Cabot Park, Bristol Final Report

Plot P9, Cabot Park, Avonmouth Post Excavation Analysis May 2009 GGAT report no. 2009/034 Project no.P1264 National Grid Reference: ST 53896 79886

A report for ProLogis by Rowena Hart BSc (Hons) MA

1

2

4 3

6

TE GI S R

ED

RE

N

OR

I A

IO

5

G

A N I SAT

The Glamorgan-Gwent Archaeological Trust Ltd Heathfield House Heathfield Swansea SA1 6EL

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Contents

Page

Background ..................................................................................................................3 1. Palaeotopographic Background .............................................................................4 1.1

Geoarchaeological investigation at Stinkums................................................4

1.2

Palaeoenvironmental Summary .....................................................................6

2. Investigations of Late Bronze Age Sites.................................................................8 2.1

Introduction......................................................................................................8

2.2

Stratigraphic Description................................................................................8

2.3

Finds Reports .................................................................................................12 2.3.1 Prehistoric Pottery - Lorraine Mepham ...............................................12 2.3.2 Unworked Stone - Jana Horák .............................................................13 2.3.3 Animal bone - Martin Locock..............................................................19 2.3.4 Lithic Analysis - Richard Lewis ..........................................................20 2.3.5 Molluscs - Andy Sherman ...................................................................22

3. Discussion................................................................................................................24 Figures Figure 1. Site location (not to be included in final publication). .................................26 Figure 2. Trench and borehole locations at Stinkums – P9. ........................................27 Figure 3. Illustrations of prehistoric pottery 1-6 (scale 1:1) ........................................28 Figure 4. Photomicrographs of stone samples 1. .........................................................29 Figure 5. Photomicrographs of stone samples 2. .........................................................30 Figure 6. Grain sorting classification...........................................................................31 Figure 7. Grain shape classification.............................................................................31 Tables Table 1. Radiocarbon dates..........................................................................................32 Table 2. Bone condition (after Locock et al. 1992, 209).............................................33 Table 3. Summary of analysed bone............................................................................33 Table 4. Animal bone catalogue. .................................................................................34 Table 5. Grain size classification scale ........................................................................33 Table 6. Summary of flint assemblage ........................................................................35 Bibliography ...............................................................................................................36 Appendix I: Palaeoenvironmental Evidence at Stinkums (Martin R. Bates, M. Allen, J. Crowther, S. Davis, K. Griffiths, S. Jones, P. Robinson, E. Tetlow, J. Whittaker ) .................................................................................................................38 Appendix II: Notes on the Archive...........................................................................79

Stinkums Palaeoenvironmental Assessment Figures (not for final publication) SPA Figure 1. Borehole 9 stratigraphy ........................................................................59

1

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Stinkums Palaeoenvironmental Assessment Tables (not for final publication) SPA Table 1. Analytical data.....................................................................................62 SPA Table 2. Pollen data from Cabot Park BH9 .......................................................63 SPA Table 3. Diatom assessment results...................................................................66 SPA Table 4. Plant macro fossil remains from borehole 1........................................67 SPA Table 5. Plant macro fossil remains from boreholes 8 and 9 ............................68 SPA Table 6. Foram and ostracod assessment from borehole 1................................70 SPA Table 7. Foram and ostracod assessment from borehole 9................................73 SPA Table 8. Mollusc from the Cabot Park 9 borehole samples...............................75 SPA Table 9. Insect assessment of the samples from Borehole 2 .............................75 SPA Table 10. Insect assessment of the samples from Borehole 7 .............................75

2

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Background Prologis have commissioned the Glamorgan-Gwent Archaeological Trust Ltd (GGAT Projects) to undertake the post-excavation analysis and reporting work arising from discoveries made during archaeological and geoarchaeological investigations undertaken in 2007 at Plot P9 Cabot Park, Avonmouth. GGAT Projects undertook an earlier evaluation on Plot P9 as part of extensive site wide investigations carried out in 1998 for Burford and subsequently have undertaken further evaluations, surveys, excavations, post-excavation and watching-brief works at Cabot Park for Burford and Cabot Property Partnership as part of the mitigation works required in fulfilment of conditions attached to the main and various reserved matters planning consents. This document is the outcome of post-excavation analysis and reporting arising from the investigations at Plot 9, Cabot Park, Avonmouth undertaken in 2007. Postexcavation assessments for the two phases of fieldwork have previously been prepared (Askew 2007, Corcoran 2007). The following report fulfils the aims and objectives as outlined in the associated Project Design (Marvell 2008). The site at Stinkums is introduced by outlining the palaeotopographical phasing as initially outlined in the report produced for Kites Corner. There was an absence of phase 7 (post later Bronze Age post-depositional disturbance), phase 8 (post later Bronze Age continued or renewed soil formation), phase 10 (Iron Age/Roman land surface) and phase 11 (post-Roman alluviation) at Stinkums. A palaeoenvironmental assessment of the boreholes was undertaken by staff at the University of Wales, Lampeter Archaeology Service (UWLAS) and will be summarised in the final report and included here as Appendix I. This is followed by a description of all archaeological works that have been undertaken at Stinkums, assimilating results from previous work at the site. All classes of finds were subject to specialist analysis and specialist reports are provided. For publication purposes the main body of the report will be assimilated into the final report on works at Cabot Park which will describe the wider landscape at Cabot Park and include sites at Kites Corner, Little Googs 1 and 2 and the pond at P11 which have recorded more extensive remains than was recorded at Stinkums where much of the site has been preserved through agreed construction design. The full palaeoenvironmental report in Appendix I and the archive notes in Appendix II will not form part of the publishable material. The physical remains and records relating to these investigations will be incorporated into the main Cabot Park Archive, which will in due course be deposited with Bristol Museum.

3

Plot P9, Cabot Park, Avonmouth: post excavation analysis

1. Palaeotopographic Background 1.1

Geoarchaeological investigation at Stinkums

Phase 1: Mesolithic and Neolithic alluviation (10,000 – 4500 cal BP) Evidence of the earliest sequences was represented in boreholes 5-9. These sedimentary deposits form the lowest layers of the Wentlooge Formation. The highest level at which these were encountered was at 4.66m OD in borehole 5 and the lowest level at which they were encountered was in borehole 6 at 4.50m OD. The stratigraphic sequence was similar in each of the boreholes. Borehole 9 recovered data from the greatest depth (-0.7m OD) and provided two additional deposits/facies (1 and 2). The lowest was an inter tidal sand flat and was radiocarbon dated to 6980 – 6560 cal BC i.e. the middle Mesolithic (Table 1: Beta 253801). Overlying this is an intertidal mudflat and sequentially a reed bed, which was also present in boreholes 7 and 8 as the basal deposits recovered. Within the reed bed in boreholes 6-9 was a deposit of humic silty clay (facies 3). A radiocarbon date was obtained for this organic rich facies and measured 3940 to 3860 cal BC (Table 1: Beta 253800). The pollen and plant macro fossil remains of this deposit represent a period of stability and relief from the salt-water estuarine environment with carr-woodland and evidence of agricultural activities with pine, elm, oak, lime, alder and hazel growing in the vicinity along with cultivatable and edible herbs and grasses suitable for human and animal consumption (Chenopodiaceae, Poaceae etc). Sealing this organic layer was a sequence of inter tidal mudflats and salt marshes indicating a return to the estuarine conditions of the Mesolithic and early Neolithic periods. Phase 2: Neolithic land surface Overlying the Phase 1 alluvium as recorded by auger and evaluation trenches was a black organic clay encountered at the base of Trenches 1-3. No finds or archaeological features were found in this layer. This deposit is the ‘N-layer’, also termed the Baras-layer, and has been dated to 3700 – 2200 cal BC (the Neolithic Period see Table 1) and recorded during many archaeological works in the area. The uppermost level of this deposit was found at depths ranging from 4.56 – 4.28m OD, which is consistent with depths encountered at other sites in the Cabot Park region. The clays of this phase represent the remains of a ripened soil horizon, with vegetation cover and prone to periodic inundations. The pollen record for the N-layer in borehole 9 is indicative of a salt-marsh environment with pine, oak, alder and hazel growing in the vicinity (the pollen and plant macro fossil quantities were lower for this layer than for the earlier organic layer of Phase 1 (Tables 3, 5 and 6). Phase 3: Neolithic and Bronze Age alluviation (4500 – 3800 cal BP) A gleyed clay deposit and an alluvial silty clay deposit represented the stratigraphy of this phase. These deposits were indicative of yet another return to estuarine and associated salt marsh environments. Pollen preservation was poor with some indication of arboreal pollen (Table 3). Diatom floras from these deposits are indicative of coastal and brackish conditions. The few molluscs present also indicate salt marsh habitats. Forams and ostracods confirm an estuarine inter tidal mudflat environment, with initially quite a high degree of marine influence, which tails off over time, as the site appears to become more and more peripheral to marine access through regression of the sea and/or aggradation of the land. The environment is gradually presenting itself as a hospitable area once again.

4

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Phase 4: Palaeochannel A palaeochannel (08) was partially recorded on site although the details are not conclusive and the feature not excavated. A depth was attributed to the upper level of the palaoechannel as varying between 4.68 – 4.78m OD. It is noted in the evaluation report (Askew 2007: 18) that the eastern edge of the palaeochannel was cut into the gleyed clay (15) of Phase 3. During the 1998 evaluation (Locock 1998: 78) two palaeochannels were recorded in Trench 4 (GGAT Trench No. 44). They were both aligned north - south and were cut through the lower gley (502). One of the palaeochannels had a width of approximately 1.5m and the other a more substantial 15m. Phase 5: Late Bronze Age soil formation Overlying the estuarine alluvium and gleyed clays of Phase 3 was a gleyed buried soil comprising a stratified sequence of late Bronze Age occupation deposits (Phase 6). This soil formation respected the palaeochannels at Stinkums (GGAT Trench 44) and Little Googs 1 where it is probable that the palaeochannels were open and active during this period. By contrast, the palaeochannels at Kites Corner were sealed by this soil horizon and associated occupation deposits of late Bronze Age date. It is clear and perhaps not unexpected that not all palaeochannels were contemporary, with new channels developing as others silted up and become inactive. Phase 6: Later Bronze Age occupation layer (3800 – 3400 cal BP) The contexts attributed to Phase 6 represent an episode of human activity at P9. The occupation deposits comprise three thin deposits (measuring 10mm, 60mm and 100m in depth), which can be equated to the B-layer seen elsewhere at Cabot Park. These occupation deposits are referred to collectively as the B-layer. The compressed date range of the pottery and the shallow depths of the deposits in this phase suggests a short period of occupation. This is consistent with the evidence from Kites Corner and Little Googs 1 and 2. The upper level for this phase was encountered at 5.07m OD (approximately 1.20m below current ground level). This occupation deposit has been dated (C14 method) to 1120–410 cal BC (Table 1) as part of the current work. Previous dating at Rockingham Farm of the B-layer provided date ranges 1398–812 cal BC. This could suggest that the occupation at Rockingham Farm (mid to late Bronze Age) was somewhat earlier than that at Stinkums (mid Bronze Age to mid Iron Age). An overlap in the dates does not make this conclusive. Phase 9: Later alluviation This phase is represented by a deposit of estuarine alluvium. This comprises a greybrown mottled iron and manganese-stained silty clay with some organic inclusions. The upper level of this period of alluviation was encountered at between 5.34 - 5.48m OD, which is within the range of values encountered across earlier Cabot Park investigations. Phase 12: Current land surface (1000 – 0 BP) This final phase comprises the modern topsoil and subsoil. The depth of the subsoil varies considerably between 0.42m in the north of the evaluation area to 0.9m in the central part of the area and then shallows again toward the south The upper surface of the subsoil was encountered between 5.80 and 6.14m OD. The overlying topsoil has an average depth of 0.3m and represents the most modern deposit.

5

Plot P9, Cabot Park, Avonmouth: post excavation analysis

1.2 Palaeoenvironmental Summary (after Bates et al. Appendix I) The key aims of the study, which was undertaken by the University of Wales, Lampeter Archaeological Service (UWLAS) were to provide a chronological framework for the palaeoenvironmental sequences and to assess samples from cores inside and outside the palaeochannel. This work included assessing the presence of pollen, diatoms, plant macrofossils, molluscs, forams and ostracods and insects. In addition loss on ignition and magnetic susceptibility determinations were undertaken. The assessment of the material was undertaken in order to attain a clearer idea of the potential of the sequences to provide information on past landscapes. Key findings of the assessment are described below: Facies 1. Inter-tidal sand flats. Pollen is limited from this facies. Diatoms are present and indicate marine and coastal conditions, some associated with littoral communities. Forams and ostracods indicate a brackish, saltmarsh and true marine components exist that are explained by reference to a large open estuary or embayment around the site at this time. This infill incorporated not only the local estuarine sediments that were forming at the time, but a great deal of marine sediment thrown in by storm surges, including possible reworked Pleistocene material. A single date was obtained from forams collected from the lowermost sample that provides an age for onset of accumulation of this fossiliferous material of 7900+/-60 B.P. Facies 2. Inter-tidal mudflats (1st phase). Pollen from this facies is limited. Diatoms are present and indicate marine and coastal conditions, some associated with littoral communities. Forams and ostracods indicate that the marine influence begins to diminish through this facies. Also, perhaps significantly, there is no further input of “exotic” material. The sandflats turn to mudflats as the area becomes more and more peripheral to marine influence and as the local estuarine intertidal scenario gradually takes over. Facies 3. Reedbeds/wetlands. Pollen preservation is good with saltmarsh locally indicated with mixed oak woodland in the vicinity below 4.6m. Plant macrofossils suggest reed swamp environments with brackish conditions at this time and some charcoal. Up-sequence carrwoodland occurs as well as taxa associated with agricultural activities. This transformation is supported by a change in diatom floras with the appearance of supra-littoral species and species of freshwater preferences. The change in environmental conditions occurs around 4990+/-40 B.P. based on the 14C date obtained on peat at a depth of 4.61m. Facies 4. Inter-tidal mudflats (2nd phase). Pollen preservation is poor with some indication of arboreal pollen. Diatom floras from these deposits are indicative of coastal and brackish conditions. Forams and ostracods indicate an estuarine intertidal mudflat environment, with initially quite a high degree of marine influence, which tails off over time as the site appears to become more and more peripheral to marine access through regression of the sea and/or aggradation of the land. The few molluscs present indicate saltmarsh habitats. Facies 5. Palaeochannel fills. Plant macrofossils associated with these deposits are those of wet, maritime conditions but also included charred and non-charred ericaceous material (possibly reworked). Forams and ostracods are limited but are typical of estuarine, tidal mudflats. The few molluscs indicate saltmarsh habitats.

6

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Facies 6-8. Saltmarsh. Poor pollen preservation. Charred and waterlogged monocotyledonous plant remains dominate in the Baras-layer seen in facies 7. Facies 9. Palaeochannel fills (upper). Poor pollen preservation. Forams and ostracods are limited but are typical of estuarine, tidal mudflats with some degree of marine access. Facies 10. Occupation horizon. Poor pollen preservation. Facies 11. Terrestrial environments. Poor pollen preservation. Conclusion The results of this investigation have supplied information on the nature and timing of environmental changes associated with the earlier Holocene habitats represented at the site. However, poor preservation of contained palaeoenvironmental material in the palaeochannel fills (facies 5 and 9) as well as the saltmarsh facies (6-8) and the uppermost occupation horizon (10) has made it difficult to determine the nature of these environments. The results of the assessment undertaken indicate that further investigation of the sediments would be unlikely to add materially to the existing information already identified at Stinkums. The results add some finer detail to the palaeotopographic sequence identified at Cabot Park but do not lead to any revision.

7

Plot P9, Cabot Park, Avonmouth: post excavation analysis

2. Investigations of Late Bronze Age Sites Stinkums (Figure 2) 2.1

Introduction

In 1998 the Glamorgan Gwent Archaeological Trust undertook a program of archaeological works as part of the Cabot Park phase 2 site wide investigations. The full program of works included 45 evaluation trenches, 6 boreholes and 8 trial pits. As part of the works an evaluation trench (Trench 44) measuring 125m in length was excavated along the northwest-southeast axis of the proposed Plot P9, referred to as Stinkums on the Tithe map of 1772 (Locock et al 1998, 78). Trench 44 revealed three separate habitation surfaces separated by two palaeochannels. These, as on the other prehistoric sites found in the area, consisted of charcoal-rich horizons within a gleyed clay (B-layer), from which small quantities of burnt stone, animal bone and prehistoric pottery were recovered (ibid). These stratigraphic deposits were neither excavated nor fully uncovered due to their known fragility and their distinct similarity to other deposits at Cabot Park (ibid). The archaeology and geoarchaeology of the Stinkums site was further evaluated in 2007 by the Museum of London Archaeological Service (MoLAS). A program of four evaluation trenches and seven boreholes was undertaken (Askew 2007: Corcoran 2007). One of the MoLAS trenches was positioned in such a way to enable the reexamination of the 1998 GGAT trench (44). Only c.67m of the original 125m was reexcavated. MoLAS trenches 2 and 3 were excavated at right angles to and adjoined the trench 4 at each end and ran at right angles from it. Trench 1 was excavated parallel to trench 4 it but set some 100m to the northeast. A transect of nine Terrier Rig boreholes was proposed with the intention of extending to the base of the alluvial sequence. Ground conditions would not allow two of the boreholes to be undertaken. The remaining seven were drilled and core samples recovered with initial examination conducted on-site (Corcoran 2007). Wherever possible the boreholes were excavated to 7m depth and cores were retained for subsequent assessment and palaeoenvironmental assessment. The purpose of the drilling was to determine the extent of the horizons of archaeological interest within the area of proposed building construction (P9). The work was carried out by MoLAS in 2007. The borehole survey demonstrated that the gleyed occupation layer extended into the building footprint and can also be identified in the palaeochannel, whereas the Baras-layer extended across the entire area examined, apart from the palaeochannel.

