Sarah Barbier
Leiden University
THE INFLUENCE OF CLIMATE AND
ENVIRONMENT ON EARLY HOMININS
AT THE MIDDLE PLEISTOCENE SITE OF
‘MARATHOUSA 1’
A BOTANICAL INVESTIGATION CONTRIBUTING TO
A MULTIPROXY APPROACH
1
Cover picture: Marathousa 1 taken from the edge of Area A looking out on the Megalopolis basin (Barbier 2018).
2
THE INFLUENCE OF CLIMATE
AND ENVIRONMENT ON EARLY
HOMININS AT THE MIDDLE
PLEISTOCENE SITE OF
‘MARATHOUSA 1’
A BOTANICAL INVESTIGATION
CONTRIBUTING TO A MULTIPROXY
APPROACH
Supervisor: Dr. M. H. Field Sarah Barbier 1542141 MSc thesisSpecialisation: Archaeobotany and Archaeozoology
Leiden University, Faculty of Archaeology
Final version
3
Table of contents
Acknowledgements ... 5
1. Introduction... 6
1.1 Out of Africa ... 6
1.2 The Gates of Europe ... 8
1.3. Research aims ... 11
2. The context of Marathousa 1 ... 13
2.1 Geological context ... 13
2.1.1. Megalopolis basin ... 13
2.1.2. Stratigraphy and Micromorphology of Marathousa ... 14
2.1.3. Age models of Marathousa 1 ... 16
2.2 The Archaeological context ... 17
2.2.1. Archaeology of the lower Palaeolithic ... 18
2.2.2. Lithics and bone tools ... 18
2.2.3. Evidence for cut-marks ... 21
2.3 Palaeoenvironmental proxies ... 22
2.3.1. Paleoecology of Greece ... 22
2.3.2. Avians and small mammals ... 22
2.3.3. Large Mammals ... 23
2.3.4. Carpological and diatom remains ... 24
3. Methods and techniques ... 26
3.1. In field sampling ... 26
3.2. Description of the sampled section ... 26
3.3. Data generation in the lab ... 28
4. Results... 31
4.1 Pollen data ... 31
4.2 Plant macrofossil data ... 34
4.2.1. Biodiversity and amount of remains ... 34
4.2.2. Indicator taxa and other interesting taxa... 36
4.2.3. Analysis per sample ... 37
4.3 Combining the pollen data with the macrofossil data ... 41
5. Discussion ... 44
5.1. Interpreting the vegetational and environmental changes at Marathousa 1 ... 44
5.1.1. The landscape of Marathousa 1 ... 44
4
5.1.2. Trampling and its effect on the vegetation ... 47
5.2. The climate of Marathousa 1 and the Megalopolis basin and its potential... 48
5.2.1. The climate of Marathousa 1 ... 48
5.2.2. Influence of climate on hominins ... 49
5.3. Potential exploitation of the landscape ... 50
5.3.1. Animal exploitation ... 50
5.3.2. Plant exploitation for food ... 51
5.3.3. Plant exploitation for tools and fire ... 52
5.4. Limitations of the research ... 54
5.4.1. Sampling strategies... 54 5.4.2. Taphonomy ... 54 6. Conclusions ... 56 Abstract ... 58 Bibliography ... 59 Internet pages ... 64 List of Figures ... 64 List of Tables ... 65 List of Appendices ... 65
5
Acknowledgements
I would like to personally thank Dr. Field for his guidance and help. He has learned me everything I know about archaeobotany and remained helpful with my thesis when I was rushing to get it finished. He was always available and never asked me to come a different time. Thank you for all your help.
I also want to thank everybody who worked on the Marathousa 1 samples, especially Lars den Boef and Oda Nuij, who both kindly provided me with the information they already generated. I would like to thank Rory Granleese for editing and revising my language and structure and for his support throughout this period. Lastly I would like to thank all the people who took the time to discuss the topic of my thesis with me during breaks, this helped me think about the subject (you know who you are).
6
1. Introduction
Climate change and environmental change are important topics in modern society and it is commonly agreed that these factors have a huge impact on human society. Not only do they play a role in contemporary society, but they have also played a major role in animal and plant evolution, and almost certainly in hominin evolution. Climate and environment have been driving forces in changes in morphology, distribution and migration, of groups of living organisms. The impact of climatic and environmental change should not be underestimated. This topic is particularly important in prehistoric archaeology. When examining early migrations of hominins across the globe, the archaeology should always be considered alongside the environmental and ecological context (Muttoni et al. 2018, 4).
1.1 Out of Africa
The hypothesis that human evolution took place mostly in Africa is long standing. The earliest evidence of hominin presence comes from Africa and up until 1.8 million years (Ma) ago all hominin evidence was found in the African continent. The first evidence of hominin migration out of Africa occurs in the Early Pleistocene (from around 2.58 Ma to around 0.78 Ma), but the exact timing of it remains unclear (Macdonald et al. 2012). The migration of hominins into Europe has been vigorously discussed by many specialists from different fields (Muttoni et al. 2018; Roebroeks et al. 2006; Macdonald et al. 2012). The evidence displays huge temporal gaps, making it difficult to precisely reconstruct a route or timeline. Climate changes at around 1.8 Ma induced by Heinrich events1 were
likely very influential to hominin migration out of Africa. Aridification and cold temperatures that are associated with these climate changes would have rendered large parts of the African continent inhabitable for hominins, forcing them to relocate (Carto et
al. 2009, 149).
Early migration into Europe seemingly took place around 1.7-1.8 Ma. The earliest site found in Europe is Dmanisi in Georgia, with a date of 1.8 Ma. It contains a great number of hominin fossils, however the species definition is still debated. In addition to this, a number of mode 1 tools2 were also located at this site (Ferring et al. 2012, 10432). After
1 Heinrich events are phenomenon where large armades of icebergs break off from glaciers and this
causes fresh cold water pulses in the sea disturbing the thermohaline circulation and causing climatic fluctuations.
2 Mode 1 tools are simple tools with one or two flakes chipped off and are part of the Oldowan
7
Dmanisi, the oldest site known in Europe is Sima del elefante. This site is part of the Atapuerca cave system located in southern Spain (Cuenca-Bescós et al. 2013, 29). These sites indicate an early migration into Europe. However, there are only a small numbers of these early sites, and they are only present in southern Europe and the Balkans. Considering the small number of sites this migration event into Europe is often interpreted as a short-dispersal event, which did not lead to permanent migration into Europe (Roebroeks et al. 2006, 428). These early dispersal events in southern Europe are often correlated with climate, suggesting that the surrounding climates and thus the environments were only suitable in southern Europe at that time, rendering this area a refugia. In essence a refugia is an area where a certain species or several species survive during an entire glacial-interglacial cycle. On the base of zoological and botanical finds, the southern peninsula of Europe (Iberia, Balkans, and Italy) are interpreted as refugia (Stewer and Stringer 2012, 1317).
Between 1 and 0.6 Ma a thinly spread presence in southern and western Europe is attested for. Most sites in this period are found in southern Europe and migration into the Northern latitudes cannot be confirmed (Roebroeks et al. 2006, 429). A major climatic transition occurred around 0.9 Ma, which is interpreted as being a part of Marine Isotope Stage 22 (MIS)3. This cold phase led to a drop in sea levels and created a conduit formed by the Po
Plain in Northern Italy and the Danube Plain in Romania, Serbia and Hungary. Because of the sea level drop these river valleys became continental plains that were highly suitable for large grazing mammals (Muttoni et al. 2018, 7).
