Archaeological and palaeoenvironmental investigations at Pannel Bridge, near Pett Level, East Sussex

Introduction The deposits of the Pannel valley (Fig. 4.1) have been the subject of two palaeoenvironmental studies within the last four years. A borehole survey to record the lithostratigraphy of the valley was undertaken between 1983 and 1984 (Woodcock 1984), while a more detailed study around Pannel Bridge (TQ 882152) was in progress (Waller 1982-1985) as part of research into the Flandrian vegetational history of the Brede and Pannel valleys (Waller et al., 1988). The investigations at Pannel Bridge revealed one of the few locations in south-east England with organic deposits dating back to the early Flandrian period. The pollen and plant macrofossil assenlblages at this site record the pattern ofvegetational change in this valley through the Flandrian period. An absolute chronology has been established with a series of radiocarbon assays (Table 1.3, Waller et al. 1988). Fieldwalking within the Pannel valley by the Hastings Area Archaeological Research Group (HAARG) has located a number of places rich in archaeological material dating from the mesolithic period (Fig. 4.2). One such place, adjacent to a spring on the northern edge of the field immediately east of Pannel Bridge, was also sampled during Woodcock's borehole survey and found to contain a considerable quantity of mesolithic flints, possibly in situ, buried under a thin layer of colluvium. Given the potential for recording an undisturbed mesolithic site and other archaeological material in close proximity to contemporary peat deposits, a field project was undertaken to investigate the impact of prehistoric human activity on the landscape of the Pannel valley. This paper is a statement of the results achieved so far. The Archaeological Survey Aims and objectives The initial aims of the project were to reconstruct the patterns of settlement and land use of the Pannel valley to the east of Pannel Bridge during and since the mesolithic period and to relate this sequence to the vegetational history of the valley. A two-week fieldwork season was carried out in September 1986 with the following objectives: 1. 2. 3. to investigate by excavation colluvial deposits on the lower valley slopes; to define the extent and character of past human activity by using sample excavation and surface collection survey to record the distribution of discarded artefacts; and to relate the evidence of human activity located by excavation and survey to the pollen and plant macrofossil record. Excavations The first excavation, trench A (Figs. 4.1 C and 4.3), was positioned on the northern edge of the Pannel valley at the place where mesolithic material had previously been found, with the intention of recording the relationship between colluvial and peat deposits, and determining whether or not in situ mesolithic material was preserved. In trench A, a series of drainage ditches dug during and after the Second World War, along with earthmoving activities associated with trackway construction (Fig. 4.3) have destroyed all traces of colluvial deposits adjacent to the peat (layer 23). Trenches AI and A2 were then excavated to the north of the 'zone of destruction', revealing a thin layer of colluvium which is likely to be of medieval or post-medieval date. A dense scatter of mesolithic flints was located beneath the disturbed layers in trench A (Fig. 4.3). The presence of fragments of late medieval pottery amongst this material shows that artefacts have been moved downwards in the soil profile by natural processes e.g. earthworm activity. It was not possible to assess how far these flints have moved laterally since they were deposited, but the presence of small chips, a variety of flakes and blades, a core and a hammerstone, suggest that flintworking took place at or near to this locality. Flints were also recovered from the disturbed layers in trench A. In order to define the extent of the mesolithic flint scatter alongside the peat deposits, trenches B1-B10, D1 and El were excavated (Fig. 4.1C). Trenches B1-B4, B7-B 10 and E l all produced mesolithic flints; those from trenches B7-B10 and E l were from disturbed contexts, but the flints from the lower levels of trenches B1-B4 might rest in a similar position to those retrieved from layer 4 in trench A. Archaeological and Palaeoenvironmental Investigations 7 3 74 Robin Holgate and Andreui Woodcock Fig. 4.2 Pannel Valley: Surface Collection Survey 1986. Trenches Cl-C6 (Fig. 4.1 C) were excavated to define the northern limit of colluvial deposits and to search for undisturbed mesolithic material. In all cases, recent ploughing had truncated