Investigating Neolithic land use in Swifterbant (NL) using micromorphological techniques

In the Swifterbant area in The Netherlands, a complete Neolithic landscape is preserved, buried in a wetland environment. A dozen sites (dating from ca. 4300-4000 cal. BC) on levees of a former creek system are characterized by a black layer containing large amounts of carbonized plant remains, burnt bone, flint and pottery. These sites are usually interpreted as occupation sites with accumulated refuse of a society in transition from a Mesolithic to a Neolithic lifestyle (hunting and herding), in an area that was too wet for crops. In the context of a new research campaign in the area, we investigated the site- and land-use on two locations (S2 and S4) using micromorphological techniques. On S2, the soil matrix in the archaeological horizon has a heterogeneous, non-sedimentary appearance due to the ubiquitous presence of rounded and subrounded aggregates. These aggregates indicate that the sediment was disturbed after deposition, but it is unclear by what human activity. This disturbance is not restricted to the archaeological site horizon, but extends also into the under- and overlying layers. On S4, three layers can be distinguished. Only the lower and central parts are usually regarded as archaeological layers. The lower layer shows thorough mixing of the (partly decalcified) groundmass and the incorporation of anthropogenic materials (carbonized plant remains, burnt bone, etc.). Most probably, the thorough mixing is a result of tillage. The central layer consists of thinly laminated phytoliths and carbonized plant remains with ample fragments of (burnt) bone, angular sediment clods and some dog coprolites. Mineral sediments are absent. Apparently, the layer was formed by human activities that resulted in a massive accumulation of burnt organic materials and anthropogenic debris. The preservation of the microlayering formed by phytoliths and carbonized is remarkable, since trampling would have deformed or disturbed the layering. The surprising conclusion therefore has to be that this location was not the main settlement area-as was implicitly thought until now. Rather, the area should be interpreted as a location where specific activities resulted in the accumulation of burnt plant material, e.g. a waste dump. The upper layer again has indications of anthropogenic soil disturbance. Samples from the basin adjacent to the levee site show sediments that are deformed by trampling. The transition of the site to the channel shows no signs of trampling, but rather interfingering of site layers with channel sediments. Lack of trampling in the central zone of S4 and the adjacent channel shore is at odds with the classical interpretation that the site represents a settlement area. The black layers that up till now were thought to encompass a complete settlement, could very well be a very prominent part of larger sites with remains of human activities during the Neolithic. (C) 2009 Elsevier B.V. All rights reserved