Dungeness and Romney Marsh: Barrier Dynamics and Marshland Evolution

By Antony Long, Martyn Waller and Andrew J. Plater

The Romney Marsh / Dungeness Foreland depositional complex comprises an extensive tract of marshland and associated sand and gravel barrier deposits, located in the eastern English Channel. This monograph presents the results of a programme of palaeoenvironmental investigation aimed at improving our understanding of this internationally-significant coastal landform. The focus is on the evidence for landscape change during the late Holocene, from c. 3000 BC onwards, and on identifying the local, regional and global driving mechanisms responsible for the changes observed. The research details the results from two related projects, each funded as part of English Heritage's Aggregate Levy Sustainability Fund scheme. The first project concerns the late Holocene evolution of the port of Rye, located in the southeast part of the complex, and the second the depositional history of the gravel foreland. Topics explored include the vegetation and land-use history of the study area, methodological issues relating to the collection and interpretation of radiocarbon dates from coastal lowlands, the role of compaction in influencing landscape and sea-level change, and the effects of medieval storms on coastal flooding and landscape change. This monograph is intended for students and researchers interested in Holocene coastal evolution and sea-level change, coastal vegetation history and land-use history, and the development of new techniques for reconstructing past environmental change in coastal lowlands.

• Language: English
• Publisher: Oxbow Books
• Release date: Sep 20, 2007
• ISBN: 9781782974871

Book preview

1

Introduction

Antony J. Long, Martyn P. Waller and Andrew J. Plater

1.1 Introduction

At the height of the last glacial maximum, about 24000–18000 years ago, global sea-level fell up to 120 m below its present level as water from the oceans became trapped on land as ice. This lowering of sea level exposed vast areas of continental shelf around the world and, in northwest Europe, the North Sea shrank such that Britain became part of the European land mass. The subsequent melting of the ice sheets caused the re-flooding of these continental shelves, a time transgressive process that meant that the shallow Strait of Dover only became flooded relatively recently, after c. 6000 calibrated years (cal. yrs) BC, and Britain regained its island status. Most of the great ice sheets had melted by c. 4000 cal. yrs BC or were close to their present size. From this time onwards, changes in the height of the sea with respect to the land (known as relative sea-level (RSL)) was determined by the complex interplay of subtle on-going changes in ocean volume due to minor climate changes, as well as other processes such as land uplift or subsidence. In southern England, beyond the former margins of the British and Fennoscandanavian ice sheets, isostatic sinking of the land meant that RSL continued to rise by about another 5 m or so from 4000 cal. yrs BC to present.

The rise in global sea-level since the end of the last glacial maximum resulted in the reworking of extensive volumes of sediment that had accumulated on the floor of the continental shelves during the sea-level lowstand. The exposed floors of the North Sea and the English Channel were mantled with a variety of sediments; the former included extensive glacial deposits whilst the latter, lying almost entirely beyond the former ice limits, was covered in terrestrially-derived fluvial sediments deposited by the great rivers that once meandered across its exposed floor, including the former Thames, Rhine, Meuse and Somme. Some of these sediments infilled former river valleys incised into the floor of the Channel, others accumulated as spreads of sediment in terraces and sheets, several tens of metres thick in places, that were reworked and transported landwards as the Channel floor was flooded.

Computer modelling suggests that during the initial stages of flooding, the eastern English Channel was a shallow, relatively low energy embayment, with a low tidal range (Austin 1991). However, as water depths increased and as the Strait of Dover became flooded, so the tidal range increased, and the area became transformed into a much higher energy environment. Strong nearshore currents, moving in an easterly direction along the south coast of England from the Isle of Wight and through the Strait of Dover, initiated and sustained the nearshore drift of sediment from the west to the east. These sediments began to infill a series of small estuaries that discharge from the Hampshire, Sussex and Kent coasts into the English Channel, with small sand and gravel barriers developing locally at estuary mouths (Jennings & Smyth 1991).