2.2

Stratigraphic Description1

The earliest stratigraphy encountered at Stinkums was a deposit of fine brownish-grey clayey sand. This deposit was encountered in borehole 9 only and referred to as facies 1 with the upper surface of this deposit encountered at 0m OD. The palaeoenvironmental results indicate that this was a sand flat comprised of brackish saltmarsh and true marine components. Radiocarbon investigation of a foraminifera 1

GGAT context numbers are given in three digits, MoLAS context numbers are given in two digits.

8

Plot P9, Cabot Park, Avonmouth: post excavation analysis

sample recovered from this facies was dated to 6980 - 6560 cal BC (Table 1, Beta Ref. 253801) i.e. the middle Mesolithic period. These are the oldest dates measured by radiocarbon dating for the Cabot Park region. Overlying the sand flat deposit was an intertidal mudflat comprising sand and clay and recorded as facies 2 in borehole 9 only. The upper surface of this deposit was encountered at 1.10m OD. Forams and ostracods data indicate that the marine influence of the earlier facies 1 begins to diminish through this second facies. The sandflats turn to mudflats as the area becomes increasingly peripheral to marine influence as a local estuarine intertidal environment gradually becomes established. The mudflat of facies 2 was overlain in borehole 9 by an organic clay (facies 3). This facies was also encountered in boreholes 6-7. The uppermost levels at which this was encountered was 1.56m OD, 1.79m OD, 1.86m OD and 1.90m OD respectively. These levels represent a gradual slope downwards towards the southwest. These deposits comprise humic clays that contain the highest concentrations of pollen recovered during the borehole investigation. The taxa associated with this deposit suggest an association with agricultural activity although established agriculture on this estuarine environment is unlikely. Taxa describing a mixed deciduous woodland, with Oak, Pine, Alder, Birch, Lime and Hazel was also found during the palaeoenvironmental assessment. Both of these plant macrofossil results might be interpreted to indicate that the flora recorded persisted in the vicinity of the site probably on the higher ground to the northwest and the foothill regions rather than on the estuarine areas themselves. Radiocarbon investigation of the peat in this facies was dated to 3940 – 3660 cal BC (Table 1, Beta Ref. 253800). This dates the deposit to the early to middle Neolithic. The subsequent deposit is facies 4, encountered in boreholes 5-9. The uppermost levels at which this deposit was encountered was 3.26m OD, 3.76m OD, 3.89m OD, 3.36m OD and 3.0m OD respectively. This indicates a highpoint at borehole 7 and gentle slopes downward to both the east and to the southwest. This facies was a deposit of laminated silty clay representing a return to inter-tidal mudflat conditions. Overlying the second phase of inter-tidal mudflats was a sequence of inter-tidal salt marsh deposits (facies 6 and 8; MoLAS (015) GGAT (508)). This sequence included an organic clay (facies 7) deposit laid down between the two periods of salt marsh environments (facies 6 and 8). This organic clay was encountered during trial pitting, watching briefs, evaluation trenches and borehole investigations across the Cabot Park site at depths ranging between 4.15 - 5m OD. This deposit is referred to as the Baraslayer, or the N-layer and has been suggested to represent a freshwater lagoon between adjacent marshy areas (Locock 1998: 36) preceding a ripening of the soil horizon, with vegetation cover and prone to continual periodic inundations. The pollen record for the N-layer in borehole 9 is indicative of a salt-marsh environment with pine, oak, alder and hazel growing in the vicinity most probably toward the foothills to the northwest. This deposit was encountered in trenches 1-3 (MoLAS) at depths of 4.50m, 4.33m and 4.56m OD respectively. These trenches did not produce finds from this context and samples were not presented for C14 dating from these trenches. Previous radiocarbon dating of this deposit has produced results ranging between 3700 – 2200 cal BC i.e. the Neolithic period (Table 1). Across the Stinkums site this deposit is quite flat with only minimal undulation (within 0.14m).

9

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Subsequent to the salt marsh deposit is a return to a stabilised land surface along with evidence for human occupation. This organic occupation deposit is referred to as the B-layer. It is most probable that this occupation was both short lived and limited to a localised area. During the excavation of GGAT evaluation trench 44 (Locock 1998: 78) this deposit was encountered and extended for approximately 109m. The occupation deposit was separated into three distinct surfaces (503, 504 and 505) divided by two palaeochannels (506 and 509). These palaeochannels had widths of 15m and 3m respectively. It has been considered that palaeochannel 506 might be the same palaeochannel as 531 in trench 43 located at approximately 100m to the southwest of trench 44. The occupation layers did not reach the edges of the palaeochannels suggesting that they were active during the period of occupation. MoLAS trench 4 was a partial re-excavation of GGAT trench 44 (Askew 2007: 1114). MoLAS excavated 64.70m at the southern extent of the GGAT Trench 4 which measured 125m in total. The records from the MoLAS excavation partially records a single palaeochannel as being revealed during excavation. It was described as being filled and then overlain by the B-layer. If this description is correct it cannot be either of the palaeochannels encountered by GGAT in 1998. The B-layer in the GGAT results respected the edges of the palaeochannels, suggesting that the palaeochannel was still open during the occupation period of the Bronze Age. Within trench 4 the general occupation stratigraphy comprised three thin deposits (03, 04 and 05). Held within the uppermost occupation deposit (03) was a small hearth (02) measuring 0.80m by 0.60m and 10-20mm in depth. Charcoal was present in the hearth, surrounding a central patch of orange baked clay. The charcoal was observed at a height of 5.08-5.12m OD. Charcoal samples taken from the deposit have produced radiocarbon dates (Table 1) in the range of 1120 to 910 cal BC (i.e. the later Bronze Age). This not only gives us a date for the hearth feature, but also a terminus post quem for the occupation layer below it (04). Each of the occupation deposits contained similar groups of finds, each comprising late Bronze Age pottery (see Mepham 2.3.1 below), animal bone and small quantities of unworked stone. The very limited date range of the pottery (Mepham ibid.) and the shallow depths of the deposits in this phase suggest a short period of occupation following the trend seen at other sites at Cabot Park. The upper level of this occupation was encountered at 5.07m OD (approximately 1.20m below current ground level). Further evidence for occupation was encountered some 25m to the south of the hearth (02) in trench 4. An oval-shaped spread of blackened, fragmented stones (06) measuring 1.80m long by 0.85m wide and 50mm in depth was found with adjacent charcoal fragments at a height of 5.03m OD. The central area of the stones was densely packed, compared to its periphery, which was sparsely spread and each fragment ranged between 40-50mm in diameter. Specialist analysis of the stones has concluded that they are unburned and so cannot be interpreted as a hearth (see Horak 2.3.2. below). Charcoal from this deposit produced radiocarbon dates (Table 1) of 780 to 410 cal BC (i.e. late Bronze Age through to early Iron Age). A small waste flint flake (see Lewis 2.3.4 below) and eight fragments of animal bone (see Locock 2.3.3 below) were also found in close proximity to context 06. The B-layer was then sealed by a further series of estuarine alluvium deposits. Directly overlaying the occupation deposit in GGAT trench 44 was an upper gleyed deposit (501), followed by weakly gleyed silty alluvial clay (500) and this in turn is

10

Plot P9, Cabot Park, Avonmouth: post excavation analysis

sealed by a silty clay alluvium (499). All three of the alluvial deposits were recorded in borehole 6; in all other boreholes only the uppermost one or two of these were present. MoLAS trenches 1-4 recorded an alluvial silt deposit (021 in trench 1 and 019 in trench 2-4). Overlying the alluvial deposits is a subsoil, followed by the topsoil.

11

Plot P9, Cabot Park, Avonmouth: post excavation analysis

2.3

Finds Reports

2.3.1

Prehistoric Pottery -Lorraine Mepham (Figure 3)

Introduction The pottery assemblage recovered from P9 at Cabot Park amounts to 131 sherds (628g). Apart from one medieval sherd, all of this material is of later prehistoric date. The condition varies from fair to poor; apart from a few modern breaks, sherd edges are worn, and calcareous inclusions have leached out, at least from surfaces. Mean sherd weight overall is 4.8g. The pottery derived from nine contexts. The medieval sherd (a small, glazed body sherd in Bristol Redcliffe ware, from deposit/backfill layer 01) is not discussed further here. Description The pottery has been analysed following the standard Wessex Archaeology recording system for pottery (Morris 1994), which follows nationally recommended guidelines for prehistoric pottery (PCRG 1997). Details of fabric, form, surface treatment and decoration have been recorded; data are held in an Excel spreadsheet, which forms part of the project archive. Seven fabric types were identified, based on the range and frequency of macroscopic inclusions. These fall into three groups, on the basis of the dominant inclusion type: calcareous (LI), flint-tempered (FL) and sandy (QU). This small type series incorporates and extends that created for previously analysed pottery from Cabot Park (Mepham 1998), which comprised four calcareous fabrics (LI/1-4) and one grogtempered fabric (GR1). The assemblage from P9 includes three calcareous fabrics, as previously identified (LI/1, LI/2 and LI/4), as well as three sandy fabrics (QU1-3) and one flint-tempered fabric (FL1). Fabric descriptions are given below. In all cases, firing is irregular, and surface colouring is patchy. FL/1

Soft, fine matrix, slightly micaceous, containing sparse, very poorly sorted, subangular flint <5mm, frequently protruding through surfaces, giving irregular appearance. Rare, subrounded quartz <0.5mm; rare fine mica flecks.

LI/1

Soft, fine clay matrix, slightly micaceous and with a soapy feel; containing sparse to moderate (3-20%), poorly sorted, subangular calcite fragments <3mm; rare (1-3%) subrounded quartz <0.5mm; rare iron oxides and very rare mica flecks.

LI/2

Soft, moderately coarse clay matrix, slightly micaceous with a slightly soapy or slightly sandy feel; containing sparse to moderate, poorly sorted, subangular calcite <3mm; sparse (3-10%), fairly well sorted, subrounded quartz <0.5mm; rare iron oxides and very rare mica flecks.

LI/4

Soft, moderately fine clay matrix, slightly micaceous with a soapy feel; containing moderate, poorly sorted, irregular calcareous inclusions (probably calcite) <1.5mm; rare subrounded quartz <0.5mm; very rare fine mica flecks.

12

Plot P9, Cabot Park, Avonmouth: post excavation analysis

QU/1 Soft, moderately coarse, poorly wedged clay matrix, with a fairly smooth feel; containing sparse, fairly well sorted, subrounded quartz <0.25mm; rare voids (probably from calcareous inclusions); rare iron oxides. QU/2 Soft, fairly coarse matrix with a gritty feel; containing moderate, fairly well sorted, subrounded quartz <0.5mm; sparse small voids (probably from calcareous inclusions; rare fine mica flecks. QU/3 Soft, coarse matrix with a fairly smooth feel; containing sparse, fairly well sorted, subrounded quartz <0.35mm; sparse iron oxides; rare fine mica flecks. Petrological analysis of the calcareous fabrics from the previously analysed assemblage (D. Williams, archive report) suggested a local source for these fabrics, and there is nothing in any of the other fabrics to indicate anything other than local manufacture. The calcareous fabrics are sufficiently similar in hand specimen to suggest that they are in fact variants of a single fabric type, varying only in the size of the calcareous inclusions. There are very few diagnostic features amongst this small assemblage – 14 rim sherds (from a maximum of 11 vessels) and one decorated body sherd are present. Four rim profiles were identified, of which three (inturned or hooked rims (Fig. 3, 1-2); everted, thickened rims (Fig. 3, 3-4); and plain, upright rims (Fig. 3, 5)) conform to types previously identified. The fourth type is upright and flat-topped (Fig. 3, 6). In no instance, however, could the rims be attributed to overall vessel form. Two of the hooked rims, both in the same fabric (LI/4) and both from stone spread 07, possibly from the same vessel, have fingernail impressions on top of the rim, and one body sherd in QU/1, from deposit/backfill layer 01, carries fingertip impressions, and probably comes from a vessel shoulder. These are the only instances of decoration. Discussion This small assemblage confirms and extends the range of types identified from Cabot Park, but is unlikely to add any significant chronological information. Sandy wares have a fairly wide chronological range within the Severn area; they were found in all phases at Brean Down (Woodward 1990), for example, and have been identified in Middle/Late Bronze Age contexts at Chapeltump II (Locock et al. 2000, 23). Rim forms seen at Cabot Park, however (all everted rims on necked forms), would fit more comfortably with a Late Bronze Age date. Most of the sandy wares came from deposit/backfill layer 01, with a few sherds in hearth 02, deposit 03 and possible hearth 06. Their absence from other contexts (alluvial deposit 04, fill 05 of feature 28, stone spread 07 and deposit 17) may not be significant, given the small quantities of pottery involved.

2.3.2

Unworked Stone - Jana Horák (Figures 4-7, Table 2)

Introduction The samples submitted for examination comprise 41 pieces of rock from six contexts, divided into 11 lots. The samples are characterised by low lithological variation, with

13

Plot P9, Cabot Park, Avonmouth: post excavation analysis

white quartoze sandstone dominating, but also included samples of pink dolomite, and two impure carbonate rocks. The samples examined were examined following the principles outlined under British Standard EN 12407:2000. A hand lens, and where necessary illumination, was used to observe the following features: Assessment of the general colour of fresh and/or weathered surfaces of the rock. Where appropriate a colour values was attributed using the Munsell Rock Colour Chart.  Fabric (grain composition, grain shape, and sorting Figure 6 and 7)  Grain size (expressed in mm and converted to grain size classification scale, see Table 5).  Grain composition  Presence of features such as macrofossils, igneous textures/structures In addition to the above an evaluation of the natural or man-made form of the lithic samples was made.