It has been attested by many early hominin sites that hominins depended on animal proteins and animal fat from large herbivores to survive. In addition, sites that contained mammal bones with cut-marks associated with butchering practices, confirm the theory that hominins were dependent on large mammals (Muttoni et al. 2018, 9-10). The continental plains, formed around 0.9 Ma, in the Levant and South-eastern Europe most probably attracted large herbivores, because of their open grasslands and reduced woody cover. It is very likely that hominins followed large herds of mammals in order to survive (see figure 1). This theory is called the Galerian migration hypothesis (Muttoni et al. 2018, 11). Considering the evidence presented above it is highly possible that hominins migrated into Europe around 0.9 Ma through the Levantine corridor into southern Europe (Tourloukis 2010, 21).
3 Marine Isotope Stages are alternating warm and cool periods. Even numbers represent glacials,
8
Figure 1. A geographic map showing key sites of Early Pleistocene hominins and Galerian mammals (Muttoni et al. 2018, 2).
1.2 The Gates of Europe
The Galerian migration hypothesis holds that early hominins entered Europe around 0.9 Ma, together with other mammals (Muttoni et al. 2018, 4). This hypothesis suggests that the early hominins immigrated from Africa through the Levantine corridor. During the early and middle Pleistocene the Aegean sea would not have always acted as a barrier. On the contrary, the landmass was probably highly accessible during periods of low sea level (see figure 2). This would have created a passage to the southernmost part of the Balkan and the Greek peninsula (Tourloukis 2010, 21). Considering this accessible route into southern Europe, it is remarkable that Pleistocene sites from this region are lacking (see figure 3). The Greek peninsula is not only fit for hominin presence because of the likely dispersal routes, but also because it sustained refugia for floral, faunal, and possibly human populations during periods of climatic stress (Tourloukis and Karkanas 2012, 2).
9
Figure 2. Palaeographic reconstruction of the Aegean and Ionian Seas during several glacial stages (Tourloukis and Karkanas 2012, 10).
10
The evidence of hominin presence in Greece in the Pleistocene is very scarce. The exceptionally well preserved cranium from the Petralona cave is one of the clearest evidences of hominin presence in Greece. The skull dates to between about 150 and 250/350 thousand years ago (Ka). The only other potentially Middle Pleistocene hominin remain is an isolated upper third molar from the Megalopolis basin (Tourloukis and Karkanas 2012, 2). Several sites with lithic remains have been found in Greece from the lower Palaeolithic, like Rodafnidia, which is an open air site associated with lithic artefacts that occurred inside a deposit of river conglomerates (Tourloukis and Harvati 2018, 50). However, these sites do not contain any hominin remains.
Greece has become a larger focus for paleoanthropologists and new sites are being researched extensively. One of these sites is Marathousa 1, an open air site situated in the Megalopolis basin. It is located in one of the lignite mines present in the basin (see figure 4).
Figure 4. Map of the Megalopolis basin and the location of Marathousa 1 in the lignite mine (Panagopoulou et al. 2018, 34).
Marathousa 1 (MAR 1) was discovered in 2013 through a survey initiated by a collaboration between the Ephoreia of Paleoanthropology and Speleology (Greek Ministry of Culture) and the University of Tübingen in the framework of the ‘Palaeoanthropology at the Gates of Europe’ (PaGE) Project (Thompson et al. 2018, 22). Systematic excavations
11
have revealed the presence of lithic artefacts in correlation with faunal remains, including the cranium and several postcranial elements of a straight tusked elephant (Palaeoloxodon antiquus) (Tourloukis and Harvati 2018, 49). The straight tusked elephant and other faunal remains contain cut-marks associated with the lithics found at the site. This suggests hominin presence and indicates that hominins were exploiting (or even hunting) the fauna in the area (Panagopoulou et al. 2018, 39).
1.3. Research aims
Environment and climate hugely impact humans, and as earlier stated it can influence complete mammal migrations. The focus of this research lies on the Marathousa 1 site, since this site provides key data that can potentially fill in the hiatus of knowledge regarding what we know about hominin migration into Europe. The environment and climate of the site before, during and after the archaeological horizon are incredibly important to study in order to understand the nature of the site. The environment and climate can help us understand why hominins were present in this area around that time. This has led to the following research question:
What was the potential for exploitation and habitation for hominins and large mammals in the local and regional environment of Marathousa 1, during the Middle Pleistocene? In order to answer this main research question a number of sub-questions have been posed: What was the local and regional vegetation of Marathousa 1, as seen in the sequence? What plants are present that have been exploited by hominins and what change, if any, is visible in the sequence?
What animals have been found that show hominin exploitation?
What can be inferred about the climate during the sequence and what effect could that have on hominin or large mammal habitation of the area?
In order to answer these questions two datasets are created and used (they are presented in the results chapter). The first dataset is a pollen data set of a sequence of 60 cm sampled in the vicinity of the skeleton of the straight tusked elephant. This dataset has been created by Lars den Boef MSc, for his master thesis and is used with permission of the author. The other data set is a macrofossil dataset taken from the same sequence as the pollen data set. This set was generated and processed by the author. The reason for using both botanical macrofossils and microfossils, is related to the differences in taphonomic processes for both types of botanical data. Pollen, which are a type of botanical microfossil, are
12
generally distributed very far, as they weigh very little and can be transported by the wind, water or animals. Because of this large distribution area, pollen can originate from areas that are over 100 km away. They do not indicate the local vegetation, but instead give an idea of regional vegetation. Pollen can often not be identified with taxonomic precision and are mostly identified up to genus level. Macrofossils on the other hand can identify up to species level very often and thus using them to determine the local vegetation and indicating local environments is very useful. Some plant species have poorly preserved pollen, whilst their macrofossil remains do occur in the sedimentary record (Birks et al. 2000, 31). Using both macrofossil and pollen data is incredibly useful to reconstruct a complete environment with both the local vegetation and regional vegetation. Moreover, it is also good scientific practice to include more than one type of data, to strengthen the results and integrate different forms of knowledge. On that note, the zoological data that has been previously studied for this site will also compliment this research.
13
2. The context of Marathousa 1
The site of Marathousa 1 has been extensively researched from 2013 onwards. In this chapter previous research conducted at the site is presented as well as background information to support this. This information will provide the context necessary to fully understand the botanical data presented in chapter 4 and the background information will be regarded in the discussion.
2.1 Geological context
The geology of Greece is incredibly complex due to its location at the junction of the European and African tectonic plates. This causes a lot of seismic activity throughout the region, disturbing sedimentation (Higgins and Higgins 1996). Below the geological context of the Megalopolis basin and Marathousa 1 are presented as well as its age models.
2.1.1. Megalopolis basin
The Megalopolis basin is located in the middle of the Peloponnesus peninsula in the south of Greece. It is an intramontane lacustrine basin, meaning that the basin lies in between mountains and consists of lake environments. The basin was partly formed by NE-SW trending normal faults4, that have been active since the Miocene (Karkanas et al. 2018,
124; Van Vugt et al. 2000, 70).
