Sowing the Seed? : Appendices [1 ed.] 9789048512607, 9789087280772

Citation preview

Sowing the seed? Human impact and plant subsistence in Dutch wetlands during the Late Mesolithic and Early and Middle Neolithic (5500-3400 cal BC) Appendices W. A. Out

Copyright © 2010. Leiden University Press. All rights reserved.

Appendices to ARcHAeoLoGicAL stUdies Leiden UniVeRsitY 18

Archaeological Studies Leiden University is published by Leiden University Press, the Netherlands Series editors: C.C. Bakels and H. Kamermans

Copyright © 2010. Leiden University Press. All rights reserved.

ISBN: 9789087280772 e-ISBN: 9789048512607 NUR: 682 © W.A. Out / Leiden University Press, 2010 All rights reserved. Without limiting the rights under copyright reserved above, no part of this book may be reproduced, stored in or introduced into a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the written permission of both the copyright owner and the author of the book.

Copyright © 2010. Leiden University Press. All rights reserved.

Contents Appendix i. Archaeobotany of Hardinxveld-Giessendam polderweg and de Bruin, the netherlands I.1 Geology of Hardinxveld-Giessendam Polderweg I.2 Archaeology of Hardinxveld-Giessendam Polderweg I.3 Archaeobotany of Hardinxveld-Giessendam Polderweg I.3.1 Pollen analysis I.3.2 Macroremains analysis I.3.2.1 Reconstruction of the vegetation I.3.2.2 Crop plants and disturbance indicators I.3.2.3 Carbonised macroremains of non-cultivated plants I.3.3 Wood and charcoal analysis I.3.4 Other sources I.4 Geology of Hardinxveld-Giessendam De Bruin I.5 Archaeology of Hardinxveld-Giessendam De Bruin I.6 Archaeobotany of Hardinxveld-Giessendam De Bruin I.6.1 Pollen analysis inside the excavation trench I.6.2 Pollen analysis outside the excavation trench I.6.2.1 Introduction I.6.2.2 Materials and methods I.6.2.3 Dates I.6.2.4 Results and discussion I.6.2.5 Human impact I.6.2.6 Correspondence with the pollen analysis inside the trench I.6.3 Macroremains analysis I.6.3.1 Reconstruction of the vegetation I.6.3.2 Crop plants and disturbance indicators I.6.3.3 Carbonised macroremains of non-cultivated plants I.6.4 Wood and charcoal analysis I.6.5 Other sources I.7 Hardinxveld-Giessendam Polderweg and De Bruin: comparison of the archaeobotanical results I.7.1 Reconstruction of the natural vegetation I.7.2 Human impact based on pollen diagrams I.7.3 Crop plants and weeds I.7.4 Carbonised macroremains I.7.5 Wood and charcoal

7 7 7 13 13 15 15 20 21 22 25 25 29 30 30 34 34 34 34 35 40 40 41 41 44 44 45 48 48 48 49 50 50 51

Appendix ii. Archaeobotany of Brandwijk-Kerkhof, the netherlands II.1 Introduction II.2 Materials and methods II.3 Results II.3.1 Macroremains analysis II.3.1.1 Layer 30 (4610-4550 BC) II.3.1.2 Layer 45 (4470-4370 BC) II.3.1.3 Layer 50 (4220-3940 BC) II.3.1.4 Sample box samples layer 50

53 53 56 58 58 58 61 61 63

Copyright © 2010. Leiden University Press. All rights reserved.

II.3.1.5 Lab samples layer 50 II.3.1.6 Layer 60 (3940-3820 BC) II.3.1.7 Crop plants II.3.1.8 Carbonised macroremains of non-cultivated plants II.3.1.9 Arable weeds II.3.2 Wood and charcoal analysis II.4 Discussion II.4.1 Reconstruction of the natural vegetation II.4.2 Human impact II.4.3 Seasonality and site function

67 69 71 72 74 76 81 81 84 85

Appendix iii. Archaeobotany of the Hazendonk, the netherlands III.1 Introduction III.2 Materials and methods III.3 Results III.3.1 Southern section of square 57 (Hazendonk 0, 1, 2a, 2b and 3) III.3.2 Core 3, near unit C (Hazendonk 0 and Hazendonk 1) III.3.3 Core 2, near unit C (Vlaardingen 1a and Vlaardingen 1b) III.3.4 M86, eastern section of square 25, unit B (Vlaardingen 1b) III.3.5 M87, eastern section of square 25, unit B (Vlaardingen 1b) III.3.6 Southern section of square 41, unit C (Hazendonk 1, Hazendonk 2, Hazendonk 3 and Vlaardingen 1b) III.3.7 Comparison of proportions of dryland versus wetland vegetation III.3.8 Macroremains from the excavation III.3.8.1 Crop plants III.3.8.2 Carbonised macroremains of non-cultivated plants III.3.8.3 Arable weeds III.3.9 Wood analysis III.3.10 Moss analysis III.4 Discussion III.4.1 Reconstruction of the natural vegetation III.4.2 Development of the natural vegetation III.4.3 Human impact III.4.3.1 Indications of human impact in the diagrams III.4.3.2 Further interpretation of indications of human impact III.4.4 Plant subsistence III.4.5 Local cultivation III.5 Acknowledgements

99 99 106 111 111 118 125 131 136

Appendix iV. synthesis of archaeobotanical sources of the Hazendonk IV.1 Introduction IV.2 Sources IV.3 Reconstruction of the natural vegetation IV.4 Human impact IV.4.1 Mesolithic occupation IV.4.2 Hazendonk 0

215 215 215 219 221 221 221

140 144 145 148 158 161 164 168 168 168 169 170 170 173 176 177 179

Copyright © 2010. Leiden University Press. All rights reserved.

IV.4.3 Hazendonk 1 IV.4.4 Hazendonk 2 IV.4.5 Hazendonk 3 IV.4.6 Vlaardingen 1a IV.4.7 Vlaardingen 1b IV.5 Summary of human impact IV.6 Distance between human activity and sampling point

222 223 223 224 225 226 227

Appendix V. Archaeobotany of Bergschenhoek, the netherlands V.1 Introduction V.2 Materials and methods V.3 Results V.3.1 Pollen and macroremains analysis V.3.2 Wood and charcoal analysis V.3.2.1 Wood and charcoal from the excavation in 1976 V.3.2.2 Wood from the excavation in 1978 V.3.2.3 Other wood data V.3.3 Moss analysis V.3.4 Mollusc analysis V.4 Discussion V.4.1 Reconstruction of the natural vegetation V.4.2 Human impact V.4.2.1 Human impact on the vegetation and deposition processes V.4.2.2 Plant subsistence V.4.3 Seasonality V.5 Acknowledgements

229 229 231 233 233 242 242 244 245 246 246 250 250 251 251 252 254 254

Appendix Vi. Archaeobotany of the Late neolithic site Vlaardingen, the netherlands VI.1 Introduction VI.2 Materials and methods VI.3 Results VI.3.1 Macroremains VI.3.2 Information from earlier publications VI.4 Discussion and conclusions VI.5 Acknowledgements

255 255 255 255 255 257 259 259

Appendix Vii. References of appendices

261

Copyright © 2010. Leiden University Press. All rights reserved.

Appendix I. Archaeobotany of Hardinxveld-Giessendam Polderweg and De Bruin, the Netherlands

Copyright © 2010. Leiden University Press. All rights reserved.

I.1 GeoloGy of HardInxveld -GIessendam PolderweG The site of Hardinxveld-Giessendam Polderweg is located on a Late Glacial inland dune in the central river area (coordinates 116.104/427.636; see fig. I.1). The top of the dune is at about 4.5 m -NAP. Its surface measured c. 80 x 50 metres (4000 m2) at the start of occupation and became smaller through time. The relative height of the dune decreased from 3 to 2.5 metres during occupation phase 1, and to 1.2 m during phase 2 (the occupation phases are discussed below). The former landscape around the site was reconstructed by means of hand corings in an area of 1700 x 600 metres (see fig. I.2). During phase 1, marshes and open water surrounded the dune, while marshes and carr became increasingly important during the later phases. An open crevasse channel, active between c. 5500 and 5100 BC, was present approximately 500 metres to the south. Its presence resulted in the sedimentation of peaty clay with some local thin sand layers. At the start of occupation, Polderweg was one of four dunes protruding in the vicinity, located in the middle of a large marsh area without channels or large patches of open water. At the end of occupation, most of the other dunes had become submerged, while Polderweg would follow soon (Mol 2001a, 2003). I.2 arcHaeoloGy of HardInxveld -GIessendam PolderweG The discussion of Hardinxveld-Giessendam Polderweg and De Bruin is primarily based on Louwe Kooijmans (2001a, b and 2003). Excavation of a part of the southern slope of the dune at Polderweg took place in 1997-1998. The excavation trench (28 x 16 metres) revealed a sequence of refuse layers formed during several occupation phases, large oval pits interpreted as huts (phases 0 and 1), round pits interpreted as fire pits, and postholes (see fig. I.3). The Holocene deposits (clay, colluvia and refuse layers: fossil anthropogenic horizons) comprised artefacts made of flint, stone, pottery as well as organic remains including human remains. Four occupation phases have been distinguished: phase 0 before 5500 BC, phase 1 around 5500-5300 BC, phase 1/2 around 5100 BC and phase 2 around 5000 BC.1 Phase 0 may only be the initial stage of phase 1. Phase 1 is the major phase and is represented by much more material than the other phases. Figure I.4 shows the extent of the refuse layers of the main phases. Archaeological finds in the corings indicated that the occupation during phases 0 and 1 extended over an area of 110 x 25 metres, while occupation during phases 1/2 and 2 together extended over an area of 80 x 20 metres. The site can be characterised as Late or even Final Mesolithic in view of the absence of domesticates, and moreover represents the initial stage of the neolithisation process in the region in view of the first occurrence of pottery in phase 2. Phase 1 is contemporaneous with the älteste LBK, and the material of this phase contained three arrowheads in LBK style, which indicates some contact with these Neolithic communities and illustrates the start of the availability phase as defined by Zvelebil (1986; see chapter 1). The nearest known LBK settlements at this time were located in Hessen, Germany. Other stone and flint artefacts point to contact with communities within Southern Limburg and the Ardennes. Phase 2 is contemporaneous with the jüngere LBK and ended before the start of the Rössen culture (see fig.1.1). In phase 2, the scarce presence of characteristic pottery indicates that Polderweg was part of the initial phase of the Swifterbant culture. The subsistence of the site Polderweg can be characterised as a broad-spectrum economy, with a special focus on the hunting of furred animals and on fishing (especially for pike). The consistent character of the faunal assemblages from all phases suggests a stable use of resources through time. 1 Recalculation of the 14C dates of Polderweg resulted in the new conclusion that phase 1 dates between 5430±90 and 5350±100 cal BC and phase 2 between 5200±140 and 5069±140 cal BC (Mol and Van Zijverden 2007). See also fig. 2.5.

7

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

coastal barriers

salt marshes

tidal flats

pleistocene coversand

river deposits

inland dunes

fen peat

Hardinxveld - Giessendam

Copyright © 2010. Leiden University Press. All rights reserved.

upland peat bogs

0

50km

Figure I.1 Hardinxveld-Giessendam, the Netherlands, location plotted on a palaeogeographical map (c. 5700 BC; NITG).

8

-7

.0

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

0

200 m

0

open water and marsh

marsh

inland dune

inland dune

excavation trench road

200 m

open water crevasse channel excavation trench road

Figure I.2a Hardinxveld-Giessendam Polderweg, the Pleistocene subsurface (m –NAP) and palaeogeographical reconstructions for phases 0 (left) and 1 (right) (after Mol 2001a).

9

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

-5.0

Copyright © 2010. Leiden University Press. All rights reserved.

0

200 m

0

alder carr

alder carr

inland dune

inland dune

excavation trench

excavation trench

road

road

200 m

Figure I.2b Hardinxveld-Giessendam Polderweg, the Pleistocene subsurface (m –NAP) and palaeogeographical reconstructions for phases 1/2 (left) and 2 (right) (after Mol 2001a).

10

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Crop plants and domestic animals other than dog are absent during all phases. For phase 1, the site is interpreted as a winter base camp. It could concern a main occupation in late winter and additional occupation in early autumn or occupation from September to March. There is not enough evidence for both other phases to make any conclusions about seasonality, but continuity in function seems most plausible. The interpretation of the site as a winter base camp implies that it represents only a part of a residential mobile settlement system. The indications of southern contacts make the southern coversand region the best candidate for occupation during other periods of the year, especially the summer. unfortunately this is hard to prove due to bad preservation of sites on the southern sand grounds (Louwe Kooijmans 2001a, b, 2003).

P

P

TF

C P

G

G

G

P G

S20

C C P

P

C C

P

C

TF

P

TF

G TF

s P

not excavated sand (phase 1) marsh (phase 1) features wood eastern section square 11 C = concentration TF = tree fall P = pit

Copyright © 2010. Leiden University Press. All rights reserved.

G = grave

0

4m

Figure I.3 Hardinxveld-Giessendam Polderweg, features of all phases (after Hamburg and Louwe Kooijmans 2001, adapted by L. Amkreutz).

11

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

a

P

M

0

50 m

0

50 m

0

50 m

b

P

M

c

Copyright © 2010. Leiden University Press. All rights reserved.

P

M peat clay sand P = pollen core M = macroremains column

find layer excavation trench road

Figure I.4 Hardinxveld-Giessendam Polderweg, the Pleistocene subsurface (m –NAP), the extent of the surface of the dune and the refuse layers for phases a) 1, b) 1/2 and c) 2 (after Mol 2001a).

12

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

I.3 arcHaeobotany of HardInxveld -GIessendam PolderweG The archaeobotanical data of Polderweg consist of a dated pollen diagram and information on macroremains, worked and unworked wood, charcoal and fungi published by Bakels and van Beurden (2001), and an unpublished macroremains diagram (De Kort 1998). All source material was collected in the excavation trench, with the exception of the pollen core. I.3.1 Pollen analysIs The location of the pollen core was chosen at the northern side of the dune where the influence of the crevasse channel was small (see fig. I.2 and I.4). The core was located at 4 metres from the attachment point2 of phase 1, at 26 metres from the attachment point of phase 2, and at c. 40 metres from the refuse layer (i.e. the settlement area, at the other, southern site of the dune) during all phases. The sediment of the core consists of sandy peat with a layer of gyttja at 7.55 to 7.66 m -NAP. The pollen analysis is based on an upland pollen sum of at least 300 pollen grains. The pollen diagram (see fig. I.5) provides the development of the vegetation during phases 1, 1/2 and 2 (between 5584 to 5230 BC and 4916 to 4540 BC, see the publication for 14C dates). Periods before and after these occupation phases are not represented in the pollen diagram because such a long sequence of peat was not available. The pollen diagram indicates that the natural vegetation on the dune consisted of mixed deciduous woodland and woodland edge vegetation, with some open patches of herb vegetation. The dry dune was surrounded by alder carr, marsh and open water. For the wetlands surrounding the dune, the lower part of the diagram shows relatively open riparian vegetation, which changed into alder carr approximately around phase 2. It is not possible to discern clear signals of human impact in the diagram (Bakels and van Beurden 2001). The diagram shows a decreasing percentage of dryland trees and an increasing percentage of dryland shrubs that could be related to human impact, but this may also be caused by the submergence of the dune. There are some other changes in the diagram that may be related to human activity as well3, but a strong causal relationship cannot be demonstrated. In addition sudden decreases in dryland trees or peaks in dryland herbs are absent, and the different occupation phases are not separately recognisable. One would expect signals of occupation phase 1 in particular since the distance between the dune and the sample point is very small during this phase (4 metres) and since phase 1 is the longest and most intensive occupation period, although the phase is possibly not reflected completely in the diagram. However, even signals of this phase cannot be discerned accurately. This indicates that any human impact was probably of restricted scale. The weak human impact is rather unexpected in view of the length of occupation at the dune and the evidence of use of the vegetation as demonstrated by the analysis of macroremains, wood and charcoal. A possible explanation is that the evidence of human impact at Polderweg is weak due to the absence of agriculture. There are however several other explanations for the weakness of the anthropogenic signal. Firstly, possible evidence of human impact may be blurred by the changes in the vegetation that were caused by the rising water table, which led to the continuous submerging of trees, and resulted in the development of open patches in the vegetation and growth of secondary vegetation. Secondly, the distance to the dune and the distance to the refuse layer may play a role. The distance between the sample location and the dune (phase 1: 4 m; phase 2: 26 m) is, however, not exceptionally large and is not suspected to explain the weakness 2 Attachment point: location where the peat is attached to the dune, indicative of the height of the ground water level and representing the transition from the dune to the surrounding wetlands. 3 In the first spectrum of the diagram (8.00 m -NAP, phase 1), Urtica-type and Poaceae as well as monoletae psilatae fern spores show small peaks, while Filipendula sp. shows a peak in the top of the diagram. Other possible anthropogenic signals are a decrease in Quercus sp. (7.50 m -NAP, phase 1/2) and Tilia sp. (7.30 m -NAP, phase 1/2), a gradual increase in dryland shrubs and herbs in the middle and upper part of the diagram indicating increased openness of the vegetation (phases 1/2 and 2, indicated by pollen of Corylus sp., Rhamnus cathartica, Viburnum opulus, Cornus sanguinea-type, Sambucus nigratype and Chenopodiaceae, Artemisia sp., Plantago lanceolata and Pteridium sp.).

13

Copyright © 2010. Leiden University Press. All rights reserved.

14 or yl us

800

Figure I.5 part 1. 20 40 60 80 100

20 40

200400600800

Peat

Wood remains

Sandy sediment

Detritus-gyttja

Sa ulu li x s lu pu M yr lu ic s a R ha m A l nu is m s Er a - f r a n ic t yp g ul a a Fi l e s e lip en d H o r ula Ir i de s u Ly ps m-t s e y c f ima uda pe .L c c Ly ys hia oru t h im t s C ru ac hyr a m h s R lt ha s a ia i f l o um li t h r a Sa ex pal car yrs u g s ia if lo S o it t a aqu t r i s ra r i a at - t y l a Sp n ic p ar um s ag us - e Sp g d it g a u ti ro T h ar g niu l c a fo li up al an m e ma a U ic iu m r r ti tr u m e a ca m e rs re u di ct m oi um Va ca -t y l ty Ty eri pe pe ph an a a la tif ol ia

P3: 6480 ± 110

um

P1: 5870 ± 80

pl pl

an

d

20

Upland shrubs, herbs and spore plants 20 40 20 40 60 80

Wetland trees and shrubs

20

lm

rc

us

20

xi n B e us tu c f la . H Po p ed u V i er lus sc a um al bu

us

ue

Fr a

U

Q

Pi c Ti ea lia

m

an sh d ru he bs Pi rb nu s an s d

U

U

D

14 C

sp

or

D a e p te s th ( (c yrs m B Li - N P) th ol AP og ) U pl y an d t re es 660

H

R h V i am b n C ur n us c or u a S a nu m o t ha s p r A r mb s a ulu t i c a t uc ng s C e m us uin he is n e Pl n o i a i g r a - t a - yp a p ty e M nt a o d pe er go ia c P t c u la ea er r ia n e Po i d l i s c e l iu p o A l y p o m a e r l at a nu d q e n s ium uil ni s in um

C

e

pl

an

ts

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Upland trees

680

700

720

P2: 6150 ± 70 740

760

780

20 Wetland herbs

20

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

of the signal (compare with appendix II and III). In contrast, the distance to the refuse layer is relatively large (c. 40 metres). It is very probable that this hampers the distinction of evidence of human impact, since dispersal of the pollen signal related to human occupation at the settlement may have been restricted by the presence of high and dense vegetation between the settlement and the sample point. I.3.2

MacroreMaIns analysIs

I.3.2.1 Reconstruction of the vegetation The main analysis of macroremains concerns 65 samples, of which 51 were sieved on a mesh width of 0.25 mm and 14 on a mesh width of 2 mm. The samples were collected in pits and refuse layers dating to phases 1 and 2 from sediments consisting of sand, peat and clay, and were collected from different parts of the slope of the dune in the excavation trench. During all phases, several deposition processes may have played a role in the formation of the macroremains assemblages, resulting in assemblages indicative of diverse ecological conditions (e.g. in situ deposition, colluviation resulting in transport down the slope, flooding of the slopes of the dune resulting in transport in upwards direction, and transport by people and/or animals).

Open water

ac

ea e Po ac As e t a R era e > an c 4 R un e ae 0 m a c R nun ula tub u um c c e u C e ulu ae li f l o ar x s ra C yo a c e a c e h a p h to r i s B r m y ll sa - t y a e a - p A p s si r i on c ea gro e i c - e up G ac e ac e t y p al a a e M ium e e e Tr nth -typ il a e Tr eta -typ il e e Ly eta , wi c o e, t h p o ps pe di i la r in um t a e e (m ar ke r) Po lle n su m Zo ne

Ecologically indeterminate

Po

M

on

ol

et ae Th ,p e si S p ly p la ta h a te e N g r is up n p N ha um al us ym r tr i Po p l u t s t a hae ea Pe m a t y o p a d S p i a s g et l b a e i t o C r o g r um n t y p yp y e er ra ac ea e

Wetland spore plants

517

343 370 359 363

III

326 314 342 329

Copyright © 2010. Leiden University Press. All rights reserved.

376

II

331 311 403

I

328 20

20 40 60 80

20

100 200 300

307

Analyst: L. van Beurden, 1998

Figure I.5 Hardinxveld-Giessendam Polderweg, pollen diagram based on an upland pollen sum, exaggeration 5 x (after Bakels and Van Beurden 2001), part 2.

15

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

N

S

m - NAP

m - NAP

5.00

5.00

6.00

6.00

7.00

7.00

peat 8.00

8.00

sand

macroremains column

clay

9.00

9.00

wood colluvium

10.00

10.00 0

2.5 m

m -NAP

7.00

clay

8.00

clayey peat

Copyright © 2010. Leiden University Press. All rights reserved.

humic clay

clayey peat 9.00

sandy peat

0

20 cm

sand

Figure I.6 Hardinxveld-Giessendam Polderweg, eastern section of square 11, position of the sample boxes (after Louwe Kooijmans and Mol 2001, fig. 3.2, and De Kort 1998, photograph: W.J. Kuijper).

16

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The assemblage of macroremains mainly consisted of waterlogged remains.4 The macroremains assemblage of phase 1 indicates the presence of eutrophic open marsh and bank vegetation at the location of the excavation trench, combined with open water. During phase 2, the slope of the dune was covered with alder carr, marsh and bank vegetation with some open patches. These results correspond with the reconstruction of the wetland vegetation based on the pollen diagram. The vegetation on the slope was less varied than during phase 1. The increasing water levels resulted in a decrease in dryland vegetation at the lower parts of the dune and in the spread of Alnus vegetation upward over the slopes of the dune. The unpublished macroremains diagram of De Kort (1998) consists of nine samples collected with four sample boxes in the eastern section of square 11, just outside the local colluvium but still relatively close to the dune and the location of human activity (see fig. I.4 and I.6). The analysed section is c. 150 cm long, the sample size of the samples was 200 cm3 (0.2 litres), and the mesh width of the sieves was 0.25 mm. The 14C dates (see table I.1) as well as the stratigraphy indicate that the diagram corresponds with phases 1 and 2.

sample

depth lab code (m -NAP) HG-Pw 11-783 7.83 GrA-9802

age (yrs BP) 6050 ± 50

HG-Pw 11-818

6320 ± 50

8.18

GrA-9798

age phase dated material (yrs cal BC, 2σ) 5210-4790 2 Alnus sp., cones 5470-5210

t.a.q. 1 Cornus sanguinea, fruit stones

HG-Pw 11-864

8.64

GrA-9803

6380 ± 50

5480-5220

1 late Iris pseudacorus, seeds, Schoenoplectus lacustris, fruits

HG-Pw 11-919

9.19

GrA-9799

6540 ± 50

5620-5370

1 early Quercus sp., fruits

t.a.q. = terminus ante quem

Copyright © 2010. Leiden University Press. All rights reserved.

Table I.1 Hardinxveld-Giessendam Polderweg, 14C dates of the macroremains diagram of square 11 (Louwe Kooijmans and Mol 2001, 68).

4 Identifications that are supplementary to the results published in Bakels and Van Beurden (2001) are Torilis sp. (probably representing Torilis japonica), Glechoma hederacea, Persicaria cf. hydropiper (Jongste and Verbruggen 1998), Betula sp. and Stellaria neglecta (unpublished data Faculty of Archaeology, Leiden University).

17

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The results of the macroremains analysis by De Kort are shown in figure I.7. Table I.2 provides the relevant stratigraphy. The lowest two spectra show the presence of disturbed dryland vegetation (Tilia sp., Quercus sp., Chenopodium album, Solanum nigrum and Galeopsis bifida-type) as well as marsh vegetation (Phragmites sp., Typha sp., Alisma plantago-aquatica and Urtica dioica). Subsequently, slowly running water reached the sample location, as indicated by clayey sediments, followed by the development of marsh vegetation dominated by Schoenoplectus lacustris, Typha sp. and Alisma plantago-aquatica. The presence of the disturbance indicators Persicaria lapathifolia, Solanum nigrum and Chenopodium album at 8.45 m -NAP is probably related to the activity of the nearby crevasse channel. At 8.18 m -NAP, the diagram shows a major change in the vegetation: alder carr gradually developed, initially combined with Carex species and Lythrum salicaria. Dryland trees and shrubs reappear in the diagram while the values of water plants are low. These changes can all be explained by a diminished influence of the channel and partial recovery of the dryland vegetation in the extra-local vegetation. Urtica dioica shows maximal values just before and after the activity of the crevasse. The first three spectra are rich in charcoal, implying human activity. The diagram does not, however, show clear indications of anthropogenic influence on the vegetation, with the exception of a carbonised seed of Iris pseudacorus at 8.64 m -NAP (late part of phase 1). Although in situ deposition can be questioned, this carbonised seed indicates the horizon of anthropogenic influence and it is therefore not excluded that the taxa at 8.45 m -NAP indicative of disturbance (Persicaria lapathifolia, Solanum nigrum and Chenopodium album), interpreted as the result of disturbance by the crevasse, result instead from human activity. The absence of other signals of anthropogenic influence on the vegetation is remarkable in view of the sample location in the excavation trench. The absence of more carbonised macroremains corresponds, however, with the small number of taxa found in a carbonised state at the excavation.

Copyright © 2010. Leiden University Press. All rights reserved.

depth (m -NAP)

sediment

sample (m -NAP)

7.64-7.43

peaty clay poor in wood remains

7.57

7.75-7.64

clayey peat poor in wood remains

8.15-7.75

peat containing wood remains

8.21-8.15

clayey peat

8.40-8.21

peaty clay poor in wood remains

8.51-8.40

clay rich in wood remains

8.62-8.51

humic clay poor in wood remains

8.87-8.62

sandy peat rich in wood remains

8.92-8.87

sandy peat poor in wood remains

9.02-8.92

peat rich in wood remains

8.95

9.12-9.02

peaty sand poor in archaeological remains

9.11

9.26-9.12

humic sand poor in wood remains

8.02, 7.83, 7.78 8.18 8.45 8.64

Table I.2 Hardinxveld-Giessendam Polderweg, stratigraphy of the macroremains diagram of square 11 (De Kort 1998).

18

Copyright © 2010. Leiden University Press. All rights reserved.

20 40

50 100150 200

um

a pl an ta

50 100 150

20

50 100150200

Peat

Sand

Clay

Humic sediment

Peaty sediment

Sandy sediment

Clayey sediment

Wood remains st

ar

re

li c ia

C

pa t ar or iu ex m C a r s p. c a n e , n S o x p t r ic ab la seu ar p inu S p num do e ll m a ar d c y te C gan ulc per ic u iu am us C ta m al vi e ara t r r G ha os ec t al pa a um iu m lus pa t r lu is

sa

a 50 100150200

Eu

um

te

ic

20

hr

la

at

D a e p te s t h (y (c rs Li m B th - N P) ol AP og Ti y ) lia Q pl u e at rc yp C u s hy l or s lo R ylu p.,f s ha s ru it m a C o r n u ve s a n R nu s c l l a n d cu ub s a a pu G u s san t h a la le c g r e C c h ae uin t i c a he om s e i u a l Fa id a s llo on he p iu d U ia d m era rti m ca um a j c e a et u s di o oi ca r um

50 100 150

nu

s s um p., c u O lu one en s lu s P h a nt p u ra he lus g a R mit qua um e s t i M ex a c en h us a t h yd t r a a ro l aq la is ua pa tic th a um H

Al

oe Pe h r i n r s gia So ic a tr la r ia ine C num lap r v ia he a n n th Pe o p i g r i f o r s o d um lia A l i c a ium nu r ia a l s sp ma bum . cu lo sa

M

D

14 C

757

Ly t

qu

el

-a

rp

go

20

B e la r u tifo la liu er m C ar e ex c t a sp ., bi ca

Si

m

911

is

6540 ± 50

Al

lip e Ir i nd s ps ula Ly e ul c o ud ma a S c pus c o r ia ho e r us en uro , c op pa ar le eu bo ct ni Ty us s se ph d la a cu sp s tri . s

Fi

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Upland trees, shrubs and herbs Wetland trees and herbs

778 6050 ± 50 783

802

6320 ± 50 818

845

6380 ± 50 864

895

20

Wetland trees and herbs

20 40 60 80 100

Figure I.7 part 1.

19

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Open water

Ecol. indet.

Varia

Va l R er ia an n Pe u n a o r c f Po s i c u l u f i c i te a r i s r n a l H m a ep is yd o m e L e r o c g et i n o n s m h o r -t y N na ar n s pe ym s is p S a p h p. m o . rs lv ae in a us -r R ia n alb an an at a ae S t un an el cu s S c lar lu r ia s S i o p h sp aq le u . ua til G ne lar is a s ia -t y P o l e o p. s p pe . a psi Po s p s a ., bif Ep s p l a i d ilo ., s rge a-t yp bi m u a e C ha m ll rc hir o a su tu l Fl m in -t y B o t re pe n m Fi e r e a i n sh m s B u rem ain s d sc ain al s es To ta ln um be ro fs ee ds

Wetland herbs

206 200 149 228 497

408 382

279 20

20

112

Analyst: J.-W. de Kort, 1998

Figure I.7 Hardinxveld-Giessendam Polderweg, macroremains diagram of the eastern section of square 11, + = few (1-10), ++ = some tens (10-49), +++ = many tens (50-99), ++++ = some hundreds (100-499), +++++ = many hundreds (500-999) (after De Kort 1998), part 2.

Copyright © 2010. Leiden University Press. All rights reserved.

Since both the pollen diagram (Bakels and van Beurden 2001) and the macroremains diagram (De Kort 1998) are dated, the diagrams can be linked to each other. The diagrams are contemporaneous, while the sequence of De Kort ends earlier. The clay in the middle of the macroremains diagram probably corresponds with the gyttja in the pollen diagram, and the increase in Alnus sp. after phase 1 can be found in both diagrams. The difference between the sediments can be explained by the presence of the crevasse channel at the southern side of the dune where the macroremains sequence was collected. Neither diagram shows strong indications of human impact. I.3.2.2 Crop plants and disturbance indicators Cultivated plants were absent in the samples of all phases despite the analysis of 74 macroremains samples and an active search for crop plants. On one hand, the absence of crop plants at Polderweg corresponds with the absence of crop plants at De Bruin (Bakels et al. 2001) and the early phases of Brandwijk-Kerkhof (see appendix II), and corresponds with the absence of pottery in phase 1 at Polderweg and the absence of domestic animals other than dog. On the other hand, the absence of crop plants can be considered as unexpected since the people at Polderweg had contact with people of the agricultural LBK (see paragraph I.2). Despite the absence of crop plants, there are various finds of waterlogged macroremains and pollen identifications of taxa of which the ecology is similar to that of arable weeds, i.e. taxa that show a preference for open, disturbed, eutrophic soils (see table I.35). How can we explain the presence of these taxa? 5 The selection of taxa in table I.3 is explained in chapter 10.

20

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

category

w

P

Arctium sp.

+

-

Artemisia sp.

-

+

Atriplex patula/prostrata

+

-

Capsella bursa-pastoris

+

-

Chenopodiaceae

-

+

Chenopodium album

+

-

Chenopodium ficifolium

+

-

Chenopodium murale

+

-

Cirsium sp.

+

-

Galeopsis-type

+

-

Galium spurium

+

-

Lychnis sp./Silene sp.

+

-

Moehringia trinervia

+

-

Persicaria lapathifolia

+

-

Persicaria maculosa

+

-

Plantago lanceolata

-

+

Ranunculus acris-type

-

+

Rumex acetosa-group

-

+

Solanum nigrum

+

-

Sonchus sp.

+

-

Stellaria media

+

-

Stellaria neglecta

+

-

urtica dioica

+

-

urtica dioica-type

-

+

urtica urens

+

-

Copyright © 2010. Leiden University Press. All rights reserved.

taxon

w = waterlogged macroremains P = pollen + = present - = not present Table I.3 Hardinxveld-Giessendam Polderweg, taxa of which the ecology is comparable to that of arable weeds (after Bakels and Van Beurden 2001; De Kort 1998).

Firstly, zones at the border of open water and rivers where drift litter is deposited form the natural habitat of certain taxa that developed into arable weeds after the introduction of agriculture. Secondly, the occupation of the dune resulted in disturbance of the vegetation. As a result, the dune formed an excellent habitat for taxa that prefer disturbance, the presence of nutrients (nitrate and ammonia) and light. The taxa indicative of disturbance were probably native, representing the natural vegetation of the river area, although there is a theoretical possibility that they were imported (unintentionally) from locations elsewhere where agriculture was practised. I.3.2.3 Carbonised macroremains of non-cultivated plants Non-cultivated taxa that are known as plant food sources of which remains were found in a carbonised state in the material of phase 1 are Malus sylvestris, Corylus avellana, Trapa natans and Ranunculus ficaria (tubers). The material of phase 2 comprises only carbonised remains of Trapa natans. The carbonised state suggests that people used these taxa in some way and their edibility suggests that they were consumed. The excavation at Polderweg did not reveal strong contextual evidence of consumption of the taxa mentioned above except for T. natans. The distribution of remains of this species of phase 1 shows concentrations on the higher part of the slope where the plant will not have grown since it is a water plant. On the basis of this distribution pattern, it has been concluded that T. natans was collected and consumed by people during phases 1 and 2 (Bakels and van Beurden 2001). Taxa that have edible parts and that were present at Polderweg in a waterlogged state only are Quercus sp., Prunus spinosa, Rosa sp., Crataegus monogyna, Cornus sanguinea, Sambucus nigra (/racemosa), Rubus caesius, Viburnum opulus, Nuphar lutea, Nymphaea alba. Furthermore some taxa with edible tubers were attested and a large variety of herbs and grasses of which the seeds, fruits and leaves are edible (see chapter 9). The presence of people at the dune during the early autumn can be deduced from the presence of carbonised seeds, fruits or tubers of Ranunculus ficaria, Corylus avellana, Iris pseudacorus and Trapa natans.

21

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The first three species can be collected in the autumn while tubers of R. ficaria can be collected from autumn to early spring (discussed in Bakels and van Beurden 2001). In addition, the carbonised fruits of Hedera helix may indicate the presence of people at the dune in March and/or April. These periods all correspond with the assumed period of occupation of Polderweg, as mainly based on archaeozoological evidence. The presence of Malus sylvestris is not informative of the season of occupation since it is concluded that this species is not part of the local vegetation (see paragraph I.7.1). This implies that wild apples could have brought to the site from elsewhere, possibly in a dried state. I.3.3 Wood and charcoal analysIs The discussion of unworked wood, charcoal and wooden artefacts is based on Bakels and van Beurden (2001), Hänninen et al. (1999) and Louwe Kooijmans, vermeeren and van waveren (2001). Material from phase 1 has the better representation in the analysis. A selection of 225 samples of unworked and presumably unworked wood trunks, branches and twigs has been identified (see table I.4). Alnus sp. and Quercus sp. were the most common trees in the vegetation, followed by Salix sp. and Fraxinus excelsior. Other identified taxa are Tilia sp., Ulmus sp., Acer sp., Corylus avellana, Viburnum sp., Cornus sp., Prunus padus6 and Pomoideae. Comparison of the unworked wood assemblages of the different phases generally shows continuity through time (Bakels and van Beurden 2001). Only Salix sp. forms an exception, being especially important during phase 1, when the distance to the crevasse channel was smallest. Salix sp. might therefore have been present in the local vegetation at the southern slope of the dune during phase 1. 0

1

N 1/2

taxon Acer sp.

-

-

1

1

2

-

-

3

1

1

Alnus sp.

-

40

11

35

86

-

38

30

43

38

Cornus sp.

-

3

1

1

5

-

3

3

1

2

Corylus avellana

-

6

-

1

7

-

6

-

1

3

Fraxinus excelsior

-

7

-

6

13

-

7

-

7

6

Pomoideae

-

-

3

-

3

-

-

8

-

-

Prunus padus

-

1

-

-

1

-

1

-

-

-

Quercus sp.

2

17

6

20

45

x

16

16

25

20

Salix sp.

-

16

3

3

22

-

15

8

4

10

Tilia sp.

-

1

1

-

2

-

1

3

-

-

ulmus sp.

1

5

-

3

9

x

5

-

4

4

viburnum sp.

-

-

1

2

3

-

-

3

2

1

Indet.

-

8

10

9

27

-

8

27

11

12

total

3

104

37

81

225

Copyright © 2010. Leiden University Press. All rights reserved.

phase

- = not present

2

total

0

1

% 1/2

2

total

x = no meaningful percentages

Table I.4 Hardinxveld-Giessendam Polderweg, unworked and presumably unworked wood for each phase, twigs included (after Bakels and Van Beurden 2001). 6 But see the remarks on the possibility to identify wood of Prunus species other than Prunus spinosa in chapter 7.

22

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The charcoal analysis is based on the investigation of 577 fragments. The selection of samples included a single concentration of each phase as well as a few additional samples from the different refuse layers, which were considered to represent scattered charcoal material without a specific context. The results are shown in table I.5. The dominant taxa are Alnus sp., Quercus sp., Fraxinus sp. and Corylus sp. Several taxa are present in limited numbers (Salix sp. and Ulmus sp.), while many others are scarce (Cornus sp., Rhamnus cathartica, Rhamnus frangula, Viscum album, Pomoideae and Viburnum sp.). The number of taxa in the charcoal assemblage is slightly higher than that in the unworked wood assemblage, probably as a result of the larger number of investigated charcoal fragments. The charcoal assemblage does not allow an analysis of changes through time since the samples of phase 0 were all collected from one concentration and since the number of identifications of phase 1/2 and phase 2 is rather small (contra Bakels and van Beurden 2001, 352). Characteristics of the charcoal indicate the use of old wood (Alnus sp., Corylus sp., Salix sp. and Alnus sp./ Betula sp.) and moist wood (Alnus sp., Quercus sp., Fraxinus sp., Rhamnus frangula and Viburnum sp.). In general, the charcoal assemblage does not indicate the selective use of taxa as fuel, but there are two possible exceptions.

N phase

%

0

1

1/2

2

total

0

1

1/2

2

total

Alnus sp.

1

114

25

94

234

2

32

50

78

41

Alnus sp./Betula sp.

-

-

1

-

1

-

-

2

-

0

Betula sp.

1

-

-

-

1

2

-

-

-

0

Cornus sp.

-

2

-

1

3

-

1

-

1

1

Corylus avellana

31

18

17

-

66

62

5

34

-

11

Euonymus europaeus

-

42

2

1

45

-

12

4

1

8

13

53

-

-

66

26

15

-

-

11

Pomoideae

-

1

1

-

2

-

0

2

-

0

Quercus sp.

4

95

3

13

115

8

27

6

11

20

Rhamnus cathartica

-

3

-

-

3

-

1

-

-

1

Rhamnus frangula

-

2

-

-

2

-

1

-

-

0

Salix sp.

-

8

-

7

15

-

2

-

6

3

ulmus sp.

-

9

-

2

11

-

3

-

2

2

viburnum sp.

-

7

-

2

9

-

2

-

2

2

viscum album

-

1

-

-

1

-

0

-

-

0

Indet.

-

2

1

-

3

-

1

2

-

1

total

50

357

50

120

577

taxon

Copyright © 2010. Leiden University Press. All rights reserved.

Fraxinus excelsior

- = not present Table I.5 Hardinxveld-Giessendam Polderweg, charcoal for each phase (Bakels and Van Beurden 2001).

23

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Firstly, one pit contained a relatively abundance of Corylus avellana charcoal (up to 62% of the contents of the pit). Secondly, one sample from phase 1 contained a large quantity of charcoal of Euonymus sp. (26% of the contents of the pit), which may indicate selective use in view of the absence of the taxon in the assemblages of macroremains, pollen and wood at Polderweg except for a single fragment of worked wood. It could concern selective use of the wood for fuel or for a skewer (Bakels and van Beurden 2001, 352). The assemblage of worked wood and artefacts yielded 57 identifications (see tables I.6 and I.7). The assemblage comprises a variety of objects interpreted as posts, fragments of a bow, hafts, paddles, pointed roundwood including possible points of spears, planks including possible fragments of canoes, skewers and digging sticks. Selective use of wood for artefacts based on the quality of the wood and the function of artefacts is indicated by the limited use of Alnus sp. and Quercus sp., which were dominant elements in the vegetation, and by the high correlation between artefact type and the wood species (see tables I.6 and I.7). In addition to the wood and charcoal, the finds comprised several knots and a fragment of a net made of the bark of Acer sp. and/ or Ulmus sp., Tilia sp. and an unknown (probably herbaceous) plant.

N phase

%

interpretation

1

1/2

2

total

1

1/2

2

total

Acer sp.

-

1

1

2

-

13

10

4

Alnus sp.

8

5

5

18

30

63

50

40 planks, post, other

Cornus sp.

-

-

1

1

-

-

10

2

roundwood

Corylus avellana

3

1

-

4

11

13

-

9

post, pointed roundwood, skewer, other

Euonymus europaeus

-

-

1

1

-

-

10

2

skewer

Fraxinus excelsior

3

1

1

5

11

13

10

11 skewer, roundwood, other

Quercus sp.

6

-

-

6

22

-

-

13 anvil, beams, plank, roundwood

Salix sp.

3

-

-

3

11

-

-

7

digging stick, plank, other

Tilia sp.

2

-

-

2

7

-

-

4

planks

ulmus sp.

1

-

1

2

4

-

10

4

other

viburnum sp.

1

-

-

1

4

-

-

2

digging stick

total

27

8

10

45

Copyright © 2010. Leiden University Press. All rights reserved.

taxon roundwood

- = not present other = other interpretation of the function Table I.6 Hardinxveld-Giessendam Polderweg, identifications and interpretation of presumably worked wood for each phase (Louwe Kooijmans, Vermeeren and Van Waveren 2001).

24

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

N phase

%

interpretation

1

1/2

2

total

total

Acer sp.

1

-

-

1

8

haft

Fraxinus excelsior

7

-

1

8

67

paddles, haft, points

Salix sp.

1

-

-

1

8

haft?

ulmus sp.

2

-

-

2

17

bow

total

11

-

1

12

taxon

- = not present Table I.7 Hardinxveld-Giessendam Polderweg, identifications and interpretation of wooden artefacts for each phase (Louwe Kooijmans, Vermeeren and Van Waveren 2001).

Copyright © 2010. Leiden University Press. All rights reserved.

I.3.4 other sources The botanical analysis included six species of fungi that indicate the local presence of trees including old or weak trees (Bakels and van Beurden 2001, see table 2.7). The analysis of arthropod remains indicated the presence of Caltha palustris, Phragmites australis, Solanum dulcamara, Equisetum sp., Alnus sp., and the presence of dying or recently cleared trees, dead wood, dead branches that are still attached to trees, dead trees with loose bark, trees that are strongly decayed, fungi, wood floating in water, and running water (Hakbijl 2001). This information all corresponds with the archaeobotanical results. I.4 GeoloGy of HardInxveld -GIessendam de bruIn Hardinxveld-Giessendam De Bruin is located on an inland dune in the central river area (coordinates 115.200/427.170). The top of the dune is at about 4.0 m -NAP and is slightly higher than Polderweg, which is located at 1 km distance to the northeast. During phase 1 the surface of the dune was 44.200 m2 and the dune protruded 4 metres above the landscape. During phase 2, the surface of the dune was 4200 m 2 and the relative height 1.5 metres. During phase 3, the dune was divided into two tops with a total surface of 940 m 2 and a relative height of 0.8 metres. The former landschape around the site was reconstructed by means of hand corings in an area of 1500 x 1000 metres around the dune, directly west of the Polderweg study area (see fig. I.8). During occupation, most of the area around the site of De Bruin was submerged and consisted of open water and marsh with a maximal depth of 1 metre. Carr vegetation was only present on the slightly higher patches in the surroundings. A few other dunes protruded above the floodbasin in the vicinity. The Holocene sediments at De Bruin largely resemble those of Polderweg. Here, however, occupation lasted longer and peat formation was interrupted twice by clay deposition resulting from fluvial activity. The first fluvial phase dates to c. 5500-5100 BC. The second fluvial phase started at c. 4700-4450 BC (Mol 2001b, 46). The influence of crevasse channels was slightly stronger at De Bruin than at Polderweg. Three crevasse channels were situated at a distance of 250-500 metres from the dune, one of them directly east of the dune. It is, however, not certain whether all crevasse channels were active simultaneously during the second phase of clay deposition. There are no channel belts situated in the investigated area around Polderweg and De Bruin (Mol 2001b, 2003).

25

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

-7.

Copyright © 2010. Leiden University Press. All rights reserved.

0

marsh

0

200 m

open water crevasse channel inland dune excavation trench road

26

Figure I.8a Hardinxveld-Giessendam De Bruin, the Pleistocene subsurface (m –NAP) and palaeogeographical reconstructions for phase 1 (after Mol 2001b).

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

marsh

0

200 m

open water crevasse channel inland dune excavation trench road

Figure I.8b Hardinxveld-Giessendam De Bruin, the Pleistocene subsurface (m –NAP) and palaeogeographical reconstructions for phase 2 (after Mol 2001b).

27

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

marsh

0

200 m

open water crevasse channel inland dune excavation trench road

28

Figure I.8c Hardinxveld-Giessendam De Bruin, the Pleistocene subsurface (m –NAP) and palaeogeographical reconstructions for phase 3 (after Mol 2001b).

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

I.5 arcHaeoloGy of HardInxveld -GIessendam de bruIn Excavation of a part of the eastern slope of the dune of De Bruin took place in 1998. The excavation trench (15 x 23 metres) revealed refuse layers formed during several phases, large pits with an unknown function though comparable with the pits at Polderweg that were interpreted as huts, round pits interpreted as fire pits, postholes, a feature interpreted as a small landing place for canoes and some intentional depositions, all covered by layers of clay and peat (see fig. I.9). The Holocene deposits contained flint, stone, pottery and organic remains including scarce bones of domestic animals (phase 3) and human remains.

canoe

fish trap

Copyright © 2010. Leiden University Press. All rights reserved.

canoe

sand

0

5m

marsh pit tree fall not excavated hearth concentration grave

Figure I.9 Hardinxveld-Giessendam De Bruin, features of all phases (after Louwe Kooijmans and Nokkert 2001, adapted by L. Amkreutz).

29

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The Holocene stratigraphy enabled to distinguish three occupation phases: phase 1 between 5475 and 5100 BC, phase 2 between 5100 and 4800 BC, and phase 3 between 4700 and 4450 BC.7 Phase 1 yielded relatively few finds and is possibly separated from phase 2 in view of a level with a low find density, despite the overlap of c. 250 calibrated years in 14C dates (Mol and Louwe Kooijmans 2001b, 69). Phase 2 yielded an extensive find assemblage and is separated from phase 3 by a more distinct archaeological hiatus. Fig. I.10 shows the extent of the corresponding refuse layers. The site can be characterised as Late Mesolithic to Early Neolithic and covers the availability phase as defined by Zvelebil (1986) in view of the presence of pottery from phase 2 onwards and the small percentage of domestic animals during phase 3. Phase 1 is contemporaneous with the ältere and jüngere LBK, while phase 2 is contemporaneous with the jüngere LBK and the Großgartach culture. Phase 3 is contemporaneous with the Rössen culture (Louwe Kooijmans 2001b, 527). The pottery from phases 2 and 3 is characteristic of the Swifterbant culture, and shows some Blicquy elements in phase 2 (Raemaekers 2001). The bones of domestic animals in the assemblage of phase 3 (cattle, pig, goat and sheep) are the oldest in Northwestern Europe to the north of the loess zone. The number of bones, the distinct concentrations and the fact that they mainly consist of limb bones suggest that it probably concerns intentional depositions, interpreted as offerings, and that there were no living domestic animals present at De Bruin, with exception of dogs and possibly pigs. The introduction of pottery and domestic animals shows a gradual neolithisation process. The Blicquy pottery and characteristic flint and stone (phases 2 and 3) indicate contacts with southern regions. In addition, the flint indicates some contact to the north during phases 2 and 3. The nearest finds of similar material originate from the utrechtse Heuvelrug area at c. 50 km distance. The site is interpreted as a base camp, possibly with a shift in function from a winter base camp (similar to Polderweg during phase 1) to a year-round extraction site. For phase 2, there are indications of occupation during all seasons except for late autumn and early winter (November-January). For all phases, there are some minor indications of occupation during summer. During phase 3, the site was probably too small to function as a base camp (Louwe Kooijmans 2001b, 515-517).

Copyright © 2010. Leiden University Press. All rights reserved.

I.6 arcHaeobotany of HardInxveld -GIessendam de bruIn The archaeobotanical data of De Bruin consist of pollen columns collected inside and outside the excavation trench and macroremains, wood, charcoal, coprolites and fungi collected inside the excavation trench (Bakels et al. 2001; verhoeven 2003). I.6.1 Pollen analysIs InsIde the excavatIon trench For the pollen analysis inside the excavation trench, material of phase 1 was sampled in the western section of square 16, and material of phases 2 and 3 was sampled in the western section of square 24. The distance to the dune top is similar for both squares. The sample locations are situated within the refuse layers, and the distance between the sample locations and concentrations or considerably large quantities of refuse is less than 4 metres for all phases (see fig. I.10). Sampling was performed with the sampling boxes DB. The pollen diagrams are based on an upland pollen sum of at least 300 grains. The material from square 16 consisted of three spectra that reflect the vegetation but do not represent changes of the vegetation before, during and after occupation phase 1 (as a longer sequence of peat that reflects the relevant period was not available). The material of square 24 represents at least the period during and after phase 2 and the period during phase 3, and is directly dated between 5279-4959 BC and 4685-4367 BC (see fig. I.11 and the publication for 14C dates). The top of the investigated sediment of square 24 may be 200 14C years younger, assuming a continuous sedimentation rate. 7 Recalculation of the 14C dates of De Bruin resulted in the new conclusion that phase 1 dates between 5230±150 and 5110±90 cal BC, phase 2 dates between 5040±80 and 4850±170 cal BC and phase 3 between 4560±100 and 4480±160 cal BC (Mol and Van Zijverden 2007). See also fig. 2.5.

30

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

a -10.5

-1

0.

5

-9.5

DB .5

-6.5 -8.0

0

-10 .

.5

-10.5

-5.0 -5.5 -6.0

-9.5

.0 -7 5 . -7

-6.

-9 .0

0 -4.5

-4.

-4.5

- 10

-4.0

0

-5

-9.0

HB .0

0

-8

50 m

b -10.5

-1

0.

5

-9.5

DB .5

-8.0

-6.5 -7 .0

-10

-10.5

-5.0 -5.5 -6.0

-9.5

.5

0

.5

-7

-6.

-4.5

-9

0.0

0

-4.

-4.5

-1

-4.0

.0

-5

-9.0

HB .0

0

-8

50 m

c -10.5

-1

0.

5

DB -9.5

.5

.0

-5

-10

-10.5

0

-6.5 -7. 0

-5.0 -5.5 -6.0

-9.5

.5

0

-6.

HB

clay loam

excavation trench

sand

find layer

-9.0

peat

-8.0

.5

0. -1

0 -4. -4.5

-9

-7

Copyright © 2010. Leiden University Press. All rights reserved.

-4.5

.0

-8

.5

-8

0

50 m

Figure I.10 Hardinxveld-Giessendam De Bruin, the Pleistocene subsurface (m –NAP), the extent of the surface of the dune and the refuse layers for phases a) 1, b) 2 and c) 3 (after Mol 2001b). The location of the pollen sample locations is indicated by DB and HB. dune at Hardinxveld-Giessendam De Bruin for phases 1 (a), 2 (b) and 3 (c) (after Mol 2001b). The location of the pollen sample locations is indicated by DB and HB.

31

Copyright © 2010. Leiden University Press. All rights reserved.

32

100

yr

nu s

lu pu lu s

DB3: 5685 ± 50

300 500

Figure I.11 part 1. 800

Peat

Sand

Humic sediment

Sandy sediment

pl

Li d

ol an

th

es

y t re

og

AP )

BP ) -N

rs

20 40 60 80 100

20 40 60 80 20

Clayey sediment

20 40

ue

rc

us

20 40 60 80 20

Wetland trees, shrubs, herbs and spore plants

m

or

e

pl

an

R h V i am b n C ur n us or u c Po n u m at h s o A r l yg s a p u a r t i t o n lu c C em num g u s a h i i P t eno sia pe n e a rs -t e Po r i d p o d ic y p ar e ly ium iac po ea ia - t a di q e yp um ui e lin um

Fr a U xin lm u u s Be s t u Fa la g H us e V i de s r C cu a or m y l al us b u

Q

U p U lan pl d a Pi n d s hr c e h ub e Pi a r b s s nu an Ti s d lia sp

U

(y

(c m

es

th

at

ep

D

D

14 C 620

Eu U pho r ti r c a bia di oi ca -t y Ly pe t R hr u um m Fi e s li x a C p en aq li c a a u r C l t h a dul at i i a ir a cu A l c ae pa slu i gr Lo sm a st r ou a is -t y p Ly tus -ty si ul pe p e Sa m ig a in S ogit t c hi os u a a Splan r ia t h s u s y Sparg m d ag r sif it lo a T h ar g niu ulc t i fo r a an m am lia a Va lic iu e a t m m r Ty leri rum e er a a re su Er p h a n a ct m um - t i yp S p c a l l at i e f e ol h s T h ag iu m n O e ly p um s C m u te r i e n s M r ato d a p a y r l M rio phy egaust ri y p l M r i o y llu lum lis s y p m N r i o p hy l a u lu l N p h hy l m t e r n y m ar lu sp if Tr p lu m ic lor a h t v a u Po p a a e e a - e r t t u m m n a t ic Petam ata albype illa t a o d M ia ge ns - t y u m o u st to pe g e r um n ot ia

Sa ic a li R x ib e H s um ul us

M

Al

ts

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Upland trees, shrubs, herbs and spore plants

640

660

680

700

720

DB4: 6010 ± 55 740

DB5: 6120 ± 50 760

DB6: 6170 ± 50 780

20 40

Open water

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The pollen diagram of square 24 (see fig. I.11) indicates that the natural vegetation on the higher parts of the dune consisted of mixed deciduous woodland including Tilia sp., Quercus sp., Fraxinus excelsior, Ulmus sp., Hedera helix, Viscum album, Corylus avellana, Rhamnus cathartica, Viburnum opulus, Cornus sanguinea, dryland ferns, Artemisia sp. and Chenopodiaceae. The diagram further points to the presence of carr vegetation around the dune dominated by Alnus glutinosa, open patches with marsh and forb vegetation including Phragmites australis and Carex vegetation, open water with a minimum depth of 1.5 to 2 metres and shallow open water. The material of square 16 (phase 1) does not give explicit information on human impact on the vegetation. Indications of human impact in the pollen diagram of square 24 are only present at 7.50-7.40 m -NAP, showing a decrease in Quercus sp. and Tilia sp. and an increase in Pinus sp., Hedera helix, Corylus avellana, Rhamnus cathartica, Viburnum opulus, Chenopodiaceae, Poaceae, Humulus lupulus and Urtica dioica-type, as discussed in Bakels et al. (2001, 378). It can be added that the curve of Alnus sp. decreased and the curve of Cyperaceae increased. The disturbance can be more easily recognised in the curves of wetland taxa than in the curves of the dryland taxa since the wetland taxa give higher peaks. The changes indicate the replacement of closed (dryland and wetland) woodland vegetation by more open shrub vegetation, but are not strong and are very temporal of character. It is difficult to distinguish whether the changes are caused by natural factors of disturbance, such as the submerging of the dune, or by human impact. The recovery of the wetland vegetation, as indicated by a peak of Alnus sp. and the return of other taxa to previous values, nevertheless indicates that it probably concerns human impact. For phase 3 there are no indications of human impact, despite the similar distance to the refuse layer and despite the small distance to the dry surface of the dune (max. 8 metres during phase 3, see Louwe Kooijmans and Nokkert 2001, fig. 4.1).

Ap A s i ac e A s te r a ae R te r a c e a a e R nuncea tu os c e b C ac ulu li g uli f a e u l B r r yo a e s a l i f l o r a cr or e a p is ae P o s s hy -t ic ll R te n ac ac e y p e um t il ea a M e la - e e e x t G nt h a c y p e a a e M lium - t y to s o p a Tr no -ty e - g ro i le p up Ly leta tae e c o e, , p po ps si di i l a l at um t a ae (me ar ke r) Po lle n su Zo m ne

C

yp e Po r a c ac e ae ea e

Ecologically indeterminate

329

Copyright © 2010. Leiden University Press. All rights reserved.

486 353 676 354 341 576 318 466 398 312

20

20 40

319 319 331 366 294 50 100150200

IV

III

II

I

Analyst: L. van Beurden, 1999

Figure I.11 Hardinxveld-Giessendam De Bruin, square 24, pollen diagram DB based on an upland pollen sum, exaggeration 5 x (after Bakels et al. 2001), part 2.

33

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Summarising, possible human impact can only be distinguished during phase 2. Importantly, the sediments of phase 2 contain a considerable concentration of sturgeon remains (Acipencer sturio), located more or less at the same spot as the pollen sample location (Beerenhout 2001, 316). This concentration of fish remains can be interpreted as the waste from the local cleaning of fish. Furthermore, a structure interpreted as a landing place for canoes was present at only few metres away from the pollen sample location. It therefore appears that the relative strong signal of possible human activity in phase 2 indeed reflects human impact. The signal thus forms an exception to other changes in the pollen diagram, which is caused by sampling in the middle of a specific activity zone at the border of the dune. I.6.2

Pollen analysIs outsIde the excavatIon trench

I.6.2.1 Introduction In addition to the boxes sampled inside the excavation trench, a pollen core was sampled south of the dune outside the excavation trench, beyond the area of human activity (coordinates 115.168/428.144, see fig. I.10). This new core HB was investigated by K. verhoeven and w.J. Kuijper (verhoeven 2003). The additional data enable the reconstruction of the natural vegetation and human impact, and additionaly allow a comparison of the two pollen diagrams. The core was located further away from the main find concentration than the on-site sample boxes DB, and is expected to show relatively weak evidence of human impact. However, the distance between the two sample points and the dry surface of the dunes is comparable, possibly reducing the difference. The environmental conditions at the two sample locations were probably different, since the location of the sample boxes DB was strongly influenced by the crevasse channel, while core HB was located further away from this channel.

Copyright © 2010. Leiden University Press. All rights reserved.

I.6.2.2 Materials and methods The sediment of the core HB consisted of amorphous peat and detritus-gyttja, in the lower part of the core mixed with sand from the dune and in the upper part of the core mixed with clay deposited during the Middle Ages. The pollen samples with a volume of 1 cm3 were prepared according to the standard methods, and pollen, non-pollen palynomorphs and macroremains from these samples were identified. To each sample, a Lycopodium tablet containing c. 12077 spores was added. The sample interval of the analysis is 20 cm in the lower and middle part of the core and 40 cm in the upper part of the core.8 The pollen diagram is based on an upland pollen sum of at least 300 pollen grains (dryland trees, shrubs, herbs and spore plants including crop plants). The classification of plants in ecological groups is based on van der Meijden (1996) and weeda et al. (1985-1994). Veronica beccabunga-type represents V. beccabunga, V. anagallis-aquatica and V. catenata. I.6.2.3 Dates In order to correlate both pollen diagrams of De Bruin, four samples of the core were dated (see table I.8). The dates demonstrate that the core represents at least a part of the (Late) Atlantic and Sub-Boreal. In addition, pollen of Juglans sp., Fagopyrum sp., Centaurea cyanus and Zea mais in the very upper part of the diagram indicate that the Sub-Atlantic is represented in the diagram as well. This last period is not relevant for this study and will therefore not be discussed here. The reconstruction of the ground water level indicates that peat accumulation at the location of the core started at c. 5200 BC (Louwe Kooijmans and Nokkert 2001, 75),

8 The identification was based on literature (Fægri and Iversen 1989; Moore et al. 1991; Punt and Blackmore 1991; Punt and Clarke 1980, 1981, 1984; Punt et al. 1976, 1988, 1991; Reille 1992, 1995) and the reference collection of the Faculty of Archaeology, Leiden University.

34

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

sample

depth

lab code

(m -NAP) HG De Bruin 4

3.29-3.30

GrA-30181

age

age

(yrs BP)

(yrs cal BC, 2σ)

3975 ± 45

2620-2330

dated material Alnus sp., three fruits, a bud and a bud scale

HG De Bruin 3

4.28-4.29

GrA-30180

4565 ± 45

3500-3090

Alnus sp., four fruits, Lythrum salicaria, three seeds

HG De Bruin 2

5.28-5.29

GrA-30178

5400 ± 60

4350-4050

Alnus sp., seven fruits

HG De Bruin 1

6.68-6.69

GrA-30177

5680 ± 60

4690-4360

Alnus sp., four fruits and a cone, Humulus lupulus, one fruit, Ranunculus repens-type, one fruit

Table I.8 Hardinxveld-Giessendam De Bruin, 14C dates of core HB (Center for Isotope Research Groningen).

Copyright © 2010. Leiden University Press. All rights reserved.

presumably corresponding with phase 1.9 The large age difference of 500-800 years between the estimated age of the lowest level of the core and the first date at 6.68 m -NAP can be explained by compaction of the peat and possibly by absence of sedimentation and/or erosion since the presence of sand in the lower part of the core suggests that only restricted compaction could have occurred. Date HG De Bruin 1 corresponds with the period during or after occupation phase 3 and with the youngest date from the on-site samples boxes of De Bruin (date DB 3, GrA-14864), indicating that the lower part of diagram HB corresponds with the upper part of diagram DB. I.6.2.4 Results and discussion Figure I.12 shows the results of the pollen analysis and macroremains of core HB. The diagram is divided into four zones: zones I (7.27-5.97 m -NAP), II (5.97-4.12 m –NAP, divided into zone IIa and IIb at -5.37 m), III (4.12-2.67 m -NAP) and Iv (2.67-1.87 m -NAP). The diagram presented here is highly similar to the unpublished diagram (verhoeven 2003)10 and I refer to this primary source for an extensive discussion of the diagram. In general the pollen diagram of core HB is similar to the diagrams of Polderweg and De Bruin and indicates similar vegetation for the Atlantic and Sub-Boreal. At the dune, dryland vegetation consisting of deciduous woodland was present, mixed with open patches with shrubs, herbs and spore plants. The wetland environment consisted of alder carr and open marshes. Alder carr was probably present in the earliest period, as indicated by a high percentage of Alnus sp., Cyperaceae, monoletae psilatae fern spores and scalariform perforation plates (type 114). In the remaining lower part of the diagram, representing the period of occupation of the dune, the importance of alder seems to be restricted. From 4.62 m -NAP onwards (the middle part of the diagram), the local presence of alder carr is indicated by the same curves as earlier.

9 Peat accumulation at the sample location could only start when the ground water level reached the surface. As a result, models of the rise of the ground water level can be used to date the lowest part of the core. 10 The presented diagram differs from the original diagram in the combined presentation of pollen and macroremains, and in the zonation.

35

Copyright © 2010. Leiden University Press. All rights reserved.

36 or yl us

1: 5680 ± 60 750

20 40 20 40 60 80 100

Upland shrubs

Peat

20

Detritus-gyttja

Figure I.12 part 1. 20 20 20 40 60 20

s g nu lut i n s A l sp osa nu ., c s Sa sp one lix ., f s ra gm R ha en H m ts um nu of s ul f m us ra al e lu ng ca pu ul a tk lu in s s

2: 5400 ± 60

nu

3: 4565 ± 45

2000

Al

4: 3975 ± 45

Al

C o R r y lu ha s V i m sp bu nu ., r s an C num c a t he th r or Sa nus opu ar t i s m s a lus c a Li bu n g c g C ust us u i n he r u n e A r no m igr a-t a - yp t p ty e Pl e m o d pe an i si iac Tr ta a ea ifo go e Pl liu la a m n A l nt a - t y c e li g p o Po u m o m e l at a aj l or Po yg o /m nu l y C go m ed er n s ia Pr ea um e c un lia a . p C e t y vi e e lla p c r s Pa nt a - t y e u l a i c a re r i p u p a S a av r e a e m e r jac C bu ea e c -t y Fa nta us pe g o ure eb Ze p a u a y c lu S p m r u m ya n s a us Po e r g i s l u P t ypo la a e d r Al ridi iumven nu um si s s a qu ili nu m

C

14 C

th

S

ep

ue

rc

us Q u U erc lm us us s p. Fr ,b ax ud in s us Be t Ac ula e Fa r g C us ar J u pin g u H lan s ed s V i er sc a um al bu m

Q

U p U lan pl d a s Pi n d hr u nu he bs s rb s an Pi d ce sp Ab a or e i pl A b es / an P ie ic ts Ti s e lia a

Li

D (c ate m s( - N yr th ol AP s B o P) ) U gy pl an d t re es

D

AM

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Upland trees

200

250

300

350

400

450

500

550

600

650

700

20

Upland herbs and spore plants Wetland trees and shrubs

4168

20

Copyright © 2010. Leiden University Press. All rights reserved.

20

20

20 40 60

200 400 600

20 40 60

ce

>3 5

ae

e μm

Open water

O e B e nan r t A p ula he e a A s i ac e r e qua te a e c t a t i c ra a c G al eae Er i u m tu i b A s c al e - t y p uli f lo te s e ra ra e c M en eae C th l ig ar a u S t yo - t y p l i f l or ac p h e ae Br hy yll as s - ac V i si t y ea c c p Eu i a a e e e ae Ve pho ro r b ni ia ca

ia

ea

20

Ap

ac

20 40

Po

N y G mp lo ha bo e a N se c e y m m ae ph ic ro , m ae fo u ac s s c i la Sp ea il ( gi iro t y no Sp g e, y t r i pe us c f iro g ra s c ho 128 hai . y p ) rc sc M Spi ra ., z el le o ro sp y ls r id Pe uge gy ., r go (ty es ot r a et s p d pe D ias ia sp i c u o r ( t ap t s ., l es 12 y p 7) e Zy hn rum p., re ata (t 12 gn ia s z y tic e ( yp 9 C e sp p. go ul ty e ) yp m ., sp ata pe 130 er a - e or e 13 ) ac t y p p h ( ea e ipp es cf. 2) 13 e (ty ia 2) pe (t y 31 pe 4) 1 C 89 ar E) Po ex ac sp ea . e la nd he

Wetland herbs and spore plants

rb s an Sp d ar sp Sp g or a a e Fi r g n i u pl lip an m an Ly en ium em ts t h du e T h r u la e r r su e a m ct m R li c s a um - t y u tr l pe S o m e um i c ar x ia l U anu cf. r ti m p c a d alu U - t y p ul c st r i r ti e am s Al ca ar i s di a A l ma oic is - t a m y R a p u e Ty me sp. p x I r i ha ac e s la t Lo p s e t i fo o s a t l Ly us uda ia gro up si uli c o Sa ma gin r u g c o s R it t hia su u a s C me ria thy al x s a r s Va tha aqu git iflo l p a ti r S y e r ia alu t i c fo li a m na st us a r is - g Sp ph - ro h y M ag tum t y p up e o n nu ol m et ae Tr ,p il si Tr eta la ta i l e e, e t p Tr a s e ile , ila N t a sc t a up e a e N h a , ve b r a up r r r t a N h a l u te u c e u r a N p ha lut a - t y t ae ym r ea p e s N p h p. ym a / Po p e a N y t ho a m C am i de lba p h al o s - ae C l i t r g et p e t y p a ha i c o lt e sp . r a h e n at a sp sp ., . oo go ni a

W et

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

20 40

Open water

20 Ecologically indeterminate

20

20

Figure I.12 part 2.

37

Copyright © 2010. Leiden University Press. All rights reserved.

38 358 452

373

331

369

50 100 150

366 350 362 379 413 401 423 340 351 349 353 404 404 383 407 341 381

Zo

lle n

op

ne

Po

Ly c m

m

iu

su

od (m ar ke r)

yp

er o s i c um ac R an ea u e R nc an ul un uc c u ea lu e s ac r is Ve -t y r pe In o ni de c a t. b ec ca G bu e ng c f lasi a. C no ty T i la sp pe ll e st o Ac t ia ero r a s a s s U r i ( p h p o p. ( st t y a r i t y A n alin p e gni um p e i a 3 ( c 1 Ar ma de 6) typ arc ) e in an ls us 27 u Te e ha ta ) m t a ir , (ty G r a p e, c s ( a s c pe ae l o la c f o . D um a a ws t y sp 25 ip a r ( p o ) Lo o r o nn i st a t y p e 5 r e s p t o t e 2 ( Ar ho hec my a (t 71 ) typ e c p a c y ) 44 R ella us r h e s pe ha d c r iz sp 8 9 ) o i R b d i s st a p h . ( ) ha o c o l il t y Fu bd coe ide linu a ( pe ng oc la s s, typ 12 al oe , c (cf sta e 6) po la, oc . ty to 14 ly co oo pe bla 3) U he c ns 3 s ni dr oo (t 52 ts Pe s e p ic ns yp ) (ty r fo t a ce ( e pe t ra e lls t y p 3 5 19 t i o as (ty e 3 3A 2) n co pe 5 3 ) pl sp at o 36 B) es re 1) (A s ln us U sp ni .? de )( nt ty if i pe ed 11 fu 4) ng al re m ai Sa ns nd R

H

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Ecologically indeterminate

20 40 20

Varia

20 50 100 150 50 100 150

IV

378

III

IIb

IIa

I

Analysts: K. Verhoeven and W.J. Kuijper, 2003

Figure I.12 Hardinxveld-Giessendam De Bruin, core HB, pollen, non-pollen palynomorphs (NPP) and macroremains diagram, pollen and non-pollen palynomorphs percentages based on an upland pollen sum, exaggeration 5 x, + = present (after Verhoeven 2003), part 3.

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

An important difference with other pollen diagrams of Hardinxveld-Giessendam is that this diagram represents the development of the dune vegetation relatively short before the submergence of the dune, and the subsequent developments. The point of submergence of the dune roughly corresponds with date HG De Bruin 2 (4350-4050 BC) at the start of zone IIB (5.29 m -NAP), since this date can be placed shortly after the end of occupation at De Bruin (presuming that occupation continued as long as possible). The submergence is supported by various changes in the relevant spectra of the pollen diagram. Firstly, the diagram shows the stabilisation of the curves of the total percentage of dryland trees and shrubs at the transition from zone IIa to zone IIb (after 5.37 m -NAP), while the percentage of dryland shrubs starts to decrease. Secondly, pollen grains of Picea sp. and Abies sp. are regularly present from the same level onwards, both indicating the influence of river water. Thirdly, the percentage of dryland shrubs is relatively high in zone IIa, and sand and macroremains of Quercus sp. and Corylus sp. are still present, indicating the local presence of dryland vegetation and erosion respectively. There are however no longer any indications of the local presence of dryland vegetation in zone IIb. The relatively high value of Ulmus sp. and the increase in Quercus sp. in zone IIb probably represent transport by river water since the peak values correspond with small peaks of Abies sp. and Pinus sp. In addition, the curve of the total percentage of wetland herbs shows an increase at 5.37 m -NAP (related to the increase in S. dulcamara, Urtica-type, Alisma-type, Rumex acetosa-group and others). Finally, changes in the curves of various non-pollen palynomorphs indicate the submergence of the sample location. From the start of zone IIb (slightly higher than 5.37 m -NAP), the percentage of non-pollen palynomorphs increases strongly and starts to fluctuate, while the curves of Spirogyra sp. and Mougeotia sp. appear, indicating the presence of stagnant shallow eutrophic water. Type 128 shows a continuous increasing curve as well, indicating shallow, mesotrophic or eutrophic, stagnant or slowly moving water. In addition, the curve of unidentified pollen grains starts at the same level, also indicating environmental change. The macroremains support the occurrence of environmental change at the levels 5.28 m -NAP and in a minor degree at 5.08 m -NAP (both lower part zone IIB), characterised by a significant increase in the number of taxa and the number of macroremains, especially macroremains of wetland (riparian) taxa. The increased seed production is probably related to the final submergence of the (dryland) woodland of the dune, resulting in the drowning of the trees, which created major gaps and led to the absence of living woodland vegetation. This resulted in improved access of herbs and wetland taxa to light at the sample location, leading to increased seed production. Human impact as the cause of increased seed production is unlikely since level 5.29 m -NAP corresponds with date HG De Bruin 2, which clearly postdates the last occupation phase. If the changes at the beginning of zone IIb represent the submergence of the dune including the sample location, the fluctuations of dryland shrubs and herbs in zone II have to be explained, especially since there were no other dryland patches left in the direct surroundings of De Bruin during/after the last occupation phase and since the dispersion of pollen of dryland herbs is generally limited. Pollen grains of dryland taxa present in the middle of zone II may represent the last elements of dryland vegetation at the dune since the top of the dune is located higher than the mowing field at the location of core HB. The macroremains and the dates however strongly indicate that this argumentation cannot be used to explain presence of remains of dryland taxa in the upper part of the zone. The rise of Quercus sp. in zone IIb therefore definitely does not represent vegetation on the dune De Bruin, while similarly the pollen of dryland shrubs cannot represent (extra-) local vegetation. The small peaks of Artemisia sp. and Plantago lanceolata may simply represent pollen that was transported by the river water since these taxa are part of the natural vegetation along levees.