Petrological Descriptions Context 01 - Stone(3 pieces of cracked stone, described in decreasing order of size) 1.1 Grey, quartzose sandstone. Medium-grained, grain-supported, moderately well sorted with sub-rounded to sub-angular grains of pale white frosted quartz, occasional darker grey quartz grains, and cream cement/interstitial infill. The stone is cracked and shows some red colouration from iron oxide staining. Some of the surfaces show more substantial films of iron oxide. In thin section the hand specimen observations are supported; the grain composition is seen to be dominated by moncrystalline quartz, with minor polycrystalline (some mylonitic) quartz, minor quartz-rich lithics, altered feldspar, iron oxide cubes (0.3 mm (Figure 4, A-B). The stone shows no evidence of cracking or features that might be attributed to heating. The form of the stone and the presence and distribution of iron oxide are considered natural in origin. 1.2 Grey, quartzose sandstone. Medium-grained, grain-supported, moderately well sorted texture, with sub-rounded to sub-angular grains of pale white frosted quartz, occasional darker grey quartz grains, and cream cement/interstitial infill. The ironoxide areas are on the stone are post-excavation, having been derived from the larger sample in the same sample bag. 1.3 Grey, quartzose sandstone. Coarse-grained, grain-supported, moderately well sorted texture, with sub-rounded to sub-angular grains of pale white frosted quartz and occasional darker grey quartz grains. There is a higher content of cream

14

Plot P9, Cabot Park, Avonmouth: post excavation analysis

cement/interstitial infill than the other two samples in this group. Minor red staining by iron oxide is present on one face. Theses three samples are all of the same lithology and show a variable presence of natural iron oxide. The pale colour of the rock precludes attribution of a Munsell colour code. Context 01 – Stone 1.4 Pink, pebbly sandstone [7.5YR7/3 pink]. Natural, angular fragment, composed of smaller pink (iron stained) quartz grains and larger rounded white vein quartz pebbles up to 2mm in diameter. The colour variation between different faces of the fragment reflects different degrees of weathering. The form and characteristics of the stone are natural in origin. Context 04 – Stone 4.1 Grey quartzose sandstone. Natural angular fragment of grey, coarse-grained, moderately sorted, quartzose sandstone with minor, but variable, iron staining. Iron staining is also present on one surface, coating slickensides. The form of the fragment and the iron oxide are natural in origin. Context 05 – Stone 5.1 Pale grey, quartzose sandstone. Angular fragment of medium-grained sandstone with heterogeneous pink staining. The form and texture of the sample are natural in origin. Context 06 – Lithic Sample 6.1 Grey quartzose sandstone. A moderately sorted, grain-supported, quartzose sandstone. Part of the outer surface of the fragment is darker. In thin section the grains are seen to be composed dominantly of moncrystalline quartz, with minor white mica flakes, polycrystalline quartz, altered lithic fragments and small grains of detrital tourmaline and epidote (Figure 4, C-E). The form of the stone and the black colouration are considered natural in origin. 6.2 Grey quartzose sandstone. Fractured pebble of medium-grained, quartzose sandstone, showing grain-supported texture and moderate sorting. The outer surfaces of the stone have a black coating of variable thickness. Thin section examination of this sample confirms the above observations. The grains are dominated by moncrystalline quartz and the stone is variably infiltrated by iron oxide (probably goethite, α-Fe3+O(OH), (Figure 5, A & B). There appears to be no difference between this iron and the outer black surface. The form and black coating to the stone are considered natural in origin. 6.3 Grey quartzose sandstone. Angular fragment of medium-grained, moderately sorted, grained-supported quartzose sandstone. Two surfaces of the stone are extensively coated by a film of iron oxide. The form of the fragments and the distribution and presence of the iron oxide are natural in origin. 6.4 Grey quartzose sandstone. Angular fragment of medium-grained, grainsupported, well sorted quartzose sandstone. Much of the outer surface of the fragment is coated by a thin film of iron oxide, and fractures cutting the rock can also be seen to

15

Plot P9, Cabot Park, Avonmouth: post excavation analysis

be iron coated. The form of the fragments and the distribution of iron oxide are natural in origin. 6.5 Grey quartzose sandstone. Angular fragment of medium-grained, grainsupported, well sorted, quartzose sandstone. The two surfaces of the stone show variable iron oxide staining, and a zone of blackening. The form of the stone and the red and black iron oxide colouration are considered natural in origin. 6.6 Grey quartzose sandstone. Angular fragment of variably grained (fine-coarse), grain-supported, quartzose sandstone with an interstitial infilling of cream illite/white mica. The form and colouration of this stone are natural in origin. Context 06 - Stone 6.7 Grey quartzose sandstone. Naturally rounded, oblate cobble of coarse-grained, well-sorted, grain-supported, quartzose sandstone. The surface is variably mottled red by iron oxide. One end of the cobble shows a fractured and weathered surface. Although the overall form of the cobble is natural, one face of has a flattened surface making it slightly concave. There is no indication of polishing of this surface but it may have been produced by use as a whetstone. The lithology is no different to other samples in the suite, so has not been especially selected, other than for its convenient shape and size, which fits nicely within a hand. Context 06 – Lithic Sample 6.8 Grey quartzose sandstone. Angular fragment of grey-weathered mediumgrained, quartzose sandstone. The sandstone is very well sorted and shows signs of pressure solution and the infilling of interstitial spaces with illite/white mica. The form of the stone is natural. 6.9 Grey quartzose sandstone. Angular fragment of medium-grained, grainsupported, well sorted, quartzose sandstone. The sample is cut by several fractures. The form of the sample is considered natural in origin Context 06 – Stone 6.10 Quartzose sandstone. Rounded fragment of coarse-grained, grain-supported, poorly sorted quartzose sandstone with cream matrix/infill. Iron staining is heterogeneously distributed over the surface of the sample. X-ray diffraction has confirmed that the infill is of illite/white mica composition. 6.11 Quartzose sandstone. Rounded fragment of coarse-grained, grain-supported, poorly sorted quartzose sandstone with cream matrix/infill. This lithology is identical to that described above. In thin section the grains are dominated by moncrystallaline quartz and minor polycrystalline quartz, and altered feldspar (Figure 5, C & D). The outer rim of the stone has a thin (0.08mm) coating of iron oxide, and the iron has also percolated through the stone to fill interstitial spaces. The texture and mineralogy of the stone are natural in origin. 6.12 Small sample of altered limestone. The limestone is fossiliferous containing fragments of crinoid stems, but also contains scattered quartz grains. The samples shows local iron staining by goethite (α-Fe3+O(OH), but this colouration is not attributed to artificial heating of the sample, but results from natural processes.

16

Plot P9, Cabot Park, Avonmouth: post excavation analysis

6.13 Fragmental rocks. This sample comprises rounded clasts of the white quartzose sandstone, cemented by a large volume calcium carbonate precipitate. The composition of the precipitate was confirmed by X-ray diffraction and this lithology probably formed by the cementation of fragments, perhaps in a scree. Context 07 – Cracked stone 7.1 Grey quartzose sandstone. Weathered, sub-angular fragment of coarsegrained, grain-supported, well sorted, grey quartzose sandstone. Grains are composed of frosted quartz and less abundant grey quartz grains. A small area (10 mm across) of iron staining is present on one surface. The form of the stone and the area of iron staining have formed by natural processes. 7.2 Quartzose sandstone. Angular fragment of weathered, medium-grained, grainsupported, quartzose sandstone. The sample is highly sheared and the shearing is associated with iron oxide staining. This close association indicates that the form and distribution of iron oxide in this samples are natural in origin. 7.3 Pink quartzose sandstone. Weathered, irregular, rounded fragment of coarsegrained moderately sorted, grain-supported, quartzose sandstone. This shows variable light red staining by iron-oxide which imparts a pale pink colour. Part of the fragment is also has a black coating. X-ray diffraction shows negligible iron or manganese mineral content and therefore this coating may be of organic origin. 7.4 Grey quartzose-sandstone. Weathered, angular fragment of fine-grained, grain-supported, well-sorted, quartzose sandstone. Two surfaces of the fragment are coated by orange iron oxide, and this is also seems to continue along fractures in the rock. The form of the fragment and the distribution of iron oxide are natural. 7.5 Quartzose sandstone, veined. Small angular fragment of weathered, coarsegrained, quartzose sandstone. One surface of the rock is coated by iron oxide and secondary quartz. The form of the fragment and the iron and quartz distribution are of natural origin. Context 07 – Probable Stone Artefact 7.6 Oblate pebble of grey [GLEY1/5/N, grey], fine –grained, sandstone. Slight mottled on the surface of the pebble results from the presence of iron pyrites. The samples shows evidence of bedding and there is a slight groove cross one surface. The form of this stone, including the groove, are considered natural in origin. The groove may follow a less resilient layering in the bedding. Context 17 – Stone 17.1 Grey quartzose sandstone. Angular fragment of medium-grained, well-sorted. One surface is partially coated iron oxide. The shape of the block and the presence of the iron oxide are the result of natural processes. 17.2 Orange/pink dolomite. Pink impure dolomite, containing grains of quartz (Figure 5, E). The fragment has a partial black coating of MnO. There is a variation in colour from the core of the rock [10Y 5/3 weak red] to the outer margin, on one side of the stone [10R6/ 3 pale red]. This is considered the result of natural weathering processes. The form of the stone and the variation in colour is natural in origin.

17

Plot P9, Cabot Park, Avonmouth: post excavation analysis

17.3 Orange/pink dolomite. [5YR 6/4, light reddish brown] Angular fragments, similar to 2.11.2, but with fewer large white quartz grains. The form and colour of this fragments are natural in origin. 17.4 Pebble. Naturally rounded pebble cracked and stained by weathering. and burial within superficial deposits. The lithology is white when unweathered, and a coarse-grained quartzite. The form and darkening of the outer layer of the pebble are the product of natural processes.

Provenance of samples With the exception of four samples (contained within 6.12, 6.13 above), all rocks in the suite from Plot 9 Cabot Park (BT-CBK06) submitted for examination, are sandstones. Most of these are weathered, being originally cream coloured quartzose sandstones that have weathered a grey colour. Local Geology The Cabot Park site is located within an area of tidal flat deposits (BGS, 2004), with bedrock geology exposed a few kilometres to the south west, south and west. The strata are dominated by a variety of sandstones, limestone, dolomite, and mudstone with a minor component of volcanic rocks present within the Pembroke Limestone Group (Carboniferous Limestone). The quartzose sandstone identified within this report resembles lithologies found within the Carboniferous sequence, in particular the quartz-rich sandstones of the Quartzitic Sandstone Group (Kellaway & Welch, 1993). These are hard, grey quartzrich sandstone dominated by vein quartz (mono-crystalline quartz), with outcrops occur on the western side of Bristol from Kingsdown to Long Ashton. Grey sandstone is also present within the Portishead Beds (Upper Old Red Sandstone), in particular the Black Nore Sandstone exposed at Portishead. Insufficient information is available on the texture of these sandstones to assign a precise provenance to the samples studies, however it is highly likely that they are derived locally, from within a few kilometers of the site. Similarly the dolomite samples are most likely derived from the Dolomitic Conglomerate of the Mercia Mudstone Group. This unit is exposed in the upland area immediately to the east of the site at Cabot Park. Iron staining is present in many samples, typically as an orange/brown or black coating (goethite) of hydrous iron oxide on surfaces and filling fractures in the rock fragments. Iron ores (in the form of iron oxide, hematite, hydrous oxide goethite and ochre deposits) are widely distributed throughout the Bristol area, providing a source of iron that has been mobilised and reprecipitated by the movement of fluids. The iron oxide occurrences described from the Cabot Park samples are all considered to have resulted from these natural processes. Discussion of the use of stone samples from Cabot Park Of the samples submitted for examination, with the exception of one (6.7 above) all are considered to have a form that has resulted from the fracturing of larger pieces of rock. It is assumed that this fracturing, generally producing angular fragments has occurred by natural processes. Many of the fragments show evidence of weathered surfaces.

18

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Examination was also made of the samples for evidence of fire damage. The features observed in rocks which have been heated vary with the maximum heat of the fire, the duration of heating and the composition of the rock. The typical petrological features observed after heating of rocks are:Colour change, in siliceous igneous rocks this is often a lightening in colour, clay rocks vary in colour depending on the actual temperature reached. Crack initiation – this may in the form of micro- cracking, accurate cracks, or thermal induced stress cracks be parallel to fire damage surface. In limestone cracks may occur parallel to layers of impurity, particularly those of clay. In limestone where quartz is also present the mineral portlandite (Ca (OH) 2.) may form under conditions of atmospheric humidity and may result in the decay of the outer layer of the rock. Dehydration/dehydrozylation. In sandstone samples containing hydrous minerals such as illite/muscovite, dehydrozylation occurs around 400-500°C depending on the precise composition of the mineral present. Dehyroxilation of micas and clay mineral typically released some iron, resulting in the red colouration of the rock, and the formation of hematite, (Hajpál & Török, 2004). In hand specimen investigation, no cracking suggestive of fire damage was observed. In thin section examination of selected samples, no cracking from thermal expansion of the quartz grains was observed. Although this is most noticeable in specimens that have been heated to in excess of 573ºC, thermal expansion of approximately 1% would be expected in quartz grains of 0.2mm when heated to 210ºC, a temperature comfortably attained in a small campfire (Audley, 1921). Furthermore, studies on the effects of fire of building stone (Hajpal & Török, 1998) have found cracks at all grain boundaries in sandstones heated to 450ºC. Examination of the clay/mica content of the sandstones in two samples (6.1 and 6.2 above) indicated that the white infilling was of unmodified illite/white mica, providing an additional line of evidence for the lack of heating of these samples. As mentioned previously, the presence of iron oxide coating on many of the samples is considered the result of geological processes and it not attributed to the effects of man produced heating of the rocks. As noted in section 2, an oval cobble (6.7 above), has one concave surface, which is not considered to have formed by natural processes. Although there is no obvious polishing of this concave surface, the shape suggests that it may have been worn into its current form by use as a whetstone.

2.3.3

Animal bone - Martin Locock (Tables 3-5)

Introduction During archaeological investigations of P9, a small assemblage of animal bone was recovered from nine prehistoric contexts within the sequence of alluvial silts.

19

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Normal The animal bone was inspected and identified by comparison with known specimens and atlases (Schmid 1972; Hillson 1992; Cohen and Serjeantson 1986; Amorisi 1989). Where species could not be established, elements were assigned to size classes (large mammal, medium mammal), based on Shackley (1981). No distinction was made between sheep and goat and rabbit and hare. Completeness was recorded using a code in which the parts of the bone were numbered from 1-5, from proximal to distal. Bone counts include all fragments larger than 10mm; joining parts within a single context were counted as one. Ribs and vertebrae were assigned to size class only. Metrical data was recorded using Driesch (1976) and compared to published data (University of Southampton 2003). Age classes follow Noddle (1977, 381) (N neonatal, A juvenile, B immature, C adult); tooth wear was recorded by Grant’s wear stages as republished in Amorisi (1989). The bone count corresponds to NISP. Bone condition was assessed visually based on the definitions in Table 3: Summary The analysed assemblage comprises 48 bones, of which 16.7% were identified to species level (see Table 4); all is from presumed domestic sources. Condition and retrieval The bone varies in condition from II to III (Table 3). The surface survived in fair condition, with eight retaining evidence of pre-depositional gnawing; one bone's surface had corroded, perhaps the results of passing through a dog or pig's digestive system. There is no indication of recovery bias. Butchery and craft evidence Most of the bone was fragmentary and some retained evidence of butchery. No craft evidence was noted. Species present Three cow bones are immature (Age class B/C), probably 30 months of age (the two teeth may have come from the mandible, this representing a single individual). The sheep/goat bones appear to be mature. Discussion Previous prehistoric sites at Cabot Park have produced bone assemblages dominated by cattle and sheep/goat, with a significant proportion of immature animals, and this assemblage falls well within that pattern, perhaps reflecting the slaughter and consumption of animals while summer grazing on the salt marsh.

2.3.4

Lithic Analysis - Richard Lewis (Table 6)

Introduction A small assemblage of lithic material recovered during an evaluation of Plot 9, Cabot Park, Avonmouth has been presented for analysis. The evaluation produced lithic material from three contexts within a single trench (Table 6). The analysis of this small assemblage forms the subject of the present report.

20

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Methodology The assemblage is composed of flint and was examined under x10 magnification and recorded using a typological recording system (Andrefsky 2000). Brief details including raw materials, condition and pertinent technological information was also recorded. No further technical analysis was undertaken. Raw material and condition There is wide variation as to what is commonly classified as flint or chert (Andrefsky 2000, 52-58; Whittaker 2007, 70). However, the most common definition used by archaeologists limits flint to material formed in bedded chalk deposits while chert can be found forming in limestone environments. Both materials belong to the cryptocrystalline (or microcrystalline) group of silicates. All of the material examined from the present assemblage can be classified as medium quality grey flint. Two pieces of flint were recovered from context 004; one was of medium quality grey flint and the other was glossy and of a higher quality but essentially the same material. The third piece (06) has a significant proportion of cortex, which is quite pitted indicating exposure to the elements. The quality of the material and the evidence of exposure may suggest that these pieces originated from flint pebbles recovered from streams or glacial moraine deposits. Description of material Trench 4, context 04, produced a flint scraper and a single piece of flint debitage. Both pieces are of grey flint with light grey and brown mottling. The scraper is long (43mm) and narrow, and has a slight curve with a contracted waist. It has one blade edge with retouch on the dorsal surface and limited retouch on the ventral surface; the opposing edge is flat. The scraper is abraded with smooth rounded edges. The debitage fragment from this context (04) has a smooth ventral surface containing several isolated dark spots, most likely mineral impurities in the flint. Radiating ripples marginally indistinct but present, however, the bulb of percussion is missing. The dorsal surface has two small crushed platforms at proximal end relating to previous flake removal and two localised areas of radiating ripples. Trench 4, context 06, produced a single piece of flint debitage dark grey in colour with light grey mottling. Cortex was found to cover around 50% of the dorsal surface. A crushed platform is present at the proximal end, the result of previous flake removal, along with two small step fractures. The ventral surface has a clear bulb of percussion and eraillure facet below a small platform at the proximal end, with radiating ripples marginally indistinct but also present. Discussion The assemblage is too small to produce heuristic typological data sets other than the most basic attribute and dissection strategies. Of the two debitage pieces one (04) is glossy with well defined edges and no patina on its flaked surfaces indicating deposition fairly rapidly once struck from the parent core. The second piece of debitage (06) shows some minor abrading but the cortex exhibits pitting rather than abrading/rolling indicating a short period of exposure before deposition once struck from the parent core. The scraper (04) is well abraded and is likely to have been exposed to the elements for a significant period, possibly as a result of attrition in a streambed.