The Pliocene sequence consists of lacustrine marls of the Markrision formation and this is followed by the sediments of the Trifolon formation which are fluvial, which in turn is followed up by the Apiditsa alluvial fan formation (table 1). The Choremi formation consists of two members, the Megalopolis member and the Marathousa member, the latter consists of a lacustrine sedimentary environment. The Marathousa member has a cyclic sedimentation of dark lignite and bluish gray mud, fine sand, and in some cases marl sediments (Van Vugt et al. 2000, 70).
Formation Member Sedimentary
environment
Period
(Holocene) Fluvial terrace Holocene
Thoknia Fluvial terrace Pleistocene
Potamia Fluvial terrace Pleistocene
Choremi Megalopolis Marathousa Fluvial Lacustrine Pleistocene Pleistocene
Apiditsa Alluvial Fan Pleistocene
4 Normal faults are breaks in the rocks, due to this fracture one part of the rock sinks down relative to
14
Trilofon Fluvial Pliocene
Markrision Lacustrine Pliocene
Table 1. Stratigraphy and chronology in the Megalopolis basin (after Karkanas et al. 2018, 124).
Lignite mining has been done at the Marathousa member since 1969. The lignite is used as an energy source for the powerplant next to the mine. Due to the mining activity up to 150 m of the Marathousa member has been exposed, making it an ideal place to study Middle Pleistocene hominin presence. There are four lignite seams present in the Marathousa member, each of them are between 10-20 meters thick and they are labelled I-IV from the bottom up. The lignite seams alternate with 12-25 meters thick layers of sand, silt and clay. The lignite layers are thicker in the western part of the mine and wedge out towards the East. The detrital layers do the opposite and are thicker in the East and wedge out towards the West (Van Vugt et al. 2000, 80) (see figure 5). According to palynological research done on lignite seam I, the lignite seams were deposited in interglacial periods, whilst the detrital layers were deposited in glacial periods (Okuda et al. 2002, 152).
Figure 5. Schematic cross section of the Megalopolis basin showing the lignite layers (van Vugt et al. 2000, 72).
2.1.2. Stratigraphy and Micromorphology of Marathousa
The archaeological site of Marathousa 1 has been found due to the mining activity, which resulted in archaeologists and paleoanthropologists only having to remove four meters of sediment to get to the archaeological level. The find bearing horizon is situated in between lignite seam II and III. The detrital bed in which the find bearing horizon lies is made up of lacustrine clay, silt and sand beds. Ostracods and freshwater bivalves are found in this detrital bed, indicating a freshwater environment (Okuda et al. 2002, 145).The excavation of Marathousa 1 has been divided in two areas, Area A and Area B. Area A lies towards the North of the basin and Area B towards the South, there is 60 meter inbetween the two excavated areas. Both areas contain lithics and zoological remains, however Area A has a greater quantity of zoological remains and contains an almost complete skeleton of Palaeoloxodon antiquus (the straight tusked elephant mentioned previously) with cut marks. Area B is characterised by lithic tools and zoological remains with cut marks. The stratigraphic units are separately named for Area A and Area B (UA
15
and UB) and then numbered. This means that in the two areas the names of stratigraphic units can have a different name, while they can be correlated with each other (Karkanas et
al. 2018, 125).
Area A and B show very similar sedimentary units, however Area A consists of a relatively thinner sedimentary sequence than Area B. Depositions seem to have been continuous in both Area A as well as Area B, however one major erosional contact has been identified. This contact lies between UB10-6 and UA7 to 4 and the environment of deposition was quite variable. In the South of Area B the deposition was mostly characterized by grey to bluish organic-poor silts and sands, and the area between A and B mostly by silty sands with fluctuating amounts of organics. The lower part of the erosional contact shows massive organic rich clayey sands and graded bedded sand/organic-rich silty sands. It has a low organic input of only (<4-5%). The upper part of the erosional contact is characterized by depositional units bounded by erosional contacts, which consists of coarser sediments at the base and progressively finer sediments at the top (Karkanas et al. 2018, 125-126).
The skeleton of the elephant in Area A has been found lying on top of the eroded boudinaged layer that consists of bluish gray, massive muddy sand with load deformation structures (UA4). It is interpreted as being sedimentation caused by hyperconcentrated flows deposited into ephemeral lake-fringing marsh areas, occasionally affected by waves and currents. The skeleton is buried by the first depositional mud flow unit (UA3). In Area B the find bearing layer lies on top of UB5 and is buried by UB4, which is at the erosional contact layer (Karkanas et al. 2018, 126-128) (see figure 6).
16
Figure 6. Stratigraphy of Area A and Area B with the finds bearing horizon indicated and the correlations between major stratigraphic markers (Karkanas et al. 2018, 126).
UB4 and UB5 can be correlated due their interpretation of facies and sedimentological observation, which suggest that both these units were deposited by mudflows. These mudflows were characterized by variable concentrations and grain-aligning capacity. Their overall interpretation is consistent with the rest of the stratigraphy of the area (Karkanis et
al. 2018, 134). According to Tourloukis et al. 2018 and Karkanas et al. 2018, the
archaeological units must have been deposited close to the western paleolake margin and they must have been influenced by variable energy and density of sedimentation events (mudflows), which came from the northwest and flowed in the form of turbidites5 to the
east of the paleolake depocenter (Tourloukis et al. 2018; Karkanas et al. 135).
2.1.3. Age models of Marathousa 1
There are two age models that have been considered for the Marathousa 1 stratigraphy, which are both based on the correlation of the Brunhes/Matuyama boundary and 6 chronostratigraphic surfaces. These chronostratigraphic surfaces consists of six identified lignite seams that were represented on the drill core logs as black bars. In order to correlate the lignite detritus alternations to glacial-interglacial cycles, two assumptions need to be
17
made. The first assumption is that every individual lignite unit or seam represents a full interglacial stage. The second assumption is that a single lignite unit or seam represents more than one interglacial or interstadial stage (Tourloukis et al. 2018, 162).
The first age model that has been made on the base of these assumptions is the most preferable. It correlated the LIa (Lignite layer Ia) with MIS 19, based on the position of the Brunhes/Matuyama boundary (0.78 Ma). This means that the base of LIa represents Glacial Termination IX, which is dated to around 0.79 Ma. According to this age model the archaeological layers (UB4c-UB5 and UA3c-UA4) combined should have an age of between 0.48 Ma and 0.42 Ma. This date agrees well with radiometric assays, which dated between 500 Ka and 400 Ka. Electron spin resonance (ESR) dates from mollusc samples in UA2 show a minimum age of 400 Ka, which agrees with the other dates considering UA2 is overlying the find bearing layer (Tourloukis et al. 2018, 161-163).
Due to these other dates confirming the hypotheses, the date of between 0.48 Ma and 0.42 Ma for Marathousa 1 has been chosen as the preferred option (Tourloukis et al. 2018, 163-165). These dates place Marathousa 1 in the Middle Pleistocene (see figure 7) and in the Lower Palaeolithic.
Figure 7. Subdivisions of the Geological time periods in the Quaternary (www.wikipedia.nl).