39

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

I.6.2.5 Human impact Anthropogenic influence on the vegetation in the pollen diagram of core HB related to the investigated occupation periods can be expected in the samples before date HG De Bruin 2 (5.29 m -NAP: c. 4350-4050 BC), since this date likely represents a terminus ante quem of occupation. The pollen diagram shows that anthropogenic influence on the vegetation may be present in the lower part of zone I (first two samples), indicated by the relatively high value of Pinus sp. and the relatively low value of Quercus sp., the presence of small peaks of Artemisia sp., Plantago lanceolata, a small peak of Cornus sanguinea-type, and the presence of Euphorbia sp. The strength of this signal is however limited. In the next spectrum, Corylus sp. shows a peak volume, which may represent human impact, but this does not correspond with other changes. Zone IIa may reflect anthropogenic influence on the natural vegetation as well, indicated by a decrease in Fraxinus sp., a peak of Viburnum sp., a slightly later decrease in Quercus sp., the presence of Allium sp. (cf. the Hazendonk, appendix III), a peak value of Apiaceae and a decrease in Cyperaceae. This potential signal is however not supported by the presence of any of the classical anthropogenic indicators. In addition, the curves of water plants show relatively high values as well, which may explain some of the observed changes. Summarising, the changes that may be related to human impact at the dune are rather weak and are only moderately supported by changes in other curves, as was the case in the earlier published diagram DB (see paragraph I.6.1). This can be related to the character of the activities and the distance between the sample location and the activity zone. Due to the weak evidence of disturbance, it cannot be excluded that other disturbance processes, such as abiotic factors or the activity of wild animals, resulted in the observed changes in zones I and IIa. Important causes of natural disturbance may concern the rising water level and channel activity, as indicated by the almost continuous curve of Chenopodiaceae in both zones and the presence of pollen grains of Abies sp. and Picea sp. in zone IIa. I.6.2.6 Correspondence with the pollen analysis inside the trench This paragraph compares core HB with the earlier published pollen diagram DB. Zone B-Iv of the earlier published diagram (Bakels et al. 2001) is contemporaneous with zone I and possibly with the lower part of zone II (below 5.29 m -NAP, date HG De Bruin 2) of core HB. The upper part of zone B-III of the published diagram may additionally be contemporaneous with the very lower part of zone I of core HB (below date HG De Bruin 1, 6.68 m -NAP). A difference between both sample locations during phase 3 is the deposition of clayey peat at the location of the published diagram, related to presence of a nearby crevasse channel. This results in relatively smooth curves in this diagram (zone B-Iv). At the sample location of core HB, the deposition of clay started later and the influence of the crevasse channel was weaker. The results of core HB show that the results on the development of the vegetation from both sample locations are highly similar on a general level. Analysis and comparison of the contemporaneous zones of the pollen diagrams shows that it is not possible to compare the development of the vegetation in detail. This is on the one hand due to the large distance between the sample points, and on the other hand due to the large sample interval and the smooth curves. The first factor, the distance between the sample points, reduces the possibility of comparing the presence of those taxa which produce little pollen. For the earlier published diagram DB it is furthermore difficult to reconstruct which taxa were present in the local environment and which taxa represent pollen transported by the channel. The second factor restricts the possibility to discern changes in the vegetation that occurred at both sample locations and which may link the diagrams to each other. Interestingly, there is one exception, since there is one example of a change in the vegetation that can be observed in both diagrams. The earlier published pollen DB diagram shows a peak of Corylus sp. in zone B-Iv that is contemporaneous with occupation phase 3. The diagram of core HB also shows a peak value of Corylus sp. in the lower part of zone I. These peaks, probably indicative of more open vegetation, may indicate that the diagrams correspond at these levels.

40

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

I.6.3

MacroreMaIns analysIs

Copyright © 2010. Leiden University Press. All rights reserved.

I.6.3.1 Reconstruction of the vegetation The analysis of macroremains concerned grid samples (N = 12, sample size 0.5 litres, minimal mesh width 0.25 mm), samples from features (N = 15, sample size 0.5 litres, minimal mesh width 1 mm), samples analysed in order to investigate the presence of carbonised cereal remains (N = 50, sample size 0.5 litres, minimal mesh width 1 mm), and a high number of handpicked samples and remains collected from the residue of sieving (minimal mesh width 4 mm) from phases 1, 2 and 3. The grid samples were collected in the zone that was waterlogged during occupation. In addition, macroremains in the pollen samples from the sample boxes collected in square 24 were analysed (sample volume 1 cm3, mesh width 0.25 mm), as well as larger samples collected from the same sample boxes (N = 10, mesh width 0.25 mm, sample volume 0.15 litres). The analysis of the last group of samples (volume 0.15 litres) unfortunately did not include those spectra where the pollen diagram shows the strongest indications of human impact. The representativity of the macroremains analysis of phase 2 is restricted due to the small number of samples and the applied mesh width. The variety of ecological groups in the taxa is large, which can be explained by the presence of a water channel next to the dune and the rolling down of macroremains from dryland taxa downwards along the slope of the dune, while anthropogenic influence may have played a role as well. Comparison of the macroremains of the different phases indicates that the water level rose through time, resulting in a shift of the vegetation belts in an upslope direction. Macroremains of Tilia sp. are very common in the material of the first phase. During phases 2 and 3, the relative importance of the macroremains of of dry terrain woodlands and woodland edge vegetation, including Tilia sp., decreased, while the importance of marshland and riparian taxa increased. The marsh taxa indicate moderate to highly eutrophic conditions in an open or moderately shaded environment. The macroremains of phase 1 contained a theca of Betula sp. with pollen grains inside, strongly indicating the local presence of Betula sp. The analysis of the 0.15 litre macroremains samples (fig I.13, next page) confirms the gradual submerging of the dune at the end of phase 2 as concluded above. Nevertheless, the remains of Tilia sp., Quercus sp. and various shrubs indicate that woodland vegetation of dry terrain still remained present in the extra-local vegetation. The macroremains analysis of the pollen samples (fig I.14, next page) shows the presence of charcoal and sand in large parts of the core, indicating colluviation and/or human activity. The results of both diagrams correspond with each other.

41

Copyright © 2010. Leiden University Press. All rights reserved.

42

20 40

Figure I.13 part 1.

20

20 40 60

20 40

ha

.

lu

sp

cu s aq ua til

is

-t y pe C ar R ex u a Be me cu r x tif C u l a hy d o r m al e r i Ly tha re ola s c p c Va op palu ta ath um l e us st Eu r i a e u r i s n p r O at a o o p e o f a S c nan r ium f i c i eus n h t Ph o e h e c an ali s no aq n a l S i ar p ua ab u is le t in S p m l ar c tu i c a um a a u s M r g t i fo n di lac en an liu na . / Fi t h i u m c e t a b li a m a . S o pen aq er la du ua e c t Po n u l a t i c u a m ul a / m p d m ar S c a l u u l c a r i ve st a a ns h e St oe r is ma ra ac n o hy p l s ec pa tu lu s l st ac r is u st r is

Ty p

un

820

an

th (c m

20 20 40

20 40

Peat

Sand

Clay

Humic sediment

Sandy sediment

Clayey sediment

nu s A l s p. ,c nu o R s ha sp ne s m ., nu f r . s of fra m ng al ul e c a a

Al

tk

-N AP o ) Ti gy lia Ti p li la Q a p t yp ue la h C rc t yp y llo o u h s V i r n u s sp y llo , f r b s . s u C ur n s a , ju , s it s o r u n g ve e e C y lu m u i ni ds ra s op n le U t a e av u l u e a r ti g el s c a us la d m na M oe ioic on a og Po h r yn a in a C c f gia he . t c Fa lid ha rin e o i l A n lop niu xii r via t ia m C hr i du m he sc m aj n o u s et u s p o sy o r u di lv m um e s al t r is St bu el m l Pe a r r ia So sic m l a e A l anu r ia dia nu m m s n ac sp ig ul . r u os m a ol

ep

th

D

Li

660

R

C ic Ph ut a v R rag iro an m s S t un i te a e c s A l llar ulu aus is ia s t r m a f la al a qu m is pl a m an t ic u t a a la Al go is -a m qu a Ly s at t h p. ic a ru m sa li c ar ia

in

s

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Upland trees, shrubs and herbs Wetland trees

680

700

720

740

760

780

800

100 200 300 20

Wetland herbs and spore plants

20

Copyright © 2010. Leiden University Press. All rights reserved.

620

660

720

760

800

ol

og

th (c m -N AP )

Analyst: T.J.J. Vernimmen, 1998

y Ti lia U sp r ti ., c j A l a d u ve nu i o n i Ly s s ica le th p. Ep r u m il s Ph o b i u a li ra m c ar C gm hi ia ar it r s e e u R x s s a tum an p u . Tr unc , bi stra typ ap u c a li s e l u r Sa a n s pe l v at a q l l a I n i n i a n s u a te d e a n , t ili S a t . at a s p i n s ns e s nd C ha C rc o r is a l Fi t ate sh ll a In rem m se a uc c M t r ins ed o os em (ty s pe N rem ain ec s 39 a 0B Bu ker ins ) ds a c r is Bu pa d ,l Le s c ea af a l e fr re s em m ai ai Zo ns n ne s

th

ep

ns

Open water

Li

D

N u Tr pha a r N pa lut ym n ea S a p h at a l a n C vin ea s ar ia a ex n lb s p at a a ., ns C bi ar ca e R x rp um s el p la R ex ., te an s t r i c u A p n p. a rp ia c u el N c e lu la ec a s te ke e re p en Zo ra c s r is ne -t y pa pe ,l ea fr em ai

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Ecol. indet.

IV

III

II

I

Figure I.13 part 2 Hardinxveld-Giessendam De Bruin DB, square 24, macroremains diagram from samples with a volume of 0.15 litres, + = few (after Bakels et al. 2001, 393). Part of the sample column was not analysed.

20

640

IV

680

700

II I

740

II

780

I

820

Analyst: L. van Beurden, 1999

Figure I.14 Hardinxveld-Giessendam De Bruin DB, square 24, macroremains diagram from the pollen samples with a volume of 1 cm3, + = few (1-10), ++ = some tens (10-49), +++ = many tens (50-99) (after Bakels et al. 2001, 377). See figure I.13 for the lithology key.

43

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

category

C

Copyright © 2010. Leiden University Press. All rights reserved.

taxon Artemisia sp. Chenopodiaceae Chenopodium album Fallopia convolvulus Galeopsis-type Galium aparine + Moehringia trinervia Persicaria lapathifolia Persicaria maculosa Poa annua Ranunculus acris-type Rumex acetosa-group Silene sp. Solanum nigrum Stellaria media urtica dioica urtica dioica-type C = carbonised macroremains w = waterlogged macroremains P = pollen + = present - = not present

w

P

+ + + + + + + + + + + + -

+ + + + +

I.6.3.2 Crop plants and disturbance indicators Cultivated plants were absent in all phases despite the active search for crop plants by sieving 50 extra samples on a 1 mm sieve. On the one hand, the absence of crop plants at De Bruin corresponds with the absence of crop plants at contemporaneous Dutch wetland sites and the early phases of Brandwijk-Kerkhof. On the other hand it is remarkable that the introduction of pottery and domestic animals at Hardinxveld-Giessendam did not result in the introduction of crop plants. Table I.9 shows the taxa of which the ecology is comparable with that of arable weeds11 (based on Bakels et al. 2001; verbruggen and verpoorte 1998). The interpretation of these taxa is the same as for similar taxa found at Polderweg (see discussion above). The single species found in a carbonised state is Galium aparine.

I.6.3.3 Carbonised macroremains of non-cultivated plants Non-cultivated plants of which macroremains were present in a carbonised state are Tilia platyphyllos, Quercus sp., Corylus avellana, Cornus sanguinea, Trapa natans (all five all phases), Crataegus sp. (phases 2 and 3), Malus sp. (phase 2 or 3), Rosaceae (phase 2), Acer campestre (phase 2), Galium aparine (phase 2), Ranunculus ficaria (phase 3), Schoenoplectus lacustris/tabernaemontani (phase 2), S. lacustris (phase 2) and Poaceae (small seeds, phase 3) (Bakels et al. 2001). Macroremains of Tilia sp., Table I.9 Hardinxveld-Giessendam De Bruin, Corylus avellana, Galium aparine, Ranunculus ficaria, taxa of which the ecology is comparable to that of arable weeds (after Bakels et al. 2001). Schoenoplectus lacustris and Trapa natans were found in anthropogenic contexts (pits, deposition pits or surface hearths) or in concentrations of charcoal. In addition, the analysis of a concentration of charcoal additionally resulted in a find of cf. Quercus sp. All taxa found in a carbonised state are known as food plants, except for Acer sp., Tilia sp., Galium aparine and Schoenoplectus sp. (though see chapter 9 for a discussion on the function of Galium aparine). Taxa that have edible parts and that were present at De Bruin but that were not found in a carbonised state are Rubus caesius, Ribes sp., Nymphaea alba, Nuphar lutea and a large variety of herbs and grasses of which the seeds and fruits, leaves and tubers may have been consumed as well (see chapter 9). The carbonised state of many taxa indicates occupation during the autumn (September /October) for all phases. The carbonised remains of Tilia sp. indicate a somewhat earlier occupation for all phases (July and/or August), while Galium aparine indicates summer and/or autumn occupation (July to September) for phase 2. winter occupation cannot be excluded on basis of botanical remains. The charcoal assemblage indicates occupation during spring for phase 2. 11 The selection of taxa is explained in chapter 10.

44

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

I.6.4 Wood and charcoal analysIs The discussion of wood and charcoal remains is based on Bakels et al. (2001), Hänninen and vermeeren (1999) and Louwe Kooijmans, Hänninen and vermeeren (2001). wood remains were categorised as unworked and possibly worked wood remains12 or as presumably worked remains and artefacts.13 The assemblage of unworked wood remains consists of trunks, branches and twigs (N = 188). The identifications are shown in table I.10. In this assemblage, Alnus sp. is the dominant taxon, followed by Fraxinus sp. and Ulmus sp. Other identified taxa are Cornus sp., Quercus sp., Viburnum opulus, Salix sp., Corylus avellana, Betula sp., Euonymus europaeus, Tilia sp., Pomoideae and cf. Rhamnus frangula. The assemblage is diverse and contains a diverse array of shrubs. Tilia sp. is absent in phase 3, in contrast to earlier phases. N

Copyright © 2010. Leiden University Press. All rights reserved.

phase

%

1

2

3

total

1

2

3

total

taxon Alnus sp. cf. Alnus sp. Betula sp. cf. Betula sp. Cornus sp. Corylus avellana cf. Corylus avellana Euonymus europaeus Fraxinus excelsior Pomoideae Quercus sp. cf. Quercus sp. cf. Rhamnus frangula Salix sp. cf. Salix sp. Tilia sp. cf. Tilia sp. ulmus sp. viburnum sp. Indet.

19 3 1 3 3 1 7 2 1 2 3 6

22 1 6 4 3 1 1 5 1 7 1 1 2 7 8

22 5 1 1 8 2 6 7 10 5

63 1 3 6 10 7 1 2 16 1 10 7 1 9 1 2 2 17 10 19

37 6 2 6 6 2 14 4 2 4 6 12

31 1 9 6 4 1 1 7 1 10 1 1 3 10 11

33 7 1 1 12 3 9 10 15 7

34 1 2 3 5 4 1 1 9 1 5 4 1 5 1 1 1 9 5 10

total

51

70

67

188

- = not present Table I.10 Hardinxveld-Giessendam De Bruin, unworked wood for each phase, twigs included (after Bakels et al. 2001). 12 The possibly worked wood remains comprised finds without indications of working, for which the original orientation of the wood in the trunk indicated that it may concern worked wood. 13 The presumably worked wood remains comprised finds that did not show traces of working but that were nevertheless distinguished due to non-natural characteristics. They were interpreted as semi-finished artefacts and refuse of artefact manufacture.

45

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

The absence of Tilia sp. at the dune can however be rejected since several macroremains of Tilia sp. have been found in the macroremains assemblage of phase 3 (contra Bakels et al. 2001, 411). The charcoal from two charcoal concentrations of each phase was identified (N = 607; 47 to 95% of the charcoal of each concentration). Three of the six concentrations are interpreted as surface hearths, and the other concentrations may represent similar hearths (Louwe Kooijmans and Nokkert 2001, 82). The results of the charcoal analysis are shown in table I.11. Alnus sp., Quercus sp. and Viburnum opulus are present in all phases. Alnus sp. is dominant, and the percentage of this taxon increases in every phase. Other identified taxa are Fraxinus sp., Ulmus sp., Corylus sp., Cornus sp., Pomoideae, Rhamnus cathartica, Viscum album, cf. Betula sp., Rhamnus frangula, and Salix sp. The variety of taxa is greatest in phase 1, which may be explained by the large extent of the dune during phase 1. The variety may alternatively be related to the disturbed context of both concentrations of phase 1 (in contrast to the concentrations of the other two phases). The assemblage contains two taxa that were not identified in the pollen and macroremains assemblage of both De Bruin and Polderweg: cf. Sorbus sp. (though pollen grains of Sorbus sp. might have been identified up to family level as pollen of Rosaceae in the pollen analysis) and Prunus avium.14 The broad spectrum of the charcoal assemblage as well as the importance in individual contexts of taxa that were dominant in the local vegetation

Copyright © 2010. Leiden University Press. All rights reserved.

N

%

phase taxon Alnus sp. cf. Alnus sp. cf. Betula sp. Cornus sp. Corylus avellana cf. Corylus avellana Fraxinus excelsior Pomoideae Prunus avium Quercus sp. Rhamnus cathartica Rhamnus frangula Salix sp. cf. Sorbus sp. ulmus sp. viburnum sp. viscum album Indet.

1

2

3

total

1

2

3

total

42 1 1 15 2 1 2 3 12 11 4 16 8 13

98 2 7 8 107 5 1 8 1 14

167 9 14 4 6 4 3 1 17

307 3 1 16 23 2 121 1 2 12 12 11 5 6 20 19 2 44

32 1 1 11 2 1 2 2 9 8 3 12 6 10

39 1 3 3 43 2 0 3 0 6

74 4 6 2 3 2 1 0 8

51 0 0 3 4 0 20 0 0 2 2 2 1 1 3 3 0 7

total

131

251

225

607

- = not present Table I.11 Hardinxveld-Giessendam De Bruin, charcoal for each phase (Bakels et al. 2001). 14 But see the remarks on Prunus sp. in chapter 7.

46

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

(Fraxinus sp. or Alnus sp., 62 up to 85%) indicates that there was no selection of wood for fuel. During phase 1, people often used fresh or moist wood as a source of fuel. During phase 2, people may have used old and mainly dry wood. During phase 3, people often used dry, dead wood as a source of fuel (Bakels et al. 2001). The results of the analysis of possibly worked wood remains (N = 98) and wooden artefacts (N = 79) are shown in tables I.12 and I.13. The assemblage of wooden artefacts consisted of a variety of artefacts: fragments of a bow, hafts, paddles, pointed roundwood, skewers, posts, planks, planks interpreted as fragments of canoes and a fragment of a fish trap. Taxa that were present in the assemblage of artefacts are Tilia sp., Quercus sp., N phase

%

1

2

3

total

1

2

3

total

taxon Acer sp. Alnus sp. Cornus sp. Corylus avellana Fraxinus excelsior Quercus sp. Tilia sp. ulmus sp.

6 1 1 2 6 1 2

1 36 1 5 6 1 1

6 1 1 12 7 1

1 48 2 3 19 19 2 4

32 5 5 11 32 5 11

2 71 2 10 12 2 2

21 4 4 43 25 4

1 49 2 3 19 19 2 4

total

19

51

28

98

- = not present Table I.12 Hardinxveld-Giessendam De Bruin, possibly worked wood (after Louwe Kooijmans, Hänninen and Vermeeren 2001).

Copyright © 2010. Leiden University Press. All rights reserved.

N

%

interpretation

phase taxon Acer campestre Alnus glutinosa Cornus sp. Corylus avellana Fraxinus excelsior Quercus sp. Tilia sp. ulmus sp.

1

2

3

total

1

2

3

total

3 2 1 5 3 6 1

19 2 1 9 6 -

1 7 3 1 5 4 -

1 29 7 3 19 13 6 1

14 10 5 24 14 29 5

51 5 3 24 16 -

5 33 14 5 24 19 -

1 37 9 4 24 16 8 1

total

21

37

21

79

haft plank, point, skewer fish trap, plank, post, point, skewer point, skewer, wedge beam, haft, paddle, plank, point, post, trunk haft, plank, post, point canoes, plank, other bow

- = not present other = other interpretation of the function Table I.13 Hardinxveld-Giessendam De Bruin, identification and interpretation of wooden artefacts for each phase (after Louwe Kooijmans, Hänninen and Vermeeren 2001).

47

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Acer sp., Fraxinus sp., Ulmus sp., Corylus sp., Cornus sp. and Alnus sp. The dominant taxa are Alnus sp., Fraxinus sp. and Quercus sp., which corresponds with their dominance in the assemblage of unworked wood except for the importance of Quercus sp. instead of Ulmus sp., suggesting that people did not apply selective use based on the quality of the wood and the function of artefacts. The choice for certain taxa for specific artefacts nevertheless indicates such selective use (Ulmus sp. for a bow, Tilia sp. for canoes, Alnus glutinosa and Quercus sp. for planks, Cornus sanguinea for the fish trap and Fraxinus excelsior for paddles and spears). Some of the pits that are considered as deposition pits also contained wood. One pit contained two branches of Fraxinus excelsior (length: 24 and 26 cm) and a root of Euonymus europaeus (length: 25 cm). Another pit contained a block of Ulmus wood of 20 x 20 cm and a branch of Fraxinus excelsior (length: 20 cm). The finds did not show indications of wood working. Intentional deposition is nevertheless assumed for the content of both pits due to the location and contexts of the remains, because of the equal length of the sticks in the first-mentioned pit, and since it was excluded that the block of wood found in the second pit represented a post. I.6.5 other sources At De Bruin four species of fungi were found (see table 2.7). Two of the identified species prefer old or dead wood as a growing medium. The analysis of macroremains demonstrated the presence of the moss species Neckera crispa, indicative of moist and dense woodland. I.7

HardInxveld -GIessendam PolderweG and de bruIn: comParIson of tHe arcHaeobotanIcal results

Copyright © 2010. Leiden University Press. All rights reserved.

The distance between the sites Polderweg and De Bruin is only 1 km. Polderweg was occupied between 5500 and 5000 BC while De Bruin was occupied between 5500 and 4500 BC. At both sites, occupation started at approximately the same period and ended when sediments covered the dune. In view of this synchronicity and the close proximity both dunes are considered to have been used complementarily by one local community in the same territory in the early phases (Louwe Kooijmans 2003). The occupation level was estimated for each occupation phase of each site based on the concentrations of archaeological finds. According to the resulting model, Polderweg was occupied the most intensively and De Bruin was used only complementary in phase 1. After 5300 BC, the main occupation shifted to De Bruin while Polderweg was used complementarily. After 5000 BC, occupation continued at De Bruin, while the Polderweg dune had become submerged as a result of the gradual water-level rise. There is however no explanation why people selected the dunes Polderweg and De Bruin and did not occupy the other, larger, dunes in the near surroundings. The reconstruction of the palaeogeography indicates that the presence/absence of crevasse channels and the distance to these channels alone cannot explain this pattern of settlement choice. I.7.1 r econstructIon of the natural vegetatIon The archaeobotanical information of Polderweg and De Bruin can be compared in detail since the distance between both sites is rather small and since the palaeogeography, the landscape and the natural vegetation of both sites are very similar. The pollen diagram of Polderweg partly reflects an earlier period than the pollen diagram of De Bruin, but there is an overlap in chronology between the pollen diagrams. Although the natural vegetation at both dunes is generally the same, there are indications of subtle differences. At both sites, the dryland vegetation consisted of deciduous woodland vegetation consisting of Tilia sp., Quercus sp., Ulmus sp., Fraxinus excelsior, Corylus avellana, Rhamnus cathartica, Viburnum opulus, Cornus sanguinea and possibly Humulus lupulus (this last species might have grown in the alder vegetation as well). The characteristics of the wood and charcoal remains, the fungi and the mosses indicate that the woodland was generally moist, dense and of considerable age. At De Bruin, Crataegus monogyna was probably present,

48

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

and Betula sp. and Ribes sp. possibly as well, while this is unlikely for Polderweg15 (Betula sp.: contra Bakels and van Beurden 2001, 355). The presence of the following taxa in the local and extra-local vegetation is uncertain: Populus sp. (Polderweg), Ilex sp. (De Bruin), Sorbus aucuparia (De Bruin: charcoal identifications), and Acer campestre (found at both sites). At both sites, there are indications that Euonymus europaeus¸ Prunus avium, Prunus padus, Prunus spinosa, and Malus sylvestris were not part of the local vegetation.16 The natural vegetation at the dunes in the central river area nevertheless forms the natural habitat of these taxa so they may have been present before occupation, or at other dunes. The wetland vegetation at both sites consisted of Alnus carr vegetation, forb vegetation, marsh vegetation and lake vegetation (both submerged and floating water plants). The wetland environment was very eutrophic with only minor mesotrophic patches. The alder carr was generally dense and old, though it might have been less stable than the dryland vegetation due to dynamic conditions in the wetland environment. The dune gradually submerged due to the gradual rise of the water level, resulting in a decrease in the dryland woodland vegetation and an increase in wetland vegetation though time. I.7.2 huMan IMPact based on Pollen dIagraMs It is not possible to accurately recognise signals of human activity on the vegetation in the pollen diagrams of Polderweg and De Bruin. The stable AP/NAP ratio indicates that the dryland vegetation does not show important changes that could have resulted from human impact. The percentage of dryland herbs remains very low and stable, and the number of anthropogenic indicators as defined by Behre (1981) is small. There are no indications of the recovery of the vegetation after most of the occupation phases. It is possible to observe an increase in the secondary shrub vegetation during occupation (Viburnum opulus, Rhamnus cathartica, Cornus sanguineatype and Sambucus nigra-type), but the increase is mostly gradual and not supported by other anthropogenic indicators. It is furthermore not possible to distinguish whether the changes in vegetation are caused by human activity or by changes in natural conditions, such as tree falls, activity of wild animals or changes in the water table or water activity. Pollen grains of crop plants are absent at both sites. It must therefore be concluded that the strength of human impact was very restricted. Large-scale deforestation did not occur. The scarce evidence of human impact in the pollen diagrams appears to be in contrast to the indications of gathering of plant food and wood. The only exception to the possibility of recognising anthropogenic influence on the vegetation in pollen diagrams is a major change in the vegetation 7.40 m -NAP in the diagram of De Bruin (Bakels et al. 2001) corresponding with phase 2. The change is characterised by peaks of especially wetland taxa that indicate the presence of light, nutrients and disturbance (Poaceae, Humulus lupulus, Urtica dioica-type, Lythrum salicaria). There are three arguments that support that it concerns evidence of human impact: 1) the strength of the signal, 2) the taxa that show peaks are taxa that are indicative of human impact at other sites as well (see for example appendix II and III) and 3) the vegetation recovered afterwards. The fact that it concerns mainly wetland taxa indicates that the anthropogenic influence affected mainly the local vegetation, and that it concerned small-scale disturbance.

15 Finds of Betula sp. at Polderweg comprise a fruit and a single charcoal identification both dating to phase 0 only. The presence of Crataegus monogyna at De Bruin is indicated by the presence of pollen of Rosaceae and by the number of finds of macroremains that is relatively high compared with the number of finds at Polderweg (Polderweg N = 4; Bakels and Van Beurden 2001; Jongste and Verbruggen 1998). See also the remarks on Prunus avium in chapter 7. 16 At Polderweg, the number of macroremains of Malus sylvestris is cf. 1, and the number of wood and charcoal remains of Pomoideae is small as well (N = 4; Bakels and Van Beurden 2001). At De Bruin, only a single seed of Malus sylvestris was found. The presence of the attested finds can be explained by gathering of fruits for consumption and selective gathering of wood for fuel.

49

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

It can be questioned whether occupation pressure played a role in the evidence of human impact on the vegetation. The density of archaeological finds at Hardinxveld was highest at the site Polderweg during phase 1 (Louwe Kooijmans 2001b). The pollen diagram of Polderweg does however not show any clear signals of anthropogenic influence on the vegetation during this phase, despite the limited distance between the dry surface of the dune and the sample location. During later phases, occupation is concentrated at De Bruin. The pollen diagram of De Bruin DB shows a very local signal related to phase 2 but no signal related to phase 3. The absence of a signal related to phase 3 is unexpected since the extent of the refuse layer of phase 3 is only a little bit smaller than that of phase 2, while the sample location was still located in the refuse layer. It can therefore be tentatively concluded that it is not possible to reconstruct a relationship between occupation intensity and anthropogenic influence on the vegetation at Hardinxveld. The absence of accurate signals of human impact on the natural vegetation is further discussed in paragraph 2.8.3. I.7.3 croP Plants and Weeds At both sites, crop plants are absent, despite an extensive sampling program and active search for cereal grains (discussed above). Archaeological finds that indicate cultivation of crop plants, such as sickle blades, sickle gloss or quern stones are absent as well. Although domestic plants are absent, it cannot be excluded that people did not know about crop plants as the result of contact with fully agricultural communities. Furthermore, the people who visited the Hardinxveld-Giessendam sites spent part of the year elsewhere, presumably the southern sand region. It is unknown when domestic plants where introduced there (see chapter 2.8.1). Both sites show the presence of various taxa that would be identified as potential arable weeds if crop plants were present (see tables I.3 and I.9). Most macroremains were found in a waterlogged state only, except for fruits of Galium aparine and small fruits of Poaceae. The waterlogged state of most taxa does not provide evidence of a direct relationship with anthropogenic activities.

Copyright © 2010. Leiden University Press. All rights reserved.

I.7.4 carbonIsed MacroreMaIns Most taxa found as macroremains in a carbonised state at Polderweg were present in a carbonised state at De Bruin as well: Malus sylvestris, Corylus avellana, Ranunculus ficaria and Trapa natans. These taxa may all represent food plants, and the similarity of the finds at the two sites therefore suggests similarity of plant food. At De Bruin, however, macroremains of seven more potential food sources were found in a carbonised state (5 taxa at Polderweg versus at least 13 food plant taxa at De Bruin). This difference can probably be partially explained by the higher number of samples of macroremains analysed at De Bruin (including the samples sieved on a coarse mesh width). Seasonality may have played a role as well, since there are scarce indications of summer occupation at De Bruin only. Other factors such as deposition processes and taphonomical processes may also have played a role. Overall, the evidence of plant consumption in the macroremains assemblages of Hardinxveld is small in view of the variety of potential food plants present at both sites (further discussed in paragraph 2.8.3).

50

APPENDIx I - ARCHAEOBOTANy OF HARDINxvELD-GIESSENDAM POLDERwEG AND DE BRuIN, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

I.7.5 Wood and charcoal The assemblages of wood and charcoal indicate exploitation of both dryland and wetland vegetation in the near surroundings of the dunes. At both sites, the variety of taxa is high (N = 17, N = 18). Most wood could have been collected at the dunes themselves. Alnus glutinosa, Quercus sp. and Fraxinus excelsior are the dominant taxa in the assemblage of both unworked and charcoal remains. In addition to local resources, people also used non-local retrieved resources. Both at Polderweg and De Bruin, remains of Acer campestre and Euonymus europaeus were only present in the assemblages of worked wood, possibly worked wood or as whole trunks and branches, and not in the assemblages of unworked wood and charcoal, which suggests that they were brought in from elsewhere. However, three carbonised macroremains of Acer sp. were found at De Bruin. This suggests that local presence cannot be excluded, although the macroremains may also have been transported to the site on a branch. Characteristics of the charcoal indicate the use of both old wood and moist, possibly fresh wood at both sites. The variety in the assemblage of wood and charcoal of both sites indicates that people usually did not select wood taxa for fuel. However, the analysis of concentrations of charcoal at Polderweg indicates some possible exceptions: one concentration of charcoal contained a high percentage of Corylus avellana and another concentration contained a high percentage of Euonymus sp. At De Bruin, the selection of Euonymus europaeus is also indicated by the find of worked wood remains of Ulmus sp., Fraxinus excelsior and Euonymus europaeus in pits interpreted as deposition pits. The types of artefacts found at the sites are very similar and the assemblages of both sites demonstrate that people selectively chose wood taxa because of their quality for the manufacture of specific artefacts continuously through all phases.

51

Copyright © 2010. Leiden University Press. All rights reserved.

Copyright © 2010. Leiden University Press. All rights reserved.

Appendix II. Archaeobotany of Brandwijk-Kerkhof, the Netherlands II.1 IntroductIon The site Brandwijk-Kerkhof, was partially excavated in 1991 by the Institute of Prehistory Leiden (now the Faculty of Archaeology) under the direction of A.L. van Gijn and M. Verbruggen, in the context of an investigation of the Neolithic occupation at Late Glacial inland dunes. Brandwijk-Kerkhof is a small dune (0.5 m -NAP) in the Dutch Rhine/Meuse river area, located in the central river area (coordinates 114.163/433.735, see fig. II.1). The site, its cultural setting, geology and the general development of the regional landcape are presented in Out (2008a). It can be added that open water was present south of the dune during all phases and that in addition to the large dune Brandwijk-Donk, a third smaller dune is present in the near surroundings of Brandwijk-Kerkhof. The excavation trench (3 x 15 metres, down to 6.5 m -NAP) was situated on the southern slope of the dune where occupation was concentrated. Several refuse layers (fossil anthropogenic horizons) were distinguished in the excavation trench (see table II.1 and fig. II.2). Layers 50 and 60 are the most important in extent and archaeological remains, while the other layers are of minor importance.17 In the text below, the abbreviation L will be used to indicate a layer. The site was occupied periodically from c. 4600 to 3550 BC. The age of the dated layers is shown in table II.2 (see also fig.2.5). The dates are reconstructed by conventional 14 C dating and by interpolation of ground water levels of other sites in the region (see Verbruggen 1992). Swifterbant pottery was found in all layers at the site. The flint assemblage indicated contact with southern regions. Layer 30 contained a sherd with characteristics of pottery of the Blicquy group; sherds similar to layer sediment Blicquy pottery were also found at Hardinxveld80 peaty clay Giessendam De Bruin (Raemaekers 1999, 2001, 142-147). Fragments of flint axes in the base and top 70 very clayey peat; find layer of L50 furthermore indicate incorporation of the 60 amorphous black peat; find layer Michelsberg material culture. There were no structures 56 peat with flat positioned fragments recognised except for a series of pointed posts, dating to L50 or later. The bone assemblage was dominated of wood; possible find layer by bones of beaver, otter, deer and wild boar. Bones of 55 amorphous peat domestic animals (one bone of goat/sheep and one bone 50 sandy peat; find layer of possibly domestic pig) were present from the initial occupation (L30) onwards but only formed a minority 45 peat layer in layer 40 characterised of the assemblage. It is only during the upper part of by small lenses of sand; find layer L50 that bones of cattle are present and bones of pig/ 40 peat wild boar can be identified as domestic pig bone with 30 sandy peat; find layer certainty (this may be related to the scarcity of material in the lower part of L50). Preliminary results indicate 20 slightly clayey peat that the site was used as a temporary hunting/fishing 15 top of the dune camp. There are indications of summer and winter 10 clean dune sand occupation, while year round occupation is not excluded (Van Gijn and Verbruggen 1992; Raemaekers 1999). Table II.1 Brandwijk-Kerkhof, (based on Verbruggen in prep.).

stratigraphy

17 Of the minor find layers, layer 45 did not contain any sherds in contrast to layers 30 and 70.

53

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

coastal barriers

salt marshes

tidal flats

pleistocene coversand

river deposits

inland dunes

fen peat

Brandwijk - Kerkhof

Copyright © 2010. Leiden University Press. All rights reserved.

upland peat bogs

0

50km

Figure II.1 Brandwijk-Kerkhof, the Netherlands, location plotted on a palaeogeographical map (c. 4200 BC, NITG)

54

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

m - NAP

m - NAP

L70 L60 4.50

-

- 4.50

L56 L55 L50

L40

L45

L15 5.50

-

- 5.50 L40 peat

L30

refuse layer

L20

dune sand

0

1 m

X

Y

Figure II.2 Brandwijk-Kerkhof, eastern section of the excavation trench, the layers (L) (based on Verbruggen in prep.). The letters X and Y correspond with figure II.3. The vertical scale is by estimation. The arrow indicates the expected location of the sample boxes of L45 and L50.

Copyright © 2010. Leiden University Press. All rights reserved.

layer

age (yrs cal BC)

surface (m2)

70

3760-3630

60

3940-3820

1600

50 top

4030-3940

1500

50 base

4220-4100

1500

45

4470-4370

30

4610-4550

attachment point (m -NAP) 3.45 4.00 4.60 5.30

200

5.75

Table II.2 Brandwijk-Kerkhof, the age, extent and attachment point of the refuse layers (Verbruggen in prep.). The dates are reconstructed by conventional 14C dating and by interpolation of ground water levels of other sites in the region (see Verbruggen 1992). The attachment point is the height where the peat is attached to the sand of the dune, influenced by the ground water table.

55

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

See Van Gijn and Verbruggen (1992), Out (2008a) and Raemaekers (1999) for more information.18 Please note that the final site report is forthcoming and that all non-archaeobotanical data represent preliminary results and interpretations. In order to study the developed stage of neolithisation at Brandwijk-Kerkhof, data on botanical remains will be presented here. Questions considered here are: • • • • •

what did the natural vegetation look like? what was the influence of occupation on the vegetation? Did people use plants that grew in the near surroundings as a food source and if so, which plants did they use? which crop plants were present? Do the botanical remains give information on seasonality or site function?

Copyright © 2010. Leiden University Press. All rights reserved.

The investigations included the analysis of pollen, macrofossils and wood and charcoal remains. The results of the analysis of pollen and macroremains of four cores taken at the northern side of the dune have been published separately, together with preliminary results on the crop plants of the excavation trench and a discussion on local crop cultivation (Out 2008a; see also paragraph II.4 for a short summary of the results of the cores). Here the complete data on macroremains, wood and charcoal collected in the excavation trench will be presented. The questions of both publications are roughly the same, but the emphasis here is put on the new results. The results on wood and charcoal analysis were published in Hänninen and Vermeeren (1998). II.2 MaterIals and Methods The macroremains samples were selected in five ways. The layers investigated are the layers 30, 45, 50 and 60. Firstly, macroremains samples were collected from the pollen cores; these results are already published and are not discussed here. Secondly, two sample boxes were collected from the eastern section of the excavation trench; their precise location is unfortunately unknown. Layer 45 (archaeological importance unclear) was sampled with a box of 20 x 10 x 15 cm next to the sample box of layer 50. Layer 50 was sampled with a box of 50 x 10 x 15 cm at the eastern section of the excavation pit. The top and bottom of this box also contained small quantities of sediment from the layers 40 and 70.19 From the boxes samples with a volume of 0.15 litres (1 x 10 x 15 cm) were selected. For L45, 15 samples were selected (interval 1 cm). For L50, 11 samples were selected (interval c. 5 cm). The sediment of both boxes consisted mainly of peaty sediment mixed with sand, but the middle part of the box containing layer 50 contained minor quantities of clay while the upper part consisted of clay completely. The results of the sample box of layer 50 are partly based on multiplications since only fractions were counted of the residue from the fine mesh widths. Thirdly, lab samples were collected during the excavation from the layers L30, L50 and L60. These samples have a volume of 1 litre and were sieved on several mesh widths of which the finest was 0.25 mm. The presented results of these samples are partly based on multiplications of true numbers since only fractions were counted of the residues from the fine mesh widths. Fourthly, botanical remains from the layers L30, L50 and L60 were collected from the sieves that were used to sort material during the excavation (mesh width 1.5 mm). The fifth category of macroremains samples consists of finds that were picked by hand from layer L30, L50 and L60 during the excavation. Figure II.3 shows the excavation trench and the numbers of the excavated squares, corresponding with the sample numbers. Samples of layer 50 are not assigned to either the base or top of layer 50 since these sublayers were distinguished only after excavation. 18 More information is also to be published in Verbruggen (in prep.). 19 The assignment of samples to layers is based on a drawing by W.J. Kuijper.

56

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

31 16

1

32 17

2

33 18

3

34 19

4

35 20

5

36 21

6

37 22

7

38 23

8

39 24

9

A

C

B

D

X

40 25 10 41 26 11 42 27 12 43 28 13 X

0

20 m

1m

44 29 14 45 30 15

Y

0

1 m

Y extent of the dune at the time of L50 extent of L50 eastern section

Copyright © 2010. Leiden University Press. All rights reserved.

Figure II.3 Brandwijk-Kerkhof, location of the excavation trench and the plan of the excavation (based on Raemaekers 1999 and Verbruggen in prep.). The numbers in the figure represent square numbers that correspond with the sample numbers.

w.j. Kuijper and D. Paetzold analysed most of the macroremains in 1991, while w.j. Kuijper and w.A. Out identified the macroremains of some samples of layer 45 in 2006. The diagrams show absolute numbers of macroremains.20 The plant taxa are separated into ecological groups, based on combinations of modern plant community classes (Schaminée et al. 1995-1999) and on the interpretation of the vegetation. Table II.3 shows the ecological groups that are distinguished. This classification is meant only as a method to analyse changes in the vegetation, since the distinction between some groups is not sharp and since some taxa are characteristic of more than one plant community. Prehistoric plant communities may furthermore possibly have differed from modern plant communities and modern ecology can only function as a comparison. wood and charcoal were collected by hand during excavation and from sieve residues samples (mesh width 2 mm). The investigated charcoal represents two samples of the layers 30, 50 and 60 that were all collected in different squares. D. Paetzold, C.e. Vermeeren and K. Hänninen identified the wood remains, while K. Hänninen identified the charcoal remains. In 2004 C.e. Vermeeren identified a knot of plant material with help of Schweingruber (1978). 20 Epilobium hirsutum-type represents E. hirsutum, E. montanum, E. parviflorum, E. obscurum, E. roseum and E. tetragonum. Galeopsis bifida-type represents G. bifida, G. speciosa and G. tetrahit. Juncus effusus-type represents 14 different species (Körber-Grohne 1964). Ranunculus aquatilis-type represents Ranunculus subgenus Batrachium. Ranunculus repens-type represents R. acris/lingua/repens but it probably refers to R. repens here. Veronica beccabunga-type refers to V. anagallisaquatica, V. beccabunga and V. catenata.

57

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

ecological groups 1) Taxa of woodland and woodland edges of dry terrain. 2) Dryland ruderals and taxa that indicate recent disturbance of dry to slightly moist sediment. 3) Crop plants. 4) wetland trees and shrubs. 5) wetland herbs and spore plants of marsh and forb vegetation. 6) wetland pioneers and wetland herbs that indicate disturbance. 7) Taxa indicative of open water. 8) Taxa that are not associated with specific ecologic conditions. 9) Varia, including archaeological remains, macroscopic plant remains such as bark, leaves and moss, animal remains and sand.

Copyright © 2010. Leiden University Press. All rights reserved.

Table II.3 Brandwijk-Kerkhof, ecological groups used for the classification of macroremains.

II.3

r esults

II.3.1

MacroreMaIns analysIs

II.3.1.1 Layer 30 (4610-4550 BC) Table II.4 shows the results of eight samples collected from layer 30. Figure II.4 shows the number of macroremains and taxa of the lab sample for each ecological group. The macroremains of layer 30 were collected in only three adjacent excavation squares, since the refuse layer was only present in a few squares. The small number of samples and the relatively large number of handpicked samples restrict the representativity of the results. The assemblage consists of waterlogged and carbonised taxa (carbonised remains are further discussed below). Crop plants are absent in this layer, despite the presence of pottery and a few bones of domestic animals. All other ecological groups are represented by some taxa. Most plant species indicate moderate to very eutrophic conditions. Macroremains of trees, shrubs and herbs of woodland and woodland edges of dry terrain (group 1) are well represented, especially Cornus sanguinea, Corylus avellana and Crataegus monogyna that are probably overrepresented due to handpicked sampling. The sieve residue samples indicate that alder vegetation, marsh vegetation and open water were present nearby. The scarce taxa from groups 2 and 6, like Chenopodium album and Persicaria species, are characteristic of disturbance of the vegetation, probably due to anthropogenic influence.

58

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

sample sample type mesh width volume (litre) taxon Group 1 Quercus sp. Quercus sp., buds Quercus sp., cupulae Quercus sp., juvenile fruits and cupulae Tilia platyphyllos Tilia sp. Malus sylvestris Viburnum opulus Cornus sanguinea Corylus avellana Crataegus monogyna Glechoma hederacea Alliaria petiolata Chelidonium majus Urtica dioica Galium aparine Group 2 Atriplex patula/prostrata Chenopodium album Persicaria lapathifolia Group 4 Alnus glutinosa Alnus glutinosa, buds Alnus glutinosa, male catkins Alnus glutinosa, cones Group 5 Alisma sp. Carex acutiformis Cladium mariscus Schoenoplectus lacustris/ tabernaemontani

12 hp ?

26 C hp ?

26 D hp ?

27 D hp ?

12 sieve 1.5 ?

26D sieve 1.5 ?

27D sieve 1.5 ?

26D lab 0.25 1

5 cf. 4 7 2 29 12 14 -

2 1 -

14 9 2 -

10 6 4 1 11 15 3 -

+ 1 ++ 1 + 6, 2 c 3 ++ 1c

2 1 1c 1 1c

4 1 2c + -

20 2, 1 c 1 ++ -

-

-

-

-

8 2

6 3 cf.

1 -

-

-

-

-

-

+, 1 c + 1 2

+, 2 c -

+ + -

++ + 1

-

-

-

-

1 1 1

-

1 -

-

-

-

-

-

6

-

14, 2 c 4, 2 c

Table II.4 part 1.

59

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

12

26 C

26 D

27 D

12

26D

27D

26D

-

-

-

-

2 4, 1 c 1 cf.

1 -

1 -

-

-

-

-

-

++ +

-

-

-

Group 5 (cont.) Solanum dulcamara Sparganium erectum Stachys palustris Group 6 Persicaria cf. minor Persicaria hydropiper Group 7 Nymphaea alba Potamogeton sp. Ranunculus aquatilis-type Trapa natans Group 8 Carex sp. Fallopia sp./Persicaria sp./Polygonum sp. Ranunculus repens-type Rumex sp. Galeopsis bifida-type

-

-

-

-

1 2 1 -

-

1

-

-

-

-

-

1 4 1 6

1 -

1 -

+ 1 -

Stachys sp.

-

-

-

-

3

-

3

-

hp = handpicked sample

c = carbonised

sieve = sample from sifting residue

x, yc = x macroremains of which y are carbonised

lab = botanical sample

+ = few (1-10) ++ = some tens (10-49)

Table II.4 Brandwijk-Kerkhof, layer 30, macroremains.

- = not present

1 sample, 57 seeds and fruits

Copyright © 2010. Leiden University Press. All rights reserved.

9%

1 sample, 7 taxa

29 %

8

8 1

44 %

47 %

1 57 % 4

4

14 %

a

b

Figure II.4 Brandwijk-Kerkhof, layer 30, macroremains, lab sample, a) number of identifications and b) number of taxa, ordered by ecological group (see table II.3).

60

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

II.3.1.2 Layer 45 (4470-4370 BC) Figure II.5 shows the macroremains diagram of layer 45. The lower part of the diagram probably shows clearance of Alnus glutinosa since fruits of this species are temporarily absent, and especially since Urtica dioica increases, combined with the presence of the wetland herbs Galium palustre, Solanum dulcamara and Berula erecta. The curve of sand, resulting from erosion and downwash, indicates the disturbance and open character of the southern slope with maximal values in the middle of the diagram. At this time alder carr mixed with Cornus sanguinea develops again, indicating that the diagram shows the rather local clearance of alder. In the upper part of the diagram the sedimentation of sand decreases, and Moehringia trinervia becomes an important element in the local vegetation. M. trinervia is a species common in woodland and shrub vegetation, preferring locations where organic material is decomposed at a high rate, for example due to changes in the water table or due to the increased presence of light (weeda et al. 1985). The presence of Tilia sp. in the most upper samples probably demonstrates further recovery of the vegetation from human impact on the higher parts of the slope. In comparison with layer 30, the diversity of marsh taxa (group 5) has increased, while the diversity of dryland woodland taxa has decreased. The changes in the diversity of marsh taxa and dryland woodland taxa can be related to the increasing water level. The only species found in a carbonised state is Corylus avellana, which may be related to restricted human impact during this occupation period and/or at the location of the sampling boxes. Crop plants were not found in this layer despite active searching in the available samples. II.3.1.3 Layer 50 (4220-3940 BC) The macroremains analysis of layer 50 is based on 11 sample box samples, five lab samples, and 20 samples that were handpicked or collected from sieve residues from 11 squares. The number of samples is large since this layer was rich in archaeological material, indicating intensive occupation. The results of all samples except for the box samples are shown in table II.5 (at the end of appendix II). The presentation of the results starts with a discussion of all samples together, followed by a discussion of the sample box samples and lab samples separately. Compared to the earlier layers, the relative importance of the vegetation of woodland and woodland edges of dry terrain has decreased, while there is an increase in the diversity of taxa and number of macroremains of ruderals and disturbance indicators of both dry and moist to wet terrain (groups 2 and 6, especially Chenopodium album, Solanum nigrum, Brassica rapa and Persicaria lapathifolia). Both changes can be related to human activity and to the rise of the ground water level. The presence of remains of dryland taxa including Tilia sp. nevertheless indicates that at least patches of woodland remained present on the slope of the dune, and that anthropogenic influence probably did not lead to large-scale deforestation of the dune. The common presence of Alnus remains indicates the local presence of alder vegetation at the southern slope of the dune. Alnus glutinosa, taxa of forb and marsh vegetation (group 5) and indicators of open water have increased compared with the previous layers, which points to an increase in the ground water level. This must have led to gradual submerging of the dune and a shift of plant communities along the slope of the dune in an upward direction. An important difference with the previous layers is the presence of crop plants Triticum dicoccon, Hordeum vulgare var. nudum and Papaver somniferum ssp. setigerum, further discussed below. Layer 50 provides extensive information on the marsh vegetation and water plants. Most marsh taxa point to stagnant to slowly running water of c. 50 cm depth, although a few species like Apium nodiflorum and Berula erecta tolerate quickly running water. The group of taxa that indicate open water includes both floating and submerged species, as well as rooting and non-rooting species. Nymphaea alba and Nuphar lutea point to a water depth of 1 to 2 metres. Other species, like Ceratophyllum demersum (pollen) and Zachinellia palustris, prefer water with a limited depth that warms in summer.

61

Copyright © 2010. Leiden University Press. All rights reserved.

M e C nt h ar a Si ex aq le s u R ne p., ati um s t r c a R e p. ic a /a a x r p r ve el n c f nun sp. la si .R c u te s C an lu en u s D o c nc re r y o u pe s a c c lus ns nd um re - t y > ge pe pe 0. o n s 25 p h - t m i lu y p e m m C ha (g , sc ra le B rc on o m ro a ) t W e/ l ia oo f is Bu d h r d r e Bu s em ma ai in d ns s M sc os a s les re m ai ns

62

Peat

th

(c m ) y

og

th

ol

ep

50 100 150

oe

hr in gi a G tr le in c C h er he om va Fa lid a l o h Pe l o p n i u e d r ia m e A s i c c o m r ac ln a n a e A u s r i a vo j u s a ln g m lv C us lut i ac ulu a g n u s Eu rex luti osa los a p c n G ato f. a osa al r c u , i i f So um um tif r. o o l B e anu pa c a r m f m r u m lus nna i s al e A la d t r b ca li s u e in tk um Sa ma ere lc a in c m lv sp t a a s in . r a ia na ta ns

M

Ti li cf a s . T p. C ili o a C r y l u s p. o s C r y l u av or s e c f n u av l l a n . s e a U C o s a llan r t r n ng a ic u u , a s in c a di s a e r oi n a bo c a gu ni se in ea d

Li

D

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Upland trees, shrubs and herbs

Ecol. indet.

Wetland taxa

2

4

6

8

10

12

14 20 Varia

Analysts: W.J. Kuijper and W.A. Out

20 40 60

Figure II.5 Brandwijk-Kerkhof, eastern section of the excavation trench, layer 45, macroremains diagram, + = few (1-10), ++ = some tens (10-49).

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

The group of water plants contains thermophilous species, including Trapa natans and Najas minor. T. natans needs a temperature of 20 °C for flowering and 12-15 °C for germination of the fruits (Karg 2006). The remains of T. natans include bristles of spines. These fine bristles usually do not remain preserved after transport or collection, and their presence therefore indicates that T. natans grew in the extra-local vegetation. T. natans is no longer present in The Netherlands while Najas minor is very rare. The marsh taxa Elatine triandra and Elatine hydropiper were probably favoured by high summer temperatures as well (Brinkkemper et al. 2008). II.3.1.4 Sample box samples layer 50 The stratigraphy of the complete layer 50 is somewhat complex. In the first place layer 50 was divided into two sublayers after excavation, the base of layer and the top of layer 50. The sublayers however cannot be recognised in the macroremains diagram and will therefore not be discussed. In the second place small cracks were present in the peat of layer 50, identified by micromorphological analysis and not visible during excavation. The presence of these cracks resulted in mixture of sediment of different occupation periods that together form layer 50. The results of the analysis of the sample box of layer 50 do however not give indications of mixing of layer 50 at the location where the box was sampled since the curves are rather smooth. There are moreover no indications that the peat suffered from intensive oxidation despite the presence of cracks in the peat (pers. comm. Verbruggen 2006). Figure II.6 shows the macroremains diagram of layer 50. The sediment in the sample box consisted of peat (50-40 cm), sandy peat (39-8 cm) with minor quantities of clay in the upper 10 cm of the peat, and clay (7-0 cm). The sample location was c. 4 metres away from the dry surface of the dune during the formation of layer 50. The sample box samples reflect the development of the vegetation on the dune before, during and after occupation of layer 50. The two lower samples of the diagram represent layer 40, a non-anthropogenic layer. This layer nevertheless contains charcoal, bone and pottery remains. These archaeological remains probably represent material from the start of layer 50 or the transition to layer 50. The two samples show the local presence of dry woodland terrain with Tilia sp., Corylus avellana, Cornus sanguinea, Glechoma hederacea and Moehringia trinervia. Marsh vegetation, including alder and herb taxa, is present as well, but the variety of taxa is restricted and water plants are scarce. The remaining part of the diagram represents layer 50 except for the highest sample. The curve of the sand combined with high values of bone, fish and pottery remains indicate that human impact is relatively strong in the middle part of the diagram, resulting in the disturbance of the vegetation and erosion of the dune. Tilia sp. has disappeared, which may indicate clearance or submerging of the woodland of dry terrain at the sampling location. Some other dryland trees and shrubs appear instead: Quercus sp., Malus sylvestris and Viburnum opulus. The finds of these taxa probably indicate the transition of relatively closed woodland into more open shrub vegetation. It can however not be excluded that all macroremains of shrubs have been transported by people to the spot together with resources (e.g. fuel or food). Remains of crop plants are present from 40 cm in layer 50, but only in very small numbers per sample. Ruderals, disturbance indicators and marsh taxa show increasing values and variety. These particularly include the taxa Urtica dioica, Chenopodium album, Persicaria lapathifolia, Brassica rapa, Veronica beccabunga-type, Lythrum salicaria, Elatine hydropiper, Stellaria aquatica/media and Alisma plantago-aquatica. The increase in these taxa in the middle of the diagram is probably related to human impact at the sample location, while the maximum at 7 cm is probably related to a decrease in the human impact that had previously prevented maximal growth of the taxa during occupation. The plants Capsella bursa-pastoris and Polygonum aviculare that are indicative of tread are present as well, representing open patches in the dryland vegetation caused by human impact. By 17 cm the assemblage of woodland of dry terrain taxa already changes, while Alnus glutinosa also increases, suggesting recovery.

63

Copyright © 2010. Leiden University Press. All rights reserved.

64

20 40 20

ap

m

ca

nu

si ra pa Upland herbs and spore plants

ni gr um s Po e l l a l b y Fa g ur o s Pellop num a-p i a r So sic c av ast o a o i Sonch ria nvo cula ris us c f lv re n H c h a . m ul o u u Tr rde s c spe ac s i u r ul o Tr ticu m f. a sa i m v s Tr ticu d ulg per it i m ic o ar Tr cu d c e i m ic c v Tr ticu c oc on ar. i m f. c , n Paticu m dic on car ud m o o , b u Al pav s no cc spi on m, nu er p., co on ke ise ca s s o i n c c , c l et d r b g l m te u h b ut n r n m af as o ni se in if o -t f, e o s e r di y p c a s d a u m u m e, r b c a on ss r b is p. se o ni e d Al se t nu ig d er Al s g um nu lu A l s g t ino nu lu s a A l s g t ino , c nu lu s a ar A l s g t ino , c b on nu lu s a o n i s C s g t ino , c e s ed ar lu s a a t d H am t in o , b k ins um in s u d a Ve ulu e p , b s r o s l r at u d ni up en sc ca u s a b e lus i s l e s cc ab un ga -t y Ly pe th ru m sa li c ar ia

C

la

as th

ep

ol

D th (c m

-N AP T i o gy ) lia T i sp lia ., f Q sp r uit ue . s C rc u or s y C l u s s p. or a C y l u s ve l or a lan n M u s ve l a al s la u a n M s s n g a, al y l ui c a u s ve n e r V i s s a bo bu y t r ni l i se R r n u ve s s d ub m t r i u R s o p s, f ha fr u r. u l G mn t i c o us end le us s oc c h c us ar M o a p oe m th To hri a h ar ti r il ng ed c a c f is j ia t era . T ap r in c e c f o r il o ni er v a . A is c a ia L a lla jap ps r ia on To an pe ica r il a c tio G i s s o m l at a al p i . mu G um ni al a s iu pa U m r r t i a in c a pa e di r in oi e ca , c ar bo ni se d Pt er i d Po i u ly m C p o aq h e di u n o ac i lin p o e a um d i e, , l um sp ea a l o r a ve s bu ng m ia C he Pe n o p rs od ic iu ar m ia f i la c if pa o l t h ium ifo lia

Li

525

So

Br

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Upland trees and shrubs

20

Sandy sediment

Peat

Figure II.6 part 1.

Clay

Upland herbs and spore plants

530

535

540

545

550

555

560

565

570

575 Crop plants

100 200 300 200 400 600

50 100 150

100 200 300

Alder carr/marsh

100 200 300

50 100 150

Copyright © 2010. Leiden University Press. All rights reserved.

100 300 500 50 100 150

Alder carr/marsh Wetland pioneers

20

20 40

N aj a Po s m t a in R mo or an g S c u n c et o ro u n s P o p h l u s p. a u l re P o s p. a r i a p e n a sp s-t Po s p . yp . e , a c Ju c ea a r nc e b o ni J u us se n c ef d f R us us um s u s p R ex . - t y um s pe e p. M x s , pe e p C nt h . r i a n a th C r ex a a s q a C r ex s p . u a t ar s ic B r ex p . a/ , ar G a s s s p. b i c a i , a ve M l e o c ac t r i c r p e ns i s y p e a A p o s o s i s ae r p l l at ia t is bif ella e c e s id t a e p. a - e ty pe

Th Spely p Pe h a g t e r is S t r si c num pa e a l A p lla r ia sp ust ria h ., ri i u R m a yd l e s , a le u q Pe me cf. ua rop ves av ip es t x r i r e C sic c o pe c a er al a n n lit r i g l s r ic a o he mi m e N sp nor r at aj us a . Tr s ,p a m R pa a er an n r in ia un at a a nt hs c u ns lu s N aq y ua C mp ha h til is Z a r a ae -t y a pe Sannicsp. alb a lv he in ll ia ia na pa t a lu ns st r is at in e hy dr op ip

is

ha

sp

.

a/

m

er

ed

ia

m a pl an R um ta Sc e go x -a Euhoe hy qu dr n p at C at o p o l ic ar or le ap a Ly ex iu ct a t m u hu co cf s Sc p . c a la m a u Phirpu s e cut nna cus u if b t P h r a g s sy r o p o r m i n u r i s m P h r a g i t e l va t a e u i s m m s i Ph r a g i te au c u s s , c ar m s Ep r a g i te a u s t r a bo m s l ni C i l o b i te au st r a i s se la iu s s li d Ep d i u m a u t r a s , c l m G ilob mhirs stra is, arb a i u l r o Va liumum aris tumis, hizo nis s e s c l A n e r i p a p. u s - t y t e m m e d pe f b g a l ra ud O e li c n a u st e gm s o r C nana sy f fic e . al t lv in B e t h a h e e st a l r i a s Fi r u l a p a q u i s / P lu a e li Ly pen ere str tic uce c o d c t i s a da S o p ul a nu u a m C lan s e ulm pa ar um u lu S i ex d r o p a r i st re um a u a a pp lc eu la ro am s tif p a ol . / ra iu di m an dr a/ pa ni c.

Al

Ty p

at St ine el tr la ia r ia nd aq ra ua tic

El

El

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Alder carr/marsh

20 40 20

Open water

100 300 500 Ecologically indeterminate

20

Figure II.6 part 2.

65

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Varia

So A s lan u S t te r a m ac c sp V i h ea . o y e S i la s s l s p D e n e p. . ry s s a p. nd C > ha 0. 25 Bo rc o m a ne l m Po / f (g is t ra Fl te r h r m in y em ) W t r re a e o o m m in Bu d r ai ain s e n s Buds ma s in d M s s os c a s le s r L a em ye ai ns r

Ecologically indeterminate

L70

L50

L40 20 Analysts: W.J. Kuijper and D. Paetzold

Figure II.6 Brandwijk-Kerkhof, eastern section of the excavation trench, layer 50, macroremains diagram (partly based on Paetzold unpublished data), + = few (1-10), ++ = some tens (10-49), +++ = many tens (50-99). The results are partly based on multiplications, part 3.

5 samples, 37312 seeds and fruits 4% 19 %

8

11 %

1

7

5 samples, 89 taxa

2

19 %

16 %

4 5%

1% 6

12 % 3%

Copyright © 2010. Leiden University Press. All rights reserved.

1

8

6%

5 54 %

a

2 13 %

7 6

4

3 3% 3%

5 31 %

b

Figure II.7 Brandwijk-Kerkhof, layer 50, macroremains, all lab samples together, a) number of identifications and b) number of taxa, ordered by ecological group (see table II.3). The number of identifications is partly based on multiplications.

66

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

The open water taxa increase in the upper part of the diagram as well, indicating the presence of shallow pools and possibly of open water. The decreased occupation intensity at the locality may therefore be caused by the rise of the water level. The occupation does not end completely, as indicated by carbonised chaff of Triticum sp., some flint remains and a high value of charcoal at 7 cm (though vertical transport and colluviation processes may play a role as well; see Amkreutz et al. 2008). The upper sample represents the clayey layer 70 (3760-3550 BC). The strongly deviating macroremains assemblage suggests that some erosion preceded the deposition of clay. Dryland woodland is replaced by alder carr with herb undergrowth. This sample interestingly contains a carbonised fruit of Galium aparine. A moss species that was found regularly in the sample box samples of layers 45 and 50 is Neckera crispa. This species was probably a common epiphyte on old deciduous trees in moist woodland from 5000 BC until some centuries ago. In the Mesolithic and Neolithic the moss was probably used for domestic purposes. Another moss present at Brandwijk-Kerkhof is Anomodon viticulosus. Comparable with Neckera crispa, this species grows on bark of trees and also on calcareous rocks (but rocks were not naturally present at Brandwijk-Kerkhof). II.3.1.5 Lab samples layer 50 Figure II.7 shows the number of macroremains and taxa from each ecological group of all the lab samples of layer 50. Figure II.8 shows the number of macroremains from each ecological group for each lab sample separately (N = 5). In contrast to the other layers, the number and distribution of lab samples makes it possible to compare the distribution of macroremains of layer 50 on the slope of the dune, although the number of samples is still small. The assemblage of sample 36.45, sampled relatively close to the top of the dune, contains mainly fruits of Urtica dioica, Chenopodium album and Chenopodium ficifolium (groups 1 and 2). The dominance of these very strong, resistant fruits indicates that selective corrosion influenced the composition of the sample. Sample 9.69, 2 to 3 metres further away from the top of the dune than the first sample (sample 36.45), shows a strong increase in the diversity of ecological groups. Marsh taxa (group 5) are the dominating ecological group, followed by undisturbed and disturbed dryland vegetation and indicators of open water (groups 1, 2 and 7), while Alnus glutinosa increases somewhat. The more important taxa are Urtica dioica, Chenopodium album, Capsella bursa-pastoris, Lythrum salicaria, Salvinia natans, and Stellaria aquatica/ media, probably representing the transition from dryland to wetland vegetation. Sample 27.7, 5 to 6 metres further away from the top than the first sample, shows again increasing percentages of water plants (group 7). The importance of dryland vegetation (groups 1 and 2) has decreased. The percentage of group 1 is still rather high, but this is misleading since it mainly represents fruits of Urtica dioica, which is not necessary part of vegetation of dry woodland and woodland edges. Group 8 (no ecological meaning) is well represented, suggesting that in this sample this group is associated with wetland vegetation. The samples 13.124 and 43.542 are both located at the same distance from the top of the dune (6 to 7 metres further downwards than the first sample). Plants of fresh water, marshes and riparian areas dominate in both samples. In conclusion, the lab samples from the top of the dune are rich in ruderals and indicators of disturbed dryland terrain as well as marsh taxa, and appear to be strongly influenced by selective corrosion. On the lower part of the dune slope the importance of marsh plants, riparian plants and water plants increases. This pattern can be linked to the level of the ground water level, resulting in a change in vegetation and preservation.

67

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample 36.45, 838 seeds and fruits

39 %

a

2 1 61 %

sample 9.69, 4402 seeds and fruits 10 %

3%

8

7

1

a

22 %

b 42 %

5

b

2 19 % 4

3%

1%

3

c sample 27.7, 9245 seeds and fruits 10 %

1

8

c

d

15 %

18 % 7

2

6%

e

1m

4 5%

0

1 m

5 46 %

sample 43.542, 15656 seeds and fruits 1 2 2% 1% 4 1%

Copyright © 2010. Leiden University Press. All rights reserved.

23 %

d

sample 13.124, 7171 seeds and fruits

2%

3%

8

16 %

7

e

1% 6

5

70 %

7

8

5% 6

11 %

1

5%

2 1 %3 4 12 %

5 47 %

Figure II.8 Brandwijk-Kerkhof, layer 50, macroremains, individual lab samples (a-e), their location and the number of identifications, partly based on multiplications and ordered by ecological group (see table II.3).

68

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

II.3.1.6 Layer 60 (3940-3820 BC) The analysis of layer 60 is based on three lab samples from two squares and sieving samples from seven squares (see table II.6). Figure II.9 shows the number of macroremains and taxa from each ecological group from all the lab samples of layer 60. The number of taxa in layer 60 is limited in comparison with layer 50, which is probably related to the lower number of samples. Urtica dioica, Chenopodium album, and additionally Solanum nigrum and Stellaria aquatica/media are present with numerous macroremains, indicating forb vegetation of disturbed, open, nutritient-rich terrain. Their dominance might partly be the result of selective corrosion. Both cereal species are present that were attested in layer 50, but Papaver somniferum ssp. setigerum is not present anymore. Figure II.10 shows the number of macroremains from each ecological group for each separate lab sample. Sample 21.26 is not representative since it contained few macroremains (the number of macroremains of water plants (group 7) for example is equal between sample 21.26 and sample 7.15). It is not possible to distinguish a gradient along the slope within the samples of layer 60 due to the small number of samples. sample 7.15 20 C 20 D 21 C 22 C 22 D 23 D 24 C 24 D 25 D 7.15

7.54

21.26

sample type

hp

hp

hp

hp

hp

hp

hp

hp

hp

hp

lab

lab

lab

mesh width

-

-

-

-

-

-

-

-

-

-

0.25

0.25

0.25

volume (litre)

?

?

?

?

?

?

?

?

?

?

1

1

1

Cornus sanguinea

-

-

-

25

3

7

-

-

-

1

2

-

-

Corylus avellana

-

3

1

4,

-

2,

3,

2

5

3,

1c

1c

1c

1c

2c

taxon Group 1

1c

1c

Crataegus monogyna

-

-

-

1

-

-

-

1

-

-

-

-

-

Prunus spinosa

-

-

-

1

1

-

-

-

1

-

-

-

-

Urtica dioica

-

-

-

-

-

-

-

-

-

-

2096

1637

++

1c

-

-

-

-

-

-

-

-

-

-

-

-

Brassica rapa

-

-

-

-

-

-

-

-

-

-

12

2

-

Chenopodium album

-

-

-

-

-

-

-

-

-

-

748 255, 1 c +++

Persicaria lapathifolia

-

-

-

-

-

-

-

-

-

-

4

-

-

Solanum nigrum

-

-

-

-

-

-

-

-

-

-

24

46, 2 c

1c

Hordeum vulgare var. nudum

-

-

-

-

-

1c

-

-

-

-

-

-

-

Hordeum vulgare, internodia

-

-

-

-

-

-

-

-

-

-

-

8c

-

-

-

-

-

-

-

-

-

-

-

4c

-

Galium aparine

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2

Group 3

Hordeum vulgare/Triticum sp. 4 c Triticum dicoccon

-

-

-

-

-

-

-

-

-

-

-

1c

-

Triticum sp., glume bases

-

-

-

-

-

-

-

-

-

-

-

6c

-

Table II.6 part 1.

69

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample 7.15 20 C 20 D 21 C 22 C 22 D 23 D 24 C 24 D 25 D 7.15

7.54

21.26

Group 5 Alisma sp.

-

-

-

-

-

-

-

-

-

-

16

-

-

Carex acutiformis

1

-

-

-

-

-

-

-

-

-

-

-

-

eupatorium cannabinum

-

-

-

-

-

-

-

-

-

-

-

1

-

Iris pseudacorus

-

-

-

-

1

-

-

-

-

-

-

-

-

Lycopus europaeus

-

-

-

-

-

-

-

-

-

-

20

-

-

Mentha aquatica/arvensis

-

-

-

-

-

-

-

-

-

-

16

-

-

-

-

-

-

-

-

-

-

-

-

-

1

-

Salvinia natans

-

-

-

-

-

-

-

-

-

-

++

-

++

Trapa natans

-

-

-

-

-

-

-

-

-

-

-

1c

-

Rumex sp.

-

-

-

-

-

-

-

-

-

-

16 c

-

-

Stellaria aquatica/media

-

-

-

-

-

-

-

-

-

-

80

66

-

Schoenoplectus lacustris/ tabernaemontani Group 7

Group 8

Group 9 Cenococcum geophilum, sclerotia hp = handpicked sample lab = botanical sample

c = carbonised x, yc = x macroremains of which y are carbonised

+++ + = few (1-10) ++ = some tens (10-49) +++ = many tens (50-99) - = not present

Table II.6 Brandwijk-Kerkhof, layer 60, macroremains, part 2.

3 samples, 5227 seeds and fruits 7

3 samples, 18 taxa

1%

5

1%

3

1% 3%

11 %

Copyright © 2010. Leiden University Press. All rights reserved.

8

8

11 %

22 % 2

1

17 %

7 2 22 %

1

a

72 %

28 %

5 3

b

11 %

Figure II.9 Brandwijk-Kerkhof, layer 60, macroremains, all lab samples together, a) number of identifications and b) number of taxa, ordered by ecological group (see table II.3). The number of identifications is partly based on multiplications.

70

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample 7.15, 3060 seeds and fruits

sample 21.26, 127 seeds and fruits

sample 7.54, 2040 seeds and fruits

7

3

1%

5

2% 20 %

7

1

2

20 %

60 %

26 %

1% 3%

3%

8

8

15 %

2

1

68 %

2

1 81 %

Copyright © 2010. Leiden University Press. All rights reserved.