21

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Stratigraphically, the flint pieces from context 04 were recovered from an occupation deposit. This might account for the abraded nature of the scraper. Overlying 04 was an additional occupation deposit (03) which contained a hearth (02). Charcoal from this hearth deposit (02) has been radiocarbon dated to 2840+/-40 BP indicating late Bronze Age occupation sealing the context from which the scraper and debitage piece were recovered. This date is further corroborated by the presence of late Bronze Age ceramics found within contexts 02 and 03 and is indicative of the B-Layer horizon. The abraded scraper and debitage fragment must therefore be earlier in date. Bronze Age occupation activity along palaeochannels is attested at Stinkums from an evaluation by GGAT 1998 (Locock et al 1998) and at similar sites on the Welsh side of the Bristol Channel at Rumney Great Wharf on the Wentlooge Level (Allen 1996). Context 06 contained a single piece of flint debitage and has been interpreted as an occupation layer containing a spread of fragmented stones. Charcoal from this deposit has been radiocarbon dated to 2490+/-40 BP indicating very late Bronze Age to early Iron Age occupation. The presence of a single piece of flint debitage may be residual from earlier phase of occupation activity. The discussion of lithic assemblages at Cabot Park is under represented and in the main restricted to rather short and limited summaries on artefact totals (see Locock et al 1998 and Locock 2001). However, Waban-Smith (in Yates et al, unpublished) discusses the lithic assemblage from Kites Corner, Cabot Park, in particular detail. It is interesting to note that of the 86 flint pieces identified 84 were of medium quality grey flint from pebbles possibly sourced from glacial moraine deposits. This may indicate a common source for the current Plot 9 assemblage and those recovered from the adjacent Kites Corner site. Stratigraphical sequences also appear synonymous; both assemblages have been recovered from the B-Layer, which is known to belong to the late Bronze Age/early Iron Age. The radiocarbon dates recovered from Plot 9 certainly appear to support this date range.

2.3.5

Molluscs - Andy Sherman

Introduction A small sample of terrestrial mollusca (small find number 2007/31) was recovered from context 04 for analysis. This sample consisted of a single near intact snail shell and several fragments of snail shell. Analysis The sample consisted of a single near intact Cepaea hortensis (White-lipped banded snail) shell with a diameter of 13mm. Cepaea hortensis has a glossy, smooth shell, which is typically yellow in colour but may be pink, brown or red, with up to five variable spiral dark bands and an obvious white lip around the aperture (Janus 1982). This particular specimen had a pinkish-white shell with five purplish-brown spirals, (it should be noted that the original colouring of the specimen would probably be somewhat darker). A total of seven fragments of similarly coloured shell (weighing 0.09g) from the same species were also recovered from context 04.

22

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Discussion This species is common throughout Western Europe, and in England can be found in a range of habitats, including waste ground, woodland, hedgerows and grassland, and is often found in dense vegetation. Cepaea hortensis is active during the day in damp, mild conditions; its preferred food plants include nettles (the Urticaceae family), ragwort (Senecio jacobaea) and hogweed (Hercaleum mantegazzianum) (National Biodiversity Network Species Dictionary 2008).

23

Plot P9, Cabot Park, Avonmouth: post excavation analysis

3. Discussion The archaeological information provided by the 2007 evaluation at P9 has added little to the existing knowledge of the Stinkums site. Stinkums comprises distinct organic markers within gleyed horizons, which represent periods of stability within the more unsettled intertidal environment. The site at Stinkums has not been fully investigated and its extents are not known. Future examination of the site will be fundamental to further understanding the occupation patterns. The nature of the site at Stinkums has direct correlation with the neighbouring site at Little Googs by means of Bronze Age occupation horizons clearly divided by palaeochannels. This was clearly seen during the 1998 evaluation at Stinkums with two palaeochannels dividing three occupation areas. This stratigraphy is in contrast to the larger, more diverse site at Kites Corner where the Bronze Age occupation overlies an already filled palaeochannel. Finds encountered at Stinkums fit into the broad range of finds types at other sites at the Cabot Park. The prehistoric pottery types recovered add to the known range outlined from investigations at Little Googs and Kites Corner. An additional four pottery fabric types were identified. One was a flint tempered ware and three were differing quartz tempered wares (Mepham, above). The animal bone assemblage from Stinkums falls well within the general patterns encountered at Cabot Park dominated by cattle and sheep/goat with a significant proportion of immature animals. This pattern has been suggested to reflect the slaughter and consumption of animals during summer grazing on the salt marsh (Locock, above). The assemblage of unworked stone had low lithological variation with white quartoze sandstone dominating, but also included samples of pink dolomite, and few impure carbonate rocks. This introduces no further rock types to those already recorded at Cabot Park (Horak above). The similarity in rock types is sufficient to suggest that they are derived from the same source pointing to deliberate collection and transport to the site. The most likely origin is a homogenous source such as a nearby scree slope. The closest comparison for these rocks are in the Portishead beds of the Upper Old Red Sandstone. These are well exposed at Woodhill Bay, northwest of Portishead approximately 5km from the site. During the Bronze Age these would have been clearly visible and easily accessible from the esturine plain. The artefactual assemblages at Stinkums reflect general domestic activities. The importing of stone to the area, as at other sites in Cabot Park might have served to stabilise the land surface, although the quantities recorded to date do not fully support this. Ceramic petrology analysis undertaken using Kites Corner samples rules out the possibility that the stone was used as temper for pottery manufacture on site (Yates 2002: 37-39). The potential for the stones having been used as the base of fires/hearths has also been ruled out due to the lack of thermal expansion in the rocks quartz grains (Horak above). The assemblage also points to periods of occupation that were seasonal and extended over several generations with local ecotone dwellers migrating coastward during dry seasons to exploit the resources of the salt marshes bringing livestock to graze on the exposed vegetation.

24

Plot P9, Cabot Park, Avonmouth: post excavation analysis

These environmental and social characteristics are well known and have been previously established by the archaeological investigations at Kites Corner. The evidence from Stinkums corroborates that from Kites Corner but adds no significant further information.

25

Plot P9, Cabot Park, Avonmouth: post excavation analysis Based on the 2007 Ordnance Survey 1:5000 Landplan with the permission of The Controller of Her Majesty’s Stationery Office, © Crown Copyright, Glamorgan-Gwent Archaeological Trust Ltd, Licence number Al10005976

GRID N

Area shown in Figure 2

Area shown in enlargement

0

0

1.00

250

500metres

Based on the 1999 Ordnance Survey 1:25000 Explorer 166 map with the permission of The Controller of Her Majesty’s Stationery Office, © Crown Copyright, Glamorgan-Gwent Archaeological Trust Ltd, Licence number Al10005976

2.00km Figure 1. Site location (not to be included in final publication) 26

Plot P9, Cabot Park, Avonmouth: post excavation analysis

E353980 N179930

Trench 3 Trench 1 BH9

BH8 BH6 508

BH5 BH4 BH3 BH2

BH1

505 509 510

Trench 4

Trench 2

502

E353840 N179850 504

507 506

Trench 44

Key MOLAS Trenches

Boreholes

GGAT Trench

Boreholes not drilled

GRID

503

N

Palaeochannel

0

Figure 2. Trench and borehole locations at Stinkums - P9. Palaeochannels also shown (with GGAT context locations)

27

25

50metres

Plot P9, Cabot Park, Avonmouth: post excavation analysis

1

2

4 3

5

0

6 5.00

Figure 3. Illustrations of prehistoric pottery 1-6 (scale 1:1)

28

10.00cm

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Figure 4. Photomicrographs of stone samples 1.

29

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Figure 5. Photomicrographs of stone samples 2.

30

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Figure 6. Grain sorting classification

Figure 7. Grain shape classification

31

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Table 1. Radiocarbon dates Sample Reference

Site Data

Measured Radiocarbon Age

Beta 129454 Beta 118379

Kites Corner 5.10m OD Rockingham Farm Upper B-Layer. 5.20m OD

2610 +/- 70 BP

890 to 530 Cal BC

2830 +/- 70 BP

1158 to 812 Cal BC

Locock 1997, 88

Beta 118378

Rockingham Farm Lower B-Layer 5.10m OD

3060 +/- 50 BP

1398 to 1132 Cal BC

Locock 1997, 88

Wk 5804

Baras-layer 4.7m OD P9 Trench 4 Context 02 5.10m OD

3930 +/- 50

2580 to 2290 Cal BC

BARAS 1998, 11

2840 +/- 40 BP

-25.2 oo/o

2840 +/- 40 BP

1120 to 910 Cal BC (Cal BP 3070 to 2860)

Askew 2007,

Beta 231526

P9 Trench 4 Context 06 5.03m OD

2530 +/- 40 BP

-27.5 oo/o

2490 +/- 40 BP

780 to 410 Cal BC (Cal BP 2740 to 2360)

Askew 2007

Beta 253800

P9 BH9 – Facies 3 1.69m OD

4890 +/-40 BP

-18.9 o/oo

4990 +/- 40 BP

3940 to 3860 Cal BC (Cal BP 5890 to 5810), 3810 to 3660 Cal BC (Cal BP 5760 to 5620)

Current report

Beta 253801

P9 BH9 – Facies 1 -0.21m OD

7900 +/-60 BP

-8.9 o/oo

8160 +/- 60 BP

6980 to 6560 Cal BC (Cal BP 8920 to 8500)

Current report

Beta 231525

13C/12C Ratio

Conventional Radiocarbon Age

32

Sigma – 2 Calibrated Result

Reference

Plot P9, Cabot Park, Avonmouth: post excavation analysis Table 2. Grain size classification scale

Table 3. Bone condition (after Locock et al. 1992, 209)

Group Description I Fresh, greasy surface II Surface dulled, no longer greasy III Some surface deterioration, pitting or powdering IV Severe surface deterioration; whole thickness of bone affected; cracking and splitting V Disintegration, losing cohesion when handled

Table 4. Summary of analysed bone

Species Cow Sheep/goat Large mammal Medium mammal Total

NISP

% of identified to species 75.0 25.0

6 2 12 28 48

100.0

33

% all bone 12.5 4.2 25.0 58.3 100.0

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Table 5. Animal bone catalogue. context

species

catalogue

count

1

Medium mammal

Unid x 3, gnawed Unid x 5

8

1

Cow

Tooth

1

1

Large mammal

Unid, gnawed by rodents

1

4

Medium mammal

Unid x2

2

4

Large mammal

Tooth

1

5

Medium mammal

Unid

1

6

Medium mammal

7

6

Sheep/goat

Unid, gnawed by dog Unid, chopped Unid, x5 Scapula, 2 Scapula, 1

7

Cow

Tooth, deciduous, Age B, x2 Astragalus, surface corroded

3

7

Medium mammal

Unid, x9

9

11

Large mammal

Unid x3

3

11

Cow

Mandible, M3 a, M2 c, M1 d (Age B/C = 30 months), gnawed by dog

1

12

Medium mammal

Skull

1

12

Large mammal

Vertebra Unid x4

5

13

Large mammal

Unid, gnawed by dog

1

14

Cow

Calcaneum

1

16

Large mammal

Unid, possible gnawing

1

34

2

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Table 6. Summary of flint assemblage Trench

Context

Worked stone

Flakes

Retouched forms

Chips

Period

Summary

Total

1

Cores and core fragments -

Trench 4

004

-

-

-

Prehistoric

1

-

-

-

1

Prehistoric

-

-

-

-

1

Prehistoric

Unpatinated flint scraper, grey with light and pale brown mottling. Scraper has one blade edge with retouch on the dorsal surface and limited retouch on the ventral surface. Opposing edge is flat. Proximal end is pinched to a point with a small bulb of percussion; eraillure facet present. Distal end is slightly tapered but also flat. Scraper length has a slight curve, contracted at the waist. Weight = 4.1g, Length = 43mm, Width = 6mm (proximal end), 17mm (distal end) and overall thickness 5mm (flat edge). Unpatinated flint debitage fragment, grey in colour with a small light grey segment. Ventral surface is smooth with several isolated dark spots visible. Radiating ripples marginally indistinct but present. Dorsal surface has two small crushed platforms at proximal end relating to previous flake removal and two localised areas of radiating ripples. Small area of radial fissures also present. Weight = 2.3g, Length 20mm, Width = 25mm and overall thickness = 5mm (proximal end), 3mm (distal end). Flint debitage fragment, glossy dark grey in colour with light grey mottling. Cortex covers c50% of dorsal surface. Crushed platform from previous flake removal present at proximal end and two small step fractures. Ventral surface has a clear bulb of percussion at proximal end with radiating ripples marginally indistinct but present. Eraillure facet present below small platform. Weight = 5.3g, Length = 20mm, Width = 34mm, and overall thickness = 5mm (proximal end), 10mm (distal end).

Trench 4

004

-

Trench 4

006

Group Totals

0

0

1

0

2

Assemblage Totals

35

1

1

3

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Bibliography Amorisi, T 1989 A postcranial guide to domestic neo-natal and juvenile mammals: the identification and aging of the Old World species BAR International Series 533 Allen, JRL 1996 Three Final Bronze Age occupations at Rumney Great Wharf on the Wentlooge Level, Gwent. Studia Celtica 30, 1-16 Allen, MJ & Scaife, RG 2001 The physical evolution of the North Avon Levels; a review and summary of the archaeological implications Unpubl. Ms Wessex Archaeology Andrefsky, W 2000 Lithics – Macroscopic Approaches to Analysis. Cambridge Manuals in Archaeology, Cambridge University Press Askew, P 2007 Plot 9, Cabot Park, Avonmouth, City of Bristol: An archaeological evaluation report. MoLAS unpublished report. Audley, JA, 1921 Silica and the Silicates in: Rideal, S (Ed) Industrial Chemistry Bailliere, Tindall and Cox, p34. Baker, J and Brothwell, D 1980, Animal diseases in archaeology Academic Press, London Bell, MG 2007 Mesolithic activity at about the time of the Lower Submerged Forest in Bell, MG (Ed) Prehistoric Coastal Communities: The Mesolithic in western Britain York: CBA Research Report 149 36-45 British Geological Survey 2004 1:50 000 Bristol, England and Wales Sheet 264 Solid and Drift Geology, 1:50 000 (Keyorth, Nottingham, British Geological Survey) BS EN 12407:2000, Natural stone test methods. Petrographic examination. ISBN 0 580 36350 Corcoran, J 2007 ‘Plot 9, Cabot Park, Avonmouth, City of Bristol: A report on the geoarchaeological borehole survey’ MoLAS unpub. report Dark, P 2007 Plant communities and human activity in the Lower Submerged Forest and on Mesolithic occupation sites in Bell, MG (Ed) Prehistoric Coastal Communities: The Mesolithic in western Britain York: CBA Research Report 149 169-182 Driesch, A von den 1976 A guide to the measurement of animal bones from archaeological sites Harvard University: Peabody Museum Bulletin 1, Harvard Hajpal, M & Török A 1998 Petrophysical and Mineralogical studies of Burnt Sandstones in Proceedings of the 2nd International PhD Symposium in Civil Enginerering, Budapest, Hungary. (Online). Hajpal, M & Török A 2004 Mineralogical and colour changes of quartz sandstones by heat Environmental Geology, 46 31 1-322. Hillson, S 1992 Mammal bones and teeth: an introductory guide to methods of identification Institute of Archaeology, University College London, London Janus, H 1982 The illustrated guide to molluscs Harold Starke Ltd, London Kellaway, GA & Welch, FBA 1993 Geology of the Bristol district. Memoir of the British Geological Survey 199pp Locock, M, Currie, C K and Gray, S 1992 ‘Chemical changes in buried animal bone: data from postmedieval assemblage’ International Journal of Osteoarchaeology 2, 297-304. Locock, M, Robinson, S and Yates, A., 1998, ‘Late Bronze Age sites at Cabot Park, Avonmouth’, Archaeology in the Severn Estuary 9, 31-6