Most of the sand and gravel barriers that developed around the British Isles during the current interglacial (or Holocene, the time period since c. 11430 ± 140 ¹⁴C years BP, or 9560 and 9300 cal. yrs BC) are relatively modest, ephemeral features. These barriers have often passed through several phases of growth, stability and destruction in response to variations in the rate of RSL, changes in storm magnitude and frequency, as well as variations in the supply of sediment (Orford et al. 1991). Despite their variability, these barriers have played a critical role in shaping the patterns of coastal and estuary evolution around many parts of the British Isles by moderating the wave and tidal energy that enters an estuary and distorting long-shore and cross-shore sediment transport pathways. Insights into the early configuration of these barriers can be seen from the preserved remnants of these early structures, especially in northern Britain where land uplift has resulted in a fall in RSL since the mid Holocene (e.g. Dawson et al. 1999). However, the upwards trend in Holocene RSL in southern England has meant that most of the early barriers here are now drowned and either lie buried or have been entirely reworked by wave and tidal processes. Most, that is, except for one exceptional stretch of coastline located in the eastern English Channel. Follow the coast from the Isle of Wight eastwards and one traces a shoreline crenulated by shallow embayments and projecting headlands. Many of these headlines are made of bedrock, most famously the 160 m-high Chalk cliffs known as Beachy Head (Figure 1.1). However, to the east of Beachy Head is an equally impressive headland, this time comprising the largest expanse of exposed and buried sand and gravel barrier in the British Isles, a vast expanse of low-lying sediment, never exceeding +7 m Ordnance Datum (OD, present sea-level), that projects some 20 km into the waters of the eastern English Channel. Known as the Romney Marsh/Dungeness Foreland depositional complex, this unique landform owes its existence to a particular interplay of RSL and sediment supply that has created one of the largest and best-preserved coastal barriers and associated hinterland marshes in the British Isles and, arguably, in northwest Europe.

The first attraction of the Romney Marsh/Dungeness Foreland depositional complex to any coastal geomorphologist is the corrugated gravel beach ridge plain that comprises Dungeness Foreland (Figures 1.2a, b). The oldest of the exposed beaches are located in the central part of the study area, where they form low, subdued landforms, partly submerged by marshland sediments that abut against and in places overlap them. These beaches have a typical surface elevation of about +1 m OD, some 6 m or so below their contemporary equivalents at the present foreland ness. The beaches rise in elevation as they become younger, tracing the rise in RSL that has occurred during the mid and late Holocene and which partly controlled their height of deposition.

Compared with the bold geomorphology of Dungeness Foreland, the lower relief of Romney and Walland Marshes (although Romney Marsh is used as a generic name, it strictly applies just to the area north of the Rhee Wall (Figure 1.1)) which lie in its protective lee may, at first glance, appear of secondary interest. But this is not so. The subtle surface relief of the marshland tells a rich story of changes in tidal flooding, land claim, as well as differential land movement due to variable compaction of the sediments that underlie the marshland surface. These marshland sediments are thick, locally reaching 30 m or more, and contain within them a rich archive of coastal and sea-level change that is intimately linked to changes in the dynamics of the barrier itself, as well as variations in processes within the Wealden catchments of the rivers (Rother, Tillingham, Brede and Pannel) that discharge into the study area from the west.

What is it, therefore, that makes this particular depositional complex so interesting to study? Firstly, the area contains an exceptionally well-preserved suite of sediments and landforms that tell a fascinating story of the interplay between physical and human processes during the last 6000 years or so. This resource contrasts with the often incomplete record preserved in smaller coastal landforms that have experienced several phases of barrier development and breakdown during the mid and late Holocene. Secondly, the interplay between barrier and marshland landforms and sediments makes the study area ideal for examining the processes that control barrier evolution over a variety of timescales. For example, here it is possible to assess the impact of an individual storm event (or cluster of storms, such as those of the 12th and 13th century AD) on the complex, or the response of the landform to century to millennial-scale changes in RSL. Thirdly, the back-barrier marshland complex contains a very extensive suite of minerogenic (clays, silts and sands) as well as organic (saltmarsh, freshwater reedswamp, fen carr, and raised bog) deposits preserved as peats. Palaeoenvironmental analyses of these deposits provide insights into past environmental changes in the study area, including shifts in climate and prehistoric and more recent land-use history. Fourthly, human occupation and resource use during historic and pre-historic times is strongly coupled with the environmental and landscape histories of the barrier and back-barrier systems, as well as their contiguous river basins and coastal waters. In this respect, unravelling the natural history of the region presents us with detailed insights into the lives of former peoples who lived and worked on the gravel foreland and marshland–perhaps being opportunistic in the first instance but exerting progressively greater imprint on the landscape during the archaeological past. These and other reasons that we hope to explain in this monograph, make the Romney Marsh/Dungeness Foreland depositional complex of more than simply local interest; indeed we believe that its sediments and landforms record the interplay of complex global, regional and local processes that combine to shape one of the most fascinating parts of the British coastline.

