP. The open place in which the Niersen barrows were built was larger with
an ADF of more than 100 m. Both open spaces were dominated by Calluna
heath and grasses. The arboreal pollen percentage is dominated by Alnus, which
is probably the result of an alder carr in the lower and wetter parts of the area.
The forest of the drier area consisted mainly of Quercus and Corylus, the latter
likely to be found at the forest rim. The vegetation of the open space seems stable,
since the barrow spectra from all represented periods show similar vegetation
patterns: an open place with species-poor grassy heathland surrounded by oak
forest with an alder carr nearby. Some Neolithic finds underneath barrow V1,
together with the relatively high percentage of anthropogenic indicators in the
samples from V1 might indicate that the open space of the Vaassen barrows was
used as a settlement area prior to the barrow building. After the barrow was built
archeologically visible human activity decreased, leading to the decreased amount
of anthropogenic indicators present at the when time barrow V2 was built. This
could be an indication of change in function: a place for the living changed
into a place for the dead with only the necessary management activities being
maintained. The continued maintenance of the heath vegetation from when the
oldest barrows (V1, V3) were built continuing to when the younger barrow (V2)
was constructed strongly indicates conscious management. This also accounts
for the Niersen barrow area in an even more pronounced way. The Niersen
barrows formed a long alignment of barrows11 (Bourgeois 2013, 51-66). From
this alignment only two barrows were analysed for pollen. However, based on the
results of barrows that formed part of other alignments (see Chapter 9) and on the
palynological data of all other barrows in the southern and central Netherlands
(see the remaining of this chapter and Chapters 9-12), it can be assumed that all
barrows belonging to the Niersen alignment were built amongst heath vegetation.
During the earliest phase (late Neolithic A) the alignment was at least 1.6 km
long containing 6 barrows. With an ADF of 100-200 m it is very likely that the
heath areas the barrows were built in were connected to each other, forming a
long-stretched heath area. The alignment was extended in the Bell Beaker phase
implying an even more extended heath area; Heath that had to be managed to
remain in existence. Comparable to the Echoput, barrow management could
have taken place by grazing, burning and/or sod cutting. It is not clear from the
results whether there are indications of burning the heath. Grazing is indicated
by the presence of Poaceae in combination with Plantago lanceolata, Asteraceae
liguliflorae and Succisa (Hjelle 1999). A notable difference between the Vaassen
11
The alignment might even have been more extended while part of it might have been destroyed by
modern land use
northern and central veluwe
123
and the Niersen barrows are the high amounts of Betula pollen at Vaassen and the
almost absence of this taxon at Niersen. Betula is a pioneer tree, meaning that it
is one of the first to appear when no management is applied to prevent the tree
from establishing. Young Betula trees are easily removed by grazers. This could
indicate a difference in grazing intensity or management method (grazing versus
not grazing) between the two barrow locations. Either this could mean that the
barrows of Niersen belonged to another community with different management
regimes or perhaps this could mean a difference in importance between Niersen
and Vaassen is indicated. Niersen being part of a barrow alignment, while the
Vaassen barrows might not be related to this.
The next phase is represented by the DOS (denuded old surface) layer, the
surface at which the first cultivation of crops started. This phase shows a higher
percentage of arboreal pollen compared to the barrow phases. Although the dating
of this layer is very coarse it is likely that this pollen spectrum represents the
phase prior to the arable activities, since the amount of cereals and arable weeds
is still very low. Probably forest was present at this site, which might very well
be the forest that has been recorded in the barrow pollen spectra. The amount
of anthropogenic indicators is very low. This could indicate that there was not
a lot a human activity in the area. The absence of human influence in the area
is also indicated by the sparseness of archaeological finds in the area. From the
Middle Bronze Age period onwards there is hardly any evidence for the building
of new barrows (Bourgeois 2013). However, older barrows have been frequently
used for secondary graves indicating not a total absence of humans in the area. In
addition urnfields have been found in the area, including one in the Celtic field
of Vaassen.
At the Celtic field sections of Vaassen III the first agricultural activities have
been recorded (OAL). The forest had probably decreased in size and at least this
site was cleared of trees. The amount of anthropogenic indicators, including
Cerealia, and arable weeds like Artemisia, is a clear indication for crop cultivation
and more specifically the cultivation of cereals. Heather is well represented in
the pollen spectrum. Since this spectrum probably represents a longer period, it
is likely that heath vegetation was present at the site before agricultural activities
started or perhaps during times when the arable fields were abandoned. Another
possibility is the presence of heath very close to the agricultural field.
At the Pre-Celtic Field phase the forest that was first present (Vaassen I, DOS)
was cleared and agricultural activities were started. The amount of cereals is not
very high, but considering that prehistoric cereal pollen do not spread (Diot 1992)
it is likely that this spot was used for crop cultivation. The agricultural activities
were probably expanded during the next phase, when the Celtic Field system was
created. The forest clearance had been furthered at this stage. At Vaassen III cereal
cultivation was continued (started at the OAL) and at Vaassen I other crops might
have been cultivated.
8.3 Ermelo
In the area of Ermelo over a hundred barrows are known to be located, of which
55 have been excavated. During a great campaign in 1952, Modderman excavated
34 of these barrows (Modderman 1954) providing high-quality information on
the mounds (Bourgeois 2013). In 1971 a re-excavation took place by Lanting and
van der Waals during which two barrows (Ermelo I and III) were sampled and
analysed for pollen (Casparie and Groenman-van Waateringe 1980, 29-30, 31).
124
ancestral heaths
Ermelo III
rmelo
of E
ment
align
w
o
r
Bar
0
125
250
Figure 8.14. Detailed map
of the Ermelo area with the
locations of the barrows from
the Ermelo barrow alignment
that were sampled for pollen
analysis. The map is based on
digital elevation model of the
AHN (copyright www.ahn.
nl).
Ermelo I
Ermelo III
rmelo
fE
ent o
nm
w alig
Barro
500 m
0
125
250
500 m
Sampled barrows
Sampled barrows
Other barrows
Other barrows
m NAP
m NAP
60 m
0m
60 m
0m
8.3.1 Site description and sample locations
Several barrow alignments were recognized in this region. The two investigated
barrows formed part of one of these alignments and are situated about 125 m
from each other (Bourgeois 2013, 78-88; figure 8.14). This barrow alignment is
situated at the bottom of a valley on the northern slope of the ice-pushed ridge of
Garderen. Ermelo I is a single period barrow, originally excavated by Modderman
(1954). The mound probably was surrounded by a palisaded ditch (diameter=5.5
m), that consisted of a broad trench which was filled up after posts were placed. Part
of an AOO-beaker was found in the upper part of the ditch fill (see figure 8.15),
dating the barrow to the late Neolithic A. The barrow was re-excavated by Lanting
and van der Waals in 1971 (Lanting and van der Waals 1971b, 1976). Samples for
pollen analysis were taken from the old surface in and outside the encircling ditch,
from the ditch fill (referred to as turfs by Casparie and Groenman-van Waateringe
1980) and from upper part of the ditch fill (referred to as the old surface by
Casparie and Groenman-van Waateringe 1980, 31; see figure 8.15). Ermelo III
northern and central veluwe
125
Figure 8.15. Ermelo barrow
I with the probable sample
locations. 1= old surface
inside encircling ditch, 2=
old surface outside encircling
ditch, 3= ditch fill, 4= upper
part ditch fill. Figure redrawn
after Modderman (1954, plate
XXXIV).
is a single period barrow. The barrow was originally excavated by Modderman
(1954). Two PF-beakers and a flint blade have been found, dating the barrow to
the Neolithic A. This barrow is like Ermelo I approximately 0.5 m of height and
has a diameter of about 6.5 m. The barrow was re-excavated by Lanting and van
der Waals in 1971 (Lanting and van der Waals 1971b). Samples for pollen analysis
were taken from the old surface underneath the mound (Casparie and Groenmanvan Waateringe 1980, 29-30).
8.3.2 Results
Results will be described per barrow in chronological order. See figure 8.16.
Ermelo III (2900-2500 cal BC)
The pollen spectra from the old surface of Ermelo III show an arboreal percentage
of approximately 50%. This arboreal pollen percentage consists mainly of Alnus.
Corylus is present in considerable amounts of approximately 35%. Other trees
are Quercus (5-10%), Tilia (10-15%) and Betula (5%). The herbal vegetation is
dominated by Calluna vulgaris and Poaceae. Some anthropogenic indicators are
present in the form of Chenopodiaceae and Asteraceae tubuliflorae. A few pollen
grains of Cerealia were also noticed. Grazing indicators are mainly represented by
Poaceae and Plantago lanceolata.
126
ancestral heaths
northern and central veluwe
127
2900-2500 BC
2600-2500 BC
20 40 60
20 40 1
1
1
20
20 1
20 5
20 40
50
100 150 5
ris
lga
vu
1
5
5
20 40
1
5
1
1
5
1
1
1
5
1
1
20 40 5
5
686 1357
659
329
Ermelo III_os1
Figure 8.16. Pollen spectra from samples taken from
the Ermelo barrows. Spectra are given in % based on
a tree pollen sum minus Betula pollen. In the total
AP (=arboreal pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are included, non
pollen palynomorphs are excluded. Different scales
have been used, indicated with different colours.
969
459
Ermelo III_os2
20
478 1084
Ca
na
llu
Ermelo III_os3
r
Co
us la
m tu
Ul Be
686 1644
n
Al
s
us
s inu
s
erc
lix ia
gu x ea u
Fa Fra Pic Pin
Qu
Sa Til
Ermelo I_os (in)
20 40 60 80 100
P
NA
s
ylu
Upland herbs
Anthr. ind. Grazing indicators
Ferns and mosses
pe
t- y
lla
pe
se
-ty
)
e
to
ae
r a ta
um
u la
ce
or
e
t
o
a
fl
et
fl a
.
ar
ra
-B
uli
ae
/R
lg
uli eol
m
fo
a e ceae
P
e
r
b
e
a
u
g
e
a
i
s
e
c
A
c
v
su
u
l n
(
e a
o
p
ae llac ia
et
et
e ac r i s
m
ae la
m
len
um
ea lia ace go ae
ac a eae ca ce p hy pod eae icum cea cul h ula teri
ol
di ium gnu n su
c
x
o
p
r
a
s
o
a
n
i
r
p
e
a
c
i
e
d
a
t
l
r
lyp e ri ha o lle
m c c iac ss ryo en ba pe mi n u r op yo
ta
te re s te a n ac
P
Ru Su Ap Bra Ca Ch Fa Hy La Ra Sc Dr
Po Pt Sp
To
As Ce A Pl Po
594 1647
Heath
1186 2626
AP
us
Trees and shrubs
Ermelo I_os (out)
Ermelo I_ditch
Ermelo I_ditch (up)
Ermelo
Ermelo I (2600-2500 cal BC)
Compared to Ermelo III the arboreal pollen percentage seems to have slightly
increased to 55%. The main tree is still Alnus. Corylus is also still present in
high amounts (35%). The amount of Tilia seems to have decreased to 5-10%;
the amount of Quercus seems to have slightly increased to 10-15%. Fagus has
appeared, although still in very low numbers. Betula expanded from 5% at Ermelo
III to 20% at Ermelo I. The heather seems to have expanded with percentages up
to 125% at cost of Poaceae. No indications of Cerealia have been found. Other
anthropogenic indicators such as Asteraceae tubuliflorae and Chenopodiaceae are
present in low amounts.
The size of the open space
Based on the measurements of the barrows the minimum size of the open area
has been calculated (see also table 8.2). Since the height of the sods is not known
a standard height (known from the Echoput and Vaassen barrows) of 0.25 m has
been applied. This gives the following estimates of open area:
Ermelo I: 33.4 m2, ropenarea≈3.3 m, based on a circular open spot
Ermelo III: 24 m2, ropenarea≈2.7 m
Based on the ratio AP:NAP, the open space had an ADF of approximately 50-100
m. The open spot might have decreased a little at the time Ermelo I was built
(AP=55% for Ermelo I and AP=50% for Ermelo III). The relation found between
arboreal pollen percentage and size of the open space (see 7.2) is not detailed
enough to explain this difference in percentage by a difference in distance to the
forest.
8.3.3 Discussion
The vegetation composition in the area of the Ermelo barrows in the late Neolithic
seems to be quite similar to the late Neolithic phase of Vaassen (8.1.2). The barrows
were built in an open space with an ADF of 50-100 m with a vegetation cover
of mainly heather and grasses. When the first barrow (Ermelo III) was built the
heath seemed to more grassy than when Ermelo I was built. The two investigated
barrows were part of a barrow alignment implying that they were built in a longstretched heath area (see also 8.2.3 and chapter 9). Management is required to
maintain such areas of heath. The increased amount of Betula could indicate a
change in management regime making it possible for Betula to expand. This is
also indicated by a slight decrease in anthropogenic and grazing indicators. An
extensive alder carr must have been present in the stream valleys close to the
barrows indicated by Alnus pollen percentages of approximately 45%. The dry
forest was most likely quite open with mainly Corylus and some Quercus and
Tilia.
8.4 Putten
8.4.1 Site description and sample locations
Close to the village of Putten, approximately 5 km to the southwest of the
Ermelo barrows, a burial mound is situated (see figure 8.1). This barrow was
excavated by van Giffen in 1947 and a sample from the old surface was analysed
for pollen by Waterbolk (1954, 93-94). During this excavation a PF-beaker was
found together with a battle axe, a Grand Pressigny dagger, a flint axe and four
flint flakes. Three secondary interment Bell Beakers were buried in the mound
128
ancestral heaths
northern and central veluwe
129
40 60
40
60
r
Co
s
20
ylu
40
s
cu
er
Qu
ia
Til
60
80
40
60
60 80 100
m
m
iu
um g nu
od
di
lyp te ri
ha
o
p
S
P
P
e
ar
a)
ul
et
-B
m
P
(A n su
m
le
su p ol
n
l
ria olle ota
P
Va
T
193 240
80 100
20
20
5
20
Vierhouten
AP
Vierhouten
5
20
1
40
5
60
1
5
80 100
P
NA
1
20
s
nu
Al
20
40
20
s
lu
ry
Co
20 40
40
1
20
1
1
1
5
20
20
1
5
20
20
40
20
e
ea
la
ac
tu
ic
e
r
B
E
Trees and shrubs
1
s
s
cu
s
gu x nu er
lia
Ti
Fa Ile Pi Qu
5
Figure 8.18. Pollen spectra from the samples taken from
the Vierhouten barrow. Spectra are given in % based on
a tree pollen sum minus Betula pollen. In the total AP
(=arboreal pollen) Betula is included. In the total NAP
(= non arboreal pollen) spores are included, non pollen
palynomorphs are excluded. Different scales have been
used, indicated with different colours.
LNEO-B
20 40
80
20 40
20
Graz. ind.
Ferns and mosses
Anthr. ind.
Upl. herbs
20
1
1
5
1
1
5
1
5
ae
a)
ul
or a
e
et
ar
lifl lat
e
B
g
u
a
m
l
e e
P
b eo
su
vu
(A
ea a c
tu c
e lan
en
ac lari ris ium m um
l
l
a
l
o
u
e cu u e d n
o
s
ce g
lp
ra ta ea n ph pt po g n
te an ac nu ro ryo ly ha lle ota
As Pl Po Ra Sc D Po Sp Po
T
545 1090
Heath
20
319 1022
20
s
lg
vu
Ferns and mosses
Putten_os2
nu
Al
Upland herbs
361 1234
P
NA
e
yp
-t
lla
se
to
e
c
ia
ar
.a
e
sic
/R
ea
er eae
sa
c
p
e
o
e
a
c
et
ea yll
ea m la is
ac
ac h
ac nu cu er
ni o n pt
ex isa ic op
ra olyg an u ryo
m uc c rass ary
e
u
R S
B C
G P R D
Heath Anthropogenic indicators Grazing indicators
ae
ae
or
or ta
i fl
li fl ola
ul eae
is
u
b
e
r
tu iac
lig nc
ga
ul
ae od
ae o la e
a
av
isi race op alia race tag
ea
us ula llun
m
n
ac
m t
te te e re te a n
Ar As Ch Ce As Pl
Ul Be Ca
Po
Trees and shrubs
Putten_os1
AP
Figure 8.17. Pollen spectra from the samples taken
from the Putten barrow. Spectra are given in %
based on a tree pollen sum minus Betula pollen. In
the total AP (=arboreal pollen) Betula is included.
In the total NAP (= non arboreal pollen) spores are
included, non pollen palynomorphs are excluded.
Different scales have been used, indicated with
different colours. LNEO-A= Late Neolithicum A.
LNEO-A
Putten_os(W)
Putten
(Waterbolk 1954, 93). The old surface contained fragments of PF-Beakers that
might indicate a former settlement site (Casparie and Groenman-van Waateringe
1980, 30). Re-excavation of the barrow took place in 1971 for pollen sampling.
Samples were taken from the old surface. Results have been published by Casparie
and Groenman-van Waateringe (1980, 30). Measurements of the mound are not
known.
8.4.2 Results and discussion
See figure 8.17
The first thing to notice is the difference in pollen spectra from the sample published
by Waterbolk and those published by Casparie and Groenman-van Waateringe.
The Waterbolk spectrum shows an arboreal pollen percentage of approximately
75%, while the arboreal pollen percentage in the spectrum published by Casparie
and Groenman-van Waateringe is only 30%. The differences seem mainly to
have been caused by high percentages of Poaceae and ferns in the Casparie and
Groenman-van Waateringe spectra, which are very low or absent in the Waterbolk
spectrum. Waterbolk mentioned the bad conservation of pollen in his sample. He
did not reach a pollen sum of 300 arboreal pollen grains and as a consequence
this spectrum might not be representative. However, it is difficult to conclude this
being the cause of the dissimilarities. Yet, it is difficult to interpret these results.
Some similarities can be seen. All pollen spectra show very low percentages of
Calluna vulgaris, indicating that the open space did not contain a lot of heather.
This could be the result of a small open space (as in the Waterbolk spectrum)
or a larger open space that was dominated by grasses (as in the Casparie and
Groenman-van Waateringe spectra).
8.5 Vierhouten
8.5.1 Site description and sample locations
Close to Vierhouten (see figure 8.1) a single period barrow was excavated in 1939
by A.E. van Giffen. Two Veluvian Bell Beakers and a wrist guard were found
dating the barrow to the late Neolithic B period (2500-2000 cal BC, see table
2.1). Measurements of the mound are not known. In 1972 a re-excavation took
place by Lanting and van der Waals (1972c). At that time samples for pollen
analysis were taken from the old surface. One sample was analysed and published
by Casparie and Groenman-van Waateringe (1980, 36).
8.5.2 Results and discussion
The pollen spectrum (see figure 8.18) shows an arboreal pollen percentage of
approximately 56%, which indicates an open place with an ADF of approximately
50-100m. Trees in the surroundings are dominated by Alnus and Corylus, which
both occur with pollen percentages of approximately 40%. Quercus, Tilia and
Betula are present in less but still considerable amounts of circa 10%. The open
spot was mainly covered with Ericaceae, most likely Calluna vulgaris. Other herbs
were almost absent. The situation is comparable to the late Neolithic B-phase of
Niersen-Vaassen.
130
ancestral heaths
northern and central veluwe
131
20 40
60
1
1
1
5
P
20
1
us
5
1
20 40
s
ylu
50
e
ea
ac
100 150
1
5
1
1
1
1
5
20 1
5
1
1
1
1
1
1
1
1
5
1
20
1
1
5
529 1218
Uddelermeer 2
AP
20 40
60
80 100
NA
n
Al
20
40
60
80
r
Co
20 40
60
1
5
20
s s
gu nu er
Fa Pi Qu
s
cu
5
1
5
na
20
us la u
lia m t u ll
Ti Ul Be Ca
Trees and shrubs
Figure 8.20. Pollen spectra from the samples taken from the Uddelermeer
barrows. Spectra are given in % based on a tree pollen sum minus Betula
pollen. In the total AP (=arboreal pollen) Betula is included. In the total NAP
(= non arboreal pollen) spores are included, non pollen palynomorphs are
excluded. Different scales have been used, indicated with different colours.
LNEO-B
Uddelermeer 1
Uddelermeer
vu
40
is
ar
lg
60
80
Anthr. ind. Grazing ind.
Upl. herbs Ferns and mosses
Algae
5
5
1
20
40
1
1
20
1
1
20
40
337 737
es
or
sp
a)
ae ae
n
r
r
ul
o r
re
fe
et
ifl o ta
e
-B
u l ulifl e ola
te ulga
m
e
a
P
a
l
b
l
e
i
v
c
( A n su
tu lig c
ab
ps
m
e e la n
lla
in
um olle
ea cea go ae
hy zia lete odiu 14 rm
s
c
a
p
p
n
is o er no
t3 te
ra ra ta e
lle otal
lyp G de
cc ry p
te te an ac
Su Ca Hu Mo
Po
Po Bv In
A s A s Pl P o
T
332 535
Heath
620 1219
60 80
ic
Er
Ferns and mosses
Emst_os_per1
20 40
s
s
s la
s in u ra s
cu
gu ax de nu er alix ili a lmu et u
Fa Fr He Pi Qu
S T U B
Upland herbs
417 807
1
s
lu
ry
Co
pe
-ty
lla
e
os
et
ac
Grazing indicators
Emst_os_per2
60 80 100
s
ies nu
Ab Al
Anthr. ind.
)
ae
la
ae
or
e
tu
ta
or
ar
lifl
Be m
la
e
R.
e l ifl
g
e
a
/
u
o
l
a
a
P
u
e
e
b
e
sa
ce
su
vu
(A
ae ea
tu
e ace
ac lig e nc
ia
to
n
di ae typ la
m m
e a yll eae l a ae ace lac e lar m ris
le
um
ce
ae
a
di iu m nu n su pol
isi ace lia opo ace m- tag o eae ex a isa icac ph rac ndu ace ver ncu ce a hu ct ru pte
o
g
d
a
l
r
r
o
p
e
p i
tem te re en te liu an ac um ucc rass ary ype i lip ami apa anu osa c ro hal ry o oly t eri ph a olle ota
P P S P
R S
Ar As Ce Ch As Ga Pl Po
B C C F L P R R S T D
T
580 1642
Heath
594 1096
20 40
P
NA
Trees and shrubs
Emst_os_per3
AP
Figure 8.19. Pollen spectra
from the samples taken from
the Emst barrow. Spectra are
given in % based on a tree
pollen sum minus Betula
pollen. In the total AP
(=arboreal pollen) Betula is
included. In the total NAP (=
non arboreal pollen) spores
are included, non pollen
palynomorphs are excluded.
Different scales have been
used, indicated with different
colours.
LNEO-B
Unknown
Emst_os_per4
Emst_sur_per4
Emst
8.6 Emst
8.6.1 Site description and sample locations
Near Emst a barrow of probably four periods is situated. The barrow was first
excavated in 1932 by J. Butter. The first period was dated to the late Neolithic B
period based on the bodies being buried semi-flexed (Hulst 1972). The original
measurements of the barrow are not known. Samples for pollen analysis were
taken from the old surface of all periods. The results were published in Casparie
and Groenman-van Waateringe (1980, 36-37).
8.6.1 Results and discussion
See figure 8.19
The pollen spectrum of the late Neolithic B period shows an arboreal pollen
percentage of approximately 65% dominated by Alnus (60%) and Corylus (30%).
This indicates an open space with an ADF of approximately 50 m at the oldest
phase, which is very small compared to most of the other barrows in this region.
This open spot is mainly covered with heath vegetation and grasses and most
likely some Betula trees. In the next periods (which are not dated) the amount of
arboreal pollen decreases, accompanied by an increase of heath. This indicates an
increase of the open spot to an ADF of approximately 150m.
8.7 Uddelermeer
8.7.1 Site description and sample locations
Two barrows at the edge of the Uddelermeer (see figure 8.1) were excavated in
1911 by Holwerda. Uddelermeer 1 measured approximately 20 m in diameter
and 1.0 m in height. Uddelermeer 2 was approximately 18 m in diameter and 1.5
m high. Both barrows were dated to the late Neolithic B period based on sherds
from Bell Beaker pottery (Holwerda 1912), however, since these finds were small
this dating could be questioned (Q. Bourgeois pers. comm., October 2012). In
1989 both mounds were the focus of conservation by the ROB (presently known
as Cultural Heritage Agency of the Netherlands). Samples for pollen analysis have
been taken from the profile in trenches during consolidation. The soil samples
were taken in small glass tubes, which were sealed and sent to the University of
Amsterdam, to Prof. Groenman-van Waateringe. The samples were stored until
July 2009 and then taken to Leiden University for analysis. From both mounds a
sample from the old surface was prepared and analysed by the method described
in Chapter 4. It should be noted that samples were derived from trenches. This
makes it is difficult to relate these samples exactly to the barrow, since only a small
part of the barrow was exposed. Therefore properly dating of the pollen spectra
is difficult as well, what with the dates of the barrows themselves being already
in doubt.
8.7.2 Results and discussion
The preservation of pollen was poor in both samples resulting in a high amount
of indeterminable pollen grains. The ratio arboreal versus non arboreal pollen
is approximately 65-35% for Uddelermeer 1 and approximately 45-55% for
Uddelermeer 2 (see figure 8.20). When an average thickness of 0.25m for the sods
is assumed (see 8.2.2) the area that needed to be stripped to build Uddelermeer 1
132
ancestral heaths
is approximately 630 m2 indicating an open space with a radius of approximately
14 m. To build Uddelermeer 2 approximately 770 m2 (radius ≈ 15.5 m) was
necessary. Based on the arboreal pollen percentage the size of the open area had
an ADF of approximately 50 m for Uddelermeer 1 and approximately 150 m for
Uddelermeer 1. This might indicate that Uddelermeer 1 was built first in a small
open space and that Uddelermeer 2 was constructed later when the open space
had expanded. Both barrows were built in heath and grass vegetation. The forest
in the surroundings was probably quite open and consisted mainly of Corylus.
Alder carr was present in the wetter areas.
8.8 Boeschoten
8.8.1 Site description and sample locations
In de area of Boeschoten (see figure 8.1) a barrow was excavated by Glasbergen
and van der Waals in 1952. The old surface contained lots of charcoal particles.
The excavators dated the barrow to the Early Bronze Age or the late Neolithic B
period, based on sherds of ceramics found in the old surface. Measurements of the
barrow are not known. Samples for pollen analysis were taken from the old surface
underneath the mound and from the fill of the ditch surrounding the barrow. The
results of the pollen analysis were published by Waterbolk (1954, 93-95).
8.8.1 Results and discussion
The pollen spectra from both samples show rather similar results (see figure 8.21).
The arboreal pollen percentage is approximately 65%. This indicates a small open
space with an estimated ADF of approximately 50 m. The surrounding forest
consisted mainly of Quercus and Corylus with nearby an alder carr in the wetter
parts of the area. The amount of herbal pollen is low and consists of 13-28%
Calluna and approximately 15% Poaceae. The pollen spectra show very poor
variety of species. Some anthropogenic indicators are present, however, in such
low amounts that they cannot be linked to the activity of man.
8.9 Ugchelen
8.9.1 Site description and sample locations
Near Ugchelen four barrows were excavated in 1947. All barrows were heavily
damaged prior to the excavation and the original measurements of the barrows
could not be reconstructed. Two of the barrows (Ugchelen 1 and 4) could be
sampled for pollen analysis. These barrows could not be dated. The obtained
pollen spectra from the old surfaces of both barrows have been published by
Waterbolk (1954, 94-95).
8.9.1 Results and discussion
See figure 8.22
Like the barrow of Boeschoten, the barrows of Ugchelen were built in a small
open space with an ADF of approximately 50 m covered with heather and grasses
surrounded by a forest of mainly Quercus and Corylus. Alder carr in the wetter
surroundings was probably responsible for the high percentage of Alnus pollen
in the spectra. A remarkable difference in Tilia pollen between the two barrows
(approximately 20% for Ugchelen 1 and approximately 2% for Ugchelen 4) makes
northern and central veluwe
133
134
ancestral heaths
Figure 8.22. Pollen spectra from the
samples taken from the Ugchelen
barrows. Spectra are given in % based
on a tree pollen sum minus Betula
pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (=
non arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours.
Figure 8.21. Pollen spectra from the
samples taken from the Boeschoten
barrow. Spectra are given in % based on
a tree pollen sum minus Betula pollen.
In the total AP (=arboreal pollen) Betula
is included. In the total NAP (= non
arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours.
Ugchelen 4
Ugchelen 1
AP
20 40
Boeschoten_ditch
Ugchelen
LNEO-B/EBA
Boeschoten_os
Boeschoten
60
AP
P
NA
40
80 100
20
20
40
80 100
s
nu
Al
60
P
NA
60
Co
40
Co
20
s
lu
ry
40
5
20
5
1
20
lix lia
Sa Ti
Anthr. ind.
Grazing ind.
Upl. herbs
Ferns and mosses
5
20
40
1
1
Heath
1
5
20
1
20
1
1
20
pe
-ty
Anthropogenic ind. Grazing ind.
5
1
5
476 866
Upland herbs Ferns and mosses
20
20
20 40
1
5
1
5
20
20
1
5
5
1
20
1
20
20
5
476 866
a
ell
e
os
a)
et
ra
ae
r
ul
c
o
a
e
a
t
et
ifl
.
a
ar
flo
e e
-B
ul
/R
m
ae u l i
lg
ol
is
a
P
b
e
r
e
a
u
e
a
s
su
a
v
c
(A
tu
nc
ac lig
to
n
lg
la ce
m
e
la
di e
ce
yl ula eris
vu
iu um
um olle
o
um
ea ia po cea ae
a
h
s
d
g
n
c
a
c
t
a
i
x
p
o d
o ra e
n
g
n
n p
e is
la
ta
lp
ra al
m cc ryo nu yo
llu
lyp eri
ha
ria lle ota
tu
te re en te ac lan
Ca
Be
P
Sp
Po Pt
Ru Su Ca Ra Dr
As Ce Ch As Po
T
Va Po
160 264
Trees and shrubs
5
s
s rcu
nu e
Pi Qu
20 40
s
s rcu
nu e
Pi Qu
60
20 40
s
lu
ry
20
s
nu
Al
Heath
pe
-ty
le la
s
to
ae
a)
ae
ul
ce
or
e
o r ta
et
.a
ar
lifl
e lifl ola
B
e
R
u
g
s
m
a
/
l
a e
i
u e
P
b
e
a
su
ar
vu
ce ea
(A
tu
ac lig nc
os
lg
m
lla lac is
en
et
di ae o la
m
m
l
ae
y
u
c
l
vu
o
i
u
r
m
u
u
e
e
h c e
a
o
s
d u
c ia op ac ag eae
a
gn n
ex isa
op n pt ypo ridi
lp
ra al
r t
la un
ry nu yo
l e
ha lle ota
te re en te an ac um ucc
lia tu ll
Ti Be Ca
Po Pt
R S
Sp Po
T
As Ce Ch As Pl Po
Ca Ra Dr
211 316
Trees and shrubs
Stroe
Unknown
LNEO-B
Er
i ca
As cea
te e
ra
c
As eae
te
tu
r
a
Pl ce bul
an a
ifl
or
e
t
ae
Po ago lig
ac la ulifl
e a nc o
r
e
e
o l ae
Su
at
cc
a
i
Ca sa
ry
o
Ra ph
nu y l
l
Dr ncu ace
yo la ae
pt ce
er ae
is
Po
lyp
P t od
er ium
id
Sp ium vu
lg
ha
ar
e
Po gnu
lle m
n
su
To
m
ta
l p (A P
-B
ol
et
le
ul
n
a)
su
m
HeathAnthr. ind. Grazing ind. Upl. herbs Ferns and mosses
Be
tu
la
Qu
er
cu
s
Til
ia
ry
lu
s
Co
Al
nu
s
AP
NA
P
Trees and shrubs
Stroe_os1_per2
346 554
Stroe_os2_per2
335 550
Stroe_os_per1
20
40
Figure 8.23. Pollen spectra
from the samples taken from
the Stroe barrow. Spectra are
given in % based on a tree
pollen sum minus Betula
pollen. In the total AP
(=arboreal pollen) Betula is
included. In the total NAP (=
non arboreal pollen) spores
are included, non pollen
palynomorphs are excluded.
Different scales have been
used, indicated with different
colours.
60
80 100
20
40
60
80
20
40
20
20
20
40
5
20
1
5
20
1
1
1
20
5
5
5
326 528
it unlikely that they were built in the same period. Although the barrows have not
been dated, Barrow 1 can be assumed the older of the two based on the rather
high percentage of Tilia pollen. However, based on only one sample this cannot
be concluded with certainty. Also remarkable are the high percentage of Plantago
lanceolata in the spectrum from Barrow 1 and the presence of Cerealia. This may
indicate an increase in human activity in the area around the period Barrow 1 was
constructed than when Barrow 4 was built.
8.10 Stroe
Near Stroe (see figure 8.1) a barrow is located that was excavated several times
(by Pleyte and Nairac in 1877, by Westendorp in 1926-1929 and by Lanting
and van der Waals in 1971). The barrow might contain two periods, although
this cannot be confirmed with certainty based on the excavation data. The first
(?) period of the barrow was dated to the Late Neolithic B, based on the find of a
copper tanged dagger, a wrist guard and a Veluvian Bell Beaker. Below the mound
some PFB sherds were found. Some fragments of charcoal that were scattered
on the old surface were 14C-dated to 3955 ± 55 BP (GrN-6350; 2600-2287 cal
BC, calibrated with Oxcal 4.2) and might be associated to the PFB material. The
barrow was re-excavated by Lanting and van der Waals in 1971 (Lanting and
van der Waals 1971c, 1976). Samples for pollen analysis were taken from the
old surface of the primary mound and from the old surface of the presumably
secondary mound in 1971. The results of the pollen analysis were published by
Casparie and Groenman- van Waateringe (1980, 34).
8.10.1 Results and discussion
The barrow was built in a very small open place with an ADF of less than 50 m. In
contrast to all other analysed barrows this mound was not built in heath vegetation
(see figure 8.23). Instead, the vegetation at the open spot was probably covered
with grass, indicated by the relatively high percentages of Poaceae found in the
pollen spectra (ca. 15%). The forest in the surrounding area mainly consisted of
Corylus, Quercus and probably also Betula. Alder carr was present in the wetter
parts of the area. The barrow was possibly built on a former settlement, given the
finds of PFB material. The presence of heath and grazing indicators suggests that
the site was used as pasture before the barrow was built and after abandonment
of the settlement.
northern and central veluwe
135
diameter (m)
height (m)
Echoput 1
19
10.8
Echoput 2
14.5
Niersen 4
Niersen 6
Vaassen I
Vaassen II
sod thickness diameter 2nd
(m)
period (m)
height 2nd
period (m)
Sod area (m2)
Radius (m)
0.25
615.06
13.99
1
0.25
332.35
10.29
28
1.65
0.25
2041.39
25.49
19
1.5
0.25
857.65
16.52
13
1
0.25
267.56
9.23
8
0.3
0.25
30.22
3.10
Vaassen III
7.5
0.3
0.25
26.56
2.91
Ermelo I
6.5
0.5
0.25
33.44
3.26
5.5
0.5
0.25
24.02
2.77
Ermelo III
Putten
unknown
Vierhouten
unknown
Emst
unknown
15
1.4
Uddelermeer 1
20
1
0.25
630.41
14.17
Uddelermeer 2
18
1.5
0.25
770..48
15.66
Ugchelen
unknown
Boeschoten
unknown
8.11 Palynological results from peat and lake sediments
8.11.1 Site description and sample locations
The Uddelermeer is one of the largest pingo ruins in the Netherlands created
in the Last Glacial period of the Pleistocene. It is very deep, around 17 m, and
has slowly been filled up with organic mud. Pollen was caught in every layer of
organic mud and an archive of vegetation development was formed. Polak took
samples for pollen analysis at four places, the results of which were published in
Polak (1959).
8.11.2 Results and discussion
Polak (1959) made several pollen diagrams that show the regional vegetation
development of the area. In figure 8.24 a summarized pollen diagram of the Polak
diagrams is shown. The diagram is based on the results of two different sample
locations: a deeper location with the older organic layers, the results of which
are shown in the part below the dashed line. The part above the dashed line
shows the more recent vegetation development, derived from the upper organic
layers. The pollen sum used in this diagram is based on the arboreal pollen sum
minus Betula to be able to compare it to the barrow pollen spectra. The total
arboreal and non arboreal percentages are based on a total pollen sum of which
the aquatic vegetation has been left out (Poaceae are included although the marsh
plant reed belongs to this family and could have been locally present). Although
the diagrams have not been 14C-dated, pollen zones according to Jessen and
Iversen have been applied to the lake samples, based on the stratigraphy of the lake
sediments and the palynological results. The results from the Preboreal (zone IV)
until the Subatlantic (zone IX) will be discussed here. The pollen diagram shows
the regional vegetation development of the area where the barrows described
above are situated in. Although not directly linked in time to the barrows due to
the lack of exact dating, the pollen diagram shows the general development of the
environment of the barrows.
136
ancestral heaths
Sod area 2nd
period (m2)
Radius 2nd
period (m)
470.33
12.24
1422.91
21.28
Table 8.2. The minimum size
of the open space per barrow
based on the sods used to build
the barrows.
northern and central veluwe
137
(9000)-8000 BC
8000-7000 BC
5.1
5 1 1
5.1
Heath
Anthropogenic indicators
Grazing indicators
20 1 5
20 40 60 80 1
20 40 60 20 1 5
20 40 60 80 1
Upland herbs
20 1 5
20 40 20 40 1
20 1 5 5
Aquatic herbs
50 100 150 20 40 60 20 1
20 1 1
20 1 1 5
Ferns and mosses
20 20 40 20 1
Algae
5 1 1
20 1
20 1
20 40 1 1 1 1
20 40 1 1 1 1
VIIAtlantic
1800 2367
VIBoreal
20 20 1610
40 602153
20 40 20 40 60 5
V Boreal
1078 1326
1231 1582
159 622 IV Preboreal
20 20 40 60 20 40 20 40 60 5
20 40 5 1 1 5 1 1 1 5 1 1 5 1
20 40 5 1 1 5 1 1 1 5 1 1 5 1
20 1 1 1 1 1 1 1 1 5 1 1 1 1 1 1 1 1 1
20 1 1 1 1 1 1 1 1 5 1 1 1 1 1 1 1 1 1
50 100 150 1 1 1 1 5 1 5 1
907 1230
1053 1732
862 1263
985 1604
927 1367
1065 1471
1127 1692
1092 1788
1521 2433
2043 3140
859 1435
754 1307 VIIISubboreal
VIBoreal
VIIAtlantic
V Boreal
1078 1326
1231 1582
159 622 IV Preboreal
1800 2367
1610 2153
907 1230
1053 1732
862 1263
985 1604
927 1367
1065 1471
1127 1692
1092 1788
s
20 40 iu m20 40 1 20 1 5 5 50 100 150 20 40 60 20 1 20 1 1 20 1 1 5um 20 20 40 20 1
re
s
ar a
po
or
)
de
s
fl
m
i
u l a ti v
i
a
Upland herbs
Aquatic
herbs
Ferns
and
mosses
Algae
n
io
a rn
m s
ia
ula
at tinu tu m er
re
m is
ch ra
et
ta a r
l ia
iat l te
nd no ava te f lg a
or rsic
ra ma
s
nu ab
fo
ol a
-B u m
i
m
u
t
t
f
e
n
i
a
P
m
d
n
s
u
e
l
l
n
a ca
i e
iu
i a c si
s
un
tr um
v
n
s
(A
o r l ia
s
m pe b p
um n
us
yp
apr e a
s b ul pe
o r tu
ae a e
le
m
l la m m p
es yll ea etmo
cu
dea
m /Ca
a
m ll en
s pm ifo m
utly tiav e la s-t
yru -ty um um la ul u d -ty um a
on
e ce ul ea
ty
th p h
um la) s m
nifl ro
oihi
ata gu m n iu t i u m ie iu iu te a iu m l l u m o c
mli s- sarb i ea ndu psi iu m nthe lus ae lrlaia
dr n su po
erera o niaoc o es
eas ic a p anerac
str etu ede
hai carraum mp tha gon gon ntill ianc numaria ic tr rian haiatai ulm
an rio har pdha motin atu fgeran a laaryec h isetu pod pod pod o le undpo di diu g ineagn yo c
e
s
e
o
t
c
n
m
n
c
u
a
i
r
t
a
l
c
o
e
y
s
a
t
a
h
r
r
m
a
t
i
o
u
a
e
a
l
u
y
n
i as m p
ll
nig i aupb ab il ip ale era eu eli um astio itrtsoic otu tyrasimmyth ela en o ly o ly o tediafon ol a tell ha al e aliftolladaros ott yd oë
ta
tr
d -B em
en M up ynmuno ntanol av latpear yphulgo tr qu yco yco yco on s m o ly ter ela ph o tr
n
a a
r m
s
is e
Zone
Pe ( AP Sscu
To
A ep Br C a C y pe
Te
Po
N aNi P a c si S T v B E L L sL M O P P S Sus B
u mDan E F F G Ga G uHm Hpe bJ pL L s b Lul c Lype M M P P Ptun R Sl e S T V esCt ylCl u D H toHn I m olMia
y
u
l
p
540 1472
p
a ge
n u m ll en
m
i
o a ty
m
ty ia ae u la si s-t m
h ph
iu atif i umu m d iel d iumd iumte d a diu mm el la u m o cc
emus /C
e
yr -ty n umn um l la ul um d a-t rum na
r
o
s
h
t
u
r
a
a
s
h
r
i
b
o
l
o
s
n
e
p
c
r
n
l
t u l n e rel
t
d
m m ha o o nti nc u ri c t ia a um era n oc es
li or c e end o p niu
o a ph mo
a a yc h set po po po o le un po diu in gn yo c
p
d
i
s
a
a
858
2105
n
e
u
n
r
a
e
l
i
n
i
r
g
g
g
o
t
y
h
s
r
a
a
h
t h
r t
m a
918 2120
en
tra o lle ota
ar ph tr u i co co co on m ly eri lag h a tr
di
g i p ba ip le ra u li m s io to tu sim th ela en ly ly te nu l a ell al l e l t ad os tt d ë
en y p m ta
636 1478 IXSubatlantic
Zone
Pe
T
Te
Sp Ty Bo Eq Ly Ly Ly M Os Po Pt Se Sp Bo
Di Eu Fa Fil Ga Ge Ge He Hu Ja Lit Lo Ly Ly M M Po Po Po Ra So St Th Va C a C l Dr Ho Hy Is o
P
M M Nu Ny Po
Sc
540 1472
439 1056
391 891
858 2105
635 1215
918 2120
644 1105
636 1478 IXSubatlantic
1168 1891
439 1056
391 891
1932 3165
635 1215
644 1105
1521 2433
2043 3140
1168 1891
859 1435
754 1307 VIIISubboreal
1932 3165
20 40 60 20 1 5
20 40 60 80 100 20 40 60 5 1
50 100 150 1 1 1 1 5 1 5 1
ae a e
e ce ul ea
ea c a a n a c
ac ass i mp p er
i
A p Br C a C y
Trees and shrubs
e
ae
or
Trees and shrubs
Heath
Anthropogenic indicators Grazing
indicators o ra ta
a
m
li fl
ifl
a
ti c
e
tu
la
l
e
r
a
u
x
u
a
i
e
b
s
eo
in
m
th
tu
l ig
e
nu
nc
gu
iac m
a
l
)
a
e
c
p
e
e
l
s
n
d
m
a
a
s
o u
e
u
a ea
s sa
a a s us -ty
(m
cu s
or
go ae
ce
isi rac e ia a p yr
ea
et
us us
nu
u l u er is
h
nu s us
ex isa
s
bu
m ifleora eal at no o p ale c a era i um nta c e
rc
i ca ul rc
li fl
u u la
m x
aec
u s ng xin ip h n a us tum
pt
p i rnu ryl
m cc
eaP
nu
ue
art Po p Que
rteguAlst C ereolC he Fag Sec Urti A st Gal Pl a Po a
h aSali SamTil iaUlm Bet
ag Fra Fra Ju n Lyc Pi ce Pi n mix
bu Ercicea
s ar C o C o
De A P
iu Nn A
R
C
Ru Su
A
F
Al
Q
h
u
u
c
i
t
t
l
e
a
n
n
g
i
a
l
1.8
a
l
m
a
yp
ae
ae
sc
od um
a
s san
us
ae
tu
la s us -t
s s
go ae
ce
isi c e ia p yr
nu
uc
s la
s gu inu er nis
l u cu
ce
ce
inu us lus
ex isa
s
m ra al no p le a ra um ta e
er
amali x ambil ia lmu etu
m cc
pu uer
gu ran rax u nip ych i cea i nu
rp orn ory
te ste ere he ago ec a rtic ste ali l an o ac
ic a
u
a
r
r
h
a
o
E
R S
C C C
Ru Su
S T U B
A A C C F S U A G P P
F F F J L P P
P Q
Q
20 40 60 80 100 20 40 60 5 1
Figure 8.24. Pollen diagram Uddelermeer
After Polak (1957)
from the Uddelermeer,
composed and redrawn from
three pollen diagrams by Polak
)
(m
s
th
(1959, diagrams I, VI and
p
P
nu
De A P
NA
Al
850 BC-(present)
VII). Pollen zones have been 1.8
applied according to850Jessen
and
BC-(present)
Iversen. A percentage diagram
is shown, with % based on a
3750-850 BC
tree pollen sum minus Betula.
In the AP (= arboreal pollen)
BC
Betula is included. In3750-850
the total
NAP (= non arboreal pollen)
7000-3750 BC
spores are included, non pollen
palynomorphs are excluded.
7000-3750 BC
8000-7000 BC
Different scales have been used,
(9000)-8000 BC
indicated with different colours.
Uddelermeer
After Polak (1957)
In the Preboreal (IV) the percentage of arboreal pollen increased due to a
decrease of herbs like Poaceae, Cyperaceae and Artemisia. The arboreal pollen
percentage consisted mainly of Pinus and Corylus, of which the latter appeared
in the preceding period. In the Boreal (V&VI) period percentages around 80%
of total pollen (minus aquatic plants) were reached. During this period Alnus
and Quercetum-mixtum (e.g. tall deciduous dryland trees) appeared. Alnus reached
percentages of around 25-30% and Quercetum-mixtum increased even further until
35-40%. At the same time Pinus, probably a long distance element at this time,
decreased to around 5% (of ∑AP-Betula). Corylus decreased as well, although less
dramatically until around 20%. Tilia, Ulmus and Fraxinus appeared in this period.
Ericaceae were present with percentages of approximately 2-4%. This situation
remained quite stable until the last part of the Subboreal period, although the
amount of herbs gradually increased. This is mainly due to the increase of Poaceae
until percentages of around 10-15%. The amount of Ericaceae increased slightly
too up to 10%. Anthropogenic indicators like Plantago lanceolata and Rumex rose
up to around 3%. Cereal pollen grains increased until 3-4%.
Towards the end of the Subboreal period (which ends at 800 cal BC, see table
2.1) the arboreal-non arboreal ratio changed in favour of the non-arboreal pollen.
The percentage of non-arboreal pollen increased until around 45%. This is mainly
caused by the further increase of grasses (until around 30%) and Ericaceae (around
25%). A slight decrease in Quercetum-mixtum pollen can be seen, while the
percentage of Alnus pollen seemed to increase slightly. This change in vegetation
composition could be indicative for the influence of humans in the area.
The Subatlantic (zone IX, from 800 cal BC, see table 2.1) started with a further
decrease of total arboreal pollen and an expansion of Ericaceae and Poaceae.
Cereal pollen grains continued to increase slowly as well. Halfway through this
zone Secale appears in the diagram, which probably coincides with the Roman
Iron Period (Behre 1992). At this time there seems to be a slight regeneration
of the forest (mainly Quercetum-mixtum) and some decrease of heath. Then the
non arboreal vegetation expanded again at cost of the forest, with further increase
of heath and cereal pollen, including Secale. The end of the diagram probably
represents the early Middle Ages (according to Polak 1959).
8.12 Summary: the barrow landscape of northern and central
Veluwe
In this chapter the palynological results of barrows at the northern and central
part of the Veluwe have been discussed in order to answer the question: What did
a barrow landscape look like before and after the barrows were built? And, what
was the role of prehistoric human?
Barrows from the late Neolithic A period until the Iron Age were built in open
spaces that generally had an average distance to the forest (ADF) of approximately
50-100m, shown by arboreal pollen percentages of 55-60%. Most herb pollen is
coming from local vegetation. All barrows except one (Stroe; 8.10) were built in
a heath vegetation type, according to the percentages of Calluna vulgaris found
in all pollen spectra. These percentages are on average lowest in the oldest barrow
spectra (around 20%) and highest in the youngest, with percentages up to 100%.