2.2 The Archaeological context
In this paragraph an overview of archaeology in the Palaeolithic is given. The archaeological data from Marathousa 1 is presented below and gives information on the lithics, bone tools and the cut marks that have been found.
18
2.2.1. Archaeology of the lower Palaeolithic
The Acheulean is a Lower Palaeolithic complex that is mostly based on bifacial handaxe technology. It follows up on the older Oldowan technology (Rocca et al. 2016, 2). Acheulean artefacts are found throughout Africa, the Near-East, the subcontinent of India and several places in Europe and range from c. 1.6 Ma to less than 200 Ka (Lycett and Gowlett 2008, 295). The complex supposedly mostly consists of the bifacial tradition of handaxes, however the technology necessary to create these tools can also be reflected in different tools, that then also belong to the Acheulean. Especially sites in central and southern Europe during the period between 0.8/0.7 Ma and 0.5 Ma, show a sparse presence of handaxes and Large Cutting Tools (TCL’s) and a larger presence of flakes and small tools (Rocca et al. 2016, 408).
Lithic variability in the Middle Pleistocene of Europe is still under investigation. Around 900 Ka the first instances of the Acheulean technology appeared. After these first instances there is a scarcity of sites containing Acheulean artefacts and a gap can be observed up until around 600-500 Ka. In this time period the Acheulean is well established in Europe and the hominins going into Europe have an archaeologically visible momentum and also grow with important behavioral innovations (Tourloukis et al. 2018, 47).
2.2.2. Lithics and bone tools
At Marathousa 1 lithics were recovered from both Area A and Area B, however Area B consists of a significantly larger amount of lithics. Excavation Area A yielded a small amount of artefacts and an even smaller number of tools. 4 flakes were found in this area together with 63 chips and debris, which were found in close proximity to the elephant (Tourloukis et al. 2018, 51).
In Area B the bulk of the lithic assemblage was found, over 1100 pieces of lithics. Next to a higher amount of lithics Area B also consists of a much higher degree of fragmentation of fossils, that also show cases of anthropogenic influence, fractures, percussion marks and cut-marks. In both Areas the lithics are found in stratigraphic association with the faunal fossil remains and in both Areas this includes elephant bones (Tourloukis et al. 2018, 50-51).
The raw materials for the lithics found at Marathousa 1 are, in order of frequency of occurrence, red radiolarite, brown radiolarite, quartz, limestone, grey flint and black flint. Small radiolarite pebbles were used for most of the lithic material. The primary sources of this material were at the fringes of the Megalopolis basin. The accessibility to this material was easy and considering its locality this must have been the best option for the hominins in the area. However, the radiolarites mechanical properties are of low quality in most
19
cases of the lithics. The inclusion of quartz and cleavage plane fractures in the radiolarite make it more prone to fractures. Considering the close proximity and the abundance of the sources of radiolarite (4-5 km from the site) which were chosen most frequently by early hominins at the site, suggests that the other ‘better’ lithic materials were harder to find or the sources were too far away. It is suggested that the black flint was introduced in the form of finished artefacts, instead of being created locally (Tourloukis et al. 2018, 50). The lithic assemblage consists of chips (63.3%) and flakes (15.1%). 135 specimens were analysed on edge damage and from this 13% was classified as having possible edge damage. The use-wear traces on these lithics are commonly found on the working edges of the tools and consists of chipping scars that resemble micro-nibbling. This all points towards a “small tools” lithic industry. The tool kit consists of minimally prepared small cores and the selection of flakes for use, which were either not worked at all or slightly modified with retouch to produce a notched or pointed edge (see figure 8). Backed edges are also visible in the lithic assemblage, which suggests that manual prehension or hafting took place. The Marathousa 1 assemblage shows evidence for a special-purpose site, where knapping was focused on creating tools and maintaining them (Tourloukis et al. 2018, 54-56).
20
Figure 8. Several retouched tools found at Marathousa 1 (Tourloukis et al. 2018, 54).
Next to the vast evidence of the lithic assemblage accumulated at Marathousa 1, evidence for organic technology has been found in the form of bone tools. The evidence consists of bone flakes, that show a very high similarity to the lithic flakes, bone tools and a bone percussor6. The bone flakes and tools are identified as anthropogenic by the flake scars,
cut marks and percussion marks that are found on them. However, this evidence is still preliminary, more research need to be done to actively confirm that these were in fact from an anthropogenic origin (Tourloukis et al. 2018, 56-58).
One of the most peculiar finds is the bone percussor. It is made from a diaphysis fragment of a large mammal. Percussion marks found on the tool suggest it is a soft hammer (see figure 9). The bone tools indicate that hominins at Marathousa 1 must have had a good
21
understanding of bone properties and knapping techniques alike (Tourloukis et al. 2018, 59).
Figure 9. Bone tools from Marathousa 1. 1) bone flake; 2) denticulated bone flake; 3) bone percussor (Tourloukis et al. 2018, 59).
2.2.3. Evidence for cut-marks
Cut marks were found at Marathousa 1 on elephant bones in Area A as well as on elephants and other mammals in Area B. In Area A an astragalus (tale bone) and tibia (shinbone) of an elephant show cut marks. The astragalus shows parallel striations on the distal side of the bone. The tibia presents a single wider groove that is orientated to the long axis of the bone. The characteristics of these cut marks are indicative of cut marks created during exploitation of elephant carcasses. Carnivore gnawing is not found on the elephant bones present in Area A (Konidaris et al. 2018, 71).
Elephant bones in Area B show cut marks as well, one of which is a broken rib fragment that contains a single V-shaped mark and a cluster of scrape marks. Next to elephants ungulates seem to have been exploited as well. A distal radius epiphysis (part of the wrist bone) of a fallow deer shows four short parallel striations. The spiral fractured humerus diaphysis (main section of the long bone in an arm) fragment of a small to medium sized mammal, contains oblique cut marks relative to the axis of the bone (Konidaris et al. 2018, 71-72).
The cut marks found on elephants and other mammals at Marathousa 1 indicate butchering activities by humans. This is in line with the lithic industry found at the site and its spatial association with the mammal bones (Konidaris et al. 2018, 81).
22
2.3 Palaeoenvironmental proxies
This paragraph presents several paleoenvironmental proxies that have been studied previously at Marathousa 1. In order to introduce these, the paleoecology of Greece is roughly presented and explained. This information is necessary to reconstruct a complete picture of the environment of Marathousa 1.
2.3.1. Paleoecology of Greece
Multiple long term sequences have been studied in Greece, to understand the fluctuations of climate on a large scale. Orbital forcing is the effect on climate of the slightly tilt of the Earth’s axis, which influences the Earth’s orbit around the sun. Essentially this causes the cycles of glacial-interglacials (van Vugt 2000, 9). The Pleistocene is known for its climatic fluctuations and in the past these climatic fluctuations were identified by sedimentological research. In a number of researches of paleolake environments in Greece from van Vugt 2000, the stratigraphy would show that the lake levels fluctuate between being a dry reed swamp and an actual shallow lake. When the climate was cold the environment would become dryer and the lake would dry out letting the reed swamp establish, however when the climate warmed up and precipitation would increase, the water levels would become higher and drown the reed swamp, resulting in the lignite layers (van Vugt 2000, 13). The megalopolis basin is one of these fossil lakes, that shows the warm-cold cycles in its stratigraphy. However, to further enforce this idea palynological research of the lignite-detrital layers is necessary. Okuda et al. 2002 did palynological research on a section of the Marathousa member at the Megalopolis basin, including 4 lignite seams. The research concludes that the lignite layers are indicative of an interglacial with oak forests and the detrital layers in between were interpreted as glacials, because of the large amount of
Artemisia pollen that indicate dry cold steppe environments (Okuda et al. 2002, 155).