1.2 This monograph

This monograph reports the results of two interdisciplinary research projects conducted under English Heritage’s Aggregate Levy Sustainability Fund research programme. The projects concern the landscape history of two areas of the Romney Marsh/Dungeness Foreland depositional complex (Figure 1.1), with a particular focus on the interplay between natural and human processes during the late Holocene (from about 3000 cal. yrs BC to present). Although conducted as distinct projects, the two programmes of research are complementary and this monograph provides an opportunity to present the results of each in a single integrated volume. The field and laboratory aspects of the research presented here were completed over an 18 month period, commencing in November 2002. This research generated a large data archive, a brief public outreach report (Long et al. 2004), and two technical reports (Roberts & Plater 2005; Plater et al. 2006). There followed a 12-month publication contract that enabled the formal publication of our research findings, first in a series of specialist research papers (Long et al. 2006a; Long et al. 2006b; Roberts & Plater 2007; Schofield & Waller 2005; Stupples & Plater 2006; Waller & Schofield 2007; Waller et al. 2006). The aim of this monograph is to draw together the research conducted under the Dungeness Foreland and Rye projects, thus far published in a series of separate papers, into a single publication, and also to present aspects of the two projects not previously published.

e9781782974871_i0002.jpg

Figure 1.1 Location map of the Romney Marsh / Dungeness Foreland depositional complex.

A large number of individuals and organisations have been involved in the Dungeness Foreland and Rye area projects. The core academic staff of Antony Long (Durham University), Andy Plater (Liverpool University) and Martyn Waller (Kingston University), were supported by research staff located at each respective institution (Damien Laidler, Paul Stupples and Ed Schofield). An Optically Stimulated Luminescence (OSL) dating programme was conducted by Helen Roberts (Aberystwyth University) in conjunction with Andy Plater, radiocarbon dating support was provided by Alex Bayliss and John Meadows of the English Heritage Scientific Dating Section, and palaeomagnetic (PSV) dating was undertaken in collaboration with John Shaw and Sigrid Hemetsberger (Geomagnetism Laboratory, University of Liverpool). Peter Wilson was project manager for English Heritage and provided us with excellent support throughout. Much of the work presented here has benefited from the support of the Romney Marsh Research Trust, both during the project itself but also in the years prior to this work when much of the groundwork for our investigations was laid.

1.3 The study area

The Romney Marsh/Dungeness Foreland depositional complex comprises three main landscape units; first a series of valleys that drain from the Weald and enter into the western margin of the study area (the Rother, Brede, Tillingham and Pannel, Figure 1.3), second an expansive coastal lowland of now reclaimed marshland (Romney, Walland and Denge Marshes), third an area of sand and gravel beaches (Dungeness Foreland). The deposits at Dungeness Foreland, along with sand and gravel that occurs submerged beneath younger marshland sediments to the west, are the present incarnation of a once extensive barrier system behind which the low-energy back-barrier sediments of the marshland accumulated.

e9781782974871_i0003.jpg

Figure 1.2a Photographic montage of the gravel beach ridges on Denge Beach, with Dungeness Nuclear Power Station behind (Roland Gehrels).

e9781782974871_i0004.jpg

Figure 1.2b Photographic montage of the gravel beaches viewed from the Dungeness Lighthouse (Roland Gehrels).

e9781782974871_i0005.jpg

Figure 1.3 Typical valley marshland landscape. The view is from Cadborough Cliff in a northwest direction towards Rye.