Figure II.10 Brandwijk-Kerkhof, layer 60, macroremains, individual lab samples, number of identifications, partly based on multiplications and ordered by ecological group (see table II.3).

II.3.1.7 Crop plants Layer 50 and layer 60, dating to 4220-3820 BC, contained small quantities (c. 100 remains) of Triticum dicoccon and Hordeum vulgare var. nudum. These two cereals are commonly found at early and Middle Neolithic Dutch wetland sites. In the samples of layer 50, containing most cereal remains, T. dicoccon is clearly dominant to H. vulgare. Layer 50 contained waterlogged and carbonised grains, glume bases and internodia of Triticum dicoccon, and waterlogged and carbonised grains and internodia of Hordeum vulgare var. nudum. Layer 60 contained a carbonised grain and carbonised glume bases of Triticum dicoccon, and carbonised grains and internodia of Hordeum vulgare var. nudum. A single grain of Triticum monococcum-type found in the material of layer 50, probably represents a grain of Triticum dicoccon that developed in the top of an ear. Triticum monococcum-type is very scarce in the Dutch wetlands (see chapter 11). The practice of local cereal cultivation is discussed in Out (2008a) and in the discussion below (paragraph II.4.3). It is concluded that crop cultivation was either practised on a small scale or that cereals were imported from elsewhere. Interestingly, the introduction of crop plants coincides with indications of the introduction of cattle and possibly domestic pig (see paragraph II.1), after the earlier presence of sheep/goat. The question raised is whether these patterns are related to each other and whether crop cultivation was more often introduced as part of a mixed farming system. Alternatively, both indications of increased agricultural activity at Brandwijk-Kerkhof could simply be related to increased occupation intensity during the formation of layer 50. The presence of cereals in layers 50 and 60 is in contrast to the absence of cereals in layers 30 and 45. The number of analysed samples from layers 30 and 45 is somewhat small and the sampling strategy too restricted to establish the absence/presence of crop plants, which is due to the restricted thickness and extent of these layers. Layer 30, however, is contemporaneous with occupation at Hardinxveld-Giessendam where an absence of crop plants has been demonstrated (Bakels and Van Beurden 2001; Bakels et al. 2001). The sampled material from layer 45 consists of a monolith, but the analysis of 15 samples (volume 10 x 0.15 litres) sieved on a mesh width of 0.25 mm did not show any remains of crop plants. The transition to the use of crop plants as shown by the results of this analysis therefore seems to reflect the true transition to the adoption of crop cultivation at Brandwijk-Kerkhof. It is argued that the introduction of crop plants at Brandwijk-Kerkhof also represents the introduction of crop plants in the central river area, under the influence of contact with the Michelsberg culture (Out 2008a).

71

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

Layer 50 also contained waterlogged seeds of the crop plant Papaver somniferum ssp. setigerum. The significance of the absence of Papaver somniferum ssp. setigerum in layer 60 can be questioned in view of the selective corrosion mentioned above that affected the macroremains assemblage of layer 60 and the fact that Papaver seeds rarely remain preserved in a carbonised state. The importance and function (crop or weed) of the species at Brandwijk-Kerkhof is unclear; this is problematic for other Neolithic sites in Northwestern europe as well (Bakels 1982). Firstly, poppy may have been part of the occupation waste instead of a plant that grew at the site, since its presence in the sample box samples corresponds with the presence of sand, charcoal and bone remains. Secondly, the scarcity of the species at other early and Middle Neolithic Dutch wetland sites interestingly indicates that it was not a common species. The species was absent at Hardinxveld-Giessendam and the Hazendonk despite detailed archaeobotanical research there. Other finds from Dutch Neolithic wetlands are known from Schokland-P14, Vlaardingen and western Flevoland (see paragraph 11.2.1), but these may all represent younger finds. These results imply that opium poppy may have been imported to BrandwijkKerkhof. In regions south of the central river area, Papaver somniferum ssp. setigerum is known from the LBK (Bakels 2000), the Blicquy group (Belgium; Bakels et al. 1992) and the Michelsberg culture (northern France; Bakels 1999), of which the latter is contemporaneous with occupation at Brandwijk-Kerkhof. II.3.1.8 Carbonised macroremains of non-cultivated plants Presence of carbonised remains of edible parts of plants is generally used as an indication of use and/or consumption (see chapter 9). Table II.7 shows the carbonised plant remains of Brandwijk-Kerkhof. Taxa other than crop plants that were found in a carbonised state in more than one layer are Corylus avellana, Galium aparine, Solanum nigrum, Alnus glutinosa, Schoenoplectus lacustris/tabernaemontani and Trapa natans. Layer 50 contains the largest number of taxa found in a carbonised state, correlated with the high number of archaeological remains and botanical samples. Concentrations of carbonised plant remains were absent. Taxa found in a carbonised state at Brandwijk-Kerkhof of which seeds and fruits, roots and/or leaves may have been consumed are Corylus avellana, Quercus sp., Cornus sanguinea, Trapa natans, Schoenoplectus sp., Phragmites australis and Nymphaea alba, and possibly Chenopodium album, Rumex sp. and Poa sp. Indications of consumption are the strongest for C. avellana since this species is present in most samples in a carbonised state, followed by remains of Triticum dicoccon (analysis with and without handpicked samples gives similar results). Nuts (Corylus avellana, Quercus sp. and Trapa natans) and tubers may have functioned as staple food since they have a high energetic value and can be stored after roasting and/or grinding. The carbonised remains of Quercus sp. however comprised only a single fragment of a cupula, which may represent waste from fuel instead of waste from food. Taxa that were only present in a waterlogged state at BrandwijkKerkhof and that also contain edible parts (seeds, fruits and/or roots s.l.) are Malus sylvestris, Prunus spinosa, Crataegus monogyna, Rubus fruticosus, Allium sp., Pteridium aquilinum, Typha sp., Scirpus sp. and Nuphar lutea (the selection of taxa that is presented here as edible is based on suggestions on edible plants from Bakels and Van Beurden 2001; Brinkkemper et al. 1999; Perry 1999). Ranunculus ficaria, another species with edible tubers, was not found in the assemblage, but is nevertheless expected in the environment of Brandwijk-Kerkhof as well, as suggested by finds at Hardinxveld-Giessendam and the Hazendonk. The following taxa are found in a carbonised state at Brandwijk-Kerkhof but are not known as food plants: Alnus glutinosa, Galium aparine, Galium palustre, Lycopus europeus, Solanum nigrum and Sparganium erectum. The presence of such taxa in a carbonised state is comparable with other Dutch wetland sites and is further discussed in chapter 9. Plants may of course have been used in many other ways than for consumption, which may result in the carbonisation of macroremains as well. Use-wear analysis indeed indicates the use of plants in ways other than consumption, such as plant working including working of taxa like grasses or reed (Van Gijn et al. 2001, 190; Van Gijn and Verbruggen 1992).

72

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

layer

L30

L45

L50

L60

L70

Quercus sp., cupulae

-

-

+

-

-

Cornus sanguinea

-

-

+

-

-

Corylus avellana

+

+

+

+

-

Galium aparine

+

-

+

+

+

Solanum nigrum

-

-

+

+

-

Chenopodium album

-

-

-

+

-

Hordeum vulgare var. nudum

-

-

+

+

-

Triticum dicoccon

-

-

+

+

-

Alnus glutinosa

+

-

+

-

-

Alnus glutinosa, cones

-

-

+

-

-

Galium palustre

-

-

+

-

-

Lycopus europaeus

-

-

+

-

-

Phragmites australis

-

-

+

-

-

Phragmites australis, stem fragments

-

-

+

-

-

Schoenoplectus tabernaemontani

-

-

+

-

-

Schoenoplectus lacustris/tabernaemontani

+

-

+

-

-

Sparganium erectum

+

-

-

-

-

Nymphaea alba

-

-

+

-

-

Trapa natans

-

-

+

+

-

Poa sp.

-

-

+

-

-

Rumex sp.

-

-

-

+

-

taxon Group 1

Group 2

Group 3

Group 4

Group 5

Copyright © 2010. Leiden University Press. All rights reserved.

Group 7

Group 8

+ = present

- = not present

Table II.7 Brandwijk-Kerkhof, carbonised macroremains from various layers. The data are based on lab samples, finds collected from the 1.5 mm sieve and handpicked finds.

73

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

II.3.1.9 Arable weeds The analysis of arable weeds could possibly give more information as to whether crop plants were cultivated in the exploitation area of Brandwijk-Kerkhof or whether they were imported from elsewhere (see the discussion in Out 2008a and below). A major problem for the interpretation of potential arable weeds, however, is in distinguishing arable weeds as such. Many of the ruderals and indicators of disturbance found at BrandwijkKerkhof, represented by both pollen and macroremains, indicate ecological conditions that are more or less similar to the ecological conditions of arable weeds: disturbed and open soil, the presence of nutrients and the presence of light. These conditions are created and favoured when anthropogenic disturbance and/or arable farming takes place, both in modern and in prehistoric times. In the river area, however, natural processes also result in these conditions, especially at water edges. The river water is a source of nutrients, while water activity is a source of disturbance. Tree falls and wild animals cause disturbance of the soil as well. As a consequence, ruderals do not necessarily represent arable weeds. Potential arable weeds may therefore represent the natural vegetation, or arable weeds that were part of the weed vegetation of small arable plots, or weed macroremains that were imported to the site together with the crop product, or taxa that were introduced to the site together with the crop but that became part of the synanthropic vegetation despite the absence of arable fields. This complexity makes it difficult to reconstruct the status of potential arable weeds. Carbonised macroremains that are found within concentrations of carbonised crop plants are generally considered as arable weeds, assuming that the crop plants and other taxa became carbonised during the same event of crop processing (Hillman 1981). Concentrations of crop plants were not found at Brandwijk-Kerkhof, however. This classic method can therefore not be used to identify the arable weeds. Instead of the weed analysis as is commonly applied (based on Hillman 1981), it was first investigated which taxa were present in a carbonised state in samples that contain carbonised remains of crop plants. These include the taxa Solanum nigrum, Chenopodium album, Poa sp., Corylus avellana, Cornus sanguinea, Alnus glutinosa, Schoenoplectus lacustris/tabernaemontani and Trapa natans. All these species probably grew in the local and/or extra-local vegetation of Brandwijk-Kerkhof since they were also recovered in a waterlogged state and since their ecology is similar to the ecology of the vegetation present at Brandwijk-Kerkhof. Only the first three species (Solanum nigrum, Chenopodium album, Poa sp.) could represent potential arable weeds, considering their preferred ecological conditions. The other species, and possibly also Chenopodium album and Poa sp., could represent plant food or other use plants (discussed above). As a second approach, a selection of potential arable weeds was made, including the herb taxa of all habitats where one could locate an arable field: woodland and woodland edges of dry terrain, grassland, and open and/or disturbed dry to slightly humid terrain.21 Table II.8 shows the list of potential arable weeds found in the pollen and macroremains assemblages of Brandwijk-Kerkhof, separated by preservation type. The table shows the presence of carbonised and waterlogged macroremains of potential weeds for the site for layers without and with crop plants.

21 The selection of taxa is explained in chapter 10.

74

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

layer crop plants present category

taxon Alliaria petiolata Arctium cf. lappa Artemisia sp. Atriplex patula/prostata Brassica rapa Capsella bursa-pastoris Cardamine pratensis Chaerophyllum bulbosum/temulum Chelidonium majus Chenopodiaceae Chenopodium album Chenopodium ficifolium Fallopia convolvulus Polygonum convolvulus-type Fallopia dumetorum Galeopsis-type Galium aparine Galium sp. Glechoma hederacea Hypericum cf. tetrapterum Lapsana communis Luzula sp. Moehringia trinervia Persicaria lapathifolia Persicaria cf. maculosa Polygonum persicaria-type Plantago lanceolata Polygonum aviculare Silene sp. Solanum nigrum Sonchus asper Sonchus sp. Stellaria media Torilis japonica Torilis sp. Urtica dioica Valeriana officinalis

C

L30 + L45 w

P

+ -

+ + + + + + + + + + + + -

+ + -

C

L50 + L60 + w

P

+ + + -

cf. + + + + + + + + + + + + + + + + + + + + cf. + + + + + + + + + +

+ + + + + -

Table II.8 Brandwijk-Kerkhof, potential arable weeds from the layers without crops plants and the layers with crop plants. C = carbonised macroremains, W = waterlogged macroremains, P = pollen, + = present, - = not present.

75

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

The potential arable weeds that were found in a carbonised state are more likely to represent arable weeds since the carbonised state indicates that they were handled by people. The potential arable weeds found in a carbonised state are S. nigrum, C. album, Poa sp., Galium aparine and Rumex sp. All taxa found in a carbonised state except for Rumex sp. were present in samples that contained carbonised remains of cereals as well, supporting that they may represent arable weeds. Only S. nigrum and G. aparine were present in a carbonised state in more than one sample that contained carbonised cereal remains. The taxa that probably represent arable weeds are not indicative of specific soils and/or locations (cf. Bakels 2000). All potential weeds found in a carbonised state were furthermore also found in a waterlogged state at Brandwijk-Kerkhof, which suggests that they were part of the local vegetation, and were already present in the region before the introduction of crop plants (only Poa sp. is found only once, but Poaceae were common in the region; see Bakels et al. 2001 and Bakels and Van Beurden 2001). The possibility that these taxa represent field weeds does therefore not argue against local crop cultivation or non-local crop cultivation. Comparison of the potential arable weeds present in the layers without crop plants (layers 30 and 45) and layers with crop plants (layers 50 and 60) shows that the latter layers contain 19 taxa that were not present before (see table II.8). Some of these taxa are the oldest finds from the central river area: Brassica rapa, Cardamine pratensis, Chaerophyllum bulbosum/temulum, Hypericum cf. tetrapterum, Lapsana communis, Luzula sp. and Polygonum aviculare. The first presence of some of these taxa may be the result of the introduction of crop plants, and it is therefore possible that some of the taxa represent arable weeds. The evidence is weak, however, since the taxa are only found in a waterlogged state, and since alternative explanations such as increased human impact could also explain the presence of some taxa. Therefore, the introduction of these taxa at the wetlands remains a subject for future research (see also chapter 10). II.3.2 Wood and charcoal analysIs The analysis of wood and charcoal remains is based on Hänninen and Vermeeren (1998). Table II.9 shows the results of the analysis of wood remains. The analysed wood remains (N = 71) originate from seven different layers.22 Only the number of remains of layer 50 approaches a representative sample size. Alnus sp., Quercus sp. and Corylus sp. dominate the wood assemblage of all layers together. The finds include pointed wood remains interpreted as posts, worked wood remains (apart from the pointed remains) and unworked wood remains. Twelve of the wood remains were partly carbonised, this mainly includes unworked wood remains that may represent fuel. The worked wood remains comprised all taxa that are represented in the wood assemblage and likely represent (the remains of) artefacts or the waste material of wood working, but artefacts with a known function other than posts were not recognised. The assemblage of layer 50 contains two pieces of Quercus sp. that are cleaved in a radial direction (see fig. II.11). These wood remains may possibly represent elements of the bottom of small containers comparable with casks (see Leuzinger 2002, 95-98, discussing wooden artefacts of Quercus sp. found at the wetland site Arbon Bleiche 3 in Switzerland with similar dimensions as the artefact found at Brandwijk-Kerkhof) or alternatively a canoe partition (see eberschweiler 2004; Louwe Kooijmans and Kooistra 2006).

22 Comparison with table II.2 shows that dates are not available for all layers.

76

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

layer category

L15

L30

L40

L50

Nw

P

w

Nw

P

w

Nw

P

w

Nw

P

w

Alnus sp.

-

1

-

-

-

-

1

1

1+1?

13

1?

1+1?

cf. Alnus sp.

-

-

-

-

-

-

-

-

-

2

-

-

Cornus sp.

-

-

-

-

-

-

-

-

-

4

-

1

Corylus sp.

-

1?

-

-

-

-

2

-

-

5

1

-

euonymus sp.

-

-

-

-

-

-

-

-

-

-

-

1

Fraxinus sp.

-

-

-

-

1+1?

-

1

-

-

1

-

-

Pomoideae

-

-

-

-

-

-

1

-

-

-

-

1

Quercus sp.

-

2

-

-

-

-

-

-

1?

-

2

2+1?

Viburnum sp.

-

-

-

-

-

-

-

-

-

-

-

1?

Indet./bark

-

-

-

-

-

-

2

-

-

8

-

-

7

1

1+2?

33

taxon

total

layer category

3+1?

1+1?

L55

L60

L80

3+1? 6+3?

all layers

Nw

P

w

Nw

P

w

Nw

P

w

Nw

P

Alnus sp.

-

-

-

1

1

-

-

-

-

15

cf. Alnus sp.

-

-

-

-

-

-

-

-

-

2

-

Cornus sp.

1

-

-

-

-

-

-

-

-

5

Corylus sp.

1

-

-

-

-

-

-

-

1?

euonymus sp.

-

-

-

-

-

-

-

-

Fraxinus sp.

1

-

-

-

-

-

-

Pomoideae

-

-

-

-

-

-

Quercus sp.

-

-

-

2

-

Viburnum sp.

-

-

-

-

Indet./bark

-

-

-

total

3

total w

N

%

23

32

-

2

3

-

1

6

8

8

1+1?

1?

11

16

-

-

-

1

1

1

-

-

3

1+1?

-

5

7

-

-

-

1

-

1

2

3

-

-

-

-

2

4

2+2?

10

14

-

-

-

-

-

-

-

1?

1

1

-

-

-

-

-

-

10

-

-

10

14

3

1

1?

46

Copyright © 2010. Leiden University Press. All rights reserved.

taxon 3+1? 2+2?

9+3? 7+6?

Nw = non-worked

- = not present

P = pointed

? = possibly belonging to a specific category

71

w = worked in another way than pointed Table II.9 Brandwijk-Kerkhof, wood from various layers (Hänninen and Vermeeren 1998).

77

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

0

10 cm

Copyright © 2010. Leiden University Press. All rights reserved.

Figure II.11 Brandwijk-Kerkhof, layer 50, worked piece of wood of Quercus sp., cleaved in radial direction from the tree, drawing: E. van Driel after Jeltje Schreurs.

0

5 cm

Figure II.12 Brandwijk-Kerkhof, layer 50, pointed post of Quercus sp., drawing: E. van Driel after Jeltje Schreurs.

78

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

The identified pointed posts are from Alnus sp., Corylus sp., Fraxinus sp. and Quercus sp. Most posts are pointed relatively simply (see fig. II.12 for an example). In a preliminary publication a row of pointed posts is mentioned (Van Gijn and Verbruggen 1992). The structure consisted of pointed posts of Alnus sp., Corylus sp. and Quercus sp., found in the adjacent squares 10 and 25 at a distance of c. 0.5 metres from each other23 (Van Poecke 1991). The function of the posts is unclear; they may have functioned as a landing stage since the posts were probably located at the water edge. wattle work was not found in between the posts. There are only weak indications of the selective use of wood based on the quality of the wood and the functions of artefacts, since artefacts were not recognised in the wood assemblage. Firstly, Quercus sp. may have been selected for its qualities concerning construction since the number of pointed and worked wood remains made of Quercus sp. is larger than the number of unworked wood remains made of Quercus sp. (see table II.9). Secondly, Euonymus sp. (representing E. europaeus) was only found in a worked state. It is possible that Euonymus sp. was imported from elsewhere, since pollen and seeds of this species were not found at the site, which could indicate selective use. On the other hand, the species might have been attested in the assemblage of the unworked wood if the number of identified wood remains had been larger. Finally, the evidence of the selective use of wood from contemporaneous sites in the region (Louwe Kooijmans et al. 2001a, b; appendix III) indicates that the selective use of wood based on the quality of the wood was probably not uncommon at Brandwijk-Kerkhof. The season of cutting was determined for two wood remains of layer 15 as summer/autumn and probably autumn, and for two wood remains of layer 50 as spring/summer and probably autumn. For both layers the results do not necessarily mean that the site was occupied in both seasons during the same year. Table II.10 shows the results of the charcoal analysis. The analysis included 109 pieces from three layers (other layers contained charcoal remains as well but these were not investigated). The number of samples of layer 50 is too small to be representative (Hänninen and Vermeeren 1998) and the number of charcoal

N layer

%

L30

L50

L60

total

L30

L50

L60

total

Alnus sp.

24

18

41

83

92

64

75

76

Cornus sp.

-

1

1

2

-

4

2

2

Corylus sp.

-

4

3

7

-

14

5

6

Fraxinus sp.

2

1

4

7

8

4

7

6

Quercus sp.

-

-

3

3

-

-

5

3

Viburnum sp.

-

1

-

1

-

4

-

1

Indet.

-

3

3

6

-

11

5

6

total

26

28

55

109

Copyright © 2010. Leiden University Press. All rights reserved.

taxon

- = not present Table II.10 Brandwijk-Kerkhof, charcoal from various layers (Hänninen and Vermeeren 1998). 23 The series of pointed posts dates to L50 or L60. It is not certain whether the information on the identifications of the posts of the structure is complete.

79

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

identifications of individual phases is too small to reconstruct changes in the vegetation through time. Alnus sp. is the dominant taxon in the assemblage of all layers together. Layer 30 contained only remains of Alnus sp. and Fraxinus sp. Layer 50 contains an increased number of taxa, dominated by Alnus sp. and afterwards Corylus avellana. The assemblage of layer 60 is highly similar to the assemblage of layer 50. There are indications of the use of wood affected by fungi and insects, probably dead wood. Interestingly, carbonised ants were even present in a piece of alder wood; these probably represent Lasius fuliginosus, see figure II.13 (the shining black ant, nestling in decaying or dead trees, pers. comm. Hakbijl in Hänninen and Vermeeren 1998). Comparison of the wood and charcoal assemblages indicates that Euonymus sp. and Pomoideae are present in the wood assemblage but absent in the charcoal assemblage. There are not enough data to conclude whether Euonymus sp. was present in the local vegetation or not. Pomoideae may represent Malus sylvestris, Pyrus pyraster, Crataegus monogyna and/or Sorbus aucuparia. Only seeds and fruits of Malus sylvestris and Crataegus monogyna have been found at Brandwijk-Kerkhof, so the wood probably represents one of these two species. The numbers of macroremains indicate that both were probably present in the local vegetation although collection probably played a role as well. Sample 13.124 contained a knot made of bark of cf. Acer sp./Ulmus sp. Bark of Acer sp./Ulmus sp. was used for knots at Hardinxveld-Giessendam Polderweg as well (Louwe Kooijmans et al. 2001b) and thus might indicate the selective use of taxa. The many fish scales in the sample and the location of this sample indicate that the knot may have been part of fishing gear.

Copyright © 2010. Leiden University Press. All rights reserved.

Figure II.13 Brandwijk-Kerkhof, layers 30 and 50, carbonised ant, cf. Lasius fuliginosus, magnification 20 x.

80

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

II.4

dIscussIon

Copyright © 2010. Leiden University Press. All rights reserved.

II.4.1 r econstructIon of the natural vegetatIon The main results on the natural vegetation based on the pollen and macroremains of the four cores taken north of the dune are summarised here, while the results on human impact are discussed in the next paragraph (both based on Out 2008a). Four cores were sampled north of the dune at 1, 5, 10 and 20 metres distance respectively from where the peat that was formed during deposition of layer 50 was attached to the dune sand. Figure II.14 schematically shows the development of the vegetation based on the analysis of these cores. The top of the dune (0.5 m -NAP) is not shown in the figure. The four reconstructions correspond with the period of the formation of layer 30 (4600-4550 BC), the period before layer 50 (4550-4200 BC), layer 50 and/or 60 (4200-3800 BC), and the period of recovery after layer 50 and/or 60 (after 3800 BC). The reconstruction of the vegetation based on the four cores indicates that the natural vegetation of the higher parts of the dune at Brandwijk-Kerkhof consisted of deciduous Tilia/Quercus/Corylus woodland. The vegetation on the lower slopes and in the surrounding wetland area consisted of alder carr and eutrophic freshwater marsh vegetation. This vegetation was present during all registered occupation phases. The increasing ground water level resulted in the general succession of the vegetation of the dune slope from woodland of dry terrain into woodland of moist terrain and/or marsh and open water. This succession resulted in a decrease in the quantity of woodland of dry terrain at the dune. Open water became present in the near surroundings of the dune (west of the sample locations) at 4200-3800 BC. After 3800 BC, the open water was still present but probably further away from the sampling points. The presence of macroremains of dryland trees (Tilia sp., Quercus sp., Fraxinus excelsior and Corylus avellana) indicates that dryland woodland was still present at the dune during the later occupation phases, although alder carr probably dominated on the lower and middle parts of the slopes. The analysis of macroremains, wood and charcoal from the excavation that is presented here supports the earlier presented reconstruction of the natural vegetation and does not result in important alterations of the reconstruction. woodland of dry terrain, alder carr vegetation, marsh vegetation and water plants form the major elements of the natural vegetation. Most taxa indicate moderately to highly eutrophic conditions. Some parts of the vegetation may have been rather dense, as demonstrated by the presence of woodland of dry terrain and indicators of partly shaded terrain (Polypodium vulgare, Galeopsis-type and Moehringia trinervia). Comparison of layer 30 and layer 50 shows a decrease in the importance of dryland vegetation and an increase in marsh vegetation, which can be related to the rise of the ground water table. The assemblage of layer 60, which contains several macroremains of dryland shrubs in a primarily waterlogged state, supports the presence of woodland of dry terrain after the main occupation. During formation of layer 70 however, woodland of dry terrain on the slopes was replaced by alder carr at the location of the excavation trench as a result of the submerging of the dune.

81

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

3

After 4200-3800 BC

T

Q

Q

A

A

A

S

A

Q

O

A

O

S

A

4

5 6

4200-3800 BC

Q 3

S

Q

A

4

S

A

A

O

S

A

5 6

3

4550-4200 BC

T S

Q

4

S

T

Q

A

A

5

Q

S

A

A

Q

S

A

A

6

3

T

4600-4550 BC

Q

T T

4

Q S

Q

O

Q

A

A

Q S

A

A

A

6 m - NAP

Copyright © 2010. Leiden University Press. All rights reserved.

O T

5

1 A Sand Peat Open water

82

5 B T = Tilia sp. Q = Quercus sp. O = Other trees

10 C A = Alnus glutinosa S = Shrubs = Cleared tree

20 m. D = Upland herbs = Wetland herbs = Water plants

Copyright © 2010. Leiden University Press. All rights reserved.

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

The presence of especially Quercus sp. and in lesser degree other trees (possibly Fraxinus sp., Ulmus sp. and Acer campestre) in the marsh, as drawn in figure II.14, is the subject of debate since parallels for bog oaks are not known from recent vegetation types. Oaks and other trees of softwood alluvial woodland Alno-Padion and/or Filipendulo-Alnetum, see references below) can grow in marshes when growing on a mineral subsurface or on peat containing a small quantity of clay. Germination of seedlings of at least oak and ash is possible during phases of increased drainage of the marsh, possibly as the result of flooding or channel activity (Kooistra et al. 2006; Sass-Klaassen and Hanraets 2006). The pollen diagrams of the two cores sampled at 10 and 20 metres away from the dune show values of Quercus sp. that are relatively low compared with values of Alnus sp., but the macroremains show considerable quantities of buds and/or bud scales of Quercus sp. It can be questioned whether the bud scales were blown away from trees present at the dry surface or that they represent the local presence of oaks. It is unclear whether the absence of other macroremains is indicative of absence of local oaks. The most parsimonious interpretation therefore is to reject local presence of Quercus sp. in the marsh. The study of wetland wood at Zwolle-Stadshagen however represents an interesting parallel that supports that local presence of oak in the marsh at Brandwijk-Kerkhof is possible. At Zwolle, the pollen curves of Quercus sp., Fraxinus sp. and Ulmus sp. were again very low compared with Alnus sp., and the diagram was interpreted as representing alder carr (Alnetea glutinosae). Complete trees of Quercus sp., Fraxinus sp. and Ulmus sp. were however found in the peat, indicative of local presence. Clay was only observed in the relevant horizons during micromorphological analysis (Kooistra et al. 2006; Sass-Klaassen and Hanraets 2006). Interestingly, the germination of seedlings at Brandwijk-Kerkhof during later phases could have been triggered during relative dry periods. Micromorphological analysis of the sediment in the excavation trench indeed supported that the peat dried out during a short period (pers. comm. Verbruggen 2006, see the results of layer 50 in paragraph II.3.1). The sample box of layer 50 furthermore contained minor quantities of clay, and even smaller quantities of clay may similarly have been present in the peat north of the dune during that phase. The conditions therefore may have enabled the occurrence of Quercus sp. In addition, clearance of vegetation during occupation may have decreased competition, resulting in better possibilities for germination as well (cf. Sass-Klaassen and Hanraets 2006). The presence of bog oaks at Brandwijk-Kerkhof could have been studied in better detail if the wood remains found in the cores had been identified. The systematic identification of wood from pollen cores is therefore suggested for future research. The macroremains from the excavation show the presence of a shrub species that was not present in the cores: Crataegus monogyna. The species was probably present in the local vegetation in moderate or low quantities. It is unclear whether the wood and charcoal data support a local presence since it is not possible to identify wood remains of the species on species level. The local presence of Tilia sp., Rhamnus cathartica, Acer campestre, Ulmus sp., Ligustrum vulgare, Sambucus nigra, Prunus sp. is additionally indicated by the pollen and/or macroremains assemblage, but these taxa were not found in the wood and charcoal assemblage either. This difference between the various assemblages may be related to the limited size of the excavation and the limited number of samples of wood and charcoal. The scarcity of some shrubs in the natural vegetation may additionally play a role as well. Tilia sp. is usually found in low quantities at excavations at Dutch wetland sites, probably due to the soft wood. Remains of Ulmus sp., Betula sp. and Salix sp. were only found in the pollen assemblage. Figure II.14 opposite page, Brandwijk-Kerkhof, periods corresponding with the various layers, schematic vegetation reconstruction of the vegetation at the northern side of the dune, based on (partial) analysis of pollen and macroremains of four cores. The vegetation icons are not to scale. The group of ‘shrubs’ includes Corylus avellana, Rhamnus cathartica, Cornus sanguinea, Crataegus monogyna, Prunus spinosa, Sambucus nigra, Viburnum opulus, Ligustrum vulgare, Rubus fruticosus, Rhamnus frangula and Myrica sp. C. avellana, a dryland shrub, may also have been present as a tree. The group of “other trees” include Fraxinus excelsior, Ulmus sp. and Acer campestre, which may have been more common than suggested in the reconstruction. Salix sp. may additionally have been present along the open water. The group of wetland herbs also include grasses and sedges. See the text for further explanation.

83

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

Ulmus sp. and Salix sp. may nevertheless have been present in the extra-local vegetation in low quantities, since macroremains of these taxa usually do not remain preserved. The scarcity of remains of Betula sp. probably reflects the scarcity of the taxon at Brandwijk-Kerkhof. Sambucus sp. was only found in the pollen assemblage and not in the macroremains, wood or charcoal assemblage. The absence of fruits of Sambucus sp. remains unexplained. A larger data set of wood and charcoal might have revealed Sambucus sp. after all. Marine influence at Brandwijk-Kerkhof seems to have been absent or very small. Pollen of Armeria maritima was present in one of the pollen cores (Out 2008a), which is characteristic of high salt marshes. It was probably transported by water during a heavy storm or by wind and does not represent the local vegetation since it is represented by a single pollen grain only. Other pollen diagrams of sites in the central river area occasionally contain pollen of halophytic taxa as well (see chapter 2.8.2). At Brandwijk-Kerkhof there are also some species that do not need but nevertheless tolerate brackish water (Najas marina, Zachinellia palustris, Phragmites australis and Schoenoplectus tabernaemontani). The presence of major marine influence at Brandwijk-Kerkhof is however excluded by the absence of macroremains of taxa that exclusively grow under brackish/marine conditions and by the dominance of taxa that do not tolerate marine influences (Berula erecta, Hydrocotyle vulgaris and many others; Cappers 1994). The presence especially of Elatine triandra suggests fluctuating water levels (Brinkkemper et al. 2008). II.4.2 huMan IMpact The indications of human impact based on the pollen cores are discussed elsewhere (Out 2008a), but the main results will be briefly mentioned here (see also figure II.14). At 4600-4550 BC, human impact was restricted, resulting in a slight increase in herbs and shrubs. It cannot be excluded that natural processes played a role as well in the observed changes in the vegetation. At 4550-4200 BC, alder carr became more dominant on the lower slopes of the dune. The vegetation became more open due to restricted human impact and/or natural disturbance, as indicated by the increase in shrubs. At 4200-3800 BC, open water became present in the near surroundings of the dune (west of the four cores). Human impact was maximal, resulting in the clearance of trees and more open vegetation. Only the changes in the pollen diagram contemporaneous with this period (the top of layer 50 and/or 60) can accurately be attributed to human impact. Signals of human impact are characterised by a slight decrease in Tilia sp., a decrease in macroremains of Alnus sp. and an increase in the dryland shrubs, dryland herbs that are characteristic of human impact (Artemisia sp., Plantago lanceolata, Polygonum persicaria-type and Polygonum convolvulus-type), wetland herbs that indicate disturbance and open patches (Filipendula ulmaria, Urtica dioica, Berula erecta and Eupatorium cannabinum), Poaceae and Cyperaceae. There is no relationship between the strength of the signal of human impact and the distance between the dry dune surface and the sample point, at least over a distance of 20 metres. After 3800 BC, the open water was still present but probably further away from the sampling points. Human impact decreased, resulting in the partial recovery of the woodland. The macroremains data from the excavation show primarily the submergence of the dune and the general disturbance of the vegetation. A sound comparison of human impact through time based on macroremains, wood and charcoal is restricted by the unequal availability of samples from the different layers (occupation periods) since most material is available from the most intensive occupation period (layer 50). Reconstructions of human impact on the southern slope of the dune before, during and after occupation are available for the periods contemporaneous with layers 45 and 50. The data show the clearance of alder carr during both occupation periods, and clearance of Tilia sp. and an increase in shrubs and anthropogenic indicators contemporaneous with layer 50. Both occupation periods are followed by the (partial) recovery of the vegetation. These developments through time correspond to the results obtained at the northern side of the dune.

84

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

Comparison of the indications of human impact from the cores and from the excavation shows many similarities. At both sides of the dune indications of human impact appear to be largest during layer 50 (for the cores this concerns layer 50 and/or 60 since these layers could not be recognised separately), including the clearance of Tilia vegetation and alder carr, resulting in the increased presence of shrubs and anthropogenic indicators. The presence of people did not result in large-scale deforestation, and dry terrain woodland recovered and remained present after occupation. An apparent difference between the results from the cores and the excavation is the presence of crop plant macroremains and carbonised macroremains at the excavation, which may of course be related to the differences in the number of samples and sample volume but also with the stronger human impact at the southern site of the dune (as indicated by the distribution of archaeological remains). The importance of Alnus glutinosa in the wood and charcoal assemblages is in accordance with indications of the clearance of Alnus glutinosa vegetation by people in the diagrams of the cores (especially concerning layer 50 and/or 60) and in the macroremains diagram of layer 45. The additional importance of Quercus sp. and Corylus sp. in the wood assemblage suggests that these taxa may have been cleared as well. They were indeed present in the local vegetation, and fluctuations in the curves of these taxa in the pollen and macroremains diagrams may therefore represent clearance as well (see the pollen diagrams of core B and D and the macroremains diagrams of core B, C and D in Out 2008a). The wood and charcoal data suggest that most wood was collected locally, although it may have been collected at the other nearby dunes as well. There are some indications of the selective use of Quercus sp. and possibly Euonymus europaeus. The find of a knot demonstrates the use of the bark of trees, although this way of using plant material probably remains strongly underrepresented in our knowledge. II.4.3 seasonalIty and sIte functIon Carbonised macroremains can provide information on the season of occupation at the site. Carbonised remains of Corylus avellana, Cornus sanguinea and Trapa natans point to occupation during September/October, although macroremains (nuts) of these taxa can be stored when roasted and in that case may give a wrong indication on seasonality. Remains of Galium aparine point to occupation between july and September, while Solanum nigrum and Nymphaea alba indicate occupation between August and October. In conclusion, the carbonised plant remains indicate that occupation occurred at least in September or August-October during the phases contemporaneous with the layers 30, 50 and 60. The carbonised cereals do not give information on seasonality since there is no indisputable evidence of local crop cultivation. The wood remains demonstrate human activity in spring/summer and probably autumn for layer 50. Particular fish remains confirm the summer occupation for layer 50 and 60, but less strongly for layer 3024 (Ball 1997; Raemaekers 1999). Together the data support autumn and possibly summer occupation for layer 30 and summer and autumn occupation for layer 50 and 60. The information on seasonality is too general to tell whether people were present at the site for months continuously or visited the site intermittently. The archaeobotanical data indicate that the collection of edible parts of plants and the possibility to collect storable fruits may have been one of the reasons to visit Brandwijk-Kerkhof. The practice of local cereal cultivation is discussed elsewhere (Out 2008a, see also paragraph 11.6.13). It is concluded that crop cultivation was either practised locally on a small scale, or elsewhere. This location elsewhere may be another dune (e.g. the higher and more extensive dune Brandwijk-Donk), but could also be in a different region, like the Pleistocene sand soils to the south of the river area. A new argument on local crop cultivation, resulting from the data presented here, is that the assemblage of potential arable weeds does not exclude local crop cultivation. 24 In the fish assemblage of layer 30, some of the summer indicators of layer 50 and 60 are absent (mullet family and sturgeon). Remains of the salmon are present in layer 30 and may represent summer indicators, but may also represent small individuals that remain present in the river area during the whole year.

85

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

The indications of summer and autumn occupation from layer 50 onwards do not exclude local crop cultivation either. However, the indications of local crop cultivation would be stronger if indications of occupation during late spring would be available. Moreover, the unique presence of opium poppy can be interpreted as an indication of the importation of crop plants. Analysis of preliminary data from the excavation resulted in the conclusion that both the mammal and fish spectra through time suggest a continuity of site function at Brandwijk-Kerkhof (Van Gijn and Verbruggen 1992; Raemaekers 1999). The data however show some shifts from layer 50 onwards. The archaeological remains are spread over a larger area than before and the refuse layer is generally thicker, suggesting more intensive occupation (more people or longer visits). There are stronger indications of summer occupation besides autumn occupation, as shown by changes in the fish spectrum (see discussion on seasonality above). In contrast to layer 30, there is evidence of the presence of cattle, and stronger evidence of the presence of domestic pig, which was possibly absent before. The number of plant taxa found in a carbonised state increased and crop plants were introduced. In addition the pollen diagrams show maximal human impact during the formation of layer 50 and/or 60. Although hunting, fishing, fowling and gathering probably remained very important for the people who visited the dune, it is clear that some changes took place in local subsistence and occupation intensity, which may indicate a shift in site function. A final conclusion on the site function should, however, be based on a broader range of evidence.

86

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

22 C

23 C

23 D

24 C

24 D

25 C

25 D

26 C

26 D

sample type

hp

hp

hp

hp

hp

hp

hp

hp

hp

mesh width

-

-

-

-

-

-

-

-

-

volume (litre)

?

?

?

?

?

?

?

?

?

Tilia sp.

-

-

-

-

-

-

-

-

-

Quercus sp.

-

1

-

-

-

-

-

-

-

Quercus sp., cupulae

-

-

-

-

-

1

-

-

-

Quercus sp., juvenile

-

-

-

-

-

-

-

-

-

Fraxinus excelsior

-

-

-

-

-

-

-

-

-

Cornus sanguinea

-

1

-

1

-

9

3

9

21

Corylus avellana

2

-

8, 1 c

++

+, 1 c

5

3

19

26

Corylus avellana, catkins

-

-

-

-

-

-

-

-

-

Crataegus monogyna

-

-

-

-

-

-

1

-

1

Malus sylvestris

-

-

-

-

-

-

-

-

-

Malus sylvestris, parenchyma

-

-

-

-

-

-

-

-

-

Viburnum opulus

-

-

-

-

-

-

-

1

-

Prunus sp.

-

-

-

-

-

-

-

-

-

Prunus spinosa

-

-

-

-

-

-

-

-

1

Rubus fruticosus

-

-

-

-

-

-

-

-

-

Fallopia dumetorum

-

-

-

-

-

-

-

-

-

Urtica dioica

-

-

-

-

-

-

-

-

-

Galium aparine

-

-

-

-

-

-

-

-

-

Lapsana communis

-

-

-

-

-

-

-

-

-

Chaerophyllum bulbosum/temulum

-

-

-

-

-

-

-

-

-

Arctium cf. lappa

-

-

-

-

-

-

-

-

-

Atriplex patula/prostrata

-

-

-

-

-

-

-

-

-

Brassica rapa

-

-

-

-

-

-

-

-

-

Capsella bursa-pastoris

-

-

-

-

-

-

-

-

-

Chenopodium album

-

-

-

-

-

-

-

-

-

Chenopodium ficifolium

-

-

-

-

-

-

-

-

-

Fallopia convolvulus

-

-

-

-

-

-

-

-

-

taxon Group 1

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2

Table II.5 part 1a.

87

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

22 C

23 C

23 D

24 C

24 D

25 C

25 D

26 C

26 D

Persicaria cf. maculosa

-

-

-

-

-

-

-

-

-

Persicaria lapathifolia

-

-

-

-

-

-

-

-

-

Polygonum aviculare

-

-

-

-

-

-

-

-

-

Solanum nigrum

-

-

-

-

-

-

-

-

-

Stellaria media

-

-

-

-

-

-

-

-

-

Hordeum vulgare var. nudum

-

-

-

-

-

-

-

-

-

Hordeum vulgare, internodia

-

-

-

-

-

-

-

-

-

Triticum dicoccon

-

-

-

-

-

-

-

-

-

Triticum dicoccon, glume bases

-

-

-

-

-

-

-

-

-

Triticum dicoccon, rachis segments

-

-

-

-

-

-

-

-

-

Triticum dicoccon, spikelet forks

-

-

-

-

-

-

-

-

-

Papaver somniferum ssp. setigerum

-

-

-

-

-

-

-

-

-

Alnus glutinosa

-

-

-

-

-

-

-

-

-

Alnus glutinosa, buds

-

-

-

-

-

-

-

-

-

Alnus glutinosa, bud scales

-

-

-

-

-

-

-

-

-

Alnus glutinosa, cones

-

-

-

-

-

-

-

-

-

Alnus glutinosa, male catkins

-

-

-

-

-

-

-

-

-

Salix sp., buds

-

-

-

-

-

-

-

-

-

Humulus lupulus

-

-

-

-

-

-

-

-

-

Alisma plantago-aquatica

-

-

-

-

-

-

-

-

-

Angelica sylvestris/Peucedanum palustre Apium nodiflorum

-

-

-

-

-

-

-

-

-

Carex acutiformis

-

-

-

-

-

-

-

-

-

elatine triandra

-

-

-

-

-

-

-

-

-

elatine hydropiper

-

-

-

-

-

-

-

-

-

eupatorium cannabinum

-

-

-

-

-

-

-

-

-

Filipendula ulmaria

-

-

-

-

-

-

-

-

-

Galium palustre

-

-

-

-

-

-

-

-

-

Hypericum cf. tetrapterum

-

-

-

-

-

-

-

-

-

Group 2 (cont.)

Group 3

Group 4

Copyright © 2010. Leiden University Press. All rights reserved.

Group 5

Table II.5 part 1b.

88

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

22 C

23 C

23 D

24 C

24 D

25 C

25 D

26 C

26 D

Iris pseudacorus

-

-

-

-

-

-

1

-

-

Lythrum salicaria

-

-

-

-

-

-

-

-

-

Mentha aquatica/arvensis

-

-

-

-

-

-

-

-

-

Oenanthe aquatica

-

-

-

-

-

-

-

-

-

Oenanthe sp.

-

-

-

-

-

-

-

-

-

Phragmites australis, rhizome buds

-

-

-

-

-

-

-

-

-

Phragmites australis, stem fragments

-

-

-

-

-

-

-

-

-

Rumex hydrolapathum

-

-

-

-

-

-

-

-

-

Sagittaria sagittifolia

-

-

-

-

-

-

-

-

-

Schoenoplectus lacustris

-

-

-

-

-

-

-

-

-

Schoenoplectus lacustris/ tabernaemontani Schoenoplectus tabernaemontani

-

-

-

-

-

-

-

-

-

Scirpus sylvaticus

-

-

-

-

-

-

-

-

-

Sium latifolium

-

-

-

-

-

-

-

-

-

Sparganium erectum

-

-

-

-

-

-

-

-

-

Stachys palustris

-

-

-

-

-

-

-

-

-

Thelypteris palustris, leaves

-

-

-

-

-

-

-

-

-

Thelypteris palustris, sporangia

-

-

-

-

-

-

-

-

-

Typha sp.

-

-

-

-

-

-

-

-

-

Valeriana officinalis

-

-

-

-

-

-

-

-

-

Apium cf. repens

-

-

-

-

-

-

-

-

-

Persicaria hydropiper

-

-

-

-

-

-

-

-

-

Stellaria aquatica

-

-

-

-

-

-

-

-

-

Callitriche sp.

-

-

-

-

-

-

-

-

-

Ceratophyllum demersum

-

-

-

-

-

-

-

-

-

Chara sp.

-

-

-

-

-

-

-

-

-

Najas marina

-

-

-

-

-

-

-

-

-

Najas minor

-

-

-

-

-

-

-

-

-

Nuphar lutea

-

-

-

-

-

-

-

-

-

Nymphaea alba

-

-

-

-

-

-

-

-

-

Group 5 (cont.)

Group 6

Copyright © 2010. Leiden University Press. All rights reserved.

Group 7

Table II.5 part 1c.

89

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

22 C

23 C

23 D

24 C

24 D

25 C

25 D

26 C

26 D

Potamogeton sp.

-

-

-

-

-

-

-

-

-

Ranunculus aquatilis-type

-

-

-

-

-

-

-

-

-

Salvinia natans

-

-

-

-

-

-

-

-

-

Trapa natans

-

-

-

-

-

-

-

1

-

Zannichellia palustris

-

-

-

-

-

-

-

-

-

Apiaceae

-

-

-

-

-

-

-

-

-

Brassica sp.

-

-

-

-

-

-

-

-

-

Carex sp.

-

-

-

-

-

-

-

-

-

Carex sp., bicarpellate

-

-

-

-

-

-

-

-

-

Carex sp., tricarpellate

-

-

-

-

-

-

-

-

-

Carex sp. in utricle

-

-

-

-

-

-

-

-

-

Galium sp.

-

-

-

-

-

-

-

-

-

Galeopsis bifida-type

-

-

-

-

-

-

-

-

-

Sonchus sp.

-

-

-

-

-

-

-

-

-

Hypericum sp.

-

-

-

-

-

-

-

-

-

juncus effusus-type

-

-

-

-

-

-

-

-

-

juncus sp.

-

-

-

-

-

-

-

-

-

Luzula sp.

-

-

-

-

-

-

-

-

-

Poa sp.

-

-

-

-

-

-

-

-

-

Poaceae

-

-

-

-

-

-

-

-

-

Ranunculus repens-type

-

-

-

-

-

-

-

-

-

Rumex sp.

-

-

-

-

-

-

-

-

-

Rumex sp., perianths

-

-

-

-

-

-

-

-

-

Stellaria aquatica/media

-

-

-

-

-

-

-

-

-

Stellaria sp.

-

-

-

-

-

-

-

-

-

Indet.

-

-

-

-

-

-

-

-

-

Group 7 (cont.)

Copyright © 2010. Leiden University Press. All rights reserved.

Group 8

hp = handpicked sample + = few (1-10) sieve = sample from sifting residue ++ = some tens (10-49) lab = botanical sample - = not present c = carbonised x, yc = x macroremains of which y are carbonised Table II.5 Brandwijk-Kerkhof, layer 50, macroremains, part 1d.

90

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

27 D

13.124

13.125

24.523

26.13

26.27

26.633

27.26

sample type

hp

sieve

hp

hp

hp

hp

hp

hp

mesh width

-

1.5

-

-

-

-

-

-

volume (litre)

?

?

?

?

?

?

?

?

Tilia sp.

-

-

-

-

-

-

-

-

Quercus sp.

-

-

-

1

-

-

-

-

Quercus sp., cupulae

-

-

-

1

-

-

-

-

Quercus sp., juvenile

-

-

-

-

-

-

-

-

Fraxinus excelsior

-

-

-

-

-

-

-

-

Cornus sanguinea

4

1

1

-

-

-

-

-

Corylus avellana

5

10, 1 c

5, 1 c

-

-

10, 2 c

-

1c

Corylus avellana, catkins

-

-

-

-

-

-

-

-

Crataegus monogyna

-

-

1

-

-

-

1 cf.

-

Malus sylvestris

-

-

-

-

-

-

-

-

Malus sylvestris, parenchyma

-

-

-

-

-

-

-

-

Viburnum opulus

-

-

-

-

-

-

-

-

Prunus sp.

-

-

-

-

-

-

-

-

Prunus spinosa

-

1

-

-

-

-

-

-

Rubus fruticosus

-

1

-

-

-

-

-

-

Fallopia dumetorum

-

-

-

-

-

-

-

-

Urtica dioica

-

-

-

-

-

-

-

-

Galium aparine

-

1c

-

-

-

-

-

-

Lapsana communis

-

-

-

-

-

-

-

-

Chaerophyllum bulbosum/temulum

-

-

-

-

-

-

-

-

Arctium cf. lappa

-

-

-

-

-

-

-

-

Atriplex patula/prostrata

-

-

-

-

-

-

-

-

Brassica rapa

-

-

-

-

-

-

-

-

Capsella bursa-pastoris

-

-

-

-

-

-

-

-

Chenopodium album

-

7

31

-

-

-

-

-

Chenopodium ficifolium

-

-

-

-

-

-

-

-

Fallopia convolvulus

-

-

-

-

-

-

-

-

taxon Group 1

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2

Table II.5 part 2a.

91

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

27 D

13.124

13.125

24.523

26.13

26.27

26.633

27.26

Persicaria cf. maculosa

-

-

7

-

-

-

-

-

Persicaria lapathifolia

-

2

6

-

-

-

-

-

Polygonum aviculare

-

-

-

-

-

-

-

-

Solanum nigrum

-

-

-

-

-

-

-

-

Stellaria media

-

-

-

-

-

-

-

-

Hordeum vulgare var. nudum

-

1c

-

-

-

-

-

-

Hordeum vulgare, internodia

-

-

-

-

-

-

-

-

Triticum dicoccon

-

1c

-

-

-

-

-

-

Triticum dicoccon, glume bases

-

-

-

-

-

-

-

-

Triticum dicoccon, rachis segments

-

-

-

-

-

-

-

-

Triticum dicoccon, spikelet forks

-

-

-

-

-

-

-

-

Papaver somniferum ssp. setigerum

-

-

-

-

-

-

-

-

Alnus glutinosa

-

-

1

-

-

-

-

-

Alnus glutinosa, buds

-

-

-

-

-

-

-

-

Alnus glutinosa, bud scales

-

-

-

-

-

-

-

-

Alnus glutinosa, cones

-

-

-

-

-

-

-

-

Alnus glutinosa, male catkins

-

-

-

-

-

-

-

-

Salix sp., buds

-

-

-

-

-

-

-

-

Humulus lupulus

-

4

-

-

-

-

-

-

Alisma plantago-aquatica

-

-

-

-

-

-

-

-

Angelica sylvestris/Peucedanum palustre Apium nodiflorum

-

-

-

-

-

-

-

-

Carex acutiformis

-

-

-

-

-

-

-

-

elatine triandra

-

-

-

-

-

-

-

-

elatine hydropiper

-

-

-

-

-

-

-

-

eupatorium cannabinum

-

-

-

-

-

-

-

-

Filipendula ulmaria

-

-

-

-

-

-

-

-

Galium palustre

-

-

-

-

-

-

-

-

Hypericum cf. tetrapterum

-

-

-

-

-

-

-

-

Group 2 (cont.)

Group 3

Group 4

Copyright © 2010. Leiden University Press. All rights reserved.

Group 5

Table II.5 part 2b.

92

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

27 D

13.124

13.125

24.523

26.13

26.27

26.633

27.26

Iris pseudacorus

-

-

-

-

-

-

-

-

Lythrum salicaria

-

-

-

-

-

-

-

-

Mentha aquatica/arvensis

-

-

-

-

-

-

-

-

Oenanthe aquatica

-

-

-

-

-

-

-

-

Oenanthe sp.

-

-

-

-

-

-

-

-

Phragmites australis, rhizome buds

-

-

-

-

-

-

-

-

Phragmites australis, stem fragments

-

-

-

-

-

-

-

-

Rumex hydrolapathum

-

-

-

-

-

-

-

-

Sagittaria sagittifolia

-

-

-

-

-

-

-

-

Schoenoplectus lacustris

-

-

8

-

-

-

-

-

Schoenoplectus lacustris/ tabernaemontani Schoenoplectus tabernaemontani

-

-

-

-

-

-

-

-

Scirpus sylvaticus

-

-

-

-

-

-

-

-

Sium latifolium

-

-

-

-

-

-

-

-

Sparganium erectum

-

-

-

-

-

-

-

-

Stachys palustris

-

-

1

-

-

-

-

-

Thelypteris palustris, leaves

-

-

-

-

-

-

-

-

Thelypteris palustris, sporangia

-

-

-

-

-

-

-

-

Typha sp.

-

-

-

-

-

-

-

-

Valeriana officinalis

-

-

-

-

-

-

-

-

Apium cf. repens

-

-

-

-

-

-

-

-

Persicaria hydropiper

-

-

-

-

-

-

-

-

Stellaria aquatica

-

-

-

-

-

-

-

-

Callitriche sp.

-

-

-

-

-

-

-

-

Ceratophyllum demersum

-

1

-

-

-

-

-

-

Chara sp.

-

-

-

-

-

-

-

-

Najas marina

-

1

-

-

-

-

-

-

Najas minor

-

-

-

-

-

-

-

-

Nuphar lutea

-

2

-

-

-

-

-

-

Nymphaea alba

-

17

12

-

-

-

-

-

Group 5 (cont.)

Group 6

Copyright © 2010. Leiden University Press. All rights reserved.

Group 7

Table II.5 part 2c.

93

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

27 D

13.124

13.125

24.523

26.13

26.27

26.633

27.26

Potamogeton sp.

-

9

3

-

-

-

-

-

Ranunculus aquatilis-type

-

-

-

-

-

-

-

-

Salvinia natans

-

-

-

-

-

-

-

-

Trapa natans

-

-

-

-

1c

-

-

-

Zannichellia palustris

-

-

-

-

-

-

-

-

Apiaceae

-

-

-

-

-

-

-

-

Brassica sp.

-

-

2

-

-

-

-

-

Carex sp.

-

4

2

-

-

-

-

-

Carex sp., bicarpellate

-

-

-

-

-

-

-

-

Carex sp., tricarpellate

-

-

-

-

-

-

-

-

Carex sp. in utricle

-

-

-

-

-

-

-

-

Galium sp.

-

-

-

-

-

-

-

-

Galeopsis bifida-type

-

1

1

-

-

-

-

-

Sonchus sp.

-

-

-

-

-

-

-

-

Hypericum sp.

-

-

-

-

-

-

-

-

juncus effusus-type

-

-

-

-

-

-

-

-

juncus sp.

-

-

-

-

-

-

-

-

Luzula sp.

-

-

-

-

-

-

-

-

Poa sp.

-

-

-

-

-

-

-

-

Poaceae

-

-

-

-

-

-

-

-

Ranunculus repens-type

-

-

-

-

-

-

-

1

Rumex sp.

-

1

-

-

-

-

-

-

Rumex sp., perianths

-

-

-

-

-

-

-

-

Stellaria aquatica/media

-

-

-

-

-

-

-

-

Stellaria sp.

-

-

-

-

-

-

-

-

Indet.

-

-

-

-

-

-

-

-

Group 7 (cont.)

Copyright © 2010. Leiden University Press. All rights reserved.

Group 8

hp = handpicked sample - = not present sieve = sample from sifting residue lab = botanical sample c = carbonised x, yc = x macroremains of which y are carbonised Table II.5 Brandwijk-Kerkhof, layer 50, macroremains, part 1d.

94

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

28.93

36.45

43.542

9.69

13.124

27.7

36.45

43.542

sample type

sieve

sieve

sieve

lab

lab

lab

lab

lab

mesh width

1.5

1.5

1.5

0.25

0.25

0.25

0.25

0.25

?

?

?

1

1

1

1

1

Tilia sp.

-

-

-

-

-

1

-

-

Quercus sp.

-

-

-

-

-

-

-

-

Quercus sp., cupulae

-

-

-

1

-

1

-

1c

Quercus sp., juvenile

-

-

-

-

-

1

-

-

Fraxinus excelsior

-

-

-

-

-

-

-

1

Cornus sanguinea

6

-

13, 1 c

3

3

2

-

5

Corylus avellana

14, 1 c

-

-

6, 1 c

3

4, 1 c

1c

8, 1 c

Corylus avellana, catkins

-

-

-

-

-

-

-

3

Crataegus monogyna

-

-

-

1

3

-

-

1

Malus sylvestris

-

-

-

2

5

-

-

1 cf.

Malus sylvestris, parenchyma

-

-

-

-

2

-

-

-

Viburnum opulus

-

-

-

-

1

-

-

-

Prunus sp.

-

-

-

-

-

-

-

1

Prunus spinosa

-

-

-

-

-

-

-

-

Rubus fruticosus

-

-

-

-

-

-

-

-

Fallopia dumetorum

-

-

-

4

4

-

-

-

Urtica dioica

-

-

-

956

784

1376

+++++

248

Galium aparine

-

-

-

-

-

1

-

-

Lapsana communis

-

-

-

-

-

1

-

1

Chaerophyllum bulbosum/temulum

-

-

-

-

-

1

-

-

Arctium cf. lappa

-

-

-

-

1

-

-

-

Atriplex patula/prostrata

-

-

-

-

-

8

-

-

Brassica rapa

-

-

-

4

56 cf.

68

-

28

Capsella bursa-pastoris

-

-

-

272

-

-

-

-

Chenopodium album

8

-

3

428

136

340

++++

96

Chenopodium ficifolium

-

-

-

20

-

-

+++

-

Fallopia convolvulus

-

-

-

4

-

-

-

-

volume (litre) taxon Group 1

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2

Table II.5 part 3a.

95

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

28.93

36.45

43.542

9.69

13.124

27.7

36.45

43.542

Persicaria cf. maculosa

-

-

2

-

-

4

-

-

Persicaria lapathifolia

4

-

4

32

38

99

-

40

Polygonum aviculare

-

-

-

32

-

-

-

Solanum nigrum

-

-

-

20

40

6, 3 c

40

Stellaria media

-

-

-

36

160, 32 c -

-

-

-

Hordeum vulgare var. nudum

-

-

1c

-

-

-

1

-

Hordeum vulgare, internodia

-

-

-

4c

-

-

-

1c

2c

-

-

2c

-

3c

Triticum dicoccon, glume bases

-

-

-

-

40 c

1 3 c, 2 c cf. 12 c

-

4c

Triticum dicoccon, rachis segments

-

-

-

-

8c

-

-

-

Triticum dicoccon, spikelet forks

-

-

-

-

2c

2, 1 c

-

-

Papaver somniferum ssp. setigerum

-

-

-

16

32

-

-

-

Alnus glutinosa

-

-

-

148

816

467

-

300

Alnus glutinosa, buds

-

-

-

2

2

3

-

14

Alnus glutinosa, bud scales

-

-

-

-

17

5

-

15

Alnus glutinosa, cones

-

-

-

4

12

8, 1 c

-

12

Alnus glutinosa, male catkins

-

-

-

+

1

2

-

-

Salix sp., buds

-

-

-

-

9

-

-

-

Humulus lupulus

1

-

-

4

7

-

-

-

Alisma plantago-aquatica

-

-

-

48

112

56

-

28

Angelica sylvestris/Peucedanum palustre Apium nodiflorum

-

-

-

-

-

1 32

-

-

Carex acutiformis

-

-

2

4 cf.

8 cf.

8

-

-

elatine triandra

-

-

-

-

-

192

-

768

elatine hydropiper

-

-

-

-

1888

1536

-

5776

eupatorium cannabinum

-

-

-

-

8

64

-

-

Filipendula ulmaria

-

-

-

8

-

1

-

4

Galium palustre

-

-

-

8, 4 c

-

-

1c

4

Hypericum cf. tetrapterum

-

-

-

-

64

-

-

-

Group 2 (cont.)

Group 3

Triticum dicoccon

Group 4

Copyright © 2010. Leiden University Press. All rights reserved.

Group 5

Table II.5 part 3b.

96

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

28.93

36.45

43.542

9.69

13.124

27.7

36.45

43.542

Iris pseudacorus

-

-

1

-

-

-

-

-

Lythrum salicaria

-

-

-

1664

1056

1968

-

3256

Mentha aquatica/arvensis

-

-

-

32

64

4

1

-

Oenanthe aquatica

-

-

-

2

-

1

-

-

Oenanthe sp.

-

-

-

-

-

-

-

1 cf.

Phragmites australis, rhizome buds

-

-

-

+

-

-

-

-

Phragmites australis, stem fragments

-

-

-

1c

-

-

-

-

Rumex hydrolapathum

-

-

-

12

34

4

-

2

Sagittaria sagittifolia

-

-

-

-

2

-

-

-

Schoenoplectus lacustris

4

-

-

12

54

-

-

-

Schoenoplectus lacustris/ tabernaemontani Schoenoplectus tabernaemontani

-

-

-

-

-

40 cf.

2c -

12 10 c

Scirpus sylvaticus

-

-

-

-

-

1

-

-

Sium latifolium

-

-

-

8

4

2

-

4

Sparganium erectum

-

-

-

-

-

2

-

1

Stachys palustris

-

-

-

4

-

-

-

4

Thelypteris palustris, leaves

-

-

-

1

-

3

-

5

Thelypteris palustris, sporangia

-

-

-

-

-

1

-

-

Typha sp.

-

-

-

-

-

208

-

256

Valeriana officinalis

-

-

-

4

2

1

-

-

Apium cf. repens

-

-

-

-

144

-

-

20

Persicaria hydropiper

-

-

-

4

-

1

-

4

Stellaria aquatica

-

-

-

-

216

-

-

124

Callitriche sp.

-

-

-

-

-

-

-

128

Ceratophyllum demersum

-

-

13

-

-

-

-

-

Chara sp.

-

-

-

-

64

64

-

256

Najas marina

-

-

-

-

2

2

-

-

Najas minor

-

-

-

-

-

-

-

4

Nuphar lutea

1

-

8

-

-

-

-

1

Nymphaea alba

7

-

-

-

28

5

1c

11

Group 5 (cont.)

Group 6

Copyright © 2010. Leiden University Press. All rights reserved.

Group 7

Table II.5 part 3c.

97

APPeNDIx II - ARCHAeOBOTANy OF BRANDwIjK-KeRKHOF, THe NeTHeRLANDS

sample

28.93

36.45

43.542

9.69

13.124

27.7

36.45

43.542

12

-

-

-

14

5

-

4

Ranunculus aquatilis-type

-

-

-

-

360

52

-

40

Salvinia natans

-

-

-

448

640

1552

-

3200

Trapa natans

-

1c

-

1

10, 1 c

2

-

3

Zannichellia palustris

-

-

-

-

8

-

-

-

Apiaceae

-

-

-

-

2 cf.

-

-

-

Brassica sp.

-

-

-

-

-

-

-

-

Carex sp.

2

-

-

-

-

-

-

-

Carex sp., bicarpellate

-

-

-

-

34

4

-

8

Carex sp., tricarpellate

-

-

-

-

12

-

-

-

Carex sp. in utricle

-

-

-

-

1

-

-

-

Galium sp.

-

-

-

-

-

8

-

-

Galeopsis bifida-type

4

-

-

-

12

5

-

2

Sonchus sp.

-

-

-

-

2

-

-

-

Hypericum sp.

-

-

-

-

-

1

-

-

juncus effusus-type

-

-

-

-

64

-

-

-

juncus sp.

-

-

-

-

-

192

-

128

Luzula sp.

-

-

-

-

-

-

-

Poa sp.

-

-

-

-

96

68 cf.

-

Poaceae

-

-

-

-

-

2

-

8 24 cf., 16 c cf. -

Ranunculus repens-type

-

-

-

8

-

-

-

-

Rumex sp.

-

-

-

4

-

-

-

-

Rumex sp., perianths

-

-

-

4

-

-

-

-

Stellaria aquatica/media

-

-

-

128

-

596

-

-

Stellaria sp.

-

-

-

-

-

-

-

1

Indet.

-

-

-

-

-

4

-

16

Group 7 (cont.) Potamogeton sp.

Copyright © 2010. Leiden University Press. All rights reserved.

Group 8

hp = handpicked sample + = few (1-10) sieve = sample from sifting residue ++ = some tens (10-49) lab = botanical sample +++ = many tens (50-99) c = carbonised x, yc = x macroremains of which y are carbonised Table II.5 Brandwijk-Kerkhof, layer 50, macroremains, part 3d.

98

++++ = some hundreds (100-499) +++++ = many hundreds (500-999) - = not present

Copyright © 2010. Leiden University Press. All rights reserved.

Appendix III. Archaeobotany of the Hazendonk, the Netherlands III.1 IntroductIon This appendix presents new archaeobotanical data of the Hazendonk (Alblasserwaard, The Netherlands, coordinates 116.780/430.460, see fig. III.1). The National Museum of Antiquities excavated the site during 1974-1976 under direction of L.P. Louwe Kooijmans. The excavation focussed on Holocene deposits containing discarded prehistoric remains (colluvia, clay deposits and refuse layers: fossil anthropogenic horizons) dating to the Early and Middle Neolithic. A series of preliminary publications discuss various aspects of the site (e.g. Van den Broeke 1983; Louwe Kooijmans 1974, 1987; Raemaekers 1999; Zeiler 1987, 1997). The excavation also included archaeobotanical research that has only partly been published (presented below). The site is located on a late Pleistocene inland dune in the central part of the Rhine-Meuse river delta. The top of the dune is at c. 0.10 m +NAP. The Holocene geological history of the region reflects the indirect influence of the sea and lagoonal-deltaic sedimentation. A gradual rise of the ground water level resulted in the continuous accumulation of fluvial sediment and freshwater peat. As a result, the dune slopes became gradually covered by Holocene deposits. The outcropping surface decreased from 1.2 ha during phase Haz 1 (c. 4000 BC) towards 0.4 ha during phase Vlaardingen 2b (c. 2500 BC). An anastomosing network of river channels was active in the area during occupation, resulting in steady sedimentation conditions. A crevasse channel of this network was situated within a few hundred metres of the dune during c. 2900-2100 BC (i.e. also during phase Vlaardingen 2b). Other potentially inhabitable terrain in the exploitation area existed of other dunes, levees, channel belts and crevasse splays (see chapter 2). The excavation strategy was based on test pits located at some fifteen metres apart all around the dune, halfway along the slope (see fig. III.2). The test pits with good research prospects were extended with additional squares. This resulted in five major units: A, B, C, D and E. The surface of the excavated area is approximately 900 m2. The excavation revealed features and refuse layers on the slopes of the dune and in the surrounding Holocene deposits (clay and peat), corresponding with several Neolithic occupation phases. The features comprised postholes, pits, hearths, a path of branches stretching into the marsh and dated to phase VL 1a, and a palisade dating to phase VL 1b that probably enclosed c. 1000 m2 (see Out 2008c, figure 14.2). Huts or house structures were not attested but this can be related to erosion of the top of the dune and to modern soil disturbance. The refuse layers revealed flint, stone, pottery, beads, bones of wild and domestic animals, human skeletal remains and botanical remains. Material on the top and slope of the dune could not be assigned to a single phase, but refuse from the successive occupation phases was separated in distinct layers in the Holocene stratigraphy as a result of the gradual rise of the ground water level. The age of the dated layers is shown in table III.1 (see also figure 2.5). Hazendonk phases will here be abbreviated as Haz and Vlaardingen phases will be abbreviated as VL. The dating of the occupation phases is based on conventional 14C dating and by reconstruction of the ground water level rise (see Verbruggen 1992), representing mean values. Single occupation phases may represent continuous occupation or intermittent use. This last option is especially relevant for phase VL 1a that includes all remains found between the Haz 3 and VL 1b layers and probably reflects multiple visits to the site. Van Dijk et al. (1976) investigated the extension and thickness of the refuse layers by coring. The same principle is still used to investigate the presence and extent of sites in the Dutch regions with a comparable Holocene subsurface. The resulting estimates of the extent of the refuse layers are given in table III.1 and figure III.3. Distinct refuse layers are those of phase Haz 1, Haz 3 and VL 1b. The occurrence of phase Haz 2 was underestimated during excavation. All layers had a restricted spatial extent.

99

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

coastal barriers

salt marshes

tidal flats

pleistocene coversand

river deposits

inland dunes

fen peat

Hazendonk 0

upland peat bogs

50km

Figure III.I The Hazendonk, the Netherlands, location plotted on a palaeogeographical map (c. 4200 BC, NITG).

phase

Copyright © 2010. Leiden University Press. All rights reserved.

Bell Beaker

age (yrs cal BC, 2σ)

surface (m2)

attachment point (m -NAP)

1800

Vlaardingen 2b

2570-2470

Vlaardingen 1b

3260-2960

Vlaardingen 1a

3500-3300

Hazendonk 3

2.10 760

2.55

3670-3610

730

3.50

Hazendonk 2

3910-3790

300

3.80

Hazendonk 1

4020-3960

800

4.30

Table III.1 The Hazendonk, age, extent and the attachment point of the refuse layers in unit C (Raemaekers 1999 based on Verbruggen in prep.). The dates are reconstructed by conventional 14C dating and by interpolation of ground water levels of other sites in the region (see Verbruggen 1992). The attachment point is the height where the peat is attached to the sand of the dune, influenced by the ground water table.

100

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Hazendonk 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

14

15

29

12 11

28

10

33 23 30

II

45

24

E

52

44

51

22

-5.0 -4. 0 -3.0 -2.0

32

31 34 9

20

D

1 2

19 21

17

I

III 4

18

26

.0

-7.0 .0 -6

-1.0

43

36 37 40 35 8 42 39 41 53 57

7

-8

25

3

-9.0

6

5

B

A

-8.

C

0

-7.

0

0

5

36

37

40

35

8

42

39 M86

26

41 57

25 Van der Wiel

Copyright © 2010. Leiden University Press. All rights reserved.

Voorrips

M87

southern section square 57

0

53

20 m

southern section square 41 core 2

core 3

20 m

Figure III.2 The Hazendonk, the Pleistocene subsurface (m –NAP) and the plan of the excavation showing the location of the units (A-E), squares, cores and sample boxes (archives National Museum of Antiquities).

101

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Hazendonk 0 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

14

15

29

12 11

28

10

33 23 30

II

45

24

52

44

51

22 -1.0

43

31 34 9

32 20

1

-5.0 -4. 0 -3.0 -2.0

-7.0 .0 -6

2

17

19 21 I

III 4

18

.0

26

36 37 40 35 8 42 39 41 53 57

7

-8

25

3

-9.0

6

5

-8.

0

0

-7.

0

20 m

Hazendonk 1 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

14

15

29 II

45 44

12 11

28

10

33 23 30

51

22

-5.0 -4. 0 -3.0 -2.0

31 34 9

32 20

1 2

17 18

19 21 I

III 4 26

-8

.0

-7.0 .0 -6

-1.0

43

-9.0

24

52

3

5 25

6

7

36 37 40 35 8 42 39 41 53 57

-8.

0

-7.

Copyright © 2010. Leiden University Press. All rights reserved.

0

0

20 m

Figure III.3a The Hazendonk, the Pleistocene subsurface (m –NAP) and the interpretation of the distribution of refuse for the various occupation phases, based on finds of archaeological indicators obtained by coring and on the presence of datable pottery in the squares (after Van Dijk et al. 1976 and archives National Museum of Antiquities). The refuse layers are indicated in grey, while high densities of finds are indicated in dark grey. Phases Haz 0 and 1.

102

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Hazendonk 2 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

14

15

29

12 11

28

10

33 23 30

II

45

24

52

44

51

22 -1

43

31 34 9

32 20 1

-2

-3

-4

-5 -6

2

17

19 21 I 4

18

-7

III

-8

26

36 37 40 35 8 42 39 41 53 57

7

25

3

-9

6

5

-8

0

-7

20 m

Hazendonk 3 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

14

15

29 II

45 44

52

12 11

28

10

33 23 30

51

22

-5.0 -4. 0 -3.0 -2.0

31 34 9

32 20

1 2

17 18

19 21 I

III 4 26

-8

.0

-7.0 .0 -6

-1.0

43

-9.0

24

3

5 25

6

7

36 37 40 35 8 42 39 41 53 57

-8.

0

-7.

Copyright © 2010. Leiden University Press. All rights reserved.