36

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Locock, M, Trett, R, and Lawler, M 2000 ‘Further late prehistoric features on the foreshore at Chapeltump, Magor, Monmouthshire: Chapeltump II and the Upton trackway’ Studia Celtica 35 17-48 Locock, M 2001 A Later Bronze Age Landscape on the Avon Levels in Brück, J, Bronze Age Landscapes, Tradition and Transformation Marvell, AG 2008 Cabot Park, Avonmouth P9: Post-excavation Project Design GGAT unpublished report. Mepham, L 1998 ‘Prehistoric potter y’ in Locock et al. 1998, 34 Morris, EL 1994 The Analysis of Pottery, Salisbury: Wessex Archaeology Guideline 4 National Biodiversity Network Species Dictionary (accessed December 2008): http://www.nhm.ac.uk/nbn Nayling, N and Caseldine, A E 1997 ‘Excavations at Caldicott, Gwent. Bronze Age Palaeochannels in the Lower Nedern Valley’. CBA Research Report 108, 348pp. Noddle, BA 1977 The animal bone in H Clarke and A Carter Excavations in Kings Lynn, 1963-1970 (Society for Medieval Archaeology Monograph 7), 376-389. PCRG 1997, The Study of Later Prehistoric Pottery: general policies and guidelines for analysis and publication, Prehistoric Ceramics Research Group Occas. Papers 1/2 (revised ed.) Schmid, E 1972 Atlas of animal bones for prehistorians, archaeologists and Quaternary geologists Elsevier, London Shackley, M 1981 Environmental archaeology: an introduction George Allen and Unwin, London. Tuck M 2006 ‘Cabot Park, Strategic Drainage Proposals; Second pond at P11 (Pond 6) and pond to the north of the former site of Poplar Farm (Pond 7), Avonmouth, Bristol: archaeological field evaluation’ GGAT Report 2006/092 University of Southampton, 2003 Animal Bone Metrical Archive Project (ABMAP, accessed from http://ads.ahds.ac.uk/catalogue/specColl/abmap/search.cfm) Waban-Smith, F unpublished, Kites Corner Lithic Analysis. In Yates, AM, Locock, M and Walker, MJC A Later Bronze Age coastal landscape at Cabot Park, Avonmouth, Bristol Whittaker, JC 2007 Flint Knapping - Making and Understanding Stone Tools. University of Texas, Austin Yates, AM 2001a Cabot Park Phase 2: Plot P10 Avonmouth, Bristol, Archaeological Field evaluation. GGAT Report 2001/055. Yates, AM 2001b Cabot Park Phase 2: Plots P5/P6 Avonmouth, Bristol, Archaeological Field evaluation. GGAT Report 2001/077. Yates, AM Locock, M and Walker, MJC 2002 A Later Bronze Age coastal landscape at Cabot Park, Avonmouth, Bristol. GGAT unpublished report Yates, AM, Locock, M and Walker, MJC Unpublished, A Later Bronze Age coastal landscape at Cabot Park, Avonmouth, Bristol

37

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Appendix I: Palaeoenvironmental Evidence at Stinkums (Martin R. Bates, M. Allen, J. Crowther, S. Davis, K. Griffiths, S. Jones, P. Robinson, E. Tetlow, J. Whittaker )

Executive summary This investigation was undertaken on a series of samples from boreholes previously drilled at Cabot Park, Avonmouth Plot 9 site. Project guidelines were outlined in the Post-excavation Project Design prepared by the Glamorgan-Gwent Archaeological Trust in June 2008 (GGAT, 2008). This report focuses on the sampling and palaeoenvironmental investigation of the samples reported on by Corcoran (2007). Full details of previous work and the geoarchaeological context of the site are provided by Corcoran (2007). Key aims of the study were to provide a chronological framework for the sequences and to assess samples from cores inside and outside the palaeochannel. This work was to include assessing the presence of pollen, diatoms, plant macrofossils, molluscs, forams and ostracods and insects. In addition loss on ignition and magnetic susceptibility determinations were to be undertaken. The assessment of these materials was to be undertaken in order to attain a clearer idea of the potential of the sequences to provide information on past landscapes. Key findings of the assessment are described: Facies 1. Inter-tidal sand flats. Pollen is limited from this facies. Diatoms are present and indicate marine and coastal conditions, some associated with littoral communities. Forams and ostracods indicate that brackish, saltmarsh and true marine components exist that are explained by reference to a large open estuary or embayment around the site at this time. This infill incorporated not only the local estuarine sediments that were forming at the time, but a great deal of marine sediment thrown in by storm surges, including possible reworked Pleistocene material. A single date was obtained from forams collected from the lowermost sample that provides an age for onset of accumulation of this fossiliferous material of 7900+/-60 B.P. Facies 2. Inter-tidal mudflats (1st phase). Pollen from this facies is limited. Diatoms are present and indicate marine and coastal conditions, some associated with littoral communities. Forams and ostracods indicate that the marine influence begins to diminish through this facies. Also, perhaps significantly, there is no further input of “exotic” material. The sandflats turn to mudflats as the area becomes more and more peripheral to marine influence and as the local estuarine intertidal scenario gradually takes over. Facies 3. Reedbeds/wetlands. Pollen preservation is good with saltmarsh locally indicated with mixed oak woodland in the vicinity below 4.6m. Plant macrofossils suggest reed swamp environments with brackish conditions at this time and some charcoal. Up-sequence carrwoodland occurs as well as taxa associated with agricultural activities. This transformation is supported by a change in diatom floras with the appearance of supra-littoral species and species of freshwater preferences. The change in environmental conditions occurs around 4990+/-40 B.P. based on the 14C date obtained on peat at a depth of 4.61m. Facies 4. Inter-tidal mudflats (2nd phase). Pollen preservation is poor with some indication of arboreal pollen. Diatom floras from these deposits are indicative of

38

Plot P9, Cabot Park, Avonmouth: post excavation analysis

coastal and brackish conditions. Forams and ostracods indicate an estuarine intertidal mudflat environment, with initially quite a high degree of marine influence, which tails off over time as the site appears to become more and more peripheral to marine access through regression of the sea and/or aggradation of the land. The few molluscs present indicate saltmarsh habitats. Facies 5. Palaeochannel fills. Plant macrofossils associated with these deposits are those of wet, maritime conditions but also included charred and non-charred ericaceous material (possibly reworked). Forams and ostracods are limited but are typical of estuarine, tidal mudflats. The few molluscs indicate saltmarsh habitats. Facies 6-8. Saltmarsh. Poor pollen preservation. Charred and waterlogged monocotyledonous plant remains dominate in Baras-layer in Facies 7. Facies 9. Palaeochannel fills (upper). Poor pollen preservation. Forams and ostracods are limited but are typical of estuarine, tidal mudflats with some degree of marine access. Facies 10. Occupation horizon. Poor pollen preservation. Facies 11. Terrestrial environments. Poor pollen preservation. The results of this investigation have supplied information on the nature and timing of environmental changes associated with the earlier Holocene habitats represented at the site. However, poor preservation of contained palaeoenvironmental material in the palaeochannel fills (facies 5 and 9) as well as the saltmarsh facies (6-8) and the uppermost occupation horizon (10) has made it difficult to determine the nature of these environments. The assessments undertaken suggest that no further investigation of molluscan remains and insects are merited due to their general scarcity in the record. The investigation of the forams and ostracods already provide a detailed picture of environmental change and consequently sufficient evidence already exists on these sources; additional work is unlikely to add materially to the story already extracted from this source. Both the pollen and diatom records have provided information on the nature of the sediments. In both cases however, preservation is locally poor and the case for additional work can only be made for diatoms with regards facies 1-3 (where preservation is good) and facies 3 for pollen. In both cases preservation was poor in the archaeologically sensitive parts of the sequence and therefore additional work cannot be justified at present. Finally samples which possibly warrant more detailed work on the plant macrofossils are the two organic rich levels from facies 3 and level 161-166cm from the Baras-layer in facies 7. However, in all cases sample size is small due to the limitations of the core width. Introduction This investigation was undertaken on a series of samples from boreholes previously drilled at Cabot Park, Avonmouth Plot 9 site. Project guidelines were outlined in the Post-excavation Project Design prepared by the Glamorgan-Gwent Archaeological Trust in June 2008 (G-GAT, 2008) following a series of investigations undertaken at the site during 2007. This report focuses on the sampling and investigation of the

39

Plot P9, Cabot Park, Avonmouth: post excavation analysis

borehole samples taken by staff from the Museum of London Archaeological Service (MoLAS) are previously reported on by Corcoran (2007). Full details of previous work and the geoarchaeological context of the site are provided by Corcoran (2007) and the aims and objectives of this study were outlined by Corcoran and are summarised in Section 2 of the current report. Pertinent to the current study a number of key points can be summarised from these previous works: • This is the latest in a series of archaeological and palaeoenvironmental studies undertaken in the area (Locock et al., 1998; Yates, 2001a and b; Tuck, 2006; Askew, 2007; Corcoran, 2007). • These studies have identified evidence of human activity in the area from the Prehistoric period onwards. • Two relict landsurfaces (the Baras or N-layer and the B-layer or Rockingham layer) are known to exist in the area. • Old channels are known to dissect this landscape. Geophysical survey at Plot 9 identified the presence of such features. • As part of the mitigation strategy for the site a series of boreholes were drilled and these boreholes form the basis for the current assessment report. This report attempts to consider the potential of the cores recovered for providing information suitable for landscape reconstructions.

Aims and objectives A number of key aims of the study were outlined by Corcoran (2007) and form the basis for the Project Design for this investigation. These were: • Radiocarbon dating. A chronological framework for the sequences should be constructed focusing on the organic rich horizons where possible. • Assessment of samples from a number of cores inside and outside the palaeochannel. This work was to include assessing the presence of pollen, diatoms, plant macrofossils, molluscs, forams and ostracods and insects. In addition loss on ignition and magnetic susceptibility determinations were to be undertaken. The assessment of these materials was to be undertaken in order to attain a clearer idea of the potential of the sequences to provide information on past landscapes. In particular (although not specified) consideration was to be given of the usefulness of the information to address questions regarding sea level change, habitat change and periods of stability and instability in the landscape. In this report reporting of the results from the specialist assessments are related to core depth and to the facies type previously assigned by Corcoran (2007). Sampling locations for boreholes 1 and 9 are shown in Figures 1 and 2 of that report.

40

Plot P9, Cabot Park, Avonmouth: post excavation analysis

4.1 Sediment assessment 24 bulk sediment samples from borehole 9 were analysed (as detailed in Table 2). Sample locations were taken as close as possible to the pollen and diatom samples to allow cross reference, and included at least one from each of the nine facies identified. Determinations were made of LOI (loss-onignition) and magnetic susceptibility, with a view to assessing their potential for characterising the sediments and for identifying possible phases of sub aerial exposure with associated sediment ‘ripening’ and/or soil development within the sedimentary sequence. Organic matter concentration (estimated by LOI) and magnetic susceptibility are both likely to increase as a result of exposure, the former as a result of plant growth/soil development and the latter as a result of natural fermentation processes within the developing soil (Le Borgne, 1955). Burning would potentially further increase the susceptibility (see reviews by Clark, 1996; Scollar et al., 1990). It should be noted that the sediments have been affected by waterlogging/gleying and are therefore likely to have been subject to postdepositional Fe (iron) mobilisation. These conditions will not only have inhibited organic decomposition, but may also have favoured the preservation of organic remains through pseudomorphic replacement by Fe (as might be evident in soil thin section analysis). On the other hand, because of leaching and possible localised deposition of Fe, the magnetic susceptibility recorded may poorly reflect the character of the sediments/soils at the time of burial – i.e. the magnetic susceptibility data need to be interpreted with caution. Methods Analysis was undertaken on the fine earth fraction (i.e. < 2 mm) of the samples. LOI (loss-on-ignition) was determined by ignition at 550oC for 16 hours. A Bartington MS2 meter was used for magnetic susceptibility measurements. In addition to c (low frequency mass-specific magnetic susceptibility), determinations were made of cmax (maximum potential magnetic susceptibility) by subjecting a sample to optimum conditions for susceptibility enhancement in the laboratory. cmax broadly reflects the overall Fe content of the soil. cconv (fractional conversion), which is expressed as a percentage, is a measure of the extent to which the potential susceptibility has been achieved in the original sample, viz: (c/cmax) x 100.0 (Tite, 1972; Scollar et al., 1990). In many respects this is a better indicator of magnetic susceptibility enhancement than raw c data, particularly in cases where soils have widely differing cmax values (Crowther and Barker, 1995; Crowther, 2003). Under UK conditions cconv values of _ 5.00% are often taken as being indicative of some degree of susceptibility enhancement. cmax was achieved by heating samples at 650°C in reducing, followed by oxidising conditions. The method used broadly follows that of Tite and Mullins (1971), except that household flour was mixed with the soils and lids placed on the crucibles to create the reducing environment (after Graham and Scollar, 1976; Crowther and Barker, 1995).

Results and discussion The analytical results are presented in Table 1. Here, a broad overview/assessment is presented of the two properties investigated.

41

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Loss-on-ignition As would be anticipated the lowest LOI values (1.64 and 2.41%) were recorded in the intertidal sandflats (Facies 1) at the base of the sequence, and the highest value (18.0% at 459 cm) was recorded for one of reedbed/wetland samples (Facies 3). The intertidal mudflat (Facies 2 and 4) and saltmarsh (Facies 6 and 8) sediments have a broadly similar LOI, with values ranging from 2.82–4.39%, whereas in the more recent terrestrial alluvium (Facies 11) at the top of the sequence LOI increases to a maximum of 4.93%. It should be noted that within each of these facies there is some variability in LOI (e.g. 2.97–3.79% in Facies 4). Of the two facies known to be associated with periods of exposure, the Baras-layer (Facies 7) has a notably high value of 10.3% at 164 cm (though only 3.92% at 171 cm), whereas the sample from the occupation horizon (Facies 10) has a low value of 3.30%, which is well within the range recorded for the saltmarsh sediments (Facies 8) that it directly overlies. These results suggest that more organic-rich sediments are likely to be associated with periods of exposure, though not all sediments that have been exposed show clear signs of organic enrichment. On this basis, the variability in LOI recorded within individual facies (e.g. Facies 4, as noted above) could possibly be related to minor phases of exposure – examination of the relationship between LOI and the results of the pollen and diatom analyses should provide further insight into this. Magnetic susceptibility (c, cmax and cconv) Of the samples investigated, the most organic-rich sample from 459 cm in the reed beds/wetland sediments (Facies 3) stands out as having by far the highest cmax value (1750 x 10-8 SI), which is probably attributable to an accumulation of Fe within the reed swamp environment. Otherwise, the cmax values are low, but quite variable (range, 255–741 x 10-8 SI). This likely reflects the presence of generally low and variable concentrations of Fe, which is consistent with the gleying noted above. The c_values are consistently low (range, 4–20 x 10-8 SI) and the neither the samples from the Baras-layer (9 and 11 x 10-8 SI) nor those from the occupation horizon (11 x 10-8 SI) stand out as having higher values than the adjacent sediments. While several of the cconv values, including the occupation horizon (7.14%) exceed 5.00%, it seems likely that both c and cmax will have been affecting by post-burial gleying processes. The cconv values are therefore probably poorly indicative of the magnetic properties of the soils/sediments at time of burial. Conclusions and recommendations The results of the present assessment clearly demonstrate the utility of LOI data in characterising the sedimentary sequence: not only in characterizing the major differences that are apparent from visual inspection (e.g. the more organic-rich reedbed/swamp sediments), but also in identifying more subtle variations both between and within facies that would not otherwise be apparent. On this basis, it is recommended that LOI be included as a component of any further sedimentary analysis undertaken at the site. The fact that the Baras-layer stands out as having a notably high LOI (10.3%) suggests that at least some phases of sub aerial exposure are associated with the development of more organic-rich sediments. If the pollen and/or diatom results demonstrate that some of the minor peaks in LOI within facies (e.g. Facies 4) can also be linked with periods of exposure, then this would further strengthen the case for a fuller programme of LOI to be undertaken on the sedimentary sequence.

42

Plot P9, Cabot Park, Avonmouth: post excavation analysis

In contrast, the magnetic susceptibility values recorded (c, cmax and, consequently, cconv) seem likely to have been affected by post-depositional gleying and provide no meaningful insight into sediment/soil conditions at the time of burial. A case cannot therefore be made for further magnetic susceptibility work on the sediments. It is unfortunate that the potential of phosphate analysis was not assessed, as this is perhaps more likely than magnetic susceptibility to provide evidence of soil ripening/soil development within sedimentary sequences of this type. It is recommended that an assessment of phosphate be undertaken as part any future programme of work at Cabot Park.