The solid geology of the study area comprises the Lower Cretaceous Hastings Beds Group (Allen 1975) which is divided into three formations (Ashdown Beds, Tunbridge Wells Sands and Wadhurst Clay) all of which include sandstones, siltstones and clay. The bedrock topography of the valleys is reasonably-well known, at least in their middle parts where hand-coring has penetrated through the entire alluvial (Holocene) sediment sequence to depths of c. 10 m to 15 m (Waller et al. 1988). In the lower reaches, occasional deep boreholes penetrate to bedrock and demonstrate considerably greater sediment thicknesses, up to 30 m in the Rother, Brede and Tillingham valleys (Shephard-Thorn et al. 1966; Smart et al. 1966; Long et al. 1996; Waller et al. 1988; Waller & Kirby 2002). Across much of Walland and Romney Marshes there is little information regarding bedrock depths, although boreholes sunk close to Dungeness Point reached bedrock at depths of c. -30 m OD (Greensmith & Gutmanis 1990).

The marshland sediments comprise four main stratigraphic units above bedrock; a lower sand, blue clay, the main marsh peat and an uppermost deposit of young alluvium (Green 1968). Extensive shallow boreholes (typically less than 10 m depth) confirm that this general sequence is extensive across Walland Marsh and the western parts of Romney Marsh (Waller et al. 1988; Long & Innes 1995; Long et al. 1998; Spencer et al. 1998a). Microfossil analyses demonstrate that the lower sand and blue clay accumulated under intertidal sand or mudflat conditions, that were replaced from the mid-Holocene onwards, by saltmarsh and then freshwater peat-forming communities (e.g. Waller et al. 1999). In general, the main marsh peat thins in an easterly direction and is absent across much of Romney Marsh proper, probably due to post-depositional erosion. The young alluvium is a complex deposit comprising various amounts of sands, silts and clays (Green 1968). These uppermost sediments document several phases of late Holocene tidal inundation and include extensive tidal channel deposits; laminated sands and muds that attain thicknesses of up to 6 m or more that in places accumulated very rapidly (e.g. Waller et al. 1988; Long & Innes 1995; Spencer et al. 1998a; Stupples 2002a). The marshland sediments abut against, and in places interdigitate with (e.g. on Broomhill Level and Scotney Marsh), the sands and gravels of Dungeness Foreland (Tooley & Switsur 1988; Spencer et al. 1998a).

Evidence for human occupation of the marshland is often found in association with the young alluvium. For example, pottery scatters from Romney Marsh proper suggest settlement during the Romano-British period (Reeves 1995), excavations near Lydd have revealed an extensive Medieval field system (Barber 1998b) and mounds on Denge Marsh provide evidence of Medieval salt-making (Vollans 1995). Most remarkable are an extensive series of earthworks associated both with the reclamation of the former intertidal areas and their conversion into agricultural land (e.g. the Wainway Wall, the Midley Wall; Allen 1996; Gardiner 2002) and sea defence (Eddison 2000). The region is also rich in documentary sources which can provide complementary evidence for landscape development, mostly notably a series of early Medieval charters (Brooks 1988).

1.4 The projects

1.4.1 Dungeness Foreland

One of the two projects was concerned with the depositional history of Dungeness Foreland. The foreland beaches have long attracted humans, with the oldest evidence for human activity being a group of five bronze low-flanged axes recovered from a quarry north of Lydd in 1985 (Needham 1988). Fire-cracked and worked flints, as well as some Bronze Age pottery, have also been identified as surface scatters on the gravel near Lydd (Barber pers. comm. 2002). During the Romano-British period saltworking developed as an important local industry, with saltwater trapped between the shingle beaches and evaporated over fires positioned on the higher beach crests (Barber 1998a). There followed a gap in activity, which may be real or an artefact of sampling, until the onset of land claim from the 12th century AD which continued through until the early 16th century AD (Barber 1998b). These beach deposits are attractive today as a major source of aggregate for UK industry, and a long history of sand and gravel extraction has resulted in the partial destruction of the beach ridges and their associated potential as a palaeoenvironmental resource.

Prior to this project, we knew remarkably little about the absolute age and depositional history of the 500 or so beach ridges that comprise Dungeness Foreland. Existing models suggested that the beaches east of Holmstone were emplaced from the Romano-British period onwards, but the chronology for beach ridge deposition was reliant on limiting radiocarbon dates and archaeological finds on the gravel surface, as well as palaeoenvironmental and historical records relating to the age of marshland landscapes which abut the gravel (Lewis 1932; Lewis & Balchin 1940; Eddison 1983a; Long & Hughes 1995; Long & Innes 1995; Plater 1992; Plater & Long 1995; Plater et al. 2002). Consequently, not only was it unclear how the foreland developed, but also the impact (and inter-dependence) which the evolution of the Dungeness Foreland had on the wider landscape history of the depositional complex.