0

0

20 m

Figure III.3b The Hazendonk, the Pleistocene subsurface (m –NAP) and the interpretation of the distribution of refuse for the various occupation phases, based on finds of archaeological indicators obtained by coring and on the presence of datable pottery in the squares (after Van Dijk et al. 1976 and archives National Museum of Antiquities). The refuse layers are indicated in grey, while high densities of finds are indicated in dark grey. Phases Haz 2 and 3.

103

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Vlaardingen 1b 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

29

14

15

12 11

28

10

33 23 30

II

45

24

52

44

51

22 -1.0

43

31 34 9

32 20

1

-5.0 -4. 0 -3.0 -2.0

-7.0 .0 -6

2

17

19 21 I

III 4

18

.0

26

36 37 40 35 8 42 39 41 53 57

7

-8

25

3

-9.0

6

5

-8.

0

0

-7.

0

20 m

Vlaardingen 2b 56 55

48

47

-4.0 -3.0

49

50

46

13 38

16

54

14

15

29 II

45 44

12 11

28

10

33 23 30

51

22

-5.0 -4. 0 -3.0 -2.0

31 34 9

32 20

1 2

17 18

19 21 I

III 4 26

-8

.0

-7.0 .0 -6

-1.0

43

-9.0

24

52

3

5 25

6

7

36 37 40 35 8 42 39 41 53 57

-8.

0

-7.

Copyright © 2010. Leiden University Press. All rights reserved.

0

0

20 m

Figure III.3c The Hazendonk, the Pleistocene subsurface (m –NAP) and the interpretation of the distribution of refuse for the various occupation phases, based on finds of archaeological indicators obtained by coring and on the presence of datable pottery in the squares (after Van Dijk et al. 1976 and archives National Museum of Antiquities). The refuse layers are indicated in grey, while high densities of finds are indicated in dark grey. Phases VL 1b and VL 2b.

104

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

Phase VL 1b is characterised by a trampling horizon stretching into the marsh over a distance of c. 20 metres. Unit C is the only unit where all occupation phases except for phase VL 2b were attested. In addition to the established phases, the prospection by coring suggested the presence of a phase Haz 0, which was earlier recognised in the pollen diagram of Voorrips (Louwe Kooijmans 1974). The estimated age of this phase is 4550-4495 BC (5700 BP; Van Dijk 1979, 40). Even earlier occupation dating to 6000-5650 BC may have occurred as well (6900 ± 100 BP, charcoal, Van der Woude 1983, 47). Archaeological remains of these early phases are not known since the water table did not allow excavation at the relevant depths. The cultural interpretation of occupation at the Hazendonk has developed through time in correlation with the general understanding of the neolithisation process in the Dutch wetlands. The pottery found in layers 1 and 2 is nowadays considered as representing the southern group of the middle phase of the Swifterbant culture, the pottery found in layer Haz 3 as characteristic of the Hazendonk group, and the pottery found in the upper layers as characteristic of the successive phases of the Vlaardingen group and Bell Beaker culture. Some pottery of the phases Haz 2, Haz 3 and VL 1a shows influences of the Michelsberg culture (Louwe Kooijmans 2005, 258). During all phases, subsistence at the Hazendonk was based on a combination of hunting, fishing and some animal husbandry. Domestic animals present are dog, cattle, pig, and sheep/goat. Predominantly beaver and otter were hunted during all phases, and wild boar, red deer and roe deer additionally. Wild animals became more important during each following phase except for phase VL 2b (Zeiler 1997). Detailed seasonality data are available from the zoological research (Zeiler 1997). These data give indications of occupation during all seasons for almost all phases. The site may represent a supportive special activity site that was occasionally occupied more permanently. Occupation on a year-round basis and the use of the site as a seasonal or permanent base camp may have been possible, but this cannot be demonstrated (Louwe Kooijmans 2007; Raemaekers 1999, 117). Several archaeobotanical studies on the Hazendonk have already been published. Published pollen diagrams of the Hazendonk are diagrams of Voorrips (Louwe Kooijmans 1974), Van der Wiel (1982) and Van der Woude (1983). These diagrams are based on cores that are located on a distance of some metres up to km away from the dune. furthermore, a selection of pollen curves was published (Louwe Kooijmans 1987), as well as a macroremains diagram from phase VL 1b (Bakels 1981, Out 2008c). A considerable quantity of data analysed in the years after excavation remained, however, to be published. All available results were reinterpreted in 2003-2007 (see also appendix IV). The aim of this appendix is to reconstruct the natural vegetation and human impact on the dune and its surrounding slopes, and to reconstruct the plant subsistence. This will be investigated through the analysis of pollen, non-pollen palynomorphs and macroremains from on-site sampled cores and sections, and macroremains, wood, moss and molluscs from the excavation. All presented material is sampled at a short distance from the spot(s) where human activity took place, in contrast to most data published earlier. Research questions discussed are: • • • • • •

What was the composition of the natural vegetation on the top and slopes of the dune? How did human occupation influence the natural vegetation? How do occupation intensity and distance influence the signal of human impact in the pollen diagrams? Which plants functioned as use plants, as food plants and as arable weeds? Did crop cultivation take place at the dune? Do the botanical remains give information on seasonality or site function?

105

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

III.2 MaterIals and Methods In the first place, A. Louwe Kooijmans-Bouhuijs collected samples from refuse layers with sample boxes and cores. This material was subjected to the analysis of pollen, non-pollen palynomorphs and macroremains in 1974-1981 in order to reconstruct the development of the vegetation and human impact before, during and after separate occupation phases. The analysed material comprises sample boxes from square 57, core 3, core 2, and sample boxes M86 and M87. The locations of the cores and sample boxes are shown in figures III.2, III.4 and III.5. It was not possible to sample a single continuous core that contained all refuse layers because not all occupation layers were present everywhere and because of natural irregularities in the sediment such as tree falls and large quantities of wood remains. The southern section of square 57, unit C, was investigated to provide a detailed sequence of the phases Haz 1 up to and including Haz 3. It was sampled with two sample boxes (100 x 10 x 10 cm and 50 x 10 x 10 cm). The sample interval of this column was mainly 4 cm, while a part of the material was sampled every 2 cm. Core 3 (length c. 1 metre) was analysed to cover the phases Haz 0 and 1 in detail. It was sampled with a corer on a distance of 3.5 metres from unit C and was analysed every 4 cm. Core 2 (length 1 metre) was analysed to provide the sequence of the phases VL 1a and VL 1b. It was sampled with a corer just outside square 57, next to and above the location of the sample boxes of square 57, and was analysed every 4 cm. The top of the sample boxes of square 57 and the bottom of core 2 are stratigraphically correlated. M86 and M87, which both cover phase VL 1b, were collected from the eastern section of square 25, unit B with sample boxes (20 x 20 x 10 cm) and were analysed every 2 cm. The distance between M86 and M87 is c. 3 metres. Dating of the cores and sample boxes was primarily based on the age of the refuse layers, while additional 14C dates were obtained from the section of square 57 and from M87. The pollen samples, prepared according to the standard methods, had a size of 1 cm3. A. Louwe Kooijmans-Bouhuijs identified the pollen using a magnification of 400 and 1000 times.25 The pollen sum consists of 300 to 400 upland pollen grains, including dryland trees, shrubs, herbs, spore plants and crop plants. Data were analysed and converted into percentage diagrams with the software programs Tilia (2.0.b.4) and TGView (2.02) (Grimm1991-1993, 2004). The diagrams are divided into biostratigraphical zones, indicated with straight lines, which are based on changes in the diagrams and on comparison with the pollen diagram of Van der Wiel (1982). The zones are also applied in the diagrams of non-pollen palynomorphs and macroremains from the same sections and cores. Occupation phases, indicated with grey zones in the diagrams, were recognised by observations of the sediment during excavation, the presence of charcoal, sand and bone remains in the sediment of the sample box and changes in the curves of the analysed diagrams. A. Louwe Kooijmans-Bouhuijs also analysed non-pollen palynomorphs (NPP’s) of square 57, core 3 and core 2.26 Data were analysed and converted into percentage diagrams, based on the pollen sum of the pollen diagrams. The NPP diagrams only include identified types. Each diagram nevertheless contains a curve of the total of NPP’s that shows the total percentage of both identified and unidentified NPP’s. 25 Identification was based on literature (Beug 1963; Erdtman et al. 1961; Fægri and Iversen 1964; Grohne 1957; Nilsson and Praglowski 1963; Punt et al. 1976) and the reference collection of the Institute of Prehistory Leiden (now the Faculty of Archaeology, Leiden University). Revision resulted in removal of Araceae and Cornus suecica, lumping of Populus cf. nigra and Populus cf. tremula, lumping of Sparganium emersum and Typha angustifolia in Sparganium emersum-type, lumping of Lotus corniculatus and L. uliginosus in Lotus-type and changing Cerealia into Cerealia-type. The remaining old nomenclature is mainly left unchanged for optimal correspondence with the identification literature, although family names are given according to the modern standards (e.g. Brassicaceae instead of Cruciferae). Cerealia-type pollen grains were identified with a phase-contrast microscope, and their identification was based on the diameter and shape of the pollen grain, the characteristics of the pore and annulus, the thickness of the wall and the shape of the pollen grains. Large pollen grains of grasses (> 37 or 40 μm) that were identified as wild grasses are included in the curve of Poaceae. 26 Identification of NPP’s was based on Van Geel (1978), Van Geel and Aptroot (2006), Van Geel et al. (1981), Van Geel et al. (2003), Pals et al. (1980), Prager et al. (2006) and Van der Wiel (1982). The identifications of Spirogyra sp., Mougeotia sp., Zygnema-type and Pediastrum sp. include various subgroups (including single species). These subgroups are not included separately in the diagrams since the restricted ecological information on modern species does not allow relating the identifications of subfossil subgroups to specific ecological conditions.

106

unit C

57

B

41 2 3

A

0 peaty clay sand humic sand

peat clayey peat clay

1m

bark refuse layers

wood

57

2

Ha

b

Figure III.4 The Hazendonk, north-south section across unit C and cored extension, showing the location of the sample boxes from the southern section of squares 41 and 57, and the location of cores 2 and 3 (archives National Museum of Antiquities).

4

3

41

1a z3

VL Ha

2

A

z2 Ha

1b VL 2a Ha z

m - NAP

z1 z0 Ha

Copyright © 2010. Leiden University Press. All rights reserved.

3

B

7

6

5

4

3

2

m - NAP

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

107

0

50 c m

peaty clay sand peaty sand

peat

clayey peat clay

refuse layer VL1b

87

Figure III.5 The Hazendonk, unit B, eastern section of square 25, showing the location of the sample boxes M86 and M87 (archives National Museum of Antiquities).

5.5

5.0

4.5

4.0

3.5

3.0

2.5

m - NAP

86 M

108 M

Copyright © 2010. Leiden University Press. All rights reserved.

5.5

5.0

4.5

4.0

3.5

3.0

2.5

m - NAP

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

W.J. Kuijper analysed the macroremains of the section samples and cores. The sample size of the samples of square 57 was 1 cm3. The thickness of the samples of core 3 varied between 3-6 cm, and the thickness of the samples of core 2 was 5 cm, while the diameter of both cores was 4 cm. The results of the cores 2 and 3 were calculated for a standard volume of 12.5 cm3. The volume of samples of M86 and M87 was 50 cm3. In addition to the macroremains of the sections and cores analysed for pollen and non-pollen palynomorphs, macroremains were also analysed from the southern section of square 41 in unit C, located 2 metres north of the southern section of square 57, c. 1 metre higher on the slope (see fig. III.2). This section yielded macroremains from the largest number of phases from a single sample location. The section was sampled with two sample boxes, each measuring 100 x 20 x 10 cm. The thickness of the samples of this section varied between 3 and 10 cm, while the volume of the samples varied between 350 and 1400 cm3. The number of macroremains was calculated for a standard volume of 1000 cm3. All macroremains samples from sections and cores were wet-sieved on a 0.25 mm sieve. Details on identification methods of the macroremains are discussed below. All data were turned into diagrams that show absolute values of macroremains. In addition to macroremains from cores and sections, macroremains were collected during excavation in three ways. The first class of samples consists of visible concentrations of carbonised cereals of phase Haz 1, some of which were wet-sieved in the lab on a 0.25 mm sieve. The second class consists of a part of the residue that was collected in search of archaeological finds by sieving with mesh widths of 10, 5 and 1.5 mm. The botanical samples were primarily collected from the 1.5 mm sieve while a selection of finds of the other mesh widths was sometimes added. Some of these samples were only scanned on the presence of cereal grains resulting in the identification of relatively large macroremains only. furthermore, the relation of these samples to the excavated volume is not documented and the density of macroremains could thus not be calculated. The third class consists of handpicked finds and finds collected from the coarser meshes (10 and 5 mm) of the field sieving procedure. The number of samples from phases Haz 2 and VL 2b is small and there are no samples of the Bell Beaker period because of limited occurrence of these horizons. Some samples could not be dated precisely by stratigraphy. The samples of phase VL 2b were divided in two activity periods, which is based on stratigraphy but which is not supported by 14C dates that provide similar results for both periods. W. J. Kuijper identified all macroremains between 1974 and 1981.27 C.C. Bakels and R.T.J. Cappers revaluated all cereal remains in 2004. Seeds of Elatine sp. were identified up to species level in 2006 (see Brinkkemper et al. 2008). Names of plant species are according to Van der Meijden (1996). W.J. Kuijper also identified molluscs from macroremains samples. A. Touw identified moss remains that were found in the macroremains samples. The macroremains are divided into ecological groups, based on the modern plant communities (Schaminée et al. 1995-1999) and on the interpretation of the vegetation. Table III.2 shows the distinguished ecological groups. This classification is only a method to analyse changes in the vegetation, since the distinction between some groups is not sharp and since some taxa are more or less characteristic of more than one plant community. The taxa of groups 4 and 5 in particular can occur in the same plant community as well as in other plant communities. Prehistoric plant communities are additionally not necessarily comparable with modern plant communities. Group 5, representing marsh taxa, also contains taxa of sedge vegetation and 27 Identification was based on literature (Aalto 1970; Bang 1973; Berggren 1969; Brouwer and Stählin 1975; Dickson 1970; Jessen 1955; Katz et al. 1965; Knörzer 1970, 1973; Kowal and Rudnicka-Sternowa 1969; Körber-Grohne 1964, 1967; Lange 1979; Langhe et al. 1978; Nilsson and Hjemquist 1967; d’Olivat and Pals 1974; Ooststroom and Reichgelt 1964; Villaret-von Rochow 1967; Van Zeist 1974) and the reference collection and literature of the Institute of Prehistory Leiden. A single find of Alisma plantago-aquatica was initially identified as Damasonium alisma (section square 41, 331-326 cm -NAP). This is presented as A. plantago-aquatica since the difference between the two species is not always clear and since it concerned only a single find (see also De Roller et al. 2002). The Tilia remains were identified as Tilia platyphyllos but presence of T. cordata could not be excluded. Bromus secalinus-type represents B. arvensis, B. mollis and B. secalinus. Galeopsis bifidatype represents G. bifida, G. speciosa and G. tetrahit. Ranunculus aquatilis-type represents Ranunculus section Batrachium. Ranunculus repens-type represents R. acris/lingua/repens but it probably concerns R. repens here. Veronica beccabungatype represents V. anagallis-aquatica, V. beccabunga and V. catenata.

109

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

ecological groups 1) Taxa of woodland, woodland edges and shrubs of dry terrain. 2) Dryland ruderals and taxa that indicate recent disturbance of dry to slightly moist sediment. 3) Crop plants. 4) Wetland trees and shrubs. 5) Wetland herbs and spore plants of marsh and forb vegetation. 6) Wetland pioneers and wetland herbs that indicate disturbance. 7) Taxa indicative of open water. 8) Taxa that are not associated with specific ecologic conditions. 9) Varia, including archaeological remains, macroscopic plant remains such as bark, leaves and moss, animal remains and sand. Table III.2 The Hazendonk, ecological groups that have been distinguished.

depth (m -NAP)

sediment

3.87-3.84

coarse peat

4.01-3.88

peaty clay with wood remains

4.02 4.30-4.03

fine-layered peat layered peaty clay, wood remains from 4.15 m -NAP,

Copyright © 2010. Leiden University Press. All rights reserved.

darker layers at 4.17 and 4.11 m -NAP 4.38-4.31

sandy peat with charcoal, wood remains and bone remains: Hazendonk 3

4.42-4.39

peaty clay with roots

4.47-4.43

sandy peat with charcoal remains in the upper part, burned bone at 4.47 m -NAP

4.55-4.48

peaty clay, disturbed by a tree fall, some sand and charcoal in the lower part

4.62-4.56

fine peat with charcoal remains

4.67-4.63

black peat with sand and bone remains: Hazendonk 2b

4.69-4.68

clayey peat

4.72-4.70

peat with charcoal remains

4.85-4.73

clayey peat, Hazendonk 2a at 4.78 m -NAP

4.99-4.86

black peat with sand, charcoal, bone remains and a sherd: Hazendonk 1, bark remains at 4.97 to 4.95 m -NAP

5.03-5.00

transition

5.22-5.02

peat with wood remains

5.28-5.23

sandy peat

5.33-5.29

humic calcareous sand, Hazendonk 0 at 5.30 m -NAP

Table III.4 The Hazendonk, unit C, southern section of square 57, lithostratigraphy.

110

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

eutrophic/mesotrophic grasslands, since the number of taxa in these groups is low and since most of these taxa can also grow in the classes that are included marsh vegetation. R.G. van den Berg identified wood remains from the excavation, and the results are based on an unpublished manuscript of his hand. Wood remains were retrieved from 14 squares that were mostly located around unit C, but also near units D and E and at the southwestern side of the dune. The data set consists of unworked wood, worked wood with an unknown function and artefacts with a known function. Most wood remains were collected individually and are described. There are furthermore three samples of an unknown volume that consists of several wood identifications on which additional information is absent. The assemblage of wood consists of trunks, branches and twigs, bark and roots. Bark and roots were not identified but are included in the group of indeterminate instead. Wood remains that were partly carbonised were not identified either, although there are a few exceptions. further charcoal identifications are absent. III.3 r esults III.3.1 Southern SectIon of Square 57 (hazendonk 0, 1, 2a, 2b and 3) Table III.3 (below) shows the radiocarbon dates from the section of square 57 (Lanting and Van der Plicht 2000). The results correspond with the dates of the refuse layers. Table III.4 (opposite page) shows the lithostratigraphy of the southern section of square 57. figures III.6-8 show the results of pollen analysis, non-pollen palynomorphs and macroremains of the southern section of square 57. The recognised occupation phases are Hazendonk 0, 1, 2a, 2b, 3, and possibly Vlaardingen 1a.

sample

depth (m -NAP)

lab code

age (yrs BP)

age (yrs cal BC, 2σ)

dated material

Haz 29f

3.86-3.84

GrN-8243

4720 ± 70

3640 (95.4%) 3360

peat

Haz 29e

4.33-4.31

GrN-8350

4755 ± 35

3640 (81.7%) 3500

peat

3430 (13.7%) 3380 Haz 29d

4.45-4.43

GrN-8242

5080 ± 70

4040 ( 1.1%) 4020

peat

4000 (94.3%) 3700 Haz 29c

4.64-4.62

GrN-8349

4995 ± 35

3940 (22.7%) 3850

peat

Copyright © 2010. Leiden University Press. All rights reserved.

3820 (71.5%) 3690 3680 ( 1.2%) 3660 Haz 29b

4.86-4.84

GrN-8241

5385 ± 45

4340 (61.8%) 4220

peat

4210 (17.1%) 4150 4140 (16.5%) 4050 Haz 29a

5.03-5.01

GrN-8240

5330 ± 35

4320 ( 2.0%) 4290

peat

4260 (93.4%) 4040 Table III.3 The Hazendonk, unit C, southern section of square 57,

14

C dates (Lanting and Van der Plicht 2000).

111

Copyright © 2010. Leiden University Press. All rights reserved.

112 or

20 40

yl us

20 40 60 80 100

20

Figure III.6 part 1. 20 20 40

Upland shrubs

20

Peat

Sand

Clay

Peaty sediment

Sandy sediment

Humic sediment 20 40 20

li x M y R r ic a h H am u n R mu us ib lu f r C es s an al gu la Lo yst ni e g c e ia ra

535

lm

nu

s

us

or

e

pl

an

us Be t u Po l a Ac pul e u Ta r s cf .n xu Ile s ig ra x Ju /c f. ni t re Fa pe m gu r u ul C s s a ar H pin ed u V i er s sc a um

U

rc

xi

ue

Fr a

Q

Ab Ti ies lia

U p U lan pl d a s Pi n d hr u c e he b s Pi a r b s nu an s d

sp

D a e p te s t h (y (c rs B Li m th - N P) ol AP og U ) pl y an d t re es D

14 C

4720±70 385

Sa

bu rn u R ha m Li m gu nu S a st s r c C m b um at h o u ar tic H r n u c us ip s a p s Al o a liu p h n g m ae u i ne a Al liu Ar m t C em vine e is C rea ia ale -t y h li Ja eno a - t pe si po yp Pl o n d e an e ia ce Pl t a ae a g G nt a o l a g a Po leo o mnce p o G l yg o s i s a j o l at en n r a O is um x ta R alis - t y p e es a p r s R ed c e e ic a u a t r ia R me lu ose -t y a x te l pe S p nun ac a - la e r c u et o t y p Po g l u s e u s a P t ly p o la - t ar - t y p e r d y p ve e A l i d iu e ns nu ium m is s

Vi

C

ts

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees

395

405

415

425 4755±35 435 5080±70 445

455 4995±35 465

5385±45 485 475

495 5330±35 505

515

525

20

Upland herbs and spore plants Wetland trees, shrubs and climbers

1000 2000 3000 20

Copyright © 2010. Leiden University Press. All rights reserved.

20 40

20 40 60 pe

Ecologically indeterminate

ia c A s eae te A s ra t c Br era ea a c e C s s ea li g a ic e u C r yo a c e t u b l i f l o e p u r C n t a hy l a e l i f l a e am ur la or ae Er p e a c e ae ic an a Eu l u l e a Eu p h r s a Eu p h o s i a r G r um bia e e H nt i a x y n M per a p y ic n M o s o um eu e t m L i nt h i s p e r o n lli a fo an a ra th O c ty do ea pe tu e m -t y Fa nt e -t y p b it pe e R ac e s os e a R ac e a e R nun ae u c S t bia ula a c c S e c h eae ea y e Sc dums s yl r va Tr op tic i h aTe foli ula um r i ty u pe V i cr - t a-t c i ium y p y p a e e -t y

Open water

Ap

C a H llit r o ic N t to n h e u i N pha a y r Po m p h Le t am ae a m og Eu n a e to n S t p ot cf r a .p Tr atio mo ec ap te g e t in M a s to y r na n at us M iop t a yr h n Po i o p y l l s u a c hy m e a l l u ve e m rt sp ic i ic lla at t u C um m yp er ac ea e 20 20 20 40

U rt Va ica , Spleri lar a g M hag na e on n o um M l et a on e, cf o . le ps O D r y t ae i l at p o ae O h i o pte sm gl r i un o s s s c da um ar th us

ia

na

Al is Bu ma c f to m . u Fi C o s li r r H pen igio yd d l c f ro ula a . c H El a h a r yd t i i s Ir i ro ne s c Ly ps tyle t e Lo hr u u da tu m c o s ru Ly -typ s si e Li m t to ac M re hi e l a Po l a m l a v u lg p ar Pa l yg o y r u is -t y Pr rna numm pe im s s b Pe u i a i s la to d R i c u ve r ta a r S a nun lar i is - t s yp c S pngu ulu e i s S pa r g s o r r e a b p ar n a e g iu o n S o ani m f f ic s u e i S c l a n m em e n a l r s is he um S y u d re c um tu ze u l m Sa p r i c a m -t yp hy a m e g i t Sa t t u ar ar m a x T h if r ia a a Ty lic ga ph t r u gra U a m n r t i la ul c a ti at fo a lia

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Wetland herbs and spore plants

20

20

Figure III.6 part 2.

113

297 344 266 325 342 325 334 383 369 376 380 450 402 458 421 371 350

293 448 280 358 377 331 380 393 344 482 307 374 316 359 293 422 369 344 302 443 350 357 318 354 404

se ha /P ne Zo

Va V i ler i o a Ve la p nel r a la Ep o n l u s ic t Eq h e d a r i s ty ui r a pe H se d up t ys u D er m t a ip z c hi C ha ia s ay s e ty S e sto p i ast l a g pe r o t l u Pl agi e r i m an ne s / L Po t a l l a yc op lle g o n co od su r iu o m n m op cl us av a

Ecologically indeterminate

tu

m

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

? ? 57-IV Haz 3 57-III Haz 2b Haz 2a 57-II Haz 1 57-I Haz 0

Analyst: A. Louwe Kooijmans-Bouhuijs, 1977, 1979

Upland vegetation

Figure III.6 part 3 The Hazendonk, unit C, southern section of square 57, pollen diagram based on an upland pollen sum, exaggeration 5 x. Figure III.12 can be seen as a continuation.

ol

th

385

Li

D

ep

th

(c m

-N A o C P) o r gy n V u ib s Pe u r s a n n C r si c um g u h a o i C e n o r ia pu n e a h l T. eno p o d lap us a d A ic o p o d i u m t hi ln c iu a fo us c o m l b lia g l n, f i u m ut s c i in pik fo A o s e liu ln a l et m U us fo rt g rk ic lu a ti ,c Ly di n o ar th oi sa bo ru ca , ni c m se S o o d ne s Eu l a n a l i s c u A p at m d a r i a li o S sm r iu ulc te a m a M lla pla c a ma e r r Ve nt h ia mnt a nna a g Ty ron a a edi o - abin p i q a q um ua C h a c a u at ar s be ic tic C ex p. c c a / a al s a ab r v R lit r p. un en an ic s g un he a - is c u sp ty pe lu . Zo s aq ne ua /P til ha is se -t yp e

Wetland trees and herbs

395

?

405

?

415

57-IV

425

Haz 3

435 445

57-III

Copyright © 2010. Leiden University Press. All rights reserved.

455 465

Haz 2b

475

Haz 2a

485

57-II Haz 1

495 505 515 525

57-I Haz 0

535

20 40

20

20 Analyst: W.J. Kuijper, 1977, 1979

114

Peat

Sand

Clay

Peaty sediment

Sandy sediment

Humic sediment

Figure III.7 The Hazendonk, unit C, southern section of square 57, macroremains diagram. Figure III.13 can be seen as a continuation.

Copyright © 2010. Leiden University Press. All rights reserved.

C lo A s ste s r A s uli ium n s Ty uli a m idio p e na us sp Ti 5 s c or ll 9 em o r u Ty etia in um m ( p ul (t ty C e 1 sph um y p p e la 5 a (ty e 3 60 c f ste 0 g n . D ro i( pe 2 A ) ty H ip sp 32 ) p el lo o e B) B a ic o c la r iu 27 m o d ) H c c il n p iel c a ys la lu la r c f t r i r i o r is sc ic in . c p p a u C Typ hos hyc eta lar m ha e p e tu oi (t c f et 6 6 h a a e m d e y p e er ( (t s, . o i d di a y p c 25 M Pl e m i a ae to e o n ) yc o s l e m 3 0 i di r s A l o po s) ) a t l n hy ra (E on r i s M -p um p. A ( ol s t y ty le p. p pe n (t e pa y p 3 B 19 ) ly e ) no 8 m B) or ph s Zo ne /P ha se ol

ep th (c m

-N AP o Sp gy ) Spiro g y Spiro g ra y s M irog ra p., ou y sp ps g ra ., ila M e o t s p. n o t t a e ou ia , o p Zy ge la rn sil e a a g c f n e o t i a t ev m e t a e ir n .T m s Pe yp a- p., ens tat io di e 3 t y p ot Al as 4 e he (typ n u ga tr 1A e nk r s 37 n Ty e um 3) ow p , c f e 2 u n i s p. n . T 8 de nt c f yp .C e if i ed C a 14 e ly 5 c f r ato s te . p g Va Typ hy ia s l s e 7 l l u p. ar 0 m , c ia 8 sp a va . tk r i o , sp ins Ty in sp es pe or (ty a18 Ty ty pe pe pe cf 1 (ty 137 . 2 ) pe Ty Typ 8B 14 p e D e 1 12 0) ip 4 2 c f oro 2 . T th U yp ec st e a ul 3 r i 0 h H n a 7 i zo ys d ph te e u ila G r ium st a (ty ae ( pe um cf. typ e 14 p R 3) ha ann ulic 44) Ty bd om ar pe oc yc e ( es t yp 3 6 oe G e cf e l 1 la ( 1 . T a G si ur ca 21) el no r b el ic is c f as sp la . in ria (ty Ty Typ os ora p ,t pe e po c f yp e 1 Zo 7 12 ra . r 2 e e 4 8 p 35 6) A s p. t i c D f ie 3) in lla (ty uli of s pe po la lun ge d 1) ra lla qv (ty t a i st pe ii 2) (ty pe 50 1)

th

385

Li

D

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

395

405

415

425

435

445

455

465

475

485

495

505

515

525

535

20

20 40

Peat

Sand

Clay

Peaty sediment

Sandy sediment

Humic sediment 20

Haz 0

50 100 150

Analyst: A. Louwe Kooijmans-Bouhuijs, 1977, 1979 20 20 20 20

?

57-IV ?

Haz 3

Haz 2b 57-III

Haz 2a

57-II Haz 1

57-I

Figure III.8 The Hazendonk, unit C, southern section of square 57, non-pollen palynomorph diagram based on an upland pollen sum, exaggeration 10 x, + = few (1-10), ++ = some tens (10-49), +++ = many tens (50-99). Figure III.14 can be seen as a continuation.

115

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Zone 57-I (5.33 to 5.06 m -NAP) is based on high values of Pinus sp. and Tilia sp. and low values of Corylus sp. and Betula sp. Important taxa are Tilia sp., Quercus sp., Corylus sp., Poaceae, Cyperaceae and ferns (monoletae psilatae spores, possibly from Thelypteris palustris). Tilia sp. was certainly present in the local and extra-local vegetation, as indicated by the high percentage of pollen. Pinus sp. shows high values but was probably an element of the regional vegetation only, since local presence would result in much higher percentages of pine pollen and since macroremains and wood remains are not found at HardinxveldGiessendam Polderweg, Hardinxveld-Giessendam De Bruin and Brandwijk-Kerkhof. A variety of dryland shrubs was present in the (extra-) local vegetation, such as Viburnum sp., Rhamnus cathartica, Sambucus sp. and Cornus sanguinea. Alnus sp. was probably not present in the local vegetation yet, but grew nearby (see core 3). The variety of pollen of wetland herbs indicates that marsh vegetation was already present (for example Urtica dioica, ferns, Sparganium emersum-type and Sphagnum sp.), although the low curves indicate limited importance. Macroremains are relatively scarce in this zone. NPP’s that dominate this zone are Valsaria variospora-type (type 140), type 1828, type 128B and Diporotheca rhizophila (type 143). Type 128B, 140 and 143 indicate mesotrophic to eutrophic, moist to wet conditions (Van Geel et al. 200329). The total number of NPP’s shows a peak in the upper part of the zone. This peak corresponds with an increase in Pinus sp., Fraxinus sp. and Betula sp., and low values of Quercus sp. and Corylus sp., and may be related to changes in the water regime (see also core 3). The base of zone 57-I shows a weak signal of phase Haz 0 (5.30 to 5.27 m -NAP). During this occupation phase, the pollen diagram shows a decrease in Tilia sp., relatively high values of Quercus sp., Corylus sp. and Viburnum sp. and small peaks in the curves of Artemisia sp., Chenopodiaceae (delayed), Plantago major, Pteridium sp., Urtica sp., Apiaceae and Rosaceae. The anthropogenic influence on the vegetation seems to be small. The diagram of square 57 possibly reflects only a part of phase Haz 0 since the same occupation phase is represented stronger in the pollen diagram of core 3. NPP’s that show a slight increase during phase 0 are type 74, cf. type 145 (see also Van der Wiel 1982), cf. type 128A and Ustulina deusta (type 44, a parasite causing soft-rot of wood; Van Geel and Aptroot 2006). These types are not known as typical anthropogenic indicators. Zone 57-II (5.05 to 4.90 m -NAP) is based on a decrease in Pinus sp. and Tilia sp. and temporarily peaks of Quercus sp. and Corylus sp. The zone reflects human impact during phase Haz 1, such as a decrease in Tilia sp., resulting in the increased presence, flowering and/or pollen transport of Quercus sp. and Corylus sp. These changes represent the composition of the very local vegetation since the peaks of Quercus sp. and Corylus sp. are not visible in core 3 that reflects the same period. The decrease in Pinus sp. is probably a statistical decrease, resulting from the increased pollen precipitation of Quercus sp. and Corylus sp. Dryland shrubs other than Corylus sp. do not show strong changes. The summary diagram shows a major increase in dryland herbs, corresponding with increasing values of Chenopodiaceae, Allium sp., Cerealia-type and the sporadic presence of other anthropogenic indicators. The openness of the vegetation may however be overrepresented. firstly, the Cerealia-type pollen may represent pollen from plant waste since concentrations of cereals were present in square 57. Secondly, the peak of Allium sp. may represent shade-tolerant vegetation or waste of food preparation (see also chapter 9). The Allium pollen may represent A. oleraceum A. schoenoprasum, Allium scorodoprasum, Allium ursinum and/or A. vineale, some of which tolerate shaded conditions. All taxa other than A. ursinum are known from present-day inland dunes along rivers in the Netherlands (Eenkhoorn and Smit 1982; Weeda et al. 1991; Wolf et al. 2001). Overall, it can be concluded that the dryland vegetation became more open during phase Haz 1, but not as strong as the summary diagram suggests since the increase in dryland herbs is overrepresented. The signal of human activity in the wetland vegetation confirms that phase Haz 1 is 28 Type 18 is found on Eriophorum vaginatum (Van Geel 1976), a species characteristic of acid, mesotrophic soils. There are however no direct identifications of E. vaginatum and the species does not fit well in the vegetation at the Hazendonk, suggesting that the type may represent another type, may have other hosts or may be deposited after water transport. 29 See also Van Geel and Aptroot 2006; Van der Wiel 1982; Van der Woude 1983 for type 143.

116

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

not substantially different from later major phases. A small rise in the curve of Alnus sp. and Humulus sp. and increased values of Poaceae, Urtica sp., Filipendula sp., Lythrum sp., Lotus-type, and Apiaceae indicate the increased presence of open patches in the vegetation. The Cyperaceae clearly decrease. The macroremains diagram mainly indicates the local presence of wetland taxa including Alnus glutinosa, possibly representing alder carr. The NPP diagram shows that phase Haz 1 is characterised by a decrease in previously common types, a low number of identifications and an increase in certain types such as Mougeotia laetevirens (type 373), type 145, type 708, Zopfiella lundqvistii (type 501) and cf. Diplocladiella scalaroides (EMA type 19).30 These types are again not known as typical anthropogenic indicators. Zone 57-III (4.89 to 4.39 m -NAP) is based on an increase in Pinus sp., Tilia sp. and Fraxinus sp. and high values of Quercus sp. in the diagram. The first part of the zone directly after phase Haz 1 reflects the recovery of the vegetation after occupation, as shown by the curves of Tilia sp., Quercus sp., Fraxinus sp., Corylus sp. and Alnus sp. The increase of some of these taxa may partly be explained by the statistical effect of the decrease in Corylus sp. and Allium sp. after the end of phase Haz 1. Tilia sp. only partially recovers, which is related to the increasing water table and the resulting decrease in the surface of the local habitat. Chenopodiaceae, Polypodium sp., Pteridium sp., Poaceae, Urtica sp., monoletae psilatae fern spores and Brassicaceae show peak values at the end of occupation, suggesting that disturbed patches were overgrown with fern and forb vegetation before the final recovery of the vegetation. The macroremains diagram shows the local presence of Alnus glutinosa and Urtica dioica in the first half of this zone. The sediment of the upper half of zone 57-III mainly consists of peaty clay (from 4.55 m -NAP onwards, see table III.4 for details). This part of the section is investigated in less detail since a tree fall prevented regular sampling. The presence of clay corresponds with a clear increase in Fraxinus sp., of which the local presence is confirmed by finds of macroremains in the section of square 41. Other changes contemporary with the increase in Fraxinus sp. are a decrease in Pinus sp., Tilia sp., Quercus sp., Poaceae and Cyperaceae, the presence of Ilex pollen, a slight increase in Plantago lanceolata, the start of the regular presence of Plantago maior, an increase in Urtica sp. and maximal values of Eupotamogeton. The NPP diagram shows a peak of Valsaria variosporatype and considerable values of some psilate Spyrogira sp., Mougeotia sp. and M. laetevirens. These changes in the pollen and NPP’s are probably related to an increase in the water table and/or the increased flooding frequency, although the decrease in Pinus sp. and Tilia sp. probably represent a statistical result of the increase in Fraxinus sp. The presence of archaeological remains at this depth (see table III.4) may be the result of disturbance by a tree fall or other non-stratigraphical deposition since the pollen diagram does not show changes in curves similar to those that occur during the recognised occupation phases. The increased presence of the Plantago species can be related to the instability of the natural environment, caused by the treefall and flooding (cf. Groenman-Van Waateringe 1968; Louwe Kooijmans 1974, 139; Weeda et al. 1988, 252 and further). Phases Haz 2a and 2b are present at 4.77 to 4.73 and 4.66 to 4.62 m -NAP. Both phases appear to be much shorter than phase Haz 1. Quercus sp. (phase Haz 2a), Fraxinus sp. (phase Haz 2b) and Cyperaceae decreased, while Tilia sp., Corylus sp. and Poaceae slightly increased. Quercus sp. may have been cleared. The unusual absence of a decrease in Tilia sp. may be related to the small scale of human impact and the dominance of Quercus sp. and Fraxinus sp. in the (very) local vegetation at the sample point. The curves of Artemisia sp., Chenopodiaceae, Polypodium sp., Pteridium sp. and Urtica sp. and the presence of Plantago lanceolata suggest that the effect of anthropogenic influence was stronger during phase Haz 2b phase than during phase Haz 2a. The NPP diagram shows increases in Spirogyra sp., Mougeotia laetevirens (type 373), Ustulina deusta-type (type 44), Hysterium cf. pulicare (type 121), Gaeumannomyces cf. caricis (type 126), type 361, Clasterosporium caricinum (type 25) and cf. Diplocladiella scalaroides (EMA type 19) amongst others. The total percentage of NPP’s is especially large during phase Haz 2a for unknown reasons. 30 See also Van Geel and Aptroot 2006; De Klerk et al. 1997 and Barthelmes et al. 2006; Ellis and Ellis 1997.

117

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

Zone 57-IV (4.38 to 3.84 m -NAP) is based on a decrease in Tilia sp., the presence of Fagus and Carpinus pollen (transported by river water), an increase in Ulmus sp. and the relatively smooth curves of dryland trees. The regular presence of Fagus pollen is assumed to reflect the transition from the Atlantic to the Sub-Boreal, contemporaneous with the start of the Hazendonk 3 occupation (5080 ± 70 BP, 3670 BC (see table III.1); cf. Van der Wiel 1982). The changes in the pollen rain at the Hazendonk can be seen as a confirmation of the starting date of the Sub-Boreal date as suggested by Zagwijn (1986) for the Netherlands in general (5000 BP, 3850 BC). In contrast to some other regions in the Netherlands however, Ulmus sp. does not decrease (see also chapter 2.8). Phase Haz 3 (4.37 to 4.27 m -NAP) is characterised by decreases in Tilia sp., Quercus sp. and Fraxinus sp. that are probably caused by human impact. Major increases in Corylus sp. and Viburnum sp. indicate the presence of open patches in the vegetation. The increase in these shrubs probably resulted in lower curves of Pinus sp. (despite continuous pollen rain). Other dryland taxa that increase are Rhamnus cathartica, Ligustrum sp. and Viscum sp. The latter could have grown on Tilia sp., Malus sp., Sorbus sp. (local presence not demonstrated), Crataegus sp. and Populus sp. (Westhoff et al. 1973). Similar to phase Haz 1, Cerealia-type pollen is present. Poaceae, Filipendula sp., Lythrum sp., Apiaceae, Brassicaceae and Asteraceae tubuliflorae slightly increase again, while the monoletae psilatae spores and Cyperaceae temporarily decrease. The macroremains diagram shows a large variety of taxa during phase Haz 3, representing both dryland and wetland taxa and also including a carbonised spikelet fork of Triticum dicoccon. The NPP diagram shows increased values of psilate Spirogyra sp., Mougeotia laetevirens-type (type 373) and Ustulina deusta-type (type 44). After phase 3, almost all taxa recover to previous values. Some taxa initially increase at the end of occupation (Artemisia sp., Cyperaceae, Poaceae, Urtica sp. and Solanum dulcamara) to fall back to previous values afterwards, indicative of the gradual recovery of the vegetation. Tilia sp. is not able to return to previous values due to the combination of human impact and the rising water table. In the middle of the zone, Alnus sp. shows a sharp peak, and the macroremains diagram confirms the local presence of alder vegetation. In the upper part of the zone, Ulmus sp. and Salix sp. gradually increase, indicating a raise of the water table and possibly more dynamic water activity. At 4.11 and 3.99 m -NAP, there are minor changes in the vegetation that might reflect an anthropogenic signal, such as a decrease in Tilia sp., Fraxinus sp., Corylus sp. and Cyperaceae, and (small) increases in Quercus sp., Hedera sp., Pteridium sp., Lythrum sp., Apiaceae, Brassicaceae and Caryophyllaceae. These changes correspond with earlier signals of human impact. The anthropogenic signal at 4.12 m -NAP furthermore corresponds with the presence of darker sediment. Gaeumannomyces cf. caricis, type 361, Zopfiella lundqvistii and Clastorsporium caricinum show small peaks at these spectra. All these taxa showed peaks during one or more of the previous occupation phases, also supporting that it concerns occupation phases. These changes are apparently all related to phase VL 1a, which represents all occupation between phases Haz 3 and VL 1b. The macroremains diagram shows a peak of Alnus macroremains after the presumed phase at 4.11 m -NAP, possibly indicating the recovery of alder vegetation. III.3.2 core 3, near unIt c (hazendonk 0 and hazendonk 1) Table III.5 shows the lithostratigraphy of core 3. A sample in the lower part of the core consisting of reworked peat was not analysed. The upper 10 cm of the core were not analysed because the sediment consisted of clay. figures III.9-11 show the results of pollen analysis, non-pollen palynomorphs and macroremains of core 3. Recognised occupation phases are Hazendonk 0 and Hazendonk 1. The core corresponds with the lower part of the section of square 57. Zone 3-I (6.13 to 5.98 m -NAP) is characterised by relatively high values of Tilia sp. and low values of Quercus sp. and dryland herbs. The percentage of Fraxinus sp. is higher than in the diagram of square 57. The pollen diagram and the macroremains diagram both show a combination of dryland and wetland taxa.

118

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

depth (m -NAP)

sediment

5.21-5.16

clay

5.55-5.22

coarse peat with wood remains, Hazendonk 1 at 5.40 to 5.31 m -NAP

5.86-5.56

coarse peat, sometimes sandy, wood remains at 5.79 to 5.78 m -NAP

5.87

sand, Hazendonk 0 at 5.90 to 5.80 m -NAP

5.93-5.88

coarse peat

6.01-5.94

sandy peat

6.11-6.02

peaty sand, reworked peat at 6.11 to 6.06 m -NAP

6.16-6.12

humic sand

Copyright © 2010. Leiden University Press. All rights reserved.

Table III.5 The Hazendonk, unit C, core 3, lithostratigraphy.

The macroremains diagram indicates the local presence of Alnus glutinosa and Cornus sanguinea. The NPP diagram shows high values of Tilletia sphagni, (type 27), Ustulina deusta-type (type 44), Valsaria variosporatype (type 140) and Diporotheca rhizophila (type 143), which corresponds with the section of square 57, except for the presence of T. sphagni. The macroremains diagram indicates the presence of charcoal and bone remains in the sediment of zone 3-I, but this waste was not interpreted as a refuse layer during excavation, which is confirmed by the absence of clear human impact in the diagrams. The transition to zone 3-II (5.97 to 5.53 m -NAP) is based on a decrease in Tilia sp., an increase in Quercus sp. and an increase in Apiaceae. Fraxinus sp. shows a peak during the increase in Quercus sp. The zone starts with phase Haz 0 (5.94 to 5.82 m -NAP). In the pollen diagram, Fraxinus sp. and Tilia sp. decrease during this phase, while Corylus sp., Viburnum sp., Chenopodiaceae, Rumex acetosa-type, Allium sp., Alnus sp., Urtica sp., Solanum dulcamara and Mentha-type show increased values. The decrease in Tilia sp. indicates that human impact occurred somewhat higher on the slope than at the sample point, since Tilia sp. was not the main taxon at the sample point. The macroremains diagram only shows minor changes that can be related to occupation, such as the presence of charcoal, the presence of Viburnum sp. and a decrease in Cornus sanguinea. Carbonised macroremains were not found in this core. The NPP diagram shows small peaks of Valsaria valsariospora-type (type 140), type 18, type 361 indicative of erosion of sand, fragments of wood vessels (type 114), Actinopeltis sp. (type 8C), Cladocera and Pleospora sp. (type 3B). In the following part of zone 3-II, Tilia sp. partly recovers and remained present in the extra-local vegetation. A variety of dryland shrubs was present, indicating somewhat open vegetation. The pollen and macroremains indicate the presence of alder carr. The NPP diagram shows the presence of algae and types that are indicative of moist to wet conditions and regular flooding. Gaeumannomyces cf. caricis (type 126) and Clasterosporium caricinum (type 25) indicate the common occurrence of sedges (Van Geel 1976, Van der Wiel 1982), which is confirmed by the macroremains diagram. In the upper part of zone 3-II the curve of Quercus sp. shows relatively low values, which is probably related to the water table, since the pollen diagram shows a slight increase in the percentage of water plants, increased values of Pinus sp., and relatively high values of Cyperaceae, type 140, type 143 and the total percentage of NPP’s. The peak in the NPP curve can also be observed in the diagram of square 57. The low values of Quercus sp. could also be explained by anthropogenic influence since bone remains are present at this level, but the diagrams do not support this by evidence of human impact.

119

Copyright © 2010. Leiden University Press. All rights reserved.

120

Peat

Sand

Peaty sediment

Sandy sediment

Figure III.9 part 1.

Humic sediment ty

pe

20 40

100 200 300

um M ulu yr s i R ca ha m Lo nu ni s c R er f ran ib a gu es la C al ys U te g r ti c a ia Ly t Sp hr u a m Ty rga ph n S p a ium a la So rg tifo em l an l e Ar anu iumia rsu ac m m -t y A l e d er is ae ul e c pe c a tu Bu ma m m to a C m ra on u F i vo s lip lv c f en ulu .E d s H la ula yd t i n Lo r o e tu c t y s - le

20

H

20 40 60 80 100

li x

615

pl

an

t re es

(c m -N AP )

rc

us

20 40 60

Upland herbs and spore plants

Vi b C ur n or u R nu m h s Li am s an g n g S a ust us c uin m r um at e a ha bu r ti cu ca s

xi nu U lm s u Be s t Po u l a pu Ac lu e s Ju r cf. ni ni gr H pe ed r a/ cf V i er us .t sc a re C um m or ul yl a us

Fr a

ue

pl

Q

U

d s a hr Pi n d ub c e he s Pi a r b s nu an s d sp Ab or ie e Ti s pl lia an

U

th

y

d

og

ep

an

ol

D

pl

th

U

Li

525

Sa

C h R eno um p A r ex o d ia t Po e m a c e c e a ly i sia to e sa A l go li n -t y C um um pe er pe e Ja a rs si lia ic Pl o n ar a e ia -t y P l nt a an go pe Pl t a m an go e G t a la dia e g n S p ni st o m c e e a a o Po r g - t y j o r l at a l ul pe Po yg o a - t ly nu y p e P t po m e d a A l r i di ium v i c nu um ul ar s ety pe ts

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees Upland shrubs

535

545

555

565

575

585

595

605

20 20

Wetland trees and shrubs

20

Wetland herbs and spore plants

Copyright © 2010. Leiden University Press. All rights reserved.

20

20 40

e

ce a R an e R un ub c Ve iac ula r e c A s o ni ae eae te c a A s ra t c B r er a eae a c C s si e ae li gu a c li C r yo a c e t u b f l o r ir s p h a ul a e if l e i y Er u m l l or ac ic ae ea Eu a l e e s p Eu h o r rb G um ia en ex G tia e n H ran ella yp iu M er i m c am e c pe Fa nth um st ba a - t pe r is R c yp r f -t y os ea e o r p at S t ac e um e ac e a Eq hy e -t y u s pe A s i s e sy p l t u m l va Ly en tic c i aPo o p o u m ty pe ll e di n um su c m la va tu Zo m ne /P ha se

ea

e

ac

ea

er

ia

yp

ac

Ap

C

Po Ly s M ima el c Pa a m h i a r p v Pe n a s y r u u l g d s m ar R ic u ia is in la -t y pe S a a nt r i s n g hu Sc u s u is S a te l o r b g la a T h i t t a r i a of f ic al r ia in Va ic al le tr u is Ve ria m r b na Sp en h a M ag o n nu M ol m o et O n o l ae s m et , p c f u ae sil . nd at ae c f Ble a . D ch n Bo r y u t o m C r yc p t e s p al h r i ic N lit r i ium s c an up c a t N ha h e lun r t hu ar s ym r ia ia Po p h na t a Le a m e a o m Eu n a g e t on p H ot a cf ot m .p t H on og ec yd i a e t in to M roc at n yr h us i a M op r i yr h s i o y ll p h um yl v lu e m rt sp ic il i c l at at u um m

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Wetland herbs and spore plants

20

Open water

Ecologically indeterminate

353 398 315 343 384 385 409 330 292 365 306 365 334 342 380 345 297 303 301 339 Haz 1

3-III

3-II

Haz 0

308 3-I

Analyst: A. Louwe Kooijmans-Bouhuijs, 1979-1980

Figure III.9 The Hazendonk, unit C, core 3, pollen diagram based on an upland pollen sum, exaggeration 5 x, part 2.

121

Copyright © 2010. Leiden University Press. All rights reserved.

122 ho

on ic a be cc ab un ga

-t y

R

pe en an o p Pe u n c l e c r s ulu tus El i c a s s l a c at r i a c e u i s N ne hy ler tris u p t r d r at ha ian op us C ha r lu dr a ip e r Po r a s t e a ac p. M ea e n e, S c t ha c ha r o aq f f In p hu ua re m de la t ic a t . r ia a / ins a sp r v . en si s

Sc

Ve r th

ep th (c m

-N ol AP o T i gy ) lia cf s . T p. Q ili u a C erc sp o r us . C nu s o s p Vi r ylu sa ., fr b s n u R u r n av g u i i t s h u e n a R am m llan ea nd u n o a cu U bus us pul pu r t i c c a us la c a a th e di e s i u ar t oi s ic ca a Pe rs ic ar ia So m ac C lanu ul he m os A l no n a nu p i g o s d ru A l glu ium m n t A l us in o alb s u n g R us lut i a m ha g no l u m H u n t in s a, C mu us osa c o ar lu f ra , n C ex s lu ng fr. es a c u o Eu rex f. p pulu la f m al pa an s p e C at n i c ca ar o r ic u tk l u e i a u G x m la t in a al r i s t a C ium pa c a ar r ia nn ab C ex pa a p lu in Ty rex se str um p a u e G ha c u do c ly s t i f y Ly cer p. orm pe c i is r us C op a c a u f. Ph rex s e flu e u it S oa l a r l a t r o p a n s a ae i l A l anu s ar us i s m un m O a d d e u in Ly nan pla lca ac t h t h nt a m e a r u e g ar m aq o - a s a ua aq li c t i c ua ar a tic ia a

Li

D

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees, shrubs and herbs

615 Wetland trees and herbs 20 40

Peat

Sand

Peaty sediment

Sandy sediment

Figure III.10 part 1. 20 20

100 200 300

Humic sediment Ecol. indet.

Wetland trees and herbs

525

535

545

555

565

575

585

595

605

20

Copyright © 2010. Leiden University Press. All rights reserved.

nd (% )

ha r Bo c o ne al Bu / f is d h re Bu s m d ai ns s M ca os le s s Le r e af m C rem ain en s a C o c o ins oc c o c Ac o n um ge ar s op Tr i hi ic h lu m H op i r u te ,s ra cl n er Lo din , c ot p h ae as ia D op , c es ap u o s c h Pi n i a c r i o o n s sc s ta s Fa ico p., llin ec la g ep us al e o hip , s p e m p h t at l l e et i a o b la t s r a, Zo st co s ne c /P oo ha ns se

C

Sa

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Varia

Haz 1 3-III

3-II

Haz 0

3-I

20 40 60 80 100 Analyst: W.J. Kuijper, 1979-1980

Figure III.10 The Hazendonk, unit C, core 3, macroremains diagram, + = few (1-10), ++ = some tens (10-49), part 2.

123

Copyright © 2010. Leiden University Press. All rights reserved.

R ha Te bd t o c f rap c o . lo e Zo D ip a a la ( p lo r T c f f i e c la ist a ur b . G lla d t el Pl e lu i e l a ( lar e o n la t y ia G osp glo dq sc pe , t y el o ss vi a 8 p Ty as ra um sti lar 9) e 3 i( o p in s 53 Ty e 1 osp p. ( sph typ ide ) p e 5 0 o t y a e s, ra pe gn 5 0 c o cf 8 .S 2 n o 1 c f. 3B ph ) idia Ty pa A re ) il (E p d um tic M A l e 7 ic o ul A (ty l n 4 id is es po on pe t ype ra -p bi 19 7 7A na ol ) (ty le ) pe (ty n m pe 2) ic 98 ro fo ) s Zo si ls ne /P ha se

124 ol

ep th (c m

-N AP o Sp gy ) iro Sp g y Spiro g ra i r o y r s p. cf g a , p . y sp s M Mo ra s ., n ila o u p o ta Zy uge geo ., o t p e g o t r s Pe ne tia ia l namilat a d m s a c f i a s a - t p. e t e e n t e . T tr yp (n vi a c f y p um e ot r e n t i o n . e la s et (t un Ti Typ 12 sp. ev y p k lle e 2 ire e no tia 6 9 n s 37 w n Va s ) 3) p ls h U aria ag st ni Ty ulin var (ty p a io p C e 9 de sp e 2 ha 2 u o 7 st r a - ) Ty et a ty pe om ( t y pe D 1 ia pe ( ip 8 le or s 4 4 typ ot ) e1 cf he .T 40 c Ty yp a r ) hi pe e z cf 1 14 . T 4 5 oph G yp 2 ila a e (ty Ty eum 12 pe pe a 8A 14 Ty 3 nn pe 61 om 3) H 1 yc ys 2 es c f te r 8 B . iu cf .c D Typ m in e c ar ic C of la 30 f.pu is la g 7 l i s (ty c S c te e l l a pe re a l r o s at a ( a 12 ty Ac r i po pe 6) t in for r iu An o m m 12 1) im pe pe c a M a lt i r f r y l s o ic H c r o s k i s p r at i nu ys t h n . ( i o m Ty ter yri (ty typ n p (t p i u p e l y H e 8 um m s e 8 8 c ate p e ya 5 c p 4) ) s 25 (ty ) f.p . ( As los ul t y p pe c p ic e A r uli h a 11 a re 8 b an na e n 4) C e m ia (ty ) la id u s p d a e C o , sc u b 12 la c e c l or f l d o r a aw u m a v 1) ce s ( a( ( t ra t y yp t yp pe e e 71 32 46 ) A) )

th

525

Li

D

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

535

545

555

565

575

585

595

605

615 20

Peat

Sand

Peaty sediment

Sandy sediment 20

Haz 1

3-III

3-II

Haz 0

3-I

20 40 Analyst: A. Louwe Kooijmans-Bouhuijs, 1979-80

Humic sediment

Figure III.11 The Hazendonk, unit C, core 3, non-pollen palynomorph diagram based on an upland pollen sum, exaggeration 10 x, + = few (1-10), ++ = some tens (10-49), +++ = many tens (50-99).

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

The transition to the third zone of core 3 is based on an increase in Quercus sp., a decrease in Cyperaceae and monoletae psilatae fern spores, and an increase in Apiaceae. The increase in Quercus sp. probably represents a lowered water table since there is a decrease in water plant pollen and since the NPP diagram shows a recovery of types that peaked at the end of the previous zone. Valsaria valsariospora-type (type 140) shows decreased values compared with the earlier zones. The anthropogenic signal of phase Haz 1 is clearly visible in the pollen diagram at 5.43 to 5.31 m -NAP, characterised by a slight decrease in Tilia sp., Quercus sp. and Cyperaceae, and the presence and/or peaks of Rhamnus cathartica, Cerealia-type, Chenopodiaceae, Plantago lanceolata, P. major, Polygonum persicariatype, Poaceae, Urtica sp., Ranunculaceae, Apiaceae, Veronica sp. and others (see diagram). Alnus sp. shows a peak but decreases afterwards. The decrease in Alnus sp. in the macroremains diagram indicates that it only concerns increased representation of the taxon in the pollen diagram due to decreased pollen rain of dryland pollen. The pollen diagram shows the recovery of the vegetation after occupation characterised by an increase in Pinus sp. and Cyperaceae and a decrease in Alnus sp., but some taxa suggest that occupation continued to the upper part of the diagram (Quercus sp. and various dryland herbs indicative of human impact). The macroremains diagram indicates that occupation resulted in a decrease in Quercus sp. and Cornus sanguinea, and an increase in Urtica dioica, Persicaria maculosa, Solanum nigrum, Chenopodium album and Persicaria hydropiper. Lythrum salicaria and Veronica beccabunga-type show maximal values after occupation. The NPP diagram shows small peaks of psilate Spirogyra sp., type 361, Pleospora sp. (type 3B), Gelasinospora cf. reticulispora (type 2), type 150, type 82A and cf. Spadicoides bina (type 98) during occupation. Type 361 and type 2 that is related to the presence of charcoal can both be related to human occupation (Van Geel 1978; Van Geel et al. 1981), while the other types are not known as anthropogenic indicators.

Copyright © 2010. Leiden University Press. All rights reserved.

III.3.3 core 2, near unIt c (VlaardIngen 1a and VlaardIngen 1b) Table III.6 shows the lithostratigraphy of core 2. figures III.12-14 show the results of pollen analysis, non-pollen palynomorphs and macroremains of core 2. Recognised occupation phases are VL 1a and VL 1b. The core represents a continuation of the section of square 57. The dryland vegetation is quite constant during the whole diagram. Quercus sp., Fraxinus sp., Ulmus sp. and Corylus sp. dominate the diagram, while remains of the shrubs Viburnum sp., Rhamnus cathartica, Ligustrum sp. and Sambucus sp. are regularly present as well. The presence of Fagus sp. supports that the diagram represents the Sub-Boreal. In zone 2-I (3.94 to 3.72 m -NAP), the percentages of Pinus sp. and Cyperaceae are relatively high compared with the following zone, which may indicate relatively wet conditions. The macroremains diagram suggests the presence of alder carr, but is poor in macroremains in this zone. Important NPP’s are Valsaria variospora-type (type 140), Diporotheca rhizophila-type (type 143) and Apiosordaria verruculosa (type 169). Type 169 indicates the presence of decaying plant material, possibly in the form of dung (Van Geel and Aptroot 2006). depth (m -NAP)

sediment

3.03-2.94

coarse peat with many wood remains and clay particles

3.26-3.04

clayey peat with coarse plant remains

3.28-3.27

wood remains

3.63-3.29

peat with fine plant remains, Vlaardingen 1b at 3.39 to 3.28 m -NAP

3.65-3.64

wood remains

3.94-3.66

peat with coarse plant and wood remains, Vlaardingen 1a at 3.88 to 3.80 m -NAP

Table III.6 The Hazendonk, unit C, core 2, lithostratigraphy.

125

Copyright © 2010. Leiden University Press. All rights reserved.

126

Peat

Figure III.12 part 1.

Clayey sediment 100 300 200 400 600 20 20 20

n A l um is d m S p a ulc am S parg ar a a Ty arg niu ph an m Lo a iu e m t l m S y u s - at i f e r e r s m t y p o lia e c um ph e tu m t yp yt e um

um

Wetland trees and shrubs

la

Upland herbs and spore plants 20

hr

20

So

20 40

Ly t

20

H um M ulu yr s i R ca ha R mn ib us e C s fra al ng y ul Fi ste a lip gi en a d U r t i ula ca

20 40 60 80 100

li x

pl an t re

y

d

og

(c m

es

-N AP )

lm

xi

nu

s

Vi b R ur n h u Li am m g n S a ust us r c C m b um at h o r uc ar n tic H u us ip s a po s a ph n ae g u in ea

us Be t u Po l a Ac pul u Ju er s c f .n ni Fa pe ig ra g r C us us /c ar f. t re H pin ed u m ul V i er s a sc a C um or yl us

U

Fr a

U pl U and pl s a Pi n d hr u c e he bs Pi a r b s nu an s d Ab sp or Ti ies e lia pl an Q ue ts rc us

U

th

ol

ep

th

D

Li

300

Sa

C h Po e n o p A r l yg o o d i t Pl em numace a is a R nt a i a p e e um g rs o ic Al e l ar li x a ia A l um ac nc e -t y et o liu pe os la C m t er v a- a ty Pl ea ine pe a li a G nt a a l e en g ty o pe Sp is m e ta O rg - t y ajo xa ul p r Po l i s a - t e y Pt lypoace pe e t A l r i d diu o s e nu iu m lla s m

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees Upland shrubs

310

320

330

340

350

360

370

380

390 20 40 Wetland herbs and spore plants

20

Copyright © 2010. Leiden University Press. All rights reserved.

20 40

yp ea

e

ac

ea

er

ac

e Ap A s i ac e t A s e r a ae c B r te r a e a a c e R s s e a li g u ic e u R b i a ac e t ub li f l o o c r Er s a c e a e a e u l i f a e lo i e ra R c al e ae e a C nun s ar c C yo u l a e p c C nta hyl ea ir s ur la e C iu ea c e a m ae C mp y a Eu s t o n u l p a Eu p h r t e r i a s p Eu h o s i a r G r um bia en e G tia x e n M ran a p e i M nt h u m n e u yo a m on H so - t y u t p an Li p er i s e th llia zi ea ty Fa ce s pe ba ae e l St c ag ac e a o Sc h e y Te rop s sy uc hu lv Va ri la at u r ic V i ler i m ia - t a - t yp yp o l an a Ve p el e e la r a Eq o n l u s i c D ui s e a t r i s -t y ip H ha tum pe yp si G e r ast la ic r u Po u x u m m lle pe (Ly n r fo c o su ra po m tu di m um Zo -t y ) ne pe /P ha se

C

Po

Ly s Bu im a S ato m c hi u a M git t s vu lg el ar ar C am ia is hr p -t y Li ys y r pe t to o s um Sc re pl lla en u R te iu in ll m R an ar ia an t h u Va un s c Ir i l e r i ulu s a s Pa p s n a r e e pe S ar n a u d a ns ss c x H if r ia or yd a g us c f ro a . c g S p El a t y l e r a n t ul at M hag ine a on n u ol m Tr eta il e c f et a , p . D e, s O r y p s i l at p o a O h i o pte i l at a e sm gl r e i o s N u s c u n s a H p ha da um r t h o r us N t to n ia ym ia na Po p ha t Le a m e a m o Eu n a g e to n Tr pot cf ap am .p M a o ec y r na ge tin M i o p t a to at yr h n n us s St i op y ll r a h um H t i o y ll v yd t e u m e r ro s sp t ic il ch i c l at ar at u m is um

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Wetland herbs and spore plants

20 40

Open water

20 Ecologically indeterminate

335 268 339 338 348 345 321 348 355 307 347 274 353 318 401 492 331 304 315 283 327 299 319 307 2-III

VL 1b

2-II

VL 1a 2-I

Analyst: A. Louwe-Kooijmans-Bouhuijs, 1981

Figure III.12 part 2 The Hazendonk, unit C, core 2, pollen diagram based on an upland pollen sum, exaggeration 5 x. This figure can be seen as the continuation of figure III.6.

127

Copyright © 2010. Leiden University Press. All rights reserved.

n A l c us op e R ec f fu an u s Pe u n r u s u s r c R si c ulu g en an ar s ic N un ia re p ula ym c h e t L e p h u l u yd r n s u s m a s op - t C na ea sc e ip e y p e al s a le r C litri p. lba ra ha c tu s M ra he en sp sp R th . . um a C ex aq yp ua Ve era sp. tic a/ ro c e ar ni a ve ca e ns b Po is ec ca l In ypo bu de d ng S a t . i ac aea n ty C d( e pe ha % B o rc o ) a n M e/ f l o is Bu ss h r d re e Bu s ma ma in ins d s C sc e a Lo n o c l e s p o D ho c c u ap pu m C hn s c g e o ia r o Tr coo sp ista ph ic n . l il H h o s , e p linu um i r u pt hi s, , s p p st c l Ac nd era i a ato e r o ar ina , c i e as bl t ia ,c e as oc s ts oo Zo ns ne /P ha se

Ju 300

128 Li

ep

th

Upland shrubs and herbs

th

(c m -N ol AP og ) C y or S a nu m s C bu s an ra c g U t ae us uin rt g n e G ic a us i gr a a a d m So leo ioic on l ps a o g Pe a n u i s yn a r s m bif i C ic n da he ar ig r t i y n u a Pe o m pe r p m A l s i c o di ac u nu ar um l o s ia f sa A l glu lap i c i f t i n at o l n o hi iu A l us nu glu s a fo li m a Ly s tin t h glu o r u ti sa m no , c sa sa on li c , f e s ar r ag ia m C .o ar fm C ex ar r i al e Al ex par ca i s ac i a tk S o m a ut in i s l p fo A l anu lan r m is m t a is c f ma d go . u P h S t a s p. l c a a q u m a al c h ar t i C ar ys a ca ar is p e Ly x ar alu c o ps un st O p u e u din r i s e s d a S p nan eu o cy c e a t r p a C r g he o pa er u ar an aq e s G ex ium ua us ly s ti S t c e r p. e r e c a el ia ct um B e lar f lu r u ia it a S i la aq ns u e u Ty m l re ati p h at i c t a c a a fol sp iu . m

D

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

390 Wetland pioneers

Peat

Clayey sediment 20 Open water

Wetland trees and herbs

310

320

330

340

350

360

370

380

20 40

Ecol. indet. Varia

2-III

VL 1b

2-II

VL 1a

2-I

20

Analyst: W.J. Kuijper, 1981

Figure III.13 The Hazendonk, unit C, core 2, macroremains diagram, + = few (1-10), ++ = some tens (10-49). This figure can be seen as the continuation of figure III.7.

Copyright © 2010. Leiden University Press. All rights reserved.

cf . H Typ ys e U ter 12 st i 2 C ulin um la a c cf ste de f. p . r u u Ti Typ osp sta lica ll e o r Ty etia 12 riu (typ e (t pe s 8A m e yp p ca 4 e 3 6 ha ric 4 ) 12 1 gn 1 in i( um ) t y Sc pe ( ty a 27 pe Ty lar ) p if 25 Ty e 1 orm ) p 2 Ty e 1 8B pe r fo p e 24 G 1 ra a 4 tio c f eum 2 n .t a pl at c f y p nn es .D e o 7 m C in 0 y (ty er o 8 c pe f es R ato l a g 11 ha p e c f. 4) Ty bd hy lla ca p e o c llu t a r ic cf 1 o m is . 2 e s (ty Zo Typ 3 la ( p., Tu sp pe pf e 1 Sp ie 4 r b in 12 lla 5 el e e 6) la (ty c f r m lu r ia p .T a n e t d , y C p op q ty 13 lo e h v pe 7 c f ste 3 3 o r e i st i i 35 ) . T ri 3 s (t 3) Ty yp um of y p p e C e5 op 0 To e 1 30 cf. r ta 4 8 7 o s e 1 ln B tra po ) da on to -p m (ty ol ( p ty le pe e 2 n pa 37 8) ly 2 no ) m or ph s Zo ne /P ha se Li

th

th

ep (c m

ol N A og P) Sp y S piro g y S piro g ra y r sp i r cf og a ., . y s p M Mo ra p., sila o u s n ta Zy uge geo p., ot p e g o t o s c f n e t i a i a l r n a i l at . T m s ae m ae P e y p a - t p. t e e n d e y (n v ta H ias 3 4 p e o n iren t i o ys t 1 la s n u et (ty n Ty ter rum A ev p k n pe iu s i r e e 3 ow Ty 1 m p. p 5 c n s 73 n Ty e 6 0 f. p ) ) ul pe 6 ic Ty 5 ar pe 5 B e Xy 1 (ty l 2 pe Va aria 5 12 ls s a r p. 1) ia va r io sp or aty pe Ty (ty pe pe D 7 ip 4 14 or 0) ot he ca rh iz op hi la (ty pe Ap 14 io 3) so A n rd t h ar i a o c f m o ve r .D st el r uc Xy ip la loc la fu ulo r i a la eg sa d ( C ha sp. iella ian typ a e G et s ( c o e ty 16 al pe 9) ar Ty las mia oi pe ino le de 4) 18 sp s s, or co a ni sp di .( a ty (E pe M A 1) ty pe

D

19 )

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

300

310

320

330

340

350

360

370

380

390 20 40 60 80

20 40

Peat

Clayey sediment 20 40 60 80 20 20

2-II

VL 1a

2-I

Analyst: A. Louwe Kooijmans-Bouhuijs, 1981

50 100150200 20

2-III

VL 1b

Figure III.14 The Hazendonk, unit C, core 2, nonpollen palynomorph diagram based on an upland pollen sum, exaggeration 10 x, ++ = some tens (10-49). This figure can be seen as the continuation of figure III.8.

129

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Phase VL 1a (3.88 to 3.80 m -NAP) is characterised by a very weak anthropogenic signal: a small decrease in Tilia sp. and Corylus sp., a small rise of Quercus sp. and Fraxinus sp., the presence of Chenopodiaceae, Artemisia sp., Plantago lanceolata, Rumex acetosa-type and Allium sp., a peak of Poaceae, Alnus sp., and Urtica sp., and a decrease in Cyperaceae and monoletae psilatae fern spores. The occupation is not characterised by a major change in the ratio of arboreal and non-arboreal dryland pollen, indicating that deforestation at the southeastern side of the dune was very restricted. Most of the macroremains found in zone 2-I are found within the occupation horizon. NPP’s that show an increase during occupation are psilate Spirogyra sp., Zygnema-type, type 150, type 66, type 74, type 361 and type 128B. Only type 361 is known as an indicator of anthropogenic influence (Van Geel et al. 1981), although type 150 was also present during phase Haz 1 in core 3 of this study. The limited impact of anthropogenic influence on the vegetation during this occupation period corresponds with the limited number of archaeological finds and the limited possibility of distinguishing this occupation phase during excavation. The transition to zone 2-II (3.71 to 3.40 m -NAP) is based on a decrease in Pinus sp. and Cyperaceae, and an increase in Corylus sp. and Alnus sp. Alnus sp. replaced sedge vegetation as indicated by the pollen and macroremains diagram. The macroremains diagram shows (extra-) local presence of Cornus sanguinea, Sambucus nigra and Crataegus monogyna, which were probably part of the alder vegetation as well. The herbaceous undergrowth of the alder vegetation probably consisted of Filipendula ulmaria, Sparganium sp. (pollen), Urtica dioica, Alisma plantago-aquatica, Ranunculus repens-type and Mentha aquatica (macroremains). There are no clear indications of human impact in this zone despite fluctuations in some NPP curves (including type 361). In this zone, the macroremains diagram shows strong similarity with the middle part of the diagram of section 41 (after phase Haz 3) where Alnus sp., Cornus sp., Sambucus sp. and R. repens-type show peaks as well. The transition to zone 2-III (3.39 to 2.94 m -NAP) is based on a slight increase in Betula sp., a decrease in Corylus sp. and an increase in Poaceae, Cyperaceae, Salix sp. and wetland herbs. The zone starts with occupation phase VL 1b (3.39 to 3.28 m -NAP). The pollen diagram shows a decrease in Tilia sp., Quercus sp. and Corylus sp., the presence or peaks of Chenopodiaceae, Polygonum persicaria-type, Artemisia sp., Allium sp., Plantago lanceolata, Cerealia-type, Poaceae, Cyperaceae, Urtica sp., Lythrum sp., Alisma sp., Solanum dulcamara, Sparganium erectum, Asteraceae tubuliflorae and Brassicaceae. The percentage of herbs shows a major peak in the summary diagram. The macroremains diagram shows the presence of sand, charcoal and bone remains, a decrease in Alnus sp. and increased values of Urtica dioica, Lythrum salicaria, Alisma plantago-aquatica, Persicaria hydropiper and Veronica beccabunga-type. The increase in both dryland and wetland herbs indicates relatively open conditions, disturbance and the increased presence of nutrients. The NPP diagram shows increased values of psilate Spirogyra sp., a major peak of type 361, small peaks of Gelasinospora sp. (type 1, related to the presence of charcoal), type 128B and type 708, and an increase in the total NPP percentage. Types 1 and 361 can be related to human impact. After occupation, the dryland vegetation recovered, while Urtica dioica, Solanum dulcamara, Brassicaceae, Rubiaceae, Mentha-type, Veronica sp. (pollen) and Lythrum salicaria (seeds) show temporary peaks, representing an intermediate state of recovery of the natural vegetation. In the wetland vegetation, Salix sp. partly replaced Alnus sp., as indicated by a major peak of Salix sp. The high percentages of Salix sp. indicate local dominance (cf. Waller et al. 2005), and indicate more dynamic fluvial conditions. The percentage and diversity of water plants does not show major changes, but Hottonia sp. and Myriophyllum verticillatum are present more frequently. The presence of anthropogenic indicators in the pollen diagram, the presence of charcoal and the high values of type 361 in the NPP diagram in the middle part of zone 2-III suggest continued occupation, but this is probably related to erosion due to colluviation and/or water activity. The macroremains diagram indicates that woodland of dry terrain was no longer present in the local vegetation.

