Ulrich Schmölcke

Mitochondrial DNA analysis shows a Near Eastern Neolithic origin for domestic cattle and no indication of domestication of European aurochs.

The extinct aurochs (Bos primigenius primigenius) was a large type of cattle that ranged over almost the whole Eurasian continent. The aurochs is the wild progenitor of modern cattle, but it is unclear whether European aurochs contributed to this process. To provide new insights into the demographic history of aurochs and domestic cattle, we have generated high-confidence mitochondrial DNA sequences from 59 archaeological skeletal finds, which were attributed to wild European cattle populations based on their chronological date and/or morphology. All pre-Neolithic aurochs belonged to the previously designated P haplogroup, indicating that this represents the Late Glacial Central European signature. We also report one new and highly divergent haplotype in a Neolithic aurochs sample from Germany, which points to greater variability during the Pleistocene. Furthermore, the Neolithic and Bronze Age samples that were classified with confidence as European aurochs using morphological criteria all carry P haplotype mitochondrial DNA, suggesting continuity of Late Glacial and Early Holocene aurochs populations in Europe. Bayesian analysis indicates that recent population growth gives a significantly better fit to our data than a constant-sized population, an observation consistent with a postglacial expansion scenario, possibly from a single European refugial population. Previous work has shown that most ancient and modern European domestic cattle carry haplotypes previously designated T. This, in combination with our new finding of a T haplotype in a very Early Neolithic site in Syria, lends persuasive support to a scenario whereby gracile Near Eastern domestic populations, carrying predominantly T haplotypes, replaced P haplotype-carrying robust autochthonous aurochs populations in Europe, from the Early Neolithic onward. During the period of coexistence, it appears that domestic cattle were kept separate from wild aurochs and introgression was extremely rare.

New research at Riņņukalns, a Neolithic freshwater shell midden in northern Latvia

The prehistoric shell middens of Atlantic Europe consist of marine molluscs, but the eastern Baltic did not have exploitable marine species. Here the sole recorded shell midden, at Riņņukalns in Latvia, is on an inland lake and is formed of massive dumps of freshwater shells. Recent excavations indicate that they are the product of a small number of seasonal events during the later fourth millennium BC. The thickness of the shell deposits suggests that this was a special multi-purpose residential site visited for seasonal aggregations by pottery-using hunter-gatherer communities on the northern margin of Neolithic Europe.

Neolithic fish remains from the freshwater shell midden Riņņukalns in northern Latvia

The Neolithic site Riņņukalns in the Lake Burtnieks/River Salaca area in northern Latvia is the only freshwater shell midden in the eastern Baltic Sea area. An excavation carried out in 2011 revealed an intact stratigraphy with alternating layers of unburnt and burnt mussel shells and yielded various kinds of archaeological finds, among them several thousand fish remains. To gain an understanding of the fish species and specimens caught by the Neolithic settlers, and to discern any temporal development in the fish species composition, we analysed fish remains from different sections and layers. Results from both the archaeozoological and stable isotope data, give evidence for a change in the relevance of fish species during the period of use, and they also provide information for reconstructing the former river and lake hydrology in the vicinity of the midden. The Stone Age landscape seems to have been very similar to the present situation, so that the study area has been an extraordinarily stable ecosystem for more than 5000 years. Comparisons with the results of recent monitoring programmes, long-term changes since Medieval times, and written sources from the 18th century, show that the fish species community is almost unchanged since the Stone Age. This underlines the importance of the region in nature conservation.

Archaeogenetic evidence for medieval occurrence of Atlantic sturgeon Acipenser oxyrinchus in the North Sea

The sturgeon was an important dietary resource for people living in the eastern North Sea coastal area, especially in the 19th century but also in previous millennia. However, since the discovery in 2002 that not only the European sturgeon (A. sturio) but also the Atlantic sturgeon (A. oxyrinchus) occurred in northern European waters, we still do not know which of these species was dominant in the North Sea and hence of primary economic importance. The 800-year-old, well-preserved sturgeon remains presented in this paper, from the ringfort Itzehoe by the Stör River (which is a tributary of the Elbe in northern Germany), provides an opportunity to answer the question for the first time. The aDNA amplified and sequenced from seven bones of at least five different individuals derives from A. oxyrinchus exclusively. Moreover, morphological analyses of the whole assemblage of 15 bones provided no evidence for the presence of A. sturio. Even though the dataset is still too small for general reconstructions, this study demonstrates the occurrence and possibly the dominance of the Atlantic sturgeon in at least parts of the North Sea region 800 years ago. However, further research is necessary to prove if A. oxyrinchus was the only sturgeon species in the North Sea then.