The Dungeness Foreland project comprised a large scale survey of the gravel beach complex, using OSL and radiocarbon dating techniques to determine the age of gravel deposition, and a combination of palaeomagnetic dating, stratigraphic and grain size analyses to establish the age and depositional environment of the finer-grained post-gravel sediments. The aim of the work was to develop a macroscale chronology for landscape change during the mid- to late Holocene.

1.4.2 Rye area

The second project concerned the late Holocene landscape history of the southwest corner of the depositional complex (the Rye area). Here the evolution of the sand and gravel barrier system played a pivotal role in the growth and demise of the towns of Winchelsea and Rye as major ports. Their interlinked histories tell a story of the shifting balance between natural and human processes as agents of landscape change. During the prehistoric period, the former controlled coastal evolution. Thus, prior to c. 1000 cal. yrs BC, the valley and marshland areas adjacent to Rye, lying in a protected position behind the gravel barrier, saw the extensive accumulation of peat (Waller et al. 1988; Long et al. 1996). After c. 1000 cal. yrs BC, the rate of peat formation seems to have declined and then these areas were subject to marine inundation. The earliest attempts at reclamation predate well-documented episodes of flooding associated with storms in the second half of the 13th century (Eddison 1998). Thereafter physical processes (variations in the supply of sand and gravel, alterations in the magnitude and frequency of sea-level change and storminess) vied with human activity (land claim and sea defence works) as agents of landscape change.

Prior to the late-14th century, Winchelsea, in its Old and New incarnations, was the dominant port in the region. However, a combination of factors, including the growth of the gravel barrier across the lower part of the Brede Valley, undermined its prosperity (Martin & Martin 2004). Trade then shifted to Rye, which during the late medieval and Tudor periods, enjoyed the only major harbour of refuge between Portsmouth and the Thames (Mayhew 1987). Thereafter, the deterioration of the harbour as a result of further gravel accumulation and reclamation in the Rother Levels, led to a major economic crisis during the late-16th and early-17th centuries (Eddison 1988). By the Restoration the fishing industry had all but disappeared, overseas trade was greatly reduced, and Rye had contracted to a small market town (Figure 1.4).

e9781782974871_i0006.jpg

Figure 1.4 The remaining salt marsh that fringes the present Rother Estuary, with Rye in the distance.

As at Dungeness Foreland, the gravel beaches which lie both to the west and east of the River Rother have a history of sand and gravel extraction which has resulted in the partial destruction of the beach ridges and their associated potential as a palaeoenvironmental resource. The Rye area project comprised an intensive study of the late Holocene deposits, with particular attention paid to providing a chronology for the end of peat formation, examining the impact of coastal flooding and the development and subsequent reclamation of the tidal channels. The overall aim of the project was to develop a model for the evolution of the Rye area during the last 3000 years, paying particular attention to the landscape changes associated with the development of the sand and gravel beaches.

1.5 Structure of the monograph

This monograph presents the results of first the Rye area and then the Dungeness Foreland projects, before integrating the results into a single model for the evolution of the depositional complex (Figure 1.5). The two projects have generated a rather large quantity of field and laboratory data and whilst an important element of this monograph is the presentation of these data, we seek to do this whilst simultaneously discussing key issues that arise from our work. Thus, in Chapter 2, we focus on pollen-based investigations into the landscape and vegetation history of Rye area during the late Holocene (up to c. 1000 cal. yrs AD). This is followed in Chapter 3 by a detailed re-assessment of the timing of the end of peat formation and of marine inundation and review of the associated landscape changes. Included in this is a discussion of the role of peat compaction in causing rapid landscape change, as well as the history of the large tidal channels that played an important role in the development and demise of Winchelsea and Rye as ports.