any colluvial deposits that might have existed at one time, and dispersed any previously in situ artefacts throughout the plough soil. Mesolithic flints were recovered from trenches Cl-C5, whilst one piece of possible Romano-British pottery and a late medieval sherd were retrieved from the surface of the subsoil in trench Cl. Surface collection survey The surface distribution of archaeological material was recorded in three fields (Fig. 4.2). In all cases, prehistoric flints and medieval pottery were recovered. The majority of the flints were mesolithic in date and largely occur in clusters at 5-10 m OD on the valley slopes overlooking the Pannel Sewer. Artefacts The excavations produced 415 flints; these are summarized in Table 4.1. The flint was largely dark grey, dark brown or light orange in colour. Cortex, where present, was thin and heavily abraded. Most of the flint had originally been obtained from beach deposits, the location ofwhich is unknown. Nearly half of the flints, including three of the microliths, came from trench A. A number of the flakes, the flake cores, some of the scrapers, the scraper-knife combination tool and piercer are hard hammer-struck and could be of later neolithiclearlier Bronze Age date. The remaining flints, most of which have acquired a blue-white patination, probably date to the mesolithic period. The microliths are of the narrow-blade variety and include a small obliquely-blunted point, a blunted-down-one-edge microlith, a two-edge-blunted 'rod' and a small mis-hit microlith. In south-east England, these types usually occur in assemblages of later mesolithic date (Jacobi 1978). A selection of the unpatinated flints was examined under a microscope to see if use-wear traces were discernible. In all instances, post-depositional surface modification had taken place, obscuring any polishes and striations that might have resulted from use (Roger Grace, personal communication). The flints recovered during the surface collection survey consist largely of mesolithic debitage. Four scrapers were found which, along with several of the hard hammer-struck flakes and a discoidal core, could date to the later neolithic period or earlier Bronze Age. A number of small, abraded fragments of pottery were collected during the survey, including a range of medieval fabrics from flint and sand- to sand and shell- tempered wares, which were probably emplaced during medieval manuring practices. Traces of ridge-and- furrow which could possibly date back to the medieval period are visible in some of the fields on the upper slopes Archaeological and Palaeoenuironmental Investigations 75 of the Pannel valley which have not been ploughed in recent years (Fig. 4.2). Mesolithic occupation Human activity in the Pannel valley has been recorded during the later mesolithic period, the late neolithic periodlearlier Bronze Age and after the Norman Conquest. While these excavations and surface collection survey only sampled a small part of the Pannel valley, the material collected by HAARG from a total of about 15 fields (Fig. 4.2) includes a similar range of flints and medieval pottery, although a cluster of both earlier and later neolithic flints (Fig. 4.2) and a small quantity of Romano-British and Saxo-Norman pottery has also been retrieved (Vahey 1986, 45). In all, the mesolithic flints consist mostly of debitage, microliths and a few cutting blades; the only other implement of note is a small tranchet axe found in the field south-west of Pannel Bridge. These remains are suggestive of little more than a series of short-stay camps, largely sited either around springs or on the lower valley slopes overlooking the Pannel Sewer. Burnt flints and charcoal produced by the excavations in Trench A could have resulted from hearths, but the quantities involved are insufficient to enable firm conclusions to be drawn on how extensively fires were used at this site. Given the long time span over which mesolithic activity could have taken place, there is nothing to demonstrate significant interference with the landscape before the end of the mesolithic period which could have led to dramatic changes in the natural vegetational cover along the Pannel valley. Indeed, on present archaeological evidence, the impact of human activity in the valley is likely to have been minimal throughout the prehistoric and early historic periods. Flandrian vegetational changes Martyn Waller At c. 10,000 BP organic accumulation was confined to a relatively narrow zone in response to locally high water levels. Pinus woodland occurred on the surrounding slopes. Between c. 9,400 BP and c. 7,000 BP dry land environments in the Pannel valley were heavily