Lack of support for adaptation of post-glacial horses to woodlands

To the Editor — Recently, SandovalCastellanos et al.1 published an inspiring study that supports an adaptation to forests in wild horses (Equus ferus) during the mid-Holocene epoch (9,000–5,000 years ago (9–5 ka)) in Europe1. Sandoval-Castellanos et al.1 used a modelling approach for reconstructing the relative landscape openness in Europe on the basis of the percentage of arboreal plant functional types (European Pollen Database), as well as ecological niche modelling to infer co-occurrence patterns of wild horses and plants using faunal records of wild horses and pollen records of plant taxa. The results show an increasing presence of horses in post-glacial forests but no significant horse–plant co-occurrences. To evaluate the temporal change of coat colour genes in horses they analysed the agouti gene (ASIP) and the results show a selection of black phenotypes during the Holocene. Combined results were taken as evidence by the authors to support the hypothesis that horses adapted to woodlands. However, our detailed analysis of faunal data from archaeological excavations (project ‘Climate–Landscape–Fauna–Man’) revealed that adaptation to primeval woodlands in the Central European Lowlands (CEL) is not supported by the subfossil vertebrate record. Among other data, we collected detailed information on more than 500 Mesolithic and Neolithic archaeological sites containing faunal remains (Fig. 1) to investigate interactions of landscape and dynamics of the megafauna as well as human influence. The early Holocene (11.7–9 ka) has the highest percentage of sites where horse is present (Fig. 1a), which is in line with recent research suggesting that open grasslands existed in the first two millennia of the Holocene2 on fine-grained soils, where seasonal dryness prevented the establishment of deciduous trees. We interpret the ‘near-absence’ of horse records in the early Atlantic (Fig. 1b) as large-scale regional habitat loss due to the development of the mid-Holocene primeval woodlands, dominated by oak and lime3. Whereas in our previous study4 we found an increased presence of horse in the late Atlantic, mainly in southeastern and southern Central Europe, it appears very rarely in the CEL. In 39 ‘representative archaeological sites’ with 500 to more than 10,000 faunal bone remains, wild horse (E. ferus) occurs in only five sites; a frequency of less than 1% (Fig. 1c). During the period 5.5–4 ka, when the availability of open landscape was significantly greater due to increasing human activity5, wild horse is recorded from 13 of 56 ‘representative sites’; a frequency of less than 1% (Fig. 1d), but also for the first time present in a frequency of more than 10% in the faunal material (Fig. 1d). In our opinion this indicates the beginning re-colonization of CEL resulting from the opening of the landscape by Neolithic farmers, as suggested in an earlier study4. In contrast to forest-adapted species such as red deer (Cervus elaphus), roe deer (Capreolus capreolus)6,7, wild boar (Sus scrofa) and European moose (Alces alces)8, which occur in the faunal material of midHolocene archaeological sites in very high frequencies, the large-scale absence and extremely restricted appearance of wild horse (Fig. 1) do not support an adaptation to Holocene primeval forests in the CEL. Alternatively, we think that the selection of black phenotypes may have taken place as part of the adaptation of horses to open forests in southern Central Europe and peripheral regions, where the primeval vegetation during the midHolocene was largely open forest steppe9. The records of black phenotypes are exclusively from those areas1 and support our alternative interpretation of the results of Sandoval-Castellanos et al., who suggest that the Holocene survival of the wild horse in Europe is tied to either their occurrence in cryptic open habitats or their adaptation to forests. The evidence for an uninterrupted local occurrence of steppe grasslands (termed cryptic open habitats by Sandoval-Castellanos et al.) and light-demanding plant species in southern Central Europe during the 11.7–9.1 ka 9.1–7.5 ka