We shift our focus to the Dungeness Foreland depositional complex in Chapter 4 and start by presenting the results of our programme of deep drilling and OSL dating of the sub-gravel sand body, and our palaeoenvironmental examination of the post-gravel minerogenic sediments. A provisional age model for the foreland is proposed based on these new data. This is followed, in Chapter 5, by a review of a series of natural water-logged depressions on the surface of Dungeness Foreland that provide additional information on the chronology of foreland evolution as well as the vegetation history of the area.

In Chapter 6 we draw the Rye and Dungeness Foreland research together by developing a single stratigraphic model for the Romney Marsh/Dungeness Foreland depositional complex that unites the two study areas, together with previously published data from the adjacent marshland. A series of palaeogeographic maps emphasise the importance of three large tidal inlets (at Hythe, Romney and Rye) that have, at various times, enabled significant cross-barrier exchange of sediment and water and exerted a major control on the landscape history of the depositional complex. The monograph concludes (Chapter 7) by summarising our main findings and highlighting future research opportunities.

e9781782974871_i0007.jpg

Figure 1.5 Location map of the study area showing the geographical focus of Chapters 2, 3, 4 and 5. Chapter 6 addresses the wider study area of the Romney Marsh/Dungeness Foreland depositional complex.

Acknowledgements

We thank our collaborators in this project for their assistance in the field, the laboratory and in the development of the ideas included in this monograph, particularly Dr Damien Laidler, Dr Jonathan Lageard, Kate Elmore, Dr Cecile Baeteman, Dr Simon Turner and Dr Phil Woodworth.

We thank Alex Bayliss and John Meadows of the English Heritage Scientific Dating Section for their contribution to the radiometric dating programme and discussions of the chronology and Dr Geoff Duller (University of Wales, Aberystwyth) for discussions of the OSL data. Professor John Shaw and Sigrid Hemetsberger (Geomagnetism Laboratory, University of Liverpool) are acknowledged for their work in establishing a palaeomagnetic secular variation chronology for the marshland sediments of Dungeness Foreland.

Dr Andy Woodcock, Dr Mark Gardiner, Dr Luke Barber, Dr Alan Tyler and Jim Culley provided assistance with the archaeological data. Jill Eddison, Dr Gill Draper, Dorothy Beck, Beryl Coates and colleagues of the Romney Marsh Research Trust provided valuable advice on historical sources.

The deep cores from Dungeness Foreland were expertly collected by Melvyn and Dean Stupple of Strata Investigation Services. Lorraine Morrison (University of Wales, Aberystwyth) assisted with chemical preparation of the samples for OSL dating. Frank Davies, Eddie Million, Neil Tunstall (Durham University) Hilda Hull, Irene Cooper, Bob Jude, Alan Henderson and Tim Ellis (University of Liverpool) also assisted with the laboratory work.

We thank Dr Jason Kirby for a detailed and constructive review of an initial draft of this work. The diagrams were drawn by David Hume, Chris Orton and Steve Allen (Durham University), Sandra Mather (University of Liverpool) and Claire Ivison (Kingston University). Unless stated, the photographs were taken by the lead authors. Lianne Percival and Niamh McElherron provided grateful assistance in the final compilation of the text.

Many thanks to the landowners who gave us access, particularly to Jon Hickes and Major O’Reilly of the Ministry of Defence, Brian Banks of Natural England and to the Royal Society for the Protection of Birds.

This monograph is a contribution to IGCP Project 495 Quaternary Land-Ocean Interactions: Driving Mechanisms and Coastal Responses and to the INQUA Sub-Commission on Coastal Processes and Sea-level Changes.

2

The Rye Area: Pre-inundation Landscape and Vegetation History

Martyn P. Waller and J. Edward Schofield

2.1 Introduction

This chapter outlines the work undertaken on the organic deposits of the Rye area, describing in detail the stratigraphy, palaeoecology and chronology of the sites investigated. The discussion focuses on pollen-based investigations into the landscape and vegetation history of the region during the late Holocene. The pollen preserved within the organic sediments of the Rye area were derived both from the wetland communities of the valleys and marshland and the surrounding upland of the High Weald. The palaeoenvironmental data presented in this chapter address a number of unresolved issues relating to the late Holocene landscape history of both these regions. The interpretation of data collected from the upper-surface of the peat in terms of the timing of the marine inundation of the Rye area is not dealt with here but is presented and discussed in Chapter 3.