4.2 Pollen assessment Twenty four samples were selected from BH9 for pollen analysis to provide an assessment of all the key sedimentary facies identified in Corcoran (2007). These facies numbers are referred to in the description and analysis of the results. The samples were all highly minerogenic and required extensive laboratory preparation. Samples for pollen analysis were prepared using standard techniques (Moore et al., 1991), including treatment with HCl to remove carbonates, micro-sieving through a mesh aperture of 10μ, HF digestion to remove silicates and acetolysis to digest organic matter. A known quantity of Lycopodium spores were added to each sample to provide a counting marker. The residues were mounted in safranin-stained glycerine jelly and analysed under a Leica DMR microscope at a magnification of x400, with critical identifications at x630 and, where necessary, under oil at x1000. Pollen was identified using standard pollen keys (e.g. Andrew, 1980; Moore et al., 1991) and type collections. Plant nomenclature follows Stace (1997). As the analysis was to assessment level, the aim was to achieve a sum of 100 total land pollen (TLP). In sediments of low primary pollen content, the number of pollen grains to 100 Lycopodium spores were counted. Results Pollen preservation was very variable. The lengthy laboratory preparation procedure was designed to concentrate the pollen grains in the residues. Despite this, the pollen concentration was very low and the state of preservation was poor in many of the samples, with a TLP sum of 100 only being achieved in three samples from the organic horizon (Facies 3). In all other samples TLP sums ranged from 1, to a maximum of 59 pollen grains. Results are presented in Table 3. Levels 648 – 504 cm (Facies 1 and 2) The palaeoecological evidence from the lower silts and clays is limited by the sparsity of pollen, with TLP sums ranging from 1 at 520 cm to a maximum of 25 at 544 cm. Chenopodiaceae (goosefoot) dominates the pollen spectra at 544 cm and is recorded in all levels with the exception of 520 cm and 504 cm. Pinus (pine), Quercus (oak), Tilia (lime), Alnus (alder), Corylus (hazel), Salix (willow), Poaceae (grass), Lactuaceae (dandelion family), Asteraceae (daisy family) and Apiaceae (carrot family) are also recorded, albeit sporadically. Polypodium and Filicales spores are also sporadically recorded. Levels 464 – 432 cm (Facies 3) The highest pollen concentrations are recorded in the organic sediments of Facies 3. The pollen record in level 464 cm is dominated by Chenopodiaceae, which almost

43

Plot P9, Cabot Park, Avonmouth: post excavation analysis

certainly indicates the presence of salt-marsh locally. Other taxa commonly associated with halophytic habitats, such as Plantago maritime (sea plantain), Asteraceae (e.g. Aster tripolium – sea aster) and Polygonum aviculare (knotweed) are also recorded, suggesting a continued marine influence during the onset of peat accumulation. A mixed deciduous woodland, with Quercus, Corylus, Pinus, Alnus, Betula (birch) and Tilia is also likely to have persisted in the vicinity of the site. Levels 458 cm and 432 cm are dominated by tree and shrub pollen, in particular Corylus, Quercus and Alnus, and to a lesser extent, Ulmus and Tilia. Tilia (lime) is also well represented at 432cm. The pollen evidence therefore indicates a carrwoodland dominated environment, which is likely to have persisted on wetter substrates, together with mixed deciduous woodland in drier areas. Low amounts of Poaceae and Cyperaceae (sedge) indicate the presence of grass-sedge communities, and other open-habitat taxa, such as Hedera (ivy) and Pteridium (bracken) also reflecting a carr environment. Herbaceous taxa commonly associated with agricultural activities are sporadically recorded, including Cirsium (thistle) and Apiaceae (carrot family), and is accompanied by the presence of charcoal at 432cm. Chenopodiaceae is recorded throughout these levels, albeit at diminished levels, which may suggest a decline in brackish/marine influence. The occurrence of the aquatic taxa Potamogeton (pondweeds) and Myriophyllum spicatum (spiked watermilfoil) may indicate the presence of freshwater nearby, although some species of Potamogeton are also tolerant of more brackish conditions. Levels 426 cm – 238 cm (Facies 4) The transition from the organic sediments of Facies 3 to the silts and clays of the upper section of BH9 corresponds with a decline, once again, in pollen content and state of preservation, with TLP sums of between 12 and 49 being recorded. The principal components of the pollen spectra in these levels are arboreal taxa, whilst Chenopodiaceae and Poaceae are also wellrepresented. Hedera, Calluna (heather), Asteraceae, Apiaceae and Lysimachia vulgaris (yellow loosestrife) are also sporadically recorded, the latter of which is commonly associated with fens and marshes. Levels 230 cm - 182 cm (Facies 6) The samples from this silty-clay unit yielded poor concentrations of pollen, of no more than 5 TLP. Little can therefore be ascertained from the pollen data from this unit. Levels 170 cm and 163 cm (Facies 7) Pollen preservation is poor in these levels, with a maximum of 19 TLP being recorded in this horizon, which represents the Baras-layer or N-layer (Facies 7). At 170 cm, the pollen spectra are dominated by arboreal taxa, including Alnus, Corylus and Quercus. The black horizon of the Baras-layer is represented at 163 cm. Poaceae, Corylus and Polypodium dominate the pollen spectra at 163 cm, accompanied by isolated grains of Cyperaceae, Alnus, Pinus, Quercus, Lactuaceae, Chenopodiaceae, Sphagnum and Filicales. Microscopic charcoal is also recorded in this horizon. Levels 158 cm – 44 cm (Facies 8 – 11) The remaining levels of BH9 similarly yielded low pollen concentrations, thus providing little evidence of the palaeoecological contexts of these upper horizons.

44

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Discussion The sparsity of pollen in the sediments is not unusual for estuarine sequences. Similar problems with low pollen concentrations were encountered at Kites Corner (Walker et al., 1999a), Rockingham Farm (Walker et al., 1998a) and previous studies at Cabot Park (Walker et al., 1999b). The sparsity of the pollen does limit the palaeoecological inferences that can be made from sequence. The only significant results emerge from the organic horizon (Facies 3) of BH9, radiocarbon dated to 4990 ± 40 BP at 461 cm. The pollen data indicates that a saltmarsh environment was prevalent at the onset of organic sediment accumulation during the late Mesolithic/ early Neolithic. A subsequent decline in the brackish/marine influence in the Cabot Park area is indicated by increasing arboreal pollen values, reflecting the establishment of alder dominated carr-woodland on wetter substrates, and mixed deciduous woodland in drier areas. An increasing freshwater influence is further supported by the presence of freshwater aquatic taxa, which indicate the presence of still or running water in the vicinity of the site. It is unfortunate that there is no evidence in this sequence for the well-attested ‘elm decline’ documented in many pollen diagrams from North-west Europe. This may, in part, be due to the coarse sampling resolution employed for the assessment. However, poor representation of Ulmus has been reflected in pollen records from elsewhere in the Severn Estuary (Walker et al., 1997, Walker et al, 1998b) and this seems to be the case in the Cabot Park sequence. Pollen evidence which might reflect human activity is limited. However, the occurrence of cultural indicators commonly associated with pastoral farming activities (Behre, 1986), such as Cirsium and Apiaceae, together with microscopic charcoal may be indicative of anthropogenic activity in the area during the Neolithic period. Interestingly, an increase in Tilia is recorded at 432 cm. A similar event has been recorded at several sites in the Severn Estuary area, and has been dated to 4670 ± 80 BP at Caldicot, 4660 ± 80 BP at Goldcliff (Nayling and Caseldine, 1997), and a later date of 4480 ± 60 BP at Barland’s Farm (Walker et al., 1997).

Conclusions The low pollen concentrations, together with problems, such as the incorporation of secondary and reworked pollen, often associated with estuarine sediments has lead to difficulties in interpretation of the results from Cabot Park, BH9. It is particularly disappointing that pollen preservation within the Baras-layer (Facies 7) is too poor for any reliable vegetational inferences to be made. Only in the three samples from the organic horizon (Facies 3) was a pollen count of 100 TLP achieved during the assessment, and some tentative palaeoecological inferences have been made from the pollen record at these levels. These organic sediments reflect a local vegetational succession from salt-marsh to carr-woodland for the late Mesolithic/ early Neolithic period and onwards, with mixed deciduous woodland dominating the surrounding environment. The evidence in the pollen record of human activity is limited. A more detailed analysis of the organic sediments would be needed in order to clarify the palaeoecological significance of vegetational events, such as the increase in Tilia

45

Plot P9, Cabot Park, Avonmouth: post excavation analysis

within the mixed deciduous woodland in the upper organic horizon, and to reveal possible anthropogenic activity in the locality. A pollen count of 300 TLP would need to be obtained to provide a statistically significant counting level, which would certainly be achievable in theses horizons. This would also need to be supported by further radiocarbon dating. The sparsity of the pollen in the remaining levels would make full analysis a difficult task.

4.3 Diatom assessment Methods Diatom samples were prepared according to the method of Battarbee (1986), where the 24 samples from BH 9 were placed into 40ml centrifuge tubes before the addition of about 20ml of 10%HCl. Samples positions are noted in Figure 4 (sample depths referred to in sampling nomenclature below). A lively reaction was noted in all samples as perhaps would be expected from a brackish- marine environment. When all fizzing had ceased, the samples spun at 1500rpm for 12 minutes after which the supernant was discarded. Another 20ml of 10% HCl was then added to ensure the complete digestion of all carbonate present before being spun again and washed three times with distilled water and spun to remove the acid. About 20ml of cold hydrogen peroxide (H2O2) was then added and left overnight. The samples were then placed into a water-bath to accelerate the rate of organic material digestion. The temperature of the water was gradually increased to 90oC over a couple of days, fresh H2O2 was added as necessary. When all vigorous reaction had ceased, the tubes were topped up with distilled water, spun at 1500rpm for 12 minutes before discarding the supernant. The samples were washed and spun three times to remove all trace of the H2O2. The cleaned samples were transferred to a volumetric flask and made up to 30ml by the addition of distilled water. From this solution, 500μl was transferred onto a 24mm diameter cover-slip with a variable volume pipette. The cover-slips were air-dried overnight and mounted in naphrax, a permanent resin with a high refractive index. Each slide was assessed under a Nikon Labophot-2 phase-contrast microscope with an X40 objective lens and x10 eyepiece to assess homogeneity and diatom concentration. Appropriate dilutions were made for each sample where the concentration of detritus was too high to permit effective enumeration and identification of the diatoms. Diatom identification and counting of 100 frustules was conducted on a Nikon Labophot-2 phase contrast microscope using oil immersion X100 magnification with a X10 eyepiece. Species were identified using Hartley et al. (1996) and Witkowski et al. (2000) with ecological information from Witkowski et al. (2000) and the Geological Survey of Canada (GSC) diatom database (2008). Precise identification in some cases was problematic due to the low numbers of wellpreserved specimens and range of morphological diversity that blurred boundaries between individual species. For example, it was difficult to distinguish between small species of the genus Delphineis and small forms of Cocconeis species in some cases. Normally, identification was made to species level, but where this was not possible only the genus is identified. In these cases, the suffix sp. is used (e.g. Delphineis sp.). Diatom counts were undertaken by Patrick Robson and taxonomic identifications cross-checked with Sarah Davies. Results are presented in Figure 4.

46

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Results There were no diatoms in any of the samples between C1/44 and C/298 (inclusive), which covers facies 11, 10, 8, 7, 6, and the top of facies 4. Furthermore, there are no recognisable fragments of the more resilient elements of diatom valves in these samples. Diatoms are found in small numbers in samples C4/366 to C5/458 and C6/504 to C7/620. Sample C5/464 contains no diatoms whatsoever, while sample C/7648 does contain a few individual species, but too few to reach the count of 100 frustules. It is highly unlikely that any significant losses were experienced during the processing of the samples. A high proportion of samples contained fragments of siliceous sponge spicules even where no diatoms were present, (e.g. sample C5/464). The diatom counts of the Cabot Park samples are presented in Figure 4, and as outlined above, the lowermost sample does contain a few individual diatoms, but too few for a statistically robust count. Only three Paralia sulcata and two Dimeregramma minor were found in multiple passes across the coverslip, together with two badly degraded and unidentifiable frustules. The slide was also scanned at x40 magnification to ensure that a large proportion of the coverslip was checked. Facies 1, 2 and the lowermost sample of facies 3 (represented by sample C7/648), are dominated by P. sulcata, C. belgica and a small Delphineis species, (most likely D. minutissima), which are widespread marine species. Paralia sulcata is a common constituent of coastal plankton but can also occur in the benthos (MacQuoid and Nordberg, 2003) and C. belgica is similarly tychoplanktonic, whilst the genus Delphineis is characteristic of littoral environments. Thalassiosira oestrupii and Dimeregramma. minor are also present in some of the samples but at less than 10% of the total assemblage. Both are cosmopolitan marine species with the former a coastal planktonic diatom, while the later is another species of the littoral community (Witkowski et al., 2000). A couple of individual frustules of Navicula erifuga are the only other significant species to be positively identified, this is cosmopolitan species of brackish and freshwater – particularly with a high iron content, and is common on European coastlines (GSC, 2008). No diatoms were seen in sample C5/464, although a high proportion of siliceous sponge spicules were noted including a particularly distinctive pinshaped specimen. Despite sample C5/458 being the most taxonomically diverse, and ecologically different to the other samples in the sequence, diatom frequency is very low due to poor preservation, and many more traverses of the coverslip were need to attain a count of 100 individuals. It is dominated by Opephora martyi accounting for nearly half of the total sum. This marks a significant change as O. martyi can be considered to be a freshwater (GSC, 2008), littoral and benthic diatom of rivers and lakes, although using the halobian system of salinity classification (Kolbe, 1927; Hustedt, 1953), it is regarded to be indifferent and has been encountered in a range of salinities. It is accompanied by Fragilaria sopotensis, which occurs widely on the saltmarshes of the Baltic Sea, and Fragilaria subsalina a brackish water species also found abundantly on saltmarshes and in microbial mats (Witkowski et al., 2000). F. subsalina is considered one of the best indicators of a littoral/supra-littoral environment (Witkowski et al., 2000). Other freshwater indicators found in this sample, albeit in very low numbers, are Epithemia species and Rhoicosphenia abbreviata. However, P sulcata still accounts for about 20 % of the assemblage and Cymatosira belgica around 10 %.

47

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Although this is a significant reduction from sample C6/504, it suggests that there continues to be a strong influence exerted by the marine environment, likely indicating brackish rather than full marine conditions. Samples C5/432 to C4/366 are once again dominated by P. sulcata and C. belgica with the small Delphineis sp and Dimeregramma minor also present. A few individuals of Actinoptychus senarius are also seen, which like the others is widely found in coastal waters in addition to deltas, estuarineintertidal environments and characteristic of sub-littoral and littoral diatom communities (GSC, 2008).

Conclusions There is a difficulty in drawing definite conclusions from the diatom analysis alone as the question of diatom preservation is likely to strongly influence the sequence. Unfortunately, diatom preservation was too poor in a number of samples to enable any assemblages to be counted. Diatoms in an active environment such as the estuarine-intertidal zone may be subjected to intense physical degradation through mechanical breakage and chemical dissolution. Much of the sequence is dominated by heavily silicified diatom species such as Paralia sulcata and Cymatosira belgica, which are more likely to survive transport by tidal currents (Denys and De Wolff, 1999). However, the presence of other benthic marine taxa in association with these species indicates that the interpretation of a shallow marine environment is justified. Sample C5/458 presents an interesting contrast to the rest of the sequence, where the diatom remains exhibit sigs of both physical and chemical degradation but also greater diversity in species and morphology. The sample may still be dominated by robust species, but the survival of species such as F. subsalina and Epithemia perhaps indicates a change in environment to less saline conditions. When comparing the diatom record to the stratigraphic interpretation of MOLAS (2007), it seems plausible that facies 2 could well represent a high energy, intertidal zone. The top of facies 3 could easily be indicative of a fresh/brackish reed bed or wetland system with a reduction in physical and or chemical processes. The samples from the base of facies 4 are consistent with a return to a high-energy intertidal zone, but it is unclear why the preservation in the upper-most two samples of this facies should not be preserved. Above this point there is no data to draw further conclusions. Preservation problems are also known to occur in salt marshes and reed and fen peats (Denys and De Wolff, 1999), meaning that, where preserved, the diatom record of much of the sequence may have been subjected to significant taphonomic alteration. Given the overall poor preservation and low quantities of diatoms, only limited conclusions can be drawn from this dataset and further detailed work is not recommended.

4.4 Plant Macrofossil assessment 18 sub-samples were taken from boreholes 1, 8 and 9 at predetermined levels for the purpose of assessing the nature, quality and quantity of the plant macro fossil remains. Method The samples were taken over 5cm sections, due to the varying sizes of cores sampled, all the sub-samples were standardised to 100ml of sediment, apart from sample 18, borehole 9, which was sampled over a 1cm section and provided 50ml of sediment.

48

Plot P9, Cabot Park, Avonmouth: post excavation analysis

All the samples were sieved using 1mm and 250um mesh. The samples were scanned using a Wild M5 stereomicroscope and the plant remains were scored on a 5 point scale of abundance. Identifications were achieved by using standard reference texts and modern reference material. Nomenclature follows Stace 1991. Results The plant macro fossils from the sub-samples were generally low in quantity. The results of the assessment are summarised in Tables 5 and 6.

Bore Hole 1 210-215cm, facies 9. The plant remains comprised of a small quantity of monocotyledonous root fragments and very fine root and stem fragments. 255-260cm, facies 5. The plant remains comprised of a very small quantity of monocotyledonous stem and root fragments. Rush seeds (Juncus spp.) were also present in small quantities. 285-290cm, facies 5. The plant remains comprised of a very small quantity of monocotyledonous stem and root fragments. Rush seeds (Juncus spp.) were also present in small quantities. 298-303cm, facies 5. The plant remains comprised of monocotyledonous stem and root fragments. Also present in the sample, was waterlogged heather (Calluna vulgaris (L.) Hull), shoot fragment and a charred crossleaved heath (Erica tetralix L.) leaf. Rush seeds (Juncus spp.) were also present in small quantities. 310-315cm, facies 5. The plant remains comprised of monocotyledonous stem and root fragments and woody stem/root fragments. Sea arrowgrass (Triglochin maritima L.) and goose foot (Chenopodium sp.) were also present in small quantities. 375-380cm, facies 5. The plant remains comprised of fine monocotyledonous stem and root fragments and wood fragments. Oraches (Atriplex spp.) and a small quantity of wood charcoal were also present. Bore Hole 8 114-119cm, facies 10. The plant remains comprised of fine monocotyledonous stem and root fragments. 140-145cm, facies 8. The plant remains comprised of fine monocotyledonous stem and root fragments. 161-166cm, facies 7. The plant remains comprised of a fine monocotyledonous stem and root fragments and a significant quantity of charred wood and monocot material. Also noted was a charred monocotyledonous culm node and a charred sea clubrush/club- rush (Bolboschoenus maritimus (L.) Palla/Schoenoplectus sp.(Reichb.) Palla ) fragment. 168-173cm, facies 7. The plant remains comprised of a small quantity of fine monocotyledonous stem and root fragments.

