Archaeology in the Severn Estuary 11

Estuarine development has been previously modelled in relation to the rate of relative sea-level rise, suggesting intertidal surfaces attain different vertical positions within the tidal frame according the rate of sea-level rise. These model results are significant in the context of Holocene sea-level rise, which experienced a continuum of very rapid to low sea-level rise rates, in that the estuarine palaeoenvironment and its resource potential for prehistoric populations is suggested to have changed through time. The present study empirically investigates the palaeoenvironmental development of the Somerset Levels (UK), an extensive coastal wetland (palaeo-estuarine system) that is of international archaeological importance. A composite sequence of Holocene sediments is analysed using foraminiferal and molluscan-based palaeoenvironmental techniques, and dated through radiocarbon and chemostratigraphic methods. The results indicate that during the Early Holocene (Mesolithic) the estuary was dominated by low saltmarsh and mudflat palaeoenvironments dictated by the rapid rate of post-glacial sea-level rise. This intertidal surface, low within the contempory tidal frame, would have been characterised by soft mud and sparse vegetation due to frequent tidal indundation, and would have possessed a high hydraulic duty (cf the amount of tidal water accommodated by the surface) with associated large tidal creeks/channels. Towards the Middle Holocene (Neolithic) a transition to palaeoenvironments higher within the tidal frame is evident, so that mid and then high saltmarsh dominate the estuary. This transition was in response to a progressively decreasing sea-level rise rate, allowing the intertidal surface to elevate through the tidal frame. These intertidal surfaces are likely to have become progressively drier and more richly vegetated due to increasingly infrequent tidal inundation, with a corresponding reduction in hydraulic duty and creek/channel size, culminating in emergence and supratidal peat growth. The results also suggest that this sequence is reversed through the Late Holocene (late Neolithic, Bronze Age, Iron Age) so that Roman reclamation of the Somerset Levels took place against a rising sea-level. These data support previous theoretical models, and have significant implications for the resource potential of these changing estuarine palaeoenvironments for prehistoric populations.

Past environmental work on the Severn Estuary Levels, much of it in the last decade, is reviewed and areas of particular future potential are identified. The main topics in prehisto,y include sedimentary context, vegetation history (table of pollen site~), faunal evidence (table of faunal evidence), animal husbandry, seasonality, fishing practise, and invertebrates. There is evidence for seasonal (winter) activity in the Mesolithic. In the Neolithic and earlier Bronze Age activity on the wetlands is limited. Clearance associated with pastoralism and some arable is attested on adjacent dryland. From the middle Bronze Age to the Iron Age there was extensive wetland activity associated particularly with seasonal (mainly spring and summer) cattle husbandry. In this landscape bones may be compared with animal footprint tracks and a range of other sources of environmental evidence. Wetland, wetland-edge and dryland bone assemblages are compared, suggesting there are more ovicaprids in the wetland edge and on dry land. The Romano-British landscape was extensively drained: in some areas marine influence was excluded facilitating year-round activity, while elsewhere ditched landscapes seem to have been subject to seasonal marine inundation. Evidence for continuity and change in the Romano-British and later landscape is noted. It is argued that future research should give particular emphasis to the analysis of biota from settlement sites and to the study of wetland/dryland interactions. The value of comparative studies using multiple sources of environmental evidence is clear but in the future resources will need to be tmgeted on particular research questions and priorities. An attempt is made to identify these and the sources of evidence most likely to contribute to their solution.

The deep alluvial sequences that make up the Severn Estuary Levels comprise a series of stratified landscapes dating from early prehistory through to the present day. Most of these landscapes are deeply buried, and, whilst exceptionally well-preserved, are largely inaccessible and so ill-understood. It is only with the 'historic landscape', that lies on the surface of the Levels, that we can really start to reconstruct and analyse what these past landscapes were like. However; although the enormously diverse historic landscape is itself an important source of information, its full potential is only achieved through its integration with associated archaeological and documentary evidence. This presents many challenges and whilst much has been achieved in the last ten years, there is a long way to go before we can write a comprehensive history of the Severn Levels. Two techniques are vital. Historic landscape characterisation focuses on the key character defining features of different landscapes and can suggest the processes that may have led to their creation. This is almost invariably a complex story as most landscapes combine features from a number of different periods, though retrogressive analysis can disentangle this palimpsest. In the past ten years, both techniques have been successfully applied to various of the Severn wetlands, although attention has focused upon the wholly cultural process of reclamation: the potentially significant role played by certain features of the natural environment, including the belt of sand dunes that fringe much of the Somerset coast, has been neglected. Elaborate models have also been constructed, based largely on landscape morphology, and there is a desperate need to test these hypotheses through fieldwork. Above all there is a need to move beyond simple landscape charcterisation towards more detailed palaeogeographies.