wooded with deciduous trees, particularly Corylus. After c. 7,000 BP Tilia and Q~uercus were the major arboreal components until extensive forest clearance (the Tilia decline) occurs at 3,700f BP (SRR 2887). Between c. 8,000 BP and c. 6,000 BP the vegetation of the local environment (the mire communities occupying the floodplain) alternated between periods of Alnus and Cyperacea domination, probably in response to changing environmental conditions downstream. Intercalated organic and clastic sediments (of estuarine origin) were found to the east of the site, though largely organogenic accumulation was maintained at Pannel Bridge. After c. 6,000 BP, until the Tilia decline, the pollen 76 Flakes Blades Bladelets Chips Crested blade Core tablets Core rejuvenation flake Bladelet core Flake cores Shattered pieces End scrapers Side scraper Side scraper-knife combination tool Cutting flakes or blades Piercer Microliths Fire-fractured flints TOTAL Robin Holgate and Andrew Woodcock Table 4.1 Flints recovered from the excavations. assemblages at Pannel Bridge suggest that conditions prevailed in the local environment (Alder-dominated fen) which were similar to those described from the adjacent Brede valley (Waller et al. 1988). The fluctuations recorded in the pollen assemblages after the decline in Tilia appear largely to be the result of changing conditions in the local environment. However, differences in the intensity of anthropogenic activity may also account for changing proportions of arboreal and non-arboreal pollen evident between c. 3,700 BP and c. 2,600 BP. Alnus re-exerted its domination in the local environment after c. 2,000 BP, though it is not certain whether such communities continued into historic times as modern root penetration prevented the radiocarbon dating of the near surface sediment. Conclusions Although only a small part of the Pannel valley has been surveyed, archaeological information recovered to date References Jacobi, R. 1978: The Mesolithic ofSussex. In Drewett, P. L. (editor), Archaeology in Sussex to AD 1500, (London, CBA Res. Rep. 29), 15-22. Vahey, Z. 1986: Handmade Pottery from Pett and Icklesham. Journ. Hastings Area Archaeol. Res. Group 7, 45-56. Waller, M., Burrin, P. J. and Marlow, A. 1988: Flandrian suggests that pre-medieval exploitation of the Pannel valley was not particularly intensive. In the later mesolithic period short-stay camps were positioned on the lower slopes, but this activity may have had minimal effect on the vegetational cover within the valley. By the earlier Bronze Age, human activity of some description had spread on to the lower valley slopes and is likely to have played a part in the ensuing Tilia decline. Thereafter, there is limited archaeological evidence for human activity within the Pannel valley until the expansion of arable farming in the medieval period. Further work needs to be done. The surface collection survey should be extended to cover a larger area of the valley, possibly followed by sample excavation in the hope of recovering undisturbed archaeological material. The extent of ridge-and-furrow and the remains of other earthworks should also be mapped by ground survey and aerial photography. And finally, an extensive survey of the surrounding region should be undertaken to determine whether the archaeology of the Pannel valley is a microcosm for the Wealden valleys that drain into the sea between Hastings and Hythe. Clearly, the combination of systematic field survey, sample excav- ations and palaeoenvironmental studies is an effective method of reconstructing the way in which past landscapes were settled and exploited. This paper is thus a signpost, giving an indication of the kind of fieldwork that should be undertaken to establish the pattern of prehistoric and later occupation in and around Romney Marsh. Acknowledgements We are grateful to the following: Mr. Stephen Rumsey, for granting permission to carry out the fieldwork; Zoe Vahey, for her help on the site and providing access to the material collected by HAARG; Twigs Way, for supervising the excavations; David Rudling and Roger Grace, for examining the pottery and possible use-wear traces on the flint respectively; Andy Walters, for drawing the illustrations; and Martyn Waller for both contributing the section on Flandrian vegetational changes and commenting on the first draft of this paper. sedimentation and palaeoenvironments in Pett Level, the Brede and lower Rother valleys and Walland Marsh. In this volume, chapter 1. Woodcock, A. G. 1984: The Sedimentological History of the Pannel Valley, Pett, East Sussex. Unpublished M.Sc. Thesis, City of London Polytechnic and North London Polytechnic.