However, percentages over 100% did also occur during the late Neolithic, shown
by the pollen spectra from Ermelo (8.3). This implies that heath was present in the
whole area during the entire period. These heath areas varied from small to rather
large, and in general the heath areas expanded over time. Besides Calluna vulgaris,
the heath vegetation consisted for a considerable part of grasses. Anthropogenic
138
ancestral heaths
indicators are present in all barrow spectra, although in low percentages. The
most dominant anthropogenic indicator is Plantago lanceolata, indicating that
the area had been significantly disturbed by human hands. The open places with
heath vegetation where the barrows were built in were not recorded as such in the
Uddelermeer diagram, indicating the local spread of pollen of heath species. The
Uddelermeer diagram suggests that the vegetation consisted of mainly forest and
human activity was slight. The barrow pollen spectra however, indicate otherwise.
Open places with heath vegetation must have been present in considerable
numbers from the Neolithic onwards.
In all pollen spectra Alnus was the dominant arboreal pollen type. It is very
likely that alder carr forests were present in the wetter parts of the area, probably
the stream valleys. The drier forest in the surroundings consisted mainly of
Tilia, with pollen percentages of 5-20%, Quercus, with pollen percentages of
approximately 10% and Corylus at the forest edge, with pollen percentages of
30-40%. The remaining tree species occur with somewhat fluctuating but low
percentages during the entire period. This general view on forest composition
in the area is also shown by the pollen diagram from the Uddelermeer, where
zone VI-VIII probably represent the situation that has also been registered in the
barrow pollen spectra: the high percentages of Alnus in the wetter parts of the area;
the drier forest consisting mainly of Quercus and Corylus.
In the Middle/Late Iron Age the barrow landscape seems to have changed,
according to the palynological data of the Echoput barrows (8.1). These barrows
were built in much larger open spaces, with an ADF of approximately 200-300
m (arboreal pollen percentage is around 20%). Calluna vulgaris and Poaceae
are, as at the older barrow locations, the dominating species at the open space.
Percentages of Calluna vulgaris now substantially exceed 100%, while grasses
(Poaceae) fluctuate around 60%.
The forest composition in the Middle/Late Iron Age period at the Echoput
was slightly different from the forest composition shown by the older barrows in
the area. The amount of Tilia (pollen percentages of 1-2%) and Corylus (pollen
percentages of less than 20%) seem to have decreased, while Quercus (pollen
percentages until 40%) and Fagus (pollen percentages until 5%) seem to have
increased. In addition, Carpinus has appeared in the pollen spectra. Alder carr is
still present in the wetter areas.
As mentioned above, at the time the Echoput barrows were built, heath
vegetation had expanded in the area. This spreading out of heath vegetation most
likely continued. At the time posts were placed close to the Echoput barrows,
probably in the Medieval period (see 8.1.4), arboreal pollen percentages were
only around 15%. These low percentages indicate an ADF over 600 m (see table
7.2). This large scale expansion of heath in the Medieval Period is also recorded
in the Uddelermeer diagram (when Fagopyrum and Secale have appeared as well).
This is most likely due to the large scale opening up of the landscape caused by
intensified human activities.
In this chapter it has been shown that the barrows from the Late Neolithic
A period until the Late Iron Age were built in heath vegetation. It was also
shown that during the late Neolithic A period long alignments of barrows were
present (8.2 Niersen-Vaassen and 8.3 Ermelo). These barrows alignments were
probably built in long stretched heathland areas, where visibility from one barrow
to the next is likely (Bourgeois 2013, 154-155). The fact that heath and herb
vegetation had already developed at the barrow places, suggests that these long
stretched heath areas were already present some time before the barrows were built.
Moreover, these open spaces must have been kept open until the barrows were
northern and central veluwe
139
built. This also accounts for the smaller heath areas where barrows were built in
that not formed such alignments. It is important to realise that management was
required to maintain these heath areas. This indicates the activity of humans in
the area, at least and perhaps specifically at the places where the barrows were
going to be built. Some open spaces might have been used as settlement area prior
to the barrow building (8.2 Vaassen I and 8.4 Putten). These sites must have been
abandoned for some time before the mounds were raised. For the other barrows in
this region no such indications have been found, nor for the cultivation of crops.
As has been discussed extensively in paragraph 8.1.4 (Echoput) it is likely that
most of these open spaces have been kept open by grazing.
140
ancestral heaths
Chapter 9
The Renkum stream valley
In Chapter 8 palynological analyses of barrows at the northern and central part of
the Veluwe have shown a barrow landscape that was dominated by heath vegetation
that must have been managed for several millennia (from the late Neolithic A
period until the Late Iron Age). In this chapter another group of barrows will
be discussed. These barrows are located in a region further to the south that is
very much comparable to the northern and central Veluwe. This region is also
situated on the Pleistocene push moraine complexes. In Chapter 8 two alignments
of barrows have been discussed and in Chapter 9 another example of a barrow
alignment will be shown.
This alignment with a length of at least 4.5 km12 is situated in a stream valley
near Renkum (Bourgeois 2013, 67-77), in the southern part of the Veluwe (see
figure 9.1). Several of the barrows of the alignment have been excavated and
sampled for pollen analysis as well as three barrows in the same region outside
this alignment. The barrows were all analysed by Casparie and Groenman-van
Waateringe (1980, 24-36), with the exception of Bennekom 1. Bennekom 1 was
published by van Giffen (1954). Section 9.2 presents a new interpretation of the
data retrieved by the above mentioned researchers.
9.1 Site description and sample locations
Burial mounds belonging to the barrow alignment
Renkum 1: A single-period barrow that was excavated in 1929 by Bellen. The
barrow was dated to the late Neolithic A based on the find of a PF Beaker.
Originally the barrow measured approximately 9 m in diameter and 0.80 m in
height. Lanting and van der Waals re-excavated the barrow. Samples were then
taken from the old surface for pollen analysis (Lanting and van der Waals 1972b,
Casparie and Groenman-van Waateringe 1980, 28).
Renkum 2: A single-period barrow in which a PF Beaker was found when it
was excavated in 1929 by Bellen. Based on this PF Beaker the barrow was dated
to the late Neolithic A. During a re-excavation in 1972 by Lanting and van der
Waals samples were taken for pollen analysis (Lanting and van der Waals 1972b).
Samples from the old surface have been analysed (Casparie and Groenman-van
Waateringe 1980, 29).
Renkum 3: A two-period barrow that has been excavated in 1975 by Bakker
and Groenman-van Waateringe (1980, 29). A PF Beaker has been found in the
primary mound, dating it to the late Neolithic A. The barrow measured 15 m in
diameter and 1.8 m in height. Samples that have been analysed were taken from
the old surface of the primary mound and from a sod of the secondary mound.
12
According to Bourgeois (2013, 74 ), two alignments are situated in the stream valley of
Renkum, that possibly formed one long alignment of at least 4.5 km in length.
the renkum stream valley
141
Ede 1
Ede 2
Warnsborn
Bennekom Oostereng
Bennekom 1
Wolfheze
Renkum 2
Renkum 1
Renkum 4
Renkum 3
Doorwerth
Renkum 5
0
750
1500
3000 m
Sampled barrow
Other barrows
m NAP
60 m
0m
Renkum 4: A single-period barrow that has been excavated in 1929 by Bellen.
Lanting and van der Waals have re-excavated the barrow in 1972 and at that time
samples from the old surface were taken for pollen analysis (Lanting and van der
Waals 1972b, Casparie and Groenman-van Waateringe 1980, 29). The barrow
was dated to the late Neolithic A based on the find of a PF Beaker. A 14C-date
of charcoal derived from sods, 2866-2472 cal BC (4065 ± 55 BP, GrN-6712C,
calibrated with Oxcal 4.2), can be used as a terminus post quem date. The barrow
was 15 m in diameter and 1.0 m high.
Renkum 5: A two-period barrow that was excavated in 1958 by van Giffen. A
Veluvian Bell Beaker was found dating the barrow to the Late Neolithic B period.
Measurements of the barrow are unknown. A sample for pollen analysis was taken
from the old surface (Casparie and Groenman-van Waateringe 1980, 36).
Ede 1: A single-period barrow in which a Veluvian Bell Beaker was found and
twelve amber beads during its excavation in 1927 by Bellen. The barrow was dated
to the late Neolithic B. The barrow measured 11 m in diameter and approximately
142
ancestral heaths
Figure 9.1. Location of the
Renkum barrow alignment.
The sampled barrows of this
alignment have been indicated,
as well as the sampled barrows
outside this alignment.
The map is based on digital
elevation model of the AHN
(copyright www.ahn.nl).
0.70 m in height. The mound was re-excavated by Lanting and van der Waals
(1971a). At that time samples were taken for pollen analysis from the old surface
(Casparie and Groenman-van Waateringe 1980, 36).
Ede 2: A barrow that was originally excavated by Bellen in 1927 and re-excavated
by Lanting and van der Waals (1976). The barrow was dated to the late Neolithic
B based on the find of a Maritime Bell Beaker and a 14C-date of 2890-2580 cal BC
(4155 ± 60 BP, GrN-6688C, calibrated with Oxcal 4.2) as a terminus post quem
date for the grave. The diameter of the barrow was approximately 12 m and the
height approximately 0.60 m. Two samples were taken from the old surface, which
were analysed for pollen (Casparie and Groenman-van Waateringe 1980, 34).
Bennekom 1: A multi-period barrow of which the first period most probably
dates to the late Neolithic B. Measurements of the barrow’s size after the fourth
period have been determined to have been approximately 23 m in diameter and
1.70 m in height. Samples for pollen analysis were taken from the old surface of
period 1-5, from a sod of period 1 and from the ditch belonging to period 4 (van
Giffen 1954).
Bennekom Oostereng: A three-period barrow that was excavated in 1929 by
Bursch. The primary mound contained a Veluvian Bell Beaker, a wrist guard and
several flint artefacts, dating the barrow to the Late Neolithic B period. Samples
for pollen analysis were taken during re-excavation in 1972 by Lanting and van
der Waals (1972a). Samples that were analysed originated from the old surface
and a sod of period I, the old surface of the second period and the old surface of
period III (Casparie and Groenman-van Waateringe 1980, 35).
Burial mounds outside the barrow alignment
Warnsborn (Warnsborn 1-6): Six barrows that were situated near Arnhem were
excavated in 1947 and 1948 by Glasbergen and Waterbolk (Waterbolk 1954, 9599; Glazema 1951). Barrow 1 was dated to the Late Neolithic A period, based
on the find of a PF Beaker, a flint axe and a flint blade. Barrow 2 could not be
dated, but Waterbolk mentions that this barrow was similar in structure to barrow
1 and possibly also originated from the Late Neolithic A period. Both barrows
were small and the old surface was barely recognisable. Barrows 3-6 were all
dated to the Early Bronze Age, based on burial typology. This dating, however, is
questionable, given that secure dating is not possible based on burial typology. The
barrows were built of recognisable sods on a Carbic Podzol (Dutch classification:
Humuspodzol). Samples for pollen analysis were taken from the old surface of
Barrows 1-4. Barrow 5, which was a two-period barrow, was sampled at the old
surface and two sods (one sod from each period). From Barrow 6, a three-period
barrow, samples were taken from the old surface of all three periods. All samples
were analysed and published by Waterbolk (1954, 95-99). In 1972 Lanting and
van der Waals re-excavated barrow 1. Charcoal from the primary grave was 14C
dated 3822-2290 cal BC (4435 ± 320 BP, GrN-318, calibrated with Oxcal 4.2).
Samples for pollen analysis were taken by W. Groenman-van Waateringe from
the old surface and from a Bronze Age interment (Casparie and Groenman-van
Waateringe 1980, 24).
Doorwerth: A two- or possibly a multi-period barrow excavated by Hulst in 1972
(Hulst et al. 1973). Grave-goods of the first period included an AOO Beaker,
dating the mound to the late Neolithic B period. Samples for pollen analysis were
taken from the old surface and from a sod from the primary mound (Casparie and
Groenman-van Waateringe 1980, 31).
the renkum stream valley
143
Wolfheze: A two-period barrow at Wolfheze. The barrow was excavated in 1971
by Hulst (Hulst 1971). The first period of the mound was dated to the Bronze
Age based on the find of a Drakenstein urn. Samples for pollen analysis were taken
from the old surface and a sod of period 1, the old surface of period 2 and the old
surface of the secondary mound (Casparie and Groenman-van Waateringe 1980,
37).
9.2 Results and discussion
The analysed barrows belonging to the alignment are all from late Neolithic origin
and they all indicate being surrounded by a rather similar vegetation pattern. All
barrows were built in an open space with heath and grasses. For the Neolithic A
period the ratio AP versus NAP is different between barrows (see figure 9.2 and
9.3), indicating a difference in size of the open space the barrows were built in.
Arboreal percentages fluctuate from around 45% to around 75% suggesting very
small open spaces of approximately 30 m in diameter to larger open spaces with
a diameter of approximately 250 m. A barrow also dating to the late Neolithic A
period that is situated approximately 6 km to the east of the alignment (Warnsborn
1) shows an arboreal pollen percentage in the spectrum derived by Casparie and
Groenman-van Waateringe of approximately 75%. This high percentage suggests
a very small open space of approximately 30 m in diameter. The spectra of barrow
1 and 2 obtained by Waterbolk (Warnsborn 1 and Warnsborn 2) show even higher
numbers of AP, but Waterbolk (1954, 98) mentions that herbal pollen from these
barrows were not investigated with enough care. The AP might have been lower.
As expected, the composition of the forest in the surroundings is similar for all
barrows. Alnus pollen, probably coming from an alder carr in the wetter parts of
the area, dominates the arboreal pollen spectra with 45-50%. The drier forest
consists mainly of Corylus (30%), Quercus (10%) and Tilia (5%). Betula was
present in fluctuating amounts, indicating that solitary Betula trees were probably
present in the heathland area. Barrow 1 of Warnsborn (Warnsborn 1) shows a
slightly lower percentage of Alnus. This could indicate that the barrows belonging
to the alignment were situated closer to an alder carr than the Warnsborn barrow.
The open space this barrow was built in differs also from the alignment barrows
while grasses are the dominating herbs instead of Ericaceae.
In the next phase (late Neolithic B period) the vegetation composition seems
not to have changed in the alignment. Apparently the forest composition remained
unaltered and the open spaces the barrows were built in consisted of mainly heath
and grasses. The size of the open spaces seems in general to be smaller than during
the preceding period with an ADF of approximately 25-50 m. The late Neolithic
B barrow of Doorwerth, situated east of the barrow alignment shows a similar
vegetation composition as Warnsborn 1 with low percentages of Ericaceae.
The vegetation development in the following periods is hard to reconstruct.
Barrows belonging to the Bronze Age period were not investigated in the
alignment. The Bronze Age barrows at Warnsborn (Warnsborn 3-6) show an
expansion of heath at cost of mainly Poaceae compared to the Late Neolithic A
period of Warnsborn 1. The BA barrows of Warnsborn show a similar vegetation
composition as the Neolithic barrows in the barrow alignment discussed above.
Secondary and tertiary periods (undated) of the Warnsborn mounds show a slight
decrease in AP and a slight increase in Ericaceae pollen. A two-period barrow
that is situated approximately 3 km to the east of the alignment (Wolfheze) was
dated to the Bronze Age period as well (the primary mound). This barrow also
shows a similar pollen spectrum. The secondary mound has not been dated. The
pollen spectrum of this period suggests a slight decrease of forest cover and an
144
ancestral heaths
the renkum stream valley
145
AP
20
40
60 80 100
P
NA
Figure 9.2. Pollen spectra from the
samples taken from the barrows of the
Renkum alignment. Spectra are given
in % based on a tree pollen sum minus
Betula pollen. In the total AP (=arboreal
pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are
included, non pollen palynomorphs are
excluded. Different scales have been used,
indicated with different colours.
LNEO-A
LNEO-B
Unknown
Bennekom1_os_per5
Bennekom1_ditch_per4
Bennekom1_os_peri
Bennekom1_os_per3
Bennekom1_os_per
Bennekom1_sod_peri
Bennekom1_os_per1
Renkum5_os_per1
Bennekom_oost_os_per3
Bennekom_oost_os_per2
Bennekom_oost_sod_per1
Bennekom_oost_os_peri
Ede1_os2
Ede1_os1
Ede2_os2
Ede2_os1
Renkum3_os_per2
Renkum3_os2_per1
Renkum3_os1_per1
Renkum4_os2
Renkum4_os1
Renkum2_os2
Renkum2_os1
Renkum1_os2
Renkum1_os1
Renkum
alignment
1
20
s
er n u
Ac Al
40
60 80
r
Co
20 40
s
yl u
5
1
5
20 5
20
us la
m tu
Ul Be
20 5
s
s
s inu s
cu
lix ia
gu ax nu er
F a F r Pi Qu
Sa Til
Trees and shrubs
40
60 80 100
ica
Er
50
ae
ce
a
or
e
1
20 5
5
5
1
20
1
5 10
Upland herbs
20
40
20
20
to
ce
-ty
pe
Aquatic plants Ferns and mosses
60 80
os
et
ac
isa
ex
m
cc
Su
Ru
a
R.
a/
l la
se
Grazing indicators
)
la
e
ria
tu
ar
Be m
e
li a
ica
e ae
g
a
s
o
l
p
r
P
e
if
ce
vu
su
(A
-ty c e
p e ea
st
e
ia
n
m lac
m
um m
ea l a lla ae
la e
ar m gu is
um olle
d i odi u ium
ae cac anu p hy ac e eae du c ea um o nu c u eae h ul tru an ter
nu n s
o
p
e
i
c
g
n
r
n
a
p
p
p
a
d
c
c
c
l
p
i
ll e
ia as s am r yo p e ba i pe mi thr lyg nu sa rop ali ph ryo
ha
ta
co ly
er
Po
L y Po
Pt
Sp
To
Ap Br C C a C y Fa Fil La L y Po Ra Ro Sc Th Ty D
665 1625
718 1473
485 1258
601 998
550 1181
775 1421
533 1038
728 1128
636 1187
632 1196
549 956
704 1144
177 317
120 266
499 977
611 1327
251 576
231 624
285 794
577 1384
615 1214
610 1153
700 1273
496 1154
494 1090
1 5 5 5 1 1 1 1 1 1 5 5 1 1 1
20 1
20
20
20
100 150
l ifl
bu
tu
Anthropogenic indicators
ae
ta
or
la
ifl
e
ul
eo
ea
g
c
i
l
e lanc
a m
i
p
e
e
d u
a
a
a
-ty go
e
ce
ia p o y r
isi ce
ea
m r a real en o go p tica er a lium n ta
ac
te te
t
a Pl a
r
a
e
h
s
o
G
F
Ar As
U
P
C C
A
Heath
146
ancestral heaths
Warnsborn_1_os2
Warnsborn_2_os
Warnsborn_1_os
Warnsborn_3_os
Doorwerth_os_per2
Doorwerth_sod_per1
Warnsborn_5_os_per2
Warnsborn_5_os_per1
Warnsborn_4_os
Warnsborn_6_os_per2
Warnsborn_6_os_per1
Warnsborn_5_sod_per3
Figure 9.3. Pollen spectra from the
samples taken from the barrows outside
the Renkum alignment. Spectra are given
in % based on a tree pollen sum minus
Betula pollen. In the total AP (=arboreal
pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are
included, non pollen palynomorphs are
excluded. Different scales have been used,
indicated with different colours.
LNEO-A
LNEO-B
BA
Wolfheze_os_per1
Warnsborn_6_os_per3
Warnsborn_6_sod_per2
Wolfheze_os_per2
Wolfheze_os_per2
Wolfheze_sod_per1
AP
20 40
60
P
NA
Renkum
Barrows outside the alignment
80 100
s
20
nu
Al
40
60 80
1
20
us
in l us
rp ory
a
C C
40
5
1
1
5
20
5
20
s
s
s inu
cu
s
li x ia
gu ax x nu er
Sa Til
Fa Fr Ile Pi Qu
40 60
Trees and shrubs
5
20
us l a
m tu
Ul Be
40
ic
Er
ae
150
1
5
5
1
1
1
5
Upland herbs
5
20
40 60
1
-ty
pe
Ferns and mosses
5 10
ac
R.
a/
os
t
e
ac
ex i sa
m uc c
u
R S
a
e ll
os
et
Grazing indicators
1
1
1
1
1
1
1
1
1
1
20
1
20
1
5
321
235
356
223
167
224
459
126
107
774
140
121
219 322
693 1203
406 748
780
858
881
415
504
506
634 1002
472 890
505 809
)
la
e
tu
ar
Be
e
e
g
a
m
l
a
P
ae
ce
su
vu
(A
e ace e
e
ia
ce
n
um ium m um um
le
ae cea hyll cea ae cea eae ula e ular um ris
i
l
e
d
s
a
o
d
e
r
a
a
c
c
iu n
po po
lp
ra sic o p ra ce ni ia c n ce ph i ct pt
id g n
ta
co ly
te a s ry pe ba ra m nu sa ro a l ryo
er ha l le
Pt Sp Po
Ly P o
To
As Br Ca Cy Fa Ge La Ra Ro Sc Th D
653 1695
624 1189
511 885
100
10 5
50
e
ac
Anthropogenic indicators
pe
ty
adi ae
ae
e
r
r
o
a
. m fl o
at
li fl
ol
a e r/P li
bu
ce
ce aj o l igu e
tu
n
a
i
a
e
ae
a
od m a yp l
is i ce ia p go ce -t go a e
m ra a l o ta ra um ta e
te ste ere hen la n ste a li la n oa c
r
A
A C C P
A G P P
Heath
Table 9.1. The minimum size
of the open space per barrow
based on the sods used to build
the barrows.
expansion of heath. Another multi-period barrow in the alignment (Bennekom
1) of which the first period was dated to the late Neolithic B, shows in the fourth
period (which was not dated) a similar decrease of arboreal pollen. In this case it
is not the heath that increases, but the grasses and other grazing indicators (e.g.
Rumex, Asteraceae liguliflorae). While this indicates a change in grazing regime,
the development cannot be placed in time.
Most of the barrows discussed above are part of a long alignment of many
barrows. In Chapter 8 it was shown that it is likely that all barrows of a barrow
alignment were built in heath vegetation. The Renkum alignment confirms this
conclusion. Nine out of about 20 barrows of this alignment (Bourgeois 2013,
71-74) were analysed for pollen and the results have shown they were all built
in heath vegetation as well. In addition, all barrows that do not form part of
an alignment, both in the area of Renkum as well as the barrows discussed in
the previous chapter, were built in open spaces that were covered with heath
vegetation. This leads to the conclusion that all other barrows of the Renkum
alignment were built in similar heathland open spaces with an ADF up to 250
m. The barrows of the Renkum alignment were built quite close together with
distances varying between 1030 m to 500 m (Bourgeois 2013, 74) and it is likely
that the open spaces were connected to each other. If this holds true, there would
have been a long and narrow stretched area of heath vegetation with a length of at
least 4.5 km. This type of landscape existed for hundreds or perhaps thousands of
years (spanning the period the barrows were built) and during this long period the
heath must have been maintained by human interference, in spite of the pollen
spectra under discussion, which have not all been dated properly. As has been
explained in section 8.1.4, management is likely to have taken place by grazing,
sod cutting and/or burning. Based on the pollen spectra, grazing seems probable,
as indicated by the presence of Poaceae, Plantago lanceolata and Succisa in the
barrows belonging to the alignment (Hjelle 1999). Assuming an area of 4.5 km
long (Bourgeois 2013), with a width of approximately 60 m (≈27 ha), a livestock herd of 27 sheep and/or 4-5 head of cattle is indicated (see section 8.1.4).
Indicators of heath burning have not been recorded. Sod cutting is indicated at
Diameter (m)
Height (m)
Sod thickness (m)
Renkum 1
9
0.8
Renkum 2
unknown
Renkum 3
15
Renkum 4
15
Renkum 5
unknown
Diameter 4th
period (m)
Height 4th
period (m)
Sod area
(m2)
Radius (m)
0.25
102.86
5.72
1.8
0.25
648.39
14.37
1
0.25
355.52
10.64
Ede 1
11
0.7
0.25
133.76
6.53
Ede 2
12
0.6
0.25
136.17
6.58
Bennekom 1
0.25
Bennekom Oostereng
unknown
Warnsborn 1
unknown
Warnsborn 2
unknown
Warnsborn 3
unknown
Warnsborn 4
unknown
Warnsborn 5
unknown
Warnsborn 6
unknown
Doorwerth
unknown
Wolfheze
unknown
23
1.7
Sod area 4th
period (m2)
Radius (m)
1422.91
21.28
the renkum stream valley
147
least for the purpose of barrow building. Based on the measurement of the barrows,
100-1500 m2 areas were stripped by sod taking to build these barrows (see table
9.1). The method aside, management of such vast heath areas must have involved
a long lasting special interest of prehistoric man at least for the period the barrows
were built. What about the period prior to the barrow building? Heathland was
already present when the oldest mounds were constructed. Heathland vegetation
with Calluna vulgaris and other herbs had developed, indicating that the area
must have been open for some time before. It is not clear when and how this open
area had been created, neither what it had been used for until the barrows were
built. Traces of a settlement have not been found close to the barrow sites and
there are no indications that crop cultivation had taken place prior to the barrow
building. The activity of man is required however, to manage the heath and it is
likely that grazing took place before the mounds were constructed. As has been
discussed for the Echoput and surroundings in Chapter 8, this area too was most
likely part of the economic zone of a farming community, keeping the area open
before, and after, the barrows became part of the landscape.
To recapitulate, in addition to Chapter 8, in this chapter another example of
a barrow alignment has been shown that was built in heath vegetation, possibly
forming a long stretched heathland area that already was in place in the Late
Neolithic A period, an area where grazing might have been important for the
maintenance of the heath.
148
ancestral heaths
Chapter 10
Gooi
The previous two chapters have shown many examples of barrows, including
several barrow alignments, on the push moraine complexes of the Veluwe that
were all built in heath vegetation. In the following chapter another three groups of
barrows (and one solitary barrow) will be discussed. These barrows are situated in
a region in the centre of the Netherlands called Het Gooi (see figure 10.1). These
barrows were also built on a push moraine complex. A more regional vegetation
development covering most of the Holocene could be reconstructed based on a
recently investigated sequence of podzols that was discovered in a nature reserve
area, called the Laarder Wasmeren area (Sevink et al. in press). This area is situated
very close to one of the barrow groups (Hilversum, see figure 10.1).
10.1 Site description and sample locations
Baarn Group
Close to the Lage Vuursche, a small village in the municipality of Baarn, 6 barrows
are situated of which three have been sampled and analysed for pollen (Baarn 13). The results of these analyses have been published by Casparie and Groenmanvan Waateringe (1980, 30-31, 36). Baarn 1-3 were originally excavated in 1927
by van Giffen (van Giffen 1930) and re-excavated and sampled for pollen in 1965
by Addink-Samplonius and Glasbergen. Two barrows (Baarn 1 and 2) are singleperiod barrows that were dated to the Late Neolithic A period. From Baarn 1 two
samples were taken from the old surface. From Baarn 2 one sample from the old
surface and two sod samples were taken. The third (Baarn 3) barrow is according
to Casparie and Groenman-van Waateringe a two-period barrow, although this
could not be confirmed by excavation data. This barrow could not be dated since
no grave goods were found. Samples were taken from the old surfaces of each
period, a sod and a later interment.
Hilversum Group
The second group of barrows is situated in Hilversum. The barrows have been
excavated in 1934 by Bursch (Bursch 1935). Samples for pollen analysis have been
taken from the old surface and sods of three single-period barrows (Hilversum 13) during a re-excavation that has taken place in 1965 by van Giffen and Bakker
(Bakker and van Giffen 1965). Pollen spectra have been published in 1980 by
Casparie and Groenman-van Waateringe (1980, 31-32, 37). Hilversum 1 was
dated to the Late Neolithic B period based on the find of a copper tanged dagger.
Hilversum 2 was dated to the Late Neolithic A or B period based on the type of
burial (northeast-southwest orientated crouched inhumation burial; pers. comm.
Bourgeois). It should be noted that according to Casparie and Groenman-van
Waateringe (1980, 37) this barrow was dated to the Bronze Age, which is now
known to be incorrect. The third barrow (Hilversum 3) was dated to the Bronze
Age, based on a 14C-date of 1609-1436 cal BC (3240 ± 35 BP, GrN-4885, calibrated
with Oxcal 4.2). Measurements of the barrow could not be reconstructed.
gooi
149
Hilversum 1
Baarn 2
Baarn 3
Baarn1
Hilversum 2 & 3
Laarder Wasmeren V
Laarder Wasmeren II
Roosterbos
Laren 3
Laren 2
Laren 1
0
250 500
1000 m
Sampled barrow
Sample location LWM
m NAP
300 m
0m
Laren Group
The third group consists of 10 barrows and is located near Laren. Three of these
barrows (Laren 1-3) were sampled and analysed for pollen (Casparie and Groenmanvan Waateringe 1980, 30, 31, 34). The barrows were originally excavated in
1925/1926 by Remouchamps (1928). The oldest barrow (Laren 1) is a two-period
barrow of which the first period was dated to the Late Neolithic A period based
on the find of PF Beaker and a terminus post quem 14C date of 3139-2890 cal BC
(4385 ± 75 BP, GrN-6683C, calibrated with Oxcal 4.2). During re-excavation
by Lanting and van der Waals in 1971 pollen samples were taken from the old
surface and sods belonging to the first period and from the old surface beneath the
secondary mound. Laren 2 is also a two-period barrow. The old surface and a sod
belonging to the first period were dated to the Late Neolithic B period (based on
a copper tanged dagger), and were sampled for pollen analysis in 1958 (Lanting
and van der Waals 1976). The third barrow (Laren 3) is represented by a pollen
spectrum from the old surface underneath the mound. This is a single-period
150
ancestral heaths
Figure 10.1. Locations of
the barrows in the Gooi area
that were sampled for pollen
analysis and the location of the
Laarder Wasmeren. The map
is based on digital elevation
model of the AHN (copyright
www.ahn.nl).
barrow that was dated to the Late Neolithic B period (based on V-perforated
amber buttons). Sampling took place in 1958 by Bakker and Casparie (Lanting
and van der Waals 1976).
Roosterbos
Approximately 4 km to the northeast of the Lage Vuursche barrows a single-period
barrow is situated in a forest called the Roosterbos. This barrow was excavated
in 1926 by van Giffen. A PF Beaker and a flint scraper were found dating this
barrow to the Late Neolithic A period. The barrow was re-excavated in 1970 for
the collection of palynological samples only. Samples were taken from the old
surface and from a sod. The pollen spectrum of one old surface sample (other
samples were too poor in pollen for pollen analysis) was published by Casparie
and Groenman-van Waateringe (1980, 30).
The Laarder Wasmeren area
In the same region in which the above described barrows were situated, very close
to the barrows of Laren, a nature reserve called the Laarder Wasmeren is situated.
The soil in this area shows three or four podzols on top of each other developed in
layers of drift sand, which were discovered and studied by Sevink et al. (in press).
The Laarder Wasmeren data on soil and sand drifting used in the following are
derived from this study. Based on OSL dates (see table 5.1) a reconstruction of soil
formation and drift sand phases in time could be made. Profile II consisted of four
podzols (S1-S4). S1 has developed in Pleistocene cover sand that was deposited
around 11500 years BP. Around 8800-6500 years BP this soil was covered by drift
sand. In this sand layer another podzol was formed (S2) until it was also covered
by a new layer of drift sand around 6400-5800 years BP. A distinct podzol (S3)
could develop in this layer, which was marked by bioturbation in the form of
presumed beetle burrows. Around 5300-4800 years BP a third layer of drift sand
was deposited on S3. S4 developed in this layer. Profile V consists of three layers;
S1 and S2 probably have merged together at this location (Sevink et al. in press).
Both profiles were sampled for pollen analysis by van Geel (Sevink et al. in press).
The prepared samples were kindly provided to the author of the present work who
(re-)analysed the samples. The results of these analyses are shown in figure 10.3.
The theory and discussion of pollen diagrams derived from mineral soils have
been extensively described in Chapter 5. The site and methods of sampling have
been described more in detail in section 5.2.
10.2 Results and discussion
What now follows is first a reinterpretation of all barrow pollen data, followed
by a presentation of pollen data from the Laarder Wasmeren area analysed by the
author. Following this, all data is combined with the results of the study by Sevink
et al. (in press) and discussed.
Gooi area
The barrow pollen spectra (see figure 10.2a-c) represent three periods: the late
Neolithic-A period, the late Neolithic-B period and the Bronze Age period.
The oldest barrows show an arboreal pollen percentage of 30% (Roosterbos) –
55% (Baarn 1 and 2 and Laren 1). This indicates open spaces with an ADF of
approximately 100 m for the barrows of Baarn and Laren (see table 7.1). The
barrow of Roosterbos possibly was built in a large open space with an ADF that
could reach up to 500 m. However, there seems to be an overrepresentation of
gooi
151
152
ancestral heaths
LNEO-A/B
LNEO-B
BA
s
gu
Fa
20
20 1
20
50
ae
ce
u s la
m t u ric a
Ul B e
E
100 150 200
1
5
1
1
1
1
5
5 10
20 40 60 1
1
5
5
1
5
1
50
-t
lla
se
to
e
ac
R.
ae
a/
ae
os ae llace e a
ce
t
ce ce hy cea ul ula ae eris
a
a p ra nd c e t
c
x
i
n
p
c
e
o
m ass ry pe lipe nu sa y o
Ru Br Ca Cy Fi Ra Ro Dr
100 150
Ferns and mosses
20
20
787
602
151
505
653 1330
464
368
)
la
tu
re
Be m
ga
l
P
vu
(A n s u
m
le
um m
di g nu n su pol
o
l
lyp ph a olle ota
P
S
Po
T
469 1595
Upland herbs
c
i ca
Er
1
5
20 5 10 5
20 40 60
50
100 150 1
5
1
1
1
1
5
20 40 1
1
1
1
1
1
1
1
1
1
1
1
5
5
1
5
486 1201
5
Hilversum2_sod
20 40 1
265 776
Hilversum2_os2
20 40 60 80 100
656 1516
us la
m tu
Ul Be
Hilversum2_os1
s
333 758
y lu
Hilversum1_os
r
Co
lla
se
)
to
la
ce
e
tu
a
ar
Be m
li a
e
R.
g
/
o
l
a
e f
P
a
u
vu
ea sti
(A n s
e ace
os
e
m
et
la e a e la c g u s
ea ll ae
m m
le
ac a eae ca c p hy ace eae du iac cea cu an teri odiu ium nu su pol
x
i
r
n
n
s
n
a
g
c
n
a p
l
c s o
d
e i
i
m cc ia a s ry pe ba ipe ra m nu ph yo lyp eri ha lle ota
T
Ru Su Ap Br Ca Cy Fa Fil Ge La Ra Ty Dr Po Pt Sp Po
594 1515
683 1643
20 40 60 80
s
nu
ta
la
639
Upland herbs Aquatic plants Ferns and mosses
Hilversum3_sod2
Al
ae
or
e
o
bu
ea
ce
tu
ac
pe lan
di
ae
y
o
a
t
o
e
i
e
- g a
is c lia p
m ra a no a m ta e
te te re e tic liu an a c
Ar As Ce Ch Ur Ga Pl Po
lifl
pe
-ty
Anthropogenic indicatorsGrazing indicators
510 1128
P
NA
e
ea
Heath
Hilversum3_sod1
Hilversum3_os
AP
s
s
s in u r a s
cu
gu ax de nu er ilia
Fa Fr He Pi Qu
T
Trees and shrubs
278
683 1190
Baarn2_sod1
Baarn2_os
706 1387
Baarn2_sod2
474 945
5
s
us
in s rcu a
i
ax nu e
Fr Pi Qu Til
20 40 5 10 1
us
in lus
rp ry
Ca Co
20 40 60 80 5
s
nu
Al
e
e
yp
Grazing indicators
Baarn1_os1
20 40 60 80 100
P
NA
a
or
lifl
Anthropogenic indicators
pe
-ty
ia
ed
ae
m
or a
ifl lat
e P.
ul ceo
bu
e a jor/
g
c
u
i
l n
t
ia a
ae la
ae
a
od m
e
isi race alia op tag o a race tag o
ea
ac
tr em ste ere hen l an rtic ste l an
A A C C P U A P
Po
Heath
723 1429
AP
Trees and shrubs
Baarn1_os2
Roosterbos_os
Baarn3_os1
Baarn3_os2
Baarn3_sod
Hilversum
LNEO-A
Unknown
Baarn3_int
Baarn
gooi
153
r
Co
s
1
1
20 40
s
s
s in u s
cu
gu ax nu er
Fa Fr Pi Qu
20 40 1
y lu
1
20
20 40 60 80 1
ae
ce
us la
m tu rica
Ul Be
E
20 1
lix lia
Sa Ti
5
1
1
1
5
1
20
20 40 60 1
20 40 60 1
1
1
5
5
1
1
1
1
20 1
20 5
5
220 446
373 746
241 577
Laren1_os2_per1
Laren1_os1_per1
462 863
20 40 60
s
nu
Ferns and mosses
)
la
e
tu
ar
e
Be m
e
e
g
a
l
p a
P
e
ae
vu
(A n su
e -ty ace
ia c
ce
ar is
ea la yll ae e la
um m um sum olle
ae ica c anu p h ace cea cu eae hul ter etum odi
n
u
i
p
e
n
r
n
g
a
p
c
p
d
iac a ss m ryo pe mi nu sa rop ryo quis olyp teri pha olle otal
E P
P S P
Ap Br Ca Ca Cy La Ra Ro Sc D
T
606 1361
lla
se
Upland herbs
Laren1_sod_per1
Al
o
et
ac
R.
/
a
os
et
ac a
x
s
e i
m cc
Ru Su
pe
-ty
Grazing indicators
742 1035
20 40 60 80 100
P
NA
Anthropogenic indicators
ae
ae
or
ta
or
lifl ae
la
ifl
u
ul
b ce
eo
g
u
li pe anc
t ia
e
e
a y l
a od
a
ce -t go ea e
isi ce p ia
m ra o a l a ra m ta
ac
te te en re tic te liu an
Ar As Ch Ce Ur As Ga Pl
Po
Heath
Laren 1_os_per2
Laren 2_os
AP
Trees and shrubs
Figure 10.2a-c. Pollen spectra from the samples taken from the
barrows in the Gooi area. Spectra are given in % based on a tree
pollen sum minus Betula pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (= non arboreal pollen) spores
are included, non pollen palynomorphs are excluded. Different scales
have been used, indicated with different colours.
LNEO-A
Unknown
LNEO-B
Laren 3_os
Laren
Dryopteris spores, since the percentage is extremely high compared to all other
pollen spectra. When Dryopteris spores are left out of the pollen sum the arboreal
percentage is 57%. This indicates an open spot with an ADF of approximately
100 m, which is comparable to the barrows of Baarn and Laren. In the Neolithic A
period the forest consisted of mainly Corylus (with a pollen percentage of 20-30%),
Quercus (pollen percentage = 15-20%) and Tilia (pollen percentage = 2-5%), with
an alder carr (Alnus) nearby. The pollen spectra from Baarn show a higher pollen
percentage of Alnus (55-70%) than Laren and Roosterbos (45%). It is possible
that the Baarn barrows were situated close to an alder carr. The open spaces the
barrows were built in were covered with mixed heath-grass vegetation at Laren and
Baarn, while grasses and ferns are dominant at Roosterbos. In the following period
(late NEO-B) some changes are visible. Two barrows were built in the area of
Laren (Laren 2 and 3). The open spots were probably larger than around the older
Laren barrows, with an ADF of approximately 100-150 m. Two late NeolithicB barrows were built in the group of Hilversum (Hilversum 1 and 2) in an open
place with an ADF of approximately 100-150 m. The heath vegetation was still
a mixture of Calluna vulgaris and grasses, their ratio more in favour of Calluna.
Betula trees were probably present as solitary trees in the heathland, indicated
by the fluctuating amounts of Betula in the pollen spectra from Hilversum (1060%). The composition of the dry forest was comparable to the late Neolithic A
period with mainly Corylus, Quercus and Tilia. Remarkable is the high percentage
of Alnus in one of the Hilversum barrows (Hilversum 1). Perhaps an alder carr
was situated very close to this barrow, which had then retreated when Hilversum
2 was built. This barrow shows similar percentages of Alnus as the Bronze Age
barrow of Hilversum (Hilversum 3). This barrow was built in an open space with
an ADF of approximately 50-100 m in heath vegetation that was dominated by
Calluna vulgaris. Baarn 3 could not be dated, but it shows in general a similar
vegetation pattern as the dated barrows of Baarn. A difference can be noticed in
the composition of the herbal vegetation. At the time Baarn 3 was constructed
it was dominated by grasses and contained very little Calluna vulgaris. Since the
barrows of Baarn are located quite close together (about 100 m apart from each
other, see figure 10.1b) it can be assumed that they were all built in the same
open space covered with heath vegetation. This would indicate that Baarn 3 was
not built contemporary with the other two barrows, since the herbal vegetation
composition seems to have been fairly different when barrow 3 was built. Another
possibility is that the open space Baarn 3 was built in was situated separate from
Baarn 1 and 2. In that case nothing can be said about the simultaneity of the
barrows. A sample taken from the grave pit of the barrow shows an increase in non
arboreal pollen and Calluna vulgaris. However, it is not very clear where exactly
this sample came from, yet it is difficult to draw any conclusions on this pollen
spectrum. It could indicate an expansion of the open space, with an expansion of
Calluna vulgaris. However, it is also possible that the deceased was buried on top
of a layer of heather twigs.
Clearly open spaces with heath vegetation were present in this area since the
late Neolithic A period. From this period onwards to the Bronze Age not much
changed in vegetation composition. The open spaces varied from approximately 50
to 150 m ADF and consisted mostly of heath and grasses. The surrounding forest
was dominated by Corylus, Quercus and Tilia and alder carr(s) were present in
the environment. Comparable to the Echoput and surroundings (Chapter 8) and
Renkum and surroundings (Chapter 9) this was a landscape that was managed to
maintain its heath vegetation. The method of management could not be deduced
from the pollen spectra. Some anthropogenic indicators were present, but only
154
ancestral heaths
Diameter (m)
Table 10.1. The minimum size
of the open space per barrow
could not be determined for
the barrows of the Gooi casestudy, since measurements of
the barrows were unknown.
Baarn 1
unknown
Baarn 2
unknown
Baarn 3
unknown
Hilversum 1
unknown
Hilversum 2
unknown
Hilversum 3
unknown
Laren 1
unknown
Laren 2
unknown
Laren 3
unknown
Roosterbos
unknown
Height (m)
Sod thickness (m)
Sod area (m2)
Radius (m)
in very low amounts. Grazing could be indicated by the presence of Poaceae and,
although in low amounts, Plantago lanceolata and Succisa.
Laarder Wasmeren area
The pollen diagrams from the Laarder Wasmeren (LWM) area (figure 10.3) show
the vegetation development from approximately 8700 BP onwards, long before
the first barrows were built in the area. The vegetation development per soil
phase, consisting of a phase of deposition and a phase of soil development, can be
reconstructed. The soil phases have been plotted continuously after each other. It
should be noted however that each soil phase ended with a sand drifting period,
probably resulting in a gap in vegetation development between each soil phase.
LWM II – S1 (before 8700 years BP, ca. 6700 cal BC)
The first phase in profile II shows a period in which Pinus was the dominant
species. The presence of large amounts of Botryococcus and ferns suggest the
presence of shallow water at the site. When Pinus and Botryococcus decreased,
Corylus increased. More open vegetation developed with first an expansion of
Poaceae, followed by an expansion of Calluna vulgaris.
LWM II – S2
Arboreal species are dominant in the pollen diagram, with total AP percentages
around 80%. An alder carr developed, as shown by the increasing percentages
of Alnus. A dry forest was present in the surroundings, which consisted mainly
of Quercus, Tilia and Ulmus, with Corylus at the forest edge. Heath vegetation
was present, starting with low amounts (pollen percentages around 10%) and
gradually increasing to pollen percentages around 50%. At the end of this phase
AP had decreased to approximately 50%.
LWM II – S3
AP decreased further until percentages around 40%; the composition of the
forest remained unchanged with mainly Corylus, Quercus, Tilia and Ulmus in the
drier part of the area and alder carr in the wetter surroundings. Heath expanded
together with Poaceae. At the end of this phase Calluna vulgaris is represented
with percentages of more than 100% in the pollen diagram, Poaceae fluctuates
around 30%. Other anthropogenic indicators were present in the area, but only
in small amounts (pollen percentages <1%). Grazing indicators were present as
well, in slightly higher amounts (pollen percentages <5%). This part of the soil
gooi
155
156
ancestral heaths
8690 ± 430
6040 ± 330
5580 ± 290
5410 ± 320
4710 ± 250
O
d
SL
e
at
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
pt
De
)
BP
rs
0
a
ye
s(
h
o
th
Li
gy
lo
Eh
20
AP
40
Laarder Wasmeren
LWM II
60
Bh
80 100
P
NA
A
20
us
ln
40
60
BC
1
20
us
in lus
rp o ry
a
C C
40
60
Ah
80 100
1
1
1
20
lix
he
s
s in u era s
u
x
u
d n
g a
Fa Fr He Pi
40
60
a
gr
ni
1
20
1
5
1
5
5 10
e
s
yp
s
s
cu
s
lu cu
s-t
bu
pu uer
b u alix am ilia lm u
u
o
Q
T
R
S S
U
P
Trees and shrubs
20
av
ul
100
n
la
llu
tu
Be
Ca
200
ri s
ga
300
400
1268
500
Anthropogenic indicators
1
1
1
1
5
ae
or
lifl
m
ae
bu
ru
g
u
ce
t
ni
ia
e
d
a
m
o
a
e
i
t r u i s ac l i a o p
pe tem ter rea en
Em Ar As Ce Ch
Heath
gooi
157
5
1
100
e
ra
iflo
ul
g
li e
ae y p
c e -t ae
ra liu m ce
e
a
t
As Ga Po
200
-ty
pe
Upland herbs
Aquatic and wetland herbs
1
1
5
5
5
5
1
1
1
5 10
5
1
1
5
1
1
1
1
1
1
1
1
5
5 10
1
1
1
500
Algae
rm
pe
a
al
ng
90
173
120
82
450
310
68
685
750
647
598
591
725
634
1057
612
799
529
1093
923
823
622
876
2297
769
933
759
740
130
740
343
356
317
209
328
314
409
350
632
406
585
418
918
797
701
475
437
404
339
385
309
302
es
or
sp
Fungi
NPP
S1
S2
S3
a)
ul
et
B
m
s
P
su
fu
to
(A
n
yp
ed
le
m
b
a 8b 4
l
gl
fi
u
i
8
7
3
o
s
t
2 12 12 31 31 rya 10 18a 18b 165 225 33
0 342
n
lp
t
t iden
t
t
t
t
t
t
t6
t
t9
t
t
ta
lle
G
G
G
G
G G
G
G
G
G
G
G
ba G
To Soil phase
Po
Bv Bv Bv Bv Bv De Bv Bv Bv Bv Bv Bv Un
Bv Bv
341 1951
374 1728
359 1171
323 1122
313
995
300 1162
S4
308 1108
319 1248
308
1862
326
4566
366
925
340
860
390 1282
437 1826
455 1507
372
916
380
880
Ferns and aquatic mosses
e
yp
pe
e
p e -t e
-ty
a ty p . a
ty laria typ
pe
te re
t
t a ar a
y
a u aa
lo a is- / R
t
ca lga
l
e
e
e
te
es
a e
n
a h
n u lcam typ pe
as e
id
ila
rru vu
c e acr to sa
c e typ rop ro p nta
e
a
o
s
e
e
e
n
r
l
a
p
us
s
u
o
l
v
p
a
p m
m a
y
a
o
e
l
g
u Sc
ia
y
e
e
e u m u m oc c
o l u lu ac
m
-ty iu r rh a-t ae ea a m d lar a-t niu es in te
ac hy m e s /
et d i
ha thec ave nell ace iac ifrag n u rg u nic rg a tio t loc h ole
ag un c ex c is a sic yop s tiu c uta itali o ne
ol yp o agn ryoc
t
t
n
n
i
s
a
o
n n um c a
s g s
b x
r
r
h t
a ra i g o
s
l
la e
r
p u
o
en r
M Na Pa Pr Ro Ru Sa So Sp Ve Sp St Tr M
M Po Sp Bo
Pla Ra R Su Br Ca Ce Cu Di Ja
lla
se
to
ce
Grazing indicators
Figure 10.3a. Pollen diagram from the Laarder Wasmeren area: LWM II.
Percentage diagrams are shown, with % based on a tree pollen sum minus
Betula. In the AP (= arboreal pollen) Betula is included. In the total NAP
(= non arboreal pollen) spores are included, non pollen palynomorphs are
excluded. Different scales have been used, indicated with different colours.
te
ancestral heaths
5790 ± 300
4760 ± 220
D
ep
th
iL t h
ol
og
y
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
P
NA
Bh
20 40 60 80 100
Eh
AP
Laarder Wasmeren
LWM V
s(
y
e
ar
sB
P)
da
OS
L
158
BC
20 40 60
s
nu
Al
Ah
1
20
s
s nu s
gu xi nu
Fa Fra P i
20 40 60 80 5
us
r yl
Co
20
1
5
5
is
ar
lg
vu
100 200 300
na
la
t u l lu
Be C a
5 10 5
n
us
s
uc
s
cu
er ali x amb il ia lmu
u
T U
Q
S S
a
igr
Trees and shrubs
5
5
20
1
5 10 5
5
100 200
c
sa
5
20 1
l la
se
eto
-t y
pe
Grazing indicators
p e ac
.