130

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

III.3.4 M86, eaStern SectIon of Square 25, unIt b (VlaardIngen 1b) Table III.7 shows the lithostratigraphy of M86. figures III.15 and III.16 show the results of the pollen and macroremains analysis of M86, which correspond to the period before, during and after phase VL 1b (3.70 to 3.58 m -NAP). The macroremains diagram was partially published earlier (Bakels 1981; Out 2008c). The local vegetation probably consisted of alder carr with Viburnum opulus and Cornus sanguinea. Well-represented herbs are Veronica beccabunga-type, Juncus effusus and Urtica dioica. Human impact resulted in a decrease in Quercus sp., Fraxinus sp., Alnus glutinosa, V. beccabunga-type, J. effusus, U. dioica and Plantago lanceolata, and to a strong increase in dryland anthropogenic indicators (cf. Behre 1981) including Cerealia-type, Chenopodium album, Solanum nigrum and Stellaria media. The diagram also shows a moderate increase in ferns, grasses, sedges and wetland taxa, including Sparganium sp., Filipendula ulmaria, Symphytum sp., Ranunculus sceleratus and Rorippa amphibia. Together these changes indicate the disturbance of the oak vegetation and of the alder carr, the increased presence of open patches and eutrophication that was probably caused by the dumping of waste. The macroremains do not contain crop plants or carbonised food plants. The only carbonised finds are two fruits of Mentha aquatica/arvensis (sample depth unknown31). Poaceae and Cyperaceae show a strong increase at the end of occupation, which was probably the result of the recovery of the vegetation as well as the rising ground water level. After occupation, Quercus sp. recovered, certain shrubs increase (Rhamnus cathartica, Ligustrum sp. and Sambucus sp.) and the dryland herbs decreased or disappeared, indicating the recovery of the vegetation. In the wetland vegetation, Alnus sp., Salix sp., Lythrum salicaria and Mentha aquatica/arvensis increased strongly, and Sparganium erectum, Solanum dulcamara and Rubiaceae increased as well, while Poaceae and Cyperaceae gradually decreased. The changes of these wetland taxa can be explained by the increase in the water level, as indicated by the presence of clay, and may possibly also represent the development of secondary vegetation as part of the recovery of the natural vegetation.

Copyright © 2010. Leiden University Press. All rights reserved.

depth (m -NAP)

sediment

3.59-3.56

slightly clayey peat with pottery fragments

3.62-3.60

peat with charcoal remains

3.65-3.63

very sandy peat

3.67-3.66

sandy peat

3.70-3.68

very sandy peat

3.75-3.71

peat; charcoal remains at 3.73 to 3.72 m -NAP

Table III.7 The Hazendonk, unit B, eastern section of square 25, M86, lithostrigraphy.

31 The macroremains were found in the material that was left over after preparation of the pollen samples.

131

Copyright © 2010. Leiden University Press. All rights reserved.

132 he

no

20

Peat

Clayey sediment

Figure III.15 part 1.

Sandy sediment 20 40 20

100 200 300 20

Upland herbs and spore plants

20 40

H u M mu y lu R r ic a s ha C m a n C lyst us o e fr U nv o g i a a n g r ti lv ul c a ul a us Sp Ly arg t a S p hr u niu m m Fi a r g em lip an er Sy en iu su m d m m Lo p h ula e r -t y e t y ct S o us - tu um p e ty m l a Ly n pe s u A l im m d is a u Bu ma c hia lc a vu ma c f to m lg ra . u ar H El a s is yd t i -t y ro ne pe ct yl e

20 40 60 80 100

li x

375

an t re

-N

es

(c m y

d

og

th AP )

us n B e us t u Po l a Ile pulu x Ac s c f .n e Fa r ig ra g u /c H s ed f. t re V i er m sc a ul C um a or yl us C o Vi rnu b s R ur n s a ha u n Li m m g u in g n ea S a ust us m r um c at bu ha cu r ti ca s

xi

lm Fr a

U

U p U lan pl d a s Pi n d hr u c e he b s Pi a r b s nu an d Ab s sp ie or s Ti e lia pl an Q ts ue rc us

pl

ol

ep

th

D

U

Li

355

Sa

Po p o d A r l yg o i a c e te n C m um ae e is R rea ia pe u li rs ic A l me a ar liu x a ia Pl m c -t y et an pe os Pl t a aan g ty Ja t a o m pe s g c f ion o l ajo . S e an r ce R ol e a ol at Fa sed nu m a a l G lop lu nig e n i a te r a u Po i s c o - t m ta n y P t l y p o - t y vo p e er d pe lvu A l i d ium lu nu ium s s

C

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees Upland shrubs

357

359

361

363

365

367

369

371

373

20

Wetland trees and shrubs

20

Wetland herbs and spore plants

Copyright © 2010. Leiden University Press. All rights reserved.

H y M per y ic M o s o um e t L i nt h i s p e r ll a fo ra O iac e - t y tu no a pe m R ni e -t y os s pe S t ac t y p ac e a e Sc h e y Tr rop s sy if h l Ve oliu ula vat m r i ic Eqron -t a-t a-t ui ic a y p e y p e y p D se e ip t H ha um up si Po e r a s t ll e zia r um n se / su la L m g yc Ph o o (L po as yc d e o p ium od c i u l av m a ) tu Ecologically indeterminate

m

20 20 40 60 80 20 40 60

ac

ea

e

as s A p ic a c R i ac e e ae an a Eu u n e r c A s um ula t e ex c e C ra ae a c R r yo e a ub p e A s i a hy l t u b t c la u C er a eae c e li f l ir s c ae o r ae C iu ea am m e l ig Er p ul i a if l G c al nul or er es a ae an iu m

er

Open water

Br

yp

Wetland herbs and spore plants

C

Ir i s M ps e e Pr lam uda i p c R mu yr u or u a la m s S c n u n ve c r Saute ulu is-t x i llar s r y p T h f ra ia e e pe a Ty lic ga ns ph t r u gra U a m n rt l ul at Va ica atif a l e , la o lia Sp r ia rg e n h M ag a on n o l um et ae M ,p o si Tr nol la i l e et ta N t a ae e up e , N ha p ym r si S t p h l at a r e Eu a t i o a e a t H p ot e s o a Po t t o n m o ac ia g e to ea n e

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Ecologically indeterminate

20

392

408

471

396

444

391 VL 1b

489

419

411

429

394

Analyst: A. Louwe Kooijmans-Bouhuijs, 1974-1980

Figure III.15 The Hazendonk, unit B, eastern section square 25, M86, pollen diagram based on an upland pollen sum, exaggeration 5 x, part 2.

133

Copyright © 2010. Leiden University Press. All rights reserved.

134

20 40 60 80

Clayey sediment

Peat

Figure III.16 part 1. a aq

50 100 150

100 200 300

Sandy sediment

ua tic a/ R an ar ve u C n ar c ns u E u ex l u is p a s C ato c u r e ar r t if pe Ty ex ium or ns m p r A l ha ipa c a is t yp is sp r ia nn e m ab R a . or p i n S o ipp la um a n R lanu am t ag um m o H e d phi - aq y x u b u Ly per hyd lca ia ati ca m c ic S p o p um r o la ar ar us pe pa a c f g eu r th . a f u S i S iu niu rop ora m um m m ae tu O l la er us m e a ti e R nan t i fo fo li c tu an t h liu um m un e m Pe c u a q r s l u s u at Ju ic i nc ar sc e c a us ia le h e f yd r a t u fu ro N ym su p s s ipe R p um ha r A p e ea ia x s al c e p. b a ae

th

373

en

th (c m

la

e

20

20 40 20 40 60 80

la

nu

m

ni

gr

um Pe c f r sic . a Pe Pe r r i a r s sic lap ic a a ar r ia t h ia l ifo m ap lia ac at ul hif os ol a ia

So

G le c Fa hom llo a So pia he nc sp de S i hu . / r ac l e s Pe e a n a C e l sp r si h e at e r c a r ia n i C o p fo lia sp he o . n o diu sp p m p . St e l o di f i c alb la u a r ia m i fo li m alb um e d um ia

-N AP Q o gy ) u U erc r ti u ca s s di p., oi c c a up u ol

ep

th

D

Li

355

M

Al n A l us nu gl A l s ut i n g n Ve us luti osa ro sp no ni ., f sa c a ra , c b e gm o n c c . o es ab f m un ale ga c - t y at k p e ins C ar e Ly x th el r u at m a sa li c ar ia

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees and herbs

357

359

361

363

365

367

369

371

20 40 60

Wetland trees and herbs

20

20

Copyright © 2010. Leiden University Press. All rights reserved.

20 20

Ph

as

e

nd C (gra h m B o arc ) oa n Po e / l fi Batter sh y re Bur k r rem ma ds em a in ai ins s ns Bu d Le s c a a C f re les en m C o c ai o o n In c o o c c s s n u Tr ect s m g ic r eo Fa ho em ph e p a ilu S p c a te r a i n s m l ,s or p , c cl e s ell a er s e ot in t s e s ia de t.

Sa

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Varia

VL 1b

Analyst: W.J. Kuijper, 1974-1980 20 40 60 80

Figure III.16 The Hazendonk, unit B, eastern section square 25, M86, macroremains diagram, + = few (1-10), part 2.

135

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

III.3.5 M87, eaStern SectIon of Square 25, unIt b (VlaardIngen 1b) Table III.8 shows the radiocarbon dates from M87 (Lanting and Van der Plicht 2000). The results correspond with the other dates of the refuse layer but do not add additional information because of the broad ranges. Table III.9 shows the lithostratigraphy of M87. figures III.17 and III.18 show the results of the pollen and macroremains analysis of section M87, which correspond to the period before, during and after phase VL 1b (3.91 to 3.81 m -NAP). M87 is located c. 3 metres lower on the slope of the dune than M86 but is highly comparable with M86. The local vegetation probably consisted of alder carr. The quantity of sand and archaeological refuse is smaller in M87 than in M86, indicating that M87 is located further away from human activity. At the start of occupation, Quercus sp. decreased and Fraxinus sp. showed a peak. Ulmus sp. was probably more common near M87 than at M86, since the percentage is higher in M87. Chenopodiaceae, Artemisia sp., Cerealia-type and Solanum nigrum clearly indicate the period of occupation, although the dryland herb signal is weaker than in M86. In the wetland vegetation, Alnus sp. decreased while Poaceae, Cyperaceae, Urtica dioica, Lythrum salicaria increased strongly. After occupation Quercus sp. and Betula sp. increased, certain dryland shrubs increased (Rhamnus cathartica, Ligustrum sp. and Sambucus sp.), and herbs that showed high percentages during occupation show a decrease. In the wetland vegetation, Alnus sp., Salix sp. and various herbs such as Sparganium erectum, Lysimachia vulgaris-type, Apiaceae, Brassicaceae, Ranunculaceae and Rubiaceae increased after occupation, correlated with the sedimentation of clay. The changes may possibly also represent the development of secondary vegetation as part of recovery of the natural vegetation. Although systematically counted data are absent, some information on non-pollen palynomorphs from M86 and M87 is available. M86 and M87 contained the NPP’s Mougeotia sp., Spirogyra sp., Zygnematype, Diporotheca rhizphila (type 143), type 150, Zopfiella lundqvistii (type 501) and Hystrichosphaeridae. Mougeotia sp., Spirogyra sp. and Zygnema-type indicate shallow, stagnant, mesotrophic to eutrophic freshwater. Hystrichosphaeridae are indicative of marine influence, although the other available sources demonstrate that this was negligible. Due to the absence of data it is not possible to investigate the reaction of the NPP taxa observed in M86 and M87 to human impact. sample

depth (m -NAP)

lab code

age (yrs BP)

age (yrs cal BC, 2σ)

dated material

Haz 28c

3.82

GrN-8239

4220 ± 60

2930 (95.4%) 2610

peat

Haz 28b

3.87-3.86

GrN-9198

4050 ± 120

2900 (95.4%) 2200

peat

Haz 28a

3.93

GrN-9197

4390 ± 170

3550 (95.4%) 2550

peat

Copyright © 2010. Leiden University Press. All rights reserved.

Table III.8 The Hazendonk, unit B, eastern section of square 25, M87,

depth (m -NAP)

14

C dates (Lanting and Van der Plicht 2000).

sediment

3.83-3.78

clayey peat

3.89-3.84

peaty sand with charcoal remains and a bone fragment

3.91-3.90

sandy peat

3.97-3.92

peat with Phragmites remains

Table III.9 The Hazendonk, unit B, eastern section of square 25, M87, lithostratigraphy.

136

Copyright © 2010. Leiden University Press. All rights reserved.

100

a

20

300

Peat

Sand

Peaty sediment

Clayey sediment 20 20 40 60

20 40 60 20

Ly t S o hr u m A r lanu ct m A l iu d is m ul ca Bu ma m ar Fi to m a li u Lo p e n s tu du Ly s- la s ty R ima pe a S p nun c hi c a S pa r g u l u v u l ar ani s r ga g u e ri S y ani m e pen s - t y u m p m Sa p m e e s e rs hy r x u e i t Th fr u ct m um - t y a a m Ty lic ga pe ph t r u gra U a m n r t i la ul at Sp c a tif a ha , la o lia r gn ge um

ul

Upland herbs and spore plants

H u R mu h lu C am s al nu U yst s r ti e fra c a gia n g

20 40 60 80 100

li x

396

an t re

-N

es

(c m y

d

og

th AP )

us nu s Be t Po u l a Ac pulu e Fa r s cf .n C gus ig ar ra p /c H in f. ed u tre V i er s m sc a ul C um a or yl us Vi b Li ur n gu u R st m h r S a am um m nus bu c c u at s ha r ti ca

xi

lm Fr a

U

U p U lan pl d a Pi n d s h ce h ru e b Pi a r b s s nu an s d Ab sp i or Ti es e lia pl an Q ts ue rc us

pl

ol

ep

th

D

U

Li

378

Sa

C h A r eno t p C em od e is ia A l rea ia c e ae l li Poium a ly R go u n Fa me um x Pl ll o p ac p e a n i a et o r s Ja t a c o s ic a s g n a r Po i o n o l a vo - t y p i a - t e n lvu e yp e c e lu A l ly p nu o d o s s i um l at a

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees

20

Wetland trees and shrubs

20 40

Upland shrubs

380

382

384

386

388

390

392

394

20

Wetland herbs and spore plants

20

Figure III.17 part 1.

137

Copyright © 2010. Leiden University Press. All rights reserved.

20

138 20 40 60 80 100

er

ac

ea

e

od

iu

Open water

20 40 60 20 20 20

m )

ia c A s eae te A s ra t c Br era ea as c e e l s a ig C i c a e tu uli f ar c b lo C yo e a e u l i r a e ir s p f lo Er iumhyl ra la e ic c a Eu l e ea s r e H um yp e M er x y ic M o s o um e t L i nt h i s p e r llia a fo ra O c ty tu no ea pe m R ni e -t y o s pe R s ac - t y p an e e a un e cu R u la ce St bia a a c S c c h eae e y ro s Tr p sy if h l Eq o l i u u l a v a t m r i ic H ui s e - t a - t y a - t up t y p p y p Po e r u m e e e ll e zia n se su la m g Ph o (L as yc e op

Ap

yp

Wetland herbs and spore plants

C

on o Tr leta ile e N ta , p u p e, s h N a p s i l at a ym r il e a H p ta o h e Eu t t o n a e a p i Po o t a ac am ea o g e et o n

M

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Ecologically indeterminate

366

375

360

448

411 VL 1b

377

366

330

350

419

Analyst: A. Louwe Kooijmans-Bouhuijs, 1974-1980

Figure III.17 The Hazendonk, unit B, eastern section square 25, M87, pollen diagram based on an upland pollen diagram, exaggeration 5 x, part 2.

Copyright © 2010. Leiden University Press. All rights reserved.

R an S a un n c B o d ulu s n re Bu e/ f pe ds i sh ns Bu r -t y e d m pe Le s c ai af a l ns M re es os m B a s ai r re n C k r ma s en em in Tr oc ai s ic o n C ho c c u s o c pt m In o o era g s n , e Sp ec t s c a oph o r rem se s ilu m es a ,s in ins cl de er t. ot ia Ph as e 378

th

ol

ep th (c m

396

Wetland trees and herbs

e

20 40 60 80 100

Peat

Sand

Peaty sediment

Clayey sediment 20 50 100 150

e en th Al a a q is m R a u at an p ic a R un la or c nt /ar u a i C p p l u g ve n ar a s o s cf ex am sce aqu is . O r ip p le a h C e ar ib r a t i c a na ia ia tu a s H rex nth yp p e Pe e r s e u a q r s ic u do ua ic m c ti ar c yp c a ia f. er hy tet us dr r ap o p te ip r u er m

M

la Pe n u m rs n Si ic a igr l e r ia um n C e la ma he t c n if u A l o p o lia l o s nu o d s a iu pp s A l glu m a . al t in lb ba n o s um C us a g a St rex luti n ac ac o Ve hy ut sa r s if , S o o ni pa o r m c o n l c lu i e Ly anu a b str s s th m ec is ru d ca m ul b s a c a un li c m a g a ar ra - t y ia p

So

-N AP o ) C gy or Q nu u s Q erc s an u u g U e rc s sp uin r t i us . e ca s a di p., oi c u ca p ul a

Li

D

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees, shrubs and herbs Wetland trees and herbs

380

382

384

386

388

390

392

394

20

Varia

VL 1b

Analyst: W.J. Kuijper, 1974-1980

20

Figure III.18 The Hazendonk, unit B, eastern section square 25, M87, macroremains diagram, + = few (1-10), ++ = some tens (10-49).

139

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

III.3.6 Southern SectIon of Square 41, unIt c (hazendonk 1, hazendonk 2, hazendonk 3 and VlaardIngen 1b) Table III.10 shows the lithostrigraphy of the southern section of square 41 (unit C). figure III.19 shows the corresponding macroremains diagram. The diagram generally shows the change from dryland vegetation towards wetland vegetation. In the lower part of the diagram, the variety of dryland taxa is relatively large, and woodland of dry terrain may have been present in the extra-local vegetation. Herbs indicative of moist conditions are additionally present in the extra-local vegetation as well, such as Valeriana officinalis and Solanum dulcamara. In the middle part of the diagram from phase Haz 2 onwards, alder carr develops with an undergrowth of Urtica dioica, Typha sp. (T. angustifolia/latifolia), Solanum dulcamara, Carex acutiformis, Lythrum salicaria, Scrophularia sp. and other marsh plants. After 3.60 m -NAP (between phases Haz 3 and VL 1b), alder shows maximal values, Cornus sanguinea and Sambucus nigra show peaks, and Persicaria maculosa, Phalaris arundinacea and Ranunculus repens-type show high values. Both Persicaria maculosa and Ranunculus repens may indicate trampling or water activity, resulting in the presence of condensed substrate. Alder vegetation was replaced by marsh vegetation at the time of phase VL 1b. The sediment of the upper part depth (m -NAP) sediment of the diagram (after phase VL 1b) consists of clay poor 2.63-2.60 clay in macroremains. Reed vegetation may possibly have been present since reed fruits seldom remain preserved. 2.75-2.64 very clayey peat The development of the vegetation represents a reversed 3.00-2.76 clayey peat succession, controlled by an increase in the water level 3.10-3.01 peat and increasing instability of the water level. The distinction of phases Haz 1, 2a and 2b is 3.16-3.11 clayey peat difficult since sand, charcoal, and bone remains are 3.20-3.17 peat more or less continuously present in the lower part of 3.40-3.21 peat: Vlaardingen 1b the diagram, which may be related to vertical transport 3.50-3.41 peaty clay and colluviation processes (cf. Amkreutz et al. 2008). The precise start and end of these occupation phases 3.60-3.51 peat cannot be detected in this diagram. The horizon of 3.75-3.61 peaty clay phase Haz 1 is based on the presence of carbonised 3.85-3.76 peat (brown) remains. Distinction of phase Haz 2a is based on a high value of bone remains and a peak of Urtica dioica and 3.88-3.86 peat (dark brown): Chenopodium album. The high values of Urtica dioica Hazendonk 3 ? correspond with the macroremains diagram of square 3.97-3.89 sandy peat: Hazendonk 3 57 during this phase. Distinction of phase Haz 2b is based on peak values of taxa that occur before and after 4.13-3.98 peaty clay the recognised occupation. 4.20-4.14 slightly sandy peat: The Haz 1 horizon is characterised by a relatively Hazendonk 2b large number of carbonised cereal remains, corresponding with the presence of concentrations 4.25-4.21 peat rich in clay of cereals during the Hazendonk 1 phase in squares 4.31-4.26 sandy peat: Hazendonk 2a 57 and 39. The variety of carbonised taxa (see 4.40-4.32 sandy peaty with clay diagram) is maximal compared with other cores 4.50-4.41 sandy peat: Hazendonk 1 and sections, which may be related to the nearby presence of a dry surface where human activity 4.60-4.51 humic sand could have taken place. During later phases, the Table III.10 The Hazendonk, unit C, southern distance to the dry surface gradually increased. section of square 41, lithostratigraphy. 140

Copyright © 2010. Leiden University Press. All rights reserved.

460 20

100

200

100

200

e Pe l l a r r s ia s ic p ar . ia m ac Pe ul rs os ic a ar ia la pa th ifo lia Pe rs i So c a nc r ia hu la s pa At as t h r ip pe ifo C l a p ex r lia Ar s p ,c e r ar Tr ctiu lla ost bo it i m bu ra ni Tr cu n rs ta se e i d cf ticum d moa-p . T m ic r a H r it s o c o s sto o r i c p. c u m r i s Tr de um on it i u c u m sp m sp ., Tr sp . / c a it i ., Tr rbo cu c h iti n m af c u i s e sp f, m d c a sp ., ch rb . af on f is ed

Li

ep

th (c m th -N ol AP o Ti g lia y ) pl Ti a lia t y p R s hy h llo C amp. or n s u C yl s or us c V i y l a at b u v h C ur n s a ell ar t o r u ve a n i c nu m ll a a s op an s a u a, ng lu c a ui s r b ne on a is ed C or S a nu m ss bu a c u ng s uin ni e gr a C a ,c ra ar Pr t a bo eg u Fr nu u ni se ax s s m R in sp d ub u i o n R u s no o u b s ex s a g y c u a c , n R s f r es els juv a u i i U bu ut i c us o r eni rti s le ca f r u o s u di t i c s oi o ca s u s, ca rb on La is p ed G sa al n G iu a l m c G ec h ap om a o Si leo maarinmu n l p Soene sis he e is d S o l a n l at b i f e r la um i fo l i da ac e n Fa um nigia s -typ a p C l l o p n i g r u m p. e he ia r a no c um lba po on , c di vo ar um lv bo u al lus nis bu ed m

D

260

St

C h C eno he p no od po iu di m a um lb fic um ifo , c liu arb St m on el is la ed r ia m ed ia

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Upland trees and shrubs

20 40 60 80

20 40 20 40

200 400 600

Peat

Sand

Clay

Peaty sediment

Sandy sediment

Clayey sediment 20

Upland herbs

280

300

320

340

360

380

400

420

440

2 4 6 8 10 x 1000

20 20

Upland herbs 20 40

Crop plants

20

Humic sediment

Figure III.19 part 1.

141

Copyright © 2010. Leiden University Press. All rights reserved.

142

50 100 150

Figure III.19 part 2.

20 40

20

2000

6000 50010001500 20 40 60 2000 6000

en

iu

a

la

tif

un

100 200 300

iu

m

ve

ga

a/ ar

ab

tic

ol

ua

cc

ic

ns

is

pe

a

-t y

s

at

ru

qu

pe

-a

cy

go

do

be

eu

aq

a

m

th

ic

ps

ta

50 100 150 20 40 20 40 20 40

cf .S

M

on

ex

an 20

Ve r

ar

pl

20 40 60

C

a

50 100 150 20

m

1500

is

500

Al

G ly c So er i la a c nu f. m f lu du it a lc ns am ar a C ar ex ac ut ifo rm Ph is ra g C m la it Ly diu es t h m au r u m st m a ra s a r i s li s li c c u ar s ia nu

nu

nu

s

s

s

gl

gl

gl

ut

ut

in

in

os

os

a,

a

co

ne

s

ly

ce

m

ar

a

in

ce

ab

na

nn

di

ta

ca

un

ec

pa

is

er

r iu

lu r i a st r f lu i s it a ns

a

ar

la

to

th

al al G

C

Ph

ru

pa Be

Eu

um

ut Al in n os H us um g a, Va u lut ca le lus ino rb r ia l s on u a na p , is ed of ulu fr. o s f ic fm in al al G e is a ca S c lium tk in C ho e pa s ar n l C ex op us ar e l t r ex la ec e Ty ri ta tus ph pa l a r i a ac u sp st r is .

Al

Al

Al

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Alder carr and marsh vegetation

20 20 40

Alder carr and marsh vegetation

Copyright © 2010. Leiden University Press. All rights reserved.

an

50 100 150 20

20 40 60 80 100

20 40 60 80 100

20 40

t.

vi al is

100

de

e

tr i

Open water

ea

200 400

In

50 100 150

s/

200

si

100

ac

Ju n B i c us de b n u R s c fon an e iu un r nu s cu a lu s sc el

er

at

us

G ly c Ly er c o ia R pu cf. um s m Th ex eur axi el hy op ma y a C pte dro eu ar r l s Sa ex is p apa St gitt rem alu thu ac ar o str m is h i t ,l H ys a saa ea yp p g ff Sp er al it t ra u i gm S par g i c um st r fo li i a en O ar ganiu c f s ts en a m . t R a n i u e et a n nt m r r a u n h e e e c pt cu aq re tumeru lu ua c tu m s t m re ic , pe a c a rb ns - t y on Pe pe ise rs d ic ar ia hy dr op ip er Al op ec ur us ge ni cu la tu Ju s nc us ef fu su s

Alder carr and marsh vegetation

Po

20

Ap ia Po c e a ae pr at en

20

C ar R yo p um h Sc ex ylla ro sp c e p h . ae ul ar ia sp .

un cu R lu a s aq N nu y m nc ua C p ul til ha h u is r a ae s c -t y sp a a f. c pe l . b irc a in at us

R

N u C ph al ar lit lu r i c te he a sp .

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Wetland pioneers

200 20 100 200

Ecologically indeterminate

20 20

Figure III.19 part 3.

143

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Sa C nd h B o arc n o Po e / f a l i Butte sh r r B uds y re em a M d s ma ins i o c N s s ale ns e Le ckerems Baaf r ra ain e c s D r k r ma r isp ap e in a Lo h m s n a C pho ia ins s r Tr ista pus p., i e In chotella cri ph s i s W e c pte m u t a l p p i li a h t r C ite rema, c cednus o Fa co sm ain ase o, s , st o a a C e c a n s l l u s s t ato to b en l rn bl la p as st oc e s ts s o c lle t c s Sa um m ge pl e op vo hi lu lu m Ph m e ,s as (c cl e m3 er ot ) ia

Varia

350 500 500 1400 1400 250 500 1000 1000 1000 1000 1000 1000 125 500 500 250 800 1000 1000 1000 500 6001000 1000 1000 8001000 600 600 800 1000 300 500

VL 1b

VL 1a ? Haz 3 Haz 2b Haz 2a Haz 1

Analyst: W.J. Kuijper

Figure III.19 The Hazendonk, unit C, southern section of square 41, macroremains diagram, + = few (1-10), ++ = some tens (10-49), +++ = many tens (50-99), ++++ = some hundreds (100-499), part 4.

Copyright © 2010. Leiden University Press. All rights reserved.

The absence of cereals during phases Haz 2a and 2b in the section of square 41 corresponds with the scarcity of cereals dating to phase Haz 2 at the site. Both phases 2 and 3 are characterised by high values of Urtica dioica and Typha sp. The diagram indicates that phase Haz 3 may have resulted in a decrease in alder. Phase 3 contains carbonised fruits of Rubus fruticosus and cereals. Phase VL 1a is only tentatively distinguished since it was not recognised in this section during excavation. Charcoal is present below and above the VL 1a level, while bone remains are present at an even higher depth. The high values of Cornus sanguinea and Sambucus sp. suggest that the vegetation became more open. Phase VL 1b is characterised by a thick horizon rich in herbs of dry to moist terrain indicative of enriched and disturbed conditions, which corresponds with the trampling zone of this phase that stretched into the marsh. In contrast to the previous occupation horizon, cereals are absent in this horizon and carbonised macroremains are scarce, indicating that human impact had another character at the sampling point than during previous phases. It could concern a trampling zone of cattle roaming outside the palisade and into the edges of the marsh. The very upper part of the diagram could possibly reflect phase VL 2b, but the sediment contains too little macroremains to investigate this further. III.3.7 coMparISon of proportIonS of dryland versus wetland VegetatIon The geological information as well as the pollen and macroremains diagrams indicates that the vegetation at the Hazendonk gradually submerged. The summary diagrams presented above do not show the decreased pollen rain (influx) of dryland taxa and increased pollen rain of wetland taxa since they only represent a summary of the dryland vegetation. Therefore, the proportion of dryland versus wetland taxa was investigated by calculation of percentage diagrams based on an adapted pollen sum. Generally influx diagrams are used instead to investigate changes in the influx of taxa, based on the addition of marker pollen and calculations of pollen concentrations. The calculation of concentrations of pollen grain was however not possible in this study since marker pollen grains were not added. The percentage calculations used in the new calculation are based on a pollen sum that

144

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

includes all taxa that can be assigned to either dryland or wetland vegetation with some certainty. The group of dryland vegetation includes dryland trees, shrubs, herbs and spore plants. The group of wetland vegetation includes wetland trees, shrubs, herbs and spore plants, Poaceae, Cyperaceae and water plants. Taxa with an unknown ecology (Ecol. indet.) were excluded from the pollen sum. The groups of Poaceae and Cyperaceae may also include dryland taxa but it is expected that they mostly represent wetland taxa. The material from square 57, core 3 and core 2 was all sampled in the same unit of the excavation. Despite small differences in distance to the dune, the results can be compared to each other in order to reconstruct the vegetation (with some overlap). figure III.20 (next page) shows the summary diagrams of square 57, core 2 and core 3, all based on the new pollen sum that includes both dryland and wetland taxa. The summary diagrams show the percentages of pollen grains of dryland versus wetland vegetation while the major occupation periods are indicated. The figures demonstrate that there is a trend that the dryland vegetation pollen rain decreased through time from c. 40 to 20% at the expense of wetland vegetation. This is probably explained by the rise of the ground water table, causing the submergence of the dune. The resulting decrease in dryland vegetation is characteristic of the development of vegetation at dunes in the river area during the middle of the Holocene in general. The diagrams additionally show that the percentage of wetland vegetation increased during the Hazendonk occupation periods, in the first place caused by high values of Alnus sp. collected during the excavation (as indicated by the pollen diagrams presented earlier). These increases in Alnus sp. are probably caused statistically by a reduced pollen precipitation of dryland pollen grains, since the Alnus macroremains show a contrasting decrease during occupation periods. In contrast to the Hazendonk phases, the diagrams show a reduction of Alnus sp. during the phase VL 1b. This reduction might indicate that the anthropogenic influence during the Vlaardingen occupation periods had more impact on Alnus sp. than before, assuming that the Alnus vegetation was part of the local vegetation and became relatively strongly reduced. In square 57 at approximately 4.19 m -NAP, an exceptional high value of Alnus pollen disturbs the general pattern. This extreme value cannot be related to anthropogenic influence but instead is an unexplained variation in the percentage of Alnus sp. that does not necessarily indicate fluctuations of the natural vegetation (cf. Waller et al. 2005).

Copyright © 2010. Leiden University Press. All rights reserved.

III.3.8 MacroreMaInS froM the excaVatIon Table III.11 (at the end of appendix III) shows the identifications of the macroremains that were collected during the excavation in search of concentrations and by the analysis of sieve residues. Most samples were collected from refuse layers and colluvia while five samples were collected from pits and two from a channel fill. The samples date to the occupation phases Haz 1 to VL 2b except for Vlaardingen 1a, and are discussed for each period separately where possible. The samples of phase VL 2b could be separated stratigraphically in two periods. Table III.12 shows handpicked finds. figure III.21 shows a selection of macroremains. The analysis of crop plants and potential weeds is discussed separately below. phase number of samples

Haz 1 15

Haz 2 8

taxon buds Corylus avellana 21, 6 c 21 Crataegus monogyna 1c Malus sylvestris 2c 1c c = carbonised - = not present x, yc = x macroremains including y carbonised macroremains

Haz 3 16

VL 1b 6

VL 2b 3

unknown 7

32 -

7 -

41 1 -

15 -

Table III.12 The Hazendonk, various phases, handpicked macroremains.

145

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

e Ph as

De pt h Up (cm la nd -NA ve P ) ge ta tio n W et la nd ve ge ta tio n

core 2

300 310 320 330 VL 1b 340 350

square 57

e

370 Ph as

De pt h Up (cm la nd -NA ve P) ge ta tio n W et la nd ve ge ta tio n

360

385

380

VL 1a

390

395

20

40

60

80 100

405 415 425

core 3

Haz 3

445

e

455 Ph as

De pt h Up (cm la nd -NA ve P) ge ta tio n W et la nd ve ge ta tio n

435

525

465 475 485

535

Haz 1

Haz 1

505

545

515 525

555 Copyright © 2010. Leiden University Press. All rights reserved.

495

535

20

40

60

80 100

565 575 585

Haz 0

595 605 615

146

20

40

60

80 100

Figure III.20 The Hazendonk, summary diagrams of core 3, square 57 and core 2, based on a new pollen sum that includes both upland and wetland taxa.

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Figure III.21 The Hazendonk, non-cultivated macroremains. a) Acer campestre, waterlogged, M41 (VL 2b), 7.8 x 5.8 x 1.3 mm; b) Fraxinus excelsior, waterlogged, section 41 (361-351 cm -NAP), 10.5 x 5.5 x 1.0 mm; c) Prunus padus, waterlogged, M41 (VL 2b), 7.0 x 5.6 x 5.2 mm; d) Malus sylvestris, carbonised, M74 (Haz 3), 1.75 x 1.8 mm; e) Malus sylvestris, carbonised, M74 (Haz 3), 1.6 x 1.9 mm; f) Solanum nigrum, berry, carbonised, M163 (Haz 1); g) Veronica beccabunga-type, waterlogged, section 41 (VL 1b), 0.6 x 0.4 x 0.2 mm; h) Viburnum opulus, waterlogged, section 41 (410-400 cm -NAP), 7.0 x 6.0 x 1.3 mm; i) Viburnum opulus (with scale), carbonised, M74 (Haz 3), 6.3 x 6.1 x 2.3 mm, photographs: W. Meuzelaar and J. Pauptit.

147

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

The samples of phase Haz 1 were all collected in two squares in unit C. Most of the plant remains from the Hazendonk 1 horizon are carbonised, and it is therefore unclear whether they reflect the natural vegetation. The assemblage representing woodland of dry terrain (group 1) is nevertheless similar to later phases, and resembles the material from the cores and section. The taxa of woodland of dry terrain can therefore probably be interpreted as the natural vegetation. The importance of water- and marsh plants (group 4, 5, and 6) was still restricted in this phase. Phase Haz 2 is represented by very few samples (N = 1 to 4) since refuse of this phase was scarce. The finds include carbonised and waterlogged macroremains. The vegetation during phase Haz 2 was rich in shrubs, pointing to open vegetation. Quercus remains were not found but this is probably due to the limited number of samples. The number of riparian/marsh taxa increased compared with phase Haz 1, wich is probably related to the rising water table. All samples certainly dating to phase Haz 3, except for one, were collected in square 26 (unit B), which possibly restricts the representativity of the data. The finds consist of carbonised and waterlogged macroremains. Important woodland elements were Quercus sp., Malus sylvestris, Corylus avellana, Cornus sanguinea, Sambucus nigra and Rubus fruticosus. The samples indicate diverse vegetation (woodland elements of dry terrain, alder carr and marsh, but no open water). The samples of phase VL 1b contained a large quantity of fine charcoal remains, pottery remains, flint remains and fish- and bone remains. The botanical macroremains were found in both a carbonised and waterlogged state. The juvenile cupulae of Quercus sp. indicate that this species was still present on the dune. In contrast, remains of Tilia sp. were not present in this layer or younger layers. Alnus sp. was present, but the number of remains suggests limited importance. Plants of disturbed environments and moist eutrophic soils were well represented during this phase, and this is similar to contemporaneous samples from cores and sections. Important riparian/marsh species were Carex riparia, Carex acutiformis, Sparganium erectum, Lythrum salicaria and Solanum dulcamara. These taxa indicate terrestrialisation, which was possibly characterised by the presence of floating mats, and eutrophic conditions. The macroremains assemblage of phase VL 2b indicates a mixture of eutrophic marsh/riparian vegetation and some dryer elements. Interestingly, Ranunculus cf. fluitans became present, which is characteristic of running eutrophic water. This corresponds with the fluvial activity during this period as indicated by the geological reconstruction (see introduction). Apparently, the samples of phase VL 2b contained more Alnus remains than previous horizons, although an exact comparison is not possible because of the sampling methods. A pit in square 25 (sample M41) seems to have contained a concentration of waterlogged remains from Quercus sp., Acer campestre and Alnus glutinosa. III.3.8.1 Crop plants The crop plants found at the Hazendonk are the cereals emmer wheat (Triticum dicoccon) and naked barley (Hordeum vulgare var. nudum). The combination of these two species is comparable with other Dutch wetland sites from the Early and Middle Neolithic. Carbonised grains and chaff remains were present in the refuse layers phases Haz 1, 3 and VL 1b (see table III.13). phase

Haz 1

Haz 2

taxon Hordeum vulgare var. nudum + Triticum dicoccon + (+) + = present (+) = present in low numbers in a single sample

Haz 3

VL 1b

VL 2b

+ +

+ + - = not present

(+) -

Table III.13 The Hazendonk, various phases, cereal identifications, all carbonised.

148

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

One sample of phase Haz 2 contained a few carbonised grains of T. dicoccon while a single sample of phase VL 2b contained a few carbonised grains of H. vulgare var. nudum. The scarcity of the two cereal species in some phases is probably related to limited sampling and the limited presence of refuse dating to these phases. Waterlogged remains of crop plants were present in the section samples from square 41, dating to phases Haz 1 and 3. Cereal pollen was found in phase Haz 1, Haz 3 and VL 1b. The diagram of Voorrips (Louwe Kooijmans 1974) shows a single cereal pollen grain for phase Haz 0, but the absence of further data from this layer restricts interpretation. The cereal finds of phase Haz 1 consisted of concentrations of emmer wheat and naked barley that were exceptionally well preserved. Some samples consisted mainly of emmer, while other samples represent a mix of both cereals. Emmer wheat was generally dominant (see fig. III.22a). The emmer grains were often underdeveloped and sometimes unripe, and showed a variation in morphology, sometimes resembling T. monococcum or T. aestivum (see fig. III.22b-d). All the wheat was identified as emmer wheat.32 The variation in morphology is explained by unequal ripening of the ears and poor development of part of the emmer.33 The presence of T. aestivum is rejected because of the small quantities of grains of this type and because of the results of carbonisation experiments with emmer wheat (Braadbaart 2004, 2008). The presence of T. monococcum is rejected because of the small quantities of grains of this type and since the material contains twisted spikelet forks that explain this type of grains in emmer wheat (cf. Cappers et al. 2004). The good preservation (see fig. III.22e) allowed investigation of the stage of crop processing. Some of the emmer grains were still hulled in their chaff (see fig. III.22f) and the material contained many emmer chaff remains, including occasional collars (the lowest rachis segments of the ear, fig. III.22g-h) and many spikelet forks and rachis segments (fig. III.22i-j). There are occasional finds of rachis remains that were still attached to each other. The finds of grains hulled in chaff and the presence of the many chaff remains including rachis fragments and collars indicates that the crop product was hardly processed and represents complete ears of emmer wheat. The good preservation indicates that the material must have been carbonised almost in situ. It is presumed that trampling, transport or movement of the carbonised material would have led to a minor preservation state. The grains of naked barley were normally developed (see fig. III.22k-l). Interestingly, the grains of naked barley were found partly hulled in chaff as well (see fig. III.22m). The concentrations furthermore contained large quantities of internodia of naked barley (see fig. III.22n-p). Again, rachis remains were sometimes still attached to each other (fig. III.22q-r), and again collars were present in the samples (see fig. III.22s-t). These finds support the conclusion that the crop product was hardly processed and represents at least the fragments of ears, and possibly complete ears of naked barley. Three concentration samples contained considerable quantities of rye brome (Bromus secalinus-type, see fig. III.22u-w), which was more common in these samples than naked barley. The relatively large numbers of this taxon may suggest that this weed was tolerated or cultivated, as has been suggested for the Rhineland (Bakels 1981; Knörzer et al. 1999; see also Zohary and Hopf 2000, 70, see however paragraph 11.2.1). There were no finds of chaff remains of rye brome.

32 The cereals were initially identified and published as naked barley, emmer and einkorn (Louwe Kooijmans 1976, 259). The identifications of einkorn were later revised as emmer (Bakels 1981, 1986; Bakels and Zeiler 2005). 33 Bakels (1981, 143): “Apparently this particular emmer crop is not normal. In addition to well-developed kernels, there are numerous too slender ones that resemble somehow einkorn. Initially they were even identified as such. Since however many intermediate forms between the two species were found and since the T/B index, for example, shows a single peak, all kernels are considered to be emmer. Furthermore it could be seen from the rachis segments that there were initially two kernels per spikelet base but that sometimes only one had reached full development. The kernels with the morphology of einkorn are therefore probably kernels from underdeveloped spikelets.”

149

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

a

Copyright © 2010. Leiden University Press. All rights reserved.

b

d

c

e

Figure III.22 The Hazendonk, phase Haz 1, carbonised cereal remains. a) Triticum dicoccon, grains, M164. Please note the flattened ventral sides characteristic for T. dicoccon. b) T. dicoccon, grain, square 39, 6.3 x 3.5 x 3.1 mm, normal grain; c) T. dicoccon, grain, square 39, 4.6 x 3.4 x 2.4 mm, similar to T. aestivum; d) T. dicoccon, grain, square 39, 5.7 x 2.1 x 3.1 mm, similar to T. monococcum; e) T. dicoccon, grain, 39.184.

150

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

f

g

h

Figure III.22 The Hazendonk, phase Haz 1, carbonised cereal remains. f) T. dicoccon, grain in chaff, M164 (Haz 1), 8.3 x 3.3. x 2.2 mm (hairs and chaff included); g) and h) T. dicoccon, collars, 39.184 (Haz 1). Next page: Figure III.22 The Hazendonk, phase Haz 1, carbonised cereal remains. i) T. dicoccon, spikelet forks, 39.185 (Haz 1), mean size 1.8 x 0.8 mm (N = 100), please note the variable width of the forks and the presence of twisted forks; j) T. dicoccon, rachis fragments, 39.185 (Haz 1), please note the presence of several fragments with culms.

151

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

i

j

152

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

k

m

Copyright © 2010. Leiden University Press. All rights reserved.

l

n

o

Figure III.22 The Hazendonk, phase Haz 1, carbonised cereal remains. k) Hordeum vulgare var. nudum, grains, M164, mean size 4.5 x 2.1 x 2.3 mm (N = 100). l) H. vulgare var. nudum, grain, M164, size 5.9 x 3.4 x 2.7 mm; m) H. vulgare var. nudum, grain with attached remains of chaff, M164, 7.5 x 3 x 2.5 mm; n) and o) H. vulgare var. nudum, internodia, M164, 4.3 x 2.5 x 0.6 mm and 3.7 x 1.8 x 1 mm, please note the difference in width.

153

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

p

Copyright © 2010. Leiden University Press. All rights reserved.

q

r

s

t

Figure III.22 The Hazendonk, phase Haz 1, carbonised cereal remains. p) H. vulgare var. nudum, internodia, M164, magnification 5.8 x; q) and r) H. vulgare var. nudum, rachis fragments, M164, please note the size differences and the good preservation; s) and t) H. vulgare var. nudum, collars, M164.

154

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

u

Copyright © 2010. Leiden University Press. All rights reserved.

v

w

Figure III.22 The Hazendonk, phase Haz 1, carbonised macroremains. u) Bromus secalinus-type, grains, M163, mean size 4.9 x 1.8 x 0.9 mm (N = 48), v) and w) B. secalinus-type, grains, M163. All photographs in figures III.22 a to w by W. Meuzelaar; drawings by E. van Driel after W.J. Kuijper.

155

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

The identified cereal finds (grains and chaff including collars) provide information on harvesting methods. Harvesting was probably not done by the beating of ears or picking or stripping without tools, since these methods would have resulted in the collection of single and equally matured grains instead of the attested complete ears that were sometimes poorly developed. The stripping of emmer wheat with knives can furthermore tentatively be excluded since this method usually produces a break above or below the basal spikelet, which does not result in the harvesting of collars (pers. comm. Anderson 2007; see also Mery et al. 2007, fig. 15). There is however some natural variability in the location of the break points and therefore the harvesting of collars cannot be excluded. The arguments that are used to reject the practice of stripping of emmer wheat tentatively reject the practice of mesorias34 as well, assuming that this harvesting method results in the breaking of the ear at the same fragile point. Stripping and the use of mesorias was probably not practised to naked barley either since stripping is effective only for hulled cereals that have a semi-fragile rachis, while it is less effective for free-threshing cereals that have a relatively solid rachis (pers. comm. Anderson 2007; Mery et al. 2007, 1110; Zapata Peña 2007). Harvesting methods that could alternatively have been applied are harvesting with sickles and uprooting (Anderson 1992). Harvesting methods are further discussed in chapter 11. Most of the cereal samples of the Hazendonk 1 horizon were found in concentrations that may possibly represent a single large concentration. The volume of most samples is unfortunately unknown and quantitative comparison of the samples is therefore not possible. The concentrations have alternatively been investigated by comparison of the content, spatial analysis and comparison of the T/B index of emmer wheat grains of three samples ((100 x thickness)/breadth, cf. Helbæk 1952). firstly, the characteristics of the grains suggest that the samples 39.154, 39.184, 39.185, M157, M160 and M164 all contained poorly developed emmer wheat. Sample M164 and M157 are different because of the presence of considerable numbers of Hordeum remains, while sample M164 additionally stands out because of its large number of internodia of naked barley. This suggests that the various samples reflect different deposition events. Secondly, spatial analysis (see fig. III.23) shows that all concentration samples from phase Haz 1 were collected in two squares from unit C. The distribution of the samples at approximately equal height on the slope of the dune and the absence of known structures contrastively indicates that the cereals were deposited in a dump zone during a single event.

39

41

section 39 M157

Copyright © 2010. Leiden University Press. All rights reserved.

57 39.154 M160 M164

M163 39.185 39.184

0

1m

Figure III.23 The Hazendonk, unit C, phase Haz 1, location of the macroremains samples. 34 Mesorias: an implement consisting of two wooden sticks between which the ears are hold and then pulled off.

156

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

It is however possible to distinct a separate cluster consisting of 39.154, 39.184, 39.185 and M163, while M157, M160 and M164 are located outside this cluster. Thirdly, the results of the measurements of the T/B index of emmer grains of three samples are shown in figure III.24 and table III.14. figure III.24 shows that the three measured samples (39.154, 39.185 and M157) are likely to be part of the same population, although M157 could also be independent. In conclusion, the emmer wheat of various samples probably originates from a single crop product, and possibly underwent the same processing sequence. The presence and type of remains from naked barley as well as the spatial distribution however shows variation between the samples, indicating the existence of three groups of samples. Nevertheless, the close distance between the samples and the similar height suggests that the various deposition processes were comparable. The material from phases later than Haz 1 only contains small quantities of cereals that do not represent concentrations. This contrast to the samples from phase Haz 1 restricts the comparison between the cereal finds of phase Haz 1 and other phases. The only cereal species present in the Haz 2 horizon is Triticum dicoccon. In contrast to the previous phase, Haz 2 contained only half a grain of cf. Bromus secalinus-type, which may belong to phase Haz 1 after all as a result of taphonomy and site formation processes (cf. Amkreutz et al. 2008). In the material of Haz 3, Triticum dicoccon (grains and a small number of chaff remains) and Hordeum vulgare var. nudum (only grains) were present. The cereal species and the number of cereal remains of VL 1b are similar to Haz 3, containing Triticum dicoccon and Hordeum vulgare var. nudum (mainly grains but also few chaff remains of both species, only carbonised). Sample M20 (unit E) of phase VL 2b is the youngest sample that contains cereal grains, which were identified as Hordeum vulgare var. nudum (carbonised). % 30 25 20 15 10 5 0

50

60

70

Sample 39.154

80

90

100

Sample 39.185

110

120

130

140

150

160

170

180

190

200

210

Sample M157

Copyright © 2010. Leiden University Press. All rights reserved.

Figure III.24 The Hazendonk, T/B index of Triticum dicoccon of 3 samples that contained concentrations of cereals. T/B index: (100 x thickness)/breadth.

sample square

39.154 57

39.185 57

M157 39

TB mean

95.4

99.1

85.1

min

73.0

58.0

51.7

max

211.1

172.7

142.9

Table III.14 The Hazendonk, phase Haz 1, T/B index values of Triticum dicoccon of samples 39.154, 39.185 and M157, based on measurements of 100 grains of each sample. T/B index: (100 x thickness)/breadth.

157

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

III.3.8.2 Carbonised macroremains of non-cultivated plants The presence of carbonised remains of edible parts of plants is generally used as an indication of use and/or consumption (see chapter 9). Table III.15 shows the carbonised plant remains from the Hazendonk for each phase (data from cores and sections are included). The table shows that a large variety of taxa has been found in a carbonised state. The carbonised macroremains of V. opulus are unique for Dutch prehistoric sites; a parallel has been found at Doel Deurganckdok in 2003 (Bastiaens et al. 2007). The function of the macroremains found in a carbonised state is not immediately clear, and the large variety makes one question whether all taxa where all intentionally used. The unspecific context of the carbonised remains, mostly collected in refuse layers, gives little additional information. A first possible function of the taxa found in a carbonised state is plant food. Taxa known as probable plant food sources found in a carbonised state are Quercus sp., C. avellana, C. monogyna, P. spinosa, M. sylvestris, C. sanguinea, R. fruticosus, Rosaceae, R. ficaria, T. natans, N. alba, B. secalinus-type and herbs like U. dioica, C. album, P. maculosa, P. lapathifolia and F. convolvulus. Other taxa found in a carbonised state that are less well-known as food plants may have been consumed as well. Taxa that are known as a food source but were only found in a waterlogged state at the Hazendonk are e.g. Sambucus nigra, Rubus caesius, Typha sp. and Nuphar lutea (this selection of potential food plants is based on Bakels 2005; Bakels and Van Beurden 2001; Van Zeist 1970). frequency analysis of the carbonised remains in all samples shown in table III.11 demonstrates that the carbonised remains of hazelnut were most frequently found (in c. 50% of the samples), followed by the cereals. Other taxa found in a carbonised state in more than 10% of the samples are Cornus sanguinea, Prunus spinosa, Malus sylvestris, Crataegus monogyna, Nymphaea alba, Galium aparine, Bromus secalinus-type and Fallopia convolvulus. The edibility and frequency together suggest consumption for most of these taxa. The finds of carbonised halves of crap apples support that people dried and stored these fruits. Other explanations for the presence of carbonised macroremains are that taxa were used for other functions, such as fuel, thatching and wickerwork. The presence of the many marsh taxa could furthermore be explained by the burning of organic material found in the drift litter zone or the burning of local marsh vegetation during domestic activities. The presence of certain potential arable weeds may be explained by accidents and the discarding of waste during cereal preparation. The carbonised macroremains indicate occupation between summer and winter for all phases, either continuously or intermittently. Summer indications are strongest for the phases Haz 1 and 3. These results correspond with the zoological results (Zeiler 1997). There is no strong evidence against spring occupation since botanical evidence of spring occupation based on macroremains is usually scarce or absent at comparable sites. Analysis of the separate phases interestingly shows that the number of carbonised taxa that most probably functioned as food plants during the Late Neolithic Vlaardingen phases is smaller than the number of the Swifterbant culture and Hazendonk group (see fig. III.2535), although the difference is probably not significant. The difference cannot be related to the number of samples, the type of samples, the extent and thinkness of the refuse layer or site function. It therefore may possibly represent a reduction of the number of food plants related to a new phase of the neolithisation process and a shift in subsistence in the Late Neolithic. Alternative explanations other than the effect of neolithisation can however not be excluded, such as the reduced variety of the vegetation due to the reduced surface of the dune and other nearby dunes, or a shift in seasonality. The apparent decrease in carbonised food plants in the Late Neolithic therefore deserves further investigation (see also paragraph 9.3.4).

35 The selection of food plants used in the figure includes Cornus sanguinea, Corylus avellana, Crataegus monogyna, Malus sylvestris, Prunus spinosa, Quercus sp., Rosaceae, Rubus fruticosus, Ranunculus ficaria and Trapa natans.

158

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase

Haz 1

Haz 2

Haz 3

VL 1b

VL 2b

Cornus sanguinea

+

+

+

+

-

Corylus avellana

+

+

+

+

+

Crataegus monogyna

+

+

+

+

-

Malus sylvestris

+

+

+

-

-

Malus sylvestris, parenchyma

-

-

+

-

-

Prunus spinosa

-

-

+

+

-

Quercus sp.

+

-

-

+

-

Quercus, cupulae with content

-

-

-

+

-

Rubus fruticosus

-

-

+

-

-

Tilia platyphyllos

+

-

+

-

-

Tilia sp.

+

-

-

-

-

Viburnum opulus

+

-

+

-

-

Ajuga reptans

-

-

+

-

-

Chaerophyllum temulum

+

-

-

-

-

Galium aparine

-

+

+

+

+

Galium odoratum

-

+

-

-

-

Ranunculus ficaria, tubers

+

+

+

-

-

Urtica dioica

+

-

-

-

-

Brassica rapa

-

-

+

-

-

Bromus secalinus-type

+

cf. +

-

-

-

Capsella bursa-pastoris

+

-

-

-

-

Chenopodium album

+

-

-

+

-

Elytrigia repens

+

-

-

-

-

fallopia convolvulus

+

+

-

+

-

Persicaria lapathifolia

+

-

-

+

-

Persicaria maculosa

+

-

-

-

-

Solanum nigrum

+

-

-

+

-

Veronica hederifolia

-

-

+

-

-

taxon Group 1

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2

Tabel III.15 part 1.

159

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase

Haz 1

Haz 2

Haz 3

VL 1b

VL 2b

Hordeum vulgare var. nudum

+

-

+

+

+

Triticum dicoccon

+

+

+

+

-

Alnus glutinosa

+

+

-

+

-

Alnus glutinosa, cones

+

-

-

-

-

cf. Berula erecta

+

-

-

-

-

cf. Euphorbia palustris

+

-

-

-

-

Iris pseudacorus

-

-

+

+

-

Lathyrus palustris/Vicia cracca

-

-

+

-

-

Menyanthes trifoliata

-

-

+

-

-

Oenanthe aquatica

+

-

-

-

-

+

-

-

-

-

Rumex cf. crispus

+

-

-

-

-

Schoenoplectus lacustris

+

-

-

-

-

Sparganium erectum

+

+

-

-

-

Nymphaea alba

+

-

-

+

+

Trapa natans

-

-

+

-

-

Mentha aquatica/arvensis

-

-

-

+

-

Phleum sp./Poa annua

+

-

-

-

-

Rosaceae

+

-

-

-

-

Rumex sp.

+

-

-

-

-

Scrophularia sp./Verbascum sp.

-

-

-

+

-

Veronica austriaca/chamaedrys + = present

+ - = not present

-

-

-

-

taxon Group 3

Group 4

Group 5 and 6

cf. Phragmites australis, stem fragments

Group 7

Copyright © 2010. Leiden University Press. All rights reserved.

Group 8

Table III.15 The Hazendonk, taxa found in a carbonised state for each phase, group = ecological group (see table III.2), part 2.

160

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

N 9 8 7 6 5 4 3 2 1 0

Haz 1 and 2

Haz 3

VL

Copyright © 2010. Leiden University Press. All rights reserved.

Figure III.25 The Hazendonk, number of carbonised non-cultivated food plants for phases of the Swifterbant culture (Haz 1 and 2), the Hazendonk group (Haz 3) and the Late Neolithic Vlaardingen group (VL).

III.3.8.3 Arable weeds The analysis of weeds can result in reconstruction of crop cultivation practices. Such an approach starts with the distinction of arable weeds: which taxa grew in the arable plots? Discussion of the weeds from the Hazendonk is hampered by two problems: it is not certain whether arable farming was practised at the site (see below), and it is unclear which taxa represent true field weeds instead of local disturbance indicators. Natural and anthropogenic processes at the Hazendonk other than arable farming probably resulted in conditions that were similar to ecological conditions of fields, which may explain the presence of disturbance indicators (see also chapter 10). In the first place, a selection of potential arable weeds was made, including herb taxa of all habitats where one could locate an arable field: woodland and woodland edges of dry terrain, grassland, and open and/ or disturbed dry to slightly humid terrain (see chapter 10). The selection is based on all data from pollen and macroremains that are available from the Hazendonk (including taxa from the study of Van der Wiel 1982) but only includes taxa that are related to occupation phases. The upper part of table III.16 shows this group of potential arable weeds found at the Hazendonk. The resulting selection comprises a large group of taxa that were probably not all true field weeds, since such a relationship with crop cultivation is not strongly supported by context. Secondly, it has been investigated which potential arable weeds were present in a carbonised state in samples that contain concentrations of carbonised cereals. It is expected that these samples represent crop products mixed with weeds that were not removed from the crop product yet (cf. Hillman 1981), and that the macroremains found in the same preservation state as the cereals underwent the same preparation and deposition processes. Concentrations of cereal remains were only found in the material of phase Haz 1. Densities could not be calculated since the volume of most samples is unknown. All samples from phase Haz 1 were included in this analysis. Table III.16 (conc.) shows the 31 taxa found in a carbonised state in samples that contain concentrations of carbonised cereals. There are two reasons that indicate these taxa are probably not truly associated with the cultivation of cereals. In the first place, cereals are usually cultivated in a relatively high, dry and open environment, which contrasts with the varied and wet conditions that are indicated by this assemblage of carbonised macroremains.

161

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

category and context

C

W

P

conc.

fr.

Ajuga reptans

+

+

-

-

-

Alopecurus myosuroides/pratensis

-

+

-

-

-

Anthriscus sylvestris

-

+

-

-

-

Anthoxanthum odoratum

-

+

-

-

-

Artemisia sp.

-

-

+

-

-

Brassica rapa

+

-

-

-

-

Bromus secalinus-type

+

-

-

+

+

Capsella bursa-pastoris

+

-

-

+

-

Carex remota

-

+

-

-

-

Chaerophyllum temulum

+

-

-

+

-

Chenopodiaceae

-

-

+

-

-

Chenopodium album

+

+

-

+

+

Chenopodium ficifolium

-

+

-

-

-

Elytrigia repens

+

-

-

+

-

fallopia convolvulus

+

+

-

+

+

Galium aparine

+

+

-

-

-

Galium odoratum

+

-

-

-

-

Galium spurium

-

+

-

-

-

Glechoma hederacea

-

+

-

-

-

Lamium album

-

+

-

-

-

Lapsana communis

-

+

-

-

-

Persicaria maculosa

+

+

-

+

+

Persicaria lapathifolia

+

+

-

-

-

Plantago lanceolata

-

-

+

-

-

Plantago major

-

-

+

-

-

Phleum sp./Poa annua

+

-

-

+

-

Polygonum persicaria-type

-

-

+

-

-

Polygonum aviculare-type

-

-

+

-

-

Ranunculus ficaria, tubers

+

-

-

+

-

Ranunculus arvensis

-

-

+

-

-

Rumex acetosella

-

+

-

-

-

Rumex acetosa-type

-

-

+

-

-

Silene latifolia ssp. alba

-

+

-

-

-

Copyright © 2010. Leiden University Press. All rights reserved.

taxon

Table III.16 part 1.

162

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

category and context

C

W

P

conc.

fr.

Solanum nigrum

+

+

-

+

+

Stellaria media

-

+

-

-

-

Urtica dioica

+

+

-

+

-

Valeriana officinalis

-

+

-

-

-

Verbena sp.

-

-

+

-

-

Veronica austriaca/chamaedrys

+

-

-

+

-

Veronica hederifolia

+

-

-

-

-

Veronica sp.

-

+

+

-

-

Alnus glutinosa

+

+

Alnus glutinosa, cones

+

+

cf. Berula erecta

+

-

Cornus sanguinea

+

+

Corylus avellana

+

+

Crataegus monogyna

+

+

cf. Euphorbia palustris

+

-

Malus sylvestris

+

+

Nymphaea alba

+

+

Oenanthe aquatica

+

-

+

-

Quercus sp.

+

-

Rosaceae

+

-

Rumex cf. crispus

+

-

Rumex sp.

+

-

Schoenoplectus lacustris

+

-

Sparganium erectum

+

+

Tilia sp.

+

-

Viburnum opulus

+

-

taxon

cf. Phragmites australis,

Copyright © 2010. Leiden University Press. All rights reserved.

stem fragments

C = carbonised macroremains W = waterlogged macroremains P = pollen

conc. = taxa found in a carbonised state in the concentrations of phase Haz 1 fr. = taxa found in the concentrations of phase Haz 1 relatively frequent + = present - = not present

Table III.16 The Hazendonk, potential arable weeds, part 2.

163

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

In the second place, comparison of the samples with concentrations of carbonised cereals shows a trend that samples higher on the slope of the dune contain more carbonised macroremains of dryland species and less carbonised macroremains of water plants then samples lower on the slope (M160 and M164, square 57 in contrast to other samples in square 57). It therefore seems that the samples of phase Haz 1 are of a mixed origin, resulting from different deposition processes, and do not represent a pure crop product. The removal of taxa from table III.16 that were only found in a single sample of phase Haz 1, or that are represented by a single carbonised seed or fruit only, results in a list of 13 taxa (see table III.16, fr.). further removal of potential food plants and wetland taxa results in a list of five taxa that most probably represent arable weeds during phase Haz 1: Bromus secalinus-type, Chenopodium album, Fallopia convolvulus, Persicaria maculosa and Solanum nigrum. C. album and F. convolvulus are represented by a few fruits only, which may indicate that they did not function as arable weeds (but see also chapter 10) or that they were already removed from the crop product. Concentrations of carbonised cereals were absent in the material of phases later than phase Haz 1 and a comparable weed analysis was not possible. for the samples of these phases it has instead been investigated which taxa are present in a carbonised state in samples that contain carbonised cereal remains. Only the material from phases Haz 3 and VL 1b consist of sufficient samples that allow relevant analysis. Taxa that are frequently found in samples from Haz 3 with carbonised cereal remains are M. sylvestris, C. avellana and C. sanguinea, which probably represent food plants instead of arable weeds. Taxa that are frequently found in relevant samples from VL 1b are C. avellana, P. spinosa, G. aparine and N. alba, which probably do not represent arable weeds either. It must thus be concluded that this approach is not useful for the reconstruction of arable weeds at the Hazendonk. The result furthermore suggests that G. aparine may represent a food plant. The analysis of potential arable weeds from the material from phase Haz 1 indicates that the most likely candidates are B. secalinus-type, P. maculosa and S. nigrum. The latter species was distinguished as a probable arable weed at Brandwijk-Kerkhof as well (see appendix II). The plants of all three taxa have a relatively large length. The absence of shorter taxa suggests that the cereals were harvested by collecting only the upper halve of the plant, which makes uprooting unlikely. B. secalinus-type is known as a winter crop weed while P. maculosa and S. nigrum are known as summer crop weeds, which does not give a uniform conclusion on the season of crop cultivation. The last two species were already present in the area before the introduction of crop plants and were probably part of the local vegetation at the Hazendonk (Bakels and Van Beurden 2001; Bakels et al. 2001). B. secalinus-type is contrastively not found at other, earlier and later sites in the central river area and is almost only found in the material of phase Haz 1. The ecology of B. secalinus-type is not very specific and allows crop growth at the Hazendonk as well as in other regions, although it is characteristic of soils poor in lime (calcium). The presence of B. secalinus-type is therefore a minor indication of the import of cereals or sowing seed from outside the river area. The taxon is known as an arable weed from the LBK and Rössen culture (Bakels and Rouselle 1985; Knörzer et al. 1999). It is not known from Michelsberg sites but this may be related to the small data set available on the Michelsberg culture (Bakels 2007). III.3.9 wood analySIS Table III.17 shows the unworked wood remains of the Hazendonk (N = 225). Most wood dates to phase VL 1b, followed by phases Haz 1 and Haz 3, while only a small quantity of data is available for phases Haz 2 and VL 1a. The variation of the numbers of finds per phase is related to the quantity of refuse of the phases. The preservation of the wood was good. In the assemblage of unworked wood of phase Haz 1, Alnus sp. dominates, followed by Fraxinus sp. and Quercus sp. A sample collected in square 57 contained wood from Alnus sp., Fraxinus sp., Quercus sp., Cornus sp., Acer sp., Viburnum sp. and Pomoideae (N = 32). The composition of this sample corresponds with the variety of the vegetation at that location and time as indicated by the pollen and macroremains. Other unworked wood remains from phase Haz 1 are identified as Quercus sp.

164

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

N

%

phase Haz 1 Haz 2 Haz 3 VL 1a VL 1b total

Haz 1 Haz 2 Haz 3 VL 1a VL 1b total

taxon Acer sp.

3

-

-

-

1

4

7

-

-

-

1

2

Alnus sp.

13

-

27

3

50

93

29

-

51

x

42

41

Cornus sp.

3

-

3

-

6

12

7

-

6

-

5

5

Corylus sp.

-

-

5

-

2

7

-

-

9

-

2

3

fraxinus sp.

5

1

6

-

17

29

11

x

11

-

14

13

Pomoideae

1

-

2

-

4

7

2

-

4

-

3

3

Prunus sp.

-

-

-

-

2

2

-

-

-

-

2

1

Quercus sp.

6

-

-

-

4

10

13

-

-

-

3

4

Salix sp.

-

-

-

-

6

6

-

-

-

-

5

3

Ulmus sp.

-

1

10

4

6

21

-

x

19

x

5

9

Viburnum sp.

3

-

-

-

-

3

7

-

-

-

-

1

Indet.

11

-

-

-

20

31

24

-

-

-

17

14

total

45

2

53

7

118

225

- = not present x = no meaningful percentages

Copyright © 2010. Leiden University Press. All rights reserved.

Table III.17 The Hazendonk, unworked wood for each phase (Van den Berg unpublished data).

In phase Haz 3, Alnus sp. and Ulmus sp. dominate, followed by Fraxinus sp. and Corylus sp. A sample collected in square 26 contained wood of Alnus sp., Ulmus sp., Corylus sp., Cornus sp., Fraxinus sp. and Pomoideae. The other identified unworked remains from phase Haz 3 were identified as Fraxinus sp. The unworked wood of phase VL 1a consisted of wood of Alnus sp. and Ulmus sp. In the assemblage of phase VL 1b, Alnus sp. dominates, followed by Fraxinus sp. The assemblage of this phase shows maximal diversity, probably related to the large number of identifications. Identified trunks are from Fraxinus sp. (N = 2, mean diameter = 84 mm), Alnus sp. (N = 7, mean diameter = 150 mm) and Quercus sp. (N = 2, mean diameter = 70 mm). The branches and twigs show a larger variety: Alnus sp., Corylus sp., Pomoideae, Fraxinus sp., Ulmus sp., Prunus sp. and Salix sp. for 65 identifications it is not known which part of tree or shrub they represent. Most of these were part of two samples from square 8, unit C, and will probably represent small wood fragments. It was initially expected that the unworked wood would represent the natural vegetation of the site. However, the assemblage does not show a clear development from dry woodland into alder carr vegetation since Alnus sp. dominates the assemblage of several phases. Only Fraxinus sp. shows a minor increase through time. The poor correspondence of the changes in the wood assemblage with the development of the natural vegetation could be explained by the small sample size, selective preservation and/or by the interpretation of the waste of wood working as unworked wood.

165

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

The assemblage of worked wood (see table III.18) includes worked and possibly worked wood with an unknown function (N = 18). This group was distinguished in 2007 based on information in the description about the presence of traces of splitting of the wood (Dutch: splijthout). It is thus assumed that the splitting is the result of human action (cf. Louwe Kooijmans, Hänninen and Vermeeren 2001, 437) although it could also be the result of natural processes. The importance of Quercus sp. in the worked wood compared with the unworked wood possibly indicates the selection of this taxon. Table III.19 shows the wood identifications of artefacts (N = 90). The assemblage of artefacts mainly consists of artefacts from phases VL 1a and 1b, and a small number of artefacts from phases Haz 1 and VL 2b. The material from phase Haz 1 consists of a hammer of Pomoideae (head) and Fraxinus sp. (haft). The artefact of phase VL 1a is a small trackway made of unworked branches of Ulmus sp. and Alnus sp. (4535±35 BP: 3370-3100 BC; Lanting and Van der Plicht 2000). The orientation of the trackway was directed straight away from the dune. The length is unknown since the path was not excavated completely. The unworked wood of phase VL 1a was found in the same unit as the trackway and it concerned the same species. Therefore, it can be questioned whether the unworked wood of this phase represents the natural vegetation or waste deposited during preparation of the trackway. An important group of artefacts dating to phase VL 1b is a series of pointed posts of Alnus sp. (N = 27), forming a palisade (see fig. 14.2 in Out 2008c). The mean diameter of 21 of the posts was 129 mm (range 75-170 mm). The distance between the posts was c. 25 cm (Louwe Kooijmans 1977b). The description of most posts indicates that it concerns split wood, indicating that they were split or possibly that the wood was ripped off. Twenty-one posts were cut (worked) and three were certainly not. The working traces were sometimes very simple, sometimes twofold or fourfold, and sometimes all around a post. Posts other than those from the palisade were made of Alnus sp., Ulmus sp. and Corylus sp. Other artefacts from phase VL 1b are a dugout canoe of Quercus sp. (4400±60 BP: 3340-2900 BC), a paddle of Fraxinus sp., a fragment of bow of Taxus sp. (length of fragment 44 cm), spears and arrow shafts of Viburnum sp., Corylus sp. and Salix sp., a haft of Euonymus sp., a fragment of a haft of Pomoideae and a small tray of Ulmus sp. (the bottom of a container). Several artefacts are drawn in Louwe Kooijmans (1985). The spatial distribution of the unworked wood and the artefacts from phase VL 1b was approximately similar. The material from phase VL 2b consists of a pointed post of Alnus sp., found in a concentration of bone. The choice for specific taxa for the paddle, the canoe and the bow support the selective use of wood based on the quality of the wood and the function of the artefacts, since we know of similar artefacts made of the same taxa from other excavations (see chapter 8). phase

Haz 1

Haz 2

Haz 3

VL 1a

VL 1b

Acer sp.

-

-

-

1?

-

Alnus sp.

-

1

1

-

2

Corylus sp.

1?

-

-

-

-

fraxinus sp.

-

-

1

-

3 + 1?

Quercus sp.

4?

-

-

-

2

-

-

-

-

1

5?

1

2

1?

8 + 1?

Copyright © 2010. Leiden University Press. All rights reserved.

taxon

Ulmus sp. total - = not present

? = possibly worked

Table III.18 The Hazendonk, worked and possibly worked wood remains for each phase (Van den Berg unpublished data).