The recognition of prehistoric and later archaeological features exposed by erosion on the Severn foreshore in the 1980s prompted an increased interest in the potential of the equivalent deposits behind the sea wall. The main practical problems were how to locate sites masked by subsequent sediment deposition and how to define areas of high and low potential in the absence of models of site distribution and palaeotopography. In this paper the strategies adopted for a series of major construction projects on the Levels are discussed, demonstrating the shift in focus away from the long-term natural sequence towards specific horizons considered likely to relate human activity, alongside an increased confidence in our ability to define areas and periods of high potential. The main projects discussed are the Rumney Alternative Feeding Grounds (J 989), Second Severn Crossing (English Approaches) evaluation (1991), Barland's Farm/Europark (1993), Goldclif.f Wetlands Reserve (1996), Cabot Park (1994-1999) and Europark Wilkinsons (1999). The types of feature typical o.f theforeshore archaeology first recognised have proved to be unrepresentative of the full range of the buried resource, which is in many ways comparable to 'dry land' archaeology. There has been considerable experimentation with methods of site prospection, with limited results. Deep cores from boreholes and augers have proved only partially effective at locating horizons of interest, and trenching remains the main method. Aerial photography and archaeomagnetism have limited application, but geophysical techniques are beginning to have some success, particularly caesium vapour magnetometry. Despite the undoubted progress that the decade has seen, the approach to the evaluation of sites on the Levels remains at the forefront of debates about how far it is reasonable to require developers to undertake extensive archaeological work to address a possible but unproven 'interest '. Even when evaluation is carried out, planning policies which permit development on the levels effectively consign parts of a significant archaeological resource to unseen destruction.

Wood is one of the most important and overlooked of archaeological materials. The Severn estuary region is fortunate in having a larger quantity of archaeological wood- both excavated and still in situ - than any other rural area of the UK. The evidence it contains can be analysed in many ways and provides a view of past societies which is more intense and more meaning than that derived from 'normal' dry archaeology. The challenge which faces present day archaeologists is to develop strategies to overcome the forces which threaten the survival of this important resource.

This paper is a reflection on work that has been done in the past decade in the Severn Estuary and problems that face archaeologists working within it

Vol 11 2000 Introduction

Contrary to the situation in many countries, maritime archaeology in Denmark has always enjoyed the support of terrestrial archaeologists and other academic disciplines. From its beginnings in the late I950s maritime archaeology in Denmark has proven to be a valued sub-discipline of archaeology and natural science. This paper will discuss the evolution of maritime archaeology within Denmark from the Skuldelev excavations in the late I 9 50s and the subsequent development of the Viking Ship Museum, the Institute for Maritime Archaeology and culminating in the establishment of the National Museum of Denmark's Centre for Maritime Archaeology in 1993. The focus of the paper will be on how the success of maritime archaeology in Denmark is based upon the multi-faceted nature of the research which has been, and is being, conducted. The areas of research being studied reflect the progressive development of maritime archaeology in Denmark, which includes not only ships and boats, but wider aspects of the maritime cultural landscape. The close relationships between the various institutions at Roskilde, the Danish National Museum and other archaeological and academic institutions nationally and internationally will be illustrated with examples drawn from the thematic group 'Techniques and Auxiliary Sciences in Maritime Archaeology', one of the four thematic areas of research ongoing at the Centre.

This paper takes a personal view of the highlights and successes of the last ten years archaeological work in the Severn Estuary Levels. It then attempts to identify what might happen in the next decade, and how the vast body of information that has been accumulated by a succession of multi-disciplinary projects might be co-ordinated in order to provide a much more complete and evolving picture of this remarkable region.