-t y
ris sa/R
o
t
e
u
ul
ac
nc ex isa
nu um u cc
R
S
Ra
Anthropogenic indicators
ae
e
or
ra
l ifl
iflo
m
e
u
u
ul
b
ea
g
gr
c
u
i
i
l
t
ia
n
e pe
m ia eae
od
ea ty e
l ia op
t ru is ac
ac m- a
ea en
er ali u ac e
p e r tem st er
r
t
h
e
s
A G Po
C C
Em A A
Heath
gooi
159
Upland herbs
Aquatic and wetland herbs
Ferns and aquatic mosses
Algae
Fungi
1
5
5 10 5
1
5
1
20
20
1
5 10
1
1
1
5
5
1
1
5
1
5
5
5
20
380
508
682
788
580
737
647
906
715
652
609
509
475
630
441
833
704
435
533
722
452
334
941
130
177
315
337
202
259
273
411
337
311
391
318
306
520
364
426
340
263
356
314
342
230
384
S1/2
S3
es
s
or
pe
re
ty p e e
sp
e
m
po
m
e
n
la)
p
a
u
p
u
r e
e
p ar i -t y t y
t
y
ls
g
tu
e
t
y
a
a
p
f
r
a
s
t
l
g
a
e ic
r y
Be
e
lus
ae p ea ph u p en an a
sif as- t
yp sp lb a
P - su m
at vu lg
un
u
s
e
f
l
t
e
i
e
c
A
t
e
o
c
e
s
(
d
n
e ae
n
ir yp yp o r um a m
a r o ro r
u
oe yp
pe
ps
yll eu /Sc il a mo
ty
e a h -t -t ol yll ea iu in te ium m cc 2 28a 28b 14 17 a 0 8 a 8 b 6 5 25 33 tifie s um o lle
e t yp c e
ium rh -t
ap
p h t a lis p h e
ea g um ic a tr ic ph h a an c h le od n u co
e a a- ica
ec er ll a
1 3 3 y 1 1 1 1 2 3 n
en tal
ac xif ra ifo li ro n ol a yr io m p arg igl o on o lyp hag t r yo G t 6 G t 1 G t G t G t bar G t G t G t G t G t G t id e
ac llo t as s
yo sc u git a ps o sio n ent h art h p av un e
l
i
s
l
r
n
y
To
Ro Sa Tr V e V i M N Sp Tr M P o Sp Bo
Bv Bv Bv Bv Bv De Bv Bv Bv Bv Bv Bv U
C a C u Di Gy Ja
M
N Pa Pr
Ap Ba Br
Po
Zone
315 1186
313 1182
349 854
43 127
143 566
154 531
300 825
69 174
331 643
153 250
311 1153
S4
36
91
120 280
Figure 10.3b. Pollen diagram from the Laarder Wasmeren area: LWM V.
Percentage diagrams are shown, with % based on a tree pollen sum minus
Betula. In the AP (= arboreal pollen) Betula is included. In the total NAP
(= non arboreal pollen) spores are included, non pollen palynomorphs are
excluded. Different scales have been used, indicated with different colours.
profile showed bioturbation. Consequently this part of the pollen diagram could
be showing a mixture of the original vegetation development during this phase.
S3 of profile LWM V, however, shows similar vegetation development and in this
profile bioturbation was not recorded.
LWM II – S4
This phase started 5400 years BP and during it barrows were built in the
Netherlands, including in the surroundings of the Laarder Wasmeren (see above).
In this phase an open landscape existed with non arboreal pollen percentages of
approximately 70%. Heath expanded further with pollen percentages around 200300% and even a peak of over 1000%. The forest in the surroundings consisted
mainly of Corylus, Quercus, Ulmus and Tilia, with alder carr in the wetter areas, as
also shown by the barrow pollen spectra. The levels of anthropogenic and grazing
indicators had increased slightly.
LWM V
In profile V the soil phases S1 and S2 probably have merged together. The oldest
period, with a dominant Pinus presence, appears to be missing in this diagram.
Alder carr in the surrounding area had already developed, as well as the deciduous
forest with Corylus, Quercus, Tilia and Ulmus. The heathland is represented by
pollen of Calluna vulgaris with percentages fluctuating around 50%. The soil
phases S3 and S4 show, as expected, similar vegetation development as LWM II.
The (pre)barrow landscape of the Gooi
The pollen diagrams of the Laarder Wasmeren show a ‘normal’ Holocene forest
development as has been described in section 2.1, starting with high percentages
of Pinus, which decreased at the beginning of the Holocene. When Pinus decreased
Corylus expanded and a deciduous forest developed with mainly Quercus, Tilia
and Ulmus (see Chapter 2). Striking is the relatively open landscape with relatively
high percentages of Calluna vulgaris already before the first sand drift phase around
6500-8800 years BP (4500-6800 cal BC), since the landscape in the Netherlands
was assumed not to have been opened up before the Late Neolithic period (see
also section 2.3.1).
The previous chapters have mentioned the presence of considerable heathland
areas in the Late Neolithic, since the first barrows were built. This investigation
places the occurrence of heath much earlier, to the Mesolithic (Boreal). In
addition, periods of sand drifting as early as 8800-6500, 6400-5800 and 53004800 years BP (based on OSL, Sevink et al. in press; see table 5.1) are remarkable.
Sand drifting could only occur when conditions are unstable. Due to unstable
conditions vegetation becomes scarce and is not able to stabilize the soil. Under
the influence of wind the topsoil is blown away. Periods of sand drifting are
generally linked to human activities. For example due to extensive exploitation
of the soil for crop cultivation, intensive grazing by cattle or sod cutting activities
vegetation disappears, giving wind free play.
The first man-induced sand drifts in the Netherlands are known to have
occurred since the Early Middle Ages (Castel et al. 1989, Riksen et al. 2006),
but perhaps prehistoric man was inducing sand drifts long before then. This
has also been suggested by Willemse and Groenewoudt (2012), who recorded
prehistoric sand drifts along Dutch river valleys. They concluded that these sand
drifts were mainly anthropogenic in the area north of the LWM area (the Westerand Bussumerheide) some Mesolithic artefacts and flint fragments have been
160
ancestral heaths
found, indicating the use of the area by prehistoric man. For the Early and Middle
Neolithic no archaeological finds have been reported (Wimmers et al. 1993) and
also in the LWM area itself no Meso- or Neolithic archaeological artefacts were
found (Sevink et al. in press). The third sand drifting period (5300-4800 years
BP) occurred around the time the first barrows were built a few hundred metres
from the LWM area. Prehistoric man’s activities probably intensified, indicated
by the slightly increased percentages of anthropogenic indicators. It cannot be
determined whether the recorded human activities could induce sand drifting.
The pollen diagrams and barrow pollen spectra only show few anthropogenic
indicators and there are no indications that the area was used for crop cultivation.
Therefore, it is not likely that the area was intensively used. However, given the
constant presence of Calluna vulgaris, the maintenance of the heath by humans is
indicated. This might have been accomplished by grazing, burning or sod cutting,
as has been explained in Chapter 8. Grazing is slightly indicated in the LWM
pollen diagrams and the barrow pollen spectra and it is not unlikely that the heath
area was grazed. Perhaps overexploitation of the heathland was the cause of the
sand drifting. However, Jungerius and Riksen state that these agricultural activities
alone were probably not sufficient to cause large scale sand drifts (Jungerius and
Riksen 2010). They emphasize the role that climate played. A dramatic shift in
climate could bring with it adverse conditions for vegetation establishment and
maintenance, such as in the case of drought. However, in general the Holocene
climate was relatively stable and fluctuations in temperature and precipitation
were probably not sufficient to destroy the vegetation cover (Jungerius and Riksen
2010). Therefore, it is not likely that severe climate change was the cause of the
sand drifts in the LWM area. Jungerius and Riksen (2010) stress that climatic
events such as violent storms were of great importance for the origin of sand
drifts. However, this theory is purely hypothetical (Sevink et al. in press). At this
moment the origin of the sand drifts in the LWM area, anthropogenic or natural
or a combination of both, cannot be determined, although anthropogenic seems
the most plausible explanation (in accordance with Sevink et al. in press).
In the preceding chapters it has been shown that from the Late Neolithic
period onwards, barrows, including long alignments of barrows, were built in
heath vegetation that must have been kept and maintained by human activities.
In general it is assumed that before the Neolithic vegetation was dominated by
forest, with man adjusting their way of life to the landscape. In this chapter it has
been shown that the landscape was already open long before the first barrows were
built, and that Calluna vulgaris was the prevalent species in the investigated area.
This implies a landscape that was managed. The study in this chapter has also
shown that very early periods of sand drifting have occurred in this area of which
the cause may have been anthropogenic. Possibly overexploitation of the landscape
resulted in sand drifting. If Late Neolithic barrow landscape management in
itself was already a remarkable conclusion, it is even more surprising that heath
management probably took place long before. This topic will be returned to in
Chapter 13.
gooi
161
Chapter 11
Toterfout-Halve Mijl and
surroundings
In Chapters 8-10 a number of barrows in three research areas in the northern
half of the Netherlands have been discussed. In the following two chapters the
discussion on the barrows landscape will be continued by investigating several
barrows that are situated in two regions in the southern half of the Netherlands.
Chapter 11 is on the barrows of Toterfout-Halve Mijl and numerous other
barrows situated in an area of about 30 by 20 km (see figure 11.1). A large
number of these mounds have been visited by several researchers performing
palynological analyses (for references see the corresponding sections). In this
chapter the palynological data will be described and discussed to determine the
barrow landscape in the area.
11.1 Toterfout-Halve Mijl
In an area southwest of Eindhoven, close to the two villages of Toterfout and Halve
Mijl, 34 barrows are situated on high cover-sand ridges along a large lake (the
now-drained Postelse Weijer, which still existed up to the 19th century, Glasbergen
1954, 17; see figure 11.1 and 11.2). These barrows were excavated and all dated
to the Bronze Age (Bourgeois 2013, 91-92). More than half of the barrows in this
area have been sampled and analysed for pollen analysis by Waterbolk (Glasbergen
1954, 105-122; Waterbolk 1954, 101-104).
11.1.1 Site description and sample locations
The barrows of Toterfout-Halve Mijl are situated on cover-sand ridges. The old
surface underneath all barrows was the top of a Carbic Podzol (Dutch classification:
Humuspodzol). Samples were taken by Waterbolk from the old surface underneath
the barrows, the sods the mound was constructed of and from the fill of surrounding
ditches. Besides determining the surrounding landscape, the barrows were sampled
with the purpose of dating them (Glasbergen 1954, 28). The relative chronology
based on the palynological results was for a great deal rejected by radiocarbon
dates13 and the surrounding features. Following the well substantiated chronology
proposed by Bourgeois (2013, 93-96), three groups can be distinguished based on
14
C-dating. The first group represents the oldest barrows. In contrary to several
barrows that form part of barrow alignments, described in chapters 8 and 9, these
barrows are extensively dispersed (Bourgeois 2013, 102). Based on 14C dates these
barrows (14, 4 and 1B) were built roughly between 1850 and 1600 cal BC. The
second group represents the youngest barrows (8, 17, 15, 12), which were built
between 1500 and 1250 cal BC. The third group consists of 9 barrows that were
dated in between the first two groups. However, overlap with both occurs. Then
13
Theunissen suggested a relative chronological order based upon radiocarbon dates (Theunissen
1993). These radiocarbon dates have been further calibrated by Bourgeois (2013) based on the
detailed dating program developed by Lanting and van der Plicht (2001/2002).
toterfout-halve mijl and surroundings
163
Toterfout-Halve Mijl
Hoogeloon 1
Hoogeloon 2
Knegsel 1
Knegsel 2
Knegsel Moormanlaan
Steensel
Eersel
Bergeijk
0
1000
2000
4000 m
Sampled barrows
Other barrows
m NAP
40 m
0m
there are 18 barrows that have not been dated by 14C. They have been dated based
upon the surrounding features resulting in a broad spectrum of dates. Some of
these barrows may belong to the group of the oldest barrows, while others might
be relatively young. Not included in the barrow group of Toterfout Halve Mijl by
Bourgeois, but situated in this area and sampled for pollen analysis (Glasbergen
1954, 95-97), is an urnfield. The pollen spectrum of this sample is considered to
represent the youngest period (approximately 800-500 cal BC). An overview of
barrows that have been sampled and the location of the samples in the barrows (e.g.
the old surface, sod and ring ditch) is given in table 11.1. The barrows are placed
164
ancestral heaths
Figure 11.1. Location of the
barrows at Toterfout-Halve
Mijl, Hoogeloon, Knegsel,
Steensel, Eersel and Bergeijk.
The map is based on digital
elevation model of the AHN
(copyright www.ahn.nl).
4
4
9
8A
7
8
22
26
24
20 18
22a
29 28
12
25
21 19
14
23
16 15
5
29 28
30 0
125
250
500 m
30 0
Barrows Toterfout-Halve Mijl
Urn field
m NAP
35 m
Figure 11.2. Locations of
the barrows belonging to the
Toterfout-Halve Mijl group
in detail. The map is based on
digital elevation model of the
AHN (copyright www.ahn.
nl).
1b
3
je r
Wei24
else
Post 26
27
13
9
1a
10
11
17
27
2
6
125
22
20 18
22a
8A
7
8
6
5
1
10
11
12
17
25
21 19
14
23
16 15
250
13
e
Post
500 m
Barrows Toterfout-Halve Mijl
Urn field
m NAP
15 m
35 m
15 m
in chronological order as determined by Bourgeois. Based on their geographical
location the barrows can roughly be divided into three groups (see figure 11.1c).
An easterly group consists of barrows 1-3 (including 1A and 1B), a central group
of barrow 5-11 (including 8A) and a western group of barrow 12-30 (including
22A). All barrows have been extensively described by Glasbergen and Waterbolk
(Glasbergen 1954), some findings should be noted. Glasbergen mentions that two
barrows (12 and 18) were built on and of former arable soil:
“No podsolized surface was found under it (barrow 12) anywhere; like tumulus
18 to be described hereafter it was apparently situated on a plot of prehistoric
arable. No plough markings were found in the subsoil.”
“The barrow (18) was not built on a naturally podsolized subsoil but as a stratum
of made soil, of a dirty grey colour (thickness 0.10-0.14 cm), probably old arable.
(Glasbergen 1954, 62, 72).”
It is however uncertain that such disturbed soil indeed can be interpreted as old
arable, since no plough marks are present. The second finding to be noticed is the
traces of fences that have been found underneath three barrows (14, 20 and 21).
toterfout-halve mijl and surroundings
165
eije
lse W
r
166
ancestral heaths
al
(c
1710 ± 160
1750 ± 125
1730 ± 110
1630 ± 105
1625 ± 100
1575 ± 150
1560 ± 110
1525 ± 75
1540± 125
1475 ± 50
1420 ± 100
1390 ± 110
1410 ± 90
1
4C
)
BC
THM_urnfield
THM22_per4_ ditch
THM17_ os
THM15_ sod
THM8_ os_per2
THM8_ sod_per1
THM8_ os_per1
THM28_ os
THM26_ os
THM25_ os
THM24_ os
THM23_ os
THM22A_ ditch
THM22_ os_per1
THM8A_ os
THM19_os_per2
THM19_ sod_per1
THM 2_os
THM16_ sodper1
THM16_ sod_per1
THM5_ os_per2
THM5_ sod_per1
THM5_ os_per1
THM1_ os (b)
THM1_ os
THM10_ os_per2
THM10_ sod_per1
THM9_ sod
THM9_ os
THM3_ os
THM29_ os
THM21_ sod
THM21_ os
THM20_ os
THM13_ os
THM11_ os
THM7_ sod
THM7_ os
THM6_ os
THM1B_ os
THM1B_ sod
THM1B _pos
THM4_ ditch
THM4_ os
THM14_ os
AP
20
Toterfout-Halve Mijl
40
60
80 100
P
NA
20
s
nu
Al
40
60
80
1
20
us
in lus
rp r y
Ca Co
40
60
5
5
1
20
20
s
s
s in u
cu
s
er
gu ax cea nu
Qu
Fa Fr Pi Pi
40
1
20
lix lia
Sa Ti
Trees and shrubs
40
5
100
us la
m tu
Ul Be
200
l
vu
100
na
llu
Ca
200
r is
ga
300
Anthr. indicators
1
5
1
1
20
20
40
40
60
5
e
ac
ex
isa
cc
Su
20
m
Ru
a
ell
os
et
c
.a
/R
sa
to
-ty
pe
Grazing indicators
Figure 11.3a
20
e
ae
ra
or
a
ifl
flo lat
ul
ae u li
o
b
e
u
c li g
ce
t
a
n
i
e
la
ae
a
od a
e
isi ce lia p ce tago
ea
m ra a no ra
ac
te st e ere he st e
an
r
l
Po
A A C C A
P
Heath
toterfout-halve mijl and surroundings
167
1
5
5
1
1
1
5
5
20
1
1
20
40
ria
e
ar
ca
lg
rsi ae
ae eae
e
e
vu
p ce
ac llac
m
l
m
e m a
s
u
nu hy ae um ea n u cul ae eri
d i diu um
pa op c e ni iac o n ce p t
p o o di
m ar y aba era am o ly g anu osa r yo y co o ly p t er i
a
R R D
C C F G L P
L
P P
Upland herbs
60
173
246
438
381
336
411
573
492
386
280
350
619
539
318
478
581
253
539
231
325
310
246
201
499
462
211
417
275
538
83
351
307
256
)
la
tu
Be
20
Figure 11.3a
1900-1600 BC
1800-1400 BC
um sum
n
gn
lle
ha
p
S
Po
800-500 BC
207
239
304
430
328 1500-1250 BC
474
347
152
435
254
188
1700-1300 BC
478
P(A
Ferns and mosses
Figure 11.3a-b. Pollen spectra from the samples taken from
the Toterfout-Halve Mijk barrows (11.3a) and the Neolithic
settlement (11.3b). Spectra are given in % based on a tree pollen
sum minus Betula pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (= non arboreal pollen)
spores are included, non pollen palynomorphs are excluded.
Different scales have been used, indicated with different colours.
168
ancestral heaths
60
80 100
20 40
60
Co
20 40
60
1
1
5
20
1
5
1
20
20 40
1
1
5
1
20
1
20 40
60
1
1 2
us
ul
nc ca
u
i
n t
R a Ur
5
1
5
Figure 11.3b
20 5
a)
ul
e
et
ar
-B um
lg
P
u
v
(A n s
m
le
is um m
um
ol
su
er di
pt ypo ridi u agn ia len al p
o
t
h ar ol
y ol te
r
o
p
D P P
T
S V P
420 877
Upland herbs Ferns and mosses
552 861
20 40
A
ae
s
us
s la
ce
s rcu i x a
in a
ica
l ili lmu etu
ax ice inu ue
r
a
r
E
S T U B
F P P Q
Grazing indicators
Settlement_os3
N
s
lu
ry
Anthr. ind.
ae
ae
or
a
or
lifl
at
lifl
u
ol
u
b
g
ce
tu
li e
n
la
ae
ae ea
go e
ce i a ce c
ra al ra ra s
ta cea
te ere ste ype otu
an oa
s
l
A C A C L
P P
Heath
407 759
AP
us
ln
Trees and shrubs
Settlement_os2
Settlement_os1
AP
Settlement near Toterfout-Halve Mijl
old surface samples
Sitename
Sample
location
Toterfout Tumulus 14
old surface
Toterfout Tumulus 4
old surface, ditch
Toterfout Tumulus 1b
old surface, sod, present
podsolized surface
Toterfout Tumulus 6
old surface
Toterfout Tumulus 7
old surface, sod
Toterfout Tumulus 11
old surface
Toterfout Tumulus 13
old surface
Toterfout Tumulus 20
old surface
Toterfout Tumulus 21
old surface, sod
Toterfout Tumulus 29
old surface
Toterfout Tumulus 3
old surface
Toterfout Tumulus 9
old surface, sod
Toterfout Tumulus 10
period 1
sod
period 2
old surface
Toterfout Tumulus 1
Toterfout Tumulus 5
Toterfout Tumulus 16
period 1
old surface
period 2
sod
period 1
sod
period 2
sod
period 1
sod
period 2
old surface
MBA-B
LBA/EIA
old surface
period 1
old surface
period 4
ditch
Toterfout Tumulus 22A
ditch
Toterfout Tumulus 23
old surface
Toterfout Tumulus 24
old surface
Toterfout Tumulus 25
old surface
Toterfout Tumulus 26
old surface
Toterfout Tumulus 28
old surface
Toterfout Tumulus 8
MBA-A
ditch
Toterfout Tumulus 8A
Toterfout Tumulus 22
Dating range
EBA
LN B
old surface 2x
Toterfout Tumulus 2
Toterfout Tumulus 19
LN A
period 1
old surface
period 2
old surface
Toterfout Tumulus 17
old surface
Toterfout Tumulus 15
old surface
2600
2400
2200
2000
1800
1600
1400
1200
1000
500
Cal. BC
Table 11.1. Overview of samples taken at the ToterfoutHalve Mijl barrows. Dating ranges for each barrow have
been indicated. Figure after Bourgeois (2013, table 5.5).
toterfout-halve mijl and surroundings
169
11.1.2 Results and discussion
Figure 11.3a shows the pollen spectra of the sampled mounds in the relative
chronological order proposed by Bourgeois.
The oldest group shows the highest arboreal percentages from 55% to almost
80%. The open spaces these barrows were built in had an ADF that varied from 25
to 100 m. The herbal vegetation at these open spots consisted mainly of Calluna
vulgaris and grasses. An exception is barrow 4, which is actually not part of one
of the (geographical) barrow groups, but situated approximately 300 m north
of the central group. Here the vegetation in the open space is a mixture of some
Calluna, grasses and ferns. The youngest barrows show an AP of approximately
55%, so the open spaces seem to be slightly larger in this period (ADF=50-100
m), indicating an expansion of the heath in the area. The sample of the urnfield
shows that the heath at the location of one of the oldest barrows (1B) indeed
expanded (AP=35%) with a Calluna percentage of more than 100%. Not many
changes in landscape seem to have occurred in the period in between. The barrows
that were roughly dated to this period show a similar vegetation pattern. Only
tumulus 4 shows a different vegetation composition of the open space with a low
percentage of heath. This barrow might have been constructed at the edge of the
open space where the heath was grassier. This pollen spectrum was derived from a
ditch sample and the spectrum shows a remarkable high percentage of Pteridium
(bracken) spores. This is also the case for another ditch sample of Toterfout-Halve
Mijl (barrow 22A). Possibly Pteridium was one of the first species to grow on the
barrow after it was built. The ferns might already have shed spores before the
ditches of barrow 4 and 22A were filled up. Close to the barrows alder carr must
have been present, represented in the pollen spectra by high percentages of Alnus.
Surrounding forest consisted of mainly Corylus, Quercus, Tilia and Fagus. Betula
is present in all the pollen spectra in fluctuating percentages. Probably birch trees
were present in the surrounding forest. In addition they were probably also present
in the heathland area close to some of the barrows, causing percentages of over
100% in for example the pollen spectrum of barrow 13.
Open spaces fluctuated between approximately 25 m and 250 m in ADF. Barrows
1A, 1B, 2 and 3 were built very close together. So were barrows 5–8, 10 and 11.
They were most likely built in one open place with heath vegetation. Barrow 1316, 17-20 and 21-29 were also built close together and perhaps these three groups
were built in one large area with heath vegetation. It is not unlikely that all barrows
in the Toterfout-Halve Mijl group (except for barrow 4) were constructed in one
and the same heathland: in a long stretched open space with a minimum length
of approximately 1.5 km. Whether one large heath area or several smaller heath
areas, the heath must have been managed throughout the barrow building period,
as has been discussed for the more northern areas (Chapter 8-10). Grazing being
part of the heath management is likely. This is indicated by the presence of herbal
species such as Plantago lanceolata, Succisa and Asteraceae liguliflorae, although
only represented in low amounts. No evidence for burning of the heath was found.
Charcoal that was found at the site was probably related to funeral activities,
since charcoal was mostly found together with bone material (Glasbergen 1954,
Theunissen 1993). Sod-cutting could have been a heath-management activity,
while sods were cut to build the barrows (see table 11.2). Since the amount of
barrows is enormous, and that a large number of them were built in a relatively
short time period, sod-cutting must have been a regular activity.
170
ancestral heaths
Diameter (m)
THM 1
Table 11.2. The minimum size
of the open space per barrow
based on the sods used to build
the barrows.
Height (m)
Sod thickness
(m)
Sod area (m2)
Radius (m)
20018
25.4
15/22
1.46
0.25
THM 1A
10.4
unknown
0.25
THM 1B
12.2
0.86
0.25
202
8
THM 2
15.8
1.2
0.25
474
12.3
THM 4
16
0.7
0.25
282
9.5
THM 5
110.2
1.15
0.25
191
7.8
THM 7
10.8
0.9
0.25
166
7.2
THM 8
11.2 (after 4th
period)
0.86
0.25
170
7.3
THM 8A
7.4
unknown
0.25
THM 9
7.5
0.8
0.25
THM 10
9
0.6
0.25
72
4.8
THM 11
7.5
0.66
0.25
59
4.3
THM 13
69.3
0.48
0.25
65
4.6
THM 14
12.4
0.72
0.25
175
7.5
THM 16
9.2 (2nd period)
0.68
0.25
91 (2nd period)
5.4
THM 18
6.2
0.45
0.25
27
3.05
THM 19
7
0.62
0.25
48
3.9
THM 20
8
unknown
0.25
THM 21
11.3
0.6
0.25
121
6.2
THM 22
8
1.0
0.25
268
9.2
THM 22A
6.2
unknown
0.25
THM 23
7.4
0.25
0.25
22
2.6
THM 24
6.1
0.25
0.25
15
2.2
THM 25
9
0.3
0.25
38
3.5
THM 26
4.4
0.45
0.25
14
2.1
THM 28
8
0.3
0.25
30
3.1
THM 29
11.5
0.22
0.25
46
3.8
Not much is known about the open spaces for the period prior to the barrow
building. The open spaces were not created just before the mounds were
constructed, since the herbal vegetation had already had some time to develop.
Some of the barrows (12 and 18) were possibly built on of former arable land,
indicating that at least part of the area had been used for crop cultivation prior
to the barrow building. Unfortunately samples taken from these barrows were
unsuitable for palynological analysis. In some of the barrows some cereal pollen
grains and arable weeds like Rumex were found, although in such low amounts
that it cannot be concluded that they were linked to crop cultivation at or close
to the barrow spots. Traces of fences have been found underneath barrow 14,
20 and 21 and could be associated with crop cultivation as well, indicating the
boundaries of a field. Pollen analyses of these barrows show that heath vegetation
was present at the time the barrows were raised and no crops were cultivated close
before the building. Yet, another possibility is that the fences indicate grazing
within enclosures. In all cases it is clear that the area was heavily influenced by
human activities and the area was most likely part of the economic zone of a
farming community. The presence of prehistoric man in the area long before the
barrows were built is also indicated by traces of a late Neolithic B settlement that
were found approximately 60 m northeast of barrow 5 (Glasbergen 1954, van
Beek 1977). A small part of the original soil was preserved. At this location the old
surface, which was overblown by sand shortly after abandonment of the settlement
toterfout-halve mijl and surroundings
171
(for argumentation see van Beek 1977, 48-49), was still recognizable. The old
surface was sampled for pollen analysis by Groenman-van Waateringe. The pollen
spectra are likely reflection of the vegetation composition that was present shortly
after abandonment of the settlement. These pollen spectra show that heath was
already present at that time, although the herbal vegetation was dominated by
grasses (see figure 11.3b). Grazing may have already taken place by then. It is not
clear whether the presumed arable field and the fence traces underneath some of
the barrows, which were found approximately 0.5 km to the southwest, belonged
to Neolithic settlement. It is also not clear where the community moved to after
abandonment of this settlement. Evidence for a Bronze Age settlement that might
belong to the builders of the barrows was not found. Although the function of the
area changed from settlement to burial site it stayed part of the economic zone of
the community living in the area, while the heath was probably grazed.
11.2 Hoogeloon
Approximately 6 km southwest of the Toterfout-Halve Mijl barrow group two
barrows are situated close to Hoogeloon (Hoogeloon 1 and 2; see figure 11.1).
11.2.1 Site description and sample locations
A barrow near Hoogeloon, approximately 4 km from Toterfout-Halve Mijl, called
the ‘Zwartenberg’ (Hoogeloon 1) was excavated in 1950 by Brunsting on behalf
of the ROB (presently known as Cultural Heritage Agency of the Netherlands,
RCE). The mound was dated to the Middle Bronze Age A, based on the find of
a bronze axe in 1846 by Panken. The barrow was constructed of sods that were
still clearly visible during the excavation. Measurements were 18 m in diameter
and 1.4 m in height (Waterbolk 1954, 108; Beex 1964a). A sample from the old
surface was analysed by Waterbolk and published in his thesis (Waterbolk 1954,
103).
Approximately 150-200 m to the west of Hoogeloon 1 a small barrow was
located called the ‘Smousenberg’ (Hoogeloon 2). This barrow was a two-period
barrow of which the first period was dated to the Middle Bronze Age. Its diameter
was approximately 4 m. The barrow was excavated by Beex and a pollen sample
from the old surface was analysed by Waterbolk (Beex 1954).
11.2.2 Results and discussion
Hoogeloon 1 was built in an open space with the forest at an average distance
of approximately 50-100 m. The open space was covered with heath vegetation
that was dominated by Calluna vulgaris (see figure 11.4). The heath was very
poor in other herbal vegetation, including anthropogenic indicators. The area
that was used for sod cutting had a radius of approximately 15 m (based on an
average sod thickness of 0.25 m, see also 8.2.2). Hoogeloon 2, which was probably
younger than Hoogeloon 1, was built in a much smaller open space with an ADF
of approximately 25 m. Calluna vulgaris was also the dominant species in this
small open space. The surrounding forest consisted mainly of Quercus and Tilia.
Fagus was also present in low amounts. Corylus was most likely present at the
edge of the forest. Some Betula trees were probably present as solitary trees in the
heathland or were perhaps part of the forest. In the lower and wetter parts of the
area alder carr was present, represented by high percentages of Alnus in the pollen
spectra from both mounds.
172
ancestral heaths
toterfout-halve mijl and surroundings
173
Figure 11.4. Pollen spectra from the
samples taken from the barrows at
Hoogeloon, Knegsel, Steensel, Eersel
and Bergeijk. Spectra are given in
% based on a tree pollen sum minus
Betula pollen. In the total AP (=arboreal
pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are
included, non pollen palynomorphs are
excluded. Different scales have been
used, indicated with different colours.
LNEO-A
EB A/MB A
MB A-A
MB A-B
MB A
MB A/LB A
IA
Steensel_ditch
Steensel_os
Knegsel_ditchd
Knegsel_ditchb
Knegsel_ditchc
Knegsel_ditcha
Knegsel (F)_79_ditch_upper
Knegsel (F)_79_ditch_middle
Knegsel (F)_79_ditch_under
Knegsel (F)_79_os
Hoogeloon (sm)_138_os
Knegsel (E)_78_sod
Knegsel (E)_78_os
Eersel_133__old arable 2
Eersel_133_old arable1
Eersel_133_sod2
Eersel_133_sod1
Eersel_133_os2
Eersel_133_os1
Hoogeloon (zw)_137_os
Knegsel(moor)_113_o.s.
Knegsel (moor)_113_sod
Bergeijk_403_sod
Bergeijk_403_os2
Bergeijk_403_os1
AP
20 40 60
80 100
P
NA
1
1
r
Co
20 40
s
yl u
60
5
5
1
20
20 40
60
5
20
20 40
lla
se
to
ce
-ty
pe
Grazing indicators
1
s
cu
er
Qu
5 10
li x
Sa
Trees and s hrubs
T il
20 40
ia
us
1 2
m
Ul
Vi
20
lu
Ca
50
na
Upland herbs
20 40
m
la
scu et u
B
l ga
vu
100
ri s
150
E
50
e
ea
ac
ri c
20 5
1
1
1
1
20
1
1
1
1
1
1
1
5
1
1
1
5
5
1
150
ae
or
1
1
1
1
5
e
ea
tu
ac
e
di
a
o
a
si ce l ia p
mi r a a no
te st e er e he
Ar A C C
l i fl
bu
Anthropogenic indicators
20 40 60
5
1
um
gn a
ha ar i
Sp V
a)
ul
et
e P-B um
l
s
ab (A
en
in
m um pol l
er n s
t
l
ta
de ll e
In P o
To
741
196
494 1296
649 1124
517 1037
490 1095
909
470
576 1056
915
533
462 1006
546 1027
663 1005
474 1097
878
506
622 1364
526 1199
605 1027
522 1037
885 1648
539 1012
982
488
850
424
690
332
1017 1560
1005 1567
996 1403
5
Aquatic herbs Ferns and mos s es
100
Heath
e
r ia
.a
e
e
ar
ca
e
/R
r si pe e ae
e typ pe
e a cea
ulg
e
sa
c
e
o
e
m p ty l ac
a l a yl la a e
yp m- s-ty m m v
t
e
p
t
u
e
e
m
u
c
ac n h ce ae a-ty na i um cea pyr nu il la cu ea e um aniu er i di u di u ium
xa
t un c
o o d
sa sic pa op ra ce
t ia n ia
m o
li g pt
me ucci r as am ar y ype a ba eni s ent er a a m e la ol yg ot en a n osa ri fo pa r r yo ycop ol yp te ri
B C C C F
M P P R R
S
Ru
G G G L
T S D L P P
1
s
s inu ra
s
gu ax de e a u
Fa Fr He Pic Pin
ae
or
ta
l ifl
u
ola
ce
li g
e pe l an
a
ce -ty o
ae
r a m ag ace
t e l i u nt
Po
As Ga Pla
20 40 60
s
er nu
Ac Al
Surroundings Toterfout-Halve Mijl
Bergeijk, Knegsel (Moormanlaan), Eersel,
Knegsel, Hoogeloon, Steensel
11.3 Knegsel-Urnenweg
Circa 2 km south of Toterfout-Halve Mijl a cemetery complex is located. An
urnfield was constructed around and partially on top of several older barrows.
The cemetery complex is situated around a small pool, which was drained around
1930. Over several excavations the urnfield was excavated including five of the
older barrows (Braat 1936, Glasbergen 1954). Two of these barrows (Knegsel 1
and 2) and four ring ditches belonging to the urnfield (Knegsel ditch a-d) had
been sampled and analysed for pollen by Waterbolk, with results being published
in his thesis (Waterbolk 1954, 104-108; see figure 11.1).
11.3.1 Site description and sample locations
Knegsel 1 is a three-period barrow of which the first and the second period are
dated to the Middle Bronze Age B. The third period dates to the Early Iron Age.
The diameter of the first period is 7.5 m, of the second 10 m and of the third 8
m. The height of the barrow is unknown, which makes it impossible to calculate
the sod-area. Samples were taken from the old surface of the primary mound and
from a sod originating from the grave pit, belonging to the first period.
Knegsel 2 is a two-period barrow. The first period dates to the Middle Bronze
Age, the second period to the Late Bronze Age/Early Iron Age. The diameter of
the first period barrow is 8 m and of the second 5.4 m. The barrow was 0.28 m
high. Samples were taken from the old surface of the primary mound and from
three consecutive humic layers in the ring ditch.
In addition samples were taken from the fills of four ring ditches that belonged
to the urnfield. Ditch (a) was a circular ring ditch, ditch (b) and (c) belonged to
two long beds (oblong barrows, belonging to an urnfield) and ditch (d) was a
rounded rectangular ring ditch with posts.
11.3.2 Results and discussion
Knegsel 1 and 2 show similar pollen spectra (see figure 11.4). They were both
dated to the Middle Bronze Age-A and it is possible they were built (almost) at
the same time. They were built in an open space with an ADF of approximately
50-100 m. About 28 m2 of heath area needed to be stripped to build the primary
Knegsel 2 barrow (based on an average sod thickness of 0.25 m, see also 8.2.2).
The secondary mound required about 13 m2. The vegetation of the open space
was dominated by Calluna vulgaris with most likely some Betula trees nearby.
Other herbs were almost absent, also Poaceae were only present in low amounts.
Alder carr was present in the river valleys in the environment. Corylus, Quercus
and Tilia were the main trees in the forest that could be found in the drier areas.
Other samples that were taken from this site came from urnfield ditches. Three
of them (a-c) show almost similar AP as Knegsel 1 and 2 indicating an ADF
of approximately 50-100 m. The fourth ditch showed a higher arboreal pollen
percentage of 65%, indicating an open space of approximately 30-50 m. The
forest composition seemed slightly different with a relatively high percentage of
Quercus (30%) at cost of Corylus.
11.4 Knegsel-Moormanlaan
Approximately 3 km southeast of the Toterfout-Halve Mijl barrow group and
approximately 2 km east of the Knegsel barrows a tumulus is located at the
Moormanlaan, a sandy road close to Knegsel (see figure 11.1).
174
ancestral heaths
11.4.1 Site description and sample locations
The barrow at the Moormanlaan is a 2 or 3 period barrow of which the first period
was dated to the Early Bronze Age/Middle Bronze Age-A (diameter=6 m). The
second (and third) period was dated to the Middle Bronze Age (diameter=5.4 m/6
m). The barrow was excavated by Modderman, Verwers and Boogerd in 1967.
Samples for pollen analysis were taken from a sod and from the original surface in
the north-west quadrant by Bakels (Modderman and Bakels 1971).
11.4.2 Results and discussion
The pollen spectra (see figure 11.4) show an arboreal pollen percentage of
approximately 50%, indicating that the barrow was built in an open spot with
an ADF of approximately 100 m. This open spot was mainly covered with heath
vegetation (Calluna vulgaris). Other herbal species are present in very low amounts,
including Poaceae. The surrounding forest consisted of Quercus, Tilia and Fagus
with Corylus and possibly Salix at the forest edge. Alder carr was present in the
wetter parts of the area.
11.5 Steensel
Circa 4 km southeast of Toterfout-Halve Mijl, close to Steensel, an urnfield with
over 100 (urnfield) barrows is situated at a locality called the ‘Heibloem’. This
cemetery has been the subject of several excavations since the first in 1844 by
Panken. In 1948 van Giffen decided to undertake there a trial-excavation to rescue
the cemetery (Modderman and Louwe Kooijmans 1966). At that time samples
for pollen analysis were taken by Waterbolk from one of the ‘long beds’ in the
cemetery, the results of which were published in his thesis (Waterbolk 1954, 103,
109-110; see figure 11.1).
11.5.1 Site description and sample locations
The cemetery is situated on the northern half of a ridge consisting of loamy, fine
sand deposited by wind (Modderman and Louwe Kooijmans 1966). Samples for
pollen analysis were taken by Waterbolk from the old surface and the fill of a ditch
belonging to one of the long beds (Waterbolk 1954, 103, 109-110). No dating
is known for this barrow, but in general long beds are dated to the Late Bronze
Age/Early Iron Age.
11.5.2 Results and discussion
The barrow was built in an open place with an ADF of approximately 125 m,
based on the percentage of arboreal pollen observed in a sample from the old
surface. The pollen spectrum of the ditch seems to represent a slightly younger
period with a higher percentage of herbal pollen, like Poaceae, Rumex and Plantago
lanceolata (see figure 11.4). A high percentage of Pteridium in the ditch spectrum
possibly is the result of a Pteridium being present on top of the barrow, as a
pioneer species, after construction of the barrow and shedding spores before the
ditch was filled up (see also 11.1, p.46-47). Calluna vulgaris was the dominating
species at the open place, indicated by the high percentages of this species in both
samples. Compared to the other barrows in the region the heath was grassier,
indicated by percentages of Poaceae of 20-50%. Betula trees were probably present
in or close to the heathland.
toterfout-halve mijl and surroundings
175
11.6 Eersel
Approximately 5 km to the south of Toterfout-Halve Mijl, close to Eersel a ring
and ditch barrow called ‘De Gloeiende Engelsman’ is situated (Beex 1964b; see
figure 11.1).
11.6.1 Site description and sample locations
The barrow was dated to the Middle Bronze Age-A, based on 14C-dating (3460 ±
35 BP, GrN-5350; 1777-1603 cal BC, calibrated with Oxcal 4.2) and the find of
a Drakenstein urn. The barrow measured 20.2 m in diameter and approximately
1 m in height. It was built partially on an undisturbed Carbic Podzol (Dutch
classification: Humuspodzol) and partially on grey, fairly homogenous soil,
interpreted by van Zeist (1967) as former arable land. This interpretation can be
questioned, given the absence of ploughing marks (see also section 11.1.1) The
tumulus was excavated in 1966 by the ROB and sampled for pollen analysis by
van Zeist (van Zeist 1967b). Samples were taken from the old arable land, from
the old surface underneath the mound (the Carbic Podzol) and from sods with
which the mound was constructed.
11.6.2 Results and discussion
The pollen spectra show that the barrow was built in an open space that was
covered in heath vegetation (see fig 11.4). If the open space had been used for
agricultural activities as was suggested by van Zeist (see 11.6.1), the old arable
was at the time the barrow was built no longer in use as such, indicated by the
high percentage of Calluna vulgaris and the absence of cereal pollen and other
indicators of crop cultivation. Based on the arboreal pollen percentage the average
distance to the forest was approximately 150-300 m. The minimum area that
was used for sod cutting to build the barrow could be calculated. This was an
area of circa 643 m2, indicating a radius of approximately 14 m. Alder carr must
have been present in the neighbourhood of the barrow shown by percentages of
approximately 30% Alnus. Forest in the drier regions mainly consisted of Quercus
and Tilia with Corylus present at the forest edge. The pollen spectra of the old
arable land show higher percentages of Tilia than the other pollen spectra and also
Fagus is present in both samples. Since these samples came from disturbed soil, the
relatively high number of Tilia pollen can be attributed to an older sediment that
was mixed with younger sediment.
11.7 Bergeijk
Approximately 15 km south of Toterfout-Halve Mijl a barrow, close to Bergeijk is
located (see figure 11.1).
11.7.1 Site description and sample locations
The barrow is situated on a high sandy ridge. The barrow was dated to the late
Neolithic-A period based on 14C-dating (3950 ± 150 BP, GRO 381; 2707-2460
cal BC, calibrated with Oxcal 4.2). This is the oldest barrow that will be discussed
in this chapter. The centre of the barrow was sandy and had a diameter of
approximately 3-4 m. Around the centre of the barrow a small ditch was dug from
which the sand was accumulated, forming a small bank encircling the barrow. On
top of this bank a second bank was constructed with sods expanding the diameter
of the monument to approximately 8 m. On top of this bank and barrow a layer
of sand was deposited, enlarging the total tumulus to a diameter of approximately
176
ancestral heaths
20 m and a height of 0.70 m. Samples for pollen analysis were taken by Beex
from the old surface underneath the barrow, from the old surface outside the
secondary bank and from a sod belonging to this bank. The samples were analysed
by Waterbolk (Beex 1957, Waterbolk 1957).
11.7.2 Results and discussion
The mound was probably built in a small open space with an ADF of approximately
25-50 m, based on the high percentage of arboreal pollen (70%; see figure 11.4)).
Part of this open place was probably used for sod cutting. A minimum area of
approximately 630 m2 was required to build the barrow, indicating a radius of
approximately 14 m. The small open place was covered with species-poor heath
vegetation that was dominated by Calluna vulgaris. Quercus and Tilia were the
main species of the surrounding forest, with Corylus dominating at the forest
edge. Alder carr was present in the wetter parts in the surroundings.
11.8 Alphen
A barrow called ‘Op de Kiek’ (Alphen 1) is located approximately 30 km west of
the Toterfout-Halve Mijl barrow group. The barrow was excavated in 1955 by
Modderman (Modderman 1955; see figure 11.5).
Circa 3.5 km to the southwest of Alphen 1 another barrow is present called
‘The Kwaalburg (Alphen 2). It was excavated in 1964 by Beex (1964c; see figure
11.5).
11.8.1 Site description and sample locations
Alphen 1 is a multi-period barrow that was dated to the Middle Bronze AgeA period based on 14C-dating of the primary cremation (3450 ± 60 BP, GrA15479; 1922-1618 cal BC, calibrated with Oxcal 4.2). The inner diameter of the
encircling ditch was approximately 6 m and the original barrow was approximately
1 m of height. Samples for pollen analysis were taken by Modderman from the
old surface underneath the primary mound, outside the primary mound, from
the ring ditch and from the old surface underneath the secondary mound. Results
were published by Casparie and Groenman-van Waateringe (1980, 37, 40).
Alphen 2 was dated to the Middle Bronze Age-A period based on a bronze
flanged axe. This barrow was a so-called bank-and-ditch barrow, meaning that the
original barrow was surrounded by a circular bank and ditch. Alphen 2 was built of
sods and had a diameter of approximately 15 m. At a distance of approximately 1 m
a circular bank with sods of approximately 4 m wide was placed. At approximately
1.5 m from this bank another surrounding bank of approximately 3.5 m wide was
made. The complete monument had a diameter of approximately 41 m. Samples
for pollen analysis were taken during the excavation from the old surface, a sod
and from the encircling ditch and primary bank (Casparie and Groenman-van
Waateringe 1980, 38).
11.8.2 Results and discussion
The pollen spectrum of Alphen 1 and 2 both showed an arboreal pollen percentage
of approximately 70% (see fig 11.6). This indicates that the barrows were built
in a small open space with an ADF of approximately 25-50 m. The vegetation at
the open space was dominated by Calluna vulgaris with most likely some solitary
trees of Betula. Other herbs are, including Poaceae, are only present in very low
amounts. The samples from the ditch and the bank of Alphen 2 show a slightly
different (younger?) vegetation composition, with an expansion of the heath.
toterfout-halve mijl and surroundings
177
Comparable to Toterfout-Halve Mijl and surroundings, the forest in the
environment consisted of mainly Quercus and Tilia. Corylus was present in
considerable amounts at the forest edge. In the lower and wetter parts of the area
Alnus was the dominating tree.
11.9 Goirle
Approximately 2 km to the east of the barrow ‘Op de Kiek’ an alignment of
barrows on a cover sand ridge close to a river valley is situated in an area called
‘Rechte Heide’. Along approximately 1.5 km of this barrow alignment a barrow
is situated that was excavated in 1949 by Glasbergen and Waterbolk (Glasbergen
1954; see figure 11.5).
Alphen 1
Goirle
Alphen 2
0
500
1000
2000 m
Sampled barrows
Other barrows
m NAP
35 m
15 m
178
ancestral heaths
Figure 11.5. Location of the
Alphen and Goirle barrows.
The map is based on digital
elevation model of the AHN
(copyright www.ahn.nl).
11.9.1 Site description and sample locations
A two-period barrow (h=0.90 m, d=15 m) of which the primary mound was dated
to the Middle Bronze Age. The secondary mound was probably almost similar in
age (Bourgeois 2013). From the mound a number of large wall and rim fragments
of a Drakenstein urn were recovered. The monument was heavily damaged by
deep ploughing. The old surface underneath the barrow was strongly affected by
rabbits and intrusion of tree roots. A sample for pollen analysis was taken from
one of the clearly recognizable sods (Waterbolk 1954, 103, 111).
11.9.2 Results and discussion
This barrow was constructed in an open space with an ADF of approximately 100
m. The open space was covered with heath vegetation that was, when compared
to the other barrows discussed in this chapter, quite grassy with a percentage of
Poaceae of 30% (see figure 11.6). Calluna vulgaris is the dominating species with
75%. This barrow was probably situated close to an alder carr, indicated by the
high percentage of Alnus (60%). Sods were cut to build the barrow; a minimum
area of approximately 320 m2 was required to obtain the sods.
11.10 Summary: the barrow landscape of Toterfout-Halve
Mijl and surroundings
From the area around Toterfout-Halve Mijl pollen data are available from the late
Neolithic-A to the Iron Age. The vegetation in the surroundings of the discussed
barrows seems not to have differed greatly from each other during this entire period.
Barrows were built in open spaces with heath vegetation which was dominated by
Calluna vulgaris with in most cases probably some solitary Betula trees. All other
herbal vegetation, including Poaceae, was very low in number. These heath areas
formed, in the case of the Toterfout-Halve Mijl group, most likely long stretched
areas in which groups of barrows were built in the Bronze Age period. The forest
in this area could be divided into two components. In the lower and wetter parts
alder carr was present, indicated by the high percentage of Alnus in all of the
pollen spectra. The forest at the drier parts in the area consisted mainly of Quercus
and Tilia and in the Bronze Age also of some Fagus. As has been discussed in the
previous chapters as well, the activity of man is required to manage the heath. The
method of management in this area is not easy to deduce from the pollen spectra.
Anthropogenic indicators are very low in amount. Some grazing indicators have
been found in the barrows from Toterfout-Halve Mijl. There is no evidence for
burning. Sod cutting is indicated by the barrows, while they were built of sods.