166

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

The wood of Taxus sp. and Euonymus sp. is only present in the manipulated wood remains, which may indicate the import of wood. The pollen identifications however indicate that the wood of both species may also be of local origin (diagram square 57, Van der Wiel 1982 and the recalculated diagram of Voorrips, see appendix IV). Scarce identifications of wood and charcoal of Euonymus sp. are also available from other sites in the river area (see appendices I and II). Wood fragments that were partly carbonised were identified as Fraxinus sp. (Haz 3: N = 1, VL 1b: N = 1) and Alnus sp. (VL 1b: N = 3). These identifications are included in the table of unworked wood remains. A single fragment of bark was identified as Betula sp. (find number 33.553, phase VL 1b). It may concern a piece collected elsewhere by people for the production of tar or rope since Betula sp. was probably very scarce at the dune. phase

taxon

interpretation

N

VL 2b

Alnus sp.

pointed post

1

VL 1b

Alnus sp.

pointed posts

31

small plank

1

pointed post

1

spear

1

Corylus sp.

arrows Euonymus sp.

haft

1

fraxinus sp.

paddle

1

Pomoideae

head of a haft

1

Quercus sp.

dugout canoe

1

planks

3*

spears

2

arrow

1

Taxus sp.

bow

1

Ulmus sp.

small tray

1

pointed posts

2

Copyright © 2010. Leiden University Press. All rights reserved.

Salix sp.

VL 1a Haz 1

2**

Viburnum sp.

spears

2**

Indet.

plank

1

post

1

Alnus sp.

trackway of unworked branches

9

Ulmus sp.

trackway of unworked branches

24

fraxinus sp.

haft

1

Pomoideae

head of a hammer

1

* = possibly fragments of the canoe

** = possibly two fragments of the same artefact

Table III.19 The Hazendonk, identifications and interpretation of wooden artefacts for each phase (Van den Berg unpublished data).

167

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Wood identifications from a single core by Van der Wiel (1982) concern Alnus sp., Fraxinus excelsior and (Populus cf. nigra/) Salix sp. Fraxinus excelsior was found frequently around phase Haz 2 until phase Haz 3, while Alnus sp. was found frequently from phase Haz 3 onwards. These results correspond with the data of pollen and macroremains presented here, although alder was already present in the local vegetation during phase Haz 1. III.3.10 MoSS analySIS Table III.20 shows identifications of mosses from the Hazendonk, representing Neckera crispa, Homalia trichomanoides, Isothecium myosuroides and Thamnobryum alopecurum. All moss species have their habitat on roots and/or trunks of trees in dark, humid to moist deciduous woodland, on eutrophic or calcareous soil, on single trees in river forelands that get overflowed in winter, on coppices and sometimes on dead wood. The mosses favour various deciduous tree species that are present at the Hazendonk (Touw and Rubers 1989). N. crispa is commonly found at Dutch wetland sites (Kuijper 2000), while Isothecium myosuroides was also found at Bergschenhoek and Swifterbant (see appendix V). phase sample

Haz 1

Haz 3

VL 1b

VL 1b

core 3: section square 57: section square 41: section square 41: 5.28 m -NAP 4.43 m -NAP 3.41 to 3.26 3.31 to 3.26 m -NAP m -NAP

VL 2b M22

taxon Homalia trichomanoides

-

-

-

-

+

Isothecium myosuroides

-

-

-

-

+

Neckera crispa

+

+

+

-

+

Thamnobryum alopecurum

-

+

-

+

-

+ = present

- = not present

Table III.20 The Hazendonk, mosses (Touw, unpublished data).

Copyright © 2010. Leiden University Press. All rights reserved.

III.4 dIscussIon III.4.1 r econStructIon of the natural VegetatIon The extensive sampling program at the Hazendonk has resulted in a large species list, underlining the variety of the vegetation. The natural vegetation on the high-located dry parts of the Hazendonk consisted of meso- to eutrophic deciduous woodland vegetation of dry terrain, dominated by patches of Tilia sp. and Quercus sp. On the slightly lower parts of the slope Tilia sp. was of less importance while Quercus sp., Fraxinus sp., Acer campestre, possibly Ilex aquifolium, Ulmus sp. and Alnus sp. may have been growing close to each other or even mixed. Quercus sp. usually grows on dryer soil than Alnus sp., but the taxa can co-occur when organic sediment is present on top of a mineral subsurface (see also paragraph II.4.2). The vegetation on the slopes was very rich in shrubs, such as Corylus avellana, Cornus sanguinea, Ligustrum vulgare, Viburnum opulus, Rhamnus cathartica, Sambucus nigra and others. The vegetation was possibly comparable with hardwood alluvial woodland, characterised by a flooding frequency varying between less than ten days a year to once in twenty years (Wolf et al. 2001, 129). On the lowest parts of the slopes of the dune that were directly influenced by the groundwater, the natural vegetation consisted of alder carr that was rich in shrub vegetation as well.

168

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

Most taxa of trees and shrubs found in the macroremains assemblage were probably present in the vegetation of the dune. This is, however, less clear for Euonymus europaeus, Malus sylvestris, Prunus padus, Prunus spinosa and Taxus baccata. finds of Euonymus europaeus and Taxus baccata include pollen and worked wood (discussed in the paragraph on wood above). Malus sylvestris was probably present, since there are finds of waterlogged seeds and fruits from various phases. Collection at other locations than at the dune can however not be excluded. The local presence of Prunus padus remains unclear since macroremains of this species were very scarce, found in phase VL 2b only. Macroremains of Prunus spinosa were more common, but were only found during some phases as well (phase Haz 3 and VL 1b). Two identifications of unworked wood tentatively supports presence of Prunus sp. in the local vegetation during phase VL 1b, but this identification is not up to species level and could represent P. padus and/or P. spinosa. The natural vegetation in the surrounding wetlands consisted of marsh vegetation and vegetation of open water. An important part of the wetland vegetation consisted of taxa that indicate eutrophic conditions and the presence of open patches such as Urtica sp., Symphytum sp., Solanum dulcamara, Lycopus europaeus, Eupatorium cannabinum, Filipendula ulmaria and Lythrum salicaria. These taxa could have grown as forb vegetation in well-developed reed vegetation or as the undergrowth of open Alnus vegetation. The diagrams additionally show the minor presence of mesotrophic grassland taxa that probably grew in somewhat mesotrophic reed vegetation: Rinanthus sp., Gentiana sp., Parnassia sp., Pedicularis sp. and Saxifraga granulata. At the lower parts of the slope of the dune, a variety of riparian plants were present, amongst which were Alisma plantagoaquatica, Butomus sp., Sagittaria sp., Sparganium emersum, Sparganium erectum, Typha angustifolia/latifolia, Rorippa amphibia and Oenanthe aquatica. These taxa generally indicate riparian vegetation of moderate to very eutrophic water with a maximal depth of 1.5 metres. Pollen of the riparian taxa Littorella sp. and Elatine sp. additionally indicate clear water of high quality that is poor in nutrients (Weeda et al. 1987, 1988), while Elatine sp. also indicates fluctuating water levels (Brinkkemper et al. 2008). The pollen of Littorella sp. may have been transported from elsewhere, but Elatine sp. was present in the local vegetation, as indicated by finds of seeds of E. triandra (see core 3). The vegetation of the large bodies of open water consisted of a variety of taxa including Nymphaea alba, Nuphar lutea, Trapa natans, Hottonia palustris, Hydrocharis morsus-ranae, Myriophyllum spicatum, Potamogeton sp., Stratiotes aloides, Lemna sp. and Callitriche sp. These taxa indicate moderate to very eutrophic water with a depth between 0.5 and 2 metres. Non-pollen palynomorphs such as Zygnemataceae additionally indicate the presence of shallow pools, probably as part of the alder carr and the riparian zone. The archaeobotanical remains indicate that there was no direct marine influence at the Hazendonk. The only identifications indicative of saline conditions are pollen grains of Plantago coronopus (square 57, after phase Haz 1), pollen grains of Glaux sp. (core 2, after phase VL 1b) and Hystrichosphaeridae (phase VL 1b). The scarcity of identifications of these taxa indicates that the relevant remains have been transported from the coastal zone into the river area during extreme marine influxes into the river area. The freshwater conditions are confirmed by the identifications of freshwater molluscs (Anodonta sp./Unio sp., Bythinia tentaculata and Valvata piscinalis). III.4.2 deVelopMent of the natural VegetatIon The analyses of the cores and sections show that the quantity of surface at the dune covered with dryland vegetation gradually decreased by submerging. Simultaneously, the surface covered with wetland vegetation expanded (see also fig. III.20). The lower part of the section of square 57 shows that Tilia/Quercus woodland was initially present near unit C where occupation took place, while core 3 shows that alder carr was already present several metres further down the slope at that time. The Tilia/Quercus vegetation was quickly replaced by woodland dominated by Quercus sp., Fraxinus sp. and Alnus sp. at the sample points after phase Haz 1, and was gradually replaced by alder carr. The section of square 57 shows an increase in Fraxinus sp. before phase Haz 3 (see also diagram square 41), which corresponds with the wood identifications of Van der Wiel (1982).

169

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

The diagram of square 57 furthermore shows an increase in Ulmus sp. after phase Haz 3, which is unfortunately not registered in other diagrams. Core 2, following the diagram of square 57 in time, shows maximal development of alder carr between phases VL 1a and VL 1b (macroremains). The higher-located section of square 41 shows the initial development of alder carr after phase Haz 2a, the presence of Fraxinus sp. before phase Haz 3 and confirms the maximal development of alder carr between phases Haz 3 and VL 1b (macroremains). The sections M86 and M87 show the presence of oak and alder near the sample locations during phase VL 1b despite the submerging process. Salix vegetation developed at the southeastern side of the dune after phase VL 1b (cf. Vander Wiel 1982). Woodland of dry terrain probably remained present on the top of the dune during all phases. This is indicated by the remains of Corylus sp. and Quercus sp. in the VL 1b horizon (section square 41 and M87), and the presence of the remains of Acer campestre and Quercus sp. including juvenile Quercus remains in macroremains samples dating to phase VL 2b. Tilia sp. nevertheless probably disappeared from the dune after phase Haz 3, since the youngest macroremains of Tilia sp. date to phase Haz 3, and since the curve of Tilia sp. is low and stable after phase Haz 3.

Copyright © 2010. Leiden University Press. All rights reserved.

III.4.3 huMan IMpact III.4.3.1 Indications of human impact in the diagrams Occupation resulted in a decrease in Tilia sp. during all the Hazendonk phases, and Tilia sp. only partially recovered afterwards. The disappearance of Tilia sp. is probably primarily related to the combination of human impact and the rising ground water table. Comparable examples of human impact resulting in a decrease in Tilia sp. are discussed in paragraph 8.3. Quercus sp. also suffered from human impact, but this taxon suffered less from the wet conditions and remained present longer at the dune. People probably cleared trees of Fraxinus sp. during the occupation phases Haz 2 and 3. The macroremains diagrams shows that Alnus trees were reduced during most occupation phases, which is indirectly supported by the important role of Alnus sp. in the wood assemblage, although the pollen diagrams show increased values due to better pollen dispersal. The curve of Pinus sp. shows decreases during occupation phases as well but this does not represent human impact since the taxon was not present at the dune. The decreases in Pinus sp. are interpreted as a statistical result of the increased values of other taxa, which could have been demonstrated when marker spores had been added. Most occupation phases are characterised by an increase in the number and the diversity of dryland shrubs such as Corylus avellana, Viburnum opulus, Rhamnus cathartica, Ligustrum vulgare, Sambucus nigra and Cornus sanguinea. The increase in pollen of these shrubs indicates the development of secondary vegetation in reaction to the increased presence of light and open patches that were created by woodland clearance. The increased presence of shrubs could also represent the presence of hedges around small arable plots (Groenman-van Waateringe 1978) but this remains very hypothetical (see also paragraph 8.8.2). The percentage of Corylus sp. decreased during phase VL 1b in contrast to the earlier phases, while the number of macroremains decreased as well, suggesting that the number of Corylus trees/shrubs strongly decreased through time. Occupation is characterised by a relatively high diversity of herbs and spore plants. Occupation leads to the presence or peaks of pollen of Allium sp., Artemisia sp., Chenopodiaceae, Polygonum persicariatype, Rumex acetosa-type, Polypodium sp. and Pteridium sp. in the dryland vegetation. The climber Humulus sp. and the wetland herbs Urtica sp., Symphytum sp., Solanum dulcamara, Lycopus europaeus, Eupatorium cannabinum, Filipendula ulmaria, Lythrum salicaria, Lotus-type, Ranunculaceae and Menthatype also show increased values during most occupation phases. The combination of these taxa indicates disturbance, tread and the increased availability of light and nutrients. Plantago lanceolata and Plantago major/media are present from phase Haz 1 onwards but they are not strictly related to occupation phases. The peak values of Allium sp. suggest the absence of intensive grazing (Troels-Smith 1954), which appears to be in contradiction with the presence of domestic animals during all phases.

170

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

This could possibly be explained by the interpretation that the Allium pollen does not represent the natural vegetation but waste instead, at least during phase Haz 1, although the expected moment of collection – before flowering – does not correspond with the presence of pollen. The curve of Poaceae (probably mainly representing wetland taxa such as Phragmites australis, Glyceria fluitans and Phalaris arundinacea) often shows initially low values during occupation, followed by a peak at the end of occupation phases. The low values could be related to the use of Phragmites australis by people (cutting and burning) or grazing and destruction by animals, while the peak values at the end of the occupation phases indicate the recovery of the vegetation in open terrain or increased flowering in reaction to anthropogenic influence. The Cyperaceae and monoletae psilatae fern spores show a fall during most occupation phases. This could be related to human influence or grazing as well. The pollen and macroremains diagrams show the initial recovery of the vegetation after the main occupation phases (Haz 1, Haz 3 and VL 1b), characterised by high values of certain anthropogenic indicators, which seems to represent the optimal growth of herbs and ferns indicative of human impact that benefit from open, disturbed terrain and the absence of human activity, followed by the true recovery of the natural vegetation that is characterised by moderate values in the diagrams and that is related to the final complete absence of the effects of human impact on the vegetation. Taxa showing high values characteristic of such initial recovery of the vegetation are for example Chenopodiaceae, Artemisia sp., Urtica sp., Lythrum sp., Solanum dulcamara, Sparganium erectum, Apiaceae and Rubiaceae, although the precise list of taxa varies per phase. After phase Haz 1 this concerns both dryland and wetland taxa (diagram square 57), after phase Haz 3 it already concerns more wetland taxa and after phase VL 1b it mainly concerns wetland taxa. This shift of taxa is related to the increase in the water table and the changes in the natural vegetation at the sample points. for the changes after phase VL 1b it is not precisely clear in which way the development of the vegetation is influenced by changing water conditions (as indicated by the presence of Salix sp.) and the recovery from human impact. Summarising, anthropogenic influence on the vegetation resulted in the clearance of both dryland and wetland trees, the presence of open patches, the increased presence of nutrients because of the deposition of waste, and the presence of taxa that are characteristic of disturbance and human activity. It usually took some time for the vegetation to recover after occupation, which is visible after phases Haz 1, Haz 3 and VL 1b. Some taxa were not able to recover because of the repeated occupation pressure, changing environmental conditions and competition. The extensive analysis of non-pollen palynomorphs allowed the investigation of the relationship between NPP’s and occupation. Most taxa that increase during occupation, which makes them potential anthropogenic indicators, are not known as typical anthropogenic indicators. The possible anthropogenic indicators do not include types indicative of the presence of dung except for type 169 that possibly indicates the presence of dung.36 fifty percent of the 28 taxa that show increased values during occupation increase during a single phase only. for these taxa it is therefore not possible to conclude whether they are indeed indicative of human impact, or that the increase in these taxa accidentally co-occurred with occupation. Interestingly, there are four types that show a more consequent reaction to occupation. Type 361, psilate Spirogyra sp., type 44 (Ustulina deusta), type 121 and type 373 (Mougeotia laetevirens) show increased values during at least three different phases, and appear to be relatively good indicators for human impact at the Hazendonk. Type 361 was already known as an anthropogenic indicator (Van der Wiel 1982). Type 44, also known as Kretzschmaria deusta, is a parasite of deciduous trees, and it can be assumed that it benefited from clearance activities by men, resulting in the increased susceptibility of trees for parasites (cf. Innes et al. 2006 that demonstrate an increase in type 44 contemporaneous with a presumably anthropogenic decrease in Ulmus sp. and Tilia sp. in North Wales).

36 Other NPP’s indicative of dung may have been present in the case that they remained unidentified.

171

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

The precise relationship to occupation with other types is however more difficult to explain; perhaps an increase in eutrophic conditions during occupation plays a role. Spirogyra sp. and Mougeotia laetevirens indicate shallow, stagnant, mesotrophic to eutrophic water, while it has been suggested for type 121 that it shows optimal values during the presence of lakes (Pals et al. 1980). The results of Van der Wiel (1982) tentatively support the results on the use of NPP’s as indicators for human impact, since her diagram shows maxima of type 361 and Spirogyra sp. during occupation as well. Comparison of changes in the NPP curves between the various cores during separate phases shows that human impact led to different responses in NPP curves in different cores. Changes during phases Haz 0 and 1 in the section of square 57 and core 3 are not similar. The correspondence in square 57 and core 2 during phase VL 1a consists of an increase in type 361 only, while in both cores several other types additionally increase. Most types that increase during two occupation phases were furthermore found in a single core only. This result indicates that the occurrence of taxa represented by the types is strongly influenced by differences between the cores, i.e. environmental conditions (type of subsurface, water depth, natural vegetation, presence of hosts, etc.) and the strength of human impact. This sensibility makes it very difficult to predict which NPP’s to expect during the analysis of occupation phases at comparable archaeological sites. The absence of these species furthermore does not indicate the absence of occupation. The presence of the four taxa that are the best anthropogenic indicators in this study do not necessarily indicate occupation either, since they also occur under natural conditions in between occupation phases when people were probably not present at all. The main changes in the pollen diagrams related to human impact at the Hazendonk as discussed above have been summarised in a model diagram (see fig. III.26). The model shows two phases, one representative of the Hazendonk phases and one representative of the Vlaardingen phases (mainly based on phase VL 1b). future analyses can investigate whether this model can be applied to other sites as well (though see the note in paragraph 2.8.3.1). The comparison of data with this model is only possible when the environmental conditions and methodological aspects are comparable.

ps e ea ac Po

Fi li Ly pen t d A p hr u ula ia m ce a Br as e Ty si p c S p e 3 eae yr 61 og Ph i r a sp as ., e

e ea ac er yp C

Vi b C ur n he u A r no m t p A l e m i ac liu i si e a C m a e e A l rea nu lia s -t y pe

C

or

yl

us

s us

nu

lm

xi

U

Fr a

ue Q

lia

Ti

Pi

nu

s

rc

us

ila

ta

e

Upland taxa

Copyright © 2010. Leiden University Press. All rights reserved.

VL

Haz

20

20

20

20

20

20 40

20 40 60 80 100

20 40

20 40

20

Figure III.26 The Hazendonk, model pollen diagram of changes related to human impact.

172

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

III.4.3.2 Further interpretation of indications of human impact All occupation phases that were recognised as a horizon in the sediment during excavation of the squares near the pollen sample locations could be detected in the pollen diagrams, since the changes occurred contemporaneously in the curves of many taxa, since the same changes occurred repeatedly, since the vegetation recovered (partly) afterwards and since changes in the pollen diagrams could be linked to the presence of fine charcoal particles. The possibility to recognise occupation resulted in the confirmation of the existence of phase Haz 0 (cf. Louwe Kooijmans 1974) and in indications of two extra occupation periods that were not recognised in the field or during previous pollen analysis, probably related to phase VL 1a (sample boxes of unit C, square 57, after phase Haz 3). This distinction of occupation phases demonstrates the value of the archaeological application of pollen analysis. All phases that were recognised in the sediment can be distinguished as continuous occupation phases in the diagrams. This means that either all distinguished phases represent continuous occupation, or that subphases do not come to expression in pollen, NPP and macroremains diagrams. The analysis of the pollen and macroremains diagrams of Brandwijk-Kerkhof indeed showed that a two-phased occupation period recognised by micromorphological analysis resulted in a single signal of human impact (see Out 2008a). The continuous evidence of human impact during phases Haz 1, Haz 3 and VL 1b thus does not guarantee continuous occupation during these phases. One of the research questions was how occupation intensity and distance influence the evidence of human impact. figure III.27, representing the summary pollen diagrams of the sample series from unit C (based on an upland pollen sum), gives a summary of human impact on the dryland vegetation. The summary diagrams show that occupation phases Haz 1, Haz 3 and VL 1b give a rather similar, relatively strong signal, while phases Haz 0, Haz 2a, Haz 2b and VL 1a result in a relatively weak signal. The signal of phase Haz 1 seems somewhat stronger than the signal of phase Haz 3 and VL 1b. The signal of phase Haz 1 may be overrepresented (discussed in paragraph III.3.1), or may alternatively be indicative of local arable farming (see paragraph III.4.5). Comparison with the distribution and thickness of the occupation horizons (refuse layers) indicates that the strength of evidence of human impact in the diagrams can be related to the extent of the refuse layers (see fig. III.3). It is assumed that the extent of the refuse layers corresponds with occupation pressure, which can be defined as the combined effect of the number of people, the duration of visits and possibly the number of visits within single phases. Occupation and waste deposition took place near the pollen sample locations during all phases, resulting in a signal of all phases. The similarity in strength during the three main phases can be explained by a similar close distance between human activity and the sampling locations during various phases37, similar occupation intensity, as well as by the similarity of the activities during all phases, related to the continuity of site function. The development of human impact through time and the fact that the strength of human impact is related to the extent of the refuse demonstrates that human impact did not increase with gradually ongoing neolithisation (cf. Out 2008c). If neolithisation was a major factor influencing the strength of human impact, one would expect that human impact would have increased through time, and especially during the Late Neolithic. The double registration of phases Haz 1 and VL 1b enables the investigation of the effect of increasing distance between the sampling points and the anthropogenic influence on the dryland vegetation within a single period. Phase Haz 1 is represented in the diagrams of square 57 and core 3, located at 3.5 metres distance from each other. In the material of square 57, the signal of anthropogenic influence on the dryland vegetation is stronger than in core 3 (see fig. III.27). The distribution of refuse indicates that human activities 37 An exception to this is phase Haz 0. The main concentration of refuse of phase Haz 0 was located relatively far away from the sample locations, which may possibly partly explain the limited strength of the signal of this phase. The little quantity of refuse, indicative of little intensive occupation, is however considered as a more important factor explaining the limited human impact during this phase.

173

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

) AP -N ees (cm d tr pth lan De Up

core 2

nd bs rbs a ru e h ds dh e lan p lan ha s Up U P

nts pla e or sp

300 310 320 330

VL 1b

340 350 360 370 380

VL 1a

390 20 40 60 80 100

square 57 385 395 405 415 425 Haz 3

435 445

core 3

) AP -N ees m r (c d t pth lan De Up 525

455 465 475

nts pla e r po

s nd bs rbs a r hu e ds dh e lan p lan has p U U P

485 495

Haz 1

505 525

Copyright © 2010. Leiden University Press. All rights reserved.

Haz 1

545

515 535

535

555 20 40 60 80 100 top dune

565 575 585

Haz 0

595 605 615

20 40 60 80 100 foot dune

174

Figure III.27 The Hazendonk, summary diagrams of the section of square 57, core 2 and core 3.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

during both phases were concentrated on the higher parts of the slope of the dune at unit C or above unit C, which explains the relatively strong signal in the section of square 57. Phase VL 1b is represented in core 2 (unit C), M86 and M87 (unit B), and in the core of Van der Wiel (1982) (see fig. III.2 for the sample locations). Refuse was concentrated in unit C. Analysis of the summary diagrams shows that the evidence of human impact on the dryland vegetation of M87 is weaker than the signal of M86, and that the signal is even weaker in the core of Van der Wiel (1982), but there is no clear difference between the evidence of human impact between core 2 and M86 (see fig. III.28). These results can be explained by two factors. In the first place, the distance between the sample point and the top of the dune influences the strength of the signal of anthropogenic influence on the dryland vegetation, since activity was concentrated on the higher parts of the slope. This effect explains the decrease in the signal of human impact when comparing M86, M87 and the core of Van der Wiel, and corresponds with the comparison of signals of phase Haz 1. Secondly, core 2 is located further away from the top of the dune than M86, but closer to unit C where activity was concentrated, resulting in a signal of human impact in core 2 that is similar to the signal of M86. unit C

unit B

sp or e U pl a U nd pl s an hr d ub he s rb Ph s an as d e

355

U

De pt h

(c m pl an -N d tre AP ) es

pl an ts

M86

378

369

380

371

382

Copyright © 2010. Leiden University Press. All rights reserved.

pl an ts

300 310 320

373 375

sp or e

sp or e

367

Up l Up and la sh nd r he ubs rb Ph s as an e d

VL 1b

365

De pt h Up (cm la -N nd AP tre es )

363

core 2 Up la n Up d s la hr nd ub he s rb Ph s an as d e

M87

361

De pt h Up (cm la nd -NA tre P) es

359

pl an ts

357

384 20 40 60 80 100

386

top dune

388

330 VL 1b

350

390

360

392

370

394 396

VL 1b

340

380 20 40 60 80 100

390

VL 1a 20 40 60 80 100

foot dune

Figure III.28 The Hazendonk, summary diagrams of M86, M87 and core 2.

175

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

The decrease in the signal of disturbance of the dryland vegetation over a small distance of several metres during both phase Haz 1 and VL 1b in the first place indicates that the sample locations both represent very small woodland hollows (cf. Sugita 1994). The pollen rain from woodland hollows does not spread over large distances due to the presence of dense vegetation, and the signal of human impact that was created on the higher parts of the slope of the Hazendonk did therefore not spread over a large distance either. Secondly, the decrease in the signal over a small distance indicates that human impact was only of moderate strength during the main occupation phases, and thus even smaller during minor occupation phases. The data indeed indicate that human impact did not result in complete deforestation of the dune. Instead much of the woodland present on top of and around the dune remained relatively undisturbed. The precise degree of deforestation is however hard to reconstruct since the investigated cores and sections were all present at the southeastern side of the dune, and since local vegetation may have prevented the precipitation of pollen of the vegetation of the remaining parts of the dune at the sample locations. The development of particularly the western side of the dune is therefore poorly known. The sampling of pollen in sediment that contains refuse layers may have resulted in overrepresentation of the dryland vegetation and human impact, since the refuse layers might contain eroded material from the top of the dune. The refuse could moreover have contained waste plant material as well. Colluviation processes might additionally have disturbed the stratigraphy. Comparative analysis of the available pollen and macroremains diagrams however indicates that it is possible to reconstruct the development of the vegetation and to reconstruct the relative strength of human impact at the various sample locations and for separate occupation periods, while there are only little indications of overrepresentation of taxa due to concentrations of plant waste. An exception is the peak of cereal pollen in phase Haz 1 that is probably related to the presence of cereal grains close to the analysed sections and cores. The distribution of charcoal as well as the curves of some NPP’s furthermore suggest the presence of human activity during periods other than occupation phases, but these are not confirmed by changes in the diagrams, which indicates that they must be explained by other processes than human impact. The similarity of the presented results with the results of Van der Wiel (1982) and Voorrips (Louwe Kooijmans 1974), sampled at the same distance and further away from the dune respectively, further confirms the validity of sampling in refuse layers. An important advantage of sampling in the refuse layers at close distance to human activity is the possibility to detect small, local changes in the vegetation. III.4.4 plant SubSIStence The macroremains indicate that the plant subsistence at the Hazendonk was based on both cultivated and gathered plants. This corresponds with the interpretation of the site-function as a semi-agrarian site with an extended broad-spectrum economy. The identified cultivated plants, emmer wheat and naked barley, were best represented in the material of phase Haz 1 with both grains, chaff remains and rachis fragments. The cereal remains suggest harvesting by uprooting or with sickles, although alternative methods cannot be excluded completely. The weeds reject uprooting as the harvesting method, although crop processing may have biased this result. The best indications of consumption of wild plants come from the carbonised finds of edible taxa. The spectrum of gathered seeds and fruits that probably functioned as food at the Hazendonk is large, though comparable with the spectrum found at other Early and Middle Neolithic Dutch wetland sites. Taxa that were probably consumed are Quercus sp., C. avellana, C. monogyna, P. spinosa, M. sylvestris, C. sanguinea, R. fruticosus, R. ficaria, T. natans, N. alba and possibly Galium aparine, while several other taxa may additionally have been consumed as well. The data show a slight decrease in carbonised food plants during the Late Neolithic (from the Vlaardingen phases onwards), which can be explained in various ways. The macroremains do hardly provide any clear evidence of the use of plants for other purposes than consumption. Only a concentration of macroremains in a pit dating to phase VL 2b supports the use of

176

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

plant remains, but the purpose of these remains is not directly clear. The carbonised macroremains indicate occupation between summer and winter for all phases, which corresponds with the zoological results on seasonality. The wooden artefacts indicate that Alnus glutinosa was used on a large scale, which is probably related to its abundant availability in the natural vegetation. There is also evidence of the selective use of wood for specific artefacts because of the qualities of the wood and the function of the artefacts, which corresponds with information from other contemporaneous wetland sites. There are furthermore minor indications of the selective use of Quercus sp. for wood working. Most of the wood that people used was probably collected at the Hazendonk or in the exploitation area. III.4.5 local cultIVatIon Local arable farming at the Hazendonk has been the subject of debate from excavation onwards. In the years after excavation, indications of human impact resulted in the conclusion that cereal cultivation was practised (Louwe Kooijmans 1976, 255). The finds of rachis fragments and cereal pollen were considered to be indicative of small-scale local cultivation (Bakels 1981, 145). The interpretation shifted in the mid-eighties: “We strongly doubt on local crop cultivation. The dune was very small and the environment was not suitable at all” (Louwe Kooijmans 1985, 125, translated) and “crop cultivation played a minor role or did not take place at all” (Bakels 1986, translated). The most recently published conclusion is that local crop cultivation did probably not take place at the Hazendonk (Bakels and Zeiler 2005, 326; Van Gijn and Louwe Kooijmans 2005, 210-211). This paragraph aims to give an updated overview of the arguments on local crop cultivation at the Hazendonk. A first important aspect of the discussion on local crop cultivation at the Hazendonk is our western, modern-day view on arable farming in view of the environmental conditions and the scale of cultivation. It is however argued in chapter 11 that these viewpoints should not result in a conclusion on local cultivation a priori. The best evidence of local crop cultivation dates to phase Haz 1, but most finds can be explained by import as well, while there is also an important argument against local cultivation. firstly, the finds of macroremains of phase Haz 1 that suggest the presence of ears of both crops can be seen as an indication of local crop cultivation. Especially chaff remains of naked barley would only be expected in regions where naked barley is cultivated, since it is expected that chaff remains of this free-threshing cereal would be removed before transport in order to reduce the product volume and weight. This argument in favour of local crop cultivation can however be rejected by the hypothesis that both hulled emmer wheat and free-threshing barley were transported in the ear to Dutch wetland sites (Bakels 2000). The transportation of naked barley in the ear may have been possible when the barley was harvested in dead-ripe state, with (part of) the stems still attached to the ear (Cappers 2006; Cappers and Raemaekers 2008). The delay of threshing would then result in optimal ripening of the barley, and possibly also in the protection of the grains. furthermore, the transportation of naked barley in the chaff could have an economic reason after all when people used the chaff and possibly the stems (Cappers and Raemaekers 2008), or could be explained by the scarcity of working forces at the production location (fuller et al. in press). Secondly, the presence of the pollen of cereals could also be seen as an argument in favour of local crop cultivation, since the observed curves in the pollen diagrams are as expected in case of local crop cultivation. The presence of pollen can however also be explained by threshing activities after the import of cereals from elsewhere instead of local cultivation, since most pollen is produced during the threshing and processing of cereals instead of during flowering (Bottema 1992; Robinson and Hubbard 1977; Zohary and Hopf 2000, see also chapter 11). The presence of the pollen grains of cereals at the Hazendonk is indeed correlated with the presence of waste, the size of the refuse layers and human impact on the vegetation, and can be interpreted as a waste product instead of an indication of arable farming.

177

Copyright © 2010. Leiden University Press. All rights reserved.

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

A third argument in favour of local cultivation is that the material consisted of rich concentrations of cereals that have never been found at comparable Dutch sites. One could argue that these concentrations and quantities indicate that it concerns a producer site. The attested number of cereals is however not large enough to exclude the import of cereals, and there are no indications of the storage of surplus production. Comparable quantities of cereals may also have been present at other sites, and the unique preservation of the cereal concentrations from phase Haz 1 indeed indicate that deposition processes and preservation may explain this exceptional find. Botanical find assemblages of indisputable producer sites from the Swifterbant culture and Hazendonk group are not known and it is therefore not possible to compare the assemblage of the Hazendonk with material of such a site. A fourth argument is that the underdevelopment of the emmer wheat found at the Hazendonk could be explained by the suboptimal environmental conditions at the Hazendonk. Underdevelopment could however have had a variety of causes that could have occurred everywhere. A fifth argument that supports local crop cultivation during phase Haz 1 is that the pollen diagrams provide relative strong indications of deforestation in the pollen diagrams during this phase, especially expressed in the dryland herb vegetation. The strength of the signal of human impact corresponds with the relatively large surface of the dune, which would have made the dune most suitable for local crop cultivation during phase Haz 1. The indications of deforestation during phase Haz 1 may however possibly be overrepresented since the dryland herb vegetation is dominated by Allium sp., which may represent a shade tolerating Allium species. An important argument against local crop cultivation during phase Haz 1 is the presence of large quantities of B. secalinus-type. Although it could have grown at the Hazendonk, the absence of this species at older sites in the central river area and its almost complete absence (except for half a grain) during later phases at the Hazendonk indicate that this taxon did not grow at the Hazendonk but was imported from elsewhere. Large numbers of Bromus sp. are moreover not reported for any of the other Early and Middle Neolithic Dutch wetland sites. The material from the phases later than Haz 1 give much weaker indications of local crop cultivation since indications of human impact in the pollen diagrams are smaller, since the quantities of both grains and chaff remains are smaller, since there is no comparable evidence of the presence of complete ears, and since there are less chaff remains of naked barley. The difference can be seen as an argument that the samples from phase Haz 1 indicate local crop cultivation and the samples from later phases in contrast do not. The difference can alternatively be explained by the differences between the type of finds (concentrations are available from phase Haz 1 only), reflecting differential depositional processes. The archaeological information sources, such as site function, features indicative of tillage, querns and sickles do hardly provide more information than the botanical remains. The analysis of seasonality and site function do not give indications to exclude local cultivation and theoretically allow local cultivation, since the site was visited during various seasons during most phases and since the site function may have been a base camp (although this is not precisely clear). The excavation did not reveal features that are characteristic of soil tillage. The presence of fields however cannot be excluded since the top of the dune was bioturbated and eroded. There are some finds of quern stones at the Hazendonk, but these only indicate consumption and do not demonstrate local crop cultivation. Stones identified as querns that could be assigned to a single occupation phase by stratigraphy are known from phases Haz 3, VL 1a and VL 1b (N = 9). It is not known whether querns were absent during phases Haz 1 and 2 or not, since a group of 14 stones could not be assigned to a single occupation phase (pers. comm. Wansleeben 2007). The information on querns at the Hazendonk is based on the analysis of archive information only. The stone assemblage should be investigated again by modern standards (use-wear analysis and residue analysis) in order to obtain final interpretations.

178

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Analysis of all flint finds from the Hazendonk except for those from unit C has not resulted in the positive identification of flint artefacts with use-wear with gloss in a longitudinal direction indicative of cereal cutting, i.e. sickles that are as convincing as the evidence from the Middle Neolithic sites Schipluiden and ypenburg in the coastal area (Van Gijn et al. 2006, 154; pers. comm. Van Gijn, 2007). There are however some pieces of flint that show use-wear traces that could possibly represent cereal working (dating to phases Haz 2 and 3)38, and as a result the use-wear analysis does not exclude local crop cultivation either. The absence of flint artefacts that convincingly show cereal gloss is furthermore not necessarily indicative of the absence of local cultivation since the crop may have been harvested in another way. Alternatively, sickles may have been deposited at off-site locations. The assumption of harvesting without sickles is however in contrast to the tentative indications of sickle harvesting from the cereal remains. In conclusion, the available evidence does not uniformly support or reject crop cultivation at the Hazendonk. The presence of Bromus sp. in the material of phase Haz 1 indicates import during this phase, implying the transportation of emmer and naked barley in the ear. Local cultivation during phase Haz 3 is supported but not demonstrated by the finds of possible sickles. The indications of local crop cultivation during the Vlaardingen phases are less strong than the indications from the earlier phases.

Copyright © 2010. Leiden University Press. All rights reserved.

III.5 acknowledgeMents I would like to thank A. Louwe Kooijmans-Bouhuijs, W.J. Kuijper, L.P. Louwe Kooijmans, L.B.M. Verhart and C.C. Bakels for putting available the results and for discussion, A. Louwe Kooijmans-Bouhuijs and W.J. Kuijper for analysis, J.P. Pals, C.C. Bakels, R.T.J. Cappers, B. van Geel and A. Aptroot for advice on identifications, A.J. Kalis for help with the nomenclature of pollen identifications, A.L. van Gijn for analysis of use-wear analysis of the flint, M. Wansleeben for information on querns, P.C. Anderson for discussion on harvesting methods and J. van der Plicht for information on the 14C dates.

38 The flint artefacts 24.243 (Haz 2), 37.686, 37.622 and 37.202 (three fragments of a single artefact, Haz 3), 17.715 (Haz 3) and 29.402 (Haz 3) show use-wear traces that could possibly represent relevant cereal working.

179

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample unit square mesh width volume sediment context taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica Table III.11 part 1a.

180

39.154 C 57 1.5 0.25 l peat refuse

39.184 C 57 0.25 1l peat refuse

39.185 C 57 0.25 1.5 l peat refuse

1c -

1c +

++

Haz 1 M157 M160 C C 39 57 1.5 1.5 10 l 0.25 l peat ? refuse pit

2c 1c 2c 6c 16, 9c 1 4c 2 1 3

2c 7, 6 c 1c 1c

M163 C 57 1.5? 0.25 l peat refuse

M164 C 57 1.5 0.25 l peat refuse

M167 C 57 ? ? peat refuse

-

2c 5, 4 c 2c 1c -

2, 1 c 45, 40 c 11, 2 c 7c 1c -

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

39.154

39.184

39.185

Haz 1 M157 M160

48 c 1c 4.5 c -

0.5 c 2c 1c 6c 39 c -

68 c 1c 0.5 c 1c 1c 33 c 13 c -

12 c 3, 1c 1 5c -

3c -

149 c 2c 9c -

5c 1c 2c -

1c -

Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum

+c

2c

-

200 c

56 c

+c

29 c

Hordeum vulgare, internodia

-

-

4c

37 c

-

-

58c 881 c 22 c -

+++ c +++ c +++++ c -

515 c 320 c 3c -

34 c 10c 96 c 5c -

*+ c *+ c *+ c -

394 c 678 c nudum 710 c 448 c -

3c -

4c -

2 1c

1c

-

1c -

1

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma

Hordeum vulgare, chaff remains Triticum dicoccon ++++ c Triticum dicoccon, spikelet forks ++++ c Triticum dicoccon, rachis segments 125 c Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases Group 4 Alnus glutinosa Alnus glutinosa, cones

-

M163

M164

M167

16 c 35 c -

Table III.11 part 1b.

181

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type Table III.11 part 1c.

182

Haz 1 M157 M160

39.154

39.184

39.185

M163

M164

M167

-

-

-

-

-

-

-

-

-

1 c cf. -

-

-

-

-

-

-

-

+ -

1c ++ -

1c 11 5 1c 1 -

3c -

1c cf. -

1c -

-

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

Haz 1 M157 M160

39.154

39.184

39.185

M163

M164

M167

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

-

-

-

1 -

1c

-

-

-

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

-

-

3c -

4c 1 -

3, 2 c -

16 c -

-

-

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

-

1c 1c -

-

1c 1c

-

1c -

-

-

Table III.11 part 1d.

183

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 9 Sand Buds Bud scales Leaf remains Moss remains fragments of catkins indet. Charcoal Bone/fish remains * Pottery remains flint fragments Bark remains Insect remains faecal pellets Lophopus cristallinus, statoblasts Daphnia sp., ephippia Cocoons Cenococcum geophilum, sclerotia Trichoptera, cases of larvae cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula Anodonta sp./Unio sp., fragment tissue

Haz 1 M157 M160

39.154

39.184

39.185

+ + -

+ + + + +++ c -

+ + + + +++ c -

8, 8 c + + 12, 11 c +++ c -

-

-

-

-

M163

M164

M167

++ -

1c + -

+ + + -

+ -

-

-

-

-

Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy) l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 1e.

184

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase Haz 1/ 2 sample ? M156 unit C C square 39 39 mesh width 0.25 5? volume 1l ? sediment peaty sand peaty sand context colluvium colluvium taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica

1 130

3c 1 -

Haz 2 Haz 2/3 M150 M133 M25 C D A 53 9 3 1.5 / 5 ? 5 2l ? ? peat peaty sand peat refuse colluvium refuse

1 1c 8, 4 c 2c 2 11 c 1c 2c -

1c 8 1 1c 1 -

1 8 4 3 1 3 2 4 77 7 1c 15

Haz 3 M74B B 26 5 ? peat refuse

M75A B 26 5 ? peat refuse

16, 2 c 5 10, 9 j 4 16, 7 c 2, 1 c 5, 4 c 24, 14 c, 1 j 1c 87, 4 c 83 65, 1 c 11, 2 c 2 22, 13 c 1 1 3c 15

1 3 1 4, 1 c 5 -

Table III.11 part 2a.

185

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

Haz 1/ 2 ? M156

Haz 2 M150

Haz 2/3 M133

M25

Haz 3 M74B

M75A

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma

4 1c 2c 1 -

-

1 c cf. 6 6c 2 -

1 1 -

1 1 -

2c 1 2 1c -

-

Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum Hordeum vulgare, internodia Hordeum vulgare, chaff remains Triticum dicoccon Triticum dicoccon, spikelet forks Triticum dicoccon, rachis segments Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases

8c -

-

3c -

1c -

-

16 c 67 c 4c -

-

Group 4 Alnus glutinosa Alnus glutinosa, cones

1c -

-

2 3

-

29 4

52 49

1

Table III.11 part 2b.

186

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type

Haz 1/ 2 ? M156

Haz 2 M150

Haz 2/3 M133

M25

Haz 3 M74B

M75A

-

-

-

-

5 -

10 -

-

1 -

-

1 -

-

1 1 4 1

3 1 4

-

2 1 -

-

22 7 3, 1 c 1 -

-

1 1 1c 5 4 1 1 1 -

2 1 1c 4 3 1 6 2 1 -

-

Table III.11 part 2c.

187

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

Haz 1/ 2 ? M156

Haz 2 M150

Haz 2/3 M133

M25

Haz 3 M74B

M75A

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

-

-

-

1 -

-

1 12 -

-

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

-

-

1 -

-

1 1c

2 cf. c 1c

-

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

-

-

1 -

-

1 -

1 1 cf. 1 4 1c -

-

Table III.11 part 2d.

188

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 9 Sand Buds Bud scales Leaf remains Moss remains fragments of catkins indet. Charcoal Bone/fish remains * Pottery remains flint fragments Bark remains Insect remains faecal pellets Lophopus cristallinus, statoblasts Daphnia sp., ephippia Cocoons Cenococcum geophilum, sclerotia Trichoptera, cases of larvae cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula Anodonta sp./ Unio sp., fragment tissue

Haz 1/ 2 ? M156

Haz 2 M150

Haz 2/3 M133

M25

Haz 3 M74B

M75A

+++++ + ++++ c -

-

++ ++ ++ 1 -

-

18 + 5 ++ ++ + ++ ++ -

180, 1 c -

3 -

-

-

-

-

-

-

-

Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy) l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 2e.

189

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample unit square mesh width volume sediment context taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica Table III.11 part 3a.

190

M75C B 26 5 ? peat refuse

M77B2 B 26 5 ? peat refuse

4 4c 2c 7, 4 c 1 11, 1 c -

1 4j 1 3c 2c 11, 8 c 1 54 1c -

Haz 3 (cont.) M84 M88 B B 26 26 0.25 0.25 ? ? peat sand refuse colluvium

6

1c 1 3 13 1 1 9

M154 C 39 5 ? peat refuse

? C 39 ? ? peat refuse

2, 1 c 12, 2 c -

1 -

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum Hordeum vulgare, internodia Hordeum vulgare, chaff remains Triticum dicoccon Triticum dicoccon, spikelet forks Triticum dicoccon, rachis segments Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases Group 4 Alnus glutinosa Alnus glutinosa, cones

Haz 3 (cont.) M84 M88

M75C

M77B2

M154

?

-

-

1 -

-

-

-

7c -

7c 2c -

-

-

1c -

-

4

8

-

-

32

-

Table III.11 part 3b.

191

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type Table III.11 part 3c.

192

Haz 3 (cont.) M84 M88

M75C

M77B2

M154

?

-

-

-

-

-

-

-

-

-

1 -

-

-

-

1 -

-

1 1 -

-

-

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

Haz 3 (cont.) M84 M88

M75C

M77B2

M154

?

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

-

-

-

-

-

-

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

-

-

-

-

-

-

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

-

3

-

1

-

-

Table III.11 part 3d.

193

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M75C

M77B2

Haz 3 (cont.) M84 M88

Group 9 Sand ++++ Buds 3 50 Bud scales Leaf remains Moss remains fragments of catkins indet. Charcoal + Bone/fish remains * ++ + Pottery remains flint fragments Bark remains Insect remains faecal pellets Lophopus cristallinus, statoblasts Daphnia sp., ephippia Cocoons Cenococcum geophilum, sclerotia +++ Trichoptera, cases of larvae cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula 1 Anodonta sp./Unio sp., fragment tissue Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy), l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile

M154

?

-

1 -

-

-

Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 3e.

194

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample unit square mesh width volume sediment

Copyright © 2010. Leiden University Press. All rights reserved.

context taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica

M13a B 5 1.5 ? peaty sand

M73A B 26 1.5 ? peaty sand

Haz 3/VL 1b M76A2 M110 M112 B C C 26 36 35 1.5 1.5 5 ? ? ? peaty peaty peaty sand sand sand all colluvium

1 cf. 14 17 36 1 44

2 1j 2, 1 c 4, 2 c 6 16 25 77 1c 1 17

3 2j 5c 1c 19, 6 c 4c 29, 3 c -

1c 1c -

1c -

M102 D 31 5 ? peaty sand

M113 D 31 5 ? peaty sand

1c -

2c 1c -

Table III.11 part 4a.

195

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

Haz 3/VL 1b M76A2 M110 M112

M13a

M73A

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma

3 -

2 1 -

-

-

Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum Hordeum vulgare, internodia Hordeum vulgare, chaff remains Triticum dicoccon Triticum dicoccon, spikelet forks Triticum dicoccon, rachis segments Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases

-

1c -

8c -

Group 4 Alnus glutinosa Alnus glutinosa, cones

-

4 -

2

Table III.11 part 4b.

196

M102

M113

-

-

-

-

-

-

-

1

-

-

-

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type

Haz 3/VL 1b M76A2 M110 M112

M13a

M73A

M102

M113

-

-

-

-

-

-

-

-

2 1 5 1

-

-

-

-

-

3 -

1 1 6 1 -

-

-

-

-

-

Table III.11 part 4c.

197

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

M13a

M73A

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

1 -

4 -

-

-

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

-

-

-

1 3 2, 1 c cf. 1 -

-

-

.

Copyright © 2010. Leiden University Press. All rights reserved.

Haz 3/VL 1b M76A2 M110 M112

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

Table III.11 part 4d.

198

M102

M113

-

-

-

-

-

-

-

-

-

-

-

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 9 Sand Buds Bud scales Leaf remains Moss remains fragments of catkins indet. Charcoal Bone/fish remains * Pottery remains flint fragments Bark remains Insect remains faecal pellets Lophopus cristallinus, statoblasts Daphnia sp., ephippia Cocoons Cenococcum geophilum, sclerotia Trichoptera, cases of larvae cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula Anodonta sp./Unio sp., fragment tissue

Haz 3/VL 1b M76A2 M110 M112

M13a

M73A

11 + 3 + ++++ -

12 + + + + + ++ ++ + + -

20 -

-

-

-

-

-

M102

M113

-

-

-

-

-

-

Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy) l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 4e.

199

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample unit square mesh width volume sediment context taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica Table III.11 part 5a.

200

VL 1b M49 M100 M116-117 M118 M187 B C C C C 25 36 8 8 42 1.5 5 1.5? 1.5? 5 ? ? ? ? ? peat peat peat peat peat refuse refuse refuse refuse refuse

20 17 3, 2 j 1 4 3 7

1c -

1 1 1 ++, ++ cf. 3, 1 c ++++

1c 1 4c ++

1 -

M101 D 31 1.5 ? peat refuse

M103 D 9 5 ? peat refuse

M108 D 9 5 ? peat refuse

1c -

1c 1c -

3c -

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum Hordeum vulgare, internodia Hordeum vulgare, chaff remains Triticum dicoccon Triticum dicoccon, spikelet forks Triticum dicoccon, rachis segments Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases Group 4 Alnus glutinosa Alnus glutinosa, cones

M49

M100 M116-117

VL 1b M118 M187

M101

M103

M108

1 3 1 59, 1 c 4 -

-

++++, + c ++ 3c +++ +++ ++ ++ -

+++ +++ 1c ++ + -

-

-

-

-

-

-

1c 2c 1c -

1c 1c

-

-

-

7c -

19 21

-

++ ++

-

-

-

-

-

Table III.11 part 5b.

201

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type Table III.11 part 5c.

202

M49

M100 M116-117

VL 1b M118 M187

M101

M103

M108

2 -

-

-

-

-

-

-

-

2 4 31

-

++ 2 ++ 3

-

-

-

-

-

1 2 6 18 8 -

-

++ 1 1 ++ ++ ++ 1 ++

+ 1 -

1 -

-

-

-

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M49

M100 M116-117

VL 1b M118 M187

M101

M103

M108

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

2 24 1 -

-

++ ++ 1 -

-

-

-

-

-

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

2 -

-

++, 2 c -

3, 2 c -

-

-

-

-

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

6 -

-

1 1c 1

+ + -

-

-

-

-

Table III.11 part 5d.

203

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M49

M100 M116-117

VL 1b M118 M187

M101

Group 9 Sand Buds 72 Bud scales Leaf remains Moss remains + fragments of catkins indet. Charcoal + +++++ + Bone/fish remains * + *+*+ + Pottery remains + + flint fragments Bark remains Insect remains ++ faecal pellets Lophopus cristallinus, statoblasts Daphnia sp., ephippia Cocoons ++ Cenococcum geophilum, sclerotia +c + Trichoptera, cases of larvae + cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula Anodonta sp./Unio sp., fragment tissue Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy) l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile

M103

M108

-

++ -

-

-

Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 5e.

204

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample unit square mesh width volume sediment context taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica

M109 D 9 5 ? peat refuse

VL 1b (cont.) M35 M8 10 11 1.5 5 ? ? peat peaty sand refuse colluvium

M40 11 1.5 ? peat refuse

1c 2c 3, 2 c 3c -

1c 7, 5 c 2c 2 10 5

4, 2 c 1c 2c 1 3c 1 17

1

VL 2b M20 M130 E 12 24 1.5 1.5 ? ? clay clay channel fill channel fill

1 7 1c 1 2

1c 3 1

Table III.11 part 6a.

205

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum Hordeum vulgare, internodia Hordeum vulgare, chaff remains Triticum dicoccon Triticum dicoccon, spikelet forks Triticum dicoccon, rachis segments Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases Group 4 Alnus glutinosa Alnus glutinosa, cones Table III.11 part 6b.

206

M109

VL 1b (cont.) M35 M8

VL 2b M40

M20

M130

-

3 -

1c -

10 4 23 -

1 -

2 -

9c -

1c 2c 4c -

-

1c 1c -

3c -

-

2

8 18

-

-

2 1

32 +

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type

M109

VL 1b (cont.) M35 M8

VL 2b M40

M20

M130

-

14 3

-

-

-

-

-

2 2

-

-

3

1 1 1

1c -

4 2 1 14 2 -

-

1 2 -

4 1 1 -

1 2 1 2 -

Table III.11 part 6c.

207

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M109

VL 1b (cont.) M35 M8

VL 2b M40

M20

M130

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

-

10 -

-

3 -

7 -

1 2 1 -

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

-

3c -

-

2c -

1c 1 -

2 1 -

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

-

1 +, + c

1

4 -

13 3 -

2 1 -

Table III.11 part 6d.

208

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 9 Sand Buds Bud scales Leaf remains Moss remains fragments of catkins indet. Charcoal Bone/fish remains * Pottery remains flint fragments Bark remains Insect remains faecal pellets Lophopus cristallinus, statoblasts Daphnia sp., ephippia Cocoons Cenococcum geophilum, sclerotia Trichoptera, cases of larvae cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula Anodonta sp./Unio sp., fragment tissue

M109

VL 1b (cont.) M35 M8

VL 2b M40

M20

M130

1 -

13 + ++ ++ + + ++ + -

-

+ ++ ++ ++ + + 1 -

7 + +++ +++ 8 3 3

+ ++ + -

-

+

-

-

-

1

Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy) l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 6e.

209

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample unit square mesh width volume sediment context taxon Group 1 Tilia platyphyllos Tilia sp. Acer campestre Quercus sp., cupulae Quercus sp., cupulae with content Quercus sp. Malus sylvestris Malus sylvestris, parenchyma Viburnum opulus Corylus avellana Prunus padus Prunus spinosa Cornus sanguinea Sambucus nigra Crataegus monogyna Rubus caesius Rubus fruticosus Ajuga reptans Anthriscus sylvestris Carex remota Chaerophyllum temulum Galium aparine Galium odoratum Glechoma hederacea Lamium album Ranunculus ficaria, tubers Urtica dioica Table III.11 part 7a.

210

M22 A 3 1.5 ? clay refuse

VL 2b* M28 A 3 1.5 ? clay pit

M41 B 25 1.5 ? clay pit

M19C A 17 ? ? sand pit

? M48 B 21 ? 130 g sand pit

2, 1 j 4 1 4 3

14, 2 j 2 -

25 +++, ++ j +++, +++ j 21, 3 j 2 11

-

-

M114 ? ? 1.5 ?

1c 1c -

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M22

VL 2b* M28

M41

Group 2 Aethusa cynapium Brassica rapa Bromus secalinus-type Capsella bursa-pastoris Carduus crispus Chenopodium album Chenopodium ficifolium Elytrigia repens fallopia convolvulus Galium spurium Persicaria lapathifolia Persicaria maculosa Silene latifolia ssp. alba Solanum nigrum Stellaria media Veronica hederifolia Vicia cf. tetrasperma

2 1 2 2 4 -

-

Group 3 Hordeum sp./Triticum sp. Hordeum vulgare var. nudum Hordeum vulgare, internodia Hordeum vulgare, chaff remains Triticum dicoccon Triticum dicoccon, spikelet forks Triticum dicoccon, rachis segments Triticum sp. Triticum sp., rachis segments Triticum sp., spikelet forks Triticum sp., glume bases

-

404 66

Group 4 Alnus glutinosa Alnus glutinosa, cones

M19C

? M48

M114

1 -

1 1c

-

-

-

-

-

-

1c -

-

95 89

-

-

1

Table III.11 part 7b.

211

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

phase sample

Copyright © 2010. Leiden University Press. All rights reserved.

Group 4 (cont.) Alnus glutinosa, fragments of male catkins Humulus lupulus Group 5 Alisma plantago-aquatica Berula erecta Caltha palustris Carex acutiformis Carex elata Carex pseudocyperus Carex riparia cf. Phragmites australis, stem fragments Cladium mariscus Eupatorium cannabinum Euphorbia palustris Hypericum cf. tetrapterum Iris pseudacorus Lythrum salicaria Mentha aquatica Menyanthes trifoliata Oenanthe aquatica Phragmites australis Rumex hydrolapathum Sagittaria sagittifolia Schoenoplectus lacustris Sium latifolium Solanum dulcamara Sparganium erectum Sparganium erectum spp. erectum Stachys palustris Thalictrum flavum Valeriana officinalis Veronica beccabunga-type Table III.11 part 7c.

212

M22

VL 2b* M28

M19C

? M48

M41

M114

5

-

3 -

-

-

-

6 6 56

-

8 4 4 4

-

-

-

1 20 2 4 8 26 110 18 11 1 -

-

4 22 2 4 1 1 -

1 -

-

-

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M22

VL 2b* M28

M41

Group 6 Persicaria hydropiper Ranunculus repens Ranunculus sceleratus Rumex cf. crispus

5 38 -

-

Group 7 Nymphaea alba Nuphar lutea Ranunculus cf. fluitans Potamogeton alpinus Potamogeton natans/perfoliatus Trapa natans, spines Trapa natans, bristles

14 13 1 29 -

11** 1 1 3 1 1

Group 8 Alopecurus myosuroides/pratensis Anthoxanthum odoratum Apiaceae Carex sp. cf. Holcus sp. Mentha aquatica/arvensis Phleum sp./Poa annua Poa pratensis/trivialis Poaceae Potentilla sp. Rosaceae Rumex sp. Scrophularia sp./Verbascum sp. Stellaria sp. Veronica austriaca/chamaedrys Lathyrus palustris/Vicia cracca Viola sp. Indet.

M19C

? M48

M114

5 -

-

-

-

-

-

-

-

-

-

2 -

1 -

2, 1 c

-

Table III.11 part 7d.

213

APPENDIx III - ARCHAEOBOTANy Of THE HAZENDONK, THE NETHERLANDS

Copyright © 2010. Leiden University Press. All rights reserved.

phase sample

M22

VL 2b* M28

M41

M19C

Group 9 Sand Buds 37 +++ Bud scales Leaf remains + Moss remains ++ + fragments of catkins indet. 13 + Charcoal + Bone/fish remains * +++ Pottery remains + flint fragments + Bark remains + Insect remains ++ faecal pellets 12 Lophopus cristallinus, statoblasts Daphnia sp., ephippia + Cocoons ++ ++ Cenococcum geophilum, sclerotia Trichoptera, cases of larvae ++ cf. Hypoxylon sp./Ustulina sp. Valvata piscinalis Bithynia tentaculata, opercula Anodonta sp./Unio sp., fragment tissue Haz = Hazendonk VL = Vlaardingen VL 2b* = younger phase of VL 2b (based on stratigraphy) l = litre, g = gram, refuse = refuse layer, / = or * = bone and fish remains were found in a carbonised and waterlogged state ** = possibly all representing Oenanthe aquatica = not present ++++ = some hundreds (100-499) + = few (1-10) *+ = some thousands (1000-5000) ++ = some tens (10-49) +++++ = many hundreds (500-999) +++ = many tens (50-99) *+*+ = many thousands (>5000) c = carbonised x, yc = x macroremains including y carbonised macroremains j = juvenile

? M48

M114

+ + -

-

-

-

Table III.11 The Hazendonk, macroremains that were collected during the excavation by active search for concentrations and by analysis of sieve residues, part 7e.

214

Appendix IV. Synthesis of archaeobotanical sources of the Hazendonk

Copyright © 2010. Leiden University Press. All rights reserved.

IV.1 IntroductIon The vegetation around the Hazendonk has been investigated intensively by various authors. The aim of this appendix is to compare the results from appendix III with the results of other pollen diagrams. Focus points are the reconstruction of the natural vegetation, human impact and the influence of distance on the evidence of human impact. At the Hazendonk, it is possible to distinguish at least six occupation phases (see table III.1) labelled as Hazendonk phases (Haz) and Vlaardingen phases (VL), related to the Swifterbant culture, Hazendonk group, and Vlaardingen group, and an activity phase related to the Bell Beaker culture. There are indications of Mesolithic occupation and a phase Haz 0 as well (Van Dijk et al. 1976; Van der Woude 1983) but data on these phases are scarce. See the introduction of appendix III and the original sources for more information on the Hazendonk, and chapter 2 for more information on the palaeogeography in the central river area. IV.2 SourceS The analysed sources are publications and reports of Steenbeek (1980), Voorrips (Louwe Kooijmans 1974, 136-143), Van der Wiel (1982), Van der Woude (1983), and new data on pollen and macroremains from the slopes of the dune (appendix III).39 The location of the cores of Steenbeek (1980) and Van der Woude (1983) is shown below, while the location of the other cores in shown in appendix III (fig. III.2). For each source, some details will be given on the pollen sum, the distance between the sample point and the dune, the occupation phases that can be recognised and the anthropogenic influence that is described in the original publications. The information on the pollen sum (included taxa and number of included pollen) indicates the quality of the diagram, the possibility of comparison between the different cores and the degree of suitability to answer questions about changes in dryland vegetation due to human activity. The new on-site pollen, non-pollen palynomorph and macroremains diagrams (appendix III) are the result of several cores and pollen boxes sampled in units B and C of the excavation. The diagrams give clear anthropogenic signals. The pollen analyses are based on an upland pollen sum varying between 266 and 492 pollen grains. The occupation phases that can be recognised are Haz 0, Haz 1, Haz 2a, Haz 2b, Haz 3, VL 1a and VL 1b. Occupation levels are distinguished on the basis of correlation with archaeological refuse layers, changes in the curves of trees, shrubs and herbs, changes in the diversity of taxa and the total number of macroremains and (partial) recovery of the vegetation after occupation phases. The core of Van der Wiel (1982) is sampled in unit B and concerns the analysis of pollen, non-pollen palynomorphs, macroremains and wood remains. The core is sampled at the same distance from the dune as some of the material presented in appendix III. The pollen analysis resulted in two diagrams: one diagram is based on an arboreal pollen sum (300 pollen) while the other diagram is based on an upland arboreal pollen sum (300 minus the wetland tree pollen). The diagram based on the upland sum is comparable with the recalculated diagram of Voorrips (presented below) and those in appendix III, and is therefore used in the following analysis. In the diagram of Van der Wiel, the occupation phases Haz 1, Haz 2, Haz 3 and VL 1b are distinguished by changes of mainly dryland trees, shrubs and herbs, and the presence of charcoal.

39 There is an additional study of Van Dijk (1979) that is not presented here since the preservation of the pollen was poor and since the results suggested reworking or contamination with secondary material.

215

Copyright © 2010. Leiden University Press. All rights reserved.

or yl us

Upland shrubs

20 40 20

Wood remains

Peat

Clay

Figure IV.1 part 1.

216 20

Upland herbs and spore plants 20 40

1000 20 40 60

2000 20

20

Wetland trees and shrubs

20 20

Wetland herbs and spore plants

pe

20 40 60 80 100

li x R h R am ib nu C es s or fra ng H nu um s ul s a C u u a lu e U nna s - t c ic a r ti b yp c a is e -t y So pe la n Ly um d t Ly hru ulc am s m Fi i m a ar lip c a h Ty en ia p h du Sp a la a a Sp rg ng a u a R r g niu st i f um an m o l M e iu er ia e x m e A l nt h a c e m c t u m i s a et Ir i ma - t yp osa er s s e - t y um pe - t y

4935 ± 40

Sa

4480 ± 40

or C nu ra s R t ae s a n h g g V i am us u i n bu nu - t y e rn s p a Eu u m c a t e ha o A r ny r ti ca te m u m C s he is i C no a e p Pl re a o d a li ia Po nt a a c e a g Pe l yg o o l a e nu nc r s Ec i c m e o h a A l i u m r i a av i l a t a m cu liu a Po m c u lar lo e P t ly p o sa er d i i A l d um nu ium s

C

C

150

D

an

an

at

lm

rc

xi n B e us t Ac ula e Fa r g H us e V i der sc a um

s

an

bs rb

ru he

sh

us

us

ue

Fr a

U

Q

s Pi ce Ab a i Ti es lia

d

d

th

nu

pl

pl

Pi

U

U

ep

d

sp

es (y (c rs B m Li th - N P) ol AP o U gy ) pl an d t re es

D

14 C

or

e

pl

an

ts

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Upland trees

200

3630 ± 35 250

300

350

400

450

500

5320 ± 40 550

600

650 20

Copyright © 2010. Leiden University Press. All rights reserved.

Ph as e

C a Eu m p p a R h o nul ub r b ac R us ia ea a c e Sc n u n i d a e e a e ro c u u s Po ph l a t u c ty G e nt lar e a e p e e n i l l ia L a t i a a - t ce m n y a Fa ia ac pe e b c e Po a c e a e a e lle ea n e su m 20

98

20 40 20 40 60

ea

ae

ac

ce

as

ia

er

e

Open water

20 40 60

si ca ce C ae ar G yo al ph i R um y ll o a S c s ac - t y p c e a ut ea e e A s el e te l a r ra ia ce ae As tu te bu ra li f ce G lo ae eu ra Er m e l i c - t y i gili ac p f lo e ea ra e e

Br

Ap

C

Wetland herbs and spore plants

yp

Bu t R om um u Va ex s T h ler i aqu a an a S y li c a t i c us m tr Pe p u m -g ro r s hy t up H i c um yp a r i e M r a e ic a O ny a u m m p sm n hi bi Ly u the a c o nd s Sp po a T h hag diu e l nu m yp m te r is N pa ym lu N p st up ha r is H ha ea ot r t Le o n m i Po n a a ac ea e

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Ecologically indeterminate

74

214

20 20 20

Ecologically indeterminate

VL 1b

Haz 3

Haz 1

Haz 0

2 4 6 8 10 x 100 Analysts: A. Voorrips and J. Mekel-te Riet

Figure IV.1 The Hazendonk, pollen diagram based on an upland pollen sum, exaggeration 5 x (after Louwe Kooijmans 1974), part 2.

217

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

The pollen diagram of Voorrips (Louwe Kooijmans 1974) is sampled at c. 10 metres distance from unit B of the excavation. The pollen sum exists of an unspecified upland pollen sum varying between c. 200 and 500 pollen grains. The published diagram shows a selection of taxa. The occupation phases that can be recognised by changes of several dryland and wetland taxa are Haz 0, Haz 1, Haz 3 and VL 1b. The analysis of the VL 1b level is based on two samples. Anthropogenic influence on the vegetation is described as a decrease in Quercus sp. and Fraxinus sp. and an increase in Corylus avellana (during some phases), and the presence and peaks of a variety of dryland and wetland herbs including pollen of Cerealia-type (recognised during all phases). As part of this study, the pollen diagram of Voorrips was recalculated based on an upland pollen sum (including dryland trees, shrubs, herbs and spore plants, again between c. 200 and 500 grains). This calculation resulted in a new diagram (fig. IV.1). This diagram is not discussed in detail (see Louwe Kooijmans 1974 for the interpretation) and has been used for further discussions in this appendix. The grey zones indicate the recognised occupation periods. The pollen analysis of Van der Woude (1983) is based on several off-site cores sampled in the extralocal area of the Hazendonk, at a distance of c. 1-3 km from the dune (see fig. IV.2 for the location of the cores). The pollen diagrams are based on an arboreal pollen sum of 300 pollen grains including both dryland and wetland trees, which makes the diagrams less comparable with the above mentioned diagrams. Together, the cores of Van der Woude could reflect all occupation phases of the Hazendonk. Anthropogenic influence related to Quercus sp. is suggested during some phases (Van der Woude 1983, details presented below). Cores H1110, H2114, H2115, H2118 and H2178 are located near the Hazendonk and may give information on human impact. Cores H2114, H2115 and H2118 are also located near another dune, the Kweldamsdonk, and will probably reflect the vegetation development at this dune.

e wo e r

d st

re am

ridg e

Polderweg - west

H 2115

H1110

Polderweg - oost

S choonr

H 2114

H2118

Copyright © 2010. Leiden University Press. All rights reserved.

H2178

Hazendonk

in Pr o v

cial r

oad

0

300 m

Figure IV.2 The Hazendonk, the location of the pollen cores presented by Steenbeek and Van der Woude (Van der Woude 1983).

218

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Copyright © 2010. Leiden University Press. All rights reserved.

The study of Steenbeek (1980), concerning core H2114 and H2115 (see fig. IV.2), was performed in close cooperation with Van der Woude. The analysis of core 2114 resulted in three diagrams: one based on an arboreal pollen sum of 300 pollen grains (H2114), one on the same pollen sum without Alnus sp., characterised by a very low pollen sum (H2114a), and one with an arboreal pollen sum without Alnus sp. of 200 pollen grains (H2114b). The pollen sum of H2115 consisted of 300 arboreal pollen grains. IV.3 r econStructIon of the natural VegetatIon Despite the differences in methodology, pollen sum and distance to the dune, the reconstructions of the natural local and extra-local vegetation at the Hazendonk are roughly similar amongst the various sources. In the discussion below, the dryland vegetation is emphasised since the dryland vegetation is primarily expected to be influenced by human activity. The large-scale developments of the landscape and the wetland vegetation are extensively described by Van der Woude (1983) with emphasis on the dynamics of lakes and rivers. From c. 5000 BC onwards, the environment is dominated by Alnus carr vegetation combined with Phragmites vegetation, marsh ferns, Cyperaceae and calm lakes. After phase Haz 1, the lakes expanded at the expense of swamp forest (alder carr), and clastic deposition increased, presumably resulting in better accessibility of the Hazendonk (Van der Woude 1983, 89). The water depth decreased from c. 3360 BC onwards (around phase VL 1a). The alder carr remained present but was more alternated with open patches. Initial fluvial activity might have been present in the area after c. 3360 BC, while fluvial deposition strongly increased from c. 2600 BC onwards. The local environment of the Hazendonk can be characterised as a mainly eutrophic environment. There were some taxa present at the dune that indicate patches of mesotrophic conditions due to peat growth in stagnant water that were not influenced by river water, though these patches were a minor element in the vegetation. prehistoric soils were not found during the excavation of the dune (pers. comm. Louwe Kooijmans 2004). The subsurface of the dune consisted of sand that was locally calcareous on the lower parts of the dune, although soil formation had probably taken place already since the development of the dune in the younger Dryas, which presumably had affected the higher parts of the dune in particular. Traces of calcium may additionally have been supplied by seepage water and river water. In the wetlands around the dune, many humic clayey patches were present, as well as sandier patches. The Hazendonk dune was a small patch of dryland terrain in the middle of wetlands, covered with woodland vegetation of dry terrain. Before the first known period of anthropogenic influence, a patch of rather dense Atlantic woodland vegetation was presumably present on the top and slopes of the dune. The presence of the indicator species Tilia sp. and Pteridium aquilinum confirm the presence of old woodlands (Weeda et al. 1985, 1987), while taxa such as Galium odoratum, Carex remota and Neckera crispa (found in the macroremains assemblage of the excavation) indicate shaded conditions. Dryland vegetation remained present on the dune at least until the period after phase VL 1b and probably also during phase VL 2. unfortunately, there are no pollen diagrams available that reflect the vegetation at the time of occupation phase VL 2. Around the dune, a variety of marshes, lakes and small channels were present. These wetlands were covered by vegetation consisting of alder carr, riparian vegetation and open water vegetation. The increasing water table caused a shift of vegetation belts along the slope of the dune in an upwards direction. This can for example be demonstrated by the presence of Alnus sp.; Alnus carr developed at the slope of the dune between the phases Haz 1 and 2 and gradually became the dominant element of the vegetation in unit C. Shortly after phase VL 1b the influence of water activity increased strongly, resulting in the presence of Salix sp., the increased presence of water plants and a decrease in Alnus sp. at the southern and eastern side of the dune (appendix III; Louwe Kooijmans 1974; Van der Wiel 1982).

219

Copyright © 2010. Leiden University Press. All rights reserved.

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

A short summary of Van der Woude (1983, 101) illustrates his reconstruction of the natural vegetation in the area of the Hazendonk: “The arboreal vegetation of these dry sites consisted mainly of Quercus, Ulmus and Corylus, and some Tilia on the higher parts of the dunes. … Marsh herbs occurring along the margins and at shallow places of the wet basins, and at open sites in swamps forests [mainly Alnus swamp], were among others Phragmites (and other hygrophilous grasses), Typha angustifolia, ferns, Scirpus (and other Cyperaceae), Alisma and umbelliferae [Apiaceae].” Van der Wiel (1982) discusses the development of vegetation between approximately 4300 and 2600 BC at the southern side of the dune. She describes the presence of a mixed Tilia/Quercus woodland on the higher parts of the dune, which she compares with a Quercus/Fraxinus vegetation rich in Ulmus sp. (ulmion carpinifoliae), sometimes partly replaced by Fraxinus sp. and in the long term replaced by Ulmus sp. due to increasing water levels. As a result of the rise of the water table, peat formation increased and mixed Alnus-Fraxinus carr (comparable with the Circaeo-Alnion), present on the lower parts of the slope, changed into backswamp vegetation. Vegetation comparable with the nymphaeion was replaced by reed vegetation (phragmitetalia), and afterwards by Salix carr (Salicion albae). Voorrips (Louwe Kooijmans 1974) describes woodland dominated by Quercus sp. with an undergrowth of Corylus avellana on top of the dune, a transitional zone with Ulmus sp. and Fraxinus sp. lower along the slope, followed by Alnus vegetation combined with wetland herbs, ferns and Carex species. principally, these sources reflect well the vegetation present at the Hazendonk, and it should be realised that the descriptions above are only summaries of detailed studies. However, all three summaries only partially recognise the large variety of the natural vegetation that is especially shown by the new data (see also paragraph III.4.1). In addition to the underestimation of the variety, the importance of Tilia sp. is strongly neglected by most authors except for Van der Wiel. The differences seem to be caused by the fact that the studies of Voorrips (Louwe Kooijmans 1974) and Van der Wiel (1982) are only based on a single core. The cores of Van der Woude (1983) are moreover taken at some distance from the dunes and therefore focus on the wetland vegetation. The natural vegetation at the Hazendonk can be summarised as deciduous woodland vegetation. The vegetation on the top and higher parts of the dune was presumably more or less comparable with the modern-day Stellario-Carpinetum (major elements Tilia sp. and Quercus sp. but in the case of the Hazendonk no Fagus sp. or Carpinus sp.). The vegetation at the lower parts of the slope that were temporarily flooded during the year was comparable with the Alno-padion (major elements Fraxinus sp. and Ulmus sp.), and the vegetation at the lowest parts of the dune that were inundated during large parts of the year with the Alnetea glutinosae (major element Alnus sp.) (Stortelder et al. 1999). The vegetation of the various communities was probably present in mosaic-like structures consisting of a variety of patches. It should be realised that the reconstructed vegetation communities that were present at the Hazendonk (especially the Stellario-Carpinetum) differed considerably from the modern-day vegetation since the composition of the flora and plant communities has changed due to various factors including human impact. Detailed reconstruction of the development of the vegetation (including the reconstruction of human impact) at the Hazendonk is only possible for the southeastern side of the dune, despite additional indications of occupation at the northwestern side of the dune. The only pollen cores sampled at the northern side are the cores of Van der Woude (1983) of which the value for the reconstruction of human impact is discussed below.