The site Marathousa 1 lies in this member in the middle of the detrital layer in between lignite layer II and III. According to the work of Okuda et al. 2002, it should represent a glacial period, with a cold dry environment.
2.3.2. Avians and small mammals
A significant number of zoological remains have been found at Marathousa 1. Mostly large mammals, but avians and small mammals were also researched.
23
The avian assemblage in Area A is characterized by a large quantity of Common Teal (Anas crecca), this bird lives around freshwater, mostly small lakes and marshes with well vegetated margins. Its diet consists of seeds of aquatic plants and grasses during winter, during summer and spring it consumes aquatic invertebrates. Area B is characterised by an abundance of Shovelers (Spatula clypeata), which live in a range of freshwater wetlands and eat a variety of foods. The Mallard (Anas platyrhynchos) is also very abundant in Area B, it lives in rich shallow wetlands with emergent vegetation. Its diet consists of aquatic and terrestrial plants, invertebrates and in some cases amphibians and fish (Michailidis et
al. 2018, 7).
There is a clear predominance of water birds present in the Marathousa 1 assemblage. Most of the species are adapted to shallow waters and a temperate climatic zone. Interestingly some species are present that prefer a warmer climate, like the Purple Swamp Hen (Porphyrio porphyrio) and the Darter (Anhinga sp.). These might indicate a refugium status for Marathousa 1. Overall the avifaunal composition reveals a lakeshore environment, rich in emergent vegetation and with some marshy areas and trees in the near vicinity (Michailidis et al. 2018, 8) .
There is only a small number of small mammals present in the Marathousa 1 assemblage. However, more than 80 molars of Arvicola have been found and researched. The assemblage is dominated by the Water Vole (Arvicola) that inhabits semi-aquatic habitats and usually stays within 2 metres of the water edge. It is common in marshes, around ponds and lakes. Another abundant species is the Common Vole (Microtus arvalis), it prefers more open landscapes that are well-drained. An unconfirmed species of the genus Alactaga suggests the presence of areas with steppe vegetation in the vicinity of the lake.
The small mammal assemblage does not add majorly to the palaeo-reconstruction, more research is needed to have a full understanding of this zoological group (Doukas et al. 2018, 12).
2.3.3. Large Mammals
The Marathousa 1 site is characterized by the find of a skeleton of the Straight Tusked Elephant (Palaeoloxodon antiquus). Bones of other Palaeoloxodon individuals were also found in Area A. The Palaeoloxodon antiquus is very ecologically flexible and has a wide ecological adaptation. It inhabits mild humid to warm to warm-temperate regions and prefers a moderately woodlands, but also inhabits arid grasslands. It occurs in the Pleistocene in Northern and Central Europe in interglacial stages, and during glacials it
24
occurs in Southern Europe, which acted as a refugium. The Straight Tusked Elephant is a mixed feeder, but mostly consumes grasses (Konidaris et al. 2018, 11-12).
The Marathousa 1 assemblage consists in a large part of semiaquatic large mammals like,
Hippopotamus, Castor and Lutra. Especially Hippopotamus antiquus indicates a rather
aquatic habitat, considering its high dependence on water. Its diet solely consists of aquatic vegetation, meaning it can only inhabit an environment with a large water body. The Otter (Lutra) has a diet predominantly consisting of fish, making it dependant on a water body as well. However, it spends its time on land resting. The Eurasian beaver (Castor fiber) is also a semiaquatic species that inhabits ponds, lakes, rivers and streams.
Cervus elaphus lives mainly on the edges of woodlands and consumes the foliage of trees,
grasses and low herbaceous vegetation. It can inhabit different ecologies, and can occur in woodlands, or open or even treeless environments. It occurs in both interglacials and cold stages.
Fallow deer (Dama sp.) prefers more warm temperate environments and is less tolerant to open and cold conditions. It mostly occurs in open-deciduous or mixed woodlands. The presence of this animal might further confirm that the Megalopolis basin acted as a refugium in glacial stages.
Overall, the large mammal fauna indicates a temperate climate, with a substantial woodland component and it indicates the presence of a large freshwater body (Konidaris et al. 2018, 12).
2.3.4. Carpological and diatom remains
The carpological and diatom remains researched at Marathousa 1 are dominated by aquatic, waterside and damp ground plant taxa. The results suggest that the straight tusked elephant was deposited in a marginal reed swamp just of the shore of a lake. The sediment was deposited in still water with low suspended sediment. The presence of Alnus and Salix in the wood assemblage suggests that there was a flat damp area (mudflat) present at the margins of the lake (Field et al. 2018, 12).
Taxa such as Palmae in the charcoal and phytolith record, as well as carpological remains of taxa such as Salvinia natans, indicate that the climate at the time of death of the elephant was warm. However, previous research suggests that the clastic sediments represent glacials and the lignite seams represent interglacials. Warm conditions were already prevailing during deposition of clastic sediment suggesting that there is no abrupt change from cold to warm at the boundary of the lignite seam (Field et al. 2018, 14).
25
In the palynological research of Okuda et al. 2002, the results showed that the detrital units represented cold conditions, due to the high number of Artemisia pollen found. The domination of Artemisia suggests a dry, cold steppe environment and arboreal taxa are suppressed (Okuda et al. 2002, 152). It is interesting that the Marathousa 1 data shows warm conditions during the deposition of the detrital units and that the pollen data from the area shows colder, dry, conditions. This will be discussed with the results of this research in chapter 5.
26
3. Methods and techniques
In this chapter the methodology of this research will be described and explained. This is specifically for the obtaining of the macrofossils, since the author did not prepare or identify the pollen samples.
3.1. In field sampling
The samples for this research have been taken from the North facing section of Area A. The precise location is within the southern margin of excavation Area A and is right next to the excavated bones of an ancient elephant (Palaeoloxodon). Six bulk samples were taken from the section, each 10 cm in height and around 2 liters in volume. Extraction of the samples was done from the bottom up, in order to minimize disturbance of the sediments. After extraction the samples were each put into a separate plastic bag that was sealed. They were then transported to Leiden University, the Netherlands, where they were stored in a modified fridge.
3.2. Description of the sampled section
As has been noted before the find bearing layer in Area A has been found on top of UA4 and covered with UA3. This research strives to understand conditions before this archaeological layer and after, thus a bigger section has been taken of approximately 65 cm, encompassing both UA4 and UA3 in its whole. Two faces have been noted, the main (lower) face and the back face at the top. The samples are numbered from 1 to 6, from top to bottom (see figure 10).
27
Figure 10. Sampled section in Area A. The lines indicate the boundaries od the different samples. The red line indicates the contact between UA3c and UA4 and the white line represents a contact or desiccation crack. The black diamond is at an elevation of 349.48masl (Nuij 2018, 20).