Especially for the amount of barrows being built at Toterfout-Halve Mijl sodcutting could certainly been part of the heath maintenance.
One of the research questions concerns the origin of the open spaces the
mounds were raised in. For the Toterfout-Halve Mijl group the history of its open
spaces is available. Some of the barrows were built on possibly former arable land
(although questionable, see sections 11.1.1 and 11.6.1) and traces of a Neolithic
settlement have been found nearby. After abandonment of the settlement an
open area covered with grasses and some heath was left behind. Possibly the area
was grazed at that time, causing an expansion of the heath in which later the
barrows were built. The construction of the mounds in the area did not stop
prehistoric man from using the area as pastoral grounds, because the heath could
only be maintained by human interference. The destination of the area changed
through time from settlement area with agriculture, to pastoral area, to burial site
combined with pastoral area. On the other barrows discussed in this region no
toterfout-halve mijl and surroundings
179
180
ancestral heaths
Alphen (Kiek)_outsidemound
Alphen (Kwaal)_os1
Alphen (Kwaal)_os2
Alphen (Kwaal)_sod
Alphen (Kwaal)_sod_ditch
Alphen (Kwaal)_sod_bank
Alphen (Kiek)_os_per1
Figure 11.6. Pollen spectra from the
samples taken from the barrows at Alphen
and Goirle. Spectra are given in % based
on a tree pollen sum minus Betula
pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (=
non arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours.
MBA-A
Alphen (Kiek)_os_per2
Unknown
Alphen (Kiek)_ditch
Goirle_os
MBA
AP
20
40
60
Surroundings Toterfout-Halve Mijl
Alphen, Goirle
80 100
P
NA
20
s
nu
Al
40
60
80
20
s
lu
ry
Co
40
60
5
1
1
1
5
20
1
20
s
s
s in u ra
cu
s
lix lia
gu ax de cea nu er
Sa Ti
Fa Fr He Pi Pi Qu
Trees and shrubs
40
1
40
60
5
1
20
l
vu
20
40
60
80
20
ae
ce
ic a
Er
40 60
1
Upland herbs
80 100
5
5
5
1
1
1
1
1
5
5
20
40
re
1
5
949
616
516 1195
507 1068
903
521
567 1044
526 1004
519 1124
587 1224
691 1500
)
la
tu
Be
m
P
su
(A
m um
len
l
u
s
o
gn n
lp
h a lle ota
Sp P o
T
449 1207
Ferns and aq. mosses
1
Anthropogenic indicators
ae
or
ifl
ul
b
tu
ae
a
isi race alia a
m
te st e ere rtic
r
A
A C U
a
lg
ae
a
vu
ce
os
e
et
ea ylla eae e e ris ium m
c
c
a
h c
a a e d
a
ex isa ic p ra c e ce p t o diu
m cc ass r yo pe ba sa yo ly p er i
Ru Su Br Ca Cy Fa Ro Dr Po Pt
Grazing indicators
20 40
na
llu
Ca
ae
ta
or
la
ifl
ul
eo
g
li pe anc
e
l
a y
e
ce -t go
ea
ra um ta
ac
t e ali lan
s
P
G
A
Po
20
us la
m tu
Ul Be
r is
ga
Heath
data are available that can reveal the origin of the open places these barrows were
built in. It is clear that the open spaces were already present some time before
the barrows were built, since heath vegetation had already developed, a process
that in general takes approximately 40 years (Stoutjesdijk 1953). It is likely that
grazing was involved in the maintenance of the heath vegetation already before
the barrows were constructed.
Diameter (m)
Height (m)
Hoogeloon 1
18
1.4
Hoogeloon 2
4
unknown
Knegsel 1
7.5
unknown
Knegsel 2
8
0.28
Knegsel - Moormanlaan
6
unknown
Steensel
Eersel
Table 11.3 The minimum size
of the open space per barrow
based on the sods used to build
the barrows.
Sod thickness (m) Sod area (m2) Radius (m)
0.25
718.26
15.12
0.25
28.19
3.00
unknown
20.2
1
0.25
643.04
14.31
Bergeijk
20
1
0.25
630.41
14.17
Alphen 1
6
1
0.25
58.64
4.32
Alphen 2
15
unknown
Goirle
15
0.9
0.25
319.61
10.09
toterfout-halve mijl and surroundings
181
Chapter 12
Oss-Zevenbergen and surroundings
Near the town of Oss, encompassing an area of approximately 7.5 km2, several
burial complexes are situated from which palynological data have been obtained
(see figure 12.1). The palynological results of these barrows will be described
and discussed to reconstruct the barrow landscape in this area. At the end of
this chapter three pollen diagrams derived from a palaeosoil and peat sediments
(Schaijksche heide, Sint Annabos and Venloop, see section 12.5) will be discussed.
These pollen diagrams will provide more information about the vegetation in the
wider surroundings of the barrows.
12.1 Oss-Vorstengraf area and Oss-Zevenbergen
Close to the town of Oss two burial complexes are situated, Oss-Zevenbergen
and the grave field of the Chieftain’s Grave of Oss (Dutch: Oss-Vorstengraf ).
These two sites have been the subject of various excavations since 1933 when the
Chieftain’s Grave of Oss was discovered. Especially in the last 15 years detailed
research has taken place, revealing that these two sites might actually form one
large burial complex. This will be further discussed in section 12.1.3.
The sites of Oss-Vorstengraf and Oss-Zevenbergen have a long and rather
complex research history, the results of which have been published in several
publications (Verwers 1966, Fokkens and Jansen 2004, Jansen and Fokkens 2007,
Fokkens et al. 2009b, Fontijn and van der Vaart 2013). For a detailed report of all
the research on the two sites the reader is referred to those publications. A short
overview of the several research campaigns and a summary of their findings will be
given in table 12.1. Then a more detailed description per barrow will be given.
12.1.1 Site description and sample locations
Oss-Zevenbergen and the Oss-Vorstengraf area are situated on the northwest edge
of the Peel Blok, a by tectonics elevated (uplifted) area. The grave fields in this area
are for the most part located on a ridge of cover sands. Along the side of the Peel
Block area groundwater seepage wetland occur, causing locally very wet conditions
west, north and east of the cemetery area (Dutch: wijstgronden; see figure 12.2).
The higher parts of the terrain consist of a Carbic Podzol (Dutch classification:
Haarpodzol, while the lower and wetter areas consist of Gleyic Podzols (Dutch
classification: Veldpodzol). Thin layers of wind-blown sand can be found all over
the terrain, especially at the flanks of the barrows. Along the southeast-side of the
terrain an extended drift-sand layer is located (van der Linde and Fokkens 2009,
Jansen and van der Linde 2013)
oss-zevenbergen and surroundings
183
Schaijk
Oss-Zevenbergen
Chieftain’s grave
‘Bursch’ barrows
Vorssel
Schaijkse heide
Nistelrode
Slabroek
Venloop
0
500
1000
2000 m
St Annabos
Uden
Barrow
Sampled palaeosoil/peat/lake
Groundwater seepage wetland
Possible settlement location
m NAP
35 m
0m
Figure 12.1. Locations of the barrows in the casestudy area of Oss-Zevenbergen and surroundings. In
addition the location of a possible Middle Bronze Age
settlement has been indicated, as well as the locations
where groundwater seepage wetlands occur. The
map is based on digital elevation model of the AHN
(copyright www.ahn.nl).
184
ancestral heaths
Year of excavation
Excavator
Results Oss-Vorstengraf area
1933
Bursch
Discovery of the rich Oss-Vorstengraf
1935
Bursch
Excavations of 3 other mounds at Oss-VG complex, one dating to the Late Neolithic and two to the Middle Bronze Age
1964-1965
Modderman & Verwers
Analysis of the cremation remains of Oss-VG: a disabled,
older individual
1969
Beex
1972
1978
Results Oss-Zevenbergen
Research into Oss-Zevenbergen mounds: at least
2 of the 7 mounds are barrows (mound 3 and 7),
5 other mounds were not excavated.
10 barrows at Oss-Zevenbergen, of which 6 were
shown to be built of sods and 4 appeared to
be built of drift sand. 5 mounds are barrows, 5
belong to an urnfield.
Urns were found in the area. 4 ring ditches were observed,
of which one was rectangular.
Van Alphen
1994 – 1997
Discovery of post alignments just north of Oss-Z
(part of a Medieval ‘landweer’: a defence wall)
ROB decided that re-excavation would be best to preserve the archaeological information
1997
Leiden University
A survey of the VG area with test trenches:
•
Rediscovery of the VG: a Hallstatt C grave dug
into a Bronze Age barrow, covered with a new mound.
•
Six-post structure
•
Urnfield (2 ring ditches)
Re-analysis of cremation remains: male, disabled, 40-60 years.
2002
Leiden University
Discovery 3rd ring ditch, 4 urns and a post alignment
underneath the Hallstatt C burial. 4th (rectangular) ditch was
probably too recent to be part of the urnfield.
1998-2005
Leiden University
A survey with test trenches north and northwest of the VG
complex was carried out: a Bronze axe deposition was found
in 2003
2004-2007
ARCHOL BV/ Leiden
University
A survey of the Oss-Z area: all barrows were (re-)
investigated: 3 MBA barrows (4, 2, 6), 2 LBA/EIA
barrows (8,1), 1 Hallstatt C barrow (mound 3).
Barrow 7 could not be excavated yet.
Remains of 5 additional small monuments (ring
ditches) and 4 secondary burials in older barrows
were discovered.
Discovery of 5 post alignments.
2007
University of Leiden
Re-excavation of barrow 6
Excavation of barrow 7: Hallstatt C barrow.
Double post alignment underneath barrow 7
2012
1933-2012
Re-analysis of cremation remains VG: possibly younger and
less disabled than previously thought
Several
Table 12.1. Overview of the research
history of the Chieftain’s grave of Oss
and the Oss-Zevenbergen area.
Several restorations of the grave goods, new discoveries
were made each time.
Oss-Vorstengraf area
Chieftain’s grave of Oss
In 1933 a large barrow with a diameter of approximately 53 m, surrounded by a
circular ditch, was discovered in which a rich Hallstatt C (Early Iron Age) grave was
found: a bronze situla containing amongst others cremation remains, a Mindelheim
sword (an iron sword with a hilt inlaid with gold) and many small bronze objects.
Because of the grave good’s richness the grave was named the Chieftain’s Grave
(Dutch: Vorstengraf) (Bursch 1937). Later research revealed that the cremation
remains were of an older, disabled man, although recent research showed that he
oss-zevenbergen and surroundings
185
Roerdalslenk
Peelrandbreuk
Peel Blok
Nistelrode
Waardse Breuk
groundwater seepage wetland
groundwater seepage wetland
limonite band
0
medium fine and coarse sand
impermeable loam layer
5m
gravel rich coarse sand
medium fine cover sand
might have been much younger and healthier than previously thought (Lemmers
et al. in prep). The bronze objects were probably the remains of bronze horse gear
(Fokkens and Jansen 2004, Jansen and Fokkens 2007, Fokkens et al. 2012). The
Chieftain’s Grave was rediscovered in 1997, although it was heavily disturbed
at this time. The mound itself had disappeared and only the remains of ditches
and posts were preserved. The re-excavation nevertheless revealed that the grave
was a secondary burial into a smaller Bronze Age barrow, which had an original
diameter of 16 m and was surrounded by a ditch. A new barrow was built on top
with the Chieftain’s Grave positioned off-centre in relation to the Bronze Age
barrow, possibly to respect the older grave. The Hallstatt C barrow had a diameter
of 53 m. It was probably 1 m in height above the older mound and flattened at
the top. During the 1997 excavation a fallen tree that had grown on top of the
barrow was investigated. It was discovered that in its fall, the tree had retained
a small intact part of the barrow in between its roots (see figure 12.3 and 12.4).
Although the original mound was levelled in the past, the part of the barrow that
was captured by the tree roots contained a fraction of the old surface, the soil
below and some sods. Samples for pollen analysis were taken from here by de
Kort (1999): three samples from the old surface and four samples from the sods.
In addition two monolith tins were hammered into the section of which samples
could be taken from the old surface downwards to provide a pollen diagram as has
been described in chapter 5. Samples were also taken from the ditch belonging to
the original Bronze Age barrow and from the ditch belonging to the Chieftain’s
Grave. All samples were analysed and published by de Kort as part of his MA
thesis (de Kort 1999).
Three barrows
In 1935 Bursch excavated three other barrows that were situated close to the
Chieftain’s Grave (see figure 12.1). Just south of it a barrow was located that was
dated to the Late Neolithic, based on the find of a Veluvian Bell Beaker. Two other
barrows were surrounded by multiple post circles, which date them to the Middle
Bronze Age. In addition an undecorated Middle Bronze Age urn was discovered
in one of the mounds (Bursch 1937, Fokkens et al. 2012). No samples for pollen
analysis were taken from these barrows.
Urnfield
Some urns were found in 1972 and the discovery of three small circular ring
ditches and four urns without monumental structures in 1997 and 2002 indicated
the presence of a small urnfield southeast of the Middle Bronze Age barrows.
Two ring ditches were found in 1997 just east of the Chieftain’s Grave and had a
diameter of respectively 10 and 7 m. The largest ditch was located about 15 m east
186
ancestral heaths
Figure 12.2. Groundwater
seepage wetlands at the Peel
blok. Figure after van der Laan
et al. (2011).
1
1
2
2
Chieftain’s
grave ditch
Chieftain’s
grave ditch
2
Tree fall
2
Tree fall
1
1
Chieftain’s grave
Chieftain’s grave
0
10 m
Bronze Age ditch
Figure 12.3. Location of the
barrows in the Chieftain’s grave of
Oss area. Figure after Fokkens and
Jansen (2004, figure 4.5).
0
Bronze Age ditch
10 m
Excavated
Excavated
Disturbed
Disturbed
Ring ditch
Ring ditch
(Flat) grave
(Flat) grave
Postholes
Postholes
Tree fall
Tree fall
Sample location
Sample location
Figure 12.4. Tree fall at the Chieftain’s grave of Oss that had captured
a small intact section of the barrow. A fraction of the old surface, the
soil below and some sods were remained in this section. A indicates the
uprooted subsoil from underneath the tree. B, C, D and E together form the
original podsolic soil, with the original topsoil (E), a leached horizon (D)
and the zone with iron pan formation (C). F1, F2, F3 and F4 are sods from
the barrow, laid down with the turf upwards. Figure by H. Fokkens.
200
F4
F3
F1
D
C
B
E
F2
150
100
A
50
0
oss-zevenbergen and surroundings
187
416500
17.2
416400
16
416300
15
416200
14
416100
13
10
12
11
4
167800
1
3
5
9
6
12.2m
+NAP
7
8
2
167900
168000
168100
168200
168300
Figure 12.5. Location of the
Oss-Zevenbergen barrows
and surrounding features.
Figure after van der Linde and
Fokkens (2009, figure 4.4).
of the Chieftain’s Grave and had a maximum depth of 25 cm. A cremation was
found in the centre of the (now disappeared) barrow. The other ditch was located
about 25 m east of the Chieftain’s Grave. A cremation was not found in the centre,
but eccentric at only about 1 m from the ditch. This was probably a secondary
burial, dating the ditch to the Late Bronze Age/Early Iron Age. The northern part
of a third ring ditch was discovered in 2002. In addition the remains of 4 urns
without monumental structures were found (Fokkens and Jansen 2004, Jansen
and Fokkens 2007, Fokkens et al. 2012). Samples for pollen analysis were taken
from the two ring ditches that were found in 1997 (see figure 12.3). One of the
samples from the northern ditch (urnfield ditch sample 2) was useless for pollen
analysis (de Kort 1999).
188
ancestral heaths
Posts
In 1997 a double and partly triple post alignment was found. This post alignment
was located partially underneath the eastern part of the Chieftain’s Grave, dating
the post alignment before the Hallstatt C period. The alignment is probably
related to the Bronze Age burial underneath the Chieftain’s Grave (Fokkens and
Jansen 2004, Jansen and Fokkens 2007, Fokkens et al. 2012).
A six-post structure was found directly north of one of the ring ditches, which
was interpreted as a mortuary house. It was not possible to date this post structure
(Fokkens et al. 2012). None of the posts were sampled for pollen analysis.
Oss-Zevenbergen
Approximately 350 m east of the Chieftain’s Grave a barrow complex including at
least seven burial mounds and several post structures is located (see figure 12.5),
called Oss-Zevenbergen. The barrows date from the Middle Bronze Age to the
Early Iron Age. They are situated on a ridge of cover sands in a southwest to
northeast alignment. Below follows a description per barrow. All information
about these barrows is based on the publication of van Wijk et al. (2009), unless
stated otherwise.
Oss-Zevenbergen 1
In 2004 Barrow 1 was the first to be excavated. The mound itself had mostly
disappeared, but the ditches were for the greater part still recognizable. Barrow
1 is a long bed that measured 4.7 m by at least14 23.5 m. Its height was probably
between 30 and 50 cm. It probably dated to the Late Bronze Age/Early Iron Age
(van Wijk et al. 2009, 73-74). The soil underneath the barrow was a Carbic Podzol
(Dutch classification: Haarpodzol). De Kort took five samples for pollen analysis,
of which three were analysed: two pollen samples of the old surface and a sample
from the fill of the lower part of the surrounding ditch (de Kort 2009, 158).
Oss-Zevenbergen 2
Barrow 2 was recognized as a burial mound in 1964/1965, but not excavated
until 2004. It was situated on the highest part of the cover sand ridge. The barrow
appeared to be a two-period barrow. The primary mound was built on top of a pit
that was filled with thin (5-10 cm) sods. No skeletal remains have been found in
this pit. The mound was constructed of sods with a thickness of 10-15 cm and an
average length of 34 cm. The diameter of the first period was approximately 12.5
m. Its height was probably approximately 60 cm. The mound was surrounded
by a closed spaced single post circle, probably dating the mound to 1700-1300
cal BC (cf. Bourgeois 2013, 34). The old surface belonging to period 1 was,
different from what was underneath the other barrows, an Umbric Podzol (Dutch
classification: Moderpodzol). The secondary mound was also constructed of sods
of which the thickness is unknown. The mound was increased to a height of
approximately 1.2 m and a diameter of approximately 17.5 m. No grave was
found. A closely spaced double post circle was placed around the mound probably
preceding the sod placing (the mound seemed to cover the post holes), dating the
second phase of mound building also to the Middle Bronze Age (1700-1300 cal
BC, cf. Bourgeois 2013, 34). The barrow was re-used in the Iron Age, when an
Iron Age urn with cremation remains was placed in the mound. In the Medieval
14
The barrow was heavily damaged and exact measurements could not be reconstructed (Fokkens et al.
2009).
oss-zevenbergen and surroundings
189
Period another three graves were dug at the base of the mound. Barrow 2 has
been sampled for pollen analysis by de Kort (2009). Samples were taken from the
E- and B-horizon underneath the primary mound, from 2 sods belonging to the
first period, from the old surface underneath the secondary mound and from a
sod belonging to the second period. Another sample was taken from underneath a
grey layer that covered the sods of period 1, but since it is not very clear what this
sample represents it will not be discussed.
Oss-Zevenbergen 3
Barrow 3 is located approximately 40 north of the barrow alignment and situated
in a lower part of the area. It was first discovered in 1964/1965 and excavated
in 2004. It is a single-period barrow with a diameter of approximately 30 m.
Its original height is not exactly known but is conservatively estimated to have
been approximately 90 cm. The mound was constructed of sods with an average
thickness of 8-18 cm and an average length of 50 cm. The central grave consisted
of a large burnt oak plank, some smaller pieces of charcoal, a piece of burned
bone, a small fragment of a bronze sword and fragments of one bronze and two
iron objects; probably a pars pro toto deposition (only parts of an object and/or
the deceased have been buried representing a whole object and/or person). The
barrow was dated to the Hallstatt C period (Early Iron Age), based on 14C dating
of the oak plank and might be contemporary to the Chieftain’s Grave. The soil
underneath the barrow was a Gleyic Podzol (Dutch classification: Veldpodzol).
Samples for pollen analysis have been taken from the old surface and from three
sods by de Kort (2009).
Oss-Zevenbergen 4
Barrow 4 was heavily disturbed and not recognized as a barrow before the
excavation in 2004. Barrow 4 concerns a barrow that was built in four phases. The
first phase consists of a sod layer with a thickness of approximately 15 cm. Before
adding a new layer of sods (phase 2) burning seems to have taken place, indicated
by a high concentration of charcoal fragments in the old surface underneath phase
2. Phase 2 consists of another layer of 10-15 cm thick sods (length about 80
cm). After phase 2 the mound measured approximately 14.5 m in diameter and
approximately 50 cm in height. Another burning event seems to have taken place
after phase 2 as indicated by fragments of charcoal. Charcoal fragments 14C date
this layer to phase 2 to the Middle Bronze Age A (1530-1390 cal BC). The barrow
was increased to a height of approximately 60 cm in phase 3, while the diameter
of the mound was not enlarged. No sods have been recognized in this layer. In the
fourth phase the mound was probably enlarged to a diameter of about 16 m, while
the mound was not heightened. After the last period the mound was covered with
a layer of drift sand. Underneath the barrow a disturbed brown layer was found
on top of which the old surface belonging to the primary mound was situated.
This layer was probably anthropogenic and was interpreted as an old arable layer.
Five samples have been analysed for pollen by de Kort (2009). One sample was
taken from the old surface underneath the drift sand layer at the southern part of
the barrow. Four samples were taken from the old surface belonging to phases 2,
3 and 4. A fifth sample was taken from the disturbed brown layer underneath the
old surface.
190
ancestral heaths
Mound 5
Mound 5 was recorded as a barrow in 1964/1965. During the excavation in 2004
it appeared not to be a barrow but a natural hill formed of drift sand. Two samples
for pollen analyses were taken from the old surface underneath the hill (de Kort
2009).
Oss-Zevenbergen 6
Barrow 6 was first excavated in 1964/1965 by Verwers. The data from this
excavation were reinterpreted by Valentijn (2013). It was discovered that a round
mound was built on top of an oblong monument. Next to the ditch a closely
spaced multiple post setting was found. Pottery sherds together with cremation
remains were found. In 2004 the eastern part of the barrow was re-excavated.
The western part of the monument could not be excavated yet since this part was
situated in a protected zone due to a badger sett in barrow 7 (see next section). It
was concluded that barrow 6 was constructed in two or three phases, but possibly
these construction phases occurred in the same period (the Late Bronze Age/Early
Iron Age). The first (and possibly second) phase consisted of an oblong ditch and
a double ring of posts with a length of 27 m and width of 7.5 m. In the next phase
a round barrow was erected within the eastern part of the oval monument (van
Wijk et al. 2009).
In 2007 the remaining part of the monument could be excavated, revealing
that the oldest peripheral structure is the double post setting, which measures 28.5
by 8.5 m. The post setting probably dates to the Middle Bronze Age B or Late
Bronze Age. During the second phase an oval ditch was dug that cut the inner
post-setting. It was also shown that the round mound probably was the remains
of a disturbed long mound. The long mound was extended on the southern side,
covering the oblong ditch (Valentijn 2013). In 2007 one single sample for pollen
analysis was taken from the ditch, which was analysed by Bakels (Bakels and
Achterkamp 2013).
Oss-Zevenbergen 7
Barrow 7, a large mound with a diameter of about 36 m and a height of 1.5 mm,
was first discovered in 1964/1965. During the excavation campaign of 2004, when
most of the other barrows at Oss-Zevenbergen were investigated, this barrow had
to be left alone. A badger family made the barrow their home and since badgers are
a protected species in the Netherlands Barrow 7 (and part of Barrow 6) could not
be excavated before the badgers had been relocated. Finally, in 2007 the barrow
could be thoroughly investigated. It appeared to be built of sods on a naturally
formed small hill of cover sand and the actual barrow did not measure 36 m in
diameter, but 22.8 m, and was 80 cm high15. On the northern side of this hill
wind-blown sand was deposited in the Middle Neolithic16. Underneath the burial
mound a Carbic Podzol (Dutch classification: Humuspodzol) had developed in the
cover sand. An Early Iron Age urn (Schräghals type) was excavated near the centre
of the mound. The urn was half-filled with cremation remains that appeared to be
from a male in the age of 23-40 years. The bone was 14C dated to the Hallstatt C
period (794-538 cal BC) (Fontijn et al. 2013a, 96; Smits 2013). Very close to the
urn more than 1000 small, bronze studs and large amounts of scattered charcoal
15
16
The original height of the mound could not be exactly reconstructed, but it was presumed that at
least 30 cm of the original top was absent (Fontijn et al. 2013a, 70).
Based on OSL dating by Wallinga and Lemmers, reported in an unpublished thesis (Lemmers 2008),
the deposition took place around 5000 BC.
oss-zevenbergen and surroundings
191
were found. Since the bronze and wood items appeared to be very fragile it was
decided to lift the area with its finds, covering an area of approximately 10 m2, in
blocks to allow for further treatment, preservation and excavation in a laboratory
17
(cf. Fontijn et al. 2013a , 80-81).
The charred wood consisted of oak (94%), ash (5%) and willow (<1%). In the
centre of the mound three charcoal pieces were recovered from the find assemblage.
These charcoal fragments were 14C dated to the Hallstatt C period (Fontijn et al.
2013a, 115-116). Several fragments of burned bone were found in between the
pyre remains as well as two pieces of decorated (animal) bone and an undefined
iron object. The burned bone most likely belonged to the same individual as the
remains in the urn that was buried next to the pyre debris, although this cannot be
confirmed with absolute certainty (van der Vaart et al. 2013, 138-139). The bronze
items probably were the remains of a wagon/horse-gear (yoke decoration) that was
dismantled and then partly burned with the deceased. The burned remains were
partly deposited a little to the east of the pyre and partly left behind (Fontijn and
van der Vaart 2013, 191, 193). The A-horizon was missing under the centre of the
barrow, indicating that the surface was stripped before the pyre was built (Fontijn
et al. 2013a, 114). Altogether, Barrow 7 appeared to be a rich Hallstatt C burial
mound, broadly contemporaneous with the Chieftain’s Grave and Barrow 3.
In the corner of the southwest quadrant traces of an oval pit containing a
large amount of charcoal were discovered underneath the barrow. The pit was
dated to the Middle Bronze Age A based on 14C dating of a piece of charcoal
that was retrieved from the pit fill (Fontijn et al. 2013a, 111-112). Close to this
pit, traces of an (pre-barrow) eight-post structure were found. This feature was
interpreted as an allée, a corridor related to funerary activities, comparable to the
post alignment that was found underneath the Chieftain’s Grave (see 12.1). The
allée might have been related to the funeral activities of Mound 6 (Fontijn et al.
2013a, 110-111).
Samples for pollen analysis were taken from several locations in the barrow.
As a large part of the top surface of the hill was stripped before the barrow was
erected on top of it, sampling of the old surface was difficult, but on top of the
wind-blown sand dune part of the old surface was preserved. A monolith tin that
was driven in this section contained two soils on top of each other of which the
lowest probably contained the old surface underneath the dune and the upper the
old surface underneath the actual barrow. The lower soil did not reveal pollen.
From the results of the upper soil a pollen diagram was derived, based on the
theory described in Chapter 5. In addition eight samples that were taken from
sods were analysed for pollen. Sampling and analysis of these samples was done
by Achterkamp as part of her research master’s thesis (Achterkamp 2009, Bakels
and Achterkamp 2013). In 2009 a bulk sample was taken from the central grave
assemblage by Restaura, the laboratory at which the lifted blocks were investigated.
This sample was analysed for pollen by the author of the present work.
Oss-Zevenbergen 8
Barrow 8 was for a great part excavated in 1964/1965 by Modderman and
Verwers. The last part (northwest quadrant) was excavated in 2004. The results
of both excavations show that Barrow 8 is a single period barrow that measured
approximately 12 m in diameter and 0.6 m in height. It was built of sods, covering
an inhumation grave. The barrow was dated to the Early or Middle Bronze Age,
based on the stretched position of the deceased. The barrow contained two
17
Laboratory of Restaura.
192
ancestral heaths
secondary interments (urns) of which the oldest dates to the Middle Bronze Age
and the youngest to the Early Iron Age. An encircling ditch with a diameter of
9.5 m was dug into the barrow probably when the youngest urn was buried. The
ditch was most likely part of the urnfield that was located northwest of Barrow 8
(see next section; van Wijk et al. 2009, 121-126). Samples were taken for pollen
analysis: two samples from the old surface, one from the ditch fill and one from a
sod18. In addition a monolith tin was driven into the soil underneath the barrow
of which three samples have been analysed (de Kort 2009).
Urnfield, Oss-Zevenbergen 9-12
North of the barrow alignment the remains of a small urnfield were found (see
figure 12.5). In addition to the Early Iron Age ditch at Barrow 8 (the remains of )
four ring ditches were found, called Barrow 9 (d=5 m), 10 (d=7.5 m), 11 (d=4
m) and 12 (d=2.5-2.8 m). Fragments of urns were found in Barrows 10 and 11,
which were dated to the Early Iron Age (van Wijk et al. 2009, 126-131). Samples
for pollen analysis were taken from the ditch belonging to Barrow 10, 11 and 12.
One sample from the ditch of Barrow 12 has been analysed by de Kort (2009).
The samples of the ditch of Barrow 10 have not been analysed and the samples
from the ditch of Barrow 11 did not contain enough pollen for analysis.
Post alignments and post structures
Five post alignments and four post structures were revealed during the excavation
in 2004 (see figure 12.5). Post alignment 1 is situated east of Barrow 3 and
about 116 m long. At the southern part of the alignment (close to Barrow 3)
some additional posts were found, belonging to post structures 1 and 2 (see
figure 12.5). Post alignment 2 is situated in extension of alignment 1, but with
different orientation. Its length is unknown, but at least 18 m and probably 32
m. Post alignment 3 was found east of barrow 4 and has a length of 58 m. Two
extra posts were placed parallel to the alignment, forming post structure 3. A 17
m long alignment of posts is situated between Barrows 6 and 8. The fifth post
alignment is located between Barrow 8 and Mound 5. At the end of this 8 m long
alignment post structure 4 is situated, consisting of 4 posts. The dating of the post
alignments and structures is unknown, but van Wijk et al. argue that they belong
to the urnfield and that they date to the Early Iron Age (van Wijk et al. 2009).
The fill of one of the post features of alignment 1 was sampled and analysed for
pollen by de Kort (2009).
Drift sand layer
Thin layers of drift sand were present throughout the entire Oss-Zevenbergen
area, probably the result of (post) Medieval small scale sand drifting due to the
intensive use of roads. An older layer of drift sand was found in the southeast of
the area. A sample from the old surface underneath this sand layer was analysed
for pollen by de Kort (2009).
18
At the time the samples were taken it was not entirely clear yet whether the barrow was built in
one phase or in two phases. In between the sods a layer with grey-yellow sand was present. It was
not clear whether this layer represented a second building period or that sods were taken from less
developed podzol or that the barrow was built of sods and sand. It was later concluded that the
barrow was built in one phase. The pollen sample was taken from one of the sods taken from less
developed soil/sand layer.
oss-zevenbergen and surroundings
193
One barrow complex or not?
The Oss-Vorstengraf area was at first believed to be separate from the OssZevenbergen barrows. However, it has also been assumed that they formed one
large barrow complex (Fokkens et al. 2009a, 223-224). The area was probably
first used for barrow building in the Late Neolithic, when a barrow was built in
the Oss-Vorstengraf area. In the Middle Bronze Age (A) the burial complex got
its shape, with probably six barrows dating to this period: the Bronze Age mound
underneath the Chieftain’s Grave, two additional barrows nearby and Barrow 2,
4 and 8 in the Oss-Zevenbergen area. In the following period several barrows
were enlarged and/or used for secondary burials. In the Late Bronze Age/Early
Iron Age two additional mounds were constructed (Barrows 1 and 6) and in the
Hallstatt C period (Early Iron Age) three more barrows were added to the now
already extensive burial complex (Barrow 3, 7 and the Chieftain’s Grave). Two
small urnfields were probably contemporaneous to the Hallstatt barrows.
The barrow complex of the Oss-Vorstengraf area and the Oss-Zevenbergen
area might certainly have formed one barrow complex, since they are similar in
time depth. It is however not likely that they physically formed one complex.
West, east and south of the Oss-Zevenbergen area seepage of groundwater occurs,
causing these areas to be very wet. The occurrence of seepage water west of the
Oss-Zevenbergen area creates a natural boundary between the barrow complex of
Oss-Zevenbergen and the Vorstengraf area (R. Jansen pers.comm., March 2013;
see also the introduction of section 12.1.1 and figure 12.1).
12.1.2 Results
Now follows a description of the results per barrow/sampled feature of which the
data have been produced by several researchers mentioned in the previous section.
The data have been reprocessed and reinterpreted by the author. After this section
this reinterpretation will be discussed in section 12.1.3.
Oss-Vorstengraf area
Chieftain’s Grave, old surface and sods (see figure 12.6a)
The pollen spectra from the old surface and sods are very similar, indicating that
sods belong to the same environment as the barrow. They show arboreal pollen
percentages of approximately 55% (ADF= 100 m), except for sod 1 that shows
an arboreal pollen percentage of almost 70%. The arboreal pollen component is
very much dominated by Alnus pollen with percentages of over 65%, indicating
an alder carr was present in the near surroundings. Corylus is well represented
with percentages of 10-20%, being present in the drier parts of the surrounding
forest. Quercus (ca. 5%), Tilia (1-2%) and Fagus (1-3%) pollen, also representing
components of the dry forest, are present in low percentages. Herbal pollen consists
of almost only Ericales, showing that a species-poor heathland was present at the
site.
Chieftain’s Grave, tree fall section (see figure 12.6b)
The pollen diagram shows a vegetation development from the period before the
construction of the Chieftain’s Grave. Arboreal pollen percentages fluctuated
through time between 40% and 80%, indicating a fluctuating ADF between 25
and 150 m. This is mainly caused by fluctuating percentages of Ericales pollen.
Other herbs are almost absent. Alnus shows an increase from about 40% to 60%,
Corylus decreases from 45% to 25%. Tilia decreases while Fagus appears. The
194
ancestral heaths
oss-zevenbergen and surroundings
195
20
40
60
80
20
40 5
5
5
20
5
5
5
20
40
60
80 100
1
1
1
1
5
1
1
5
1
1
1
1
5
Figure 12.6a
1
437 735
80 100
Oss_ChG_sod4
60
403 677
Oss_ChG_sod3
40
360 636
Oss_ChG_sod2
20
527 770
726
Oss_ChG_sod1
Figure 12.6a-c. Pollen spectra from the samples taken from the Chieftain’s grave of
Oss area. 12.6a: pollen spectra from the samples taken from the old surface and sods;
12.6b: pollen diagram from the samples taken from section captured by the roots of
the fallen tree; 12.6c: pollen spectra from the samples taken from the ditches in the
Chieftain’s grave of Oss area. BA= Bronze Age, EIA= Early Iron Age.
Spectra are given in % based on a tree pollen sum minus Betula pollen. In the total
AP (=arboreal pollen) Betula is included. In the total NAP (= non arboreal pollen)
spores are included, non pollen palynomorphs are excluded. Different scales have been
used, indicated with different colours.
Hallstatt C
s
s
ae
s inu s
cu
us la ce
gu ax nu er
ia m tu ica
Fa Fr Pi Qu
Til Ul Be Er
510 1011
s
lu
ry
Co
a
el l
os
et
Oss_ChG_os3
s
nu
Al
-ty
pe
Anthr. ind. Grazing ind.Upl. herbs Ferns and mosses AlgaeNPP
es
or
e
)
sp
ae
or
la
r
n
r
c
o
a
sp
tu
a
fe re
lifl e olat
Be m
R.
e lga ern
u
/
t
a
P
f
b e e
e la u
sa
(A n su
tu iac nc
yp si v te
to
ae od la
ce ula a-t e p ium sila m
lle
a
um
isi ace op tago eae ex a end ntill olet od e p gnu rya t58 n s l po
m r
n lyp ilet ha ba G lle
ta
te te en an ac m lip te
Ar As Ch Pl Po Ru Fi Po Mo Po Tr Sp De Bv Po
To
423 737
Heath
404
P
NA
Trees and shrubs
Oss_ChG_os2
Oss_ChG_os1
AP
Oss Chieftain's grave
Old surface , sods and grave
ancestral heaths
20 40
20
20
5
20
50
100 150
1
1
1
1
5
5
20
1
1
1
1
5
1
1
1
5
554
671
Figure 12.6b
1
310
5
62
5
331
59
20 40 60
409 786
57
458
506
357
55
603
633
ae
ce
us la
m etu
ica
l
r
U B
E
393
s
cu
er
ia
Qu
Til
396
s
s inu s
gu ax nu
F a F r Pi
52
20 40 60 80
s
53,5
60 80 100
ylu
573
r
Co
Upl. herbs Ferns and mosses Algae NPP
350
20 40
us
Grazing ind.
338
n
Al
Anthr. ind.
pe
ty
es
aor
ell
s
re
sp
to
e
a)
ae
e
r
n
a
po
r
ul
r a
c
o
fe are rn s
et
.a e
flo lat
lifl
i
e
R
e
l
-B
g
u
/
t ul
m
a u eo
e
a
P
f
a
b
a
e
e
l
e
s
g
c
su
c
(A
tu
yp si v ate
n
to ac
ia e li lan
ae
ce yll ula a-t e p ium sil m
a
u m olle
od a
isi race alia op race tago eae ex a oph nd ntill olet od e p gnu r ya t 58 n s
p
l
t
p
m
e
a
e
ta
te te re en te an ac um ary ilip ote on oly rile pha eb vG oll
R C F P
Ar As Ce Ch As Pl Po
M P
To
T S D B P
403 703
398 710
374 691
363 645
322 570
323 808
306 485
324 587
Heath
50
P
NA
Trees and shrubs
48,5
40
41
42
43
44
45
46
47
AP
Oss Chieftain's grave
Tree fall profile
Depth (cm)
196
oss-zevenbergen and surroundings
197
BA
EIA
ChG_BAditch
ChG_urnfield_ditch2-1
ChG_urnfield_ditch1-2
ChG_urnfield_ditch1-1
Ditches
Oss Chieftain's grave
AP
20
40
60
80 100
P
NA
20
s
nu
Al
40
60
s
20
lu
ry
Co
40
60
5
5
5 10
s
s inu s
gu ax nu
Fa Fr Pi
20
20
20
ae
ce
us la
ca
m etu
i
l
r
U B
E
5 10 5
s
cu
er
lia
Qu
Ti
Trees and shrubs
40
60
80 100
a
or
e
5
1
1
5
1
1
20
ae
or
Anthr. ind.
1
5
1
1
1
l
se
to
1
-ty
la
pe
Grazing indicators
Algae
Aq. herbs
Upland herbs Ferns and mosses NPP
5
1
1
1
350 578
Figure 12.6c
1
319 633
354 695
e res
pe yp o
ty m-t n sp ore
a)
e
e
p c
ul
r sp
a
ia u
et
ar ers e fe rn
at
-ty R. a
fl
lifl
B
l
i
e
s
b
l
i
u
/
t fe
m
r
P
eo
b
ea u
su
pe lum em sila te
ac os a
(A
tu
ac lig e nc
n
e
-ty co m p la
u s et
m
di e p la
lle
ul ac ae illa sa aniu ete psi num ya 8 su
e a lia po ce a -ty go ae
o
c
c
x
l
t
e
p
m
o
r
a
n e
o ra
e
g o te g a
t
c n i
l
ra a
t5 n
te re en te aliu an ac anu um pia ote cab par on rile pha eb vG olle ota
R R A P S S M T S D B P
As C e C h As G Pl Po
T
365 738
Heath
198
ancestral heaths
291 520
OssZ1_ditch
50
100
150
1
5
20 5
5
1
1
1
1
5
1
314 533
OssZ1_os2
5 10 1
5
s
us la le
m tu ica
Ul Be Er
1
20
5
1
5
5
20 40
60
20 40
80 100
20 40 60
Figure 12.7. Pollen spectra from the
samples taken from Oss-Zevenbergen
barrow 1. Spectra are given in % based
on a tree pollen sum minus Betula
pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (=
non arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours. LBA= Late Bronze
Age, EIA= Early Iron Age.
LBA/EIA
OssZ1_os1
Oss-Zevenbergen
Barrow 1
P
NA
Oss-Zevenbergen 2 (see figure 12.8)
The oldest period, which is represented by the sample from the B-horizon, shows
a non arboreal pollen percentage (NAP) of about 70%, indicating the open space
had an ADF of approximately 300-500 m. The high NAP is mainly the result of
a very high percentage of Ericales pollen of over 200%. Some other herbs like
Poaceae are present although only in low percentages of less than 5%, indicating
that the heathland was poor in species. The arboreal pollen component consists of
mainly Alnus (ca. 45%), Corylus (ca. 30%), Quercus (ca. 10%) and Tilia (ca. 3%).
The following period, represented by the sample taken from the E-horizon, shows
a higher arboreal pollen percentage of about 70%, indicating that the open space
was probably much smaller at this time. The arboreal component is comparable
to the B-horizon, except for an increased percentage of Corylus (ca. 45%). The
percentage of Ericales decreased to about 40%. The following periods, represented
by respectively the sods of period 1 and the old surface and sod belonging to
period 2, show similar pollen spectra as the E-horizon with an AP of about 5570%. Only Corylus has decreased slightly till around 35% at the youngest period
in favour of Alnus, which has increased to approximately 50%. The old surface
belonging to period 2 shows a peak of 20% in Poaceae pollen.
s a
s
s inu cer s
cu
gu ax ni nu er alix ilia
Fa Fr Lo Pi Qu
S T
s
lu
ry
The pollen spectra taken from the old surface, and the ditch belonging to arrow
1, show arboreal pollen percentages of 45-60% (ADF=50-125 m). Tree pollen is
dominated by Alnus (ca. 50%), indicating that an alder carr was present nearby.
Corylus (ca. 30%) and Quercus pollen (5-10%) represent the forest in the higher
and drier environment. Herbal pollen is dominated by Ericales with percentages
from 60-120%. Other herbs, including anthropogenic indicators, are almost
absent.
Co
Oss-Zevenbergen 1 (see figure 12.7)
s
nu
Al
Oss-Zevenbergen
Trees and shrubs
The ditches belonging to the urnfield of the Oss-Vorstengraf area show arboreal
percentages of about 55%, indicating an open space with an ADF of approximately
75-100 m. The herbal vegetation is dominated by Ericales with pollen percentages
of 70-80%. The arboreal pollen component is dominated by Alnus (35-45%),
Corylus (35-45%) and Quercus (5-10%).
Heath
Urnfield
AP
The Bronze Age ditch (underneath the Chieftain’s grave) shows an arboreal pollen
percentage of approximately 60% (ADF is around 50 m). The two dominating
tree pollen species are Alnus (~45%) and Corylus (~35%). Quercus pollen is present
with a percentage of 10%. Other trees are present in percentages less than 2%.
Herbal pollen mainly consists of Ericales with a percentage of 55%.
es
or
)
sp
ae ae
la
r
n
r
o r
a
tu
fe are
lifl liflo olat
Be m
e
g
u
t
l
le P
b u
e
u
ila vu
tu lig anc
ab (A n s
e
ps
l
ae ae
e ium m in um
lle
ea e
a
isi race race tago eae rac cea olet od gnu term n s l po
p a e lle
m
c e
ta
te te te
an a p sa on ly h
Pl Po Cy Ro M Po Sp Ind Po
Ar As As
To
322 734
Bronze Age ditch (see figure 12.6c)
Anthr. ind. Grazing ind. Upl. herbs Ferns and mosses
oldest samples show percentages of up to 20% while this species has decreased to
less than 5% at the time the barrow was built.
oss-zevenbergen and surroundings
199
Unknown
MBA
80 100
20
40
60
s
20 40
lu
ry
Co
60
1
1
1
20
s a
s inu cer s
gu ax ni nu
Fa Fr Lo Pi
20
1
20
s
cu
er
lix lia
Sa Ti
Qu
Ul
1
m
5
100
s
us ula cale
t
i
Be Er
200
300
Upland herbs
Ferbs and mosses
1
1
1
5
20
40
5
1
1
1
1
1
1 2
5
1
es
5
or
sp
1
331
Figure 12.8. Pollen spectra from the samples taken from Oss-Zevenbergen barrow 2. OssZ2_Bh_per1 and
OssZ2_Eh_per1 are samples taken from the Bh and Eh horizons of the soil profile underneath the barrow.
Spectra are given in % based on a tree pollen sum minus Betula pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (= non arboreal pollen) spores are included, non pollen palynomorphs
are excluded. Different scales have been used, indicated with different colours. MBA= Middle Bronze Age.
OssZ2_Bh_per1
OssZ2_Eh_per1
640
571
466
609
317 1080
414
398
60
s
nu
Al
Grazing indicators
OssZ2_sod2_per1
40
P
ae
Anthr. ind.
es
or
sp
n
)
r
e
la
fe
n
ra
or
tu
er
t e re
pe
lifl ae liflo
Be m
ef
ty
c a lga
e
u
e
t
u
a
u
a P
b e
a
u
e
rr vu
sa
tu iac lig
ce (A n s
si l
ac e
to
ve m
ta
ae d ae
ce
yll ea e te p
a
lle
e
te iu um a um
is i ace opo ace eae
le pod g n em n s l po
x a is a ce a oph rac ce a ole
e
o
r
r
m
a gn lle
m cc ia ry pe sa on
ta
te te en te ac
on oly h
A r As C h As Po
Ru S u Ap C a C y R o M
M P Sp Zy Po
To
322 476
Heath
315
20
NA
Trees and shrubs
OssZ2_sod1_per1
OssZ2_os_per2
OssZ2_sod_per2
AP
Oss-Zevenbergen
Barrow 2
Oss-Zevenbergen 3 (see figure 12.9)
The sods and old surface show similar pollen spectra, indicating that the sods were
cut in the near vicinity of the barrow location. The old surface and sods of Barrow
3 show arboreal pollen percentages of approximately 55-60%, indicating an open
space with an ADF of approximately 75-100 m. The main tree species is Alnus
with percentages of more than 50%. Corylus is also present in high amounts (2045%), together with Quercus (3-15%). Ericales pollen dominates the non arboreal
pollen component with percentages of 50-80%. Other herbs are almost absent,
except for Poaceae with a percentage of 40% in sod 3.
Oss-Zevenbergen 4 (see figure 12.10)
The oldest sample, from the anthropogenic layer underneath the mound, shows
an arboreal pollen percentage of approximately 50% (ADF is around 100 m). This
arboreal pollen component consists of Alnus (ca. 40%), Corylus (ca. 40%), Quercus
(ca. 10%) and Tilia (ca. 10%). The herbal pollen component is mainly Ericales
(ca. 50%) and Poaceae (ca. 10%). There are few other herbal pollen species, which
are present albeit in very small amounts.
The pollen spectra from the barrow period and the following periods show
a decrease in the arboreal pollen percentage, indicating an increasing ADF of
the open space. At the time the barrow was built AP was around 60%, which
decreased to 15% just before the wind-blown sand covered the barrow. This is
mainly due to an increase of Ericales pollen, which increases to over 500%. At
that time some changes are visible in the arboreal pollen composition: Quercus has
increased to approximately 25%, while Corylus has decreased to around 20% and
Tilia has disappeared.
Mound 5 (see figure 12.11)
Pollen spectra from the old surface underneath this naturally formed hill show
arboreal pollen percentages of 40-55%. Arboreal pollen is mainly Alnus (ca. 45%)
and Corylus (ca. 45%). Quercus is present in percentages of about 10%. Ericales is
the dominant herbal pollen with percentages of 75 to 150%.
Oss-Zevenbergen 6 (see figure 12.12)
The ditch of Barrow 6 shows an arboreal pollen percentage of 50% (ADF is around
100 m). Ericales (ca. 90%) is dominant in the non arboreal pollen component.
Dominant trees are Alnus (ca. 50%) and Corylus (ca. 40%). Pollen from other
trees like Quercus, Tilia and Ulmus are present in lower percentages (3-5%).
Oss-Zevenbergen 7, sods and grave (see figure 12.13a)
Pollen spectra from the sods and the grave show arboreal pollen percentages of 4060%, indicative of an open space with an ADF of 75-150 m. The arboreal pollen
component consists mainly of Alnus (45-65%), indicating an alder carr in the
surroundings. Some difference between the sods in the percentages of Alnus might
indicate that some were taken closer to an alder carr than others. The arboreal
component representing the drier forest is dominated by Corylus (25-40%) and
Quercus (5-15%). The herbal pollen component is dominated by Ericales with
percentages of 65-150%.
200
ancestral heaths
oss-zevenbergen and surroundings
201
Hallstatt C
20
40 60 80 100
P
NA
n
Al
20 40
us
60
20 40
us
i n lus
rp r y
Ca Co
Figure 12.9. Pollen spectra from the
samples taken from Oss-Zevenbergen
barrow 3 Spectra are given in % based
on a tree pollen sum minus Betula
pollen. In the total AP (=arboreal pollen)
Betula is included. In the total NAP (=
non arboreal pollen) spores are included,
non pollen palynomorphs are excluded.
Different scales have been used, indicated
with different colours.