49

Plot P9, Cabot Park, Avonmouth: post excavation analysis

180-185cm, facies 6. The plant remains comprised of fine monocotyledonous stem and root fragments. 305-310cm, facies 4. The plant remains comprised of monocotyledonous stem and root fragments. A small quantity of rush (Juncus spp.) seeds and one example of a bulrush (Typha sp.) seed were also noted. 340-345cm, facies 4. The plant remains comprised of monocotyledonous stem and root fragments. 385-390cm, facies 4. The plant remains comprised of monocotyledonous stem and root fragments. Bore Hole 9 409-414cm, facies 4. The plant remains comprised monocotyledonous stem and root fragments. Beaked/spiral tasselweed (Ruppia maritima L./R. cirrhosa (Petagna) Grande) fruits were the only identifiable plant remains. 439-444cm, facies 3. The plant remains comprised of monocotyledonous stem and root fragments and wood fragments. 451-456cm, facies 3. The plant remains comprised of monocotyledonous stem and root fragments, which included a small quantity of common reed (Phragmites australis (Cav.) Trin ex Steudal). Also noted were the perianth remains from marsh/golden docks (Rumex palustris Smith/R.maritimus L.) and a small quantity of seeds including red goosefoot (Chenopodium rubrum L.), celery-leaved buttercup (Ranunculus sceleratus L.) and horned pond weed (Zannichellia palustris L.). 462-46ccm, facies 3. The plant remains were similar to those found in level 451456cm, with the presence of monocotyledonous stem and root fragments, which included a small quantity of common reed (Phragmites australis (Cav.) Trin ex Steudal). Also present were the perianth remains from marsh/golden docks (Rumex palustris Smith/R.maritimus L.). A beaked/spiral tasselweed fruit (Ruppia maritima L./R. cirrhosa (Petagna) Grande) also occurred in the sample. Also noted was a small quantity of charred monocot material. Interpretation The plant remains from the deposits were relatively sparse which was possibly due to the nature of the sediments and also due to the quantity of sediment available for analysis. The greatest quantity of remains were provided by the organic levels in borehole 9, facies 3, a level from which a radio carbon date of 4990 ± 40 BP at 461cm was obtained. The plant macro fossils from the lowest level of this facies to be sampled (462-463cm) indicated that conditions were probably of a reed swamp type environment, with brackish conditions occurring, as indicated by the presence of tasselweed fruit. There was also evidence of burning with the presence of charred monocotyledonous plant remains. The plant remains from the next level 451-456cm also indicated a similar environment with evidence of reed swamp conditions, the species present such as horned pondweed, marsh/golden dock, celery-leaved buttercup

50

Plot P9, Cabot Park, Avonmouth: post excavation analysis

and red goosefoot are associated with damp/wet coastal conditions, but which are not generally brackish. Facies 4, considered to be inter-tidal mudflats (Corcoran, 2007) contained only the minimum of plant remains, however the presence of tasselweed fruits in borehole 8 at 409-414cm would indicate brackish conditions. The other plant remains from this facies were limited to a very small quantity of rush seeds and indeterminate monocotyledonous remains. Facies 5, from borehole 1 produced plant macrofossils associated with wet, maritime conditions, also present in level 298-303 were charred and noncharred ericaceous material which may have been eroded out of peat deposits and carried into the palaeochannel. Facies 7 from borehole 8 (the Baras-layer) of which level 161-166cm provided evidence of waterlogged monocotyledonous plant remains and a significant quantity of charred monocotyledonous plant remains. The identifiable remains from this level included a charred culm node and the charred fragment of a sea club-rush/club-rush nutlet. The other sample from this facies at 168-173cm contained only fine monocotyledonous remains. The samples from facies 6-8, from borehole 8 and 1 contained very few plant remains apart from indeterminate monocotyledonous material and rush seeds. The sample from facies 9 from contained only monocotyledonous remains. The sample from facies 10 again contained only monocotyledonous remains.

Discussion The assessment of the plant remains from the boreholes indicated that at the lowest levels (facies 3) a reed swamp environment occurred with some brackish influence at the lowest level. There was also evidence of charred monocotyledonous fragments. The other facies contain too few plant macro fossils to determine the prevalent environment. Facies 4 described as intertidal mudflats produced the minimum of rush seeds and one tasselweed fruit. Facies 5 considered to be a possible palaeochannel fill produced species associated with coastal conditions for example goosefoots, oraches and sea arrowgrass, also present were fragments of heather and cross-leaved heath, which may have been reworked into the palaeochannel from surrounding deposits. The Baras-layer from facies 7, level 161-166cm indicated the presence of both charred and waterlogged monocotyledonous plant remains. To support the evidence for the presence of charred monocotyledonous remains, a charred culm node and clubrush nutlet was identified. Facies 6-8, 9 and 10 contained very little identifiable plant remains apart from rush seeds. All the levels however did contain monocotyledonous plant remains.

Conclusion The assessment of the plant remains from Unit 9, Cabot Park would seem to indicate that the environment changed from a reed swamp, with brackish conditions in facies 3 to what may be interpreted as intertidal mudflats in facies 4. The limited quantity of

51

Plot P9, Cabot Park, Avonmouth: post excavation analysis

remains in facies 5 may indicate the presence of a palaeochannel, with the presence plant remains which could have been eroded from other deposits. The Baras-layer found in facies 7 appeared to be composed of charred and non-charred monocotyledonous stem material. The other facies 8, 9 and 10 produced too few identifiable plant macrofossils to be of value. The samples from the cores were too small to provide enough identifiable remains to make any firm conclusions about the nature of the environment. As an example of other work carried out in the area, the plant macro fossil samples from Kites Corner, Avonmouth, (Walker et al., 1999) were bulk samples ranging from 0.15 to 8 litres, even in the case of this site the waterlogged remains were sparse. Recommendations The samples which possibly warrant more detailed work are the two organic rich levels from facies 3 and level 161-166cm from the Baras-layer in facies 7. However it would be preferable to have larger samples to produce more reliable results.

4.5 Foram and ostracod assessment Twenty two samples were investigated from boreholes 1 and 9. After weighing, each sample was put in a ceramic bowl. The sediment was, if need be, first broken by hand into very small pieces and thoroughly dried in the oven. Boiling water was then poured on the sample and a little sodium carbonate added to help remove the clay fraction on washing. It was then left to soak overnight. After this soaking the samples usually broke down well when washed with hot water through a 75-micron sieve. The resulting residue was finally decanted back into the bowl for drying in the oven. When dry the sample was stored in a labelled plastic bag. Examination of the residue was undertaken under a binocular microscope. First the residue was put through a nest of dry sieves (>500, >250 and >150 microns) and then sprinkled out a fraction and a little at a time onto a tray. Ostracods, foraminifera and other organic remains of interest were picked out with a fine camel-haired brush and placed in a 3”x1” faunal slide for archive purposes. Some semiquantitative measure of the abundance of each species was done by eye and experience (present, common, or abundant/superabundant). Several tables were constructed for each site: the uppermost table giving an indication of the organic remains on a presence (x) or absence basis only and below (if present), semi-quantitative representations of the ostracods and foraminifera, species by species, on which the environmental interpretations has been made. Results are presented in Tables 7 and 8. Results Borehole 1 The results of the microfossil survey through a short (c. 1.5m) sequence of BH 1 are given in Table 7. It covers the interval 2.14m below ground level (+4.13m O.D.) down to 3.60 below ground level (+3.60m O.D.). It is referred to two facies (9 and 5) by MoLAS, interpreted as representing, respectively, an upper and lower palaeochannel infill. The indigenous low-diversity foraminifera and ostracods (colourcoded in Table 7, light grey and lime green, respectively) in all the samples are typical of estuarine, tidal mudflats, with the uppermost sample (Facies 9) indicating some

52

Plot P9, Cabot Park, Avonmouth: post excavation analysis

degree of marine influence (colour-coded blue). There was little, if any, indication of saltmarsh (foraminifera colour-coded blue-green). The faunas are very similar to those of Facies 4 in BH9. Borehole 9 The microfossils found in the samples from BH9, covering a c. 5.7m core, are shown in Table 8. The sequence examined covers an interval from 0.94m below ground level (+5.36m O.D.) down to 6.60m below ground level (-0.30m O.D.), and this was split into nine separate facies by MoLAS (Corcoran, 2007) (in descending order, facies 11, 10, 8-6, 4-1). The uppermost seven samples, covering almost one metre (0.94 to 1.82m below ground level) and separated by MoLAS into five different facies - 11, 10, 8, 7 and 6 (in part), were all barren of microfossils or anything calcareous for that matter. The samples processed down to next-to-nothing, leaving just a residue of iron mineral. The interval 2.30m down to 4.32m below ground level (comprising the whole of Facies 4 and parts of facies 6 and 3 of MoLAS) indicates tidal access, with evidence of the tides finally disappearing above the former sample (2.30m below ground level). This interval indicates an estuarine intertidal mudflat environment, with initially quite a high degree of marine influence (as evidenced by microfossils colour-coded blue in Table 8), which then tails off over time as the site appears to become more and more peripheral to marine access through regression of the sea and/or aggradation of the land. The sample at 4.58m below ground level, within Facies 3, was barren of forams and ostracods. The lowermost c.1.5m of the core, as made available, comprising for the most part, facies 1 and 2 of MoLAS, is interesting, and Facies 1 (6.48-6.60m below ground level or -0.18m to -0.30m O.D.) quite the most interesting of all. These basal samples contain several components (Table 8). First, there is an indigenous brackish fauna of foraminifera (colour coded light grey) and ostracods (colour-coded lime green); it even has the only real evidence of fringing salt-marsh within the sequence (foraminifera colour-coded bluegreen). Second, there is a quite significant marine component (colour-coded blue), many of the foraminifera being clinging (e.g. Ammonia batavu, Miliolinella subrotunda) or attached forms (e.g. Planorbulina mediterranensis, Cibicides lobatulus) which would originally have been living on marine-algae or substrate. Finally, there is an interesting input of “exotics” (colourcoded bright blue and yellow). A very similar picture occurred at the base of the Plot 5000, Avonmouth Western Approaches Distribution Park BH11, situated some 3.5km to the NNE (Whittaker, 2007). The evidence suggests the presence of a large open estuary or embayment around Avonmouth in the early Holocene which was initially infilled as the sea-level rose rapidly. This infill incorporated not only the local estuarine sediments that were forming at the time, but a great deal of marine sediment thrown in by storm surges. There may also be reworked Pleistocene. The cold indicators (colour-coded bright blue under “exotics”) are typical of the “nonanalogue” ostracods which seem to occur in all temperate interglacials, but today are confined to North Britain, or in the case of Hemicytherura clathrata, totally absent from the British Isles and usually considered a cool/cold indicator based on its modern distribution. In comparison, the two “exotic” foraminiferal species (colourcoded

53

Plot P9, Cabot Park, Avonmouth: post excavation analysis

yellow), Elphidium crispum and E. fichtellianum, would indicate a warm interglacial. Unfortunately the microfossil signature of (definite) marine Ipswichian deposits remains illusive and diagnostic microfossils indicative of an earlier interglacial are also lacking. In Facies 2 and the lower part of Facies 3, therefore, marine influence begins to diminish. Also, perhaps significantly, there is no further input of “exotic” material. The sandflats turn to mudflats as the area becomes more and more peripheral to marine influence and as the local estuarine intertidal scenario gradually takes over (Table 8).

4.6 Mollusc assessment Three samples provided sufficient material for investigation. These are listed in Table 9. Methods The specimens were examined under a stereo-binocular microscope at ×10 to ×40 magnification and compared with reference material retained by Allen Environmental Archaeology and standard reference works (eg. Kerney, 1999; Macan, 1977; Janus, 1979). The shells were identified and quantified and are recorded in Table 9 where were nomenclature follows Anderson (2005). Results and discussion Shells were present and recovered from samples at 298-303 and 310-315cm (facies 5) and 409-414cm (facies 4) in BH 9. The number of shells per sample recovered is very low (6 – 18), this is a result of two factors; general poor preservation in these deposits and the very small sample size. Sample sizes of 1000-2000g and greater are normally analysed for snails (Evans, 1972), and in these alluvial intertidal sequences often larger samples of up to 5000g required (Allen & Scaife, 2001). Nevertheless, if small sample size is taken in to account then these samples probably represent c. 60 and 180 shells per kilogram, and would contain enough shells to make significant palaeo-environmental comment. Preservation, however is moderate. The low shell numbers hinder detailed palaeo-environmental interpretation, nevertheless clear differences are present in the assemblages from the two facies allowing some comment, at least to be made. Facies 5: Only one species was present, Ventrosia (Hydrobia) ventrosa; a species generally associated with salt marshes and brackish pools with lowered salinities (typically 5-25 parts per thousand NaCl; Kerney, 1999, 31), living more or less permanently submerged. Facies 4: Although the two samples from this facies show some differences, due to low numbers they will be considered together as representing a single taxocene of this facies. In contrast to the samples from facies 4. Peringia (Hydrobia) ulvae is present with no V. ventrosa, and it is dominant in the lower sample (at 310-135cm). It occurs with Galba truncatula; a ubiquitous amphibious species common in salt marshes, and

54

Plot P9, Cabot Park, Avonmouth: post excavation analysis

the essentially marine or semi-marine species Myosotella myosotis which is also common in salt marshes but usually on upper shores estuaries. There does seem to be an ecological difference between the two facies, but low numbers mean that any interpretative statements have to be made with extreme caution. Nevertheless, both are clearly commensurate with muddy salt marsh habitats. We can tentatively suggest that the presence of Myosotella in this facies may suggest that this semi-marine species has been brought up a channel, but there is no clear indication that these assemblages reflect channel as opposed to saltmarsh habitats.

4.7 Insect assessment Material from two boreholes was assessed. The samples come from Borehole 2 from within a palaeochannel/creek and Borehole 7 from the environment surrounding the channel. The insect remains assessed from Borehole 2 are from six, six centimetre spits at depths of 1.76-1.82m, 2.28- 2.34m 2.90-2.96m, 3.16-3.30m, 3.50-3.62m and 3.82-3.94m. A further sequence of seven samples were recovered from Borehole 7 and subsampled at similar intervals to those of Borehole 2. Sample depths were at 3.22m, 3.74m, 4.34m, 4.46m, 4.52m, 4.72m and 4.8m. Methods The cores were sub-sampled and processed by the standard method of paraffin flotation outlined in Kenward et al. (1980). The insect remains were then sorted from the paraffin flot and the sclerites identified under a low power binocular microscope at x10 magnification. The system for “scanning” faunas as outlined by Kenward et al. (1985) was followed in this assessment. Results are presented in Tables 10 and 11. Results A single sample from this suite of thirteen produced identifiable Coleoptera. Two sclera were observed in material from a sample recovered at a depth of 4.72m, Borehole 7 (Table 11). A single elytron of the hydrophilid, genus Cercyon sp., and a thorax of the staphylinid, Philonthus sp. were recovered from this sample, both fragments were well preserved and, in a larger assemblage would have been readily interpretable. This material was recovered from a sample which consisted of a ‘brown humic clay’ (Facies 3), interpreted as ‘reed swamp’ (Corcoran, 2007). The restricted nature of the coleopterous remains from this deposit precludes further, meaningful interpretation. Discussion The limited Coleoptera from sample 4.72m, Borehole 7, would certainly support the hypothesis that that this period was a brief episode of negative sea-level tendencies and a period of increased terrestrial activity. The extraction of well preserved and interpretable palaeoentomological material from the clays of the lower Wentlooge formation on both shores of the Severn Estuary has proved problematic. At a number of sites, the bluegrey estuarine clays have been sampled as a matter of course when samples were recovered from overlying peat beds of the middle Wentlooge formation, the lower Wentlooge deposit forming the basement of these peat beds (Bell et al., 2000; Paddock, 2003; Smith et al., 2000; Tetlow, 2005, 2007). On the Welsh shore of the estuary at Goldcliff, Goldcliff East

55

Plot P9, Cabot Park, Avonmouth: post excavation analysis

and Redwick, samples from homogenous deposits of estuarine clay were sterile, This paucity was replicated in comparable deposits on the English shore at Minehead and Westward Ho! (Jones et al., 2004; Tetlow, 2005). In contrast material from organic rich lenses within the estuarine alluvium often produced large, well preserved and readily interpretable assemblages. The closest comparable site in the Severn Estuary is Gravel Banks which is located approximately four kilometres from Cabot Park. Deposits from the organic rich clays directly below the peat beds produced well preserved, interpretable palaeoentomological material (Tetlow, 2005). Organic rich lenses within the estuarine clays of the Gwent Levels produced similar, well preserved material which, as at Cabot Park due to assemblage size, had limited interpretable value (Paddock, 2003; Smith et al.; 1997, 2000; Tetlow, 2005, 2007). A fine band of humic material, similar to that at Cabot Park, was found at Goldcliff East, in the lower foreshore at site D, -4m OD, this produced an insect and pollen assemblage indicative of transitional saltmarsh/reedswamp (Bell, 2007; Dark ,2007; Tetlow, 2005, 2007). Conclusions The paucity of remains in 12 of the 13 samples precludes any further palaeoentomological work on this material.