Sample 1 (0-10 cm) consists of dark brown organic sediment with fine sand, it mostly consists of the back face, but partially from the main face as well. It is unsure if the sediment from this sample belongs to UA3b or UA3c.
Sample 1 and sample 2 are separated by an erosional contact or desiccation crack. The exact height has been measured right at this contact at 349.48 meters above sea level. Sample 2 (10-20 cm) is made up of dark brown organic sediment with reddish/brown patches and mollusc shell fragments.
Sample 3 (20-30 cm) is roughly the same as sample 2, except it does not contain any mollusc shell fragments and it contains grey clay ‘clasts’ which are deformed.
28
Samples 2, 3 and 4 all belong to UA3c. Sample 4 (30-40 cm), is also similar to sample 2 and 3, with the exception of the mollusc shell fragments. Next to the reddish/brown patches, grey clay ‘clasts’ also occur in sample 4.
Between sample 4 and sample 5 (40-50 cm) there is a clear contact, which is visible because of the occurrence of reddish brown smears and lots of compressed wood. This is the contact were the elephant remains were found, on top of sample 5.
Samples 5 and 6 (50-60 cm) are made up of dark brown fine grained organic sediments, without any reddish/brown patches or mollusc shell fragments. Some instances of grey clay ‘clasts’ have been found in the matrix. Both sample 5 and 6 are attributed to UA4. UA4 is interpreted as an eroded boudinaged7 layer that consists of bluish grey, massive
muddy sand with load deformation structures and the overlying layer UA3c has been interpreted as a depositional mud flow.
3.3. Data generation in the lab
The bulk samples that were taken in the field, c. 2 liters in volume, contained too much sediment to process in one research period. In order to complete the entire sample a significant increase in time allotted to study the samples would have been necessary, and the results would not necessarily have generated new or additional data, when contrasted with a ‘representative subsample’. For this reason subsampling found place in the Laboratory for Archaeobotanical studies at the Faculty of Archaeology from Leiden University. A subsample of 200 cc was obtained by using a glass cylinder of 500 ml and filling this up to 200 ml with water, then the sediment was poured in until the water-sediment mixture came up to 400 ml. This was done for all six bulk samples. The remains of the bulk samples were sealed in their plastic bags and put in the fridge for safekeeping. These raw samples are useful to keep in case of sub-sample contamination or a different scope of research. The subsamples (which from now on will be named ‘samples’) were poured in glass beakers and washing up liquid was added in order to help the sediment break apart. It is worth noting that adding washing up liquid removes the opportunity to carbon data any of the samples; again, this is why it is useful to keep the remaining raw samples. The samples were kept inside a plastic bucket with lid at room temperature for a week, in order to make the sediment fall apart and get the samples ready for sieving.
7 Boudinage is when a rigid tabular body is stretched and deformed, resulting in “sausage-shaped”
29
After letting the samples soak for a week they were sieved through a nest of sieves with sizes of 1 mm, 500μm, 250μm and 150μm. Sieving was done with tap water at a special sink, with an area to rest the nest of sieves on. No rubbing, or breaking apart of the sediments has been done by hand or with any other tool, in order to preserve the very old fragile organic material. Instead of rubbing the sediment through, the sieves were shaken lightly in order for the sediment to fall through. However, this first instance of sieving did not separate all of the sediment from the organic remains, so a second round of sieving was necessary. In between sieving the sediments were poured in separate beakers for each sieve size, so that microscope work would be easier. They were filled with water to prevent oxidising. After sieving each sieve was carefully cleaned with tap water and a brush to prevent contamination between samples. The largest fractions of the samples were soaked for longer and sieved multiple times, because of the larger lumps of sediment that took long time to break down. In some specific cases fractions of samples were re-sieved at certain points in time for various reasons, this has no effect on the results, but its sole purpose was to make the microscope work easier.
Acquiring the identifiable macro-remains was done with a Leica binocular with a magnification of 6.3 to 40 times. Acquired data includes seeds, nutlets, endocarps, sporangia and megaspores, these were collected by pouring sediment from a certain fraction into a plastic petridish (adding water if necessary) and systematically scanning this picking out remains that are possibly identifiable. The picking of the macro-remains was done with different tweezers or a soft bristled brush, depending on the size and vulnerability of the remains in question. The macro-remains would then be placed into a petridish with a filter paper and preservation liquid on it. The preservation liquid consists of one third water, one third glycerine and one third alcohol.
Eventually when all the macro-remains were picked out of the sediment they were lined up in groups and determined. Determination was done with the help of the Digital Seed Atlas by Prof.dr. René Cappers, the reference collection of the Archaeobotanical lab of the Faculty of Archaeology and the consistent in-house expertise of Dr. Michael Field. Literature for information on ecology has been used in this research and is referred to in the text. Identification where possible has been done up to species level, however this is not always possible due to preservation issues. In these cases the remains will be determined up to genus or family level. In some cases when the author could not be 100 percent sure of the determination, but it is highly probable that it is a certain species, the word cf. has been used. In some cases an identification could not be made and these were marked as unidentified and saved for future advancements.
30
All of the identified remains have been counted and put into an excel table as well as in a notebook. The remains themselves are kept in tubes per species per sample with preservative liquid. These tubes will be available at the Archaeobotanical lab at Leiden University for further investigation in the future.
31
4. Results
In this chapter the results of both the pollen analysis and the macrofossil analysis are presented. Diagrams and tables are used to present the data in a clear and understandable way. The table with rough data has been added in the appendix 1.
4.1 Pollen data
The pollen analysis performed by Lars den Boef MSc resulted in the diagram below (figure 11). This data originates from the same section as the macrofossil data and can thus be confidently correlated.
The diagram shows the frequency of the pollen grains through the profile. The taxa are grouped into ecologies based on their primary habitat. It is important to note that several plant species can occur in different habitats, these are put under the group indeterminable. 66 different taxa are found through the pollen data of which most taxa are non-arboreal (‘non-forest’) pollen. However, the arboreal pollen are presented well and 12 different taxa are found. Waterside herbs and aquatic herbs have been found and these taxa will later be compared to the macrofossil data. Pollen of two wetland trees have been found, Alnus and
Salix. Next to these groupings there are some taxa that are ecologically undeterminable,
since they have large ecological ranges.
On the left of the diagram the percentage of upland trees compared to upland herbs is shown over the chronology. Overall the herbs are dominating over the trees, however around 30 cm and 10 cm the trees are as present as the herbs are.
33
Figure 11. Pollen diagram (den Boef 2018).
Den Boef 2018 recognized zones in the results of the pollen analysis and a quick summary of his zonation is given below (den Boef 2018, 33-38).
Zone 1 (65-55 cm)
This zone shows a dry open grassland region dominated by Artemisia. The presence of
Plantago coronopus helps confirms this, together with Polygonum aviculare. Amaranthaceae remains are also indicators of open environments, meaning there could
not have been a great deal of tree coverage in the area; so it was not a dense forest around this time.
Boreal pollen have been found, indicating that trees were present in the area, but not abundantly so. Especially pollen of Quercus are found that indicate a more temperate climate. Oak trees need quite some light to germinate and when growing naturally in deciduous forests they cannot compete with other trees, meaning that they grow in more open areas (www.floravannederland.nl). The combination of oak trees and an open landscape is possible and can be seen in mountainous areas.