OssZ3_sod3
OssZ3_sod2
OssZ3_sod1
OssZ3_os
AP
Oss-Zevenbergen
Barrow 3
60
5
20
1
5
1
20
20
s
us
les
in s rcu ix a
us la
l
i
m tu
ax nu e
ica
F r P i Qu
Sa Til Ul Be
Er
5 10 5
s
gu
Fa
Trees and shrubs
40 60
80
Anthr. ind.
1
1
1
1
1
20
40 60
ae
ae
or
or ta
ifl
ul eae ulifl eola
b
tu iac lig nc
ae od ae la
a
isi race op race tago eae
m
n
te te e te an ac
Ar As Ch As Pl Po
Heath
Upland herbs Ferns and mosses Algae
5
1
5
1
1
1
1
5
1
1
415 831
331 617
329 632
s
re
es o
or n sp
p
s er
a)
rn f
ul
pe
fe te re
et
e uca lga
-B um
-ty ae
t
P
a
a
u
r
il r v
(A n s
os ce
ps ve
et lla ae
e e um um a
a su m olle
ac h y ce a
et et di
ex yop era uti nol nol yp o a gn ar y nem len al p
m r
l
h b g
p a o o l
t
Ru Ca Cy Kn M M Po Sp De Zy Po
To
329 544
Grazing indicators
202
ancestral heaths
Unknown
MBA
Unknown
40
60
20
40
60
20
40
60
s
5 10
gu
Fa
5
5
20
s
us
in s rcu
ax nu e
F r Pi Q u
40
1
5 10
lix ia
Sa Til
1
5 10
us la
m tu
Ul Be
s
le
100 200
ica
Er
300 400
500 600
Figure 12.10. Pollen spectra from the samples taken from Oss-Zevenbergen barrow 4. OssZ4_brown is the
pollen sample taken from the disturbed brown layer underneath the old surface. OssZ4_os_drift is the pollen
sample taken from the old surface underneath the drift sand layer at the southern part of the barrows. Spectra
are given in % based on a tree pollen sum minus Betula pollen. In the total AP (=arboreal pollen) Betula is
included. In the total NAP (= non arboreal pollen) spores are included, non pollen palynomorphs are excluded.
Different scales have been used, indicated with different colours. MBA= Middle Bronze Age.
OssZ4_brown
80 100
s
1
5
5
1
20
20
1
1
1
5
1
1
1
1
5
1
1
352
365
20
ylu
OssZ4_os_per2
r
Co
Upland herbs Ferns and mosses Algae
326
s
Grazing indicators
719
619
648
633
s
re
s
re spo
o
sp ern
ae e
a)
f
rn
ul
or or a ta
pe
fe te re
et
i fl
ul uli fl eola
e P-B um
te r uca ulga
-ty
a
b
a
a
e (A
il
s
r v
s
tu lig nc
to
n
e
ac
ps ve
ae ae la
at um olle
a
ce
e te te ium um
ea
e
isi ace ace ago eae
x a isa cea er ac tia cea ole ole pod gn arya nem en s
p
e
l
r
r
t
m
u
ll
s a on on ly ha b g
ta
c c ia p a
te te te a n ac um
Su Ap Cy Kn Ro M M Po Sp De Zy Po
Ar As As Pl Po
R
To
159 1054
Anthr. ind.
OssZ4_os_per3
nu
Al
Heath
314
P
NA
Trees and shrubs
OssZ4_os_per4
OssZ4_os_drift
AP
Oss-Zevenbergen
Barrow 4
oss-zevenbergen and surroundings
203
Figure 12.12. Pollen spectra from
the samples taken from OssZevenbergen barrow 6. Spectra
are given in % based on a tree
pollen sum minus Betula pollen.
In the total AP (=arboreal pollen)
Betula is included. In the total
NAP (= non arboreal pollen)
spores are included, non pollen
palynomorphs are excluded.
Different scales have been used,
indicated with different colours.
LBA= Late Bronze Age, EIA=
Early Iron Age.
Figure 12.11. Pollen spectra
from the samples taken from
Oss-Zevenbergen mound 5.
Spectra are given in % based on
a tree pollen sum minus Betula
pollen. In the total AP (=arboreal
pollen) Betula is included. In
the total NAP (= non arboreal
pollen) spores are included,
non pollen palynomorphs are
excluded. Different scales
have been used, indicated with
different colours.
LBA/EIA
20
40 60
80 100
AP
OssZ6_ditch
AP
20 40 60
Oss-Zevenbergen
Barrow 6
OssZ5_os2
OssZ5_os1
P
NA
Oss-Zevenbergen
Mound 5
Co
60
Co
20 40
1
20
5
20
50
Trees and shrubs
1
100
1
5
5
1
5
5
20
20
a
tic
ar
th
a
c
us
s
s
us
le
in s rcu mn
u s la
ica
ax inu ue ha ilia lm etu
r
F P Q R T U B
Er
5
s
s
us
le
in s rcu a
us la
i
m tu
ax nu e
i ca
Til Ul Be
Fr Pi Qu
Er
s
lu
ry
20 40 60
s
lu
ry
20 40
s
nu
Al
40 60
80 100
P
NA
20
s
nu
Al
Trees and shrubs
40 60
150 200
Anthr. ind. Graz. ind. Upland herbs Ferns Algae
1
20
1
1
5
1
1
Anthr. ind.
1
Grazing ind. Upl. herbs Ferns and mosses
360 660
5
1
1
20
1
1
1
5
1
1
335
713
pe
ty
es
aor
ell
s
sp
to
e
a)
n
e
a
r
r a
ul
c
fe
et
.a
flo t
-B
te
m
ae uli eola
/R
P
a
e
a
l
g
c
su
c
(A
si
e
os
ia e li an
n
et
ea p
m m
le
od cea go l e
ac a cac lete ium nu su pol
p
a
x
o a a e
i
g n
o
e is
en ter ant ac um ucc rass on terid pha olle otal
R S
B M P S P
Ch As Pl Po
T
Heaths
1
es
or
e
sp
a)
ra a
n
r
ul
o
fe
et
ifl t
e
ul eola
e P-B um
te
a
a
a
b
e A
il
s
ce
tu c
n
ac (
la
ps
e lan
yl
at um olle
e
te
ea o e
ea oph ne ole arya em en s
p
ac tag cea
c
l
r
n
o
ia ry si on eb g ll
ta
te an a
Ap Ca Ja M D Zy Po
To
As Pl Po
330 888
80 100
Heath
204
ancestral heaths
Hallstatt C
60
80
40
5
1
5
20
1
1
5
20
50
100
150 200
Figure 12.13a-b. Pollen spectra from the samples taken from Oss-Zevenbergen barrow 7.
12.13a: pollen spectra from the samples taken from the sods and the grave. 12.13b: pollen
diagram from the section taken underneath barrow 7. Spectra are given in % based on a
tree pollen sum minus Betula pollen. In the total AP (=arboreal pollen) Betula is included.
In the total NAP (= non arboreal pollen) spores are included, non pollen palynomorphs are
excluded. Different scales have been used, indicated with different colours.
20
1
1
1
1
1
20
1
1
1
1
5
1
1
1
Figure 12.13a
1
329 580
40
OssZ7_grave
20
312 545
OssZ7_sod 8
80 100
342 629
OssZ7_sod7
60
305 783
OssZ7_sod3
40
344 657
OssZ7_sod 5
20
338 769
OssZ7_sod4
s
s
s
s la
s inu s
cu
le
gu ax nu er
l ix ia mu tu
ica
Fa Fr Pi Qu
Sa Til Ul Be
Er
Upl. herbs Ferns and mosses NPP
308 607
s
lu
ry
Co
Grazing ind.
OssZ7_sod.3
s
nu
Al
Anthr. ind.
pe
ty
es
aor
el l
s
e
sp
e
to
e
a
a)
p
e
r
n
a
r
c
ul
o
ty
or ta
fe re
ifl
et
.a
aul eae ulifl eola
- B um
/R
te ulga
ae
P
a
b
a
l
p
s
g
c
i
s
v
o
(A
tu iac li n
o
n
ps m
ur
et
m
m
le
ae d ae la
a
ac a ta e tilla lete odiu ium nu 8
su pol
isi ace opo ace ago eae
x
5
g
s
n
u n
e
i
t
d
o p
m r
l
t
r
te te en te an ac um ucc usc ote on oly teri pha vG olle ota
S
Ar As Ch As Pl Po
R S C P
B P
M P P
T
305 712
Heath
316 618
P
NA
Trees and shrubs
OssZ7_sod 2
OssZ7_sod 1
AP
Oss-Zevenbergen
Barrow 7, old surface and sods
oss-zevenbergen and surroundings
205
Depth (cm)
40
60
20
40 60
1
5
20
1
20
1
5
50
s
us la le
m t u i ca
Ul Be Er
100
150
1
5
1
1
1
20
5
1
1
317 419
Figure 12.13b
313 690
20
s
s s
cu
gu nu er
lix ia
Fa Pi Qu
Sa Til
21.5
60 80 100
s
lu
ry
304 731
Co
rn
fe
20
40
s
nu
Al
ae
or
lifl
Aq. herbs
Graz. ind. Upl. herbs Ferns and mosses NPP
es
or
sp
a)
ul
re
et
-B um
te ulga
ae
u
P
a
l
e
g
i
s
v
c
li
(A n
la
m ps um m
ae
y l m iu e
lle
i
um
u
ce e h nu n et
od agn t58 n s l po
ra cea yop ygo rga nol
p
e
a
e
l
y
l ph vG ol
t
t oa ar ol pa o
s
o
o
P S
T
M
A P C P S
B P
317 578
Heath
319 479
20
P
NA
Trees and shrubs
18
16
14
AP
Oss-Zevenbergen
Barrow 7, diagram
Oss-Zevenbergen 7, pollen diagram (see figure 12.13b)
The pollen diagram shows a vegetation development from the period before the
barrow was constructed. The arboreal pollen percentage fluctuated between 42%
and 77% and at the time the mound was built an arboreal pollen percentage of
approximately 55% can be seen, indicating an ADF of approximately 75-100
m. The herbal vegetation is dominated by Ericales, which fluctuates from over
100% to 30%. The arboreal pollen percentages fluctuate some with Alnus (3550%), Corylus (35-45%), Quercus (5-20%) and Tilia (2-10%) being the main
components.
Oss-Zevenbergen 8 (see figure 12.14)
The arboreal pollen percentage fluctuates between 35 and 60%, indicating
a fluctuating size of the open space with an ADF of 50-250 m. Percentages of
Ericales fluctuate between 65-180%. Other herbal species show only low pollen
percentages, indicating the presence of a species-poor heathland. The arboreal
pollen component is dominated by Alnus, which seems to increase from about
25% to 45-50%. Corylus is fluctuating between 35-45%. Percentages of Quercus
(5-15%) and Tilia (0.5-10%) show a slight decline. Remarkable is the high
percentage of Betula in the oldest sample, while Betula pollen only occurs in low
amounts in all other pollen spectra from all mounds. Perhaps a Betula tree was
standing nearby.
Oss-Zevenbergen 12 (see figure 12.15)
The pollen spectrum derived from the ditch that remained from barrow 12 shows
an AP of approximately 50% (ADF is around 100 m). Alnus (ca. 60%) and
Corylus (ca. 30%) are the main components of this arboreal pollen percentage,
while Ericales is the dominating herb (ca. 80%).
Post alignment 1 (see figure 12.15)
The pollen spectrum from one of the posts from the post alignment shows an
arboreal pollen percentage of 40%. The main trees are Alnus and Corylus with a
pollen percentage of 45%. Herbs are dominated by Ericales with a percentage of
150%.
Drift-sand layer (see figure 12.15)
The pollen spectrum from underneath the drift-sand layer shows an arboreal
pollen percentage of 45%. Alnus (ca. 55%) and Corylus (ca. 35%) are the main
trees, while Ericales (ca. 120%) pollen dominates the herbal vegetation.
Size of the open space
The minimum size of the open spaces can be estimated by the measurements of
the barrows and the height of the sods that had been used in the construction of
the mounds (see section 7.1 and table 12.2). This leads to the following estimates
of sod-cut area:
Chieftain’s Grave: 11036 m2, ropenarea≈59 m, based on a circular open spot
Oss-Zevenbergen 1: 442 m2, ropenarea≈12 m
Oss-Zevenbergen 2: 284 m2, ropenarea≈9.5 m
Oss-Zevenbergen 3: 24504 m2, ropenarea≈28 m
206
ancestral heaths
Oss-Zevenbergen 4I+II: 318 m2, ropenarea≈10 m, Barrow 4III: 64 m2, ropenarea≈4.5
m, Barrow 4IV: 604 m2, ropenarea≈14 m
Oss-Zevenbergen 6: 85 m2, ropenarea≈5 m
Oss-Zevenbergen 7: 817 m2, ropenarea≈10 m
Oss-Zevenbergen 8: 262 m2, ropenarea≈9 m
Based on the ratio AP versus NAP of the old surfaces the open spaces the barrows
were built in had an ADF of 50-150 m. The sizes of the open spaces will be
discussed more in detail in section 12.1.3.
12.1.3 Discussion
The barrow landscape
Middle Bronze Age (Oss-Zevenbergen 2, 4 & 8)
The oldest group of barrows at Oss-Zevenbergen that was sampled for pollen
analysis (Barrow 2: sods and old surface period1&2; Barrow 4: period2; Barrow 8:
Ah, sods) shows that the mounds were built in species-poor heathland. Based on
the similarity of the pollen spectra from the sods and the old surface, the sods that
were used to build the barrows were probably cut in the near surroundings of the
barrow location. Since the barrows were located close together they were probably
constructed in one open space with an ADF of about 25-100 m, based on arboreal
pollen percentages of 50-70%. The forest in the environment was probably quite
open with a high percentage of Corylus at the forest edge. Besides Corylus the
forest’s main components were Quercus and some Tilia.
Betula is present in all spectra, which could indicate its presence in the forest
or perhaps some individual trees in the heathland area. In the wetter parts of the
area most likely alder carr was present. This might have been a few hundred metres
north of the barrows, the lowest part in the area based, or at the ‘groundwater
seepage wetland areas’ in the area (see figure 12.2).
The Bronze Age ditch of the Chieftain’s Grave shows an arboreal percentage
of about 60%, indicating an open place at the Oss-Vorstengraf area with an ADF
of approximately 50 m. The barrow was built in heath vegetation with mainly
Ericales. The forest in the surroundings is, as expected, comparable to the forest
around the Oss-Zevenbergen barrows. The heath in which Barrows 2, 4 and 8 were
situated was most likely separate from the heath in which the Bronze Age barrow
of the Vorstengraf area is situated. As has been mentioned in 12.1.1, seepage
water occurs in between the Oss-Zevenbergen and the Vorstengraf area, causing
conditions that were probably too wet for heath vegetation. One heathland area
stretching from the Oss-Zevenbergen area to the Vorstengraf area is therefore very
unlikely.
Late Bronze Age/Early Iron Age (Oss-Zevenbergen 1&6)
Compared to the Middle Bronze Age not much seems to have changed. Barrows 1
and 6 were built in an open space covered with species-poor heath vegetation. The
estimated ADF of the open space was approximately 50-100 m (AP= 45-60%).
The composition of the forest seems unchanged with mainly Quercus and Tilia.
Corylus was present at the forest edge. Alder carr was present in the lower and
wetter parts of the area. The spectrum of barrow 6 might represent a slightly older
period than the spectrum of Barrow 1, since Fagus is not present here.
oss-zevenbergen and surroundings
207
208
ancestral heaths
EBA/MBA
1
1
5
20
5
1
20
50
100
150
200
1
1
20
5
20
1
1
1
5
1
1
1
1
1
1
1
1
1
1
Figure 12.14. Pollen spectra from the samples taken from Oss-Zevenbergen
barrow 8. OssZ8_Eh , OssZ8_EB and OssZ8_B2h are samples taken from
the profile underneath the barrow, at respectively 6 cm, 10cm and 14 cm
underneath the old surface. OssZ8_Eh_ditch is a sample taken from the E
horizon in the ditch fill. Spectra are given in % based on a tree pollen sum
minus Betula pollen. In the total AP (=arboreal pollen) Betula is included.
In the total NAP (= non arboreal pollen) spores are included, non pollen
palynomorphs are excluded. Different scales have been used, indicated with
different colours. EBA= Early Bronze Age, MBA= Middle Bronze Age.
560
60
192
40
OssZ8_B2h
20
680
342
OssZ8_EB
60
908
309
OssZ8_Eh
40
532
309
OssZ8_Eh_ditch
20
711
403
80 100
s
s
s
s la
s inu s
le
cu
gu ax nu er
lix li a mu tu
ica
Fa Fr Pi Qu
Sa Ti Ul Be
Er
Ferns and mosses Algae
OssZ8_os2
60
s
lu
ry
Co
Upland herbs
682
40
s
nu
Al
Anthr. ind. Grazing indicators
s
re
s
re spo
po rn
e
s
e
a
a)
rn fe
ra
ul
or
a
fe ate are
pe
et
li fl ae li flo
at
y
B
l
e
c
e
t
g
u
e
t
m
l
a
u
a le P
b e
a
eo
ila rr u vu
su
ce
tu iac lig
ce b (A
nc
os
ps ve m m
lla eae
ta ina m
en
et e
la
l
ae od eae
y
e
e
u
a
a
c
l
e
l
i
u
t
t
o
u
e
l
m
h c e
a e e
a a
o
a m r
s
i
d n
g ae
c op ac
lp
ra
r
ex ce op r a n nt ce ol ol o g ry e te n
t a ce
m ia ry p e sio te s a on on lyp ha b a gn d e lle ota
te en te
an a
Pl Po
As Ch As
Ru Ap Ca Cy Ja Po Ro M M Po Sp De Zy In Po
T
327 503
Heath
329
20
P
NA
Trees and shrubs
OssZ8_os1
OssZ8_sod
AP
Oss-Zevenbergen
Barrow 8
oss-zevenbergen and surroundings
209
OssZ12_ditch
EIA
20 40
60
80 100
P
NA
20
s
nu
Al
40
60
80
20
s
lu
ry
Co
40
60
5
1
5
20
1
5
5
50
s
s
s
s inu s
cu
la le
gu ax nu er
l ix ia tu ica
Fa Fr Pi Qu
Sa Til Be Er
Trees and shrubs
Figure 12.15. Pollen spectra from the samples taken from urn field barrow 12,
the posthole fillings, and from the old surface underneath the drift sand layer.
Spectra are given in % based on a tree pollen sum minus Betula pollen. In
the total AP (=arboreal pollen) Betula is included. In the total NAP (= non
arboreal pollen) spores are included, non pollen palynomorphs are excluded.
Different scales have been used, indicated with different colours. EIA= Early
Iron Age, ME= Medieval Period.
OssZ_post101
OssZ_driftsand
EIA?
ME?
AP
Oss-Zevenbergen
Urn field, post alignment and drift sand
100
150 200
Anthr. ind.
Grazing ind.
Ferns and mosses
Upl. herbs
Algae
1
1
1
1
1
20
1
1
1
5
1
1
1
1
307 596
313 821
s
re
s
re spo
po rn
e
s
)
e
a
la
n fe
ra a
or
tu
e r te
pe
lifl e lifl o lat
Be
e f uc a
ty
u
t
e
m
a
o
u
a P
a r
b e
e
sa
su
tu iac lig nc
sil er
ce (A
n
to
ta um
le
ae od eae o la e
e p te v um a
ce
a
l
a
a
t
l
e
i
a
o
l
p c g a
e e n
c
i
m s
is
lp
ex i sa nt ol ol g ary e n
m ra no ra ta e
te te e te an ac um ucc ote on on pha eb ygn olle ota
Ar As Ch As Pl Po
D Z P
R S P
M M S
T
390 905
Heath
Early Iron Age (Hallstatt C) (VG, Oss-Zevenbergen 3 & 7, urnfield)
The younger barrows in the Oss-Zevenbergen area (3, 7) were built in open spaces
that were perhaps slightly larger than the older ones were built in (AP=40-60%,
ADF≈50-150 m). The composition of the heath vegetation had not changed
considerably and it was still poor in species. The forest seemed not to have changed
very much, except for Tilia being partly replaced by Fagus. Some differences can
be seen between sods belonging to the same barrow. This could be the result of
different locations the sods were been taken from. Some sods might have been
taken closer to the alder carr than others.
The Chieftain’s Grave, about 500 m to the west, shows a similar picture. It was
probably erected at approximately 50-100 m from the forest. Especially an alder
carr must have been reasonably close, indicated by high percentages of Alnus.
In addition, the pollen spectrum of the ditch of Barrow 12, belonging to the
urnfield near Barrow 3, indicates a vegetation composition as described above.
Pre-barrow landscape (Oss-Zevenbergen 2 & 8, Oss-Zevenbergen 7
and Chieftain’s grave)
Two of the Bronze Age barrows (2 and 8) provide information about the landscape
before barrow building took place. These spectra show that the open space already
existed and that it was covered with heath vegetation before the barrows were
built. The forest does not seem to differ greatly from later periods with alder carr
in the lower parts of the region and Quercus and Tilia being the main components
of the drier forest. However, Tilia seems to have a higher share in the forest at
the oldest spectra of Barrows 2 (e.g. Bh_per1; see figure 12.8) and 8 (e.g. Eh,
EB and B2h; see figure 12.14). The size of the open space might even have been
larger than when the Middle Bronze Age barrows were built, with an ADF up
to 200-500 m (AP barrow 2 Bh_per1 and barrow 8 B2h≈30%, see figure 12.8
and section 12.14). A barrow in the Oss-Vorstengraf area that dates to the late
Neolithic B period was situated about 200 m southwest of the Chieftain’s Grave
(figure 12.1b). Although no palynological data are available from this barrow it
can be assumed that this barrow was built in heath vegetation as well, based on
palynological data discussed in chapter 8-12, showing that all barrows were built
in heath vegetation. This could indicate that all barrows were built in a narrow,
but long-stretched heath area with a length of about 1 km that was already present
in the Neolithic. However, since the area in between the Vorstengraf area and the
Oss-Zevenbergen area probably was very wet due to seepage, it is more likely that
there were two separate heath areas. In one heath area the Neolithic barrows of the
Vorstengraf area were built, in the other heath area the first barrows of the OssZevenbergen group were built.
Barrow 2 was built on an Umbric Podzol (Dutch classification: Moderpodzol).
All other investigated barrows were built on a Carbic Podzol (Dutch classicifation:
Humuspodzol). It was suggested by de Kort that the open space in which Barrow 2
was built was created from forest recently, since heath vegetation would have caused
degradation of the soil to a Carbic Podzol (de Kort 2009). This could indicate that
Barrow 2 was one of the first barrows that was built at Oss-Zevenbergen, before
the heath vegetation could change the soil into a Carbic Podzol. However, since
the formation of a Carbic Podzol underneath heath vegetation can take 250 years
(and most likely takes even a longer period; Andersen 1979), it is not likely that
the open space was created very recently or for the intention of building a barrow
there.
210
ancestral heaths
What the heath areas were used for prior to the barrow building is not known.
Flint artefacts, dating to the Mesolithic, have been found during the excavation
campaigns in 2004 and 2007 (van Hoof 2009, 186), suggesting that the area was
in use by prehistoric man long before the first barrows were erected. Indications
for a settlement have only been found for the Middle Bronze Age south of the
burial complex (see figure 12.1b; Fokkens and Jansen 2004, Jansen and van der
Linde 2013a, 43-44). It was suggested that part of the area might have been used
as an agricultural field, indicated by the brown layer that was present underneath
Barrow 4. This can however not be confirmed by pollen analysis (see figure 12.10).
All pollen spectra show that a heath vegetation was present that was comparable
to that of the barrow period: very poor in species other than Ericaceae. Based on
the arboreal pollen percentages the ADF of the open space has probably fluctuated
through time before the barrows were built and probably at some point in time,
the ADF was larger than when the barrows were built (e.g. maximum AP= 30%35% in the pollen spectra Bh_per1 of Barrow 2 and B2h of Barrow 8, indicating
an ADF of approximately 250-500 m). The fluctuating size of the open space
could have been the result of fluctuating human related activities.
As has been explained in previous chapters, to maintain the heath it was
probably managed. Changes in grazing pressure could have caused varying ADF,
although grazing indicators are not present in high amounts in the pollen spectra.
Particles of charcoal that were found in the soil profiles throughout the entire
area could indicate that heath vegetation was regularly burned. The heath could
also have been managed by sod-cutting. From the Middle Bronze Age onwards
sods were cut in the area for the building of barrows, but it is not known whether
sods were already cut in the area before. The anthropogenic activities might even
have caused local sand drifting events. A layer of wind-blown sand was revealed
underneath Barrow 7, indicating such a period of sand-drifting probably in the
Neolithic period. This could have been a direct result of (too extensive) heath
maintenance activities in the area. When sods are cut a bare soil is left behind.
Possibly in combination with intensive grazing activities vegetation was not able
to stabilize the soil and the topsoil could be blown away locally. The Middle
Neolithic sand-drifting event recorded at Barrow 7 predates the barrow building
activities and it is not known whether sod cutting already took place there at that
time. Later events of sand-drifting were recorded in the area as well. For example
mound 5 appeared to be a natural hill formed of wind-blown sand. Although the
mound has not been dated, the pollen spectra from the old surface underneath
the sand suggest that the sand was deposited in the Middle Bronze Age (Fagus is
still absent). The activity of barrow building might very well have contributed to
this sand drifting event.
Pollen spectra showing the vegetation composition from before the Iron Age
barrows were built can be obtained from Barrow 7 and the Chieftain’s Grave.
Although the pollen spectra from underneath Barrow 7 and the Chieftain’s Grave
have not been dated, the pollen composition of both diagrams suggests that they
go back to before the Bronze Age barrows were constructed. The lower parts of
the diagrams show relatively high percentages of Tilia, reaching 5-10%. These
percentages can also be observed in the older spectra of the soil underneath
Barrow 2 (e.g. Bh_per1) and 8 (e.g. B2s). Bakels and Achterkamp (2013) suggest
that the lowest spectra from the soil underneath Barrow 7 date to the Early Bronze
Age, which indeed precedes the Middle Bronze Age barrow building. Then Tilia
decreases and Fagus increases, probably representing the replacing of Tilia by
Fagus.
oss-zevenbergen and surroundings
211
The pollen diagrams of Barrow 7 and the Chieftain’s grave both show a
fluctuating ADF much like the older barrows. The species-poor heathland areas
were probably fluctuating in size and at some point in time larger than when the
Hallstatt C barrows were built (minimum AP≈40%, ADF≈150 m).
Post-barrow landscape
What happened to the area after the barrows were built is only partially known. In
the Medieval Period the barrow complex of Oss-Zevenbergen was probably used
as an execution site. Two inhumation graves dated to the 13th and 14th century cal
AD were found dug into Barrow 2 and a 15th century cal AD inhumation grave
was found in Barrow 7 (Fontijn et al. 2013b, 313). There are no archaeological
traces that could indicate what the area was used for before the Late Medieval
Period, but the continued pressure by man is indicated by the pollen spectrum
from a sample taken underneath a drift sand layer that had covered Barrow 4. This
spectrum shows that the open space had increased, possibly by increased grazing
and/or burning activities. Layers of wind-blown sand that were found at the flanks
of the barrows could have been the result of this increased activity. It is suggested
that they are related to the intensive use of roads in the (post) medieval period
(van der Linde and Fokkens 2009, 51).
12.1.4 In conclusion: the history of the Oss-Zevenbergen landscape
A species-poor heathland area was present on top and at the side of a ridge of cover
sand in the Oss-Zevenbergen area, long before the first barrows were built. Two
heath areas had probably developed, separated by a due to seepage very wet area.
The ADF of the open space at Oss-Zevenbergen was probably fluctuating through
time and might have reached a maximum of approximately 500 m already long
before the barrows were built, according to the pollen spectra of Barrows 2 and
8 (e.g. b2: Bh_per1 and b8: B2h). Grazing and burning activities and possibly
sod-cutting were probably involved in maintenance of the heath vegetation and
varying pressure in these human related activities might have been responsible for
the varying heathland size and perhaps even some local sand drifting when the
pressure by grazing, sod-cutting and/or burning became too high.
Some indications for a Middle Bronze Age settlement have been found south
of Oss-Zevenbergen (see figure 12.1b) and it can be assumed that the community
responsible for heath management activities were settled at this location. Alder
carr was present in the lower and wetter areas in the region, probably a few
hundred metres to the north of the Oss-Vorstengraf area. The forest in the drier
surroundings consisted of mainly Quercus and Tilia with Corylus present at the
forest edge. In the Middle Bronze Age barrows (barrow 2, 4 and 8) were built in
an open space with Ericaceae as the main vegetation and an ADF of approximately
25-100 m. Since they were located on one of the highest locations in the area the
barrows were probably highly visible in the landscape. The heathland had perhaps
slightly expanded when the youngest group of barrows was built during the
Early Iron Age. Not only were new barrows constructed but also present barrows
were re-used. The construction of the very rich Chieftain’s Grave emphasises
the importance of this grave field. At this time the forest had undergone some
slight changes and Fagus had partly replaced Tilia. After this period the area was
probably kept in use for grazing. There are no indications that the area had been
used as settlement area or for other activities like crop cultivation. All this time the
barrows must have occupied a prominent place in the landscape while situated on
a relatively high location with the vegetation kept low by management activities.
212
ancestral heaths
Close to Heesch a group of six barrows is located called the Vorssel. One of these
barrows was palynologically investigated by de Kort in 2005 after it was reported
disturbed (de Kort 2005).
1
1
1
1
1
317 626
es
or
)
sp
la
n re
r
tu
fe a
e ulg e P-Be m
t
a
a
v e (A
sa
il
su
tu
to
n
e
ps m ac
e
ce
lle
e ea la te iu at um
ea e
x a isa cea rac ntil ole pod em n s l po
ac cea
e
r
m cc ia pe te on ly gn lle ota
te a
Ru Su Ap Cy Po M Po Zy Po
As Po
T
417 749
Mound 2 of the barrow complex was built from sods on top of a Carbic Podzol
(Dutch classification: Humuspodzol). The mound has not been dated. Two other
barrows in the complex contained Drakenstein urns, dating them to the Bronze
Age. Samples were taken from the old surface and from one of the sods of arrow 2.
Two samples were prepared and analysed described by the methods Chapter 4.
80 100
12.2.2 Results and discussion
60
See figure 12.16
The barrow was built in an open space with an ADF of about 50-100 m (AP≈55%)
covered with heath vegetation. There were only few other herbs besides Ericales
present amongst which some grasses. The dry forest in the surroundings consisted
mainly of Quercus and Tilia with Corylus at the forest edge. Alder carr was present
in the wetter parts of the region and the main contributor to the arboreal pollen
component.
Figure 12.16. Pollen spectra from the
samples taken from the Vorssel barrow.
Spectra are given in % based on a tree
pollen sum minus Betula pollen. In the total
AP (=arboreal pollen) Betula is included. In
the total NAP (= non arboreal pollen) spores
are included, non pollen palynomorphs are
excluded. Different scales have been used,
indicated with different colours.
Unknown
Vorssel_os
Vorssel_sod
AP
20
40
60
P
NA
80 100
s
nu
Al
20
40
60
Co
s
lu
ry
20
40
60
1
1
5
20
5
1
20
20
s
s
s
s inu s
cu
le
u s la
gu ax nu er
ica
lia m tu
Fa Fr Pi Qu
Ti Ul Be
Er
40
Heath
Trees and shrubs
Vorssel
12.2.1 Site description and sample locations
1
20
5
ae
or
lifl
bu
1
1
pe
-ty
Upland herbs Ferns Algae
Anthr. ind. Graz. ind.
12.2 Vorssel
12.3 Slabroek
A grave field that is located at Uden-Slabroekse heide (see figure 12.1) has been
partially excavated in 1923 by Remouchamps (Remouchamps 1924, Jansen and
Louwen in prep.). After the excavation the area has been partially used for crop
cultivation until 2003, when it was bought by Staatsbosbeheer to turn it into a
nature reserve area. The grave complex was supposed to form part of the area. It
was supposed to be presented and to be visible to the public as an archaeological
monument and as such to contribute to the cultural tourism. The area was
therefore archaeologically investigated in 2005 (prospectively) and excavated in
2010 (van Wijk and Jansen 2005, Jansen and Louwen in prep.). Several samples
for pollen analysis have been taken and analysed.
12.3.1 Site description and sample locations
The area is centrally located on the plateau of the Peel Blok, about 4 km south
of the Oss-Zevenbergen complex (see section 12.1.1). The urn field is located on
a ridge of cover sand. The size of the complex is unknown, but based on present
knowledge it should at least have been 250 by 200 m. The soil is classified as a
Carbic Podzol (Dutch classification: Haarpodzol). During the Medieval Period
the area was covered with heath vegetation until it was used for crop cultivation
between the early 20th century and 2003 (see previous section).
The area was first excavated in 1923. At that time 38 burial mounds were
discovered. Most of them were built of sods and they were all surrounded by a
ditch. In many of the barrows urns were found that were usually placed on the
old surface (some were dug into the old surface) before they were covered with a
barrow.
In 2005 and 2010 the area was re-investigated. The area was highly disturbed
by the cultivation activities during the last century and in 2005 the remains of
only 10 of the 38 monuments recorded by Remouchamps were rediscovered. On
the other hand, 26 ‘new’ ring ditches and the remains of two burial mounds were
oss-zevenbergen and surroundings
213
found (van Mourik 2005, 43). The excavation campaign in 2005 revealed that
the preservation of all archaeological features was very poor. To document all
traces an area of almost two hectares was completely excavated in 2010, when
all archaeological features were excavated and recorded. Several ‘new’ burial
monuments were discovered in 2010, amongst which a rich Iron Age inhumation
grave and several burials from the Roman Period. This excavation revealed that the
cemetery must originally have existed of more than hundred burial monuments
and probably has been in use from the Bronze Age until the Roman Period (Jansen
and Louwen in prep.). From all the burial monuments and features found during
the excavations in 2005 and 2010 several samples for pollen analysis have been
taken and analysed of which the details will be described below.
Slabroek 39 and 40
In 2005 the remains of two burial mounds were excavated, Slabroek 39 and 40.
Slabroek 39 appeared to be a barrow with a diameter of about 30 m and a height
of about 50-60 cm. The burial mound was built of sods that had an average length
of 50 cm and were between 7 and 29 cm thick. The central grave was looted. The
base of the barrow was dated by OSL to the Middle Bronze Age (1765-1500 cal
BC; van Mourik 2005). Samples for pollen analysis were taken and analysed by
de Kort and van Mourik. De Kort analysed a sample from a sod and a sample
from the old surface underneath the mound (de Kort and van Mourik 2005).
Van Mourik analysed a sample from the old surface and two samples were taken
respectively 5 and 10 cm below the old surface underneath the mound (de Kort
and van Mourik 2005).
Slabroek 40 was heavily disturbed and only 10 cm of its original height had
been preserved. It could still be observed that the barrow was built of sods and was
surrounded by a ditch with a diameter of 12 m. Three samples were taken from
this ditch by de Kort of which two were analysed for pollen: one sample from
the base of the primary ditch fill and one sample from the B horizon that had
developed in the ditch fill (see figure 12.17; de Kort and van Mourik 2005).
Urnfield
During the excavation of 2005, 26 new ring ditches were discovered. At the north
of the burial complex three ring ditches were found, of which one, Slabroek ditch
43 (see figure 12.17) the ditch fill has been sampled for pollen analysis by de Kort
(de Kort and van Mourik 2005).
In 2010 all discovered ring ditches belonging to the urnfield have been
excavated and recorded. Ditch 43 has been sampled for pollen analysis again and
in addition samples were taken from Slabroek ditch 12 (see figure 12.17) by the
author of this thesis (see also 6.1). Although none of the ring ditches has been
dated urnfields are generally assumed to date to the Late Bronze Age/Early Iron
Age. The dating of the urn field of Slabroek can probably be further specified
to the Early Iron Age, based on the finds of several Early Iron Age pottery by
Remouchamps (Remouchamps 1924).
Slabroek ditch 43 and 12
Ditch 43 has a diameter of 13-14 m. De Kort has taken four samples from the
northern part of the ditch. Three of these samples have been analysed: one sample
from the B horizon that had developed in the ditch fill, one sample from the base
214
ancestral heaths
of the E horizon and one sample from the top of the Eh (de Kort and van Mourik
2005). In 2010 ditch 43 has been sampled for pollen again, but this time samples
were taken from the southern part of the ditch (referred to as ditch 43A).
Ditch 12 has a diameter of 7 m. This ring ditch had already been discovered
by Remouchamps in 1923, who found an urn that was most likely filled with
cremation remains19. From the section of the ditch fill of both ditches 43 and 12
samples were taken from the top to the bottom every cm downwards. Samples
from the bottom of the fill have been analysed since it has been argued that
samples from the bottom of the ditch fill will probably provide the most reliable
information about the period that is closest to the period the ditch was dug (see
for argumentation 4.1.4).
Ditch ‘landweer’
At the western part of the burial complex a 340 m long ditch was discovered in
2005 that was partially filled with sods. The ditch was probably part of a Late
Medieval defence system called ‘landweer’ generally dating to around 1400 cal
AD. Three samples were taken for pollen analysis by de Kort. One sample from
the bottom of the primary fill and one sample from bottom of the secondary fill
were analysed (de Kort and van Mourik 2005).
12.3.2 Results and discussion
All pollen spectra (fig 12.18) show heath vegetation with mainly Ericales (most
likely Calluna as has been found in the urnfield ditches 12 and 43) and some
grasses. Through time, the heath area varied in size and was probably smallest at
the time (Middle Bronze Age) barrow 39 was built with an ADF of approximately
50 m (AP=60%). The heath expanded in the following period when several ditches
were dug in the area during the Early Iron Age, although ditch 12 seemed to have
been dug closer to the forest. The heath was probably larger than in the period the
ditch of the landweer was dug. AP was only 10-30% in the samples from the Bhorizon in that ditch. The pollen spectra are in agreement with the Late Medieval
dating suggested by the excavators (see section 12.3.1). This is indicated by the
relatively high percentage of Secale, which was not commonly introduced in the
Netherlands before the Roman Period (van Zeist 1976, Behre 1992). The find
of this species and some other Cerealia indicate that crop cultivation took place
nearby. There are no indications that this was also the case in the earlier periods,
when the Iron Age ditches and the Middle Bronze Age barrow were created. Some
anthropogenic indicators were found, but only in very low numbers.
During the entire period represented by the samples the forest composition
did not seem to change much. Alder carr was present in the surroundings on the
lower and wetter locations (probably west of the area, see figure 12.1b). Corylus
and Quercus dominated the forest on the higher and drier areas. Betula might have
been part of the forest or have been present in the heathland area as individual
trees. Although not all pollen spectra can be placed exactly in time, it is likely that
the area was covered with heath vegetation for centuries and that the area must
have been kept open to maintain this heath vegetation.
19
Cremation remains were not considered interesting at that time and were often discarded.
oss-zevenbergen and surroundings
215
barrow 40
ditch 43
barrow 39
ditch 12
0
25 m
Excavation 1923
Middle/ Late Bronze Age
Trenches 2005
Early Iron Age
Excavation 2010
Middle Iron Age
Ring ditch
Roman Period
Barrow without surrounding features
Pollen samples 2005
Possible burial monument
Pollen samples 2010
Grave ‘Prince(ss) of the Maashorst
Urn
Cremation remains
Posts
Ditch landweer
216
ancestral heaths
Figure 12.17. Locations of the
samples taken at Slabroek.
Figure after Jansen and
Louwen (in prep.).
oss-zevenbergen and surroundings
217
Unknown
MBA
EIA
1400 cal AD
AP
20
40
60
P
NA
Figure 12.18. Pollen spectra from the
samples taken at Slabroek. Spectra are
given in % based on a tree pollen sum
minus Betula pollen. In the total AP
(=arboreal pollen) Betula is included.
In the total NAP (= non arboreal
pollen) spores are included, non pollen
palynomorphs are excluded. Different
scales have been used, indicated with
different colours.
Slabroek40_base_ditchfill
Slabroek40_Bh_ditchfill
Slabroek39_os
Slabroek39_sod
Slabroek_Ditch43_Bh
Slabroek_Ditch43_Eh
Slabroek_Ditch43_TopEh/Ah
Slabroek_Ditch43_base49-48cm
Slabroek_Ditch43_base45-44cm
Slabroek_Ditch12_base35-32cm
Slabroek_Ditch12_base32-31cm
Slabroek_Ditch'landweer'__baseB/Ch
Slabroek_Ditch'landweer'_topB/Ch
Slabroek
80 100
s
20
nu
Al
40
60
80
1
20
us s
in lu
rp ry
Ca Co
40
60
g
Fa
20
us
1
20
us
in s
ax u
Fr Pin
Qu
us
20
c
er
1 2
l ix
Sa
Trees and shrubs
ia
5
Til
5
20
5
5
20
5
20
5
1
Upland herbs
20
40
5
Ferns and mosses
20
Algae
Grazing indicators
5
1
1
1
5
1
5
1
1
1
5
1
1
1
5
1
5
707
1357
826
956
334
324
755
587
358
340
326 1037
356
326 1305
327 1349
407 1081
741
361
412
285 1066
s
re
es o
or n sp
p
r
s
a)
pe
rn f e
re
ul
ty
la
fe a te a re
et
au
e
B
e
c
e
e
c
e
g
t
m
i
a a
p a rru ul s
a P
su
av
v
-ty si l
e ace o p
ce ( A
ia e p e ve um ore m
m e
ta m
en
ea yll eur eae
r
l
u
a
l
e
c
a
p
i
u
t
t
u
h a
c
n a sa ul le le od
a
o
s
a a
e gn ry em n s
ce ic p ut ra tia o ce io
lp
ia a ss ryo sc pe a u l yg sa ab erg o no o no l yp ilet ha ba gn l le
ta
To
Ap Br Ca Cu Cy K n Po Ro Sc Sp M M Po Tr Sp De Zy Po
91 1175
500 1000 1500
s
le
us la
ica
m tu
Er
U l Be
Anthropogenic indicators
ae
ae
or
or ta
pe
i fl
li fl ol a
ul
ae
- ty
u
b
e
e
c
sa
l ig nc
tu pe
o
a
i
et
ae la
ae -ty
od
a
ac a
ce g o e
i si ce lia
op le
ex
ra ta ea
is
m ra a
en eca
m
cc
te l an oa c
te ste ere
s
r
h
u
A P P
Su
C S
R
A A C
Heath
The Schaijksche heide is a former heath area that was used for sod cutting in the
12th and 13th century AD. The most part of the area has been planted with pine
forest in the early 20th century, which is the main vegetation at present times. It
is located approximately 5.5 km southeast of Oss-Zevenbergen and about 2.5
km northeast of the urnfield of Slabroek (see figure 12.1). The profile that was
sampled for pollen analysis consists of a podzol, developed in wind-blown sand,
which was deposited on top of a podzol developed in cover sand. The wind-blown
sand deposits were dated to around 4700 BC by OSL (three OSL dates were
determined: 4790 ± 308 BC, 4666 ± 377 BC, 4684 ± 337 BC). Samples for pollen
analysis were taken at 5 cm interval (van Mourik 1985).
Sint Annabos
Sint Annabos is a wetland nature reserve that used to be an extensive alder carr.
It is situated about 7 km south of Oss-Zevenbergen and about 3.5 km southwest
of the urnfield of Slabroek (see figure 12.1). Peat formation has taken place in
the area, which started between 4710 and 4530 cal BC. The profile that was
sampled for pollen consisted of three layers. The 2A layer developed in cover sand;
the H2 and H1 horizons consisted of peat. The H2 layer consisted of humified
218
ancestral heaths
1
5
5
5
400
300
200
100
5 10
20
5
5
5
20
20 40
60
80 100
20
40
1
ae
ce
ica
Er
a
l
tu
Be
s
s s
cu
gu inu uer ilia
P Q T
s
nu
Al
P
NA
Schaijk
Schaijk
Figure 12.19. Pollen spectrum from the sample taken
the Schaijk barrow. The spectrum is given in % based
on a tree pollen sum minus Betula pollen. In the total
AP (=arboreal pollen) Betula is included. In the total
NAP (= non arboreal pollen) spores are included, non
pollen palynomorphs are excluded. Different scales have
been used, indicated with different colours.
Schaijksche heide
5
Upl. herbs
Aq. mosses
Graz. ind.
In the environment of the Slabroekse heide van Mourik has palynologically
investigated several palaeosols (fossilized soils), peat and lake sediments (see figure
12.1; van Mourik et al. 2012b). In addition he applied OSL-dating to these soils
and sediments and in combination with 14C-dating he could make a reconstruction
of the evolution of agricultural soils and land forms in the area. The data of his
results will be used by the author to make a regional vegetation reconstruction of
the area around Oss-Zevenbergen from around 4700 cal BC.
12.5.1 Site description and sample locations
Fa
12.5 Palynological results from palaeosoils, peat and lake
sediments
us
in lus
rp ry
Ca Co
The barrow was built in an extensive heath area with an ADF that could have
reached approximately 500 m (AP≈25%). Other herbs besides heath were
practically absent. The forest in the surroundings was dominated by alder carr in
the wetter regions. The drier regions were covered with mainly Tilia, Quercus and
Betula with Corylus present numerously at the forest edge.
AP
See figure 12.19
)
la
tu
Be m
P
u
(A n s
m m lle
nu n su po
lia eae isa
g
a
l
re ac cc ria ha lle ota
T
Ce Po Su Va Sp Po
126 578
12.4.2 Results and discussion
Anthr. ind.
Near Schaijk five barrows were excavated in 1937 by van Giffen (van Giffen 1949).
The old surface of one of these barrows (Tumulus 3) was sampled and analysed
for pollen by Waterbolk (van Giffen 1949, Waterbolk 1954). This barrow has not
been dated.
Heath
12.4.1 Site description and sample locations
Trees and shrubs
12.4 Schaijk
organic plant remains and the H1 horizon consisted of humified plant remains
with blown in mineral grains (see figure 12.21,). Samples were taken every 5 cm
(van Mourik 1987).
Venloop
The Venloop is a stream valley where peat formation had taken place on top of
a mineral soil (cover sand). Most of the peat had disappeared due to drainage of
the area. At some locations the peat was preserved and one of these locations was
sampled for pollen analysis (van Mourik and Pet 2001). Samples were taken every
5 cm. According to the 14C-dating peat formation started between 750 and 410
cal BC. The sampled profile was situated approximately 5 km southwest of OssZevenbergen and 1 km southwest of the urnfield of Slabroek (see figure 12.1).
12.5.2 Results and discussion
See figure 12.20-12.22
At the Schaijksche heide, about 3 km west of the Slabroekse heide, a deciduous
forest developed after the wind-blown sand event of around 4700 cal BC. This
forest was dominated by Corylus and Quercus. Some Alnus was present in the
environment, but probably not in the form of the extended alder carrs that were
recorded in the barrow pollen spectra. Heath was already present in considerable
amounts. Around 4700 cal BC a short period of sand drifting occurred. Very
interesting is to realise that such early events of sand drifting have also been
recorded in the Laarder Wasmeren area (see 10.2) and as has been discussed in
Chapter 10, this might indicate an over-exploitation of the soil. The cause of the
sand drifting can however not be deduced from the pollen data. Sand drifting
might have been a local event; it has not been recorded at St Annabos, which is
located about 5.5 km southwest of Schaijksche heide. At the time peat started
to accumulate (between 4710 and 4539 cal BC) at St Annabos a birch carr (cf.
high percentages of Betula) was present, which probably evolved into an alder
carr. The development of an alder carr is probably reflected in the pollen diagram
of Schaijksche heide as well, shown by an increase in Alnus pollen. Also at the
Venloop, approximately 2.5 km southwest of Schaijksche heide and 1 km south
of Slabroekse heide alder carr was dominating the local vegetation when peat
accumulation started between 750 and 410 cal BC.
The extensive heath areas that must have been present from the Late Neolithic
onwards according to the barrow pollen spectra have not been recorded in the peat
diagram of St Annabos as such. This confirms that peat diagrams are not suitable
for a total landscape reconstruction, as has already been subject of discussion
in section 6.1. Expansion of heath is not recorded before deforestation started
accompanied by an expansion of grasses and some Cerealia, shown by all three
diagrams. The appearance of Fagopyrum in the diagrams of Venloop and St
Annabos indicates that the vegetation development of the area is recorded at least
until far into the Medieval period.
