Neolithic population crash in Germany?

Journal of Archaeological Science
Volume 34, Issue 8, August 2007, Pages 1339-1345

Prehistoric population history: from the Late Glacial to the Late Neolithic in Central and Northern Europe

Stephen Shennan and Kevan Edinborough

Abstract

Summed probability distributions of radiocarbon dates are used to make inferences about the history of population fluctuations from the Mesolithic to the late Neolithic for three countries in central and northern Europe: Germany, Poland and Denmark. Two different methods of summing the dates produce very similar overall patterns. The validity of the aggregate patterns is supported by a number of regional studies based on other lines of evidence. The dramatic rise in population associated with the arrival of farming in these areas that is visible in the date distributions is not surprising. Much more unexpected are the fluctuations during the course of the Neolithic, and especially the indications of a drop in population at the end of the LBK early Neolithic that lasted for nearly a millennium. Possible reasons for the pattern are discussed.

Keywords: Mesolithic; Neolithic; Radiocarbon dates; Population history

[. . .]

In a recent paper Gamble et al. (2005) used the S2AGES database of radiocarbon dates for the period from c. 25–8 ka that they had compiled for western and northern Europe to propose an outline of the population history of the region during the Late Glacial period. The object of this paper is to follow up that study, albeit it on a more limited geographical scale, by adopting essentially the same approach to trace regional population histories in three areas of Central and Northern Europe up into the Neolithic, and in particular beyond the Neolithic transition on which earlier radiocarbon work by one of us was focussed (Gkiasta et al., 2003). In our view the results reveal some striking patterns which have significant implications for our understanding not just of the beginning of the Neolithic but more importantly what happened after it.

[. . .]

Starting with the earliest periods and working to the right, a number of features may be observed. In all cases the Mesolithic population shows fluctuations, but immediately before the beginning of the Neolithic it is actually lower than it was in some earlier phases. If one compares the maximum Mesolithic peak with the first Neolithic peak in each region, the Mesolithic peak in Denmark is proportionally the highest, which is likely to be a reflection of the significance of aquatic resources in Denmark and the high populations they are capable of supporting (cf. Schmölke, 2005). As noted above, while there may be some doubt about the comparability of the Mesolithic and Neolithic proxy population patterns, the bias against the former in favour of dates for the latter would have to be massive to alter the obvious inference to be made from the figure. The Danish pattern is basically the same as that produced in a similar radiocarbon exercise for Denmark and Sweden by Persson (1998, reproduced in Price, 2003). The beginning of the Neolithic is strikingly apparent in all three areas: the start of the LBK in Germany at c. 5500 cal BC, slightly later in Poland; and the beginning of the TRB Neolithic in Denmark at just after 4000 cal BC. In all cases there is a rapid rise in population to a ceiling; in Germany and Denmark this is basically maintained for some time; the marked dip in the Polish R_Combine data may or may not be a sampling artefact.

After 5000 cal BC the German data suggest a remarkable decline in population, to a fraction of its maximum LBK levels, lasting, with one or two fluctuations, until after 3500 cal BC. Poland shows a very similar picture although the decline is not as striking. In Denmark there is no such marked crash although there is a decline to just over half the maximum 3500 cal BC value at c. 3000 BC, roughly at the transition between the Middle Neolithic TRB and the Single Grave Culture. A slight upturn follows, with a more marked decline after 2500 BC. In Poland a sudden rise to a peak at 3500 BC is followed by a decline to a much lower level in the centuries after 3000 BC, corresponding to the various local Polish versions of the Corded Ware. Germany by contrast shows a rapid rise to a new population plateau at c. 3400 BC, maintained until 2500 BC, followed by a marked dip and then a rapid rise again at a time corresponding to the Bell Beaker culture and the beginning of the early Bronze Age. The pattern in the final centuries of the third millennium BC should be treated with some caution, since in southern Germany and Poland at least this is already the beginning of the Early Bronze Age, so it is possible that not all available dates have been included.

[. . .]

That the appearance of the LBK marked a major population increase in the areas where it is found is well established. What the data make clear is the extremely low levels of Mesolithic population prior to this arrival; the implication being that existing hunter-gatherer populations only made a significant contribution demographically, genetically and culturally to the extent that they were incorporated into the advancing LBK demographic wave.

However, the most significant result, we would argue, is the demonstration of the drastic demographic decline at the end of the LBK and the long subsequent period of relatively low population levels. Explaining the reasons for this now becomes a major issue. The decline suggested here on the basis of the radiocarbon evidence also fits in with an increasing number of indications from other sources that far from being the foundation of the subsequent Neolithic across large parts of central, northern and northwestern Europe, in some respects at least it actually left little trace. Thus, the recent ancient DNA study of LBK samples (Haak et al., 2005 W. Haak, P. Forster, B. Bramanti, S. Matsumura, G. Brandt, M. Tänzer, R. Villems, C. Renfrew, D. Gronenborn, K.W. Alt and J. Burger, Ancient DNA from the first European farmers in 7500-year-old neolithic sites, Science 310 (2005), pp. 1016–1018. View Record in Scopus | Cited By in Scopus (40)Haak et al., 2005) suggested that the most frequent mtDNA variant was one which is extremely rare in the region in modern populations. Archaeobotanical studies are also making it increasingly apparent that the LBK crop exploitation system was an unusual one which did not have any descendants (Coward et al., unpublished paper and Bakels, in press).

doi:10.1016/j.jas.2006.10.031