The waterside and aquatic herbs found in the pollen data indicate the presence of a waterbody in the near vicinity of the sample.
At 50 cm there was a pollen sample taken that contained very little pollen. Den Boef 2018 has interpreted this as being related to a short oxidation event. This is made visible in the diagram with a dotted line.
Zone 2 (45-20 cm)
Zone 2 starts at the contact level at 45 cm, were the skeleton of the straight tusked elephant (Palaeoxodon antiquus) was found. The composition of the taxa in zone 2 remains similar to that of zone 1, however the quantity of the remains changes significantly. There is a clear decrease in Artemisia pollen and subsequently almost all the other pollen types show an increase. There is a large increase in Poaceae pollen visible suggesting more grassland surrounding the lake. There is an increase in waterside herbs present and monolete fern spores appear in the samples. An increase in aquatic herbs is also visible.
Arboreal pollen also show an increase in this zone and especially around 30 cm there is an increase in Quercus robur type and the Quercus ilex type appears. Quercus ilex originates from southern Europe and indicates warm temperatures, nowadays it mostly grows in the
34
Mediterranean and indicates more temperate and dry conditions (Weeda et al. 1985, 11). A significant increase in Alnus can be attested, which grows near the water or on wetlands. There is seemingly an increase in temperature and availability of water, during this period the water body probably extended.
Zone 3 (15-0 cm )
The top of the section shows a similar result as zone 2, however there are some slight differences. At the level of 15 cm an oxidation event has been recognized. There is a decrease in the amount of taxa after this oxidation event and especially herbaceous taxa disappear from the record. Interesting is the increase of Plantago coronopus type in this zone, however not much is known about this taxa, except that it is seen as a weed that grows mostly in the Mediterranean. It does prefer silty and sandy soils, indicating that there was a level of salinity in the area.
4.2 Plant macrofossil data
The macrofossil data is presented in appendix 1. Because of its large size it is not included in the text, since this would disfigure the data and disrupt the text. Traditional diagrams have been used to visualise the data, as a Tilia diagram could not be constructed. General trends are explained below and ecology of relevant plants is presented. The ecology of the rest of the plants is presented in appendix 2. Some plants that have been found in the macrofossil data are overproducers and have to be analysed separately for a quantative analysis. These species are; Azolla filiculoides, Salvinia natans, Dryopteris and Characeae. The preservation of the botanical remains is very good throughout the section. The remains are preserved by waterlogging, which means that they were not in contact with oxygen and did not decay (Lowe and Walker 1997, 185).
4.2.1. Biodiversity and amount of remains
A few general things can be analysed from the data. A clear fluctuation in the biodiversity of the remains is visible in the diagram below (see figure 13). There seems to be an increase in the amount if taxa present throughout the section. However, it is not the top section that can be considered the most diverse, but the 20-30 cm section, right before the oxidation event recognised in the pollen data. After the 20-30 cm section the biodiversity decreases slightly, but remains prominent. The 30-40 cm section shows a sudden dip in the amount of taxa found, it is significantly less than in the other sections. The pollen data does not
35
attest for an oxidation event in this section nor does the macrofossil data (preservation remains the same).
Figure 12. Amount of taxa per sample.
Another general analysis from the macrofossil results, is the amount of remains per section. Overall the amount of remains seems to increase when going up in the section, however there are two clear dips. One of these dips is at the 10-20 cm section. This is where an oxidation event has been recognised in the pollen data. This could explain the slight decrease in remains, however no change in preservation was found. Considering this, it is more likely that this decrease is due to natural changes. However, there is also a dip visible at 30-40 cm, this is interesting considering there is also a dip in amount of taxa at this level. This section is in the sedimentological Unit UA3c, which overlays the archaeological remains in UA4. In the pollen data this section does not show a decrease in remains, suggesting it is a local phenomenon.
0 10 20 30 40 50 60 0-10 cm 10-20 cm 20-30 cm 30-40 cm 40-50 cm 50-60 cm Am o u n t o f t ax a
36
Figure 13. Amount of remains per sample.
4.2.2. Indicator taxa and other interesting taxa
Azolla filiculoides
Azolla filiculoides is a water fern that floats on the surface of the water. When the plant is
in direct sunlight the leaves turn a reddish brown colour, but in shaded areas the colour stays green. Its roots float in the water and are not anchored in the soil. The plant naturally occurs in subtropic warm regions of North and South America, but is reintroduced in Europe through botanical gardens (www.floravannederland.nl). This occurred in 1880 when specimens of the botanical garden in Bordeaux, France, where thrown into a local pond (Field 1999, 92).
Azolla filiculoides did occur in Europe but became extinct during the Middle Pleistocene.
When found in European deposits it acts as a stratigraphic marker, indicating that the deposits are of a Middle Pleistocene or earlier nature (Field 1999, 92).
This species is only found in areas where the winters are mild. It has been recorded in Ireland in summers, but when harsher winters were apparent the plant almost completely disappeared (Lucey 1998, 6).
Azolla filiculoides is presumably an invasive species that can take up big parts of the
water surface, making it difficult and almost impossible for submerged plants to grow. In the macrofossil remains Azolla filiculoides occurs in almost all the samples, except for one. Interestingly it increases throughout the section and at the top of the section it seems that it is established at the point of sampling. (maybe diagram here?)
0 500 1000 1500 2000 2500 0-10 cm 10-20 cm 20-30 cm 30-40 cm 40-50 cm 50-60 cm Am ou n t of re m ain s
37
Salvinia natans
Salvinia natans is similar to Azolla filiculoides and also floats on the waters surface. This
fern occurs in temperate, subtropical and tropical regions and is an indicator of warm temperatures. It can reproduce by sexual and asexual means, however in order for the plant to sexually reproduce it needs warm temperatures. It has been attested to expand during periods of climate warming (Szmeja and Gałka 2013, 114).
In the macrofossil samples Salvinia natans has been found in all samples. The
microspores are found in each sample throughout the section and show an increase from sample 5 (40-50 cm) onward with a peek of more than a thousand specimens in sample 2 (10-20 cm). The megaspore occur in all but one sample (sample 3, 30-40 cm) and show no clear linear increase.
Urtica kioviensis
Urtica kioviensis is a species that occupies river valleys and grows in damp areas with
species like Phragmites, Phalaris and Carex riparia reed swamps. It is an element of floodplain and lowland landscapes. Its modern day distribution lies mainly in Hungary and the north-western adjacent countries and further northwards to northeast Germany (Field and Lewis 2019, 6; Wolters et al. 2005, 520). Urtica kioviensis has been found in the British Pleistocene at Norfolk, England. This is the first indication of this species in the area in that time period (Field and Lewis 2019, 2).
In the Marathousa 1 assemblage a number of Urtica kioviensis seeds are found in the uppermost sample (sample 1, 0-10 cm).
4.2.3. Analysis per sample
Below every sample is analysed separately to make a detailed reconstruction of the environment and the changes in it.