The nature of the peat-forming plant communities and spatial and temporal trends in wetland development across the Romney Marsh region are well-documented for the mid-Holocene (Waller 2002). Little information is, however, currently available on the evolution of the Rye wetlands for the timeframe between the end of widespread peat formation, which may have occurred as early as c. 1500 cal. yrs BC (Waller 2002), and reclamation following the marine flooding of the 13th century AD. Of particular interest with regard to late Holocene wetland development in this region are:

–The environmental history of the valleys after c. 1500 cal. yrs BC. Detailed investigations are required from the upper levels of the widespread peat layer to provide information on changes in the hydrological and nutrient status of wetland areas prior to marine inundation.

–The occurrence of surface and near-surface peat in the valleys and bog sediments of Walland Marsh. These deposits potentially contain palynological evidence spanning the late Holocene for the evolution of wetland communities and coastline movements.

–The relationship between deforestation and sedimentation in the valleys and marshland. Woodland clearance on the uplands would be expected to increase both the supply of clastic sediment and water to the valleys. In the upper parts of some Wealden valleys thick sequences of inorganic sediment built up following prehistoric deforestation (Scaife & Burrin 1983; 1985; 1987). Although deforestation has been documented in the Rye area from c. 2000 cal. yrs BC (Waller 1994a) the impact of this activity on the evolution of the wetland areas is still unclear (Waller et al. 1999).

The uplands adjacent to Rye form part of the High Weald. Today woodland covers c. 30% of the High Weald making this area unusually well-wooded for lowland England. Over the past 1,000 years this woodland has been exploited for seasonal pasture and as coppice, though its long-term history is poorly understood. Palynological data from near Rye and other areas bordering the High Weald provide detailed information on woodland composition during the mid-Holocene (Tilia was at least locally dominant) prior to extensive human interference from the Early Bronze Age onwards (Smyth & Jennings 1988; Waller 1993; 1994a; Waller 2002). Pollen sequences which cover substantial periods of time within the last 4,000 years are, however, lacking. Of particular interest with regard to the upland (dryland) vegetation history in this region are:

–Variations in the timing of woodland clearance episodes and subsequent patterns of land-use, which may be related to differences in soil type and geology.

–The environmental impacts of the early iron industry. During the Roman period a distinct group of iron working sites in the eastern Weald are thought to have had an outlet to the sea, via the rivers Rother and Brede (Cleere & Crossley 1995). Although the availability of fuel, alongside iron ore, is likely to have been important in its development, little is know about the impact of the early iron industry on the extent and composition of woodland cover. In addition, it has been suggested, partially on the basis of the absence of agricultural or urban settlements, that the eastern group of iron working sites formed part of an ‘imperial estate’ from which non-iron working activity was excluded (Cleere & Crossley 1995).

–The early Anglo-Saxon period in the Weald is associated with transhumance, in particular with swine being driven into the region in the autumn. These wood-pastures (dens) were linked by a series of droves that radiated out from the High Weald to parent settlements that appear to have been located largely on the coastal plains and at the foot of the chalkland (Whitney 1976; Brandon 2003). Little is known as to the origins of this system and its impact on woodland composition.

–Two common constituents of the modern woodlands of the Weald are Fagus sylvatica (beech) and Carpinus betulus (hornbeam). The pollen of both taxa is scarce in the mid-Holocene pollen records of the region implying their populations expanded later. A consequence of the lack of long late Holocene pollen sequences is that these expansions are poorly understood.

2.2 Study area

The town of Rye stands on a promontory overlooking extensive areas of reclaimed marshland (Figure 2.1). To the east lies Walland Marsh, and to the south-west is Brede Level which is separated from the sea by a series of gravel ridges in the vicinity of Camber Castle. The modern land surface across the former marshlands lies between +1 and +4 m OD and is protected from inundation by the sea by the presence of sea walls, the gravel ridges and sluices. Traditional sheep pastures have, as a consequence of improved drainage, increasingly been replaced by cereal production.

