Early to mid-Holocene human-river interactions in the Lower Danube Valley: new research at Poiana (Teleorman County) (original) (raw)

River Dynamics and the Neolithic of the Lower Danube Catchment

The Lower Danube in Prehistory: Landscape Changes and Human Environment Interactions - Proceedings of the International Conference, Alexandria 3-5 November 2010 (edited by Steve MILLS and Pavel MIREA), 2011

In the last decade new archaeological and geomorphological research in the lower Danube catchment (LDC) has transformed our understanding of prehistoric river-society interactions, particularly with respect to the environmental context in which farming first developed in Southeast Europe at around 6100 cal. BC. This paper critically reviews these recent developments and using a new Late Pleistocene and Holocene fluvial chronology from the Teleorman Valley (TV), southern Romania, examines the interplay between river dynamics and the Neolithic archaeological record from two perspectives. First, considering the likely impacts of rapid, climate-related changes of floodplain environments on communities. And secondly, the effects of river erosion and sedimentation on the preservation of archaeological sites within river valleys. Although chronologically the Neolithic period in the LDC falls between the so called 6600-6200 and 4000-3200 cal. BC rapid climate change events, the well dated TV fluvial record allows more precise relationships to be established between changes in river dynamics and prehistoric settlement patterns. Early Neolithic Starčevo-Criş sites (c. 6100/6000 cal. BC) are located on a Late Pleistocene river terrace (36.8 ka), 10 m above modern river level (AMRL) and were unaffected by Holocene fluvial erosion and sedimentation. Later Boian sites (4810 - 4680 cal. BC) are preserved on the surface of Last Glacial Maximum (21.6 ka, 8 m AMRL) and Late Glacial (15.8 - 12.8 ka, 7 m AMRL) river terraces, the lower parts of which are covered by a thin (< 0.5 m) veneer of Holocene alluvium. No Neolithic sites have been found either on the surface or within Holocene fluvial deposits in the study reach. The development of monumental tells in the Gumelniţa period from c. 4500 cal. BC does not coincide with a change in river dynamics and suggests that this new settlement style was not enforced by an alteration in river behaviour or flooding regime. The abandonment of tells in the TV from c. 3800 cal. BC does coincide with a marked increased in river erosion and sedimentation at c. 3900 cal. BC that continued until c. 2000 cal. BC. What is perhaps most striking from our investigations in the TV is the continuity of Neolithic sites in particular riparian locations, which seem to have been facilitated by more than 2000 years of relative river quiescence, certainly when compared to the period after c. 3900 cal. BC. While this new model of Holocene river valley development needs to be tested and further refined elsewhere in the region, it does suggest a radical re-thinking of the Neolithic record in the LDC may be required.