Sample 6 (60-50 cm)
The 60-50 cm section contains a smaller number of remains, with 239 remains found. The amount of different taxa is also less than in most samples, with only 32 taxa present. These taxa and remains mostly exists out of Aquatic and Waterside vegetation, which is making up around 70% of the sample. The location of sampling was almost definitely the waterside of a lake, with species such as Nymphaea alba, Salvinia natans, Carex and a great number of Typha present. Nymphaea alba grows in quite deep, relatively calm, preferably still water. It is mostly found at water depths of 1 to 1,5 meters (Weeda et al. 1985, 219).
38
environments. Carex is a plant that pioneers8 in shallow waters and on the waters edge,
they can however occur in grasslands, as well as dune valleys and a few types of woodlands (Weeda et al. 1994, 279). These plants together with some other taxa suggest that the location was not an abrupt water edge, but rather a reed marsh like transition into deeper waters.
Some remains of Cenococcum geophilum have been found, indicating the successional stage of a forest, it uses trees and other plants as a host and it has a significant tolerance to water stress. Apart from these remains no indicators have been found of nearby woodlands, meaning the area was relatively open and dominated by pioneering vegetation. Apart from some Poaceae seeds there are no indicators of a grassland, the water must have stretched and the area was probably swamp like.
Sample 5 (50-40 cm)
Overall section 50-40 cm is similar to the section below, showing clear influence of water, by all the aquatic taxa. The number of taxa slightly increases in this sample and more aquatic taxa appear. Waterside taxa remain the same. Some of the aquatic taxa that occur are several species of Potamogeton. Potamogeton natans, Potamogeton crispus and
Potamogeton trichoides appear in small number in this sample. These are all plants that
grow under the water surface and have their roots in soil at the bottom of the water. Interestingly Potamogeton crispus is associated with thriving in water with fluctuation water levels. It can even survive when its habitat runs dry.
Section 50-40 cm contains one species that grows in grasslands, Ranunculus sardous. This species of buttercup grows in full sun and in relatively open vegetation. It mostly occurs in more temperate regions (Weeda et al. 1985, 243).
.
Sample 4 (40-30 cm)
This sampled section shows a significant decrease in taxa as well as remains. The section mostly exists of waterside plants that take up 53% of all the remains, whilst aquatic plants only take up 22% of the assemblage.
Two waterside taxa appear in this section that weren’t found in the section below,
Chenopodium rubrum/glaucum and Cladium mariscus. Chenopodium rubrum is a
multiform species, that is spread on the northern hemisphere, mostly in more temperate areas. It is rare in several Pleistocene sand areas in the Netherlands (Probably related to
8 Pioneer species are hardy species that are the first to colonize an ecosystem that was previously
39
ice). The plant grows on moist, nitrogen-rich, open areas. It mostly prefers clayey soils and places that are submerged in winter. It occurs naturally in river banks and along brackish sea inlets and creeks, on places where the vegetation is not closed off (Weeda et al. 1985, 158). Cladium mariscus is a cosmolite of the warm-temperate and warm regions, that reaches its northern border at the South of Scandinavia. It was not present in the Netherlands during the last Ice Age and it reappeared in the Atlanticum, a period that was slightly warmer than current temperatures. Furthermore, this species is often situated in areas where water from deeper layers springs up; the consistent temperature of this water ensures a degree of heat in the winter. These warm water temperatures are important for
Cladium mariscus to survive in winter, since it is a heat-loving plant (Weeda et al. 1994,
272-273).
Sample 3 (30-20 cm)
Section 30-20 cm shows a dramatic increase in the amount of taxa, as well as the amount of remains. This sample contains the most taxa and is the most biodiverse. However, it does not contain the most remains. The aquatic and waterside taxa still make up 71% of the assemblage and this indicates that the environment did not change significantly. The increase in organic material is likely a product of taphonomic processes.
There are quite some differences from the last section and this will be discussed per ecological group. In the aquatic taxa there is a notable increase in Zannichellia palustris and Schoenoplectus lacustris. Zannichellia palustris is a plant that grows in fresh or brackish shallow waters and on soil that contains clay or sand and has a high organic content. In fresh water it mostly appears as a pioneer, in other cases it is supressed by species like Myriophyllum spicatum (Weeda et al. 1991, 264).
Schoenoplectus lacustris is a real pioneer, it cannot survive in denser vegetation. It cannot
survive periods of drought and if its water source dries out it can only survive if the soil remains wet. If erosion finds place do to waterflow it will quickly disappear (Weeda et al. 1994, 253).
In the aquatic taxa a new species appears, Groenlandia densa. This species of plant is similar to plants from the genus Potamogeton. Groenlandia densa can only grow in carbonate-rich and thus basic water. It grows in sweet, clear, somewhat chalk- and nutrient-rich, however fosfate poor water with a fosfate poor soil, that exists of sand or riverclay. It seems to be bound to waters with a uniform temperature, so cold in the summers and unfrozen in the winters. It is a pioneer plant that prefers moving water, but can also occur in newly dug puddles (Weeda et al. 1991, 241-242).
40
Several taxa appear that fit under the grassland ecology. These are, Ranunculus sardous,
Polygonum, Euphorbia exigua and Chenopodium ficifolium. Euphorbia exigua is a field
plant that grows on chalk-rich clay, chalk and loess. She can be found on open spaces that are disturbed (Weeda et al. 1988, 17-18). Euphorbia exigua occurs in temperate regions of Eurasia. It can only be found at places were the soil is reworked and in open roadside verges, in building areas and on agricultural fields. It grows on chalk holding sands (Weeda
et al. 1985, 163). A large quantity of Poaceae become apparent in this section, further
suggesting that grasslands became more prominent.
Apart from the seemingly increasing grassland a high quantity of Cenoccum geophilum appears, indicating a larger presence of woodland in the region. Next to that the remains of Rubus fructicosus have been found, which is a species that thrives in woodlands with sandy soil. They prefer damp soils, but also occur in drier environments (Weeda et al. 1987, 66).
Sample 2 (20-10 cm)
In section 20-10 cm the amount of taxa decreases and the amount of remains decreases slightly as well. The Aquatic and Waterside taxa make up 75% of the assemblage, which means the environment stayed relatively similar to the sections before.
The taxa remain relatively similar to the section below, with some minor changes. In the aquatic taxa Potomogeton coloratus appears. Potamogeton coloratus is a plant that is bound to high corbonate content and nutrient poor water. It subsides in cool, chalck-rich groundwater and a uniform temperature is very important for this plant to survive (Weeda
et al. 1991, 246-247).
In the grassland taxa the species Linum perenne appears for the first time. This plant is a perennial herb that grows in open, well-drained areas on base-rich substrates, including lightly-grazed grassland, dry banks and roadsides (www.brc.ac.uk).
There is an increase visible in remains of most of the waterside taxa and this is especially visible in Mentha aquatica. This species is native to the temperate regions of Europe. It grows on places where the water reaches the surface all year around. It thrives in sunny to light shaded areas, with carbonate- and nutrient-rich soils (Weeda et al. 1988, 179-181).
Cenoccocum geophilum is not present in this sample, but Rubus fructicosus is. Nothing
can be said about the woodland increasing or decreasing in size.
Sample 1 (10-0 cm)
Section 10-0 cm shows a slight decrease in the amount of taxa present, whilst a significant increase is present of the amount of remains in the section. This section has the most remains preserved.