To the west of Rye a series of west-east trending valleys (the Tillingham, Brede and Pannel) rise up into the High Weald (Figure 2.2). The interfluves separating these valleys coalesce to form substantial plateaulands at altitudes over 100 m OD. The local geology in this area comprises the three formations (Ashdown Beds, Tunbridge Wells Sands and Wadhurst Clay) of the Hastings Bed Group. On the permeable lithologies, brown-earths and argillic brown-earths have developed while stagnogleys occur where drainage is impeded (McRae & Burnham 1975). Today the region is dominated by low grade agricultural land and woodland, with the proportion of woodland cover close to the average for the High Weald.

2.3 Previous stratigraphic investigations of Holocene deposits in the Rye area

Although the near-surface sediments of areas to the east of the Rother (Wall and Marsh) are included in the memoir of the Soil Survey (Green 1968), no investigations of the deeper deposits or the sediments to the west of the river were undertaken until the 1980s. Subsequently the Holocene deposits and palaeoenvironments of the valleys (Waller et al. 1988; Waller 1993; 1994a; 1998; Waller & Kirby 2002), Brede Level (Long et al. 1996), Pett Level (Marlow 1984; Waller et al. 1988), and the deeper sediments of Walland Marsh (Long & Innes 1995; Waller et al. 1999; Evans et al. 2001; Stupples 2002a) were investigated in some detail. In addition to academic research, a large number of boreholes have been sunk in the region over the last 30 years for commercial purposes (Waller et al. 1988; Long et al. 1996).

The deepest deposits recorded over Walland Marsh and Pett Level consist of a blue-grey silty sand which extends to a maximum altitude of -1 m OD and represents a phase of tidal flat sedimentation (Marlow 1984; Long & Innes 1995; Long et al. 1996). Data from Brede Level and the lower Brede valley (Waller et al. 1988; Waller 1994a) indicate a lateral facies change into finer grained sediments occurs landwards, while a fining upwards sequence into blue-grey silty clays is ubiquitous. After gradational contacts the clays are overlain by a laterally persistent peat layer which is only absent from areas known to have been inundated after the 13th century AD storms (i.e. the gravel complex associated with Camber Castle, the Wainway Channel and other channel systems). The peat is thickest (maximum c. 6.5 m) in the middle sections of the valleys where it generally occurs within 1 m of the present surface (Waller 1994a). Over most of the lower valleys and marshland it is more deeply buried beneath marine/brackish clays and silts (Figure 2.3). Occasional surface outcrops of peat do however occur on Walland Marsh (Green 1968) and at a few valley locations (Waller 1993).

Peat formation commenced in the valleys c. 4500 cal. yrs BC before extending out across Walland Marsh by c. 2900 cal. yrs BC (Long, A. J. et al. 1998). The upper layers of the peat are often highly humified and the contacts with the overlying sediments generally sharp. Radiocarbon dates suggest a decline in the rate of peat accumulation in the valleys occurred sometime after c. 2000 cal. yrs BC (Waller 2002). Variations in the altitude of the upper peat contact appear to be related to the depth of overlying sediment suggesting that the upper layers of the peat bed have probably been subject to post-depositional compaction. Radiocarbon dates from the upper surface of the peat show considerable spatial and temporal variation. At some locations peripheral to Walland Marsh radiocarbon dating suggests peat formation ceased as early as c. 1500 cal. yrs BC though in the lower Brede it persisted until c. 400 cal yrs AD (Waller 2002). On Walland Marsh peat continued to form locally until at least 1000 cal. yrs AD (Waller et al. 1999).

e9781782974871_i0008.jpg

Figure 2.1 Location map: (a) the location of the Weald in south-eastern England, (b) the Weald showing the location of the study area, (c) the Rye study area showing the location of the places and sites referred to in the text. The former western extent of the raised bog on Walland Marsh is likely to have been greater than the limits shown on the map as marine erosion has occurred along the western and southern margins, since c. 700 AD.

The peat accumulated in a range of depositional environments (Waller 1993; 1994a; 2002; Waller et al. 1999). At most valley locations the pollen and macrofossil remains of fen carr taxa are abundant. Alnus glutinosa appears to have been generally dominant, though macrofossils from a range of other taxa including Viburnum opulus, Osmunda regalis and Filipendula ulmaria have been recovered (Waller 2002). On Walland Marsh, taxa (e.g. Betula, Myrica