220

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

IV.4 human Impact The paragraphs below compare the signals of human activity as recognised by investigators for each occupation phase separately. The signal of specific vegetation as reflected by pollen diagrams generally decreases when the distance between the sample point and the vegetation increases, depending on the size of the catchment basin and the strength of the signal of the vegetation. Similarly, only pollen diagrams sampled close enough to human activity on the vegetation reflect the anthropogenic influence on the vegetation (Behre and Kučan 1986; Fægri and Iversen 1989). It is therefore expected that cores that are sampled close to the Hazendonk will reflect more strongly the anthropogenic influence on the vegetation at the dune than cores sampled at some distance from the dune.

Copyright © 2010. Leiden University Press. All rights reserved.

IV.4.1 MesolIthIc occupatIon The pollen diagrams of Van der Woude (1983) are sampled at 1 to 3 km from the dune, in parts of the landscape where the pleistocene subsurface is generally lower than at the dune. One the one hand, these pollen diagrams are suited to study the signals of Mesolithic occupation since peat growth (and registration of the development of the vegetation) started relatively early in the lower parts of the landscape. On the other hand, the relatively large distance between the sample points and the dune restricts the possibilities to recognise human influence on the vegetation. It is suggested that an anthropogenic signal related to Mesolithic occupation is present in the diagram of core H2118 (see figure. IV.1 for the location). The signal is characterised by a fall of the Quercus curve, dated to c. 6000-5600 BC in another core (Van der Woude 1983, 45-47). Although the possibility of an anthropogenic signal in core H2118 is supported by changes that correspond with later anthropogenic signals that are more strongly confirmed by archaeological indicators (see below), other curves indicate that it might concern a change in the water table instead of an anthropogenic signal. It is therefore not possible to distinguish the cause of the changes in the vegetation unless further archaeological investigations are performed.40 Van der Woude (1983, 46) also suggested that the signal in core H2118 possibly corresponds with indications of occupation in the diagram of core H2178 (again a fall of the Quercus curve). The diagram is however not completely published, a statistical cause for the decrease in Quercus sp. is not excluded, and the association with Mesolithic occupation is tentative (Van der Woude 1983, 45). Therefore, this diagram does not provide sufficient evidence of human activity. IV.4.2 hazendonk 0 In the diagram of Voorrips (Louwe Kooijmans 1974), phase Haz 0 can be recognised by a fall and recovery of Tilia sp., a peak of Quercus sp., low values of Ulmus sp., a small peak of Hedera sp., an initial rise of Corylus sp., the presence of Crataegus-type, Artemisia sp., Cerealia-type (a single pollen grain), Urtica sp. and Brassicaceae, and a decline of Cyperaceae. In the pollen diagrams of square 57 and core 3 (appendix III), comparable changes take place during this phase, but Ulmus sp. does not decrease, pollen of Cerealia-type and Crataegus-type are not present, pollen of Plantago major is present, Viburnum sp. shows a small peak (core 3) and the Cyperaceae show a small peak. Differences between the pollen diagrams of square 57 and core 3 are explained by local vegetation differences. 40 The possibility of an anthropogenic signal in core H2118 is supported by the following changes in the pollen diagram: a decrease in the AP curve, an increase in Corylus sp., Viburnum sp. and monoletae psilatae fern spores, and the increased precipitation of (regional) Pinus pollen that indicates the presence of open patches in the vegetation. These changes correspond with later anthropogenic signals that are confirmed by archaeological indicators. The indications that these changes are related to a change in the water table are the absence of an increase in the dryland herbs, and an increase in Ulmus sp., Salix sp., Cyperaceae and Typha angustifolia. In the case of a change in the water table, the increase in Corylus sp. and Viburnum sp. would be caused by the increased presence/flowering of these shrubs in gaps in the vegetation where dryland trees disappeared due to submergence.

221

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Copyright © 2010. Leiden University Press. All rights reserved.

The macroremains diagram of core 3 shows very limited signals of occupation on the vegetation in the near surroundings during phase Haz 0 (an increase in macroremains and species diversity), while the macroremains diagram from square 57 shows two fruits of Urtica dioica. Van der Woude (1983, 31) suggests the possibility of anthropogenic influence corresponding with phase 0 in core H111041 (see figure IV.1), consisting of a decrease in Alnus sp. and Corylus sp., although a temporarily fall of the water level is suggested as well. The signal in the pollen diagram is not strongly characteristic of anthropogenic influence at the Hazendonk. The imprecise dating of core H1110 and phase Haz 0 hampers further interpretation. IV.4.3 hazendonk 1 In the diagrams of Voorrips (Louwe Kooijmans 1974), Van der Wiel (1982) and those presented in appendix III, phase Haz 1 is characterised by a decrease in Tilia sp. which is followed by decreased values of Cyperaceae, increased values of Quercus sp., an increase in Corylus sp., Alnus sp. and poaceae and the presence of Cerealiatype, Artemisia sp., Chenopodiaceae, Plantago lanceolata, Allium sp., Polygonum persicaria-type (Persicaria maculosa), Polypodium sp., Pteridium sp., Humulus sp., Urtica sp., Filipendula sp., Lythrum sp. and Apiaceae. These changes are probably directly linked to human activity, although the increases in Quercus sp. and Corylus could be partly a statistical result of the decrease in Tilia sp. The curve of poaceae increases during occupation at the southeast side of the dune (appendix III) but this signal is already diminished in the diagrams of Van der Wiel (1982) and Voorrips (Louwe Kooijmans 1974). The same is true for the fall of the Cyperaceae (that show a pattern opposite to the grasses). This shift in the strength of the anthropogenic signal directly corresponds with the distance between cores and the location of most of the archaeological finds (unit C). The increased values of Alnus sp. during phase Haz 1 probably reflect increased flowering or the increased presence of alder trees at some distance from the dune since the macroremains do not support the increase in alder in the local vegetation (appendix III, Van der Wiel 1982). Concerning the diagram of Voorrips (Louwe Kooijmans 1974), some additional remarks can be made. The curves of Betula sp., Cornus sanguinea, Crataegus-type and Ribes-type show small peaks. The small increase in these taxa indicates the increased presence of light due to the clearance of trees (particularly Tilia sp.), which corresponds with more open vegetation. Comparably, Viburnum sp. shows a high peak, which must represent local vegetation since this peak does not correspond with other diagrams. In the macroremains diagrams in appendix III, phase Haz 1 is reflected by an increase in Urtica dioica, Persicaria maculosa, Veronica beccabunga-type, Lythrum sp., Chenopodium album, Valeriana officinalis and others, which correspond very well with the information from the above-mentioned pollen diagrams. In core H2118 (Van der Woude 1983; sampled at c. 40 metres distance from the Kweldamsdonk, see paragraph IV.2) there are some minor fluctuations, and a pollen grain of poaceae > 40 µm is present in the zones that presumably correspond with phase Haz 1 (zones 2a and 2b), though there are no changes in the diagram that can certainly be related to anthropogenic influence. This absence of a clear anthropogenic signal indicates that the occupation pressure on the Kweldamsdonk during the period that presumably corresponds with phase Haz 1 (organic layer 2-3) was probably limited. For core H2114(b) (Steenbeek 1980), it is suggested that the influence of occupation phase Haz 1 is expressed by the presence of Rumex sp. and Artemisia sp. in the upper part of the diagram. This is tentatively confirmed by an increase in Corylus sp. and a fall of Cyperaceae.

41 Zone 10, upper part of organic layer 2-3 dated roughly between 5200 and 4300 BC; 6060±80 and 5590±70 BP.

222

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Copyright © 2010. Leiden University Press. All rights reserved.

IV.4.4 hazendonk 2 The indications of anthropogenic influence on the natural vegetation during phase Haz 2 are limited. Van der Woude (1983) and Voorrips (Louwe Kooijmans 1974) do not mention this occupation phase explicitly, and indeed the phase is not easily recognised in their diagrams. Van der Wiel (1982) suggests that phase Haz 2 is characterised by a fall in the curve of Quercus sp. at the end of the phase and by increased values of Artemisia sp., Chenopodiaceae, Galium-type and occurrence of fruits of Rumex acetosella (between samples 50 and 60). In addition, the phase is characterised by the presence of pollen grains of Plantago lanceolata, both pollen and fruits of Persicaria maculosa and an increase in poaceae, Apiaceae and Urtica sp. Van der Wiel’s interpretation implies that human impact during phase Haz 2b would be stronger and would have lasted longer than during phase Haz 1. This is however unlikely since such changes are not confirmed by the results of appendix III and since this does not correspond with the restricted distribution of refuse of phase Haz 2. It is more likely that the signal of phase Haz 2 recognised by Van der Wiel probably represents phase 2a, which would be characterised by clearance of Quercus sp. and restricted presence of indicators of disturbance and open patches. In addition, the changes higher in the diagram (at spectrum 67) may represent phase 2b. This signal of phase Haz 2b consists of increased values of Chenopodiaceae, Artemisia sp., Galium-type, Plantago lanceolata, Urtica sp., poaceae and Apiaceae. The diagram does not show a strong contrast between the supposed phases 2a and 2b, except for the increase of Tilia sp. after the presumed phase Haz 2a. In contrast to the signal of previous occupation phases in this diagram, the Cyperaceae seem to increase here in reaction to occupation, indicating open vegetation. In the pollen diagram of square 57 (appendix III), the phases Haz 2a and 2b are both recognisable with considerable accuracy due to the close distance between the dune, the location of occupation and the sample location. The signal corresponds with other anthropogenic signals in the diagram and is characterised by low values of Quercus sp. (during phase Haz 2a), Fraxinus sp. and Cyperaceae, and increased values of Corylus sp., Artemisia sp., Plantago lanceolata (phase 2b), Polypodium sp., Pteridium sp., Humulus sp., Filipendula sp., Lythrum sp. and Urtica sp. In the macroremains diagrams of square 41 and core 57 (appendix III), it is possible to recognise phase Haz 2, which is characterised by the presence of Urtica dioica and to a lesser degree by P. maculosa, Chenopodium album, Solanum dulcamara, Lythrum salicaria, Typha sp. and Juncus effusus. The presence of these taxa corresponds moderately with the pollen diagrams (Juncus pollen was rarely found during analysis). In conclusion, only the material sampled at a very close distance (appendix III and Van der Wiel 1982) reflects the anthropogenic signal of phase Haz 2, suggesting clearance of Quercus sp. The limited expression of this occupation phase corresponds with the limited presence and spread of the occupation horizons and archaeological finds. Cerealia-type pollen grains were absent in all pollen diagrams during the phase, but macroremains finds from one sample indicate that cereals were present (see paragraph III.3.8.1). IV.4.5 hazendonk 3 Van der Woude (1983) does not mention phase Haz 3. In the diagram of Voorrips (Louwe Kooijmans 1974), phase 3 is characterised by a fall of Quercus sp. at the start of the phase followed by recovery, a decrease in Tilia sp., Fraxinus sp. and Cyperaceae, a rise of Betula sp. and Corylus sp., and increased values of Cerealia-type, Artemisia sp., Chenopodiaceae, Rumex acetosa-type, poaceae, Humulus sp., Urtica sp., Solanum dulcamara, Lythrum sp., Sparganium erectum, Sparganium emersum-type, Apiaceae and Asteraceae tubuliflorae. Voorrips (Louwe Kooijmans 1974) argued that the vegetation on the dune was cleared in favour of arable land while “bush vegetation” on the lower slopes of the dune was saved, resulting in “marshy underwood”. The practice of crop cultivation on the dune, the working hypothesis used during the late seventies, is however not proven (see paragraph III.4.5). The anthropogenic signal of the diagram of Voorrips corresponds well with the diagram of square 57 (appendix III). In the latter diagram, phase Haz 3 is characterised by the same taxa as in the diagram

223

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Copyright © 2010. Leiden University Press. All rights reserved.

of Voorrips, as well as by a high peak of Viburnum opulus. Interestingly, Viburnum sp. characterised phase Haz 1 in the diagram of Voorrips (Louwe Kooijmans 1974). Apparently, this species is indicative of anthropogenic disturbance, but is only locally present and spreads only small quantities of pollen, which results in the presence of peak values in single diagrams only. In the macroremains diagram of square 41 (appendix III), phase Haz 3 is characterised by Lythrum sp., Chenopodium ficifolium and Typha sp., which corresponds well with the pollen diagrams. Van der Wiel suggests that phase Haz 3 is reflected in 36 cm of sediment. She mentions a fall of Tilia sp. and the presence of sand at sample 85, and a peak of fruits of Urtica dioica at sample 115, which corresponds with presence of charcoal and sand in the sediment. Van der Wiel (1982, 70) explains the limited strength of the signal by a tentative hypothesis that “..occupation started at the northside of the ‘donk’ [dune] and gradually shifted, via the eastside, to the southside of the ‘donk’..”. If the weakness of the signal must be explained, not shifting occupation but simply the distance to occupation can be put forward. However, it can be argued that the signal is not weak since there are various other changes characteristic of anthropogenic influence at the Haz 3 level. At spectrum 85, Chenopodiaceae, Artemisia sp. and Cerealia-type are present or increase. Spectrum 115 is characterised by the presence of much sand and charcoal, a decrease in Tilia sp., Quercus sp. and Alnus sp., an increase in Corylus sp., a peak of Urtica fruits, and the presence of Solanum nigrum seeds. The curve of Plantago lanceolata starts only after phase Haz 3 and is not related to occupation (see also III.3.1). The number of spectra that is suggested to reflect phase Haz 3 in the diagram of Van der Wiel (1982) represents a very long period compared with the other occupation phases of the same diagram and the same phase in other diagrams. Two explanations are possible for the thick horizon in the core of Van der Wiel: 1) Sedimentation took place rather fast, possibly because of the erosion of material at the top of the dune and colluviation, and the anthropogenic signals at spectrum 85 and 115 both correspond with phase Haz 3. This could have occurred locally but is not confirmed by the other pollen diagrams sampled around the Hazendonk. 2) The signal at spectrum 115 corresponds with phase VL 1a (all occupation between phases Haz 3 and VL 1b). This hypothesis is however not likely since finds of this occupation phase were rarely found at this side of the dune and since the anthropogenic signal of this phase is weaker than that of Haz 3 in the other pollen diagrams sampled at a close distance to the dune (appendix III). It must therefore be concluded that the signal of phase Haz 3 in the undated diagram of Van der Wiel (1982) is difficult to correlate with archaeological finds and other pollen diagrams. IV.4.6 VlaardIngen 1a Anthropogenic signals in pollen diagrams related to phase VL 1a (all occupation between the phases Haz 3 and VL 1b) are not mentioned in Voorrips (Louwe Kooijmans 1974), Van der Wiel (1982) or Van der Woude (1983). The possibility of recognising a signal of this phase in the diagram of Van der Wiel is discussed above. Concerning the diagrams of Van der Woude (1983), it is indeed not expected that the signal of this phase can be distinguished when considering the visibility of anthropogenic influence on the vegetation during previous occupation phases and the restricted extent of the refuse layer of this phase. In the diagram of Voorrips, there are changes in between the horizons of phases Haz 3 and VL 1b horizon at several levels that may be related to anthropogenic influence but the evidence is not convincing. In the pollen diagram of core 2 (appendix III), the anthropogenic signal of phase VL 1a can be recognised by very weak changes: a small decrease in Tilia sp., a small rise of Quercus sp. and Chenopodiaceae, a slight decrease in Cyperaceae and the presence of Urtica sp. In the pollen diagram of square 57, minor anthropogenic signals visible between phases Haz 3 and VL 1b may represent phase VL 1a (see paragraphs III.3.1 and III.4.3.2). In the macroremains diagrams of core 2 and square 41 (appendix III), the VL 1a horizon is uniquely characterised by extreme values of Cornus sanguinea and Sambucus nigra, which probably reflects the development of the local vegetation. In conclusion, there is little evidence of human impact on the vegetation on the Hazendonk during phase VL 1a.

224

Copyright © 2010. Leiden University Press. All rights reserved.

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

IV.4.7 VlaardIngen 1b In contrast to phase VL 1a, the anthropological influence on the natural vegetation during phase VL 1b is comparable with phases Haz 1 and 3. In the diagram of Voorrips (Louwe Kooijmans 1974), the VL 1b level is characterised by a small decrease in Corylus sp., the presence and increase in Betula sp., Rhamnus cathartica, Artemisia sp., Chenopodiaceae, Allium sp., Cerealia-type, poaceae, Alnus sp., Urtica sp., S. dulcamara, Lythrum sp., Filipendula sp., Sparganium emersum-type, Symphytum sp., Thelypteris palustris, Apiaceae, Brassicaceae and Asteraceae tubuliflorae and a decrease in Cyperaceae. It is not possible to recognise a fall of Tilia sp. or a strong rise in the curve of Quercus sp., although small-scale changes may be lost because of the limited number of analysed samples. In the diagram of Van der Wiel (1982), phase VL 1b is characterised by a decrease in Tilia sp. and Corylus sp., low values of Quercus sp. and Alnus sp. in the early part of the phase, and an increase in Fraxinus sp., Chenopodiaceae, Artemisia sp., Cerealia-type, Persicaria maculosa, Solanum nigrum, poaceae, Cyperaceae and Urtica sp. Afterwards, several marsh herbs (Apiaceae, Alisma sp., Mentha aquatica, Lythrum salicaria) confirm the presence of eutrophic open patches in the terrain. The increase in Fraxinus sp. (visible in the upland diagram in particular) is probably related to changes in the water table and flooding frequency. In the pollen diagrams of core 2, M86 and M87 (appendix III), phase VL 1b is characterised by a slight decrease in Tilia sp. (not in M86), a decrease in Quercus sp. (in M87 only in the beginning of the phase, followed by recovery) and a decrease in Alnus sp., an increase in dryland shrubs, herbs and anthropogenic indicators including Cerealia-type, and an increase in poaceae, Cyperaceae, Urtica sp. and Solanum dulcamara. Several taxa show increased values just after the occupation phase (Lythrum sp., Solanum dulcamara, Sparganium erectum, Apiaceae and rubiaceae). This corresponds with the macroremains diagrams of square 41, M86 and M87 (appendix III), which also showed the presence of plants indicating disturbance and increased eutrophication of both the dryland and wetland terrain both during and after the occupation phase. In core H1110, located 1 km north of the Hazendonk, Van der Woude (1983, 33, zone 15 and 20) recognises anthropogenic influence consisting of decreases in Quercus sp. that could be related to phase VL 1b. repetitive occupation of such intensity that a signal is expected at 1 km distance is however not known from the excavation. The decrease in Quercus sp. in zone 15 could represent phase VL 1b since the diagram shows additional changes consisting of a peak of Tilia sp., a peak of poaceae, high values of Corylus sp., a decrease of Alnus sp., the presence of small peaks of Artemisia sp. and Asteraceae, and the presence of Chenopodiaceae and Rumex-type. The increases in Tilia sp. and Corylus sp. may be a statistical effect of the decrease in Quercus sp. and Alnus sp. Altogether, the various sources indicate that phase VL 1b is characterised by a slight decrease in Tilia sp. (only registered at some sample locations near the dune) and a decrease in Quercus sp. and Alnus sp., all pointing to clearance of the woodland vegetation. The decrease in Quercus sp. partly differs from earlier increases in the curves of this taxon, which may be related to the reduced surface of the dune and the resulting changes in the vegetation. A major characteristic of this phase is a strong signal in the macroremains diagrams, which may be related to increased or changed anthropogenic influence (possibly related to the palisade, see paragraph III.1) and the increased water level (resulting in good preservation). Further, most diagrams from the nearby sample locations show an increase in Cyperaceae. The difference in the curve of Cyperaceae between the various studies and the contrast with the earlier phases that mainly showed a decrease may be explained by differences and changes in the local vegetation, the increased presence of this taxon and increased openness of the vegetation near the sample locations during occuation, while statistical effects and variation in plant use through space cannot be excluded.

225

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

Copyright © 2010. Leiden University Press. All rights reserved.

IV.4.8 other occupatIon perIods The diagram of Van der Wiel (1982) does not show archaeological levels later than phase VL 1b. Some changes in the upper part of the diagram could be interpreted as signals of human influence (decrease in Quercus sp., rise of Corylus sp.), though peaks of herbs are absent and natural changes in the vegetation could also explain the changes. In the diagram of Voorrips (Louwe Kooijmans 1974), the next occupation period recognised is related to the Bell Beaker culture, characterised by considerable values of Cerealia-type and Plantago lanceolata. Van der Woude (1983, 33) suggests that low values of Quercus sp. and high values of Corylus sp. in core H1110 represent occupation related to phase VL 2b. It is not possible to confirm or reject this hypothesis since the sediment consists of peaty clay (leading to the possibility of reworked material in the clay) and since changes in other curves that could indicate anthropogenic influence are limited. There are pollen grains of Cerealia-type present in this zone but they do not correspond well with the changes in the curves of Quercus sp. and Corylus sp. Higher in the core, there are some other potential signals of occupation (the presence of Chenopodiaceae, Brassicaceae, Plantago lanceolata and Cerealia-type; Van der Woude 1983, 34), corresponding with the changes in the diagram of Voorrips (Louwe Kooijmans 1974). The peaks occur roughly around 1800-1400 BC (3340±80 Bp) and may correspond to activity by people of the Bell Beaker culture. IV.5 Summary of human Impact The combination of various studies sampled at different distances from the Hazendonk and at various distances from the zone of occupation (refuse) provides a relatively valid reconstruction of human impact. The pollen diagrams from appendix III, Voorrips (Louwe Kooijmans 1974) and Van der Wiel (1982) all reflect the main occupation phases and provide rather similar evidence of human impact. The data presented in appendix III provide the most precise information on human impact due to their location and the recognition of most of the refuse layers in the cores and sections. The diagram of Van der Wiel (1982) provides detailed information but the relation with the known occupation phases is not always clear. The diagram by Voorrips (Louwe Kooijmans 1974) provides information on the main occupation phases. In contrast, the diagrams of Steenbeek (1980) and Van der Woude (1983) hardly provide detailed information on human impact (further discussed in the next paragraph). Only core H2114 (Steenbeek 1980) shows indications of human impact that may correspond with phase Haz 1, although this may reflect activity at the Kweldamsdonk rather than at the Hazendonk. The general pattern obtained from the diagrams sampled at the slope of the dune is that occupation during neolithic occupation phases resulted in the clearance of particularly Tilia sp., Quercus sp. and Alnus glutinosa, the clearance of Fraxinus excelsior during phases Haz 2 and 3, an increase of Corylus avellana, the presence or increased presence of Artemisia vulgaris, Chenopodiaceae, Plantago lanceolata, poaceae, Humulus lupulus, Urtica dioica, Lythrum salicaria, Solanum dulcamara, Apiaceae, while locally/during some phases Viburnum opulus, Rhamnus cathartica, Allium sp., Cerealia(-type), Polygonum persicaria-type, Polypodium vulgare, Pteridium aquilinum, Filipendula ulmaria, Mentha sp. and Brassicaceae increased. The Cyperaceae decreased during early phases and increased during phase VL 1b, which suggests disturbance at least during the early phases. The diagrams of Voorrips (Louwe Kooijamns 1974) and appendix III show that also shrubs other than Corylus avellana, Viburnum opulus and Rhamnus cathartica show increased values during occupation. The restricted presence of similar results in other studies can primarily be related to restricted production and dispersal of pollen of many shrubs. Together, the changes of the vegetation indicate a small-scale disturbance of the dense woodland of dry terrain, as well as the disturbance of the wetland vegetation around the dune, as indicated by the increased values of wetland herbs that are not known as typical anthropogenic indicators (see also Van der Wiel 1982, 88). The disturbance resulted in a strong increase in light and was characterised by the presence of a variety of patches of shrubs, forb and herb vegetation and probably open terrain (in addition to patches of relatively

226

AppenDIx IV - SynTHeSIS OF ArCHAeOBOTAnICAL SOurCeS OF THe HAzenDOnK

undisturbed woodland). Many herbs and ferns indicative of human impact were part of the natural vegetation and did not depend on the presence of people. It is not the presence of these taxa that support the evidence of human activity but rather the changes in the curves and the contemporaneous presence of several taxa that supports their role as anthropogenic indicators. The detailed analysis of human impact enables the assessment of some of the conclusions from earlier studies. Firstly, there are clear indications of the clearance of Tilia sp. at the Hazendonk (as demonstrated in appendix III), which was relatively unknown from earlier pollen diagrams from the site and only suggested by Van der Wiel (1982) for phases Haz 1 and 3. Furthermore, Van der Woude (1983, 30, 33, 87) suggested that anthropogenic influence at the Hazendonk mainly resulted in a decrease in Quercus sp. during the Late Mesolithic, phase VL 1b and a late Vlaardingen phase. The discussion above shows that the effect of human impact affected much more taxa and occurred during more occupation phases. Thirdly, the results of appendix III indicate that occupation affected Alnus glutinosa. This was not recognised in the studies by Voorrips (Louwe Kooijmans 1974) and Van der Wiel (1982). It is indeed difficult to distinguish a clear relationship between fluctuations of Alnus sp. and occupation phases in their diagrams.

Copyright © 2010. Leiden University Press. All rights reserved.

IV.6 dIStance between human actIVIty and SamplIng poInt In addition to the results presented in paragraph III.4.3.2, comparison of the various palynological data from the Hazendonk supports that cores that are sampled close to the dune reflect anthropogenic influence on the vegetation at the dune stronger than cores sampled at some distance from the dune. The distance to the dry surface and to the location of human activity plays a major role. At the Hazendonk, human activity is best represented in pollen cores that are sampled at the foot of the Hazendonk (appendix III; Van der Wiel 1982 and additionally Louwe Kooijmans 1974). The distance to the main zone of occupation (refuse) plays a role as well (see paragraph III.4.3.2). In contrast, the diagrams from cores sampled at a distance of 1 km or further from the dune do not give a signal that can be interpreted with certainty as anthropogenic (Steenbeek 1980; Van der Woude 1983). Absence of evidence of human impact in pollen diagrams located at a relatively large distance from the dune may be explained by disturbance and blurring of the signal of human impact on the dryland vegetation of the Hazendonk due to signals of other dryland patches, regional pollen rain (in the case of large catchment basins), and dominance of the pollen rain of the wetland vegetation (which is reinforced by a pollen sum that includes Alnus sp.). The absence of clear indications of human impact in the diagrams sampled at 1 km distance from the dune (Van der Woude 1983) indicates that anthropogenic influence on the wetland vegetation had a local character and was of restricted strength.

227

Copyright © 2010. Leiden University Press. All rights reserved.

Appendix V. Archaeobotany of Bergschenhoek, the Netherlands

Copyright © 2010. Leiden University Press. All rights reserved.

V.1 IntroductIon The Early Neolithic site near Bergschenhoek was discovered in 1976 during digging works north of Rotterdam (see fig. V.1). The site was embedded in intra-coastal clay deposits at a depth of 8 m -NAP and was extremely well-preserved. It was excavated in 1978 by the National Museum of Antiquities under the direction of Louwe Kooijmans. This rescue excavation of c. 100 m2 uncovered a small short-term fishing-fowling site and its immediate surroundings. The site has been dated on the basis of three 14C dates to the period 4350-4050 BC, which is in agreement with a restricted number of pottery sherds in a distinct Swifterbant style. This appendix presents the archaeobotanical analyses combined with mollusc analysis. The introduction is based on several preliminary publications (Bloemers et al. 1981, 42-45; Casparie 1995, 212-213; Louwe Kooijmans 1977a, 1978, 1985, 92-96, 1986, 7-11, 1987, 238-242).

coastal barriers

salt marshes

tidal flats

pleistocene coversand

river deposits

inland dunes

fen peat

Bergschenhoek

upland peat bogs

0

50km

Figure V.1 Bergschenhoek, the Netherlands, location plotted on a palaeogeographical map (c. 4200 BC, NITG).

229

Copyright © 2010. Leiden University Press. All rights reserved.

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

The site was located on a piece of wood peat, measuring c. 16 m2 and c. 35 cm thick in the sections. Before compaction, it had presumably originally been up to 70 cm thick. The peat was embedded in slightly humic clay. The sharp and frayed edges of this peat indicated that it was not in primary position, but must have been torn off from a more extensive peat deposit. The possible source, a more extensive layer of similar peat of the same thickness, was discovered at a distance of c. 25 m to the north of the site during a prospection of the surroundings of the site by means of corings. on the surface of this ‘peat island’ two or three partly overlapping surface hearths had been built, which were subsequently silted over by a thin clay seam. The peat surface was repeatedly raised afterwards with reed bundles, planks, wooden poles and even parts of a fish trap, up to c. 30 cm in total, apparently to stabilise and reinforce the surface. A central feature at the site was a complex hearth located on top of the surface hearths, consisting of an eight or nine times repeated sequence of thin layers of spread-out reed covered by brown ‘peat mud’ and charcoal (Louwe Kooijmans 1987), each representing a new layout of the hearth. The peat island was embedded in a humic clay, built up in “a sequence of thin clay beds alternating with levels of plant remains, mainly reed”, interpreted as the result of “seasonal sedimentation in which the plant remains were laid down during winter and the clay beds in spring and summer” (Louwe Kooijmans 1987, 238). There is an upward decrease in the thickness of the levels, indicating that the water depth decreased through time. In the lower and middle part of the clay some seven levels could be made out. The upper part was more diffuse. If this microstratigraphy represents an annual rhythm, local sedimentation may have covered ten to twenty years. finds were recovered from the living surface itself and from all levels in the surrounding clay deposits. They are dominated by large quantities of fish bones and scales, some bones of birds, wild mammals and dog. objects other than plant and animal remains are scarce and comprise pottery sherds, three pieces of flint, a fragment of a polished stone adze, a fragment of a perforated antler axe and an antler axe blade, an awl made of a long bird bone and a lump of burned clay. Most spectacular are a series of wooden artefacts including a complete fish trap (see fig. 8.1), fragments of at least two others and pieces of rope. The site is interpreted as a short-term special activity camp with fowling and fishing as the main activities and some additional mammal hunting and gathering. The occupation probably lasted not more than about ten years. The short-lived aspect of the site is confirmed by the microstratigraphy of the surrounding clay. The microstratigraphy of the hearth indicates that the site was intermittently used, probably during seasonal visits in successive years. occupation in late autumn and/or winter is evident on the basis of the bird remains and macrobotanical evidence (Louwe Kooijmans 1986, 10). Clason and Brinkhuizen (1993) concluded in their zoological analysis to a distinct winter presence, with a minor summer indication. At the time of occupation, the site was located in an ecological zone dominated by eutrophic marshes that more or less had been cut off from former connections with the sea. The fish spectrum indicates fresh to slightly brackish conditions. According to the first interpretations the peat was in use while floating, especially in view of the occurrence of archaeological finds in levels below the island. Several arguments were however developed against this interpretation. firstly, the thin clay seam, covering the peat and the early surface hearths, and preceding the main phase of use, cannot be understood in this interpretation. This clay seam implies that the peat must already have been in a fixed position in this stage. The clay moreover links an early stage of site use with a late stage of sedimentation around the peat, showing that the clay deposition largely predates the occupation. Secondly, it is clear from the stratigraphical position of the major (complete) fish trap and from the occurrence of refuse material in the upper levels of the clay microstratigraphy that the site was at any rate used after the piece of peat was embedded in the clay and fixed at the documented location. These inferences, however, generate the problem of how to interpret the archaeological material in the lowest levels, like a large part of a fish trap below the ‘island’.

230

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

At last, it was argued that the floating peat would have been too soft to support and stand a number of people. however, the original thickness of the peat was before its compaction more than the observed 35 cm and the surface was raised and reinforced, weakening this objection. The sequence of events appears at any rate to be more complex than assumed earlier. In the present interpretation, the sequence starts with the exposure of an earlier wood peat situated at a lakeshore, where Swifterbant people had arranged a campsite. factually a hearth was discovered on the stretch of similar peat to the north of the site, attested during the prospection mentioned above. The artefacts in the lowest level, below the peat, should be explained as refuse that was lost or thrown and washed away at the foreshore in this stage. The margins of the peat along the shoreline would subsequently have started to float as a result of water fluctuations and wave action, and at last parts would have been torn off to drift away. It is assumed that the piece of peat drifted away over only a short distance and soon stranded in the coastal muds above the earlier waste. The first use of the patch of peat is represented by the two or three shallow surface hearths. These may date from a time when the patch was still occasionally floating, or even when it was still connected to the ‘mainland’. These hearths were silted over when the clay deposition was almost completed. It is only then that the main use of the site started, that the surface was raised with reed and wood, and the complex hearth was built up. The main consequences for this study are that there never has been a campsite on a ‘floating island’ and that the major part of the local clay deposition preceded the construction and use of the campsite. It is estimated that the sequence of events will have covered not more then a few decades. The goal of the botanical and malacological analyses is to reconstruct the natural vegetation, human impact and plant subsistence. Questions are:

Copyright © 2010. Leiden University Press. All rights reserved.

• • • • • • • •

What did the local vegetation look like during occupation? What was the strength of marine influence? how did occupation effect the local vegetation? Which food plants were available and used? do the plant remains give information on seasonality? Which wood taxa were used for various purposes such as artefacts and fuel? What are indications of the selective use of wood based on the quality of the wood? What are the indications of coppice practices?

V.2 MaterIals and Methods The Institute of Prehistory Leiden (now the faculty of Archaeology, Leiden University) and the BiologicalArchaeological Institute (now the groningen Institute of Archaeology, University of groningen) both participated in the archaeobotanical investigations. The analysis performed by the Institute of Prehistory Leiden concerns in the first place two sample boxes (50 x 15 x 10 cm) collected in 1978 for botanical analysis. W.J. Kuijper analysed the material in 1979 and 1980 with help of literature and the reference collection of the institute. Box 1 covered a section of the hearth complex (see fig. V.2 and V.3). Sampling of this box included seven pollen samples with a volume of 1 cm3, of which pollen and macroremains were investigated. The macroremains of the remaining sediment were investigated by quick-scan. Box 2 was collected 2 metres apart from box 1 in a clay layer next to the peat, which was considered to be contemporaneous with occupation (see fig. V.2 and V.3). Sampling of this box included 11 samples from which both pollen and macroremains were investigated (volume 1 cm3). Two additional macroremains samples were collected from the box as well. The first additional sample (a) was collected immediately below 8.37 m -NAP and had a volume of 25 cm3. The second additional sample (b) was collected at 8.37-8.31 m -NAP and had a volume of 75 cm3.

231

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

1

2

clay disturbed peat 0

section

2 m

Figure V.2 Bergschenhoek, site plan showing the main features and the location of the sample boxes 1 and 2 (archives National Museum of Antiquities, adapted by L. Amkreutz). 7.5

1 2 8.5

m - NAP

Copyright © 2010. Leiden University Press. All rights reserved.

8.0

0

50 cm

clay reed peat hearth wood molluscs

232

Figure V.3 Bergschenhoek, section showing the location of the sample boxes (see also figure V.2; archives National Museum of Antiquities).

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

The calculation of the pollen diagrams is based on an upland pollen sum including dryland trees and shrubs and excluding taxa that may have grown in the local vegetation. Taxa such as Chenopodiaceae and Artemisia sp. are not included in the pollen sum either since these could be part of the local (brackish) drift litter zone. for the analysis of macroremains samples were sieved at least on a 0.5 mm sieve. Ranunculus aquatilis-type represents Ranunculus section Batrachium. Veronica beccabunga-type represents V. anagallis-aquatica, V. beccabunga and V. catenata. W.J. Kuijper identified molluscs from 24 samples that partly corresponded with occupation, including eight judged samples with a volume of 0.1-5 litres and many single finds and handpicked finds. The sediment of most samples was fine, humic clay. The macroremains that showed up during the malacological analysis were identified as well, resulting in nine additional macroremains samples with a volume varying between 0.1 and 0.5 litres. The methodology of the analysis of macroremains of these samples is as described above. W.A. Casparie and I.L.M. Stuijts from the Biological-Archaeological Institute analysed wood and charcoal remains in the late seventies. This appendix will present the results from the samples collected in 1976 as well as preliminary results from the samples collected in the following year, based on information in unpublished documents and personal communication with W.A. Casparie. Identification was based on literature (Tjaden 1919; Schweingruber 1978; greguss 1956) and the reference collection of the institute. Charcoal analysis included the identification of fragments larger than 2 mm from seven samples. These samples contained also macroremains and waterlogged wood. W.A. Casparie kindly discussed the wood and charcoal data with the author in 2007. h.J. during from the Biological-Archaeological Institute analysed moss remains of 18 samples. This list has been put available by W. van Zeist to W.J. Kuijper and has already been published as a species list (Kuijper 2000).

Copyright © 2010. Leiden University Press. All rights reserved.

V.3

r esults

V.3.1 Pollen and macroremains analysis Box 1 represents a sequence of wood peat, the hearth complex and a cover of clay. A clay layer of 1 cm thick at 8.13 m -NAP formed the transition between the hearth complex and the underlying peat. The top of the hearth complex consisted of a thin reed layer (1 cm thick) with an even thinner clay layer (a few mms thick) underneath at 7.90-7.85 m -NAP. In the sediment of the hearth complex in the sample box, two additional reed layers were recognised (at 8.05 and 8.00 m -NAP) as well as a clay layer (at 8.10 m -NAP). figure V.4 shows the analysis of a single pollen sample of box 1. only the spectrum collected at 7.81 m -NAP in a clay layer contained sufficient pollen for analysis (pollen sum = 316). This sample was dominated by pollen of Quercus sp., Corylus sp., Alnus sp., Poaceae, Cyperaceae and monoletae psilatae fern spores. The herbs represent eutrophic marsh vegetation. The other pollen samples contained few pollen grains, charcoal and insect remains. figure V.5 shows the results of the analysis of macroremains of box 1 (please note the variation in the sample volume within this box). Interestingly, macroremains of many taxa had been preserved in the sequence of peat, clay, reed and charcoal. Most taxa represent waterlogged macroremains of eutrophic marsh and reed vegetation. Stem remains and fruits of Phragmites australis were frequently found, while macroremains of trees and shrubs were absent. The macroremains diagram of box 1 does not show taxa that indicate brackish conditions, but Atriplex sp. and Bolboschoenus sp./Schoenoplectus sp./Scirpus sp. could represent such taxa, while many of the other taxa additionally tolerate brackish conditions.

233

Copyright © 2010. Leiden University Press. All rights reserved.

C h A r eno t p N em o d y is ia Pe m p i a c e h ae Po d i a s a e a ac t r u ea m e C yp er A s ac e te a Tr ra e il c R et a e a u e e A s b i a , ps li g t c i u A p e r a e a e l at a l i f l o c e ra e I n i ac e e a e d Te et. ae tub ul tra if l pl or oa ae ar is ta ta (ty pe 89 )

779

783

20

234 20 40 20 40

vu

lg

ar

e Sa H li x u S p mu lu S par g s a a r Ty g niu ph an m Ly a iu em t l m R h r u at i f e r e r s um m o l e u ia c tu m Al e is x m - t yp -t y e Fi m a ac e lip pe t os U en el r ti d la Eu c a u l a Er r u m i e M c ale x on s ol et ae ,p si la ta e

Po A l ly p o nu d s ium

Ti li U a lm Fr us a C x in or u yl s us

D ep Pi t h nu (cm Be s -N AP Q tula ue ) rc us

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Upland taxa Wetland taxa

781

20 40 60 20 40

Ecologically indeterminate

Analyst: W.J. Kuijper, 1980

20

Figure V.4 Bergschenhoek, box 1, pollen, + = present, • = less than 1%.

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

-N A o P) J u gy nc Sp u h s Ty ag su p h nu b n Si a m od u s s u Ve m e p. p., lus le r r Ph o n e c av es r a i c a tum g Ph m sp. i t ra e s Ph g m au r a i te st gm s a ra Ph i te u s l i s, t s r U a g s a r a l i te m rt m u s s, i i c c t t C a e ra ar rem la d s l bo a C di u i o i au s i s ni ins la m c a t se r d a At iu m d l i s, r m a st At ipl e m r is em ca r ip x ar c u rb re s s i Ir i le p sc on m s x . u is ai s, C ps sp e ns d al e ., c a u y c B st da a rb ol e c r b on T h b os gia o r u on is ed is e c s Ly lyp ho ep s, c ed t h te e n i u m a r J u r u r is us , b o n m / c n R c u s a pal S c ar b ise um s li u h o o d s c St ex p. a str i en nis r ia s o e a ,l p d C c h sp. ea le ir s ys f r c tu M ium pa em s en l ai /S c Ly th sp ust ns ir r c o a . is pu Fi p a q s, sh us ua ca t C rem eur i c a rb ha o on /a a p rc in a r v is In oa s eu en ed se l s s is ct D ap rem C hn a r is ia in Ac t a sp s ar tel ., i la ep m h Sa uc ip m ed pia pl o, e vo st at lu ob m e la (c st m3 s )

The preservation of waterlogged macroremains in the hearth complex is remarkable and could be explained in various ways. firstly, the macroremains could have been deposited by the water during periods that the hearth was not used. Alternatively, the macroremains could have been brought up by people as part of the reed bundles. In both cases the macroremains would represent plant material that remained outside the influence of the fire unintentionally. A second hypothesis is that the material was brought up with the intention to restrict the fire. It could have been brought up as part of moist organic material that was intentionally brought up the surface to protect the soil underneath from the fire, or with water or wet material that was used to extinguish the fire. In addition to the waterlogged macroremains, several marsh taxa have been found in a carbonised state. The carbonised state of these taxa is in the first place considered to represent the burning of bundles of reed mixed with other marsh taxa, since all taxa found in a carbonised state could have been part of the local reed vegetation. These bundles were probably brought to the site to raise the surface level and/or improve the spatial lay-out of the hearth. The carbonised state could furthermore be the result of the carbonisation of macroremains deposited unintentionally by water, or intentionally by people with water or as part of the moist organic layer below the hearth (as discussed above). Another explanation, which is related to plant food, is discussed below (see discussion).

Varia

ol

th

780

Li

D

ep

th

(c m

Wetland herbs

25

785 790 795 800 805

Copyright © 2010. Leiden University Press. All rights reserved.

810 815 820 825 Peat

Wood remains

Hearth

Clay

Clay

25 1

Hear th

25

1

25 1

Hear th Hear th Hear th

25 25 1

Hear th Clay

25 1

Peat

25

Peat

25

20

1

1

Peat

Analyst: W.J. Kuijper, 1980

Figure V.5 Bergschenhoek, box 1, macroremains, + = few (1-9), ++ = several tens (10-49).

235

Copyright © 2010. Leiden University Press. All rights reserved.

Sp Sparg ar an g Ty anium Typha ium em e R p h a l at e r e r s u if u So me sp. olia ctummt x - t y ype Ly lan ac (m) u pe Lythru m eto d s Lothrum ulc ell ama U tus m s ar r a U t i c a p e da li c r ti ar u n c Ly a ia c M sim dio ulat (m) e a i us R nt h c h c a a i ( Si nu a-tya v m) u u n M m c ul p e l ga e e u r is Va nth rec s -t y pe R l e r i a a tum u a q Sa me na ua (m ) t x Eu m o a c i c a / l Fi rumus eto ar v en l i p e va s a R e x le - t si a n y R nu du ran pe s (m an nc la di ) I r i un ul s u Ve ps cul s s u Veron aud s scler i a c a Veron ca co ele tu i r is ra s ( Ly ron ca tu m) c o ic be s/ Th p a c R us sp c a .a a Po l i c e . b qu un ( t m u t r at M en um r o ) g ic aon til pa ao l la eu t yp ty et pe e s ae (m (m (m ,p ) ) ) si l a Tr ta i e M l et a o Po n o e, p l Po l y p e t a s i l a o d e, t a l y Sp p iu e e c o S p ha di am v hin h a g n u c e u l g at a a M g n u m e, a r e e os sp m s re sp. ora n m ,l ai ea gia ns v es

236 845

ol

og

ep th

20

20

Clay Humic sediment

Figure V.6 part 1. 20 40 20 20 40 20 40 60 80

(c m -N AP )

m

Upland trees and shrubs

20 40 60 20

20 40 60

fra C gm la J u di u en ts n m Ju cu m nc s a s r us p is cf . (m cus .b ) ( m ul ) bo su s (m )

Ph ra g Ph m i t Po r a g e s a C acemite us er a s tr e C a e a ali yp li (m u s er a-t ) str (m al ) ac y p is ea e ,s e te

H u Sa mu Fr lix lus a Po n g ac ula ea e

H e Ac de r V i er a b R ur o n A l s ac um nu e s ae

nu y s Pi c Be ea Q tula ue rc us Ti l U ia lm u Fr s a C x in or u yl s us

Pi

th

795

Li

D

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Wetland taxa

800

805

810

815

820

825

830

835

840

20 40 60

Wetland taxa

Wetland spore plants

20

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Open water

Ecol. indet.

Varia

C h A r en o R te m p o up i s di N p i a ac a i ea A l jas a m e t a Fo haema rit im r r a i A r am / M na a At me inif alv (m) r r e a At ipleia/L rae r Po iplex s imo p Pl tamx s . (mniu a n o p. ) m N t a g e /S u u M p h g o to n ae y a m / d Eu rio r ar Tri a s p i t i c h p. N pot hy m lo ( y a ll a gi m N m p m ou m n ) ym h g s S a p a e et p i c o a h n at u Po l v i n a e a m C tamia n sp -t y ha o a . pe Pe ra ge ta (m di s p to n s ) as . n (m D ) a tru (m C ph m ) r is ni Br ta a s t Asyoz ellap., e o Apter a, mu ph a st c ip i c a A s c e a at e d p i a e o C t e r a a e e t u b l a o, s a Er r yo c e a b u l s t s t a t o i Brica phy e li flor bla le st g ae s R as s s llac uli u i e f Ty bia cac ae lora c e Type ea eae 8 e Tepe 8 In t rap 7 d C et . l o a h Fi a r c a r i s s B o h r o a l t at a e r C ne ma em (ty a r pe Ac lc iu em ins ain 89 s ) Poari m r ains lle em n ai su n m s (m ol lu sc s? )

Brackish drift litter zone

263

333 384 255 317 321 337 307 229 299 328

20

Analyst: W.J. Kuijper, 1980

Figure V.6 Bergschenhoek, box 2, pollen and macroremains, m = macroremains s.l., + = few (1-9), ++ = several tens (10-49), part 2.

sample

a

b

a

b

Phragmites australis

11

-

Ranunculus sceleratus

2

-

taxon

taxon

marsh vegetation

marsh vegetation (cont.)

Atriplex sp.

-

1

Bolboschoenus sp./Schoenoplectus sp./Scirpus sp.

Copyright © 2010. Leiden University Press. All rights reserved.

sample

2

2

Solanum dulcamara

-

1

Carex pseudocyperus

1

2

Urtica dioica

1

-

Cladium mariscus

8

-

Lycopus europaeus

2

-

varia

Najas marina

1

-

Charcoal

+

+

fish remains

+

+

Moss remains

-

2

Phragmites australis, stem remains + = few (1-9) - = not present

+

+

Table V.1 Bergschenhoek, box 2, additional macroremains, all waterlogged. Sample a) depth: below 8.37 m -NAP, volume: 25 cm3 ; sample b) depth: 8.37 - 8.31 m -NAP, volume: 75 cm3.

237

Copyright © 2010. Leiden University Press. All rights reserved.

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

The sediment of box 2 consisted of very humic clay, mixed with fine plant remains especially at 8.37-8.32 and 8.18-8.12 m -NAP. The level at 8.37-8.32 m -NAP also contained a small bone fragment and a mollusc. figure V.6 (see previous pages) shows the results of the pollen and macroremains analysis of box 2. The preservation state of pollen of this sample box was relatively good. The identifications of the two additional macroremains samples of box 2 are given in table V.1 (see previous page). In the lower part of the diagram of box 2, the macroremains of Phragmites australis, Cladium mariscus, Juncus sp., Typha sp., Veronica beccabunga-type and Sphagnum sp. indicate the local presence of reed marsh vegetation. V. beccabunga-type probably represents V. catenata when considering the local vegetation and environmental conditions (Weeda et al. 1988, 217). Najas marina and Ruppia maritima indicate brackish conditions. The first phase is followed by a peak in the curve of Cyperaceae pollen, and fruits of C. mariscus are replaced by seeds of Juncus sp., suggesting a decrease in the water depth resulting from local clay sedimentation. The middle part of the diagram shows a decrease in Poaceae pollen and V. beccabunga-type seeds, and the increased presence of fruits of Ranunculus sceleratus, which grows on open, disturbed, very eutrophic and waterlogged soil (Weeda et al. 1985, 246). These changes probably correspond to human impact and represent the disturbance of the reed vegetation, since the values of charcoal and fish remains reach maximal values at these levels. At the same time, the composition of the herb taxa identified by pollen analysis shows a change, since the curves of Lythrum sp., Lotus pedunculatus, Urtica sp., Lysimachia sp., Mentha-type and Ranunculus sp.decrease, while the curves of Valeriana sp., Rumex acetosa-type, Eurumex and Filipendula sp. increase. These changes may indicate the development of forb vegetation in the reed, which may be related to human impact. Taxa indicative of brackish conditions are only represented by pollen identifications that may represent secondary deposition, and possibly by fruits from Atriplex sp. and Atriplex sp./Suaeda sp. The middle part of the diagram shows the presence of a seed of Samolus valerandi, which grows on moist, moderate eutrophic or brackish soil in dune valleys and fens, and can be found together with Phragmites australis, Bolboschoenus maritimus and Juncus maritimus (Van der Meijden 1996; Weeda et al. 1988, 371-2). This find represents the oldest find of the Netherlands (RAdAR version 2005) and is unique for dutch Mesolithic and Neolithic wetland sites. Microremains of Samolus sp. are interestingly also reported from the Late Neolithic site Vlaardingen (Van Regteren Altena et al. 1963a). Waterlogged macroremains of this species have also been found at the funnel Beaker site Wangels in north germany (Kroll 2001). In the upper part of the diagram the curve of Poaceae recovers, while seeds of Typha sp. are frequently present together with Veronica beccabunga-type. The curves of the herbs that decreased in the middle part of the diagram additionally increase in the upper part, while the curves of several herbs that were well represented in the middle part of the diagram decrease and disappear. This recovery indicates that the changes in the herb pollen composition in the middle part of the diagram are related to human impact, although the underlying process is not completely understood. In the uppermost sample a single Cerealia-type pollen grain is present. Again taxa indicative of brackish conditions are represented by pollen identifications only. The diagram indicates that the local vegetation consisted of eutrophic marsh vegetation. The smooth curves of the dryland trees, the scarcity of shrubs and the absence of macroremains of trees and shrubs indicate the absence of trees and shrubs in the local vegetation, although the fluctuations in the curve of Alnus sp. indicate that this taxon may have been present in the extra-local vegetation. Table V.2 shows the macroremains identifications from samples collected for analysis of molluscs. Three samples correspond to an early phase of occupation, three samples correspond to the period after occupation, and for some samples the context is unclear. The variation of taxa in these samples is large, which can be related to the relatively large volume and the potential variation in sample locations. The samples only contain waterlogged remains. Taxa indicative of eutrophic open water and marshes dominate. Taxa that are present in most samples are Urtica dioica, Cladium mariscus, Schoenoplectus tabernaemontani, Carex pseudocyperus, Atriplex sp. and Sphagnum sp. These were probably very common in the local vegetation, indicating reed and

238

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

sedge marsh vegetation at the site, although seed production and selective preservation could play a role in the representation of some of these taxa as well. In contrast to the sample boxes, fruits and cones of Alnus glutinosa are present in several of these samples. A relatively rare species is Ceratophyllum cf. submersum, which was also found at Swifterbant and Rijswijk-A4 (Van Zeist and Palfenier-Vegter 1981; paragraph 3.9.2). A single sample contemporaneous with occupation contained macroremains of potential food plants, such as Malus sylvestris, Corylus avellana and Prunus spinosa. The charcoal samples also contained hazelnut shell fragments and three stones of Prunus sp. (preservation state unknown), and additionally carbonised apples (Bakels 1991, 283 based on pers. comm. Van Zeist). These remains probably represent gathered plant food (see also discussion). Taxa that are (potentially) indicative of brackish conditions were present in samples that represent the period during and after occupation: Juncus maritimus, Bolboschoenus maritimus, Triglochin palustris and Najas marina, which all tolerate both freshwater conditions and moderate brackish conditions. The samples that are not dated also contain Aster tripolium that favours brackish conditions. In addition to the analysed samples, the excavators remember that they observed many rhizomes of Nymphaea alba and/or Nuphar lutea during excavation (especially in the early layer a that dates to the period before occupation), and many waterlogged pods of Iris pseudacorus, full of seeds (particularly in the reed that was used to raise the surface). These remains probably represent the natural vegetation, which is supported by their waterlogged state, although it cannot be excluded that the tubers and possibly the pods of I. pseudacorus had been collected. Carbonised seeds of I. pseudacorus have been found in the hearth at Bergschenhoek, at the hazendonk and at hardinxveld-giessendam Polderweg (de Kort 1998). The carbonised state suggests handling by people, but a likely function is not known. relation to occupation sample volume (litre)

during early occupation 1461

1645

0.1

0.25

sediment

after occupation

relation to occupation unknown

1229 trench 3 trench 3 core 29 23-05-78 0.1

detritus-

0.1

0.2

0.1

clay

0.5

WJKA/B 1

humic clay clayey

gyttja

0.5 clay

detritus

taxon

Copyright © 2010. Leiden University Press. All rights reserved.

Dryland vegetation Malus sylvestris

1

-

-

-

-

-

-

-

-

Corylus avellana

1

-

-

-

-

-

-

-

-

Prunus spinosa

1

-

-

-

-

-

-

-

-

Alnus glutinosa*

10

2

2

1

-

-

-

-

-

Alnus glutinosa*, cones

-

1

-

-

-

-

-

-

-

Urtica dioica

1

-

1

1

2

-

1

++

+

Moehringia trinervia

-

-

-

-

-

4

1

-

-

Persicaria maculosa

-

-

-

-

-

-

-

1

-

Carduus crispus

-

-

-

-

-

-

1

-

-

-

-

-

-

+

-

-

-

-

Persicaria lapatifolia, perigons Table V.2 part 1.

239

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

relation to occupation

during early occupation

after occupation

relation to occupation unknown

Marsh vegetation Alisma plantago-aquatica

2

2

cf. 2

-

-

-

-

1

-

Bolboschoenus maritimus

-

-

cf. +

3

-

-

-

-

-

Carex pseudocyperus

+

4

4

-

2

-

-

+

-

Carex sp.

-

-

4

-

4

-

-

-

-

Carex sp., bicarpellate

1

-

-

-

-

-

-

-

-

Carex sp., tricarpellate

+

6

-

-

-

-

-

-

-

Cladium mariscus

+++

++

++++

9

++

++

+

++

+

cf. Epilobium sp.

-

-

-

-

-

-

-

1

-

Eupatorium cannabinum

2

-

1

-

-

-

-

+

-

Iris pseudacorus

1

-

-

-

-

-

-

-

-

Lychnis flos-cuculi

-

-

1

-

-

-

-

-

-

Lycopus europaeus

-

-

-

-

-

-

1

+

1

Mentha aquatica/arvensis

1

-

1

-

1

-

-

+

-

Najas marina

2

-

4

1

-

-

-

-

-

Phalaris arundinacea

-

1

-

-

-

-

-

-

-

Phragmites australis

5

2

-

-

1

-

-

++

-

Rumex hydrolapathum

+

2

3

-

-

1

-

-

-

++

12

++

-

6

5

+

+

1

Solanum dulcamara

-

3

-

-

-

-

-

1

-

Sparganium cf. erectum

1

-

-

-

-

-

-

-

-

Stachys palustris

-

1

1

-

-

-

-

-

-

Typha sp.

2

1

2

-

-

-

-

+

-

-

-

-

-

-

-

1

-

-

cf. Sium erectum

1

2

2

-

1

-

-

-

-

hydrocotyle vulgaris

3

-

-

-

-

-

-

1

-

Ranunculus sceleratus

-

1

1

-

-

-

-

++

-

Triglochin palustris

1

-

-

-

-

-

-

-

-

Nuphar lutea

22

1

1

-

-

-

-

-

-

Nymphaea alba

2

3

1

-

-

-

-

-

-

Schoenoplectus tabernaemontani

Zannichellia palustris

Copyright © 2010. Leiden University Press. All rights reserved.

ssp. pedicellata Wetland pioneer vegetation

Open water vegetation

Table V.2 part 2.

240

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

relation to occupation

during early occupation

after occupation

relation to occupation unknown

Open water vegetation (cont.) Ceratophyllum cf. submersum

-

-

3

-

-

-

-

-

-

Chara sp.

-

+

-

-

-

-

-

-

-

Potamogeton sp.

-

-

-

-

-

-

-

+

-

Ranunculus aquatilis-type

-

-

1

-

-

-

-

-

-

Aster tripolium

-

-

-

-

-

-

2

-

-

Aster sp.

-

-

-

-

-

-

-

-

+

Juncus maritimus

-

1

-

-

-

-

-

-

-

Apiaceae

-

-

-

-

-

-

-

1

-

Atriplex sp.

-

-

6

-

1

5

1

++

+

Zannichellia sp.

-

-

-

-

-

-

-

-

1

Charcoal remains

+

+

-

-

-

+

-

+

-

fish remains

-

++

+

-

-

+

-

-

-

Moss remains

-

+

-

-

-

-

-

-

-

++

++

+

++

+

-

-

++

+

Insect remains

-

-

+

-

-

-

-

-

-

Electra crustulenta

-

+

-

-

-

-

-

-

-

Bryozoa, statoblasts

-

+

+

-

-

-

-

-

-

-

+

+

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

daphnia sp., ephippia

-

+

+

-

-

-

-

-

-

foraminiferae

-

+

-

-

-

-

-

-

-

Acari

-

1

-

-

-

-

-

-

-

Brackish drift litter zone

Ecologically indeterminate

Varia

Sphagnum sp., leaves

Cristatella mucedo, statoblasts Lophopus crystallinus,

Copyright © 2010. Leiden University Press. All rights reserved.

statoblasts

Nereis sp., mandibulae 1 3 + = few (1-9) +++ = many tens (50-99) - = not present ++ = several tens (10-49) ++++ = several hundreds (100-499) * Alnus glutinosa, a wetland species, is included in the dryland vegetation here as it was probably not part of the local marsh vegetation at Bergschenhoek. Table V.2 Bergschenhoek, macroremains from the samples collected for malacological analysis, all waterlogged, part 3.

241

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

V.3.2

Wood and charcoal analysis

V.3.2.1 Wood and charcoal from the excavation in 1976 The test excavation of 1976 revealed the southern part of the site Bergschenhoek (see fig. V.2). Charcoal and wood finds from this excavation all represent material that was present on top of the peat. Table V.3 shows the results of the analysis of 624 ml of charcoal. Charcoal of Alnus glutinosa (520 ml) dominates the assemblage while Fraxinus excelsior, Ulmus sp. and Prunus sp. are represented by small quantities. Tables V.4, V.5 and V.6 show the identifications of waterlogged wood, classified into unworked wood, worked wood (including artefacts) and wood from charcoal samples. The preservation of the waterlogged wood remains was relatively good. Alnus glutinosa (wood and bark) is dominant in all three categories, which indicates the presence of alder carr in the exploitation area of the site. other taxa are Fraxinus excelsior, Betula sp. and Salix sp. Betula sp. was only found in the assemblage of worked wood, while cf. Tilia sp. was only found as rope.

sample

29

30

40

43b

47b

47d

57

total

Alnus glutinosa

5

20

70

100

120

120

85

520

fraxinus excelsior

-

-

2

20

1

25

-

48

Ulmus sp.

-

-

4

7

2

5

5

23

Prunus-type

-

-

-

4

-

-

-

4

Indet.

-

-

2

15

1

1

10

29

taxon

- = not present

Copyright © 2010. Leiden University Press. All rights reserved.

Table V.3 Bergschenhoek, charcoal in milliliters (Casparie and Stuijts unpublished data).

sample

taxon

description

37a1

Alnus sp.

4 large fragments, partly carbonised

37a2

Alnus sp.

branch, 2 fragments, Ø 3.5 cm

38a

Alnus sp.

branch, Ø 1.3 cm

38g

Alnus sp.

branch, Ø 1.5 cm, three years

38j

Alnus sp.

small trunk, Ø 5 to 6 cm

38k

Alnus sp.

small trunk, Ø 4 to 4.5 cm

38l

Alnus sp.

small trunk, Ø 3 to 4 cm

28b*

fraxinus sp.

roundwood, Ø 5.7 cm

37b

Salix sp.

7 branches, Ø 2.0 to 5.5 cm

Ø = diameter Table V.4 Bergschenhoek, unworked wood used to reinforce the site surface (Casparie and Stuijts unpublished data).

242

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

sample

taxon

interpretation and description

28a

Alnus sp.

plank, 13 x 3.2 cm

28d

Alnus sp.

plank, 6 x 2.6 cm

38c **

Alnus sp.

small beam

38d **

Alnus sp.

small beam, 5 fragments

38e **

Alnus sp.

haft?, 4 fragments

38f **

Alnus sp.

beam or post, 3 fragments, pointed?

38h

Alnus sp.

2 wedges?, 4.1 x 1.8 cm

38i

Alnus sp.

pointed post, Ø 4.5 cm

38n **

Alnus sp.

plank, Ø 2 x 7.5 cm

38o **

Alnus sp.

plank, 2 x 8 cm

38p **

Alnus sp.

plank, 2.2 x 8.5 cm

39

Alnus sp.

plank, 1.8 x 12 cm

28c

Betula sp.

pointed post, Ø 4 to 4.5 cm

44

Betula sp.

twig with rope from bark of cf. Tilia

38b

fraxinus sp.

haft?, Ø c. 3 cm

38s*

fraxinus sp.

post, Ø 4.2 cm

Ø = diameter * and ** = wood in different samples is similar and may be prepared (in the past) from a single tree

Copyright © 2010. Leiden University Press. All rights reserved.

Table V.5 Bergschenhoek, worked wood including artefacts (Casparie and Stuijts unpublished data).

sample

taxon

quantity

29

Alnus sp., bark

2 ml

30

Alnus sp., bark

15 ml

47b

Alnus sp., bark

2 ml

47d

Alnus sp., bark

10 ml

40

fraxinus sp.

1 small fragment

Table V.6 Bergschenhoek, wood from charcoal samples, see table V.3 for the corresponding sample numbers (Casparie and Stuijts unpublished data). ml = milliliter.

243

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Copyright © 2010. Leiden University Press. All rights reserved.

Wooden artefacts are beams of Alnus sp., two pointed posts of Alnus sp. and Fraxinus sp., two possible hafts of Alnus sp. and Fraxinus sp., two possible wedges of Alnus sp. and a twig of Betula sp. winded with rope of bark of cf. Tilia sp. (see photograph in Louwe Kooijmans 1977a). Unpublished drawings show that the finds comprised much more rope. Several worked remains of Alnus glutinosa may originate from a single tree (Casparie s.a., see table V.5). The wedges may represent the waste of wood working (cf. Louwe Kooijmans and Kooistra 2006, 240-242). The analysis of the wood resulted in observations on growing conditions and seasonality. Two fragments of Fraxinus sp. that may originate from the same tree both showed continuous growth that was very slow, especially during the last twenty years (28b and 38s). Bad growth conditions could be related to rising water tables, brackish conditions or competition amongst others. five pieces of wood lacked wood formed during summer, indicating clearance during spring (28c, 38g, 38j, 38k and 38l). It is likely that all wood excavated in 1976 was intentionally gathered and does not represent the local vegetation, since the results indicate that the wood remains represented trunks, branches and artefacts only, while small twigs, fruits and leaves of the attested taxa were very scarce. The bark remains could represent waste of wood working. This is supported by the scarcity of macroremains of Alnus glutinosa and the limited percentage of alder pollen, which would probably have been present in larger quantities if alder had been part of the local vegetation during occupation. V.3.2.2 Wood from the excavation in 1978 Unpublished and published texts give some additional information on the further results of wood and charcoal identifications from Bergschenhoek. The finds of this excavation were collected from the peat and from the surrounding clay sediments. Unpublished notes by Louwe Kooijmans (primarily based on research by Casparie and Stuijts) mention Alnus sp., Salix sp., Fraxinus sp., Betula sp., Tilia sp., Ulmus sp., Prunus spinosa and Malus sylvestris as the attested wood species. Charcoal identifications are possibly included in this list. An unpublished manuscript (Casparie s.a.) provides information on materials and methods of the wood and charcoal analysis. The document shows that the results of the excavation in 1976 are not representative of the complete data set. Casparie discusses identification and possible confusion between the wood of Cornus sp. and Viburnum sp., Alnus sp. and Corylus sp. and Salix sp., Populus sp. and Salix sp., Malus-type and Sorbus-type, and Prunus avium-type, Prunus padus-type and Prunus spinosa-type. It is furthermore mentioned that it was very difficult to identify small branches of Sambucus sp. with certainty. Although the text presents methodology only and no results, it suggests identification and presence of all taxa that are mentioned in the discussion, with exception of the three Prunus types. The material interestingly included roots from Malus-type and/or Prunus-type. It is moreover noted that branches of Malus-type (including Sorbus sp.) and Prunus-type showed traumatic tissues (irregularly grouped cells formed in the region of wounds) that were interpreted as being indicative of the repeated removal of young branches (see also discussion). The unpublished manuscript informs us on the charcoal analysis and mentions a volume of 1.5 litres corresponding with 1000 fragments. It is concluded that the species list resulting from charcoal identifications is shorter than that of the wood identifications, and that this is probably caused by the specific use of wood for fuel and not by the number of charcoal samples or identification problems. Corylus sp., Betula sp., Viscum sp. and Sambucus sp. “for example” were lacking in the charcoal assemblage, which suggests that these species were identified during wood analysis.

244

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Copyright © 2010. Leiden University Press. All rights reserved.

Comparison of all available data on wood identifications indicates the identification of Alnus glutinosa, Betula sp., Cornus sanguinea, Corylus avellana, Fraxinus excelsior, Malus-type, Populus sp., Prunus spinosatype, possibly other Prunus types, Rosaceae, Salix sp., possibly Sambucus nigra, Sorbus-type, cf. Tilia sp., Ulmus sp., Viburnum opulus and Viscum album. The available information does not allow making separate lists of identifications of waterlogged wood and charcoal. The charcoal assemblage at least contained Alnus glutinosa, Fraxinus excelsior, Prunus sp. and Ulmus sp., and did not contain Betula sp., Corylus sp., Sambucus sp. and Viscum sp. The identification of waterlogged wood of Viscum album is remarkable since this species is only known from charcoal identifications from several Mesolithic and Early and Middle Neolithic dutch wetland sites, but not from identifications of waterlogged wood (see chapter 7). V.3.2.3 Other wood data Publications and site maps inform that wood remains, including the remains of a dug-out canoe made of Alnus wood42 and the remains of several fish traps, had been laid down on the peat in two directions, at right angles to each other, in order to stabilise and raise the surface level. This function of the wood remains explains the presence of planks, pointed posts, roundwood, beams and unworked trunks and branches at the site. The presence of one of the fish traps in horizontal position in between two sub-layers of the hearth supports intentional deposition of this trap for stabilising the surface as well. for the other traps, the reason of deposition is less clear. Some of the remains together formed a platform (Louwe Kooijmans 1986, 10, 1987, 238, see also fig. V.2 and fig. 8.6). Several posts were found in vertical position, and may represent the supporting elements of the platform, or of another structure such as a hut. The common presence of wood at the site indicates that wood was sufficiently available in the exploitation area. The use of wood representing former artefacts for the improvement of the surface level however indicates that wood was not thrown away but was re-used if possible. The most spectacular wood finds of Bergschenhoek are the remains of several fish traps. At least four fish traps are represented, and many scattered fragments suggest that the occupants of the site used even more fish traps. The fish traps date to various stages. one was found in the clay below the peat, one was found at the edge of the peat, one was found in the hearth and one was found elsewhere in the clay. Two fish traps contained funnels while one represents a funnel only. The largest fish trap, made of c. 200 withies kept together with two consolidating withies (hoops) and rope, has a length of 1.7 metres and at the opening a width of 0.6 metres. Unpublished documentation shows that the consolidating withies were pointed. The investigation of the fish traps included hundreds of identifications and thousands of observations of wood under the microscope for the determination of the age of the wood. The sample size was this large to get representative results on the taxa used for the fish traps. It is probable that all fish traps were investigated, although the number of identifications per separate fish trap is not known. All fish traps were made of unsplitted twigs of Cornus cf. sanguinea (dogwood) that had a length up to 2 metres and an age of 1 year, while some twigs had an age of 2 years. Unpublished information suggests that some elements of a trap may have been made of cf. Sambucus sp. (Casparie s.a.), but after further identification it has been concluded that all withies represent C. cf. sanguinea (pers. comm. Casparie 2007). There is no information on the identification of consolidating withies. The rope used as part of the fish traps has been identified as bark of cf. Tilia sp. (not Juncus sp. or Scirpus sp. as mentioned in other publications). In addition to the fish traps, two or three deformed, intensively coppiced stools (root systems) of dogwood were found at the site (pers. comm. Casparie 2007; precise location unknown). These stumps indicate that shrubs of Cornus sanguinea were growing within a few km distance from the site, and that they were cleared of their shoots at or near the site, indicating that some of the fish traps were manufactured locally.

42 The interpretation of several planks as a canoe is based on the shape and curve of the planks, the wood species, and the fine working of the wood (pers. comm. Louwe Kooijmans 2007).

245

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

other excavated and published artefacts are two arrow shafts with an incision at the end and double pointed hooked sticks interpreted as leister prongs (wood identifications unknown). The presumed leister prongs have a length of c. 40 cm and are comparable with those known from denmark and northern germany dating to the Neolithic (Meurers-Balke 1981). A comparable wooden artefact made from Fraxinus excelsior is known from the Middle Neolithic site ypenburg (Kooistra 2008). The danish and german finds are reported to be made of Corylus sp. and Pomoideae dominantly and additionally of Betula sp., Cornus sanguinea, Crataegus sp., Fraxinus excelsior, Malus sp., Pyrus sp., Sorbus sp., Quercus sp., Ulmus sp. and Viburnum sp. (Meurers-Balke 1981, 135; Schmölcke et al. 2006; Skaarup 1980, 7).

Copyright © 2010. Leiden University Press. All rights reserved.

V.3.3 moss analysis The data and interpretation of the mosses are based on an unpublished manuscript by during.43 Table V.7 shows the identifications of mosses from 18 samples. The richest samples were excavated in 1976 and probably derived from the southern part of the site. The analysis resulted in the identification of many epiphytes, including taxa that could have lived on Fraxinus sp., Ulmus sp. and Salix sp., taxa living on tree stumps, taxa indicative of disturbed environments, and taxa living in marshes and reed vegetation. The species indicate the presence of marsh vegetation and carr or softwood alluvial woodland. A taxon that is mentioned explicitly in the document of during is Hylocomium brevirostre, which is nowadays a rare species of dune vegetation but may have occurred in fens in the past. There is no information on quantities except that relatively large quantities were found of Anomodon viticulosus, Neckera sp. (not N. crispa) and Hylocomium brevirostre. These large quantities may represent the intentional collection of these taxa at Bergschenhoek. Ten of the moss species found at Bergschenhoek were also found at Swifterbant (Van Zeist and Palfenier-Vegter 1981). This correspondence may be the result of similar natural vegetation in both regions, similar site formation processes and/or similar use of mosses. V.3.4 mollusc analysis Table V.8 shows the results of the malacological analysis (see also the macroremains from the mollusc samples for further information on the contents of the samples). The samples, partly corresponding with the periods during and after occupation, showed the presence of 13 species, containing three species that prefer brackish conditions and ten species that prefer freshwater conditions. The number of freshwater species decreased through time. Remains of Electra crustulenta, frequently found on mollusc species that live in brackish water, were interestingly found growing on the freshwater species Anodonta anatina. The size and conditions of the molluscs as well as the ecological preferences of the taxa indicate that all taxa probably represent the local mollusc fauna. There are no explicit indications that the taxa were handled or gathered by people. The species indicate the presence of a riparian zone at the edge of a calm body of open water where tidal influence was small or absent. The taxa furthermore indicate that the water was slightly brackish as the result of scarce marine influxes. Two interpretations are possible. firstly, the water may have been slightly brackish continuously, resulting in the combined presence of taxa indicative of freshwater and brackish water. The transition between fresh and brackish water should have been very gradual to explain the presence of both fresh water and brackish water taxa. Secondly, the brackish influxes may have varied periodically (over a period of several years), and as a result indicators of freshwater and brackish water did not live at the site contemporaneously. The second scenario is considered as the most likely explanation.

