The Western Scheldt Basin in the Mesolithic Period: SPDS and Settlement Data

POPULATION DYNAMICS IN THE WESTERN SCHELDT BASIN IN THE MESOLITHIC PERIOD: SPDS AND SETTLEMENT DATA

PREHISTORY RESEARCH UNIT, DEPARTMENT OF ARCHAEOLOGY Elliot Van Maldegem; Philippe Crombé

Acknowledgements

The research presented in this poster is part of the project “In search of the pristine environments in the lowland river system of Flanders for assessing the impact of past human occupation on current phosphate concentrations”, financed by the Flemish Fund for Scientific Research (FWO project G089319N).

Contact

Corresponding author: Elliot.VanMaldegem@ugent.be https://research.flw.ugent.be/nl/elliot.vanmaldegem

Methodology

For the Mesolithic Period a multiproxy approach was utilized consisting of:

• SPD only using critically selected dates from secure contexts on materials with no reservoir effect or possible contamination. A logistic null model was fitted to account for problems with sampling error and calibration artifacts, and aid with the separation of signal and noise (Fig. 2).

• Site counts corrected for the durations of the sub-phases of the Mesolithic Period. This results in equal weight being given to each sub-phase regardless of the duration. For the Early Mesolithic Period both an early start of the phase (ca.

11.350 cal BP) and a late start (ca. 10.650 cal BP) were considered. The upper and lower Scheldt regions were also looked at seperatly to see if there were intra-regional differences in population dynamics (Fig. 4).

• Loci data was compiled from selected sites on amount of loci per sub-phase, size, artifact density and tool frequency (Fig.

3). This enabled comparisons of intersite differences in occupation duration and re-use. Loci were taken as representing individual dwelling and/or activity areas within a settlement.

Introduction

The western half of the Scheldt drainage basin (Fig. 1) in western Belgium and northern France has been a dynamic region since at least the Mesolithic period. The area has been extensively studied during the preceding decades leading to the discovery of increasing amounts of sites through developer- led archaeology and field-walking campaigns which has resulted in large radiocarbon and site databases. This provides a solid basis for paleo demographical studies. The project aims to reconstruct the paleodemographical evolutions between the Early Mesolithic and Late Medieval periods, starting here with the Mesolithic Period

Background

In recent years, several studies have dealt with the possible impact of two important, short but abrupt cooling events, known as the 9.3 ka and 8.2 ka cal BP events on Mesolithic hunter-gatherer populations. These often utilize Summed Probability Distributions of Radiocarbon Dates (SPD’s). However it is highly questionable whether changes in population can be traced solely on the basis of the radiocarbon record, because this record is potentially biased by several factors such as interregional and intersite differences in site- taphonomy, research foci, sample selection and excavation methods. A multiproxy approach is necessary.

Results

During the Early Mesolithic period all proxies point to a population increase with high residential mobility.

This is followed by a decrease both in radiocarbon density and amount of sites during the Middle Mesolithic Period. However when the lower and upper Scheldt region are considered seperately there is a striking difference between the lower Scheldt (decline) and upper Scheldt (substantial increase). This could point to intraregional population shifts, from the lowland to the upland.

The Late Mesolithic Period sees a further decrease in all proxies, which could to reduced mobility, although this is difficult to assess due to incomplete records.

Fig. 1: Study area (red) and spatial distribution of Mesolithic Sites (squares) and Mesolithic sites with radiocarbon dates (circles) (DEM © European Union, Copernicus Land Monitoring Service 2021, European Environment Agency (EEA))

Future

• Expanding the database of dates and sites up to the end of the Medieval Period

• Including further proxies such as human impact scores derived from pollen records

• Refining the site count data

Fig. 4: (A, C-D) Chronological distribution of site counts per (sub-)region (corrected for duration of sub-phase) and (B) Scheldt Drainage Basin SPD (divided into the 3 sub-phases for easier comparison).

Fig. 2: a) Summed Probability Distribution (SPD), b) not normalized SDP (black line) with fitted logistic null model (95%

confidence interval). Red blocks denote positive deviations from the logistic null model (larger growth than expected) and blue block denote negative deviations (smaller growth than expected), c) NGRIP climate curve with dark blue blocks indicating climate events defined by INTIMATE and light blue blocks indicating climate events (IRD) defined by Bond et al. (1997)

Fig. 3: a) Loci counts per sub-phase, b) spread of loci sizes per sub-phase, c) spread of artifact density (/m²) per sub-phase, d) spread of tool frequency per sub-phase

References

Bond, G., et al. (1997). "A pervasive millennial-scale cycle in North Atlantic Holocene and glacial climates." Science 278(5341): 1257- 1266.