12.6 Summary: the barrow landscape of Oss-Zevenbergen
and surroundings
The barrow landscape of Oss-Zevenbergen and surroundings is, like the barrow
landscapes discussed in the previous chapters, a landscape dominated by open
spaces with heath vegetation.
oss-zevenbergen and surroundings
219
When and how these heath areas came into existence is not known, but an
anthropogenic origin is indicated. Heath vegetation was probably part of the
landscape already long before the first barrows were built, as has been recorded
for instance in the pollen diagram of the Schaijksche heide. The area might even
have been intensively exploited, causing sand-drifting as early as around 4700 cal
BC. Such early sand-drifting is also known for the Laarder Wasmeren area (see
section 10.2) and is remarkable for this period, since man-caused sand-drifting
was assumed to have started not before the Early Middle Ages (Castel et al. 1989,
Riksen et al. 2006). Heath was able to regenerate at the Schaijksche heide and it
is likely that heath was also present at other locations in the area. This is at least
the case for the Early and Middle Bronze Age period when several barrows were
constructed. It is not known whether the heath at the burial complexes around
Oss-Zevenbergen originated from the same early period as at the Schaijksche
heide, but heath was probably present at that location since the Late Neolithic,
when the first barrows were built. This heath area remained until at least the Iron
Age, when the Hallstatt C barrows were created. At that time, when the burial
complexes of Oss-Zevenbergen and Oss-Vorstengraf were at their most extensive
size, an extensive heath area must have been present.
To conclude, heath was present in the area for thousands of years. To maintain
such areas heath management must have taken place, which probably involved
grazing and burning. Other human activities were hardly recorded in the area.
It is not exactly clear where people lived and where they cultivated their crops.
Indications for crop cultivation have hardly been found in the barrow pollen
spectra, so it is not very likely that crop cultivation took place close to the burial
complexes.
Forest was also part of the barrow landscape. Before the barrows were built,
at the time of the sand-drifting (around 4700 cal BC), forest mainly consisted
of Quercus and Corylus, with birch carr in the wetter surroundings. Alder brook
was starting to expand at the wetter areas like Sint Annabos. At the time the
barrows were built extensive alder carrs were present as has been shown by the
high amounts of Alnus pollen in all the pollen spectra. These alder carrs could
most likely be found at locations like Sint Annabos and Venloop. In the dry forest
dominating trees were now Quercus and Tilia with high amounts of Corylus at the
forest edge.
220
ancestral heaths
Figure 12.20. Pollen diagram
from the Schaijkse Heide.
Redrawn from the pollen
diagram of van Mourik
et al. (2012b, figure 9). A
percentage diagram is shown,
with % based on a total tree
pollen sum.
oss-zevenbergen and surroundings
221
Figure 12.21. Pollen diagram from St Annabos.
Redrawn from the pollen diagram of van Mourik
(1987, figure 6). A percentage diagram is shown,
with % based on total tree pollen sum.
222
ancestral heaths
Figure 12.22. Pollen diagram from the Venloop. Redrawn from the
pollen diagram of van Mourik et al. (2012b, figure 5). A percentage
diagram is shown, with % based on total tree pollen sum.
In conclusion, the barrow landscape of Oss-Zevenbergen and surroundings
was a managed landscape of heath areas that could be quite extended, surrounded
by Corylus, Quercus and Tilia forest at the drier regions and alder carr in the brook
valleys. Part of this managed landscape had its origin probably thousands of years
before the first barrows became part of it (in the fifth millennium, see section
12.5). The barrow landscape existed as such for at least several centuries and seems
to have been a very stable element in the landscape.
Oss-Chieftain’s grave
Chieftain’s grave BA-barrow
Height (m)
53
1
Sod thickness
Sod area (m2)
(m)
Radius (m)
0.1
11036.15
59.27
442.00
11.86
16
unknown
Oss-Zevenbergen 1
4.7x23.5
0.6
0.13
Oss-Zevenbergen 2
12.5
0.6
0.13
284.07
9.51
Oss-Zevenbergen 3
30
0.9
0.13
2449.75
27.92
Oss-Zevenbergen 4
14.5
0.5
0.13
318.06
10.06
Oss-Zevenbergen 6
7.5
0.5
0.13
85.46
5.22
Oss-Zevenbergen 7
22.8
0.8
0.2
817.90
16.14
12
0.6
0.13
261.86
9.13
0.5
0.18
982.11
17.68
Oss-Zevenbergen 8
Table 12.2. The minimum size
of the open space per barrow
based on the sods used to build
the barrows.
Diameter (m)
Vorssel
unknown
Slabroek 39
30
Slabroek 40
unknown
Schaijk
unknown
oss-zevenbergen and surroundings
223
Chapter 13
Ancestral heaths: understanding the
barrow landscape
In Chapter 8-12 several case-studies have been described and discussed. In this
chapter these chapters will be summarized and interpreted in relation to the role
of barrows in the landscape.
13.1 The barrow landscape
13.1.1 What did the barrow landscape look like in the central and
southern Netherlands during the 3 rd to 1 st millennium cal BC?
In Chapters 8-12, 97 barrows and 11 urnfield barrows have been discussed in 5
regions on the Pleistocene soils in the central and southern Netherlands. It was
concluded that all barrows were built in open spaces that were covered with heath
vegetation. Barrows were built in open spaces that varied in size from small, with
an average distance to the forest (ADF) of 50-100 m, to rather large (ADF=300500 m), although the latter were only been found in the relatively young MiddleLate Iron Age barrows of the Echoput. Besides the barrows that were discussed
in the case-studies (Chapters 8-12) palynological data are known from 21 more
barrows in the central and southern Netherlands (see Appendix I) which show that
these barrows too, were built in heaths. Nevertheless, palynological data are only
available for a small part of the barrows that are still present in the Netherlands.
As has been shown in figures 3.1, 8.1b, 9.1b, 10.1b, 11.1b-d and 12.1b there
are numerous barrows in the investigated regions. Bourgeois suggests that only a
fraction of the barrows has been preserved, and that the original number of barrows
in the Netherlands was higher (Bourgeois 2013, 40). All investigated barrows were
built in heath vegetation and it is therefore probable that the non-investigated
barrows on the Pleistocene coversand areas in the Netherlands were built in a
setting featuring heath vegetation as well. As a consequence, the Dutch barrow
landscape must have been dominated by patches of heathland. The open spaces
seem to be small, however, and this in itself could be misleading. Many barrows
are often situated close to other barrows and sometimes forming long alignments.
It is therefore likely that many barrows were not built in their own small patch
of heathland, but clustered in larger open spaces that were long and narrow. This
has already been found to be the case for the oldest barrows. For example at
Renkum (Chapter 9) a long alignment of barrows can be seen. Not all barrows
in this alignment have been dated, at least 12 barrows can be placed in the late
Neolithic A period. Assuming that these barrows were all built in heath vegetation,
as has been demonstrated for four of them, it is likely that the open spaces were
connected to each other, forming a long-stretched heathland area with a length of
about 4.5 km (see figure 13.1a-c). These long-stretched heath areas could possibly
be seen as corridors in the landscape, although this research has only focussed on
the barrow landscape and not on the greater landscape. Other examples of this
barrow alignment can also be seen at Vaassen-Nierssen (Chapter 8, see figure
ancestral heaths: understanding the barrow landscape
225
0
1000
2000 m
Unexcavated or undated mounds
LN A barrows
Modelled heath area
m NAP
60 m
0m
226
ancestral heaths
Figure 13.1a-b. Barrow
alignment of Renkum at
two consecutive phases
Late Neolithic A and Late
Neolithic B. The modeled
heath area around each barrow
is indicated. Based on digital
elevation model of the AHN
(copyright www.ahn.nl).
Figure after Doorenbosch
(2013), figure 11.a-b. Figure
by Q. Bourgeois and M.
Doorenbosch.
Figure 13.1b.
0
1000
2000 m
Unexcavated or undated mounds
Older barrows
LN A or B barrows
LN B barrows
Modelled heath area
m NAP
60 m
0m
ancestral heaths: understanding the barrow landscape
227
?
0
1000
2000 m
Unexcavated or undated mounds
Older barrows
LN A or B barrows
LN B barrows
Modelled heath area
Modelled forest area
228
ancestral heaths
Figure 13.1c. Barrow
alignment of Renkum,
situated in a (hypothetical)
long-stretched heath area
surrounded by forest. The
vegetation reconstruction
is based on palynological
data from barrows. An exact
reconstruction of the forest
area is therefore not possible
(indicated by the question
mark), since barrows are not
present in those areas. The
figure is based on digital
elevation model of the AHN
(copyright www.ahn.nl).
0
1000
2000 m
Unexcavated or undated mounds
LN A barrows
Modelled heath area
m NAP
80 m
0m
Figure 13.2a-b. Barrow alignments of
Vaassen-Niersen at two consecutive
phases Late Neolithic A and Late
Neolithic B. The modeled heath area
around each barrow is indicated. Based
on digital elevation model of the AHN
(copyright www.ahn.nl). Figure after
Doorenbosch (2013), figure 10.ab. Figure by Q. Bourgeois and M.
Doorenbosch.
ancestral heaths: understanding the barrow landscape
229
0
1000
2000 m
Figure 13.2b. Barrow alignment
of Vaassen-Niersen during the late
Neolithic B.
0
1000
2000 m
Unexcavated or undated mounds
Unexcavated or undated mounds
Older barrows
Older barrows
LN A or B barrows
LN A or B barrows
LN B barrows
LN B barrows
Modelled heath area
Modelled heath area
m NAP
m NAP
80 m
80 m
0m
0m
13.2a-c), Toterfout-Halve Mijl (Chapter 11) and Oss-Zevenbergen (Chapter
12). The formation of barrow alignments have been extensively investigated and
discussed by Bourgeois (2013). He has found many other examples of barrow
alignments, indicating that this was a fairly common way to spatially order burial
mounds in the barrow landscape. Not all barrows, however, were built into such
alignments. Alignments were mainly a feature of barrows constructed during
the Late Neolithic A, while from the Late Neolithic B onwards barrows are also
found outside the alignments at presumably more random places in the landscape
(Bourgeois 2013). These dispersed barrows were built in heath as well as has been
shown at for example Toterfout-Halve Mijl (Chapter 11). Altogether, the barrow
landscape must have been dominated by patches of heath, which all contained one
or more barrows, and which were possibly often connected to each other, forming
corridors in the landscape.
230
ancestral heaths
?
0
1000
2000 m
Unexcavated or undated mounds
Older barrows
LN A or B barrows
LN B barrows
Modelled heath area
Modelled forest area
Figure 13.2c. Barrow alignments
of Vaassen-Niersen, situated in a
(hypothetical) long-stretched heath
area surrounded by forest. The
vegetation reconstruction is based on
palynological data from barrows. An
exact reconstruction of the forest area
is therefore not possible (indicated by
the question mark), since barrows are
not present in those areas. The figure is
based on digital elevation model of the
AHN (copyright www.ahn.nl).
Although dominated by heath, forest was also part of the barrow landscape. This
forest can be divided into two main components. In the lower and wetter parts of
the area extensive alder carrs could be found. At the drier locations a mixed oak
forest was present. This forest was probably fairly open and consisted mainly of
Quercus. Tilia was part of the forest as well, especially during the Neolithic and
Bronze Age. Fagus appeared later. Fagus could be noticed in the Bronze Age period
at Toterfout-Halve Mijl. It had partially replaced Tilia in the Iron Age period as
shown at the Echoput barrows (chapter 8) and at Oss-Zevenbergen (chapter 12).
Corylus, a light demanding tree, was dominant close to the barrows and profusely
growing at the edge of this forest.
ancestral heaths: understanding the barrow landscape
231
As has been explained in Chapter 2 it was previously thought that differences
in barrow landscapes were culturally linked (Waterbolk 1954, van Zeist 1967a).
Casparie and Groenman-van Waateringe already concluded that this assumption
could not be true and that differences between sites were caused by differences
in soil type and hydrology (Casparie and Groenman-van Waateringe 1980). They
emphasized the uniformity of the barrow landscape for all barrows. This uniformity
has been confirmed by this research, where (part of ) the data collected by Casparie
and Groenman-van Waateringe have been supplemented and re-interpreted.
A mosaic managed heath open-forest passage landscape
As has been extensively discussed in Chapters 8-12 the heath vegetation the barrows
were built in must have been managed to persist. Management activities could
involve grazing (or mowing), burning and/or sod-cutting (Stortelder et al. 1996,
287). Indications for large scale sod-cutting have not been found, but sod-cutting
must certainly have taken place as sods were used as construction material for the
burial mounds and could therefore have contributed in the maintenance of the
heath vegetation. Grazing has been indicated in most of the case-studies. During
the Neolithic, prehistoric man in the Netherlands switched from hunter-gatherer
to farming activities, including crop cultivation and animal husbandry. Faunal
evidence from shows that the livestock of farming communities mainly consisted
of cattle and sheep (Fokkens 2005a, 409, 427; Brinkkemper and van WijngaardenBakker 2005, 493). The heath in the barrow landscape was consequently most
likely being grazed by these animals (see figure 13.3a-b). The number of livestock
belonging to late Neolithic farming communities has not been estimated, but
for the Middle Bronze Age B has been suggested that a livestock of up to 30
animals could be kept per household. It has been suggested that these animals
were mainly grazing at natural pasture areas in the stream valleys (IJzereef 1981,
Fokkens 1991, 2005a).
Another possibility is that they were grazing the barrow heath areas. To
maintain a heath area about 1 sheep per hectare and/or 1 head of cattle per 5-6
hectare is required. The heath area around a barrow with an estimated ADF of
about 100 m could be simplified to a (hypothetical) circular patch of heath with
an estimated radius of 100 m, indicating an area of about 3 ha. Each barrow
requires then about 3 sheep or 0.5 head of cattle to maintain this heath area. It
232
ancestral heaths
Figure 13.3a-b. 13.1a: Grazing
sheep at the Tafelbergheide, a
heatland area near Huizen (the
Netherlands). 13.1b: Grazing
cattle at the Zuiderheide, a
heathland area near Laren (the
Netherlands).
Figure 13.3b
has been estimated that in the area of Ermelo 52 barrows were built in the Late
Neolithic A, 26 barrows in the Late Neolithic B, 7 barrows in the Early Bronze
Age and 48 barrows in the Middle Bronze Age. This assumes that in total about
134 barrows were present at Ermelo in the Middle Bronze Age (Bourgeois 2013,
table 8.1, p. 178). Assuming that all barrows were situated in heath vegetation
in the Middle Bronze Age, this implies a total estimated heath area of about
420 ha. To maintain such a heath area about 420 sheep are required and/or 70
head of cattle. Alternatively, when assumed that one household kept 20 head of
cattle and 10 sheep, 3-4 households were able to maintain the heath. When the
average ADF in an area for each barrow is 250 m, around 2630 ha of heathland
should be maintained, requiring 20 households with each 20 head of cattle and
10 sheep. This implies that several households, forming heath communities, must
have worked together to maintain the heathland.
Grazing in relation to the barrow landscape is also mentioned in barrow research
from for other regions in Europe. Andersen showed that barrows in West Jutland
(Denmark) were built in open places that were used as pasture (Andersen 199697). The oldest barrows (3500-3300 cal BC) were built in open places in birch
woodland that was grazed and from the Early Middle Neolithic barrows (33003100 cal BC) onwards they were often built in heathland that served as pasture.
Bunting and Tipping concluded for a Middle Bronze Age barrow cemetery (1500900 cal BC) in Orkney (Scotland) that the burial mounds were constructed on
pasture land (Bunting and Tipping 2001).
Burning is the third heath management method. Indications for burning have
only been found in a few case-studies in this research by the recording of charcoal
that was probably not just related to the burial itself. At the Echoput and OssZevenbergen for example (chapter 8.1 and 12) small fragments of charcoal were
found throughout the entire profile underneath the barrows. Karg showed for a
barrow in Western Jutland (Denmark) dating to the 14th century cal BC that it
was built in a heathland where burning had taken place. In addition the heathland
had been managed by grazing and sod-cutting (Karg 2008).
It should be noticed that the heath management activities described above,
especially grazing, might not have been practiced by prehistoric man with the
aim of managing the heath. They might just have been carried out by the barrow
builders as part of their daily (agricultural) activities. Managing the heath at the
ancestral heaths: understanding the barrow landscape
233
same time might just have been incidental. Nevertheless, whether deliberately
managed or as an additional consequence of other activities, heath vegetation was
a very important if not the most important component of a barrow landscape. To
conclude, the barrow landscape must have been a very characteristic landscape. A
landscape that could perhaps best be described as a mosaic managed heath open
forest-passage landscape.
13.1.2 What was the history of the barrow landscape before the
barrows were built?
In five cases it has been shown that this heath vegetation was already present
some time before the barrow was built by pollen diagrams derived from the soil
profile underneath the barrow (2 barrows at the Echoput, chapter 8.1; OssZevenbergen barrow 2, 7 and 8 and the Chieftain’s Grave, chapter 12.1). In other
cases the presence of diverse herbal vegetation suggests that the area must have
been open for some time. Otherwise this vegetation would not have had the
chance to get established. No indications have been found that the open space
was created recently before a barrow was built. Some barrows were built on top
of an Umbric Podzol (Dutch classification: Moderpodzol; Echoput, chapter 8.1;
Oss-Zevenbergen Barrow 2, Chapter 12), which is common underneath forest
vegetation. It has been suggested that this could be an indication that heath
vegetation had not been present for a very long time, since underneath heath
vegetation eventually a Carbic Podzol (Dutch classification: Humuspodzol) would
develop (de Kort 2009). It should be noted that this is only relative to the length
of time soil development can take; heath vegetation can transform an Umbric
Podzol into a Carbic Podzol in approximately 250 years (Andersen 1979) which
is rather long relative to a human life. To conclude, in most cases the open spaces
were present well before the construction of the barrow and it is therefore unlikely
that they were created specifically for funerary purposes. When, how and why the
clearings have been created is unknown. The open spaces might have originally
been natural open spaces in the forest (see 2.3.1) and turned into heathland by
human influence. They might also have been man-made clearings from the start.
It is also not easy to reconstruct what the open spaces have been used for prior to
the barrow building. Possible traces of abandoned settlements have been found
underneath a barrow in only a few cases (Vaassen, chapter 8.1; Putten, chapter
8.4; Stroe, chapter 8.10) and only in some cases barrows were possibly (although
questionable) constructed on former arable land (Toterfout-Halve Mijl barrow 12
and 18, chapter 11.1; Eersel, chapter 11.6). Besides, Casparie and Groenman-van
Waateringe (1980) conclude that barrows were seldom constructed on or close
to arable land that was in use when the barrows were built, a conclusion that
was confirmed by the present research. It is therefore not likely that the barrow
builders had a preference for (recently abandoned) settlement sites and/or arable
fields. On the other hand it is clear that in all cases prehistoric man must have been
present at the sites prior to the barrow building, since at least some management
was required to maintain the heath vegetation. Probably most open spaces were
used as pasture already before the barrows were built, since grazing is indicated in
several cases (Echoput, section 8.1; Oss-Zevenbergen Chapter 12).
13.1.3 What does this mean?
The Late Neolithic landscape in the southern and central Netherlands is often
seen as dominated by a fairly closed forest. As has been described in Chapter
2, deciduous forest is in general assumed to be the natural landscape in the
Netherlands that had developed since the start of the Holocene. During the
234
ancestral heaths
Neolithic prehistoric man started to interfere with the landscape when they
started to clear the forest to expand their agricultural activities. This was assumed
to have happened only at local scale, in the close surroundings of a settlement site
(Waterbolk 1954, Groenman-van Waateringe 1978). In general the open spaces
were small and did not have a great impact on the landscape yet. Casparie and
Groenman-van Waateringe (1980) concluded from their research that large open
areas did not yet occur during the Neolithic in the central Netherlands. During
the Bronze Age and Iron Age the Dutch landscape was transformed into a cultural
open landscape, with heath and fields replacing the forest. During the Neolithic
period man also started to build barrows to bury their dead. Neolithic barrows
were pictured as being in small man-made open spaces in the forest, but it is not
clear how these fitted in the landscape organization at large. The results described
in Chapters 8-12 suggest that the landscape was probably already more open than
previously thought. Based on the reconstructions from barrows the landscape
must certainly have been open. All barrows were built in heath vegetation and the
surrounding forest was open in character. Barrows were numerous and plentiful
from the earliest Neolithic period. All these barrows being built in heath paint
a different picture of the landscape than a closed forest with some small, open
spaces.
For the Bronze Age it has long been thought that a barrow’s location was
determined by the location of the settlement (Roymans and Fokkens 1991).
This theory was mainly based on sites like Elp, where a barrow was located close
to Middle Bronze Age houses (see figure 13.4). Bourgeois and Fontijn showed
that this theory could not be confirmed (Bourgeois and Fontijn 2008). Most
barrows predate the Middle Bronze Age houses, and settlements dating to the late
Neolithic and Early Bronze Age, the period in which most barrows were built,
have rarely been found (see also 2.3.2). In fact, it is not known where the people
who built the barrows and who were buried in the barrows lived. Settlements have
rarely been found close to barrows (Bourgeois 2013). In addition, this research
has shown that palynological data seldom show the presence of arable fields in the
near surroundings of a barrow, which are generally assumed to be located close to
settlements (van Gijn and Louwe Kooijmans 2005, 338-340).
The barrow landscape was a managed landscape, with numerous patches of heath.
As was previously thought that prehistoric man just started to interfere with the
landscape during the Neolithic, these managed barrow landscapes assume large
scale control of the landscape by man. And even long before the barrows were
built, prehistoric man may have already overexploited some areas, as indicated
by very early sand-drifting events at Oss-Zevenbergen in de Middle Neolitic
(Chapter 12), the Schaijksche heide around 4700 cal BC (Chapter 11) and the
Laarder Wasmeren area around 4000 cal BC (Chapter 10). Although the cause
of these sand-drifts is unknown, they show that the landscape was open and that
heath vegetation was already present by then. This is further indication of the
presence and activity of man and implies a landscape that was maintained by this
activity of man and even perhaps overexploited by him.
To summarize, despite not being built very close to settlements, the barrows seem
to be integrated into the everyday life of prehistoric man. The barrow landscape
was a managed landscape, which most likely was at least partially maintained by
grazing, and seems to form as such part of the economic zone of the people living
in the area. It is however not clear where the settlements of these communities
were located. The evidence for settlements is elusive for the late Neolithic and the
first half of the Middle Bronze Age. It seems likely that settlements were located
not too far away, at ‘grazing’ distance from the barrows.
ancestral heaths: understanding the barrow landscape
235
236
ancestral heaths
Grazing grounds, ancestral grounds?
One of the questions this research is trying to answer is whether the barrow
builders had a preference for ancestral grounds to place their mounds. Based on
the data that are available now, discussed above and in the previous chapters, this
question can most likely be answered affirmatively. In general barrows were built
on grazing grounds. Grazing took place concurrently and prior to the barrows
being built. The barrows that were investigated were never built in areas that were
recently cleared and it is not very likely that the barrow builders created heath
areas especially for the construction of a burial mound. Instead, all barrows were
built in areas that had been in use by prehistoric (heath) communities for a long
period of time. These communities might very well have consisted of the ancestors
of the people who built the barrows. The heathland areas where barrows were
built in can be considered as ancestral heaths: not only did they serve as burial
places for ancestors, they had also been used by these ancestors prior to the barrow
building. The builders of the barrows built on the investment of their ancestors.
13.1.4 What was the role of barrows in the landscape?
Figure 13.4. Excavation plan
of the barrow and settlement
of Elp. The barrow was found
very close to the settlement.
Figure after Fokkens (2005a,
figure 18.4).
Barrows were often located in alignments in long-stretched heath areas. It is
not hard to imagine that visibility must have played an important role in the
placement of the mounds as has already been suggested in Chapter 2. From one
mound the next mound could be seen and so on. Such corridors/passageways in
the landscape must have been an impressive sight. Bourgeois (2013, Chapters
6 and 8) investigated visibility for several barrow alignments and clusters. He
performed view-shed analyses to determine whether barrows were built on visible
places in the landscape. How visible was a barrow in the landscape and what part
from the landscape could be seen from a barrow? Besides the land relief (elevation,
slope and orientation of terrain features) the vegetation and especially the trees
are determining factors on the degree of visibility. The vegetation data that were
derived from the pollen analyses described in Chapter 8-12 provided valuable
information in this respect. Models have been developed to get a better grip on
the relation between pollen spectra and the corresponding vegetation abundance.
In Chapter 7 these models have been applied to barrow pollen spectra to be able to
improve our visualization of a barrow landscape. Based on these models, barrows
in the view-shed analysis were placed in (hypothetical) circular heathland areas
with an average radius of 250 m. In addition the vegetation reconstructions have
shown that alder carr made up a considerable part of the forest in the lower and
wetter surroundings of all investigated barrows. For the view-shed analyses alder
carr with a height of 15 m was placed at locations with high groundwater level,
taking the recent lowering of groundwater by modern canalization and use of
groundwater into account (Bourgeois 2013, 132). The dry Quercus forest obviously
would also be of influence and although its exact location cannot be determined
a forest with a height of 30 m was placed at the places that were not covered with
heath or alder carr to get a rough impression of the visibility of barrows. Bourgeois
concluded that barrows were more visible than their environment, but not all
barrows were equally visible from their environment. Some barrows were highly
visible and could probably be seen from long distances, while other barrows were
only visible from the edge of the heath area. Also in alignments visibility varied
between barrows. Some barrows could be seen from anywhere in the alignment,
while others could only be seen from the barrow next to it and still others appeared
only at the skyline from specific positions in the landscape. As Bourgeois (2013,
156) puts it: “Especially in the case of the alignments, visibility was manipulated in
order to reveal a succession of monuments.” Although the degree of visibility seems
ancestral heaths: understanding the barrow landscape
237
to have differed between barrows it cannot be denied that visibility must have
played an important role in the placement of barrows. Even when a barrow could
only be seen when entering the heath area it was built in, it was probably still
an eye catcher within that heath area. Visibility played an important role in the
placement of the barrows. Their visibility might have been enhanced when the
sods for barrow construction were taken in the direct environment of the barrow,
as has been shown at the Echoput (Section 8.1). These barrows were located on
one of the highest places in the environment in an open area that was covered with
heath vegetation, while the direct surroundings were completely stripped from
vegetation. This might have been undertaken to emphasize their characteristic
sight in the landscape.
Barrows and the importance of visibility have also often been discussed in
barrow research outside the Netherlands. Early Bronze Age Barrows in Thy,
Denmark, were all built in a rather treeless landscape that was used as pastureland
(Andersen 1996-97). Hannon et al. showed that five Bronze Age mounds (1800500 cal BC) at the Bjäre peninsula (southern Sweden) were built in an open
landscape that was probably grazed (Hannon et al. 2008). They concluded that
these barrows were probably designed to be visible in the landscape. Also Downes
suggested that the location of barrows in Orkney (Scotland) were probably related
to visibility (Downes 1994). Dreibrodt disagreed with the theory that all barrows
were built in an open landscape (the landscape openness hypothesis; Dreibrodt
et al. 2009). He showed that some barrows were built on hilltops while the hill
flanks were probably covered with forest, since no soil erosion had taken place at
these hills. However, he also mentions the possibility of a well suited system of
pasture that could have maintained a vegetation cover preserving the hill from
soil erosion. Fyfe rejected the landscape openness hypothesis as well (Fyfe 2012).
He stated that there is no single blueprint for the vegetation composition on
and around a barrow site and that barrows were built in landscapes that varied
from very open to forested. He also mentions however, that barrows were built in
the relatively most open places in the environment. Casparie and Groenman-van
Waateringe (1980, 61) conclude: “The environment in the immediate vicinity of
a barrow varied from only slightly degraded forest to extremely degraded, heath-rich
vegetations, with all possible intermediate stages.” The research in this thesis shows
that barrows were built in open spaces that varied in size from small to large.
Besides, it was shown that visibility could still have played a role in small open
spaces especially while they might have been connected forming a narrow longstretched corridor heathland. In addition, a forested site does not necessarily imply
that visibility played no role in the barrow building. Especially when multiple
barrows were built in a region small views might even have emphasised their
special place in the landscape (figure 13.5). Visibility was maintained while the
heathland was managed. This could have been a ritual activity purely to preserve
the visibility of the barrows in the landscape, it is however much more likely that
the management also had an economic aspect. As has been discussed above and in
the Chapters 8-12 grazing was probably involved and to maintain such extensive
areas of heathland considerable livestock was necessary. It is therefore expected
that the barrow landscape was in use as part of the agricultural organization of the
prehistoric farmers.
To conclude, the role of barrows in the landscape of the central and southern
Netherlands seemed to be twofold. On the one hand they were assigned a special
place in the landscape, separate from settlements and fields, where visibility seemed
to have played an important role. On the other hand they were integrated into
everyday life, while they formed part of the economic zone of the people living in
the area. Prehistoric landscape undergoes impressive changes from the Neolithic
238
ancestral heaths
Figure 13.5. A small
alignment of barrows at
Toterfout-Halve Mijl. The
small view emphasizes the
‘specialness’ of barrows.
to the Iron Age (and further on), when prehistoric man gradually changed it to a
cultural landscape. The heaths of the barrow landscape, however, were very stable
elements in this changing landscape that existed as such for thousands of years.
13.2 The heath open-forest passage landscape as part of the
Dutch prehistoric landscape
Peat and lake sediments have been proven to be good pollen preservers, as has been
explained in section 2.2.1. Therefore, information about the Dutch prehistoric
landscape is mainly derived from palynological analyses of peat and lake sediments.
Pollen in peat and lake sediments is assumed to represent the regional vegetation.
However, this mainly accounts for the arboreal pollen component. Most herbal
pollen does not travel long distances and therefore open places in the region of
the peat or lake will be underrepresented or not be recorded at all. As has been
shown by the palynological analyses of the Venloop and Slabroek (see chapter
6.1) a peat diagram does not necessarily represent the vegetation composition of
a burial complex at only 1 km distance. It is therefore not realistic to generalize
the landscape that was shown by peat and lake sediment analyses, since they only
represent a specific type of landscape. For Late Neolithic times, when barrows
were started being built, the general view of the Dutch landscape is that it is
dominated by deciduous woodland (see 2.1). This research has shown that this
view should be reconsidered and that the landscape was probably already more
open than previously thought. In addition, palynological analyses of barrows
only show a particular part of the landscape and the landscape picture drawn
from this research can certainly not just be extended to for example settlements
sites, neither can it be applied to sites with completely different environmental
circumstances like wetland sites. Other researchers have argued that palynological
sampling of peat and lake sediments alone are not suitable for a detailed vegetation
reconstruction. Behre (1986) for example has reconstructed the development of
landscape and prehistoric habitation within an isolated (surrounded by bogs)
prehistoric settlement area called Flögeln (Northwest Germany) by creating a
dense network of ten pollen diagrams. Palynological data were collected from
ancestral heaths: understanding the barrow landscape
239
a large raised bog just north of the settlement area, which provided the history
of a regional vegetation development. However, habitation phases were hardly
reflected in these diagrams. Only pollen diagrams derived from kettle-hole bogs
within the settlement area showed a detailed overview of the several habitation
phases in the area. Behre concluded that many pollen diagrams only show part
of the (settlement) landscape, even when a settlement area was situated very close
to the sample location. To get a most complete reconstruction of landscape and
habitation development sampling at multiple locations in the area is necessary.
Also Groenewoudt et al. stated that the distance of most peat remnants to
settlement areas is too large to provide reliable data about them (Groenewoudt
et al. 2007). They collected palynological data from (man-made) pools and wells
in or close to Late Bronze Age to Medieval settlements in a small-scale cover sand
area in the eastern part of the Netherlands to get a more detailed understanding of
the vegetation development in that settlement area. They concluded that during
the Neolithic the settlements were situated in natural open spaces as islands in a
forest landscape, after which rapid deforestation reversed the landscape structure
with islands of woodland in a cultivated landscape. This was already established
during the Iron Age, much earlier than suggested by most peat pollen diagrams.
The data used in their research still do not provide a complete picture of the
total landscape, since these samples were all taken in a settlement setting and
as a consequence all represent a by humans influenced vegetation composition.
Nevertheless this research is another confirmation that peat and lake pollen
diagrams do not necessarily reflect a complete image of a landscape, since they
might miss valuable local information.
The prehistoric landscape did not just consist of deciduous woodland, neither
does deciduous woodland with settlement islands show the whole picture or is
the barrow heathland landscape representative for the total landscape. To get a
complete image of a landscape sampling of multiple locations in different settings
is necessary. The sampling of barrows has proven to be a valuable addition to
reconstruct the landscape at a more local level.
In conclusion, the barrow landscape was a landscape dominated by heath. Heath
communities worked together for many generations to maintain these heathland
areas. These heaths were not only the final resting places for their ancestors, but
they had also been used and maintained by these ancestors. These ancestral heaths
were very stable elements in the landscape and were kept in existence as such for
thousands of years, forming the most important factor in structuring the barrow
landscape.
240
ancestral heaths
Chapter 14
Conclusions: answers to the research
questions
14.1 What did a barrow landscape look like and what was the
vegetation (history) around barrows?
From the Late Neolithic onwards barrows were built in open spaces that were
covered with heath vegetation. The heath the barrows were raised in originated
from before the barrows were built and must have been maintained by heath
management activities before and after the barrows were built. Management
activities most likely involved grazing and possibly also burning and sod cutting.
On the one hand these activities might have been applied intentionally to
maintain the heath. On the other hand maintenance of the heath might have
been a side-effect to the agricultural activities prehistoric man carried out in
their everyday life. The oldest barrows were built in heath areas with an average
distance to the forest (ADF) varying from 50 m up to 150 m. These heath areas
were often connected to each other, forming long-stretched heath areas with a
length of several kilometres, while in the late Neolithic A long alignments of
barrows were formed. From the Late Neolithic B barrows onwards barrows were
also built outside these alignments. These barrows too were built in heath areas
with an ADF of 50 to 150 m. At the same time the long-stretched heath areas were
maintained as well, while barrows in the alignments were re-used or new barrows
were added to the alignments. The open spaces the youngest (Middle to earlier
Late Iron Age) barrows were built in might have been larger in size, with an ADF
that could reach 500 m. The barrow heath was surrounded by deciduous forest. In
the relatively dry parts of the environment this deciduous forest was fairly open of
character and consisted mainly of Quercus (oak) and Tilia (lime; from the Bronze
Age onwards partly replaced by beech e.g. Fagus) with probably Corylus (hazel)
profusely present at the forest edge. The forest in the wetter parts of the area was
dominated by Alnus (alder).
In summary, the barrow landscape must have been dominated by managed
patches of heath surrounded by open forest. These heath areas contained one or
more barrows and were often connected to each other, forming passage ways in
the landscape. The barrow landscape was a stable, managed mosaic heath openforest passage landscape that was must have been maintained as such for many
generations.
14.2. Were barrows built on ancestral grounds? What is the
relationship with pastoral zones?
Based on the data that are currently available, discussed in the previous chapters,
it is most likely that barrows indeed were built on ancestral grounds. Most
barrows were situated in pastoral areas that were not only grazed when the barrows
had been built, but probably also prior to the barrow building. None of the
investigated barrows was built in areas that were very recently cleared especially
conclusions
241
for the construction of a burial mound. No indications have been found that
barrows were built in the near vicinity of a settlement or an arable field, but in
all cases barrows were built on land that had been in use by prehistoric man who
could very well be the ancestors of the builders.
14.3 What was the size of the open space barrows were constructed
in and what was the distance to the forest?
Open spaces barrows were built in varied in size from small, with an average
distance to the forest of 50-100 m, until rather large, with an average distance from
the barrow to the forest of 300-500 m, although the latter has only been found in
the relatively young Early Iron Age barrows of the Echoput. Most barrows were
probably built in open spaces with an ADF of approximately 50-150 m. Although
the forest might have been rather close to most barrows, the heathland area barrows
were built in could still have been relatively extended. Long-stretched heathland
areas with a length of several kilometres were probably not exceptional. Such
extensive heathland areas already existed in the Late Neolithic and continued to
exist for thousands of years.
14.4 What was the role of barrows in the landscape? How can the
history of the barrow environment be linked to that of the natural
and cultural landscape in the surroundings?
The role of barrows in the landscape of the central and southern Netherlands
appeared to have been twofold. First, they occupied a special place in the
landscape. Barrows were built in heath areas that were probably at a distance from
settlements and arable fields. Visibility seemed to have played an important role.
Second, they were part of the daily life of prehistoric man. The barrow landscape
was included in the economic zone of farming communities in the area, while the
heath areas were used as grazing grounds. Prehistoric landscape seems to undergo
impressive changes from the Neolithic to the Iron Age (and further on), when
prehistoric man gradually changed it to a cultural landscape. The heaths of the
barrow landscape, however, probably were very stable elements in this changing
landscape that existed as such for thousands of years.
14.5 Supplying Staatsbosbeheer with advice and suggestions, to
aide in reconstructing the original environment around barrows for
purposes of tourism
In the previous chapters has been attempted to sketch what the barrow landscape
of the 3rd and 2nd millennium BC in the central and southern Netherlands looked
like. In Chapter 1 (1.2) the societal significance of this barrow research has been
stated. Combined with the theses of Bourgeois (on the genesis of the barrow
landscape; Bourgeois 2013) and Wentink (on the social and ideological identity
of the dead; Wentink in prep.) this thesis should provide a most detailed possible
story about the barrow landscape, the barrows, who and what is buried inside
the barrows and who built them: a story that could be told to the public. The
owners of Dutch nature reserves want to present the barrows to the public in
their original environmental context (if possible). Therefore they are interested in
what the original environment looked like, information which would enable them
to adjust their management and development regime [to achieve this original
environment as much as possible].
242
ancestral heaths
The barrow landscape as reconstructed in the previous chapters has provided a
general view on what it must have looked like in reality. The reconstruction pictures
with circular patches of heath are simplifications of what the barrow landscape
must have been looked like in reality. Nevertheless, they must certainly give a good
impression of the visual impact of the heathlands barrows were built in. For the
owners of nature reserve areas that want to include barrows in their development
and management this would be a good starting point. To show the public what the
barrow landscape looked like they should be situated in a heathland area in such a
way that the barrow (or barrows) is well visible when entering that heathland area.
The size of the heathland differed from case to case and the size of the area that
should be reconstructed is probably more dependent on present day environmental
and logistical circumstances. Current environmental circumstances are different
than they were in the barrow period. Present day acidification, fertilization and
dehydration have changed the soil conditions. Consequently, these factors will be
of great influence on the maintenance of heath areas and surrounding forest. As
for heritage management: only the barrow itself is considered a monument and in
some cases the area around the barrow to a maximum of 10 metres (see 1.2). This
research has shown that a barrow was inextricably linked to the heathland around
it. The heath was most likely wider than 10 m around the monument. In addition,
the excavation of Oss-Zevenbergen (Chapter 12.1) and the Echoput (Chapter 8.1)
have shown that the area around a barrow could be of great archaeological value
(post hole structures) and it does make sense to enlarge the protected environment
around the barrow to preserve valuable Dutch cultural heritage. This thesis
provides a guide line of what the barrow landscape probably looked like in general
and it is now up to the landowners (and the cultural heritage management) how
to use it.
conclusions
243
References
Achterkamp, Y. 2009. Separating heather from the Ling. An archaeopalynological investigation
and its development at Oss-Zevenbergen during and before the Early Iron Age. Universtity
of Leiden.
Adams, S.N. 1975. Sheep and cattle grazing in forests: a review. Journal of Applied Ecology
12(1), 143-152.
Andersen, S.T. 1979. Brown earth and podzol: soil genesis illuminated by microfossil
analysis. Boreas 8(1), 59-73.
Andersen, S.T. 1988. Pollen spectra from the double passage-grave, Klekkendehøj, on
Møn. Evidence of swidden cultivation in the Neolithic of Denmark. Journal of Danish
Archaeology 7, 77-92.
Andersen, S.T. 1992. Early and Middle-Neolithic agriculture in Denmark: Pollen spectra
from soils in burial mounds of the Funnel Beaker Culture. Journal of European
Archaeology 1, 153-180.
Andersen, S.T. 1994-95. Pollen analytical investigations of barrows from the Funnel
Beaker and Single Grave Cultures in the Vrouwe area, West Jutland, Denmark. Journal
of Danish Archaeology 12, 107-132.
Andersen, S.T. 1996-97. Pollen analyses from Early Bronze Age barrows in Thy. Journal
of Danish Archaeology 13, 7-17.
Arnoldussen, S., de Kort, J.W., Bourgeois, Q. 2009. A barrow for a camp-site? The
disturbed barrow of Rhenen-Elsterberg (Unitas 4). Lunula. Archaeologica protohistorica
XVII, 177-187.
Averdieck, F.-R. 1980. Zum Stand der palynologischen Untersuchungen an Erdbauten in
Schleswig-Holstein. Offa 37, 384-393.
Bakels, C.C. 1975. Pollenspectra from the Late Bronze Age urnfield at Hilvarenbeek-Laag
Spul, Prov. Noord-Brabant Netherlands. Analecta Praehistorica Leidensia 8, 45-51.
Bakels, C.C. 1978. Four Linearbandkeramik settlements and their environment. University
of Leiden.
Bakels, C.C. 2000. Pollen diagrams and prehistoric fields: the case of Bronze Age Haarlem,
the Netherlands. Review of Palaeobotany and Palynology 109, 205-218.
Bakels, C.C. 2012. The early history of Cornflower (Centaurea cyanus L.) in the
Netherlands. Acta Palaeobotanica 52(1), 25-31.
Bakels, C.C., Achterkamp, Y. 2013. The local vegetation at the time of the construction
of the Oss-Zevenbergen mounds 7 and 6. In: D.R. Fontijn, S.A. van der Vaart, R.
Jansen (eds.). Transformation through Destruction. A monumental and extraordinary
Early Iron Age Hallstatt C barrow from the ritual landscape of Oss-Zevenbergen. Leiden:
Sidestone Press, 239-248.
Bakker, J.A., van Giffen, A.E. 1965. Hilversum, grafheuvels 5 en 6 bij de Erfgooiersstraat.
Nieuwsbulletin KNOB 9, 120-121.
Bakker, R. 2003. The emergence of agriculture on the Drenthe Plateau. A palaeobotanical
study supported by high-resolution 14C dating. University of Groningen.
Bastiaens, J., Deforce, K. 2005. Geschiedenis van de heide. Eerst natuur en dan cultuur
of andersom? Natuurfocus 4(2), 40-44.
van Beek, B.L. 1977. Sporen van een Neolithische nederzetting tussen Toterfout en Halve
Mijl. Gemeente Veldhoven (N.Br.). In: N. Roymans, J. Biemans, J. Slofstra, W.J.H.
Verwers (eds.). Brabantse Oudheden (Bijdragen tot de studie van het Brabants Heem 16,
opgedragen aan Gerrit Beex bij zijn 65ste verjaardag). Eindhoven, 43-54.
references
245
Beex, G. 1954. Twee bronstijdgrafheuvels en enige graven uit de merovingische tijd te
Casteren (gemeente Hoogeloon C.A.). Brabants Heem 6, 57-65.
Beex, G. 1957. Een Neolitische grafheuvel met beker en vuursteendolk bij Witrijt (gem.
Bergeijk). In: G. Beex (ed.). Twee grafheuvels in Noord-Brabant (Bijdragen tot de studie
van het Brabanste Heem deel IX). Eindhoven, 7-23.
Beex, G. 1959. Onderzoek van Grafheuvels te Weelde. Archaeologica Belgica 47, 1-29.
Beex, G. 1964a. Archeologisch overzicht van de gemeente Hoogeloon C.A. Brabants
Heem XVI, 99-110.
Beex, G. 1964b. Archeologische vondsten te Eersel. Brabants Heem XVI, 9.
Beex, G. 1964c. Ringwalheuvel te Alphen, prov. Noord-Brabant. Berichten van de
rijksdienst voor het oudheidkundig bodemonderzoek 14, 53-65.
Behre, K.-E. 1986. Anthropogenic indicators in pollen diagrams. Rotterdam/Boston:
Balkema.
Behre, K.-E. 1992. The history of rye cultivation in Europe. Vegetation History and
Archaeobotany 1, 141-156.
Beijerinck, W. 1933. Die mikropaläontologische Untersuchung äolischer Sedimente und
ihre Bedeutung für die Florengeschichte und die Quartairstratigraphie. Proceedings
Koninklijke Akademie van Wetenschappen 36, 107-115.
Benrath, W., Jonas, F. 1937. Zur Entstehung der Ortstein-Bleichsandschichten an der
Ostseeküste. Planta 26(4), 614-630.
Berendsen, H.J.A. 2004. De vorming van het land: inleiding in de geologie en de geomorfologie,
4th ed. Assen: Koninklijke Van Gorcum.
Beug, H.-J. 2004. Leitfaden der Pollenbestimmung für Mitteleuropa und angrenzende
Gebiete. München: Verlag Dr. Friedrich Pfeil.
Bieleman, J. 1987. Boeren op het Drentse Zand (1600-1900). Een nieuwe visie op de ‘oude’
landbouw. PhD-thesis. Wageningen.
Bokdam, J., Gleichman, J.M. 2000. Effects of grazing by free-ranging cattle on vegetation
dynamics in a continental north-west European heathland. Journal of Applied Ecology
37, 415-431.
Bourgeois, Q. 2008. Grafheuvelonderzoek in Nederland: een gedane zaak? Vitruvius 4,
16-21.
Bourgeois, Q., Amkreutz, L., Panhuysen, R. 2009. The Niersen Beaker burial: A renewed
study of a century-old excavation. Journal of Archaeology in the Low Countries 1-2,
83-105.
Bourgeois, Q.P.J. 2013. Monuments on the horizon. The formation of the barrow landscape
throughout the 3rd and 2nd millennium BC. PhD thesis. University of Leiden.
Bourgeois, Q.P.J., Fontijn, D.R. 2008. Houses and barrows in the Low Countries. In: S.
Arnoldussen, H. Fokkens (eds.). Bronze Age settlements in the Low Countries. Oxford:
Oxbow Books, 41-56.
Bourgeois, Q.P.J., Fontijn, D.R. 2011. Antiquarian leftovers - Mound 2. In: D.R. Fontijn,
Q.P.J. Bourgeois, A. Louwen (eds.). Iron Age Echoes. Prehistoric land management
and the creation of a funerary landscape - the “twin barrows” at the Echoput Apeldoorn.
Leiden: Sidestone Press, 65-90.
Braat, W.C. 1936. Een urnenveld te Knegsel (gem. Vessem). Oudheidkundige Mededelingen
van het Rijksmuseum van Oudheden te Leiden 17, 38-46.
Bradshaw, R.H.W., Hannon, G.E., Lister, A.M. 2003. A long-term perspective on
ungulate–vegetation interactions. Forest Ecology and Management 181, 267-280.
246
ancestral heaths
Brinkkemper, O., van Wijngaarden-Bakker, L. 2005. All-round farming. Food production
in the Bronze Age and the Iron Age. In: L. Louwe Kooijmans, P.W. van den Broeke, H.
Fokkens, A. van Gijn (eds.). The prehisotry of the Netherlands. Amsterdam: Amsterdam
University Press, 491-512.
Van den Broeke, P.W. 2005. A prehistory of our time. In: L.P. Louwe Kooijmans, P.W. van
den Broeke, H. Fokkens, A.L. van Gijn (eds.). Prehistory of the Netherlands. Amsterdam
Amsterdam University Press, 17-32.
Brongers, J.A. 1976. Air photography and Celtic Field research in the Netherlands. PhD
thesis. University of Groningen. Amersfoort: ROB.
Broström, A., Gaillard, M.-J., Ihse, M., Odgaard, B. 1998. Pollen-landscape relationships
in modern analogues of ancient cultural landscapes in southern Sweden - a first step
towards quantification of vegetation openness in the past. Vegetation History and
Archaeobotany 7(189-201).
Broström, A., Nielsen, A.B., Gaillard, M.-J., Hjelle, K., Mazier, F., Binney, H., Bunting,
J., Fyfe, R., Meltsov, V., Poska, A., Räsänen, S., Soepboer, W., von Stedingk, H.,
Suutari, H., Sugita, S. 2008. Pollen productivity estimates of key European plant taxa
for quantitative reconstruction of past vegetation: a review. Vegetation History and
Archaeobotany 17, 461-478.
Broström, A., Sugita, S., Gaillard, M.-J. 2004. Pollen productivity estimates for the
reconstruction of past vegetation cover in the cultural landscape of southern Sweden.
The Holocene 14, 368-381.
Broström, A., Sugita, S., Gaillard, M.-J., Pilesjö, P. 2005. Estimating the spatial scale of
pollen dispersal in the cultural landscape of southern Sweden. The Holocene 15(2),
252-262.
Bunnik, F.P.M. 1999. Vegetationsgeschichte der Löβbörden zwischen Rhein und Maas von der
Bronzezeit bis in die frühe Neuzeit. PhD thesis. University of Utrecht.
Bunting, M.J., Hjelle, K.L. 2010. Effect of vegetation data collection strategies on
estimates of relevant source area of pollen (RSAP) and relative pollen productivity
estimates (relative PPE) for non-arboreal taxa. Vegetation History and Archaeobotany
19, 365-374.
Bunting, M.J., Tipping, R. 2001. ‘‘Anthropogenic’’ Pollen Assemblages from a Bronze Age
Cemetery at Linga Fiold, West Mainland, Orkney. Journal of Archaeological Science 28,
487-500.
Bursch, F.C. 1933. Het urnenveld op het landgoed Oostereng onder Bennekom. OMROL
14, 26-38.
Bursch, F.C. 1935. Oudheidkundige onderzoekingen te Hilversum, 1934. OMROL 16,
45-63.