43 Unpublished manuscript by H.J. During (1980), attached to a personal letter from W. van Zeist to W.J. Kuijper, 1 October 1998.

246

-

dicranoweisia cirrata

homalia trichomanoides

homalothecium sericeum

hylocomium brevirostre

hypnum andoi

Isothecium myosuroides

Kindbergia praelonga

Leptodictyum riparium

Leucodon sciuroides

Neckera complanata

Neckera crispa

Neckera pumila

Platyhypnidium riparioides

Sphagnum palustre

Ulota sp.

-

-

-

-

+

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

+

-

-

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

Table V.7 Bergschenhoek, mosses (During unpublished data).

- = not present

+ = present

-

Brachythecium velutinum

cf. rutabulum

-

-

Aulacomnium palustre

Brachythecium

+

Anomodon viticulosus

taxon

-

-

-

-

-

+

-

-

-

-

+

+

-

-

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

-

-

-

-

-

-

sample 1552 1566 1621 387 1223 1382 244

Copyright © 2010. Leiden University Press. All rights reserved.

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

+

-

279

-

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

271

-

-

-

-

-

-

-

-

+

+

-

+

-

-

-

-

-

-

-

296

-

+

-

-

-

-

-

-

+

+

+

-

-

-

-

+

-

-

-

42

+

-

+

-

-

-

-

-

+

+

+

-

+

-

+

+

-

-

-

-

+

-

-

-

+

-

-

-

+

+

-

+

-

-

-

+

-

-

-

-

-

-

-

+

-

+

-

+

+

-

+

-

-

-

+

-

-

47A 47B 47C

-

+

-

-

-

-

-

-

+

+

+

-

+

-

-

+

+

-

-

48

-

+

-

-

-

+

-

+

+

+

+

-

-

-

-

+

+

-

-

49

-

-

-

+

-

+

-

-

-

+

+

-

+

+

-

-

-

-

+

50

-

-

-

-

-

+

+

+

-

+

+

-

+

-

-

+

-

-

-

54

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

247

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

-

-

++ 29 -

8 2 -

Fresh water Unio pictorum Anondonta anatina Bithynia leachii Bithynia tentaculata Planorbarius corneus Planorbis planorbis Radix peregra Stagnicola palustris Theodoxus fluviatilis Valvata piscinalis

1 1 -

1 1 -

1 1 -

1 ++ 1

1 2 ++ 1 2

248

11 5, t 61 op fis h 12 tra 35 p

nc h

91 3

-

91 2

Brackish water Cerastoderma glaucum hydrobia ventrosa Scrobicularia plana

51 9

sf

28 5

16 45

sf

Table V.8 part 1.

during occupation

14 61

68 4 12 29

19 25

11 31

sf

0.1 sf

taxon

Copyright © 2010. Leiden University Press. All rights reserved.

sample volume (litre) sample type

during early occupation

sf

sf

sf

sf

sf

sf

sf

15 ++ -

33 -

-

-

-

++ 5 -

19 1 -

6 1

1 ++ -

1 3 1 10 -

1 -

1 -

++ -

3 1 -

-

-

0.1 sf

5

tre

relation to occupation

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Copyright © 2010. Leiden University Press. All rights reserved.

taxon

sf

0.25 0.1 sf

sf

sf

sf

sf

0.1

sf

JK B

W

JK A W

tre

tre

nc h

3

2

nc h

tre

nc h

tre

nc h tre

sample volume (litre) sample type

relation to occupation unknown 2, 16 nc -0 h 2, 5-19 tre 05 78 nc -0 h 5 -1 2, 97 tre 26 8 nc /3 h 02 2 04 co /2 re 08 29 W JK

after occupation

2

relation to occupation

0.5

0.5

Brackish water Cerastoderma glaucum hydrobia ventrosa Scrobicularia plana

1 1

2 26 1

21 36 3

1 -

-

12 -

-

20 ++ -

++ ++ 1

1 6 -

7 ++ -

Fresh water Unio pictorum Anondonta anatina Bithynia leachii Bithynia tentaculata Planorbarius corneus Planorbis planorbis Radix peregra Stagnicola palustris Theodoxus fluviatilis Valvata piscinalis

-

1 -

1 -

2 -

2 -

-

2 1 -

2 -

4 -

1 -

1 -

sf = single finds and hand-picked finds ++ = tens - = not present Table V.8 Bergschenhoek, molluscs (Kuijper unpublished data), part 2.

249

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Copyright © 2010. Leiden University Press. All rights reserved.

V.4

dIscussIon

V.4.1 r econstruction of the natural Vegetation The reconstruction of the natural vegetation is based on all presented data including those from the sample boxes. The results from pollen and macroremains analysis from both boxes correspond very well with each other and indicate that Bergschenhoek was located in the middle of eutrophic reed marsh vegetation. The moss analysis partly confirms this interpretation. The pollen and macroremains data strongly indicate that trees were absent in the local vegetation, since marsh taxa dominate and macroremains from trees and shrubs are very scarce. only Alnus glutinosa may have been present in the environs, since macroremains were found and since this taxon shows minor fluctuations in the pollen diagram of box 2. The macroremains of alder may have been transported by water or brought into the site with wood. The data from the wood and charcoal and also some mosses give a quite different view on the vegetation than the pollen and macroremains since the results strongly indicate the presence of alder carr and alluvial woodland vegetation. The data from the moss analysis partly confirm this reconstruction of the natural vegetation as well. The most probable explanation for this contrast between the data sets of pollen, macroremains, wood and charcoal is that the pollen and macroremains primarily represent the local and extra-local vegetation while the wood was gathered at a larger distance. The mosses indicative of woodland would primarily have been brought in together with the wood or by the intentional gathering of the moss itself. Alder, dominant in the wood and charcoal, was certainly the most common of all trees in the marshes that surrounded the site. The species probably did not grow at the excavated patch of peat itself, but may have been present at less than a km distance. Salix sp. may also commonly have been present in the marshes. Taxa such as Cornus sanguinea, Malus-type, Populus sp., Prunus sp., Rosaceae, Sambucus nigra, Sorbus-type and Viburnum opulus may have been present in patches of well-developed alder carr. They may also have grown in alluvial woodland vegetation on higher patches in the landscape, together with Corylus avellana, Fraxinus excelsior and Ulmus sp. The large variety of these taxa indicates that alluvial woodland vegetation was present within the exploitation area of Bergschenhoek, i.e. at a distance of c. 5 km (see paragraph 1.3). Locations where such vegetation may have grown were present are the dunes at hillegersberg and the central station of Rotterdam at c. 2 km distance from Bergschenhoek (guiran and Brinkkemper 2007; see also chapter 6). Tilia sp. may also have been present on the higher parts of these dunes. The absence of acorns and wood of Quercus sp. at Bergschenhoek is however remarkable since oaks are expected to have grown on the dunes as well (e.g. Voorrips 1964; see also chapter 2). A single exception to the indications of the absence of wood at the site itself are the finds of roots of Malus-type and/or Prunus-type, indicating the local presence of a species that belongs to these types, assuming that the collection and import of roots of such a species is unlikely. Local presence is tentatively confirmed by the presence of Rosaceae pollen. A possible candidate is Sorbus aucuparia, which is able to grow in reed marsh at patches with a relatively solid, dry surface and on dead wood (Weeda et al. 1987, 92). Scarce finds of macroremains of M. sylvestris and P. spinosa tentatively suggest the local presence of such trees, but these macroremains may also have been gathered elsewhere for consumption. Local presence of M. sylvestris and P. spinosa is furthermore less likely since the local environment was probably too wet for these taxa (in contrast to what was suggested in an earlier publication). The habitat and possible local presence of Cornus sanguinea needs special attention (see discussion above and below). C. sanguinea may have grown in the river flood plains, woodland edges and woodland clearances (Weeda et al. 1987) and was probably not uncommon in the coastal region and river area (out 2008b). It is however highly improbable that C. sanguinea grew at the site itself since the local environment was too wet. The unexpected presence of intensively coppiced stools of C. sanguinea at Bergschenhoek has therefore resulted in the conclusion that the shrubs were transported to the site from a nearby location.

250

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

The results of the archaeobotanical and malacological analyses indicate that the site was located in a riparian zone at the edge of a calm body of open water. The molluscs show a combination of freshwater and brackish water taxa, with a dominance of freshwater taxa, and indicate that the strength of brackish influxes may have varied periodically over a period of several years. The macroremains assemblage indicates slightly brackish conditions, since taxa that do not tolerate brackish conditions (e.g. Caltha palustris) are absent, since many attested freshwater taxa tolerate minor brackish conditions, and since most taxa that are indicative of brackish conditions also tolerate freshwater conditions. The brackish conditions must have been so moderate that alder carr could have survived, possibly on slightly elevated patches in the landscape. Macroremains of true halophilous plant taxa were probably part of the drift litter zone, where macroremains of local and nonlocal taxa are deposited during high water. This could especially concern fruits of Ruppia maritima, since the obligate preference of this species for brackish conditions does not correspond with the further results on salinity. organic material in the drift litter zone furthermore may have represented an excellent habitat for species like Chenopodiaeceae (C. glaucum and C. rubrum) and Atriplex sp. Brackish influxes occurred at least during and after occupation, as indicated by both the botanical remains and the molluscs. The location of Bergschenhoek in between the tidal flats and freshwater marshes implies marine influxes from an estuary located at some distance in western direction.

Copyright © 2010. Leiden University Press. All rights reserved.

V.4.2

human imPact

V.4.2.1 Human impact on the vegetation and deposition processes The most direct aspects concerning human impact on the vegetation at Bergschenhoek are shown by the raising of the surface with wood and bundles of reed. The available data on wood (including drawings of the site) strongly support the intentional deposition for the improvement of the surface. The idea of intentional deposition of reed bundles, such as the large quantities and the uniform direction within single layers, is mainly based on observations during excavation (Louwe Kooijmans 1977a). In earlier publications the site Bergschenhoek has been compared with Swifterbant-S3, and the evidence and interpretation for raising the surface with reed is highly comparable between both sites (see also Van Zeist and Palfenier-Vegter 1981, 139). The two sample boxes offer a unique possibility to study human impact. Box 1 represents a hearth complex on top of the peat. The results show that macroremains remained preserved in the hearth both in a carbonised and waterlogged state. despite indications of intermittent occupation, the investigated spectra of box 2, sampled two metres further away and next to the peat, do not show clear separate occupation phases. There are indications of anthropogenic reduction of the reed vegetation and the development of forb vegetation, while a decrease in alder may be related to human impact as well. The diagram of box 2 does not show other indications of deforestation since woodland was not present in the (extra-) local vegetation. Taxa of which macroremains were found in a carbonised state are Phragmites australis, Cladium mariscus, Atriplex sp., Iris pseudacorus, Calystegia sepium, Bolboschoenus sp./Schoenoplectus sp./ Scirpus sp. (all in the hearth) and Malus sylvestris (location and precise context unknown). fruits of P. australis (reed) were found in all hearth samples, and stem fragments of Phragmites australis were additionally found as well. Their carbonised state is in the first place considered to represent the carbonisation of bundles of reed mixed with other plants from reed vegetation (discussed above). In addition to this first interpretation, it can be added that some macroremains may represent remains of food preparation in the hearth (discussed below). The taxa found in a carbonised state in the hearth at Bergschenhoek are not known from hearths at other Late Mesolithic and Early and Middle Neolithic dutch wetland sites, except for Cladium mariscus (see chapter 9). Taxa that are most commonly found in a carbonised state in hearths at dutch wetland sites are Corylus avellana, Quercus sp., Galium aparine and Cladium mariscus, which are hardly found at Bergschenhoek.

251

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

Copyright © 2010. Leiden University Press. All rights reserved.

The exceptional selection of taxa found carbonised in the hearth suggests that the carbonisation is the result of very specific deposition processes, which could be related to the construction and use of the hearth, as well as the natural vegetation around the hearth. hearths at e.g. hoge Vaart contained carbonised macroremains of many other marsh taxa (see chapter 5). Comparison of carbonised macroremains from hearths from dutch Late Mesolithic and Early and Middle Neolithic wetland sites thus indicates that the available data sets on botanical macroremains are not representative yet for the variety of processes resulting in the deposition of plant material in hearths. V.4.2.2 Plant subsistence The number of macroremains from potential food plants found in a carbonised state at Bergschenhoek is remarkably small, leaving very little indications of consumption (see the following paragraph). Several explanations are possible, such as the influence of the sampling strategy, the number of samples, relevant use and preparation processes and locations, the function of the hearth and the recognition of plant food. The number of artefacts (other than wooden artefacts) is however relatively small as well, which indicates that the scarcity of plant food may be a true result of the activities at the site. This can be explained by the function of the site as a short-term hunting/fishing/fowling camp, where the collection of plant food did not play an important role, and by the seasonal human presence, since the site was occupied in autumn/winter (discussed below) when plant food was rather scarcely available in the natural vegetation. Storage of plant food presumably did not take place either at special activity sites as Bergschenhoek. Most macroremains found in a carbonised state in the hearth may represent food plants. This concerns Atriplex sp. (edible fruits and leaves), Bolboschoenus maritimus (edible rhizomes and possibly fruits; Kubiak-Martens 2006), Cladium mariscus (edible fruits; Mears and hillman 2007), Phragmites australis (edible rhizomes) and Schoenoplectus sp. (possible edible rhizomes and fruits; Mears and hillman 2007). however, it can be questioned whether any edible parts were left of these plants in autumn and winter, especially concerning the leaves of Atriplex sp. on the one hand the carbonised finds in a context of possible food preparation support that they were used as food. on the other hand, the evidence of consumption of these taxa is not very strong, since they also could represent the natural vegetation used for raising the surface. In addition to the carbonised finds of potential food plants, some of the waterlogged macroremains found at the site may represent food plants as well: macroremains of Corylus avellana, Malus sylvestris and Prunus spinosa, rhizomes of Nymphaea alba and/or Nuphar lutea, and several edible herbs (e.g. Persicaria maculosa). The evidence of consumption of these taxa is however restricted in view of the restricted evidence of handling by people. only the consumption of M. sylvestris is supported by the finds of carbonised apples, and additionally the consumption of nuts of C. avellana since hazel was not part of the local vegetation and must have been brought to the site. The wood identifications furthermore indicate the availability of fruits of Cornus sanguinea (edibility discussed in chapter 9) and Rosaceae. Local crop cultivation at the peat island Bergschenhoek was impossible in any case when considering the size of the site and the environmental conditions. Crop plants are represented at Bergschenhoek by a single pollen grain of Cerealia-type only. Its meaning remains unclear since there is no confirmation from the macroremains of cereals and since this single grain may represent one of the local grass species. Although no cereal grains were found, there is however no absolute evidence of the absence of macroremains, because systematic botanical sampling was not included in the excavation program. The absence of crop plants at Bergschenhoek is usually related to site function. The absence could however also be related to the absence of crop plants in the Swifterbant culture, but only if occupation took place at the beginning of the period 4350-4050 BC (see also chapter 11).

252

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

The reconstruction of the natural vegetation strongly indicates that wood must have been scarce in the direct vicinity of the site, and that the collection of wood must have taken place at some km distance from the site, where wood was not scarce. There are no indications that wood was imported from other macroregions over a large distance. A part of the mosses was probably collected at some km distance from the site as well. The moss data suggest the intentional collection of Anomodon viticulosus, Neckera sp. (not N. crispa) and Hylocomium brevirostre. The artefacts and worked wood show that people used Alnus glutinosa, Betula sp., Cornus sanguinea and Fraxinus excelsior for the manufacture of artefacts. Most artefacts and worked wood concern alder. Although the small number of artefacts does not enable firm conclusions, this result suggests that people did not select their wood for manufacturing artefacts based on the quality of the wood and the function of the artefacts but used primarily the species most commonly available. This would mean that the choice for alder to make a dugout canoe does not necessarily represent selective choice either. The fish traps are a major exception since these offer excellent evidence that people selectively used twigs of Cornus sanguinea to make fish traps in a very skilful way, which moreover corresponds with other fish traps from the same period and region (out 2008b). The charcoal assemblage demonstrates the use of Alnus glutinosa, Fraxinus excelsior, Prunus sp. and Ulmus sp. for fuel, while other taxa were possibly used as well. The data do not indicate the selection of wood for fuel based on the qualities of the wood. It is instead more likely that the assemblage represents the availability of taxa in the exploitation area. The investigation of the wood from Bergschenhoek resulted in the conclusion that shrubs of Cornus sanguinea were coppiced on a large scale and during a considerable period (Casparie 1995, 212-213). W.A. Casparie and I.L.M. Stuijts based this conclusion on the investigation of wood remains from Bergschenhoek on the one hand and research on the coppicing of dogwood at modern-day plant nurseries on the other hand. The aim here is to present the underlying arguments and to make them as clear as possible (based on pers. comm. Casparie 2007, 2008). • • •

Copyright © 2010. Leiden University Press. All rights reserved.

•

for the fish traps hundreds of twigs of dogwood were used that mostly had an age of one year; some two years. The unnatural, exceptional length of the twigs (up to c. 2 metres) indicates that people had developed excellent coppicing techniques. The large numbers of twigs and the short live cycle of fish traps (three to four years) suggest that people used many thousands of twigs. The cut-off stumps found at the site were completely deformed, suggesting long-lasting coppice practices.

The data set strongly supports coppicing because of the large number of twigs with a similar age, the large length of the twigs and the finds of deformed stumps. The data set is moreover exceptionally representative because of the large number of identifications and age observations. Natural processes (drifting of ice sheets) can result in ‘natural coppicing’ as well, but it is hardly possible that this would have resulted in the availability of such good material. Wood of Malus-type and Prunus-type, not used for fish traps, furthermore showed traumatic tissues (irregularly grouped cells formed in the region of wounds), also suggesting coppicing. hard evidence of coppicing practices in prehistory is very difficult to obtain (see chapter 8). The indications of coppicing at Bergschenhoek are however relatively strong, and represent the best of all indications of the coppicing of wood found at dutch Mesolithic and Early and Middle Neolithic wetland sites, especially thanks to the work by W.A. Casparie and I.L.M. Stuijts.

253

APPENdIx V - ARChAEoBoTANy of BERgSChENhoEK, ThE NEThERLANdS

In addition to coppicing, it has been proposed that people may even have cultivated dogwood (pers. comm. Casparie 2008). The assumed relatively low density of shrubs of Cornus sanguinea in the natural vegetation (not to confuse with common presence in the region) has resulted to the expectation that people would have relocated shrubs and put them together, perhaps as a kind of osier-beds, somewhere on the most suitable soils in the vicinity of Bergschenhoek in order to facilitate coppicing activities and the collection of the twigs, thus guaranteeing a sufficient supply of rods and reducing travelling time (Casparie 1995, 213). It may also be argued that the indications of the extraordinary well-developed coppicing technique imply cultivation. dogwood is furthermore considered as a species that is highly suitable for pruning, coppicing, relocation and cultivation. for example, an osier bed consisting of c. 100 shrubs in an area of 50 x 50 metres would have been sufficient to produce 2000 suitable withies a year (i.e. 50% of the total of branches grown) that could be used for the manufacture of c. ten fish traps a year (see also the discussion on cultivation in Casparie 1986). The possibilities to prove or reject such a cultivation option and calculations are restricted, since it is not possible to reconstruct the origin of individual shrubs on such a detailed scale. The scenario therefore remains an option. V.4.3 seasonality The carbonised state of some macroremains provides information on seasonality. Macroremains of Calystegia sepium and Iris pseudacorus are present between July-September and September-october respectively, indicating autumn. The carbonised finds of Malus sylvestris support the indications of autumn, although collection and storage of apples cannot be excluded. Atriplex sp. and Phragmites australis flower in late summer and autumn, and fruits can thus be expected in autumn as well. Cladium mariscus flowers in June and July and may indicate late summer or early autumn. All taxa may have been deposited before getting carbonised. Taxa that flower or produce macroremains in spring and early summer are however clearly absent. It can thus be concluded that the new information on seasonality from the macroremains indicates autumn or winter occupation. A small number of wood remains indicate clearance during spring. The indications of coppicing and local manufacture of fish traps imply activity during autumn or winter since this is the best period for the collection of twigs. overall the new archaeobotanical results confirm the earlier conclusions on seasonal late autumn and winter occupation.

Copyright © 2010. Leiden University Press. All rights reserved.

V.5 acknowledgeMents This appendix is dedicated to Wil Casparie (†) who was actively involved in the investigations on the site Bergschenhoek and gave much support during the writing of this appendix. The author would like to thank I.L.M. Stuijts, L.B.M. Verhart, L.P. Louwe Kooijmans, C.C. Bakels, W.J. Kuijper and C.E. Vermeeren for permission to use the available data and for information, suggestions and discussion.

254

Appendix VI. Archaeobotany of the Late Neolithic site Vlaardingen, the Netherlands VI.1 IntroductIon The site Vlaardingen was excavated between 1959 and 1964 by the Institute for Prae- and Protohistory (University of Amsterdam, now the Amsterdam Archaeological Centre). Archeological results have been published in Van Regteren Altena et al. (1962, 1963a, b) and Van Beek (1990). The site, located on a levee along a channel, was occupied at 3350-1950 BC, corresponding with the Vlaardingen group and Bell Beaker culture. Most finds correspond with occupation by the Vlaardingen group, dating to 3350-2550 BC (Van Beek 1990, 249). This appendix presents the results of botanical macroremains identifications of the site Vlaardingen. The text is based on a manuscript from Prof. dr. W. van Zeist sent to Dr. B.L. van Beek.44 Prof. dr. W. van Zeist kindly gave permission to publish the results. The interpretation of these finds has been discussed already in Van Beek (1990, 210-211), but the original data set was not included there. The data represent unique results from the site Vlaardingen. The new information is presented in context of a short discussion on earlier published data on pollen, macroremains, wood and charcoal. VI.2 MaterIals and Methods Researchers of the Biological-Archaeological Institute (now the Groningen Institute of Archaeology) were involved in the sampling and investigation of the botanical macroremains from Vlaardingen. In 1964, J.A. Bakker and W.A. Casparie collected three samples for analysis of botanical macroremains. W. van Zeist identified the material. The samples were collected near the channel, probably at the southern side of the excavated area. Sample 1 was collected from refuse layer Vlaardingen 1b in the channel (see fig. 10 in Van Regteren Altena et al. 1962). At the sample location, a concentration of finds was present. Sample 2 was collected next to a concentration of finds located 30 cm below refuse layer Vlaardingen 1b. Sample 3 was collected from layer Vlaardingen 1a, representing a drift litter deposit. All samples can be related to the Vlaardingen occupation period. The volume of the samples and sample preparation methods are unknown. The presence of seeds of Juncus sp. indicates that sieves with a fine mesh width were used. Plant names are according to Van der Meijden (1996). Brassica sp. probably represents B. rapa, since this is the only identified species of this genus for this region and period.

Copyright © 2010. Leiden University Press. All rights reserved.

VI.3

r esults

VI.3.1 MacroreMaIns Table VI.1 shows the macroremains found at Vlaardingen. The taxa are grouped according to ecology. The number of samples is small and as a result one can only make tentative conclusions. The samples only contained waterlogged remains. Most taxa may represent natural vegetation present near the site, although it cannot be excluded that some macroremains were gathered for use or consumption. In the original manuscript, Van Zeist tentatively concluded that the environment consisted of dryland vegetation, alder carr, marsh vegetation, humid grasslands and open water. The environment was mainly a fresh-water environment, although some minor brackish influxes occurred. Human activity resulted in the presence of ruderals of dry and humid terrain, and probably also in the clearance of the alder vegetation.

44 The title of the manuscript is De vegetatie van Neolithisch Vlaardingen.

255

Copyright © 2010. Leiden University Press. All rights reserved.

APPenDIx VI - ARCHAeoBoTAny of THe LATe neoLITHIC SITe VLAARDInGen, THe neTHeRLAnDS

sample taxon Dryland vegetation Corylus avellana Rosa canina Atriplex patula/prostrata Brassica sp. Chenopodium album Chenopodium ficifolium Cirsium cf. arvense Conium maculatum Persicaria lapathifolia Persicaria maculosa Plantago major Polygonum aviculare Rumex obtusifolius Solanum nigrum Stellaria media Crop plants Papaver somniferum ssp. setigerum Carr and marsh vegetation Alnus glutinosa Alisma plantago-aquatica Bolboschoenus maritimus Carex riparia/acutiformis Cladium mariscus eleocharis palustris epilobium hirsutum eupatorium cannabinum Galium palustre Hypericum tetrapterum Lychnis flos-cuculi Lycopus europaeus Lythrum salicaria Mentha aquatica Myosotis scorpioides Peucedanum palustre Phragmites australis + = present ++ = present in relative large numbers

1

2

3

+ + + + + + + + + ++ + ++ + +

+ + + + + +

+ + + + -

+

-

-

++ + ++ ++ + + + + + + + + + + +

+ + + + + + +

+ + + + + + +

sample taxon Carr and marsh vegetation (cont.) Rumex hydrolapathum Schoenoplectus tabernaemontani Scirpus sylvaticus Solanum dulcamara Typha angustifolia/latifolia Urtica dioica Wetland pioneer vegetation Carex obtrubae Chenopodium glaucum/rubrum Juncus articulatus-type Juncus bufonius Juncus effusus-type Persicaria hydropiper Ranunculus cf. repens-type Ranunculus sceleratus Rumex conglomeratus Rumex crispus Open water vegetation Ceratophyllum submersum najas marina najas minor Potamogeton cf. pectinatus Potamogeton perfoliatus Ranunculus aquatilis-type Zannichellia palustris Salt marsh vegetation Apium graveolens Juncus gerardii Salicornia europaea Ecologically indeterminate Agrostis sp. Alopecurus sp. odontites sp. Poa pratensis/trivialis

1

2

3

+ ++ + + +

+ -

+ + +

+ + + + + + + +

+ + + + + + + -

+ -

+ + + + + +

+ -

+ + +

+ + -

+

+ +

+ +

+ + -

+ -

- = not present

Table VI.1 Vlaardingen, waterlogged macroremains from three samples dating to the Vlaardingen period, collected from 1) the refuse layer Vlaardingen 1b, 2) a find concentration located 30 cm below refuse layer Vlaardingen 1b and 3) a drift litter deposit layer Vlaardingen 1a (Van Zeist unpublished data).

256

APPenDIx VI - ARCHAeoBoTAny of THe LATe neoLITHIC SITe VLAARDInGen, THe neTHeRLAnDS

Copyright © 2010. Leiden University Press. All rights reserved.

It can be added that the samples probably represent the vegetation from open patches along the levees. The number of taxa indicative of closed deciduous woodland is small, although this vegetation may have been present nearby. The presence of salt marsh taxa and water plants that tolerate brackish conditions indicates some marine influence. The assemblage of plants nevertheless also contains species that do not tolerate brackish conditions, such as Alnus glutinosa, Cladium mariscus, Eupatorium cannabinum, Lycopus europaeus, Lythrum salicaria and Urtica dioica, indicating that marine influence was restricted in time and/or space. The only crop plant found in the investigated samples is Papaver somniferum ssp. setigerum (opium poppy). other neolithic finds of poppy are known from Brandwijk-Kerkhof, Schokland-P14 and flevoland (see paragraph 11.2.1). other potential food plants found in the samples are Corylus avellana and Rosa canina. In addition seeds, fruits and leaves of many herbs may have been consumed (Chenopodium sp., Persicaria sp., Apium graveolens, etc.), as well as the roots s.l. of taxa such as Typha sp. and Bolboschoenus maritimus. The information from other archaeobotanical studies from Vlaardingen (see below) indicates that the material from the presented three samples only represents a part of the complete assemblage of potential food plants. An interesting find are the fruits of Conium maculatum. This species has also been found at Swifterbant-S3 (Van Zeist and Palfenier-Vegter 1981) and Hekelingen III (Bakels 1988). The presence of the remains of this plant at Vlaardingen confirms that it was probably part of the natural vegetation of levees (cf. Bakels 1988). VI.3.2 InforMatIon froM earlIer publIcatIons Published information on other macroremains from Vlaardingen concern finds of cereals and wild plants. Identified cereals are Triticum aestivum s.l. (bread wheat/club wheat), Triticum dicoccon, Hordeum vulgare var. vulgare and Avena sp., found in a refuse pit in a house (Van Zeist 1970, 55-58). other macroremains from the Vlaardingen occupation period are shells of hazelnuts and stones of Prunus sp. (Van Beek 1990, 46, 82, 103, 159). Concerning Prunus sp., Van Beek refers to cherries (kersepit), which implies it would concern Prunus avium. However, if this indeed concerns a neolithic find, it probably represents Prunus spinosa or possibly Prunus padus and not Prunus avium. finds of stones of Prunus avium are only known in the netherlands from the Roman period onwards while the other two species have been found at other neolithic sites in the same region as Vlaardingen. Moreover, Prunus spinosa was also identified in the wood assemblage of Vlaardingen. The pollen analysis by Groenman-van Waateringe and niessen-Boomgaard is based on two cores. The first core was sampled in the backswamp area some 60 metres away from the edge of the natural levee (Van Regteren Altena et al. 1962, 22, 1963a, 53). The published diagram shows a selection of taxa. The analysis indicates the presence of woodland of dry terrain present on the levees, alder carr vegetation in the higher parts of the back swamps at the transition to the levees, and marsh vegetation in the back swamps. It is concluded that the woodland at the levees represented Ulmion woodland, comprising Ulmus sp., Fraxinus excelsior, Quercus sp., Corylus avellana, Acer sp. and some Tilia sp. The results of the first core confirm the occurrence of marine influence and the presence of ruderals indicative of human impact. The second core was sampled in the channel at less than 5 metres away from the refuse layer (Groenman-van Waateringe and Jansma 1969), and is partly contemporaneous with the Vlaardingen occupation. The range of taxa of trees and shrubs is similar to the results of the first core. Comparison of the periods during and after occupation indicates that occupation resulted in the presence of Cerealia-type pollen, Artemisia sp., Polygonum persicaria-type and Polygonum aviculare-type. Local crop cultivation is suggested (Groenman-van Waateringe and Jansma 1969, 114).

257

APPenDIx VI - ARCHAeoBoTAny of THe LATe neoLITHIC SITe VLAARDInGen, THe neTHeRLAnDS

The publication of Van Beek (1990) presents the wood and charcoal identifications, based on the work of C.H. Japing (Landbouwhogeschool Wageningen, now Wageningen University). The material was collected in several squares at the location of the former channel and levee. Table VI.2 shows the identifications of both material categories dating to the Vlaardingen and the Bell Beaker occupation periods. The precise plant names are not always clear since Van Beek gives most names in Dutch. Alnus glutinosa/incana probably represents A. glutinosa. The remains of Aesculus hippocastanum and Myrica gale probably represent contamination with recent material (Van Beek 1990, 41, 184). The identification of Sorbus aucuparia is remarkable since this is usually not identified on species level. The small diversity of species from the Bell Beaker period can be related to the relatively small number of samples available from this period. In the wood and charcoal assemblage from the Vlaardingen period, Alnus sp., Fraxinus excelsior and Acer sp. are very common, Quercus sp. and Corylus avellana occur in an intermediate frequency, and Salix sp., Ulmus sp., Sorbus aucuparia, Lonicera periclymenum, Prunus spinosa, and Taxus baccata are very scarce (Van Beek 1990; Van Regteren Altena et al. 1963b). Most taxa identified in the wood and charcoal assemblage may have been part of the local vegetation in the exploitation area of Vlaardingen: Quercus sp., Corylus avellana, Fraxinus excelsior, Ulmus sp., Acer campestre, Betula sp., Crataegus monogyna, Sorbus aucuparia, Prunus spinosa, Lonicera periclynemum, Alnus glutinosa and Salix sp. The information on frequencies indicates that Alnus sp., Fraxinus excelsior and Acer sp. were probably dominant in the local vegetation, when assuming that the wood represents the local and extra-local vegetation. However, it cannot be excluded that wood was imported from elsewhere in the exploitation area of the site. cultural group

wood

charcoal

Bell Beaker culture

Alnus glutinosa/incana Corylus avellana fraxinus excelsior Aesculus hippocastanum Acer sp. Alnus glutinosa/incana Corylus avellana Crataegus monogyna fraxinus excelsior Larix decidua Lonicera periclymenum Prunus spinosa Quercus sp. Salix sp. Sorbus aucuparia Taxus baccata Ulmus sp.

Acer campestre Acer platanoides Alnus glutinosa/incana Betula sp. Corylus avellana fraxinus excelsior Myrica gale Quercus sp. Salix sp. Ulmus sp.

Copyright © 2010. Leiden University Press. All rights reserved.

Vlaardingen group

- = not present Table VI.2 Vlaardingen, wood and charcoal dating to the Vlaardingen group and Bell Beaker culture (Van Beek 1990). See the text for comments on the validity of the results.

258

APPenDIx VI - ARCHAeoBoTAny of THe LATe neoLITHIC SITe VLAARDInGen, THe neTHeRLAnDS

for some species, their presence in the (extra-) local vegetation is less likely. firstly, taxa that were scarce in the wood and charcoal assemblage (presented above) may have been collected elsewhere. furthermore, Larix sp. was probably not part of the natural vegetation of the netherlands. Van Beek (1990, 184) states that remains of Larix sp. may have originated from Switzerland and may have reached the netherlands via the river Rhine. This would represent transport over an unusually large distance. In addition, it is not certain whether Acer platanoides was present in the natural vegetation of the netherlands during the neolithic, since archaeobotanical finds are not known (Maes 2006, 91). A. platanoides was certainly not common, and it is expected that most wood and charcoal identifications of Acer sp. represent Acer campestre. Interestingly, Taxus baccata becomes part of the natural vegetation of the studied regions in the neolithic (see chapter 7). The wood of T. baccata found at Vlaardingen may have been collected in the exploitation area of the site or may have been transported over a larger distance. finds of wood of T. baccata are also known from the Middle neolithic coastal sites Schipluiden (Kooistra 2006) and ypenburg (Kooistra and Hänninen 2008). Another find from a Dutch neolithic wetland site is a bow found in one of the younger layers at the Hazendonk (Louwe Kooijmans 1987).

Copyright © 2010. Leiden University Press. All rights reserved.

VI.4 dIscussIon and conclusIons The identifications of pollen, wood and charcoal all indicate the presence of alluvial woodland and alder carr in the extra-local vegetation of the settlement at Vlaardingen. Therefore, the major part of the wood was probably collected in the near surroundings of the site. It is not possible to conclude for all species from the wood and charcoal assemblage whether they where present in the natural vegetation or not, and there are indications of contamination with recent material. The pollen data only partly support the variety of taxa that is indicated by the wood analysis. This may support the gathering of some taxa from a distance of at least some km, but it should also be kept in mind that at least one of the pollen diagrams only represents a selection of taxa. In contrast to the pollen, wood and charcoal data, the macroremains identifications do not support the presence of alluvial woodland, but emphasise the presence of alder carr, marshes and dry to moist ruderal terrain. This indicates that the macroremains probably represent the vegetation of the disturbed terrain used for occupation and of the levees along the channel, and the drift litter present in this zone. The wood and pollen data probably represent the vegetation around the settlement, such as the less disturbed vegetation of the levees and the backswamp area. In addition, the small number of samples of macroremains and changes in the vegetation through time may explain the differences between the macroremains assemblage and the other assemblages as well. The newly presented macroremains show human impact, since the assemblage contains several indicators of disturbed and eutrophic conditions. Many taxa may have possibly functioned as food plants or may have been used in other ways. However, the absence of carbonised macroremains restricts further interpretation of people’s handling of taxa. VI.5 acknowledgeMents The author would like to thank W. van Zeist for permission to publish the macroremains data.

259

Copyright © 2010. Leiden University Press. All rights reserved.

Appendix VII. References of appendices Aalto, M. 1970

Potamogeton fruits I. Recent and subfossil endocarps of the Fennoscandian species, Acta Botanica Fennica 88, 1-85.

Amkreutz, L.W.S.W. in prep. Negotiating neolithisation. A long-term perspective on communities in the process of neolithisation in the Lower Rhine Area, Ph.D. thesis Leiden University. Amkreutz, L., L. Verhart & M. Wansleeben Hazendonk layers over and over again, in: H. Fokkens, B. Coles, A. van Gijn, J. Kleijne, H. Ponjee & 2008 C. Slappendel (eds), Between foraging and farming. An extended broad spectrum of papers presented to Leendert Louwe Kooijmans, Leiden (Analecta Praehistorica Leidensia 40), 139-152. Anderson, P.C. 1992

Bakels, C.C. 1981 1982 1986 1988 1991 1999 2000 2005 2007

Experimental cultivation, harvest and threshing of wild cereals and their relevance for interpreting the use of Epipalaeolithic and Neolithic artefacts, in: P.C. Anderson (ed.), Préhistoire de l’agriculture: nouvelles approaches expérimentales et ethnographiques, Paris (Monographie du Centre de Recherches Archéologiques 6), 179-209. Neolithic plant remains from the Hazendonk, province of Zuid-Holland, the Netherlands, Zeitschrift für Archäologie 15, 141-148. Der Mohn, die Linearbandkeramik und das westliche Mittelmeergebiet, Archäologisches Korrespondenzblatt 12 (11), 11-13. Akkerbouw in het moeras?, in: M.C. van Trierum & H.E. Henkes (eds), Rotterdam Papers, a contribution to Prehistoric, Roman and medieval archaeology, Rotterdam, 1-6. Hekelingen, a Neolithic site in the swamps of the Maas estuary, in: H. Küster (ed.), Der prähistorische Mensch und seine Umwelt, Festschrift U. Körber-Grohne zum 65. Geburtstag, Stuttgart (Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Württemberg 31), 155-161. Western Continental Europe, in: W. van Zeist, K. Wasylikowa & K.-E. Behre (eds), Progress in Old World palaeoethnobotany, Rotterdam, 279-298. Archaeobotanical investigations in the Aisne valley, northern France, from the Neolithic up to the Early Middle Ages, Vegetation History and Archaeobotany 8 (1-2), 71-77. The Neolithization of the Netherlands: two ways, one result, in: A.S. Fairbairn (ed.), Plants in Neolithic Britain and beyond, Oxford, 101-106. Planten in de steentijd, in: J. Deeben, E. Drenth, M.-F.van Oorsouw & L. Verhart (eds), De steentijd van Nederland, Zutphen (Archeologie 11/12), 67-80. Nature or culture? Cereal crops raised by Neolithic farmers on Dutch loess soils, in: S. Colledge & J. Conolly (eds), 343-347.

Copyright © 2010. Leiden University Press. All rights reserved.

Bakels, C.C. & R. Rousselle 1985 Restes botaniques et agriculture du Néolithique ancien en Belgique et aux Pays-Bas, Helinium 25, 37-57. Bakels, C., C. Constantin & A. Hauzeur 1992 Utilisation de graines de pavot comme dégraissant dans un vase du groupe de Blicquy, Archäologisches Korrespondenzblatt 12, 473-479. Bakels, C.C. & L.M. van Beurden Archeobotanie, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam 2001 Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83), 325-378. Bakels, C.C., L.M. van Beurden & T.J.J. Vernimmen Archeobotanie, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam De 2001 Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88), 369-434.

261

APPENDIX VII REFERENCES OF APPENDICES

Bakels, C. & J. Zeiler 2005 The fruits of the land. Neolithic subsistence, in: L.P. Louwe Kooijmans, P.W. van den Broeke, H. Fokkens & A.L. van Gijn (eds), 311-335. Ball, E. 1997

De vissen van Brandwijk-Kerkhof, Leiden (internal report Faculty of Archaeology, Leiden University).

Bang, P. 1973

Elseviers diersporengids, Amsterdam.

Barthelmes, A., A. Prager & H. Joosten Palaeoecological analysis of Alnus wood peats with special attention to non-pollen palynomorphs, Review of 2006 Palaeobotany and Palynology 141, 33-51. Bastiaens, J., K. Deforce, L. Meersschaert & B. Klinck Two Late Mesolithic sites along the river Scheldt (Doel, Belgium): focus on woodland and the use of Viscum 2007 album and Hedera helix, poster presentation at the14th Symposium of the International Work Group for Palaeoethnobotany, 17-23 June 2007, Krakow, Poland. Beek, B.L. van 1990 Beerenhout, B. 2001 Behre, K.-E. 1981

Steentijd te Vlaardingen, Leidschendam en Voorschoten. De vondstverspreiding in laat-neolithische nederzettingen in het Hollandse kustgebied, Ph.D. thesis University of Amsterdam. Vissen, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giesendam De Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88), 299-326. The interpretation of anthropogenic indicators in pollen diagrams, Pollen et Spores 23 (2), 225-245.

Behre, K.-E. & D. Kučan Die Reflection archäologisch bekannter Siedlungen in Pollendiagrammen verschiedener Entfernung – 1986 Beispiele aus der Siedlungskammer Flögeln, Nordwestdeutschland, in: K.-E. Behre (ed.), Anthropogenic indicators in pollen diagrams, Boston/Rotterdam, 95-115. Berggren, G. 1969 Beug, H.-J. 1963

Atlas of seeds and small fruits of Northwest-European plant species with morphological descriptions 2. Cyperacae, Lund. Leitfaden der Pollenbestimmung I, Jena.

Copyright © 2010. Leiden University Press. All rights reserved.

Bloemers, J.F.H., L.P. Louwe Kooijmans & H. Sarfatij 1981 Verleden land. Archeologische opgravingen in Nederland, Amsterdam. Bottema, S. 1992

Braadbaart, F. 2004 2008

262

Cereal-type pollen in the Near East as indicators of wild or domestic crops, in: P.C. Anderson (ed.), Préhistoire de l’agriculture: nouvelles approaches expérimentales et ethnographiques, Paris (Monographie du Centre de Recherches Archéologiques 6), 95-106. Carbonization of peas and wheat – a window into the past, Ph.D. thesis Leiden University. Carbonisation and morphological changes in modern dehusked and husked Triticum dicoccum and Triticum aestivum grains, Vegetation History and Archaeobotany 17, 155-166.

APPENDIX VII REFERENCES OF APPENDICES

Brinkkemper, O., W.-J. Hogestijn, H. Peeters, D. Visser & C. Whitton 1999 The Early Neolithic site at Hoge Vaart, Almere, The Netherlands, with particular reference to non-diffusion of crop plants, and the significance of site function and sample location, Vegetation History and Archaeobotany 8, 79-86. Brinkkemper, O., E.J. Weeda, S.J.P. Boncke & W.J. Kuijper The past and present occurrence of Elatine and implications for environmental reconstructions, Vegetation 2008 History and Archaeobotany 17, 15-24. Broeke, P.W. van den 1983 Neolithic bone and antler objects from the Hazendonk near Molenaarsgraaf (prov. South Holland) (Hazendonk paper 2), Oudheidkundige Mededelingen uit het Rijksmuseum van Oudheden te Leiden 64, 163-196. Brouwer, W. & A. Stählin Handbuch der Samenkunde, Frankfurt am Main. 1975 Cappers, R.T.J. 1994 2006

An ecological characterization of plant macro-remains of Heveskesklooster (The Netherlands), Ph.D. thesis University of Groningen. The reconstruction of agricultural practices in ancient Egypt: an ethnoarchaeobotanical approach, Palaeohistoria 47/48, 429-446.

Cappers, R.T.J., T. van Thuyne & L. Sikking Plant remains from Predynastic El Abadiya-2 (Naqada area, Upper Egypt), in: S. Hendrickx, R.F. Friedman, 2004 K.M. Cialowicz & M. Chlodnicki (eds), Egypt at its Origins. Studies in memory of Barbara Adams, Leuven (Orientalia Lovaniensia Analecta 138), 277-293. Cappers, R.T.J. & D.C.M. Raemaekers Cereal cultivation at Swifterbant? About being a Neolithic wetland farmer in the western part of the North 2008 European Plain. Current Anthropology 49 (3), 385-402. Casparie, W.A. s.a. 1986 1995

The wood of Early Neolithic Bergschenhoek (The Netherlands). Unpublished manuscript, archives National Museum of Antiquities, Leiden. The two Iron Age wooden trackways XIV (Bou) and XV (Bou) in the raised bog of Southeast Drenthe (the Netherlands), Palaeohistoria 28, 169-210. Houtgebruik van paleolithicum tot middeleeuwen: voorbeelden uit de archeologie, in: K. Helfrich, J.F. Benders & W.A. Casparie (eds), Handzaam hout uit Groninger grond. Houtgebruik in de historische stad, Groningen, 212-219.

Copyright © 2010. Leiden University Press. All rights reserved.

Clason, A.T. & D.C. Brinkhuizen 1993 Bergschenhoek, in: A. Clason, S. Payne & H.-P. Uerpmann (eds), Skeletons in her cupboard. Festschrift for Juliet Clutton-Brock, Oxford, 61-73. Colledge, S. & J. Conolly (eds) 2007 The origins and spread of domestic plants in Southwest Asia and Europe, Walnut Creek. Dickson, C.A. 1970 Dijk, G.J. van 1979

The study of plant macrofossils in British Quaternary deposits, in: D. Walker & R.G. West (eds), Studies in the vegetational history of the British Isles, London, 233-254. Verslag van een palynologish onderzoek nabij de Hazendonk (Alblasserwaard), Amsterdam (internal report Faculty of Earth and Life Sciences, VU University).

263

APPENDIX VII REFERENCES OF APPENDICES

Dijk, G.J. van, R.M. van Heeringen & M. Verbruggen 1976 De Hazendonk, Leiden/Amsterdam (internal report Faculty of Archaeology, Leiden University/Faculty of Earth and Life Sciences, VU University). Eberschweiler, B. Bronzezeitliches Schwemmgut vom “Chollerpark” in Steinhausen (Zug), Basel (Antiqua 37). 2004 Eenkhoorn, J. & E.M.A. Smit Inventarisatie en beheer van het Zalkerbos, Wageningen (Afdeling Natuurbeheer Verslag nr. 511). 1982 Ellis, M.B. & J.P. Ellis Microfungi on landplants. An identification handbook, Slough. 1997 Erdtman, G., B. Berglund & J. Praglowski An introduction to a Scandinavian pollen flora, Stockholm. 1961 Fægri, K. & J. Iversen 1989 Textbook of pollen analysis, Chichester, 4th edition. Fuller, D.Q., C. Stevens & M. McClatchie in press Routine activities, tertiary refuse and labor organization: social inference from everyday archaeobotany, in: M. Madella & M. Savard (eds), Ancient plants and people. Contemporary trends in Archaeobotany, Tucson. Geel, B. van 1976 1978

A paleoecological study of Holocene peat bog sections, based on the analysis of pollen, spores and macroand microscopic remains of fungi, algae, cormophytes and animals, Ph.D. thesis University of Amsterdam. A palaeoecological study of Holocene peat bog sections in Germany and the Netherlands, Review of Palaeobotany and Palynology 25, 47-115.

Geel, B. van, S.J.P. Bohncke & H. Dee A palaeoecologocial study of an upper Late Glacial and Holocene sequence from “De Borchert”, 1981 the Netherlands, Review of Palaeobotany and Palynology 31, 415-445. Geel, B. van, J. Buurman, O. Brinkkemper, J. Schelvis, A. Aptroot, G. van Reene & T. Hakbijl 2003 Environmental reconstructions of a Roman Period settlement site in Uitgeest (The Netherlands), with special reference to coprophilous fungi, Journal of Archaeological Science 30, 873-883. Geel, B. van & A. Aptroot 2006 Fossil ascomycetes in Quaternary deposits, Nova Hedwigia 82 (3-4), 313-329.

Copyright © 2010. Leiden University Press. All rights reserved.

Gijn, A.L. van & M. Verbruggen 1992 Brandwijk-Kerkhof: Het Kerkhof, in: P.J. Woltering & W.A.M. Hessing (eds), Archeologische Kroniek van Holland 1991, Amersfoort (Holland 24, 6), 348-352. Gijn, A.L. van, Y. Lammers-Keijsers & R. Houkes 2001 Vuursteen, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam De Bruin, Amersfoort, (Rapportage Archeologische Monumentenzorg 88), 153-191. Gijn, A.L. van & L.P. Louwe Kooijmans Early and Middle Neolithic: introduction, in: L.P. Louwe Kooijmans, P.W. van den Broeke, H. Fokkens & 2005 A.L. van Gijn (eds), 203-218. Gijn, A.L. van, V. van Betuw, A. Verbaas & K. Wentink 2006 Flint, procurement and use, in: L.P. Louwe Kooijmans & P.F.B. Jongste (eds), 129-166.

264

APPENDIX VII REFERENCES OF APPENDICES

Greguss, P. 1956

Pflanzenreste aus der Kreidezeit von Tata, Acta Biologica Szegediensis 2, 39-49.

Grimm, E.C. 1991-1993, 2004

TILIA v2.0.b.4 and TGView v2.0 (computer software), Springfield.

Groenman-van Waateringe, W. & M.J. Jansma Diatom and pollen analysis of the Vlaardingen creek, a revised interpretation, Helinium 9, 105-17. 1969 1978 The impact of Neolithic man on the landscape in the Netherlands, in: S. Limbrey & J.G. Evans (eds), The effect of man on the landscape: the lowland zone, London (Council for British Archaeology Research Report 21), 135-146. Grohne, U. 1957

Die Bedeutung des Phasenkontrastverfahrens für die Pollenanalyse, dargelegt am Beispiel der Gramineenpollen vom Getreidetyp, Photographie und Forschung, Zeiss Ikon im Dienste der Wissenschaft 7, 237-248.

Guiran, A.J. & O. Brinkkemper in cooperation with D.C. Brinkhuizen & A. Klink Rotterdam-Randstadrail: archeologisch onderzoek 1 Emplacement Centraal Station, Rotterdam 2007 (BOORrapporten 318). Hakbijl, T. 2001

Arthropoden, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83), 277-281.

Hamburg, T.D. & L.P. Louwe Kooijmans 2001 Sporen en structuren, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. HardinxveldGiessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83), 73-104. Hänninen, K. and C. Vermeeren Hout en houtskool van een donk met bewoningssporen uit het midden Neolithicum in Brandwijk-Kerkhof 1998 (Z.H.), Amsterdam (BIAXiaal 61). Hänninen, K., L.I. Kooistra, P. van Rijn & C. Vermeeren 1999 Archeologie Betuweroute: hout en houtskool van een donk uit het Meso- of Neolithicum te HardinxveldPolderweg, Amsterdam (BIAXiaal 65). Hänninen, K. & C. Vermeeren 1999 Hout van de donk Hardinxveld-De Bruin, gedateerd in het meso- en neolithicum, Amsterdam (BIAXiaal 78).

Copyright © 2010. Leiden University Press. All rights reserved.

Helbæk, H. 1952 Hillman, G. 1981

Early crops in Southern England, Proceedings of Prehistoric Society 18, 194-233. Reconstructing crop husbandry practices from carbonised remains of crops, in: R. Mercer (ed.), Farming practice in British prehistory, Edinburgh, 123-163.

Innes, J., J. Blackford & F. Chambers 2006 Kretzschmaria deusta and the Northwest European Mid-Holocene Ulmus decline at Moel Y Gerddi, North wales, United Kingdom, Palynology 30, 121-132. Jessen, K. 1955

Key to subfossil Potamogeton, Botanisk Tidsskrift 52, 1-7.

Jongste P.F.B. & M. Verbruggen 1998 Aanvullend archeologisch onderzoek in het tracé van de Betuweroute, vindplaats 4: Hardinxveld-Giessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 37). 265

APPENDIX VII REFERENCES OF APPENDICES

Karg, S. 2006

The water chestnut (Trapa natans L.) as a food source during the 4th to 1st millennia BC at Lake Federsee, Bad Buchau (southern Germany), Environmental Archaeology 11 (1), 125-130.

Katz, N.J., S.V. Katz & M.G. Kipiani 1965 Atlas and keys of fruits and seeds occurring in the Quaternary deposits of the USSR, Moscow. Klerk, P. de, C.R. Janssen & J.H.J. Joosten 1997 Patterns and processes in natural wetland vegetation in the Dutch fluvial area: a palaeoecological study, Acta Botanica Neerlandica 46 (2), 147-159. Knörzer, K.-H. 1970 1973

Römerzeitliche Pflanzenreste aus Neuss, Berlin (Novaesium 4). Römerzeitliche Pflanzenreste aus einem Brunnen in Butzbach (Hessen), Mainz am Rhein (Saalburg-Jahrbuch 30), 71-114.

Knörzer, K.-H., R. Gerlach, J. Meurers-Balke, A.J. Kalis, U. Tegtmeier, W.D. Becker & A. Jürgens PflanzenSpuren. Archäobotanik im Rheinland. Agrarlandschaft und Nutzpflanzen in Wandel der Zeiten, Köln. 1999 Kooistra, L.I. 2006 2008

Wood and charcoal, in: L.P. Louwe Kooijmans & P.F.B. Jongste (eds), 363-374. Artefacten van hout en bast, in: H. Koot, L. Bruning & R.A. Houkes (eds), 270-275.

Kooistra, M.J., L.I. Kooistra, P. van Rijn & U. Sass-Klaassen Woodlands of the past, the excavation of wetland woods at Zwolle-Stadshagen (the Netherlands): 2006 reconstruction of the wetland wood in its environmental context, Netherlands Journal of Geosciences 85 (1), 37-60. Kooistra, L.I. & K. Hänninen Hout en houtskool, in: H. Koot, L. Bruning & R.A. Houkes (eds), 345-361. 2008 Koot, H., L. Bruning & R.A. Houkes (eds) Ypenburg-locatie 4. Een nederzetting met grafveld uit het Midden-Neolithicum in het West-Nederlandse 2008 Kustgebied, Den Haag. Kort, J.-W. de 1998

Zadenonderzoek van een sectie uit het veen bij de donk aan de Polderweg, Hardinxveld-Giessendam, Leiden (internal report Faculty of Archaeology, Leiden University).

Copyright © 2010. Leiden University Press. All rights reserved.

Kowal, T. & W. Rudnicka-Sternowa Fruit morphology and anatomy of the Bromus L. species of Poland, Warsaw, (Monographiae Botanicae 29), 1969 3-68. Körber-Grohne, U. 1964 Bestimmungsschlüssel für subfossile Juncus-Samen und Gramineen-Früchte, Probleme der Küstenforschung im Südliche Nordseegebiet 7, 1-47. 1967 Geobotanische Untersuchungen auf der Feddersen Wierde, Wiesbaden, (Feddersen Wierde 1). Kroll, H. 2001

Der Mohn, die Trichterbecherkultur und das südwestliche Ostseegebiet. Zu den Pflanzenfunden aus der mittelneolithischen Fundschicht von Wangels, Kr. Ostholstein, in: R. Kelm (ed.), Zurück zur Steinzeitlandschaft. Archäobiologische und ökologische Forschung zur jungsteinzeitlichen Kulturlandschaft und ihrer Nutzung in Nordwestdeutschland, Heide, 70-76.

Kubiak-Martens, L. 2006 Roots, tubers and processed plant food in the local diet, in: L.P. Louwe Kooijmans & P.F.B. Jongste (eds), 339-352. 266

APPENDIX VII REFERENCES OF APPENDICES

Kuijper, W.J. 2000 Lange, E. 1979

The former occurrence of Neckera crispa Hedw. in The Netherlands, Lindbergia 25, 28-32. Verkohlte Pflanzenreste aus den slawischen Siedlungsplätzen Brandenburg und Zirzow (Kr. Neubrandenburg), Archaeo-Physika 8, 191-207.

Langhe, J.-E. de, L. Delvosalle, J. Duvigneaud, J. Lambinon & C. Vanden Berghen Nouvelle flore de la Belgique, du Grand-Duché de Luxembourg, du nord de la France et des régions voisines 1978 (Ptéridophytes et Spermatophytes), Meise. Lanting, J.N. & J. van der Plicht De 14C-chronologie van de Nederlandse pre- en protohistorie III, Palaeohistoria 41/42, 1-110. 2000 Leuzinger, U. 2002

Holzartefakte, in: A. de Capitani, S. Deschler-Erb, U. Leuzinger, E. Marti-Grädel & J. Schibler (eds), Die junsteinzeitliche Seeufersiedlung Arbon Bleiche 3. Funde, Frauenfeld (Archäologie im Thurgau Band 11), 76-114.

Copyright © 2010. Leiden University Press. All rights reserved.

Louwe Kooijmans, L.P. The Rhine/Meuse delta. Four studies on its prehistoric occupation and Holocene geology, Ph.D thesis Leiden 1974 University (Analecta Praehistorica Leidensia 7). 1976 Local developments in a borderland. Oudheidkundige Mededelingen uit het Rijksmuseum van Oudheden te Leiden 57, 227-297. 1977a Bergschenhoek, in: H. Sarfatij (ed.), Archeologische kroniek van Zuid-Holland over 1976, Amersfoort (Holland 9), 245-247. 1977b Molenaarsgraaf: Hazendonk, in: H. Sarfatij (ed.), Archeologische kroniek van Zuid-Holland over 1976, Amersfoort (Holland 9), 248-251. 1978 Bergschenhoek, in: H. Sarfatij (ed.), Archeologische Kroniek van Holland over 1977, Amersfoort (Holland 10), 297-300. 1985 Sporen in het land, Amsterdam. 1986 Het loze vissertje of boerke Naas?, in: M.C. van Trierum & H.E. Henkes (eds), Rotterdam Papers, a contribution to Prehistoric, Roman and medieval archaeology, Rotterdam, 7-25. 1987 Neolithic settlement and subsistence in the wetlands of the Rhine/Meuse delta, in: J.M. Coles & A.J. Lawson (eds), European wetlands in Prehistory, Oxford, 227-251. 2003 The Hardinxveld sites in the Rhine/Meuse delta, The Netherlands, 5500-4500 cal BC, in: L. Larsson, H. Kindgren, K. Knutsson, D. Loeffler & A. Åkerlund (eds), Mesolithic on the move. Papers presented at the sixth international conference on the Mesolithic in Europe, Stockholm 2000, Oxford, 608-624. 2005 Hunters become farmers. Early Neolithic B and Middle Neolithic A, in: L.P. Louwe Kooijmans, P.W. van den Broeke, H. Fokkens & A.L. van Gijn (eds), 249-271. 2007 The gradual transition to farming in the Lower Rhine Basin, in: A. Whittle & V. Cummings (eds), Going over: the Mesolithic-Neolithic transition in north-west Europe, Oxford (Proceedings of the British Academy 144), 287-309. Louwe Kooijmans, L.P. (ed.) 2001a Archeologie in de Betuweroute. Hardinxveld-Giessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83). 2001b Archeologie in de Betuweroute. Hardinxveld-Giessendam De Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88). Louwe Kooijmans, L.P., K. Hänninen & C.E. Vermeeren 2001 Artefacten van hout, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. HardinxveldGiesendam De Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88), 435-478. Louwe Kooijmans, L.P. & J. Mol 2001 Stratigrafie, chronologie en fasering, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83), 55-72. 267

APPENDIX VII REFERENCES OF APPENDICES

Louwe Kooijmans, L.P. & M. Nokkert 2001 Sporen en structuren, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. HardinxveldGiessendam De Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88), 75-116. Louwe Kooijmans, L.P., C.E. Vermeeren & A.M.I. van Waveren Hout en artefacten van vezels, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld2001 Giessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83), 379-418. Louwe Kooijmans, L.P., P.W. van den Broeke, H. Fokkens & A.L. van Gijn (eds) Nederland in de prehistorie, Amsterdam. 2005 Louwe Kooijmans L.P. & P.F.B. Jongste (eds) Schipluiden. A Neolithic settlement on the Dutch North Sea Coast c. 3500 cal BC, Leiden (Analecta 2006 Praehistorica Leidensia 37/38). Louwe Kooijmans, L.P. & L. Kooistra Wooden artefacts, in: L.P. Louwe Kooijmans & P.F.B. Jongste (eds), 225-251. 2006 Maes, B. 2006

Inheemse bomen en struiken in Nederland en Vlaanderen. Herkenning, verspreiding, geschiedenis en gebruik, Amsterdam.

Mears, R. & G. Hillman Wild food, London. 2007 Meijden, R. van der Heukels’ flora van Nederland, Groningen. 1996 Mery, S., P. Anderson, M.-L. Inizan, M. Lechevallier & J. Pelegrin A pottery workshop with flint tools on blades knapped with copper at Nausharo (Indus civilisation, 2007 ca. 2500 BC), Journal of Archaeological Science 34, 1098-1116. Meurers-Balke, J. Steinzeitliche Aalstecher. Zur funktionalen Deutung einer Holzgerätform, Offa 38, 131-151. 1981 Mol, J. 2001a 2001b

Copyright © 2010. Leiden University Press. All rights reserved.

2003

Landschapsgenese en paleogeografie, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam Polderweg, Amersfoort (Rapportage Archeologische Monumentenzorg 83), 35-54. Landschapsgenese en paleogeografie, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam De Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88), 33-56. Landscape evolution and site formation of two Mesolithic sites in the lower Rhine-Meuse delta (Hardinxveld, The Netherlands), in: A.J. Howard, M.G. Macklin & D.G. Passmore (eds), Alluvial archaeology in Europe, Abingdon/Exton/Lisse/Tokyo, 147-161.

Mol, J. & L.P. Louwe Kooijmans 2001 Stratigrafie, chronologie en fasering, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. Hardinxveld-Giessendam De Bruin, Amersfoort (Rapportage Archeologische Monumentenzorg 88), 57-74. Mol, J. & W. van Zijverden 2007 Prehistorische woonplaatsen in de dynamische landschappen van de Rijn-Maasdelta, in: R. Jansen & L.P. Louwe Kooijmans (eds), Tien jaar Archol. Van contract tot wetenschap, Leiden, 89-101. Moore, P.D., J.A. Webb & M.E. Collinson 1991 Pollen analysis, London. Nilsson, Ö. & H. Hjemquist 1967 Studies on the nutlet structure of south Scandinavian species of Carex, Botaniska Notiser 120, 460-485. 268

APPENDIX VII REFERENCES OF APPENDICES

Nilsson, S. & J.R. Praglowski 1963 An introduction to a Scandinavian pollen flora II, Stockholm, 28-64. Oberdorfer, E. 1983

Pflanzensoziologische Exkursionsflora, Stuttgart.

Olivat, H.J.P. d’ & J.P. Pals Determination tables and description of the seeds and fruits of the Ranunculaceae occurring in the 1974 Netherlands, Amsterdam (IPP publication 149). Ooststroom, S.J. van & T.J. Reichgelt Alismataceae, Flora Neerlandica 1(6), 1-15. 1964 Out, W.A. 2008a 2008b 2008c

Neolithisation at the site Brandwijk-Kerkhof, the Netherlands: natural vegetation, human impact and plant food subsistence, Vegetation History and Archaeobotany 17, 25-39. Selective use of Cornus sanguinea L. (red dogwood) for Neolithic fish traps in the Netherlands, Environmental archaeology 13.1, 1-10. The scale of human impact at the Hazendonk, the Netherlands, during the Late Neolithic, in: H. Fokkens, B. Coles, A. van Gijn, J. Kleijne, H. Ponjee & C. Slappendel (eds), Between foraging and farming, an extended broad spectrum of papers presented to Leendert Louwe Kooijmans, Leiden (Analecta Praehistorica Leidensia 40), 153-165.

Pals, J.P., B. van Geel & A. Delfos 1980 Palaeoecological studies in the Klokkeweel bog near Hoogkarspel (prov. of Noord-Holland), Review of Palaeobotany and Palynology 30, 371-418. Perry, D. 1999 Poecke, M. van 1991

Vegetative tissues from Mesolithic sites in the Northern Netherlands, Current Anthropology 40 (2), 231-237. Opgravingsverslag Brandwijk-Kerkhof 1991. De stratigrafie, Leiden (internal report Faculty of Archaeology, Leiden University).

Prager, A., A. Barthelmes, M. Theuerkauf & H. Joosten Non-pollen palynomorphs from modern alder carrs and their potential for interpreting microfossil data from 2006 peat, Review of Palaeobotany and Palynology 141, 7-31.

Copyright © 2010. Leiden University Press. All rights reserved.

Punt, W., C.R. Janssen, T.J. Reitsma & G.C.S. Clarke The Northwest European Pollen Flora I, Amsterdam. 1976 Punt, W. & G.C.S. Clarke 1980 The Northwest European Pollen Flora II, Amsterdam. 1981 The Northwest European Pollen Flora III, Amsterdam. 1984 The Northwest European Pollen Flora IV, Amsterdam. Punt, W., S. Blackmore & G.C.S. Clarke 1988 The Northwest European Pollen Flora V, Amsterdam. Punt, W. & S. Blackmore 1991 The Northwest European Pollen Flora VI, Amsterdam. Punt, W., S. Blackmore & P.P. Hoen 1991 The Northwest European Pollen Flora VII, Amsterdam.

269

APPENDIX VII REFERENCES OF APPENDICES

Raemaekers, D.C.M. 1999 The articulation of a ‘New Neolithic’. The meaning of the Swifterbant Culture for the proces of neolithisation in the western part of the North European Plain (4900-3400 BC), Ph.D. thesis Leiden University (Archaeological Studies Leiden University 3). 2001 Aardewerk en verbrande klei, in: L.P. Louwe Kooijmans (ed.), Archeologie in de Betuweroute. HardinxveldGiesendam De Bruin, Leiden (Rapportage Archeologische Monumentenzorg 88), 117-149. Regteren Altena, J.F. van, J.A. Bakker, A.T. Clason, W. Glasbergen, W. Groenman-van Waateringe & L.J. Pons The Vlaardingen Culture (I), Helinium 2, 3-35. 1962 1963a The Vlaardingen Culture (IV), Helinium 3, 39-54. 1963b The Vlaardingen Culture (V), Helinium 3, 97-120. Reille, M. 1992 1995

Pollen et spores d’Europe et d’Afrique du Nord, Marseille. Pollen et spores d’Europe et d’Afrique du Nord, supplement 1, Marseille.

Robinson, M. & R.N.L.B. Hubbard The transport of pollen in the bracts of hulled cereals, Journal of Archaeological Science 4, 197-199. 1977 Roller, G.J. de, G. Korf, I.L.M. Stuijts & B. Mook-Kamps Archeobotanie, in: J. Milojkovic & E. Smits (eds), Archeologie in de Betuweroute Lage Blok, een 2002 nederzettingsterrein uit de Midden-IJzertijd bij Meteren, Amersfoort (Rapportage Archeologische Monumentenzorg 90), 179-227. Sass-Klaassen, U. & E. Hanraets 2006 Woodlands of the past – the excavation of wetland woods at Zwolle-Stadshagen (the Netherlands): growth pattern and population dynamics of bog oaks, Netherlands Journal of Geosciences 85 (1), 61-71. Schaminée, J.H.J., A.H.F. Stortelder & V. Westhoff De vegetatie van Nederland 1-5, Leiden/Uppsala. 1995-1999 Schmölcke, U., E. Endtmann, S. Klooss, M. Meyer, D. Michaelis, B.-H. Rickert & D. Rößler Changes of sea level, landscape and culture. A review of the south-western Baltic area between 8800 and 2006 4000 BC, Palaeogeography, Palaeoclimatology, Palaeoecology 240 (3-4), 423-438. Schweingruber, F.H. Microscopic wood anatomy, Birmensdorf. 1978 Skaarup, J. 1980

Copyright © 2010. Leiden University Press. All rights reserved.

Steenbeek, R. 1980

Undersøisk stenalder, Skalk 1, 3-8. Palynologisch onderzoek aan twee boringen bij de Hazendonk, Amsterdam (internal report Faculty of Earth and Life Sciences, VU University).

Stortelder, A.F.H., J.H.J. Schaminée & P.W.F.M. Hommel 1999 De vegetatie van Nederland 5. Ruigten, struwelen en bossen, Leiden/Uppsala. Sugita, S. 1994 Tjaden, M.E.H. 1919

Pollen representation of vegetation in Quaternary sediments: theory and method in patchy vegetation, Journal of Ecology 82, 881-897. Microscopisch onderzoek van hout, Amsterdam.

Touw, A. & W.V. Rubers 1989 De Nederlandse bladmossen, Utrecht. 270

APPENDIX VII REFERENCES OF APPENDICES

Troels-Smith, J. 1954 Verbruggen, M. 1992 in prep.

Ertebøllekultur-Bondekultur. Resultater af de didste 10 Aars Undersøgelser i Aamosen, Vestsjælland, Aarbøger for Nordisk Oldkyndighed og Historie 1953/1954, 5-62. Geoarchaeological prospection of the Rommertsdonk, Analecta Praehistorica Leidensia 25, 117-128. Neolithicum op de donken (Raap-rapport 1234).

Verbruggen, M. & A. Verpoorte Aanvullend Archeologisch Onderzoek in het tracé van de Betuweroute, vindpaats 3. Hardinxveld-Giessendam 1998 De Bruin, Amersfoort. Verhoeven, K. 2003

Palynologisch onderzoek op een veenboring uit de nabijheid van een mesolithische donkbewoning te Hardinxveld-Giessendam De Bruin (Nederland), Leiden (internal report Faculty of Archeology, Leiden University).

Villaret-von Rochow, M. Frucht- und Samenreste aus der neolithischen Station Seeberg, Burgäschisee-Süd, Acta Bernensia 2, 21-64. 1967 Voorrips, A. 1964

A pollenanalytical investigation of a boring made near a river dune in South-Holland, Acta Botanica Neerlandica 13, 237-245.

Waller, M.P., H.A. Binney, M.J. Bunting & R.A. Armitage 2005 The interpretation of fen carr pollen diagrams: pollen-vegetation relationships within the fen carr, Review of Palaeobotany and Palynology 133, 179-202. Weeda, E.J., R. Westra, C. Westra & T. Westra Nederlandse oecologische flora. Wilde planten en hun relaties 1-5, Amsterdam. 1985-1994 Westhoff, V., P.A. Bakker, C.G. Leeuwen, E.E. van de Voo & I.S. Zonneveld Wilde planten, flora en vegetatie in onze natuurgebieden 3, ‘s-Graveland. 1973 Wiel, A.M. van der A palaeoecological study of a section from the foot of the Hazendonk (Zuid-Holland, The Netherlands), 1982 based on the analysis of pollen, spores and macroscopic plant remains, Review of Palaeobotany and Palynology 38, 35-90.

Copyright © 2010. Leiden University Press. All rights reserved.

Wolf, R.J.A.M., A.H.F. Stortelder & R.W. de Waal 2001 Bosecosystemen van Nederland 2. Ooibossen, Utrecht. Woude, J.D. van der 1983 Holocene paleoenvironmental evolution of a perimarine fluviatile area. Geology and paleobotany of the area surrounding the archeological excavation at the Hazendonk river dune (western Netherlands), Ph.D. thesis VU University (Analecta Praehistorica Leidensia 16). Zagwijn, W.H. 1986

Nederland in het Holoceen, Haarlem.

Zapata Peña, L. 2007

First farmers along the coast of the Bay of Biscay, in: S. Colledge & J. Conolly (eds), 189-208.

Zeiler, J.T. 1987

Exploitation of fur animals in Neolithic Swifterbant and Hazendonk (central and western Netherlands), Palaeohistoria 29, 245-263.

271

APPENDIX VII REFERENCES OF APPENDICES

1997 Zeist, W. van 1970 1974

Hunting, fowling and stock-breeding at Neolithic sites in the western and central Netherlands, Ph.D. thesis University of Groningen. Prehistoric and early historic food plants in the Netherlands, Palaeohistoria 14, 41-173. Palaeobotanical studies of settlement sites in the coastal area of the Netherlands, Palaeohistoria 16, 223-371.

Zeist, W. van & R.M. Palfenier-Vegter Seeds and fruits from the Swifterbant S3 site, Palaeohistoria 23, 105-168. 1981 Zohary, D. & M. Hopf Domestication of plants in the Old World, Oxford, 3th edition. 2000

Copyright © 2010. Leiden University Press. All rights reserved.

Zvelebil, M. 1986

272

Mesolithic prelude and Neolithic revolution, in: M. Zvelebil (ed.), Hunters in transition. Mesolithic societies of temperate Eurasia and their transition to farming, Cambridge, 5-15.