Bursch, F.C. 1937. Grafheuvels te Oss. Oudheidkundige Mededelingen van het Rijksmuseum
van Oudheden te Leiden 18, 1-3.
Butler, J.J., van der Waals, J.D. 1967. Bell Beaker and early metal-working in the
Netherlands. Palaeohistoria 12, 41-139.
Casparie, W.A. 1976. Palynological investigation of the celtic field near Vaassen, the
Netherlands. In: J.A. Brongers (ed.). Air photography and Celtic field research in the
Netherlands. (PhD thesis, University of Groningen). Amersfoort: ROB, 105-113.
Casparie, W.A., Groenman-van Waateringe, W. 1980. Palynological analysis of Dutch
barrows. Palaeohistoria 22, 7-65.
Castel, I.L.Y., Koster, E.A., Slotboom, R.T. 1989. Morphogenetic aspects and age of Late
Holocene aeolian drift sands in Northwest Europe. Zeitschrift für Geomorphologie,
Neue Folge 33, 1-26.
references
247
Davidson, D.A., Carter, S., Boagc, B., Longa, D., Tipping, R., Tyler, A. 1999. Analysis
of pollen in soils: processes of incorporation and redistribution of pollen in five soil
profile types. Soil Biology and Biochemistry 31, 643-653.
Davis, M.B. 1963. On the theory of pollen analysis. American Journal of Science 261,
897-912.
Dijkstra, E.F., van Mourik, J.M. 1995. Palynology of young acid forest soils in the
Netherlands. Meded. Rijks Geol. Dienst 52, 283-296.
Dijkstra, E.F., van Mourik, J.M. 1996. Reconstruction of forest dynamics based on pollen
analysis and micromorphological studies of young acid forest soils under Scots pine
plantations. Acta Botanica Neerlandica 45(3), 393-410.
Dimbleby, G.W. 1952. Soil regeneration on the North-East Yorkshire Moors. Journal of
Ecology 40(2), 331-341.
Dimbleby, G.W. 1957. Pollen Analysis of Terrestrial Soils. New Phytologist 56, 12-28.
Dimbleby, G.W. 1961. Soil pollen analysis. European Journal of Soil Science 12(1), 1-10.
Dimbleby, G.W. 1985. The palynology of archaeological sites. London: Acedemic Press,
Inc.
Dimbleby, G.W., Evans, J.G. 1974. Pollen and land-snail analysis of calcareous soils.
Journal of Archaeological Science 1(2), 117-133.
Diot, M.-F. 1992. Etudes palynologiques de ble´s sauvages et domestiques issus de
cultures expe´rimentales. In: P.C. Anderson (ed.). Préhistoire de l’agriculture. Monographie
du CRA 6. CNRS, Paris, 107-111.
Doorenbosch, M. 2011. An environmental hisotry of the Echoput barrows. In: D.R.
Fontijn, Q.P.J. Bourgeois, A. Louwen (eds.). Iron Age Echoes. Prehistoric land
management and the creation of a funerary landsacpe - the “twin barrows” at the Echoput
in Apeldoorn. Leiden: Sidestone Press, 111-128.
Doorenbosch, M. 2013. A history of open space. Barrow landscapes and the significance
of heaths - the case of the Echoput barrows. In: D. Fontijn, A.J. Louwen, S. van
der Vaart, K. Wentink (eds.). Beyond barrows. Current research on the structuration
and perception of the prehistoric landscape through monuments. Leiden: Sidestone Press,
197-224.
Downes, J. 1994. Excavation of a Bronze Age burial at Mousland, Stromness, Orkney.
Proceedings of the Society of Antiquaries of Scotland 124, 141-154.
Dreibrodt, S., Nelle, O., Lütjens, I., Mitusov, A., Clausen, I., Bork, H.-R. 2009.
Investigations on buried soils and colluvial layers around Bronze Age burial mounds
at Bornhöved (northern Germany): an approach to test the hypothesis of ‘landscape
openness’ by the incidence of colluviation. The Holocene 19(3), 487-497.
Elbersen, B.S., Kuiters, A.T., Meulenkamp, W.J.H., Slim, P.A. 2003. Schaapskuddes in het
natuurbeheer. Alterra-rapport 735. Wageningen: Alterra, Research Instituut voor de
Groene Ruimte.
Erdtman, G. 1943. An introduction to pollen analysis. Waltham, Massachusetts: Chronica
Botanica Company.
Faegri, K. 1966. Some problems of representativity in pollen analysis. Palaeobotanist 15,
135-140.
Faegri, K., Iversen, J. 1989. Textbook of pollen analysis, 4th ed. Chichester: Joh Wiley and
sons.
Fagerlind, F. 1952. The real signification of pollen diagrams. Botaniska notiser 105,
185-224.
248
ancestral heaths
Fanta, J. 1995. Beuk (Fagus sylvatica L.) in het Nederlandse deel van het nw-Europees
diluvium. Nederlands Bosbouw Tijdschrift 1995, 225-234.
Field, D. 1998. Round barrows and the harmonious landscape: placing Early Bronze
Age burial monuments in South-East England. Oxford Journal of Archaeology 17(3),
309-326.
Firbas, F., Losert, H., Broihan, H. 1939. Untersuchungen zur jüngeren Vegetationsgeschichte
im Oberharz. Planta 30(3), 422-456.
Florschutz, F. 1941. Resultaten van microbotanisch onderzoek van het complex loodzandoerzand en van daaronder en daarboven gelegen afzettingen. Besprekingen over het
Heidepodsolprofiel. 10e bijeenkomst sectie Nederland Internationale Bodemkundige
Vereeniging. 18-19 april 1941. Utrecht: Hoitsema, 1-21.
Fokkens, H. 1991. Verdrinkend landschap. Archeologisch onderzoek van het westelijk FriesDrents Plateau 4400 BC tot 500 cal AD. PhD thesis. University of Groningen.
Fokkens, H. 2005a. Longhouses in unsettled settlements settlements in Beaker period
and Bronze Age. In: L.P. Louwe Kooijmans, P.W. van den Broeke, H. Fokkens, A.
van Gijn (eds.). The Prehistory of the Netherlands. Amsterdam: Amsterdam University
Press, 407-428.
Fokkens, H. 2005b. Mixed farming societies: synthesis. In: L.P. Louwe Kooijmans, P.W.
van den Broeke, H. Fokkens, A.L. van Gijn (eds.). Prehistory of the Netherlands.
Amsterdam Amsterdam University Press, 463-474.
Fokkens, H., van der Vaart, S.A., Fontijn, D.R., Lemmers, S.A.M., Jansen, R., Valentijn,
P. 2012. Hallstatt burials of Oss in context. Analecta Praehistorica Leidensia in press.
Fokkens, H., Jansen, R. 2004. Het vorstengraf van Oss. Een archeologische speurtocht naar
een prehistorisch grafveld. Utrecht.
Fokkens, H., Jansen, R., van Wijk, I.M. 2009a. Monumenten en herinnering; het grafveld
Oss-Zevenbergen in samenhang. In: H. Fokkens, R. Jansen, I.M. van Wijk (eds.). OssZevenbergen: de langetermijn-geschiedenis van een prehistorisch grafveld. Leiden: Archol
rapport 50, 209-224.
Fokkens, H., Jansen, R., van Wijk, I.M. (eds.) 2009b. Oss-Zevenbergen: de langetermijngeschiedenis van een prehistorisch grafveld. Leiden: Archol Rapport 50.
Fontijn, D.R. 2007. Ancestral Mounds. The social and ideological significance of barrows,
2900-1100 BC. Project Proposal N.W.O. - programmatic research. Leiden.
Fontijn, D.R. 2011a. Barrow excavations at the Echoput. Problem, research aims and
methods of the 2007 fieldwork campaign. In: D.R. Fontijn, Q.P.J. Bourgeois, A.
Louwen (eds.). Iron Age Echoes. Prehistoric land management and the creation of a
funerary landscape - the “twin barrows” at the Echoput Apeldoorn. Leiden: Sidestone
Press, 13-32.
Fontijn, D.R. 2011b. The genesis and history of the Echoput barrows. In: D.R. Fontijn,
Q.P.J. Bourgeois, A. Louwen (eds.). Iron Age Echoes. Prehistoric land management and
the creation of a funerary landsacpe - the “twin barrows” at the Echoput in Apeldoorn.
Leiden: Sidestone Press, 149-162.
Fontijn, D.R., Bourgeois, Q.P.J., Louwen, A. (eds.) 2011. Iron Age Echoes. Prehistoric land
management and the creation of a funerary landsacpe - the “twin barrows” at the Echoput
in Apeldoorn. Leiden: Sidestone Press.
Fontijn, D.R., van der Vaart, S.A. 2013. Dismantled, transformed and deposited prehistoric bronze from the centre of mound 7. In: D.R. Fontijn, S.A. van der Vaart,
R. Jansen (eds.). Transformation through Destruction. A monumental and extraordinary
Early Iron Age Hallstatt C barrow from the ritual landscape of Oss-Zevenbergen. Leiden:
Sidestone Press, 151-194.
references
249
Fontijn, D.R., Jansen, R., Bourgeois, Q.P.J., van der Linde, C. 2013a. Excavating the
seventh mound. In: D.R. Fontijn, S.A. van der Vaart, R. Jansen (eds.). Transformation
through Destruction. A monumental and extraordinary Early Iron Age Hallstatt C barrow
from the ritual landscape of Oss-Zevenbergen. Leiden: Sidestone Press, 69-118.
Fontijn, D.R., Jansen, R., van der Vaart, S.A., Fokkens, H., van Wijk, I.M. 2013b.
Conclusion: The seventh mound of seven mounds – long-term history of the
Zevenbergen barrow landscaoe. In: D.R. Fontijn, S.A. van der Vaart, R. Jansen (eds.).
Transformation through Destruction. A monumental and extraordinary Early Iron Age
Hallstatt C barrow from the ritual landscape of Oss-Zevenbergen. Leiden: Sidestone
Press, 281-316.
Fontijn, D.R., Louwen, A. (eds.) in prep. Re-united in death. Collective burials in the
Bronze Age barrow landscape of Apeldoorn-Wieselse Weg. Leiden: Sidestone Press.
Fyfe, R.M. 2012. Bronze Age landscape dynamics: spatially detailed pollen analysis from
a ceremonial complex. Journal of Archaeological Science 39(8), 2764-2773.
Gaillard, M.-J., Sugita, S., Bunting, M.J., Middleton, R., Broström, A., Caseldine, C.,
Giesecke, T., Hellman, S.E.V., Hicks, S., Hjelle, K., Langdon, C., Nielsen, A.-B.,
Poska, A., von Stedingk, H., Veski, S., POLLANDCAL-members. 2008. The use
of modelling and simulation approach in reconstructing past landscapes from fossil
pollen data: a review and results from the POLLANDCAL network. Vegetation History
and Archaeobotany 17, 419-443.
van Geel, B. 1978. A palaeoecological study of Holocene peat bog sections in Germany
and the Netherlands, based on the analysis of pollen, spores and macro- and
microscopic remains of fungi, algae, cormophytes and animals. Review of Palaeobotany
and Palynology 25, 1-120.
van Giffen, A.E. 1930. Die Bauart der Einzelgraber. Beitrag zur Kenntnis der älteren
individuellen Grabhügelstrukturen in den Niederlanden. Leipzig.
van Giffen, A.E. 1949. Een vierperioden-tumulus met primaire klokbekerheuvel bij
Schaaik, Ge. Oss, N.Br. Brabants Heem 1, 71-74.
van Giffen, A.E. 1954. Een Meerperioden-heuvel, Tumulus I te Bennekom, gem. Ede.
Bijdragen en Mededelingen van de Vereniging Gelre LIV, 1-21.
van Giffen, A.E. 1958. Renkum. Archeologisch Nieuws in het Bulletin van de Koninklijke
Nederlandse Oudheidkundige Bond 11(4), 66.
van Gijn, A., Louwe Kooijmans, L.P. 2005. The first farmers: synthesis. In: L.P. Louwe
Kooijmans, P.W. van den Broeke, H. Fokkens, A. van Gijn (eds.). The prehistory of the
Netherlands. Amsterdam: Amsterdam University Press, 337-356.
Glasbergen, W. 1954. Barrow excavations in the Eight Beatitudes. The Bronze Age cemetery
between Toterfout & Halve Mijl, North-Brabant.
Glazema, P. 1951. Oudheidkundige onderzoekingen in de provincie Gelderland,
gedurende de jaren 1946 tot en met 1950. Gelre 51, 1-120.
Grimm, E.C. 1992 Published. Tilia and Tilia-graph: pollen spreadsheet and graphic programs.
8th International Palynological Congress, Aix-en-Provence (France), September 6-12,
56.
Groenewoudt, B., van Haaster, H., van Beek, R., Brinkkemper, O. 2007. Towards a reverse
image. Botanical research into the landscape history of the Eastern Netherlands (1100
B.C. - A.D. 1500). Landscape History 29, 17-33.
Groenman- van Waateringe, W. 1977. Palynologisch onderzoek van grafheuvels te Weelde,
Belgische Kempen. Archaeologica Belgica 193, 42-49.
Groenman- van Waateringe, W. 1979. Palynological investigations of five German burial
mounds. Archaeo-Physica 8, 69-84.
250
ancestral heaths
Groenman- van Waateringe, W. 1983. Appendix II: Pollen and seeds. In: G. Eogan (ed.).
Excavations at Knowth 1: Smaller Passage Tombs, Neolithic Occupation and Beaker
Activity. Dublin: Royal Irish Academy, 325-329.
Groenman- van Waateringe, W. 1986. Grazing possibilities in the Neolithic of the
Netherlands based on palynological data. In: K.-E. Behre (ed.). Anthropogenic
Indicators in Pollen Diagrams. Rotterdam: Balkema, 187-202.
Groenman- van Waateringe, W., Voorrips, A., Van Wijngaarden-Bakker, H. 1968.
Settlements of the Vlaardingen Culture at Voorschoten and Leidschendam (Ecology).
Helenium 7, 105-130.
Groenman-van Waateringe, W. 1978. The impact of Neolithic man on the landscape in
the Netherlands. In: Limbrey (ed.). The effect of man on the landscape: the Lowland
Zone. London: Council for British Archaeology.
Groenman-van Waateringe, W. 1993. The effects of grazing on the pollen production of
grasses. Vegetation History and Archaeobotany 2, 157-162.
Guillet, B. 1970. Etude palynologique des podzols. Pollen et Spores XII, 45-69.
Hannon, G.E., Bradshaw, R.H.W., Nord, J., Gustafsson, M. 2008. The Bronze Age
landscape of the Bja¨re peninsula, southern Sweden, and its relationship to burial
mounds. Journal of Archaeological Science 35, 623-632.
Havinga, A.J. 1962. Een palynologisch onderzoek van in dekzand ontwikkelde bodemprofielen.
Dissertation. Landbouwhogeschool Wageningen.
Havinga, A.J. 1963. A palynological investigation of soil profiles developed in cover sand.
Mededelingen van de Landbouwhogeschool Wageningen 63(1), 1-93.
Havinga, A.J. 1964. Investigation into the differential corrosion susceptibility of pollen
and spores. Pollen et Spores 6(2), 621-635.
Havinga, A.J. 1967. Palynology and pollen preservation. Review of Palaeobotany and
Palynology 2, 81-98.
Havinga, A.J. 1984. A 20-year experimental investigation into the differential corrosion
susceptibilty of pollen and spores in various soil types. Pollen et Spores 26(3-4),
541-558.
Hjelle, K.L. 1999. Modern pollen assemblages from mown and grazed vegetation types in
western Norway. Review of Palaeobotany and Palynology 107, 55-81.
van Hoeve, M.L., Hendrikse, M. (eds.) 1998. A study of non-pollen objects in pollen slides.
The types as described by Dr. Bas van Geel and colleagues. . Utrecht.
Holloway, R.G. 1989. Experimental mechanical pollen degradation and its application to
quaternary age deposits. The Texas journal of science 41, 131-145.
Holwerda, J.H. 1907. Grafheuvels bij Hoogsoeren. Oudheidkundige Mededelingen van het
Rijksmuseum van Oudheden te Leiden I, 7-10.
Holwerda, J.H. 1908. Tumuli bij Nierssen. Oudheidkundige Mededelingen van het
Rijksmuseum van Oudheden te Leiden II, 1-17.
Holwerda, J.H. 1912. Opgravingen aan het Uddelermeer. Oudheidkundige Mededelingen
van het Rijksmuseum van Oudheden te Leiden VI, 1-16.
van Hoof, L. 2009. Vuurstenen en natuurstenen artefacten uit het mesolithicum. In: H.
Fokkens, R. Jansen, I.M. van Wijk (eds.). Oss-Zevenbergen: de langetermijn-geschiedenis
van een prehistorisch grafveld. Leiden: Archol rapport 50, 183-186.
Hulst, R.S. 1971. Renkum. Nieuwsbulletin KNOB 70, 114-115.
Hulst, R.S. 1972. Epe. Nieuwsbulletin KNOB 71, 13.
references
251
Hulst, R.S., Lanting, J.N., van der Waals, J.D. 1973. Grabfunde mit frühen Glochenbechern
aus Gelderland und Limburg. Berichten van de rijksdienst voor het oudheidkundig
bodemonderzoek 23, 77-101.
IJzereef, G.F. 1981. Bronze Age animal bones from Bovenkarspel. The excavation at Het
Valkje. PhD thesis. University of Amsterdam.
van Impe, L., Beex, G. 1977. Grafheuvels uit de vroege en midden-bronstijd te Weelde 193.
Brussel.
ISRIC-FAO, 2006. World Reference Base for Soil Resources 2006. World soil resources
reports 103.
Iversen, J. 1941. Land occupation in Denmark’s Stone Age. A pollen-analytical study of
the influence of farmer culture on the vegetational development Danmarks Geologiske
Undersøgelse II 66, 5-23.
Iversen, J. 1973. The development of Denmark’s nature since the Last Glacial. Danmarks
Geologiske Undersøgelse V. Række. Nr. 7-C, 1-126.
Jansen, R., van der Linde, C. 2013. The physical and archaeological “landscape” of the
Zevenbergen barrow group. In: D.R. Fontijn, S.A. van der Vaart, R. Jansen (eds.).
Transformation through Destruction. A monumental and extraordinary Early Iron Age
Hallstatt C barrow from the ritual landscape of Oss-Zevenbergen. Leiden: Sidestone
Press, 35-46.
Jansen, R., Fokkens, H. 2007. Het vorstengraf van Oss re-reconsidered. Archeologisch
onderzoek Oss-Vorstengrafdonk 1997-2005. Leiden: Archol rapport 49.
Jansen, R., Louwen, A. (eds.) in prep. Heidense heuvels op de Slabroekse heide.
Janssen, C.R. 1974. Verkenningen in de palynologie. Utrecht: Oosthoek, Scheltema &
Holkema.
Janssen, C.R., Ten Hove, H.A. 1971. Some late-Holocene pollen diagrams from the Peel
raised bogs (Southern Netherlands) Review of Palaeobotany and Palynology 11, 7-53.
Johansen, K.L., Laursen, S.T., Holst, M.K. 2004. Spatial patterns of social organization
in the Early Bronze Age of South Scandinavia. Journal of Anthropological Archaeology
23, 33-55.
Jonassen, H. 1950. Recent pollen sedimentation and Jutland heath diagrams. Dansk
Botanisk Arkiv 13(7), 1-168.
Jungerius, P.D., Riksen, M.J.P.M. 2010. Contribution of laser altimetry images to the
geomorphology of the Late Holocene inland drift sands of the European Sand Belt.
Baltica 23(1), 59-70.
Karg, S. 2008. Direct evidence of heathland management in the early Bronze Age (14th
century B.C.) from the grave-mound Skelhøj in western Denmark. Vegetation History
and Archaeobotany 17, 41-49.
de Kort, J.-W. 1999. Een reconstructie van de vegetatie rond het Vorstengraf van Oss in de
Vroege IJzertijd University of Leiden.
de Kort, J.-W. 2002. Schapen op de heide. Een vegetatiereconstructie van de omgeving van
het vorstengraf van Oss in de Vroege IJzertijd. In: H. Fokkens, Jansen, R. (ed.). 2000
jaar bewoningsdynamiek. Brons- en ijzertijdbewoning in het Maas-Demer-Scheldegebied.
Leiden: Leiden University, 341-353.
de Kort, J.W. 2005. Palynologisch onderzoek van een grafheuvel uit de Midden Bronstijd bij
de Vorssel, gemeente Bernheze. Amsterdam: RAAP-notitie 1276.
de Kort, J.W. 2009. Palynologisch onderzoek Zevenbergen. In: H. Fokkens, R. Jansen,
I.M. van Wijk (eds.). Oss-Zevenbergen: de langetermijn-geschiedenis van een prehistorisch
grafveld. Leiden: Archol rapport 50, 155-182.
252
ancestral heaths
de Kort, J.W. 2010. Palynologisch onderzoek Slabroekse Heide. In: I.M. van Wijk, R.
Jansen (eds.). Het urnenveld Slabroekse Heide op de Maashorst. Leiden: Archol bv,
57-66.
de Kort, J.W., van Mourik, J.M. 2005. Palynologisch onderzoek Slabroekse Heide. In:
I.M. van Wijk, R. Jansen (eds.). Het urnenveld Slabroekse Heide op de Maashorst. Een
verkennend en waarderend archeologisch proefsleuvenonderzoek. Leiden: Archol rapport
nummer 72, 57-66.
Koster, E.A. 1978. De stuifzanden van de Veluwe: een fysisch-geografische studie. PhD thesis.
University of Amsterdam.
van der Laan, K., Jansen, R., de Kort, J.W., van Eijk, G. 2011. Natuurgebied de Maashorst.
In: R. Jansen, K. van der Laan (eds.). Verleden van een bewogen landschap. Landschapsen bewoningsgeschiedenis van de Maashorst. . Utrecht, 8-19.
Lake, S., Bullock, J.M., Hartley, S. 2001. Impacts of livestock grazing on lowland heathland
in the UK. English Nature Research Reports, 422. Peterborough: NERC Centre for
Ecology and Hydrology and Sussex University.
Lanner, R.M. 1966. Needed: a new approach to the study of pollen dispersion. Silvae
Genetica 15, 50-52.
Lanting, J.N., van der Waals, J.D. 1971a. Ede. Nieuwsbulletin KNOB 70, 125-126.
Lanting, J.N., van der Waals, J.D. 1971b. Ermelo. Nieuwsbulletin KNOB 70, 126.
Lanting, J.N., van der Waals, J.D. 1971c. Laat-Neolithische grafheuvels bij Vaassen en
Maarsbergen. OMROL 52, 93-127.
Lanting, J.N., van der Waals, J.D. 1971d. Stroe. Nieuwsbulletin KNOB 70, 126.
Lanting, J.N., van der Waals, J.D. 1971e. Vaassen. Nieuwsbulletin KNOB 70, 53-54.
Lanting, J.N., van der Waals, J.D. 1972a. Bennekom. Nieuwsbulletin KNOB 71, 129.
Lanting, J.N., van der Waals, J.D. 1972b. Renkum. Nieuwsbulletin KNOB 71, 131-132.
Lanting, J.N., van der Waals, J.D. 1972c. Vierhouten. Nieuwsbulletin KNOB 71, 132.
Lanting, J.N., van der Waals, J.D. 1976. Beaker culture relations in the lower Rhine
basin. In: J.N. Lanting, J.D. van der Waals (eds.). Glockenbecher symposion, Oberried,
1974. Haarlem: Fibula-van Dishoeck, 1-80.
Lawson, A.J., Martin, E.A., Priddy, D. 1981. The barrows of East Anglia. East Anglian
Archaeology Report 12.
Lemmers, P. 2008. Investigating optically stimulated luminescence in naturally occurring
quartz with the purpose of dating man-made burial mounds. Unpublished MSc thesis.
University of Delft.
Lemmers, S.A.M., Janssen, M., Waters-Rist, A., Grosskopf, B., Hoogland, M., Amkreutz,
L. in prep. The Chieftain of Oss: New perspectives on an Iron-Age individual with
DISH.
van der Linde, C., Fokkens, H. 2009. Het landschappelijk kader van De Zevenbergen.
In: H. Fokkens, R. Jansen, I.M. van Wijk (eds.). Oss-Zevenbergen: de langetermijngeschiedenis van een prehistorisch grafveld. Leiden: Archol rapport 50, 37-52.
van der Linde, C., Fontijn, D.R. 2011. Mound 1- A monumental Iron Age barrow. In:
D.R. Fontijn, Q.P.J. Bourgeois, A. Louwen (eds.). Iron Age Echoes. Prehistoric land
management and the creation of a funerary landscape - the “twin barrows” at the Echoput
Apeldoorn. Leiden: Sidestone Press, 33-64.
Louwe Kooijmans, L.P. 1974. The Rhine/Meuse delta. Four studies on its prehistoric
occupation and holocene geology. University of Leiden.
references
253
Louwe Kooijmans, L.P. 1979. Opgraven in Nederland. Het werk van de archeoloog.
Staatsuitgeverij ‘s Gravenhage.
Louwen, A., Fontijn, D.R., Valentijn, P., van der Linde, C., Verbaas, A. 2011. Finds.
In: D. Fontijn, Q. Bourgeois, A. Louwen (eds.). Iron Age Echoes. Prehistoric land
management and the creation of a funerary landsacpe - the “twin barrows” at the Echoput
in Apeldoorn. Leiden: Sidestone Press, 129-141.
Mallik, A.U., FitzPatrick, E.A. 1996. Thin section studies of Calluna heathland soils
subject to prescribed burning. Soil Use and Management 12(3), 143-149.
Mellars, P. 1976. Fire ecology, animal populations and man. Proceedings of the Prehistoric
Society 42, 15-45.
Middleton, R., Bunting, M.J. 2004. Mosaic v1.1: landscape scenario creation software for
simulation of pollen dispersal and deposition. Review of Palaeobotany and Palynology
132, 61-66.
Mitchell, F.J.G. 2005. How open were European primeval forest? Hypothesis testing
using palaeoecological data. Journal of Ecology 93, 168-177.
Modderman, P.J.R. 1954. Grafheuvelonderzoek in Midden-Nederland. Berichten van de
rijksdienst voor het oudheidkundig bodemonderzoek V, 7-44.
Modderman, P.J.R. 1955. Het onderzoek van enkele Brabantse en Utrechtse grafheuvels.
Berichten van de rijksdienst voor het oudheidkundig bodemonderzoek VI, 44-65.
Modderman, P.J.R. 1964. De verspreiding van de Bekerculturen op de Veluwe. Berichten
van de rijksdienst voor het oudheidkundig bodemonderzoek 5, 7-44.
Modderman, P.J.R., Bakels, C.C. 1971. A Bronze Age barrow on the Moormanlaan near
Knegsel, prov. Noord-Brabant. Analecta Praehistorica Leidensia 4, 52-56.
Modderman, P.J.R., Louwe Kooijmans, L.P. 1966. The Heibloem, a Cemetery from the
Late Bronze Age and Early Iron Age between Veldhoven and Steensel, Prov. NoordBrabant. Analecta Praehistorica Leidensia II, 9-26.
Moore, P.D., Webb, J.A., Collinson, M.E. 1991. Pollen analysis, 2nd ed. Oxford: Blackwell
Scientific Publication.
Mothes, K., Arnoldt, G., Redmann, H. 1937. Zur Bestandesgeschichte ostpreussicher
Wälder. Schriften der Physikalisch-Ökonomischen Gesellschaft zu Königsberg in Preussen
69, 267-282.
van Mourik, J.M. 1985. Een eeuwenjong podzolprofiel op de Schaijksche heide. Geografisch
Tijdschrift XIX(2), 105-112.
van Mourik, J.M. 1986. Pollen profiles of slope deposits in the Galicean area (N.W. Spain).
PhD thesis. University of Amsterdam.
van Mourik, J.M. 1987. Laat glaciaal veen en holoceen stuif-zand bij de Peelbreuk tussen
Nistelrode en Uden. Geografisch Tijdschrift XXI, 421-436.
van Mourik, J.M. 1999. The use of micromorphology in soil pollen analysis The
interpretation of the pollen content of slope deposits in Galicia, Spain. Catena 35,
239-257.
van Mourik, J.M. 2001. Pollen and spores, preservation in ecological settings. In: E.G.
Briggs, P.R. Crowther (eds.). Palaeobiology II. Oxford: Blackwell Science, 315-318.
van Mourik, J.M. 2003. Life cycle of pollen grains in mormoder humus forms of young
acid forest soils: a micromorphological approach. Catena 54, 651-663.
van Mourik, J.M. 2005. Resultaten van het dateringsonderzoek van grafheuvel 39 op de
Slabroekse Heide. In: I.M. van Wijk, R. Jansen (eds.). Het urnenveld Slabroekse Heide
op de Maashorst. Een verkennend en waarderend archeologisch proefsleuvenonderzoek.
Leiden: Archol rapport nummer 72, 67-74.
254
ancestral heaths
van Mourik, J.M., Nierop, K.G.J., Vandenberghe, D.A.G. 2010. Radiocarbon and
optically stimulated luminescence dating based chronology of a polycyclic driftsand
sequence at Weerterbergen (SE Netherlands). Catena 80, 170-181.
van Mourik, J.M., Pet, A. 2001. Natuurontwikkeling in het dal van de Venloop. Broekbos
of ecobeek? Nederlands Bosbouw Tijdschrift, 12-15.
van Mourik, J.M., Seijmonsbergen, A.C., Jansen, B. 2012a. Geochronology of soils and
landforms in cultural landscapes on aeolian sandy substrates, based on radiocarbon
and optically stimulated luminescence dating (Weert, SE-Netherlands). InTech,
Weert, SE-Netherlands. Radiometric Dating, 75-114.
van Mourik, J.M., Seijmonsbergen, A.C., Slotboom, R.T., Wallinga, J. 2012b. The impact
of human land use on soils and landforms in cultural landscapes on aeolian sandy
substrates (Maashorst, SE Netherlands). Quarternary International 265, 74-89.
van Mourik, J.M., Slotboom, R.T., Wallinga, J. 2011. Chronology of plaggic deposits;
palynology, radiocarbon and optically stimulated luminescence dating of the Posteles
(NE-Netherlands). Catena 84(1-2), 54-60.
Müller, P.E. 1884. Studier over skovjord, som bidrag til skovdyrkningens theori. II.
Tidsskrift for Skovbrug 7, 1-232.
Munaut, A.V. 1967. Recherches paleo-ecologiques en basse et moyenne Belgique. Acta
Geographica Lovaniensia 6.
Nielsen, A.B. 2004. Modelling pollen sedimentation in Danish lakes at c. AD 1800: an
attempt to validate the POLLSCAPE model. Journal of Biogeography 31, 1693-1709.
Odgaard, B. 1988. Heatland History in Western Jutland, Denmark. In: H.H. Birks,
H.J.B. Birks, P.E. Kaland, D. Moe (eds.). The cultural landscape: the past, present and
future. Cambridge: Cambridge University Press, 311-320.
Odgaard, B.V. 1994. The Holocene vegetation history of northern West Jutland, Denmark.
Opera Botanica 123, 1-171.
Odgaard, B.V., Rostholm, H. 1987. A single grave barrow at Harreskov, Jutland. Journal
of Danish Archaeology 6, 87-100.
Out, W.A. 2009. Sowing the seed? Human impact and plant subsistence in Dutch wetlands
during the late Mesolithic and Early and Middle Neolithic (5500-3400 cal. B.C.).
Archaeological Studies Leiden University. vol. 18. Leiden: Leiden University Press.
Ovrevik, S. 1990. The second millennium and after. In: A.C. Renfrew (ed.). The prehistory
of Orkney. Edinburgh.
Pape, J.C. 1970. Plaggen soils in The Netherlands. Geoderma 4(3), 229-255.
Parsons, R.W., Prentice, I.C. 1981. Statistical approaches to R-values and the pollenvegetation relationship Review of Palaeobotany and Palynology 32, 127-152.
Pidek, I.A., Piotrowska, K., Kasprzyk, I. 2010. Pollen-vegetation relationships for pine
and spruce in southeast Poland on the basis of volumetric and Tauber trap records.
Grana 49, 215-226.
Piek, H. 2000. 30 jaar begrazing bij Natuurmonumenten.
Polak, B. 1959. Palynology of the Uddelermeer. Acta Botanica Neerlandica 9, 547-571.
Poska, A., Meltsov, V., Sugita, S., Vassiljev, J. 2011. Relative pollen productivity estimates
of major anemophilous taxa and relevant source area of pollen in a cultural landscape
of the hemi-boreal forest zone (Estonia). Review of Palaeobotany and Palynology 167(12), 30-39.
Prentice, I.C. 1985. Pollen Representation, Source Area, and Basin Size: Toward a Unified
Theory of Pollen Analysis. Quaternary Research 23, 76-86.
references
255
Prentice, I.C., Parsons, R.W. 1983. Maximum Likelihood Linear Calibration of Pollen
Spectra in Terms of Forest Composition. Biometrics 39(4), 1051-1057.
Prentice, I.C., Webb, T., III 1986. Pollen percentages, tree abundances, and the Fagerlind
effect. Journal of Quaternary Science 1, 35-43.
Punt , W., Blackmore, S., Clarke, G.C.S. 1976. Northwest European flora 1. Amsterdam:
Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 1980. Northwest European pollen flora 2.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 1981. Northwest European pollen flora 3.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 1984. Northwest European pollen flora 4.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 1988. Northwest European pollen flora 5.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 1991. Northwest European pollen flora 6.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 1995. Northwest European pollen flora 7.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 2003. Northwest European pollen flora 8.
Amsterdam: Elsevier.
Punt , W., Blackmore, S., Clarke, G.C.S. 2009. Northwest European pollen flora 9.
Amsterdam: Elsevier.
Punt , W., Hoen, P.P., Blackmore, S., Nilsson, S., Le Thomas, A. 2007. Glossary of pollen
and spore terminology. Review of Palaeobotany and Palynology 143, 1-81.
RADAR. 2006. Dutch Relational Archaeobotanical Database version 2006.
Reille, M. 1992. Pollen et spores d’Europe et d’Afrique du Nord. Marseille: Laboratoire de
Botanique historique et Palynologie.
Reille, M. 1995. Pollen et spores d’Europe et d’Afrique du Nord. Supplement 1. Marseille:
Laboratoire de Botanique Historique et Palynolgie.
Reille, M. 1998. Pollen et spores d’Europe et d’Afrique du Nord. Supplement 2. Marseille:
Laboratoire de Botanique Historique et Palynologie.
Remouchamps, A.E. 1924. Opgravingen van een urnenveld te Uden. Oudheidkundige
Mededelingen van het Rijksmuseum van Oudheden te Leiden 5, 69-76.
Remouchamps, A.E. 1928. Grafheuvelonderzoekingen. . OMROL 9, 58-74.
Renes, J. 1988. De geschiedenis van het Zuidlimburgse cultuurlandschap. Assen-Maastricht:
Van Gorcum.
Riksen, M., Ketner-Oostra, R., van Turnhout, C., Nijssen, M., Goossens, D., Jungerius,
P.D., Spaan, W. 2006. Will we lose the last active inland drift sands of Western Europe?
The origin and development of the inland drift-sand ecotype in the Netherlands.
Landscape Ecology 21, 431-447.
ROB. 1989. Consolidatierapport Uddelermeer.
Roossens, H., Beex, G. 1965. Bronstijdgrafheuvels op de Haarterheide te Hamont 81.
Brussel.
Roymans, N., Fokkens, H. 1991. Een overzicht van veertig jaar nederzettingsonderzoek
in de Lage Landen. . In: H. Fokkens, N. Roymans (eds.). Nederzettingen uit de
bronstijd en de vroege ijzertijd in de Lage Landen. Amersfoort: Rijksdienst voor het
Oudheidkundig Bodemonderzoek, 1-19.
256
ancestral heaths
Sarauw, G.F.L. 1898. Lynheden i oldtiden. Iagttagelser fra gravhöje. Aarbøger for Nordisk
Oldkyndighed og Historie.
Sevink, J., Koster, E.A., van Geel, B., Wallinga, J. in press. Drift sands, lakes and soils:
the multiphase Holocene history of the Laarder Wasmeren area near Hilversum, The
Netherlands.
Siebel, H., Piek, H. 2001. Veranderde inzichten over begrazing bij natuurbeheerders.
Vakblad Natuurbeheer 4, 45-49.
Simmons, I.G., Innes, J.B. 1987. Mid-Holocene adaptations and later Mesolithic forest
disturbance in Northern England. Journal of Archaeological Science 14, 385-403.
Simmons, I.G., Innes, J.B. 1996a. Disturbance phases in the Mid-Holocene vegetation
at North Gill, North York Moors: form and process. Journal of Archaeological Science
23(2), 183-191.
Simmons, I.G., Innes, J.B. 1996b. Prehistoric charcoal in peat profiles at North Gill,
North Yorkshire Moors, England. Journal of Archaeological Science 23(2), 193-197.
Smits, L. 2013. Analysis of the cremated bone from mound 7. In: D.R. Fontijn, S.A.
van der Vaart, R. Jansen (eds.). Transformation through Destruction. A monumental
and extraordinary Early Iron Age Hallstatt C barrow from the ritual landscape of OssZevenbergen. Leiden: Sidestone Press, 257-262.
Soepboer, W., Lotter, A.F. 2009. Estimating past vegetation openness using pollen–
vegetation relationships: A modelling approach. Review of Palaeobotany and Palynology
153, 102-107.
Stortelder, A.H.F., de Smidt, J.T., Swertz, C.A. 1996. Calluno-Ulicetea. In: J.H.J.
Schaminée, A.H.F. Stortelder, E.J. Weeda (eds.). De vegetatie van Nederland. Deel 3.
Plantengemeenschappen van graslanden, zomen en droge heiden. Leiden: Opulus press,
287-316.
Stoutjesdijk, P. 1953. Vegetatiekundig onderzoek van Veluwse heidevelden. Studiekring
voor de Veluwe, heeft onze heide nog toekomst ? De Studiekring voor de Veluwe
Arnhem, 15-32.
Sugita, S. 1993. A model of pollen source area for an entire lake surface. Quaternary
Research 39, 239-244.
Sugita, S. 1994. Pollen Representation of Vegetation in Quaternary Sediments: Theory
and Method in Patchy Vegetation. Journal of Ecology 82(4), 881-897.
Sugita, S. 2007a. Theory of quantitative reconstruction of vegetation I: pollen from large
sites REVEALS regional vegetation composition. The Holocene 17(2), 229-241.
Sugita, S. 2007b. Theory of quantitative reconstruction of vegetation II: all you need is
LOVE. The Holocene 17(2), 243-257.
Sugita, S., Gaillard, M.-J., Broström, A. 1999. Landscape openness and pollen records: a
simulation approach. The Holocene 9(4), 409-421.
Sutton, O.G. 1953. Micrometeorology: a study of physical process in the lowest layers of the
earth’s atmosphere. New York: McGraw-Hill.
Svenning, J.-C. 2002. A review of natural vegetation openness in north-western Europe.
Biological Conservation 140, 133-148.
Theunissen, E.M. 1993. Once again Toterfout-Halve Mijl. An attempt to demonstrate
vertical stratification in the burial evidence of a Bronze Age cemetery. Analecta
Praehistorica Leidensia 26, 29-43.
Tinsley, H.M. 2001. Modern pollen deposition in traps on a transect across an
anthropogenic tree-line on Exmoor, southwest England: a note summarising the first
three years of data. Review of Palaeobotany and Palynology 117(1-3), 153-158.
references
257
Tinsley, H.M., Smith, T.R. 1974. Surface pollen studies across a woodland/heath transition
and their application to the interpretation of pollen diagrams. New Phytologist 73,
547-565.
Trautmann. 1952. Pollenanalytische Untersuchungen über die Fichtenwälder des
Bayerischen Waldes. Planta 41, 83-124.
Troels-Smith, J. 1953. Ertebøllekultur-Bondekultur. Resultaten af de sideste 10 aars
undersøgelser i Aamosen, Vestsjælland. Aarbøger for Nordisk Oldkyndighed og Historie
1953, 5-62.
van der Vaart, S.A., Fontijn, D.R., Valentijn, P. 2013. The central find assemblage of
mound 7. In: D.R. Fontijn, S.A. van der Vaart, R. Jansen (eds.). Transformation
through Destruction. A monumental and extraordinary Early Iron Age Hallstatt C barrow
from the ritual landscape of Oss-Zevenbergen. Leiden: Sidestone Press, 119-140.
Valentijn, P. 2013. “Mound” 6: a post and ditch aligned long barrow. In: D.R. Fontijn,
S.A. van der Vaart, R. Jansen (eds.). Transformation through Destruction. A monumental
and extraordinary Early Iron Age Hallstatt C barrow from the ritual landscape of OssZevenbergen. Leiden: Sidestone Press, 47-69.
Valentijn, P., Fontijn, D.R. 2011. Excavating the surroundings of the barrows. In:
D.R. Fontijn, Q.P.J. Bourgeois, A. Louwen (eds.). Iron Age Echoes. Prehistoric land
management and the creation of a funerary landscape - the “twin barrows” at the Echoput
Apeldoorn. Leiden: Leiden Press, 91-110.
Vera, F.W.M. 1997. Metaforen voor de wildernis: Eik, hazelaar, rund en paard. PhD thesis.
Landbouwuniversiteit Wageningen.
Verbeek, P.J.M., Graaf, d.M., Scherpenisse, M.C. 2006. Verkennende studie naar de
effecten van drukbegrazing met schapen in droge heide. Effectgerichte maatregel tegen
vermesting in droge heide. Rapport DK, nr. 2006. Ede: Directie Kennis, Ministerie van
Landbouw, Natuur en Voedselkwaliteit.
Verwers, G.J. 1966. Tumuli at the Zevenbergen near Oss, Gem. Berghem, Prov. NoordBrabant. Analecta Praehistorica Leidensia 2, 27-32.
Verwers, G.J. 1972. Das Kamps Veld in Haps in Neolithikum, Bronzezeit und Eisenzeit.
Analecta Praehistorica Leidensia V.
Waterbolk, H.T. 1954. De praehistorische mens en zijn milieu. Een palynogisch onderzoek
naar de menselijke invloed op de plantengroei van de diluviale gronden in Nederland.
University of Groningen.
Waterbolk, H.T. 1957. Pollenanalytisch onderzoek van twee Noordbrabantse tumuli.
In: G. Beex (ed.). Twee grafheuvels in Noord-Brabant (Bijdragen tot de studie van het
Brabantse heem deel IX). Eindhoven, 34-39.
Waterbolk, H.T. 1964. The Bronze Age settlement of Elp. Helinium 4, 97-131.
Webb, T., III, Howe, S.E., Bradshaw, R.H.W., Heide, K.M. 1981. Estimating plant
abundances from pollen percentages: the use of regression analysis Review of
Palaeobotany and Palynology 34, 269-3600.
Weeda, E.J., Westra, R., Westra, C., Westra, T. 1985. Nederlandse Oecologische Flora.
Wilde planten en hun relaties 1. Haarlem.
Weeda, E.J., Westra, R., Westra, C., Westra, T. (eds.) 1988. Nederlandse Oecologische
Flora. Wilde planten en hun relaties Haarlem.
Wentink, K. in prep. The biographies of grave goods and the identity of the dead (working
title), Sidestone Press, Leiden. (PhD thesis).
Whitehouse, N.J., Smith, D. 2010. How fragmented was the British Holocene wildwood?
Perspectives on the “Vera” grazing debate from the fossil beetle record. Quarternary
Science Reviews 29(3-4), 539-553.
258
ancestral heaths
van Wijk, I.M., Fokkens, H., Fontijn, D.R., de Leeuwe, R., Meurkens, L., van Hilst, A.,
Vermeeren, C. 2009. Resultaten van het definitieve onderzoek. In: H. Fokkens, R.
Jansen, I.M. van Wijk (eds.). Oss-Zevenbergen: de langetermijn-geschiedenis van een
prehistorisch grafveld. Leiden: Archol rapport 50, 69-140.
van Wijk, I.M., Jansen, R. 2005a. Archeologisch onderzoek Uden-Slabroekse Heide 2005.
In: I.M. van Wijk, R. Jansen (eds.). Het urnenveld Slabroekse Heide op de Maashorst.
Een verkennend en waarderend archeologisch proefsleuvenonderzoek. Leiden: Archol
rapport nummer 72, 43-56.
van Wijk, I.M., Jansen, R. (eds.) 2005b. Het urnenveld Slabroekse Heide op de Maashorst.
Een verkennend en waarderend archeologisch proefsleuvenonderzoek. Leiden: Archol
rapport nummer 72.
Willemse, N.W., Groenewoudt, B.J. 2012. Resilience of Meta-Stable Landscapes? The
NonLinear Response of Late Glacial Aeolian Landforms to Prehistoric Reclamation along Dutch
River Valleys. eTopoi, Journal for Ancient Studies Special Volume 3, 1007-1037.
Wimmers, W.H., Groenman- van Waateringe, W., Spek, T.H. 1993. Het culturele erfgoed
van een natuurgebied. Historisch Geografisch Tijdschrift 11e jaargang(2), 53-74.
Yallop, A.R., Thacker, J.I., Thomas, G., Stephens, M., Clutterbuck, B., Brewer, T.,
Sannier, C.A.D. 2006. The extent and intensity of management burning in the
English uplands. Journal of Applied Ecology 43(6), 1138-1148.
van Zeist, W. 1955. Pollenanalytical investigations in the northern Netherlands. Acta
Botanica Neerlandica 25, 1-81.
van Zeist, W. 1959. Studies on the post-Boreal vegetational history if south-eastern
Drenthe (Netherlands). Acta Botanica Neerlandica 8, 156-185.
van Zeist, W. 1963. Het stuifmeelonderzoek van de grafheuvel te Mol. Archaeologica
Belgica 72, 20-22.
van Zeist, W. 1967a. Archaeology and palynology in the Netherlands Review of Palaeobotany
and Palynology 4, 45-65.
van Zeist, W. 1967b. Palynologisch onderzoek van een ringwalheuvel bij Eersel. Berichten
van de rijksdienst voor het oudheidkundig bodemonderzoek 17, 53-58.
van Zeist, W. 1976. To early rye finds from the Netherlands. Acta Botanica Neerlandica
25, 71-79.
references
259
188431/472100
188440/472080
Echoput 1
Echoput 2
Echoput
180441/473390
Uddelermeer 2
appendix 1
Vierhouten
Ugchelen
180411/473379
Uddelermeer 1
Uddelermeer
185941/483081 ?
191550/465163 ?
Ugchelen 4
Vierhouten
191550/465163 ?
Ugchelen 1
178759/466801
Stroe
Stroe
170800/476200
191226/479004
Niersen 6
Putten
191092/478801
192251/477294
Vaassen 3
Niersen 4
192262/477262
Vaassen 2
Putten
Niersen
192215/477315
Vaassen 1
174949/478453
Ermelo I
Epe
174832/478453
Ermelo III
Ermelo
191868/482135
Emst
Emst, Langeweg
174040/469844 ?
Boeschoten
Boeschoten
Coordinates
Barrow name
Northern and Central Veluwe
-
-
-
-
-
-
ID 409
ID 637
ID 635
ID 275
ID 274
ID 273
ID 324
ID 326
ID 631
-
-
-
Barrow ID
Bourgeois (2013)
S26
S21
S12
S3
S24
S2
S15
S9
S30
Synonym Casparie and
Groenman-van Waateringe
1980
Doorenbosch (this volume)
mound 2 (Fontijn et al. 2011)
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
tumulus II (Lanting and van der Waals
1971d)
tumulus III (Lanting and van der Waals
1971d)
Waterbolk 1954
Doorenbosch (this volume)
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
Vierhouten (Lanting and van der Waals
1972c)
Waterbolk 1954
Doorenbosch (this volume)
Doorenbosch (this volume)
barrow 4 (Waterbolk 1954, 95)
barrow 1 (Waterbolk 1954, 95)
barrow 224 (ROB 1989)
barrow 223 (ROB 1989)
Casparie and Groenman-van
Stroe (Butler and van der Waals 1967, 124) Waateringe 1980
Putten (Waterbolk 1954)
Galgenberg 6 (Holwerda 1908)
Doorenbosch (this volume)
Casparie and Groenman-van
Waateringe 1980
tumulus I (Lanting and van der Waals
1971d)
Galgenberg 4 (Holwerda 1908)
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
barrow I (Modderman 1954)
barrow III (Modderman 1954)
Casparie and Groenman-van
Waateringe 1980
Doorenbosch (this volume)
mound 1 (Fontijn et al. 2011)
Emst (Hulst 1972)
Waterbolk 1954
Original pollen data
tumulus Huneweg (Waterbolk 1954)
Original synonyms
Appendix 1
261
262
ancestral heaths
Wolfheze
186808/446941
Warnsborn 6
183491/446004
186891/447033 ?
Warnsborn 5
Wolfheze
186908/447054 ?
179340/443150
Renkum 5
Warnsborn 4
178975/444702
Renkum 4
187000/447080 ?
178980/444070
Renkum 3
Warnsborn 3
178933/445000
Renkum 2
187116/447009 ?
178925/444982
Renkum 1
Warnsborn 2
179234/448629
Ede 2
187220/446888 ?
179301/449114
Ede 1
Warnsborn 1
178750/446501
Bennekom Oostereng
Warnsborn, Arnhem
178638/446023
Bennekom 1
Renkum
181748/443820
Coordinates
Doorwerth
Barrow name
Doorwerth
Renkum
-
-
-
-
-
-
-
ID 4524
ID 4002
ID 4501
ID 4107
ID 4106
ID 4010
ID 4103
ID 427
ID 322
ID 400
Barrow ID
Bourgeois (2013)
S31
S1
S28
S8
S7
S6
S4
S20
S27
S25
S16
Synonym Casparie and
Groenman-van Waateringe
1980
Waterbolk 1954, Casparie and
Groenman-van Waateringe 1980
Waterbolk 1954
Waterbolk 1954
Waterbolk 1954
Waterbolk 1954
Schaarsbergen, barrow 2 (Waterbolk 1954,
95-98)
Schaarsbergen, barrow 3 (Waterbolk 1954,
95-98)
Schaarsbergen, barrow 4 (Waterbolk 1954,
95-98)
Schaarsbergen, barrow 5 (Waterbolk 1954,
95-98)
Schaarsbergen, barrow 6 (Waterbolk 1954,
95-98)
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
Schaarsbergen, barrow 1 (Waterbolk 1954,
95-98)
Wolfheze (Hulst 1971)
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
Ketsberg (van Giffen 1958)
Kwadenoord
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
Kwadenoord, heuvel S (Lanting and van
der Waals 191972b)
Keyenberg (Modderman 1964)
Casparie and Groenman-van
Waateringe 1980
Casparie and Groenman-van
Waateringe 1980
Heuvel Amber, Ede 5 (Lanting and van der
Waals 1971a)
Kwadenoord, heuvel Q (Lanting and van
der Waals 1972b)
Casparie and Groenman-van
Waateringe 1980
Bennekom Oostereng, heuvel 12 (Bursch
1933)
Casparie and Groenman-van
Waateringe 1980
Van Giffen 1954
Bennekom Kwade Oord, tumulus I (van
Giffen 1954)
Girhen (Lanting and van der Waals 1976)
Casparie and Groenman-van
Waateringe 1980
Original pollen data
Doorwerth (Hulst et al. 1973)
Other synonyms
appendix 1
263
Roosterbos
Laren
146560/468680
143400/472200
Laren 3
Roosterbos
143140/472100
Laren 2
141130/471130
Hilversum 3
143080/471890
141130/471130
Hilversum 2
Laren 1
141400/472800
144231/466122
Baarn 3
Hilversum 1
144270/466080
Baarn 2
Hilversum (Erfgooierstraat)
144240/466040
Baarn 1
Baarn (Groot Drakenstein/
Lage Vuursche)
Coordinates
Barrow name
Gooi
ID 413
ID 380
ID 388
ID 384
ID 296
ID 295
ID 291
-
ID 415
ID 414
Barrow ID
Bourgeois (2013)
S11
S19
S17
S10
S33
S32
S18
S29
S14
S13
Synonym Casparie and
Groenman-van Waateringe
1980
Casparie and Groenman-van Waateringe 1980
Original pollen data
Casparie and Groenman-van Waateringe 1980
Casparie and Groenman-van Waateringe 1980
Casparie and Groenman-van Waateringe 1980
t Bluk, heuvel 2 (Remouchamps 1928); De
Zeven Bergjes, heuvel 2 (Modderman 1954)
Roosterbos, Grabhugel II (van Giffen 1930)
Casparie and Groenman-van Waateringe 1980
t Bluk, heuvel 10 (Remouchamps 1928); De
Zeven Bergjes, heuvel 10 (Modderman 1954)
Casparie and Groenman-van Waateringe 1980
Casparie and Groenman-van Waateringe 1980
Casparie and Groenman-van Waateringe 1980
Casparie and Groenman-van Waateringe 1980
t Bluk, heuvel 6 (Remouchamps 1928)
Erfgooiersstraat, heuvel 6 (Bursch 1935)
Erfgooiersstraat, heuvel 5 (Bursch 1935)
Erfgooiersstraat, heuvel 1 (Bursch 1935)
Groot-Drakenstein, tumulus V (van Giffen 1930)
Groot-Drakenstein, tumulus III (van Giffen 1930) Casparie and Groenman-van Waateringe 1980
Groot-Drakenstein, tumulus I (van Giffen 1930)
Other synonyms
264
ancestral heaths
146220/379600
146500/379570
Alphen 2
Bergeijk
Eersel
Goirle
Hoogeloon 1
Hoogeloon 2
THM eo: Alphen, de Kwaalberg
THM eo: Bergeijk
THM eo: Eersel
THM eo: Goirle
THM eo: Hoogeloon
152242/380911
151903/380957
151912/381044
151860/380937
151511/381357
151290/380984
151242/380985
151176/380986
151165/380950
Steensel
THM 1
THM 1B
THM 2
THM 3
THM 4
THM 5
THM 6
THM 7
THM 8
THM: Toterfout Halve Mijl
150738/378384
Knegsel 2
THM eo: Steensel
150715/378274
Knegsel 1
THM eo: Knegsel Urnenweg
153275/377835
Knegsel, Moormanlaan
THM eo: Knegsel Moormanlaan
129270/389430
150620/376030
146650/367250
123935/387020
127220/389900
Alphen 1
THM eo: Alphen, de Kiek
Coordinates
Barrow name
Toterfout-Halve Mijl
ID 16
ID 15
ID 14
ID 13
ID 646
ID 12
ID 11
ID 10
ID 645
ID 79
ID 78
ID 113
ID 138
ID 137
ID96
ID 133
ID 403
ID 91
ID 92
S35
S34
Synonym Casparie and
Barrow ID Bourgeois
Groenman-van Waateringe
(2013)
(1980)
Waterbolk 1954
Waterbolk 1954
Waterbolk 1954
tumulus E (Waterbolk 1954,
104-108)
tumulus F (Waterbolk 1954,
104-108)
Steensel (Waterbolk 1954, 103,
109-110)
barrow 8 (Glasbergen 1954)
barrow 7 (Glasbergen 1954)
barrow 6 (Glasbergen 1954)
barrow 5 (Glasbergen 1954)
barrow 4 (Glasbergen 1954)
barrow 3 (Glasbergen 1954)
barrow 2 (Glasbergen 1954)
barrow 1b (Glasbergen 1954)
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Modderman and Bakels 1971
barrow 1 (Glasbergen 1954)
Beex 1954
Moormanlaan (Modderman and
Bakels 1971)
Waterbolk 1954
Smousenberg (Beex 1954)
Zwartenberg (Waterbolk 1954)
Waterbolk 1954
Van Zeist 1967b
De Gloeiende Engelsman (Beex
1964b)
barrow 1 (Glasbergen 1954)
Beex 1957, Waterbolk 1957
Bergeijk Witrijt (Beex 1957)
Casparie and Groenman-van Waateringe 1980
Casparie and Groenman-van Waateringe 1980
Alphen Op de Kiek (Modderman
1955)
Alphen de Kwaalberg (Beex 1964c)
Original pollen data
Other synonyms
appendix 1
265
Toterfout-Halve Mijl
151067/381092
151074/380864
151087/380835
150860/380713
150825/380710
150793/380711
150773/380706
150750/380759
150723/380751
150714/380771
150643/380738
150610/380731
150640/380753
150601/380713
150582/380725
150572/380712
150553/380734
150562/380704
THM 9
THM 10
THM 11
THM 13
THM 14
THM 15
THM 16
THM 17
THM 19
THM 20
THM 21
THM 22
THM 22A
THM 23
THM 24
THM 25
THM 26
THM 28
150546/380722
151176/381050
THM 8A
THM 29
Coordinates
Barrow name
ID 39
ID 38
ID 36
ID 35
ID 34
ID 33
ID 32
ID 31
ID 30
ID 29
ID 28
ID 26
ID 25
ID 24
ID 23
ID 22
ID 20
ID 19
ID 18
ID 17
Synonym Casparie and
Barrow ID Bourgeois
Groenman-van Waateringe
(2013)
(1980)
barrow 29 (Glasbergen 1954)
barrow 28 (Glasbergen 1954)
barrow 26 (Glasbergen 1954)
barrow 25 (Glasbergen 1954)
barrow 24 (Glasbergen 1954)
barrow 23 (Glasbergen 1954)
barrow 22A (Glasbergen 1954)
barrow 22 (Glasbergen 1954)
barrow 21 (Glasbergen 1954)
barrow 20 (Glasbergen 1954)
barrow 19 (Glasbergen 1954)
barrow 17 (Glasbergen 1954)
barrow 16 (Glasbergen 1954)
barrow 15 (Glasbergen 1954)
barrow 14 (Glasbergen 1954)
barrow 13 (Glasbergen 1954)
barrow 11 (Glasbergen 1954)
barrow 10 (Glasbergen 1954)
barrow 9 (Glasbergen 1954)
barrow 8A (Glasbergen 1954)
Other synonyms
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Glasbergen 1954
Original pollen data
266
ancestral heaths
Total number of barrows
in thesis
167914.44/416042.36
167930.87/416129.25
167867.76/416087.00
168006.61/416082.51
168034.35/416097.91
167972.48/416073.16
167968.60/416085.79
167922.36/416087.00
167902.44/416072.72
167913.70/416085.77
167953.77/416063.81
Oss-Zevenbergen 2
Oss-Zevenbergen 3
Oss-Zevenbergen 4
Oss-Zevenbergen 6
Oss-Zevenbergen 7
Oss-Zevenbergen 8
Oss-Zevenbergen 9
Oss-Zevenbergen 10
Oss-Zevenbergen 11
Oss-Zevenbergen 12
Mound 5
97
168105.58/416117.24
167070/415780
Oss-Zevenbergen 1
Chieftain’s grave BA-barrow
Oss Chieftain’s grave
Oss Vorstengraf
Oss Zevenbergen
Vorssel
-
Slabroek 40
Vorssel
-
Slabroek 39
-
-
-
-
-
-
-
-
-
-
ID 3
Slabroek
173060/416560
Schaijk
Barrow ID
Bourgeois (2013)
Schaijk
Coordinates
Barrow name
Oss-Zevenbergen
Synonym Casparie
and Groenman-van
Waateringe (1980)
De Kort 1999
De Kort 2009
De Kort 2009
De Kort 2009
De Kort 2009
Bakels and Achterkamp 2013
Achterkamp 2009, Bakels and
Achterkamp 2013
De Kort 2009
De Kort 2009
De Kort 2009
De Kort 2009
heuvel 1 (Fokkens et al. 2009b)
tumulus III (Verwers 1966), heuvel 2
(Fokkens et al. 2009b)
tumulus IV (Verwers 1966), heuvel 3
(Fokkens et al. 2009b)
heuvel 4 (Fokkens et al. 2009b)
tumulus V (Verwers 1966), heuvel 6
(Fokkens et al. 2009b)
tumulus VI (Verwers 1966), heuvel 7
(Fokkens et al. 2009b)
tumulus I (Verwers 1966), heuvel 8
(Fokkens et al. 2009b)
heuvel 9 (Fokkens et al. 2009b)
heuvel 10 (Fokkens et al. 2009b)
heuvel 11 (Fokkens et al. 2009b)
heuvel 12 (Fokkens et al. 2009b)
tumulus V (Verwers 1966), heuvel 5
(Fokkens et al. 2009b)
De Kort 1999
De Kort 2005
De Kort and van Mourik 2005
De Kort and van Mourik 2005
Van Giffen 1949, Waterbolk 1954
Original pollen data
Bronstijdgraf (Fokkens and Jansen 2005)
Vorstengraf (Bursch 1937)
Vorssel (de Kort 2005)
heuvel 40 (van Wijk 2005b)
heuvel 39 (van Wijk 2005b)
tumulus 3 (van Giffen 1949)
Other synonyms
Appendix 2
Trees and shrubs
Scientific
English
Dutch
Alnus
Alder
Els
Betula
Birch
Berk
Carpinus
Hornbeam
Haagbeuk
Corylus
Hazel
Hazelaar
Fagus
Beech
Beuk
Fraxinus
Ash
Es
Hedera helix
Ivy
Klimop
Pinus
Pine
Den
Populus
Popular
Populier
Quercus
Oak
Eik
Rubus
Bramble
Braam
Salix
Willow
Wilg
Tilia
Lime
Linde
Ulex
Gorse
Gaspeldoorn
Ulmus
Elm
Iep
Castanea
Chestnut
Kastanje
Sambucus nigra
Elder
Vlier
Herbs and alga
Scientific
English
Dutch
Angelica archangelica
Garden angelica
Grote engelwortel
Anthriscus sylvestris
Cow parsley
Fluitenkruid
Apiaceae
Umbellifer family
Schermbloemenfamilie
Artemisia
Mugwort
Alsem
Asteraceae
Composite family
Composietenfamilie
Asteraceae liguliflorae
Composite family (liguliflorae
refers to morphology)
Lintbloemige composieten
Asteraceae tubuliflorae
Composite family (tubuliflorae
refers to morphology)
Buisbloemige composieten
Botryococcus
Green microalga
Groene algensoort
Brassicaceae
Crucifer family
Kruisbloemenfamilie
Calluna vulgaris
Common heather
Struikhei
Cannabis sativa
Hemp
Hennep
Caryophyllaceae
Carnation family
Anjerfamilie
Cerealia
Cereals
Granen
Chenopodiaceae
Goosefoot family
Ganzenvoetfamilie
Chrysosplenium
Golden saxifrage
Goudveil
Cyperaceae
Sedges
Cypergrassenfamilie
Debarya glyptosperma
Green alga
Groene algensoort
Empetrum nigrum
Black crowberry
Kraaiheide
Ericaceae
Heather
Heidefamilie
Galium
Bedstraw
Walstro
Huperzia selago
Fir club moss
Dennenwolfsklauw
Jasione montana
Sheepsbit
Zandblauwtje
appendix 2
267
Scientific
English
Dutch
Liliaceae
Lily family
Leliefamilie
Monolete psilate fern spores
Monolete psilate fern spores
Monolete psilate varensporen
Monolete verrucate fern spores
Monolete verrucate fern spores
Monolete verrucate varensporen
Narthecium ossifragum
Bog asphodel
Beenbreek
Plantago lanceolata
Plantain
Smalle weegbree
Poaceae
Grasses
Grassenfamilie
Polypodium vulgare
Common polypody
Gewone eikvaren
Pteridium
Bracken
Adelaarsvaren
Rosaceae
Rose family
Rozenfamilie
Rubiaceae
Cleaver family
Sterbladigenfamilie
Rumex
Dock
Zuring
Secale
Rye
Rogge
Solanum dulcamara
Bittersweet
Bitterzoet
Sparganium
Bur-reed
Egelskop
Spergula arvensis
Corn spurrey
Gewone spurrie
Sphagnum
Peat moss
Veenmos
Stratiotes aloides
Water soldier
Krabbenscheer
Succisa
Devil’s bit
Blauwe knoop
Triglochin
Arrowgrass
Zoutgras
Trilete fern spores
Trilete fern spores
Trilete varensporen
Zygnemataceae
Green algae family
Groene algenfamilie
Scientific
English
Dutch
Astragalus-type
Milkvetch-type
Hokjespeul-type
Cerastium-type
Mouse-ear chickweed-type
Hoornbloem-type
Cuscuta europaea-type
Greater dodder-type
Groot warkruid-type
Digitalis/Scrophularia-type
Foxglove/figwort-type
Vingerhoedskruid/helmkruid-type
Filipendula-type
Meadowsweet-type
Spirea-type
Galium-type
Bedstraw-type
Walstro-type
Hypericum perforatum-type
St John’s wort
Sint-janskruid
Jasione montana-type
Sheepsbit-type
Zandblauwtje-type
Mentha-type
Mint-type
Munt-type
Papaver rhoeas-type
Poppy-type
Gewone klaproos-type
Prunella-type
Self-heal-type
Brunel-type
Ranunculus acris-type
Buttercup-type
Scherpe boterbloem-type
Saxifraga granulata-type
Meadow saxifrage-type
Knolsteenbreek-type
Spergularia-type
Sea-spurry-type
Schijnspurrie-type
Trifolium-type
Clover-type
Klaver-type
Vaccinium-type
Bilberry-type
Bosbes-type
Veronica-type
Speedwell-type
Ereprijs-type
Pollen types
Appendix II. Scientific names of all the taxa that have been identified in the
palynological analyses of this thesis. Taxa have been divided into three groups:
trees and shrubs, herbs and alga and so-called pollen-types. The pollen-types refer
to morphologically similar pollen-types and do not necessarily represent the taxa
the types are named after. According to Beug (2004, 33):
“Three or more taxa are possible alternatives, but further distinction is not possible
on the basis of pollen or spore morphology alone.”
268
ancestral heaths
Samenvatting
Dit proefschrift gaat over de geschiedenis van prehistorische grafheuvellandschappen in Midden- en Zuid-Nederland, gereconstrueerd door middel van
palynologisch onderzoek (onderzoek met behulp van pollenanalyses). Het
proefschrift bestaat uit drie delen. In deel 1 wordt de achtergrond van het
onderzoek behandeld (hoofdstuk 1). Vervolgens wordt een overzicht gegeven van
hoe palynologisch onderzoek van grafheuvels zich ontwikkeld heeft (hoofdstuk
2) en tenslotte worden de onderzoeksvragen die de basis vormen van dit
promotieonderzoek uiteengezet (hoofdstuk 3).
In deel 2 wordt de methodologie die gebruikt is om de onderzoeksvragen
te kunnen beantwoorden besproken en bediscussieerd. Hoofdstuk 4 geeft een
overzicht van de technieken die gebruikt zijn om grafheuvels te bemonsteren. In
hoofdstuk 5 komt de discussie over hoe pollendiagrammen gebaseerd op pollen
uit minerale bodems gebruikt kunnen worden om een vegetatiegeschiedenis te
reconstrueren aan de orde. In hoofdstuk 6 wordt de zogenaamde pollensom die
gebruikt wordt in grafheuvel-pollenonderzoek besproken en opnieuw vastgesteld.
Hoofdstuk 7 gaat over de vraag hoe je de grootte van een open plek waar een
grafheuvel in gebouwd werd kunt bepalen.
In het laatste deel, deel 3, komt het palynologisch onderzoek naar grafheuvels
in vijf gebieden aan de orde (hoofdstuk 8-12). In hoofdstuk 13 en 14 worden de
resultaten van deze deelonderzoeken samengevoegd en bediscussieerd om zo tot
een reconstructie van de geschiedenis van het grafheuvellandschap te komen.
Hieronder volgt een samenvatting per hoofdstuk.
Deel 1
H1: Er zijn in Europa honderdduizenden grafheuvels bekend, waarvan er zo’n
3000 in Nederland liggen. De meeste van deze grafheuvels dateren uit het 3e en
2e millennium voor Christus en in die tijd waren ze zo talrijk dat ze waarschijnlijk
hele ‘grafheuvellandschappen’ vormden. Maar welke rol speelden grafheuvels
eigenlijk in het landschap en hoe zag zo’n grafheuvellandschap er uit? Er zijn in de
vorige eeuw veel reconstructies gemaakt van de vegetatie in de directe omgeving
van een grafheuvel, maar een totale landschapsreconstructie ontbreekt. Om te
kunnen begrijpen welke betekenis grafheuvels hadden in het landschap is het van
belang om niet alleen te kijken naar de locale vegetatiereconstructies, maar om
het landschap waarin de grafheuvels gebouwd werden in een breder perspectief
te bekijken. Ook is het van belang meer te weten te komen over de ontstaans- en
gebruiksgeschiedenis van deze landschappen.
Vragen over deze grafheuvellandschappen komen niet alleen voort uit
wetenschappelijke, maar ook uit maatschappelijke interesse. Staatsbosbeheer,
als beheerder van vele natuurreservaten in Nederland waar grafheuvels te vinden
zijn, is bijvoorbeeld geïnteresseerd in hoe het landschap rond deze heuvels er
oorspronkelijk uitzag. De organisatie wil meer informatie aan het publiek kunnen
geven over de grafheuvels en ze, indien mogelijk, laten zien in hun oorspronkelijke
omgeving. Informatie over het oorspronkelijke landschap waarin heuvels lagen is
voor de organisatie van belang om hun (landschappelijk) beleid hierop aan te
kunnen passen.
H2: Over het algemeen wordt aangenomen dat het grootste deel van Midden- en
Zuid-Nederland (de gebieden waar dit onderzoek zich op gericht heeft) ten tijde
dat de eerste grafheuvels gebouwd werden (tijdens het Subboreaal) nog grotendeels
bedekt was met bos. In de vorige eeuw is al veel onderzoek gedaan naar de directe
samenvatting
269
omgeving van grafheuvels, waaruit is gebleken dat grafheuvels gebouwd werden op
open plekken. Over het ontstaan en gebruik van deze open plekken is nauwelijks
iets bekend. Wellicht was het landschap van nature in deze tijd al veel meer open
dan over het algemeen wordt aangenomen. Daarnaast kunnen bijvoorbeeld storm
of overstromingen de oorzaak zijn van open plekken. Een andere mogelijkheid is
dat open plekken ontstaan zijn door toedoen van de mens. Zo is in het Neolithicum
veel bos verdwenen (bijvoorbeeld gekapt of verbrand) om bijvoorbeeld ruimte
te maken voor landbouwactiviteiten, het bouwen van nederzettingen (huizen,
schuurtjes, hekwerken etc.) of misschien wel om een open plek te creëren om
een grafheuvel in te bouwen. Het is op dit moment niet duidelijk wat voor open
plekken gebruikt werden om grafheuvels in te bouwen en of de oorsprong van
zo’n open plek belangrijk was voor de grafheuvelbouwers. Wellicht had men een
voorkeur voor voorouderlijke gronden, dat wil zeggen gronden die al lange tijd
in gebruik waren geweest door de voorouders van de grafheuvelbouwers. Ook
is het onbekend hoe groot de open plekken waren die uitgekozen werden voor
grafheuvels.
H3: Om antwoord te geven op de vragen die in de voorgaande hoofdstukken naar
voren komen is er een vijftal onderzoeksvragen geformuleerd:
1. Hoe zag een grafheuvellandschap eruit en wat is de ontstaans- en
gebruiksgeschiedenis van zo’n landschap?
2. Werden grafheuvels gebouwd op voorouderlijke gronden?
3. Wat was de grootte van een open plek waar grafheuvels in gebouwd werden
en wat was de afstand van een grafheuvel tot de bosrand?
4. Welke rol speelden grafheuvels in het landschap? Hoe stond de geschiedenis
van een grafheuvellandschap in verband met het natuurlijke en culturele
landschap in de omgeving van grafheuvels?
5. Welk advies is te geven aan Staatsbosbeheer en andere instanties met betrekking
tot het herstellen van oorspronkelijke grafheuvellandschappen voor publieke
doeleinden?
Om deze onderzoeksvragen te beantwoorden is het onderzoek gericht op het
midden en zuiden van Nederland, aangezien daar veel grafheuvels te vinden
zijn. Deze grafheuvels stammen uit de periode van het Laat-Neolithicum tot de
Midden-Bronstijd (2900-1100 BC). Van deze grafheuvels zijn al veel gegevens
beschikbaar van waaruit verder onderzoek gedaan kon worden. Om de vragen
te beantwoorden is vooral gebruik gemaakt van palynologisch onderzoek. Zowel
bestaande als voor dit onderzoek nieuw gegenereerde pollendata zijn gebruikt om
uitgebreide vegetatiereconstructies te maken.
Deel 2
H4: Palynologisch onderzoek, oftewel onderzoek met behulp van pollen
(stuifmeel) analyses, is gebaseerd op het feit dat pollen over het algemeen erg goed
bewaard blijft onder de juiste omstandigheden. Pollenkorrels worden verspreid
en komen uiteindelijk terecht op het bodemoppervlak. Dit pollen zal in de loop
van de tijd verder de bodem inzakken of verdwijnen door corrosie. Omdat er een
evenwicht is tussen het verdwijnen en opnieuw neerregenen van pollen, zullen de
pollenkorrels die in de bovenste laag van de bodem te vinden zijn representatief
zijn voor de planten die in de (nabije en verdere) omgeving staan en deze
pollenkorrels verspreiden. Op het moment dat er een grafheuvel gebouwd wordt,
wordt de toplaag van de bodem waarin zich dit pollen bevindt afgesloten van de
buitenlucht. Er kunnen geen nieuwe pollenkorrels meer bijkomen en pollenkorrels
die al aanwezig zijn zullen niet zo snel meer verdwijnen. Het pollenspectrum dat
270
ancestral heaths
verkregen wordt uit pollen dat onder een grafheuvel ligt (van het zogenaamde oud
oppervlak) is dus representatief voor de vegetatie die in de omgeving stond op het
moment dat de grafheuvel opgeworpen werd. In hoofdstuk 4 wordt de techniek
van de pollenbemonstering van de bodem in en onder grafheuvels, van greppels
rondom grafheuvels en van sporen in de omgeving van grafheuvels beschreven.
H5: een relatief nieuwe methode is gebruikt om meer informatie te krijgen
over de vegetatiegeschiedenis van een open plek. Deze methode houdt in dat
een bodemprofiel verticaal centimeter voor centimeter onder een grafheuvel
bemonsterd en geanalyseerd wordt op pollen. Er wordt vanuit gegaan dat hoe
dieper in de bodem, hoe ouder het vegetatiebeeld is dat een pollenspectrum geeft.
Deze methode en de interpretatie ervan worden uitgebreid bediscussieerd in
hoofdstuk 5.
H6: Het absolute aantal pollen in een pollenmonster kan aanzienlijk variëren.
Om pollenspectra van verschillende monsters met elkaar te kunnen vergelijken
worden de pollentypes uitgedrukt als percentages van een zogenaamde pollensom.
Deze pollensom kan bestaan uit alle pollentypes of uit een selectie daarvan. Welke
pollensom het beste is om te gebruiken is afhankelijk van de onderzoeksvraag en
het onderzoeksgebied. De vraag is nu welke pollensom het meest geschikt is om te
gebruiken bij het reconstrueren van grafheuvellandschappen. De meest gebruikte
pollensom in het grafheuvelonderzoek is de zogenaamde boompollensom minus
Betula (berk). De pollentypes van de kruidenvegetatie en de Betula worden uit de
pollensom gelaten, omdat deze soorten lokaal veel voorkomen en daardoor sterk
kunnen variëren in pollenspectra, zelfs als deze spectra komen van grafheuvels
die dicht bij elkaar liggen of van een en dezelfde grafheuvel. Deze pollensom is
echter maar eenmalig vastgesteld en daarna niet meer gecontroleerd. Daarom is
besloten om nogmaals onderzoek te doen naar de meest geschikte pollensom voor
grafheuvelonderzoek. Dit onderzoek wordt besproken in hoofdstuk 6.
Er zijn twee methoden gebruikt voor dit onderzoek. Als eerste is er een
vergelijking gemaakt tussen een pollenspectrum van een monster uit veen en een
gelijktijdig pollenspectrum uit een greppel die rondom een nabijgelegen grafheuvel
gegraven is. Een pollenspectrum uit veen wordt geacht de regionale vegetatie weer te
geven en door dit spectrum te vergelijken met het greppelspectrum zou vastgesteld
moeten kunnen worden welke pollentypes de lokale grafheuvelvegetatie weergeven
en welke dus uit de pollensom gelaten moeten worden. Ten tweede zijn meerdere
pollenspectra van oude oppervlakten vanonder gelijktijdige en bij elkaar in de
buurt gelegen grafheuvels met elkaar vergeleken. Deze pollenspectra zouden een
(vrijwel) identiek beeld van de vegetatie moeten geven.
Uit het onderzoek is gebleken dat de meest geschikte pollensom voor
grafheuvelonderzoek een boompollensom is, dus een pollensom waaruit alle
kruiden weggelaten zijn. Of Betula al dan niet ook weggelaten moet worden
lijkt te verschillen per site. Om alle grafheuvelpollenspectra met elkaar te
kunnen vergelijken is besloten om voor alle pollenanalyses in dit onderzoek een
boompollensom minus Betula te gebruiken.
H7: In hoofdstuk 7 worden drie typen onderzoek beschreven naar de grootte van
een open plek waar een grafheuvel in gebouwd werd. Bij het eerste type onderzoek
wordt er vanuit gegaan dat de plaggen die gebruikt werden om de grafheuvel te
bouwen in de directe omgeving gestoken werden. Uit onderzoek is gebleken dat
de plaggen gestoken zijn in heidevegetatie. Het aantal plaggen dat nodig is geweest
om een grafheuvel te gebruiken kan dan uitgedrukt worden in de oppervlakte die
minimaal vrij geweest moet zijn van bomen.
samenvatting
271
Voor het tweede type onderzoek zijn er (oppervlakte) pollenmonsters genomen
in huidige heidevelden die zoveel mogelijk lijken op de heidevelden ten tijde van
de grafheuvelbouw. Deze pollenmonsters zijn op verschillende afstanden van
de bosrand genomen om op deze manier de relatie te kunnen bepalen tussen
een boompollenpercentage en de afstand van de monsterlocatie tot de bosrand.
Dit heeft geresulteerd in een zogenaamde ADF (average distance to the forest)
per boompollenpercentage. Op deze manier kan dus bij een bepaald percentage
boompollen in een grafheuvelmonster de gemiddelde afstand vanaf de grafheuvel
tot de bosrand bepaald worden.
Het derde onderzoek heeft zich gericht op simulatiemodellen die vrij recentelijk
ontwikkeld zijn (en nog steeds in ontwikkeling zijn). Met deze simulatiemodellen
kunnen landschappen met een bepaalde vegetatiesamenstelling vertaald worden
in pollenpercentages die daarbij horen. Voor deze modellen zijn verschillende
parameters nodig die kunnen verschillen per regio. Deze parameters zijn nog niet
beschikbaar voor Nederland. De parameters die gebruikt zijn voor dit onderzoek
zijn afkomstig uit eerder onderzoek uit Zuid-Zweden en voor dit onderzoek
getest op een Nederlands landschap met een bekende vegetatiesamenstelling en
bijbehorende pollenpercentages. Hieruit bleek dat de Zuid-Zweedse parameters
toepasbaar zijn in Nederland. Vervolgens zijn van een van de grafheuvellocaties
uit dit onderzoek verschillende landschapsscenario’s gemaakt, met gebruikmaking
van de simulatiemodellen. Het landschapsscenario waaruit pollenpercentages
kwamen die het dichtst lagen bij de werkelijk gevonden pollenpercentages uit de
grafheuvels is gekozen als het meest waarschijnlijke landschapsscenario.
Deel 3
H8-13: In deze hoofdstukken worden vijf verschillende case-studies besproken.
Ruim 100 grafheuvels in vijf verschillende gebieden zijn palynologisch onderzocht
om een antwoord te krijgen op de onderzoeksvragen uit hoofdstuk 3. Een deel van
de pollendata is verkregen uit nieuw onderzoek, gebaseerd op de methoden die
beschreven zijn in hoofdstuk 4. Het grootste deel van de pollendata is afkomstig
uit eerder onderzoek dat verricht is door verschillende andere onderzoekers. Deze
pollendata zijn voor het huidige onderzoek opnieuw geanalyseerd en geïnterpreteerd
met behulp van de methoden en theorieën beschreven in hoofdstukken 5-7.
In hoofdstuk 13 worden de resultaten van alle case-studies samengevoegd en
bekeken in een breder perspectief om een grafheuvellandschap beter te kunnen
definiëren.
Het is gebleken dat grafheuvels op de Pleistocene zandgronden van Middenen Zuid-Nederland gebouwd werden in open plekken die bedekt waren met heide.
Deze open plekken varieerden in grootte. De kleinste open plekken hadden een
ADF (gemiddelde afstand vanaf de grafheuvel tot de bosrand) van 50-100 m,
terwijl de grootste open plekken een ADF hadden van 300-500 m. Het originele
aantal grafheuvels in Nederland was nog vele malen groter dan het aantal dat
tegenwoordig nog bewaard is gebleven. Er vanuit gaande dat alle niet onderzochte
grafheuvels ook in heide opgeworpen zijn, zal het Nederlandse landschap dus vele
open plekken met heide gekend hebben. De pollendata geven aan dat de meeste
grafheuvels in redelijk kleine open plekken lagen, maar dit kan een misleidend
beeld geven. Vele grafheuvels, vooral in het Laat-Neolithicum, werden namelijk
gebouwd in zogenaamde alignments, rijen van grafheuvels, die kilometers lang
konden zijn. Het is zeer aannemelijk dat de heideveldjes waarin deze grafheuvels
gebouwd werden met elkaar verbonden waren, zodat weliswaar redelijk smalle (100200 m breed), maar kilometers lange heidevelden ontstonden. Dit is waarschijnlijk
272
ancestral heaths
het geval geweest in Renkum (hoofdstuk 8), Niersen-Vaassen (hoofdstuk 8),
Toterfout-Halve Mijl (hoofdstuk 11) en Oss-Zevenbergen (hoofdstuk 12).
De heidevelden werden omgeven door bos, dat ook deel uitmaakte van het
grafheuvellandschap. De bossen in de drogere delen van het landschap werd
over het algemeen gedomineerd door Quercus (eik) met aan de bosranden vooral
Corylus (hazelaar). In de nattere gebieden was voornamelijk elzenbroekbos te
vinden, gedomineerd door Alnus (els).
Heidevelden, waar het grafheuvellandschap voor het grootste gedeelte uit
bestond, hebben een bijzondere eigenschap, namelijk dat ze onderhouden moeten
worden om te kunnen blijven bestaan. Als heide niet onderhouden wordt zullen
andere plantensoorten de heide verdringen. Heidemanagement kan gedaan worden
door middel van begrazen (of maaien), afplaggen en/of afbranden. Het afplaggen
op grote schaal is in dit onderzoek niet aangetoond, maar aangezien plaggen
gebruikt werden om grafheuvels te bouwen zal dit zeker hebben plaatsgevonden.
Ook zijn er geen aanwijzingen dat er op grote schaal heide afgebrand is. Uit dit
onderzoek is gebleken dat de grafheuvelheidevelden waarschijnlijk voornamelijk
begraasd werden door vee: koeien en schapen. Om een heideveld te onderhouden
is 1 schaap per hectare nodig en/of 1 rund per 5-6 hectare. Een gemiddelde ADF
van 100 m per grafheuvel staat gelijk aan een heideveldje van 3 ha per grafheuvel.
Om zo’n heideveld te onderhouden zijn dus 3 schapen en/of 0.5 runderen
nodig. Er is een schatting gemaakt dat in de omgeving van Ermelo ongeveer 134
grafheuvels lagen in de Midden- Bronstijd. Deze grafheuvels lagen waarschijnlijk
allemaal in een heideveld, wat neerkomt op een totale oppervlakte aan heide van
ongeveer 420 ha. Om deze heide te onderhouden zijn 420 schapen nodig en/of
70 runderen. Waarschijnlijk bezat een huishouden in de Midden-Bronstijd B een
veestapel van ongeveer 30 dieren, waarvan 2/3 rund en 1/3 schaap. Dit houdt in
dat 3-4 huishoudens een gebied van 420 ha konden onderhouden. Een ADF van
100 m is een voorzichtige schatting. Als uitgegaan wordt van een ADF van 250
m, dan is de oppervlakte aan heidevegetatie 2630 ha geweest. Daarvoor waren 20
huishoudens nodig met elk 20 runderen en 10 schapen. Deze huishoudens zullen
samengewerkt moeten hebben als zogenaamde heidegemeenschappen om de heide
te kunnen onderhouden.
Het is niet te zeggen of het onderhouden van de heidevelden daadwerkelijk het
doel was van de begrazing, het kan ook onderdeel geweest zijn van de dagelijkse
agrarische activiteiten van de prehistorische mensen die in dat gebied woonden.
In elk geval was dan een bijkomend gevolg dat vele heidevelden onderhouden
werden, heidevelden die een zeer belangrijk onderdeel, zo niet het belangrijkste
onderdeel, vormden van het grafheuvellandschap.
Dit onderzoek heeft aangetoond dat de open plekken al langere tijd bestonden
voordat er grafheuvels in gebouwd werden. Het is niet altijd duidelijk waar deze
open plekken voor gebruikt werden, maar in de meeste gevallen lijkt er al langere
tijd sprake geweest te zijn van een begroeiing met heidevegetatie die begraasd
werd. Dit betekent niet alleen dat het landschap waarschijnlijk al behoorlijk open
geweest moet zijn voordat de eerste grafheuvels gebouwd werden, in tegenstelling
tot wat over het algemeen aangenomen wordt (zie hoofdstuk 2), maar ook dat grote
delen van het landschap (namelijk de heide) al intensief onderhouden werden. De
begrazing van heidevelden maakte onderdeel uit van het dagelijkse leven van de
prehistorische mensen, al is niet bekend waar zij dan precies woonden. Wel is
duidelijk dat ze niet in directe omgeving van een grafheuvel woonden, maar het is
aannemelijk dat ze binnen ‘begrazingsafstand’ woonden.
Een van de onderzoeksvragen is of grafheuvels gebouwd werden op
voorouderlijke gronden. Het antwoord hierop is hoogstwaarschijnlijk ‘ja’.
Grafheuvels werden gebouwd in heidevelden die al lange tijd onderdeel
samenvatting
273
uitmaakten van het dagelijkse leven van hun voorouders en de heidevelden
kunnen dus gezien worden als voorouderlijke heidevelden. Tevens is aangetoond
dat grafheuvels een belangrijke rol gespeeld moeten hebben in het landschap.
Het beeld dat we krijgen vanuit de grafheuvel pollenanalyses is natuurlijk niet
representatief voor het totale landschap en laat alleen het deel met grafheuvels
zien. Maar het maakt wel duidelijk dat grafheuvels een speciale plek innamen. Dit
onderzoek heeft aangetoond dat de ligging van grafheuvels niet gebonden is aan
de ligging van akkers en nederzettingen en dat de zichtbaarheid van de grafheuvels
vaak een belangrijke rol speelde. Tegelijkertijd werden grafheuvels geïntegreerd
in het dagelijks leven en maakten ze deel uit van de economische zone (door
middel van begrazing) van de prehistorische mens. Het grafheuvellandschap werd
gedomineerd door heide. Vele generaties heidegemeenschappen werkten samen
om deze heidevelden te onderhouden. Niet alleen vormden heidevelden de laatste
rustplaats voor voorouders, ook waren de heidevelden al lange tijd gebruikt en
onderhouden door deze voorouders. Terwijl de rest van het landschap enorme
veranderingen onderging in de vorm van cultivatie in de periode van het LaatNeolithicum naar de IJzertijd, vormden de heidevelden waar grafheuvels in lagen
een zeer stabiel en structurerend element in het landschap gedurende duizenden
jaren.
H14: In hoofdstuk 14 wordt een synthese gegeven op basis van de voorafgaande
hoofdstukken. De onderzoeksvragen die in hoofdstuk 3 gesteld zijn worden
beantwoord.
1. Hoe zag een grafheuvellandschap eruit en wat is de ontstaans- en gebruiksgeschiedenis
van zo’n landschap?
Het grafheuvellandschap werd gedomineerd door heidevelden die al langere tijd
bestonden voordat er grafheuvels in gebouwd werden. Ze werden omgeven door
loofbos. Deze heidevelden moesten onderhouden worden, wat hoogstwaarschijnlijk
gebeurde door middel van begrazing. In deze grafheuvel-heidevelden lagen een
of meerdere grafheuvels en de heidevelden waren vaak met elkaar verbonden.
Op deze manier vormden ze uitgestrekte (smalle) heidevelden, als corridors in
het landschap. Het grafheuvellandschap was zeer stabiel en werd gedurende vele
generaties in stand gehouden.
2. Werden grafheuvels gebouwd op voorouderlijke gronden?
Gebaseerd op dit onderzoek is het zeer waarschijnlijk dat grafheuvels op
voorouderlijke gronden gebouwd werden. Grafheuvels werden gebouwd in
heidevelden die begraasd werden, niet alleen toen de grafheuvel gebouwd was,
maar ook al lange tijd daarvoor. Deze heidevelden werden dus al gedurende lange
tijd gebruikt door waarschijnlijk de voorouders van de grafheuvelbouwers.
3. Wat was de grootte van een open plek waar grafheuvels in gebouwd werden en wat
was de afstand van een grafheuvel tot de bosrand?
De open plekken waar grafheuvels in gebouwd werden varieerden in grootte van
vrij klein (ADF = 50-100 m) tot behoorlijk groot (ADF = 300-500 m), hoewel
zulke grote open plekken alleen aangetoond zijn bij de allerjongste onderzochte
grafheuvels. Waarschijnlijk werden de meeste grafheuvels gebouwd in open
plekken met een ADF van 50-150 m. Deze smalle open plekken konden echter
wel vele kilometers lang zijn.
274
ancestral heaths
4. Welke rol speelde grafheuvels in het landschap? Hoe kan de geschiedenis van een
grafheuvellandschap gekoppeld worden aan het natuurlijke en culturele landschap in
de omgeving van grafheuvels?
De rol van grafheuvels in het landschap lijkt tweeledig te zijn. Ten eerste
namen grafheuvels een speciale plek in. Ze werden gebouwd in heidevelden die
waarschijnlijk niet direct bij een nederzetting of akkers lagen en waar zichtbaarheid
een belangrijke rol speelde. Ten tweede waren grafheuvels geïntegreerd in het
dagelijkse leven van de mensen. Het grafheuvellandschap was onderdeel van
hun economische zone en werd gebruikt voor begrazing. In de periode van het
Laat-Neolithicum naar de IJzertijd werden grote delen van het landschap steeds
meer gecultiveerd. In deze periode van verandering vormden de heidevelden
waar grafheuvels in lagen een zeer stabiel element in het landschap gedurende
duizenden jaren.
5. Advies aan Staatsbosbeheer en andere instanties met betrekking tot het herstellen
van oorspronkelijke grafheuvellandschappen voor publieke doeleinden.
De gebieden waar grafheuvels in liggen maken tegenwoordig vaak onderdeel uit van
natuurreservaten. De beheerders van deze natuurreservaten willen de grafheuvels
graag zo veel mogelijk in hun oorspronkelijke omgeving aan het publiek tonen.
De reconstructie van het grafheuvellandschap zoals hierboven besproken is geeft
een goede indicatie van hoe de omgeving van een grafheuvel eruit gezien moet
hebben. In elk geval lagen de grafheuvels in heide. De grootte van het heideveld
verschilde van grafheuvel tot grafheuvel. De grootte van het heideveld dat om
een grafheuvel gerealiseerd kan worden is waarschijnlijk meer afhankelijk van de
hedendaagse dan van de vroegere omstandigheden. De huidige omstandigheden
verschillen enorm ten opzichte van de grafheuvelperiode. Verzuring, bemesting
en uitdroging hebben de conditie van de bodem beïnvloed en zullen dus ook van
invloed zijn op het onderhouden van een heideveld en het omliggende bos.
Wat betreft het beheer van het Nederlandse cultureel erfgoed: nu wordt vaak
alleen de grafheuvel zelf als monument beschouwd en in sommige gevallen een
zone van 10 m rondom een grafheuvel. Dit onderzoek heeft aangetoond dat de
heide rondom een grafheuvel onlosmakelijk verbonden was met de grafheuvel en
deze heide strekte zich veel verder uit dan 10 m rondom een grafheuvel. Daarnaast
is in een aantal grafheuvelgroepen aangetoond (Oss-Zevenbergen, hoofdstuk 12.1
en Echoput, hoofdstuk 8.1) dat de omgeving van een grafheuvel van zeer grote
archeologische waarde kan zijn. ZO werden er bijvoorbeeld ceremoniële palenrijen
gebruikt. Het is dan ook belangrijk om het te beschermen gebied rondom een
grafheuvel te vergroten om waardevol Nederlands cultureel erfgoed te behouden.
samenvatting
275
Acknowledgments
I would like to thank all who helped me accomplishing the task of writing this
dissertation. I hope I’m not forgetting anyone.
First of all I would like to thank my colleagues from the Ancestral Mounds
Project, Quentin Bourgeois, Karsten Wentink, Corrie Bakels, David Fontijn en
Annelou van Gijn. As being a freshman in archeology when I started this project
almost 5 years ago they greatly helped me with getting acquainted with this
beautiful profession.
By having several discussions on the subject, helping me to solve practical
problems, structuring my dissertation, etc., they made a great contribution to the
final result of this dissertation. I would like to thank the Netherlands Organisation
for Scientific Research (Nederlandse Organisatie voor Wetenschappelijk Onderzoek
- NWO) for financing the Ancestral Mounds Project.
I am very thankful to all who made it possible for me to find suitable sample
locations for my research and getting permission to get access to these locations:
Andre ten Hoedt, Poul Hulzink, Klaas van der Laan, Kroondomein Het Loo and
Staatsbosbeheer. I would also like to thank the following for joining me (several
times) in the field and/or the laboratory: Eric Dullaert, Goof van Eijk and Wim
Kuijper. I’m also very grateful to all who provided me with many data without
which it would have been impossible to do this research: Prof. W. Groenman, Prof
H.T. Waterbolk, Marjolein Bouman, Bas van Geel and Jan Willem de Kort. Then
there are several people who helped me out with several problems and difficulties
by advising me in the methodology and having very useful discussions with me
on the interpretation of the data: Roy van Beek, Harry Fokkens, Hans Huisman,
Richard Jansen, Cris van der Linde, Jan van Mourik, Anne Birgitte Nielsen, Joanne
Porck, Jan Sevink and Maarten Wispelwey.
Thanks to several colleagues for having all kind of (ir)relevant discussions during
well needed breaks from the work: Stijn van As, Erica van Hees, Arjan Louwen,
Sasja van der Vaart and Alexander Verpoorte. Thanks go to Hylke de Jong for
proof-reading my thesis and Sidestone for editing and publishing my dissertation.
I would like to thank Jelte Rozema and Sjoerd Bohncke for introducing me to
palaeoecology and palynology, while I was still a doctoral Biology student at the
Free University of Amsterdam.
And last but not least, for all their love and support, I would very much like to
thank Zavit, Zev, family and friends.
acknowledgments
277
Curriculum Vitae
Marieke Doorenbosch was born in 1980 in Amsterdam. From 1992 to 1998 she
went to the Gymnasium of the Rijksscholen Gemeenschap Lingecollege in Tiel.
In 1998 she started studying Biology at the Free University of Amsterdam from
which she graduated in 2003. For her doctoral thesis she studied the effect of
UV-B radiation on the vegetation at Svalbard and in addition she studied the
vegetation and climate history of Svalbard through palynological analysis.
In 2003-2004 she worked as an adjunct researcher at the Faculty of Systems
Ecology at the Free University of Amsterdam and continued the research she
started during her doctoral study. As a result she co-authored two articles.
In 2004 she started with the study physical therapy at the Hogeschool Thim
van de Laan in Nieuwegein, from which she graduated in 2007. After deciding
her heart was still at Biology, she applied for a PhD position in the NWOfunded Ancestral Mounds project at the Faculty of Archeology, Leiden University.
Throughout her PhD she wrote a chapter in two books and presented several
lectures at national and international conferences. Currently she’s working as a
researcher on a project basis.
curriculum vitae
279
Barrows, i.e. burial mounds, are amongst the most important of Europe’s prehistoric
monuments. Across the continent, barrows still figure as prominent elements in the
landscape. Many of these mounds have been excavated, revealing much about what
was buried inside these intriguing monuments. Surprisingly, little is known about
the landscape in which the barrows were situated and what role they played in their
environment. Palynological data, carrying important clues on the barrow environment,
are available for hundreds of excavated mounds in the Netherlands. However, while
local vegetation reconstructions from these barrows exist, a reconstruction of the
broader landscape around the barrows has yet to be made. This makes it difficult to
understand their role in the prehistoric cultural landscape.
It is argued in this book that barrows were built on existing heaths, which had been
and continued to be maintained for many generations by so-called heath communities.
These heaths, therefore, can be considered as ‘ancestral heaths’. The barrow landscape
was part of the economic zone of farming communities, while the heath areas were used
as grazing grounds. The ancestral heaths were very stable elements in the landscape and
were kept in existence for thousands of years. In fact, it is argued that these ancestral
heaths were the most important factor in structuring the barrow landscape.
Marieke Doorenbosch studied Biology at the Free University of Amsterdam and specialized in
paleoecology. From 2008-2013 she worked as a PhD student within the NWO-funded project
Ancestral Mounds at the Faculty of Archaeology at Leiden University of which this dissertation is
the result.
Sidestone Press
ISBN: 978-90-8890-192-8
9 789088 901928
Sidestone
ISBN 978-90-8890-192-8
ancestral heaths
reconstructing the barrow landscape in the
central and southern netherlands
Marieke Doorenbosch
ancestral heaths
In this book a detailed vegetation history of the landscape around burial mounds is
presented. Newly obtained and extant data derived from palynological analyses taken
from barrow sites are (re-)analysed. Methods in barrow palynology are discussed and
further developed when necessary. Newly developed techniques are applied in order to
get a better impression of the role barrows played in their environment.
Marieke Doorenbosch
ancestral heaths