4.8 Discussion The assessment of the samples from the various boreholes and identified facies at the Cabot Park site indicates that preservation of palaeoenvironmental material is variable both within boreholes and between boreholes. Indeed no single source of information has provided a record of associated palaeoenvironmental change through a borehole. Despite this variability in preservation of organic material when taken in combination the evidence does allow the development of the site to be charted in some detail. For ease of discussion the palaeoenvironmental information is considered in relation to the original MoLAS facies descriptions previously outlined by Corcoran (2007). Facies 1. Inter-tidal sand flats. Pollen is limited from this facies. Diatoms are present and indicate marine and coastal conditions, some associated with littoral communities. Forams and ostracods indicate that brackish, saltmarsh and true marine components exist that are explained by reference to a large open estuary or embayment around the site at this time. This infill incorporated not only the local estuarine sediments that were forming at the time, but a great deal of marine sediment thrown in by storm surges, including possible reworked Pleistocene material. A single date was obtained from forams collected from the lowermost sample that provides an age for onset of accumulation of this fossiliferous material of 7900+/-60 B.P. (Table 1). Facies 2. Inter-tidal mudflats (1st phase). Pollen is limited from this facies. Diatoms are present and indicate marine and coastal conditions, some associated with littoral communities. Forams and ostracods indicate that the marine influence begins to diminish through this facies. Also, perhaps significantly, there is no further input of “exotic” material. The sandflats turn to mudflats as the area becomes more and more peripheral to marine influence and as the local estuarine intertidal scenario gradually takes over.

56

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Facies 3. Reedbeds/wetlands. Pollen preservation is good with salt-marsh locally indicated with mixed oak woodland in the vicinity below 4.6m. Plant macrofossils suggest reed swamp environments with brackish conditions at this time and some charcoal. Up-sequence carr-woodland occurs as well as taxa associated with agricultural activities. This transformation is supported by a change in diatom floras with the appearance of supra-littoral species and species of freshwater preferences. The change in environmental conditions occurs around 4990+/-40 B.P. (Table 1) based on the 14C date obtained on peat at a depth of 4.61m. Facies 4. Inter-tidal mudflats (2nd phase). Pollen preservation is poor with some indication of arboreal pollen. Diatom floras from these deposits are indicative of coastal and brackish conditions. Forams and ostracods indicate an estuarine intertidal mudflat environment, with initially quite a high degree of marine influence, which tails off over time as the site appears to become more and more peripheral to marine access through regression of the sea and/or aggradation of the land. The few molluscs present indicate saltmarsh habitats. Facies 5. Palaeochannel fills. Plant macrofossils associated with these deposits are those of wet, maritime conditions but also included charred and non-charred ericaceous material (possibly reworked). Forams and ostracods are limited but are typical of estuarine, tidal mudflats. The few mollusks indicate saltmarsh habitats. Facies 6-8. Saltmarsh. Poor pollen preservation. Charred and waterlogged monocotyledonous plant remains dominate in Baras-layer in Facies 7. Facies 9. Palaeochannel fills (upper). Poor pollen preservation. Forams and ostracods are limited but are typical of estuarine, tidal mudflats with some degree of marine access. Facies 10. Occupation horizon. Poor pollen preservation Facies 11. Terrestrial environments. Poor pollen preservation. The results of this investigation have therefore supplied information on the nature and timing of environmental changes associated with the earlier Holocene habitats represented at the site. However, poor preservation of contained palaeoenvironmental material in the palaeochannel fills (facies 5 and 9) as well as the saltmarsh facies (6-8) and the uppermost occupation horizon (10) have made it difficult to determine the nature of these environments. The assessments undertaken suggest that no further investigation of molluscan remains and insects are merited. The investigation of the forams and ostracods already provide a detailed picture of environmental change and consequently sufficient evidence already exists on these sources; additional work is unlikely to add materially to the story already extracted from this source. Both the pollen and diatom records have provided information on the nature of the sediments. In both cases however, preservation is locally poor and the case for additional work can only be made for diatoms with regards facies 1-3 (where preservation is good) and facies 3 for pollen. In both cases preservation was poor in

57

Plot P9, Cabot Park, Avonmouth: post excavation analysis

the archaeologically sensitive parts of the sequence and therefore additional work cannot be justified at present. Finally samples which possibly warrant more detailed work on the plant macrofossils are the two organic rich levels from facies 3 and level 161-166cm from the Baras-layer in facies 7. However, in all cases sample size is small due to the limitations of the core width.

58

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Figure 1. Borehole 9 stratigraphy

0 - 0.25

9.1 Firm dark brownish grey clay

(P/D) 44cm *

FACIES 11

0.25 - 1.10

9.2 Hard greyish brown iron-stained silty clay some heavily iron concreted patches and occasional modern roots.

1.10 - 1.18

9.3 Firm to hard smooth greenish grey clay

FACIES 10

1.18 - 1.62

9.4 Firm to hard grey brown silty clay. Iron and Manganese precipitations; occasional disrupted sand laminations and occasional charcoal inclusions

FACIES 8

1.62 - 1.66

9.5 Soft black clay with grey gleyed veins

(P/D/F+O) 94cm *

(P/D/F+O) 114cm *

(P/D/F+O) 122cm * (P/D) 140cm * (P/D/F+O) 158cm * (P/D/F+O) 163cm * (P/D/F+O) 170cm *

1.66 - 1.76

FACIES 7 9.6 Soft pale blue grey clay

(P/D/F+O) 182cm *

(P/D) 204cm * 1.76 - 2.34

9.7 Pinkish grey brown (sandy downwards) silty clay; faintly laminated + faint sand lenses. Deposit breaks into layers

FACIES 6

2.34 - 3.10

9.8 Brownish grey silty clay with fine sand laminae - often including shells

FACIES 4

(P/D/F+O) 230cm *

(P/D/F+O) 238cm *

(P/D) 298cm *

59

Plot P9, Cabot Park, Avonmouth: post excavation analysis

(P/D) 366cm *

3.10 - 4.30

9.9 As 9.8 but greyer. Below 4m becomes softer and laminae involuted/disturbed (possibly by drilling)

4.30 - 4.46

9.10 Pale blueish grey (cross) laminated silty clay. Some sand laminae, otherwise like 8.11

4.46 - 4.68

9.11 Brown humic clay becoming peaty humic silt downwards diffuse interfaces with 9.10 and 9.12

FACIES 4

(P/D/F+O) 426cm *

(P/D/F+O) 432cm *

(P/D/F+O) 458cm * (P/D/F+O) 464cm *

FACIES 3

4.68 - 5.15

9.12 Moderatley soft pale blue grey sandy clay. Some organics, including rooting from humic deposit above

(P/D/F+O) 504cm *

60

Plot P9, Cabot Park, Avonmouth: post excavation analysis (P/D/F+O) 520cm *

(P/D) 544cm *

5.15 - 5.84

9.13 Laminated sand and clay - wavy ripple-like laminae. Dark grey oxidises to brownish grey with black manganese patches. Possible organics preserved

(P/D/F+O) 578cm *

FACIES 2

5.84 - 6.22

9.14 Pinkish brown (possibly oxidised from dark grey) laminated sand and clay silt - becoming increasingly sandy downwards

6.22 - 6.60

9.15 Fine brownish grey clayey sand. Clods/clasts of laminated clay preserved in patches, but faint laminations throughout

(P/D) 620cm *

(P/D/F+O) 648cm *

Key (Based on MOLAS, 2007)

FACIES 11 - 'Terrestrial' alluvium

FACIES 10 - Occupation horizon

FACIES 9 - Possible palaeochannel fills

FACIES 8 - Saltmarsh

FACIES 7 - BARAS Layer

FACIES 6 - Saltmarsh

FACIES 5 - Possible palaeochannel fills

FACIES 4 - Intertidal mudflats

FACIES 3 - Reed beds/wetlands

FACIES 2 & 1 - Intertidal sandflats & mudflats

61

FACIES 1

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 1. Analytical data

Depth(cm)

LOI (%)

 (10-8 SI)

Facies 11: ‘Terrestrial’ alluvium 45 4.12 10 95 4.93 14 Facies 10: Occupation horizon 115 3.30 11

max (10-8 SI)

onv (%)

737 678

1.36 2.06

154

7.14

Facies 8: Saltmarsh 123

3.05

10

269

3.72

141 159

3.42 4.39

11 11

475 611

2.32 1.80

9 11

182 741

4.95 1.48

9 6 10

620 546 514

1.45 1.10 1.95

5 8 11 8

293 434 338 251

1.71 1.84 3.25 3.19

12 4 5 18

392 1750 674 275

3.06 0.23 0.74 6.55

20 18 18

274 323 352

7.30 5.57 5.11

6 6

299 255

2.01 2.35

Facies 7: BARAS 164 171

10.3 3.92

Facies 6: Saltmarsh 183 205 231

3.30 3.05 3.56

Facies 4: Intertidal mudflats 239 299 367 427

2.97 3.79 3.46 2.90

Facies 3: Reedbeds/wetland 433 5.40 459 18.0 465 5.44 505 3.01 Facies 2: Intertidal mudflats 521 545 579

2.82 3.09 3.31

Facies 1: Intertidal sandflats 621 649

2.41 1.64

62

Plot P9, Cabot Park, Avonmouth: post excavation analysis SPA Table 2. Pollen data from Cabot Park BH9

CABOT PARK Betula Pinus Ulmus Quercus Tilia Alnus Total Trees Corylus Salix Hedera Ilex Total Shrubs Poaceae Cyperaceae Calluna Lactuaceae Cirsium-type Caryophyllaceae Chenopodiaceae Asteraceae Plantago-type Plantago maritima Filipendula Scutellaria Thalictrum Apiaceae Polygonum aviculare Lysimachia vulgaris Total Herbs TOTAL LAND POLLEN PotamogetonMyriophyllum spicatum Total Aquatics Polypodium8 Pteridium Sphagnum Filicales undiff Total Spores Indeterminate Charcoal Lycopodium clavata*

FACIES 11 C1/44 cm 3 1 4 4 1 5 7 4 2 2 4 -

C1/94 cm 1 1 3 5 1 1 2 2 1 1 1 -

FACIES 10 C2/114 cm 1 1 2 3 3 4 7 3 -

FACIES 8 C2/122 cm 2 2 1 1 4 3 1 -

C2/140 cm 4 3 7 1 1 2 1 2 -

C2/158 cm 2 2 1 1 -

-

-

-

-

-

-

-

-

19 28

5 12

14 19

8 11

6 14

0 3

8 19

1 12

-

-

-

-

-

-

-

-

0 8 4 3 15 106

0 10 1 1 12 102

0 4 4 8 100

0 11 5 16 1 102

0 6 3 2 11 1 110

0 6 6 101

0 7 1 2 10 10 101

0 4 1 5 110

63

FACIES 7 C2/163 cm 2 1 2 5 4 4 5 2 2 -

C2/170 cm 1 5 6 5 5 -

Plot P9, Cabot Park, Avonmouth: post excavation analysis

CABOT PARK Betula Pinus Ulmus Quercus Tilia Alnus Total Trees Corylus Salix Hedera Ilex Total Shrubs Poaceae Cyperaceae Calluna Lactuaceae Cirsium-type Caryophyllaceae Chenopodiaceae Asteraceae Plantago-type Plantago maritima Filipendula Scutellaria Thalictrum Apiaceae Polygonum aviculare Lysimachia vulgaris Total Herbs TOTAL LAND POLLEN PotamogetonMyriophyllum spicatum Total Aquatics Polypodium8 Pteridium Sphagnum Filicales undiff Total Spores Indeterminate Charcoal Lycopodium clavata*

FACIES 6 C2/182 cm 0 0 1 -

C3/230 cm 4 3 7 14 4 4 1 1 -

FACIES 4 C3/238 cm 5 2 6 1 12 26 6 6 6 1 8 1 -

C4/298 cm 1 5 1 7 3 1 4 2 1 -

C4/366 cm 1 45 10 20 11 11 3 1 3 -

C5/426 cm 1 8 3 12 12 1 13 1 7 1 -

FACIES 3 C5/432 cm 4 3 34 10 19 70 22 1 23 3 3 13 2 -

C3/204 cm 1 1 1 3 2 3 5 2 -

-

1 1 1 -

-

1 -

1 -

-

-

1 -

-

-

-

-

-

1

-

-

1 1

5 13

2 20

17 49

4 15

8 39

9 34

22 115

-

-

-

-

1

-

1

-

0 2 2 100

0 3 2 2 7 1 2 105

0 5 2 1 8 104

0 7 4 11 114

1 7 1 3 11 1 1 109

0 9 1 10 1 108

1 10 10 103

0 18 1 2 21 2 128

64

Plot P9, Cabot Park, Avonmouth: post excavation analysis

CABOT PARK Betula Pinus Ulmus Quercus Tilia Alnus Total Trees Corylus Salix Hedera Ilex Total Shrubs Poaceae Cyperaceae Calluna Lactuaceae Cirsium-type Caryophyllaceae Chenopodiaceae Asteraceae Plantago-type Plantago maritima Filipendula Scutellaria Thalictrum Apiaceae Polygonum aviculare Lysimachia vulgaris Total Herbs TOTAL LAND POLLEN PotamogetonMyriophyllum spicatum Total Aquatics Polypodium8 Pteridium Sphagnum Filicales undiff Total Spores Indeterminate Charcoal Lycopodium clavata*

FACIES 3 C5/458 cm 1 2 24 2 20 49 37 1 1 39 5 1 10 1 -

C6/504 cm 2 1 3 1 1 2 4 -

FACIES 2 C6/520 cm 1 1 2 0 1 -

C6/578 cm 1 2 2 5 1 1 2 1 2 -

FACIES 1 C7/620 cm 3 2 2 4 11 4 1 5 3 1 -

C5/464 cm 1 4 1 7 1 3 16 5 5 2 2 78 1

C6/544 cm 5 1 2 8 2 1 3 1 1 11 -

C7/648 cm 1 3 4 2 2 3 4 1 -

1 -

1

-

-

1 -

-

-

-

-

-

-

-

-

1

-

-

18 106

84 105

4 9

1 3

14 25

3 10

4 20

8 14

1 7

-

-

-

1 -

-

-

-

8 0 9

0 9 1 3 14 153

0 2 4 6 111

0 0 101

1 2 1 3 102

0 0 1 100

0 2 2 101

0 0 102

65

Depth (mm)

20 20 40 60 20

De l

40

ta

60 80

Facies

20 40

De lp Di hin m ei O er e s su pe gr r ph a m ir e or m lla a a m m a r in ty or i Fr ag ila ria so po Fr ag te ns ila is ria Na cf .s v ub Rh icul a sa ap e lin r h Di o if a p l n ug o n ei a Di e s p l is nit Pl o ne sp ida eu is O ro s inte do ig r u Ni nte ma pta tz lla sp Co sch sp s ia Am con p a r p od v u Ep ho isc la ith r a us Ac en sp sp h ia Rh n an sp o th Na ico es v sp d e Pl icul h en lica an a ia tu Su oth sp ab la re ri id via Co re lla ium ta cc s lill on p je b ei or s ge co i st at a

sp

su lca

ph in ei s

Pa ra lia

Ac ti Th nop al tyc as h sio u s Cy sir se m a na at o e riu os st s ir a ru pi be i lg ica

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 3. Diatom assessment results

Cabot Park Diatom Counts

50

Facies 11

100

150

Facies 10 Facies 8

200

Facies 7

Facies 6

250

300

350

Facies 4

400

450

500

Facies 3

550

Facies 2

600

650

Facies 1

20 40

66 20 20

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 4. Plant macro fossil remains from borehole 1

Bore Hole Facies Depth

1 9

1 5

1 5

1 5

1 5

1 5

210215

255260

285290

298303

310315

375380

Vol/ml

100 -

100 -

100 -

100 -

100 +

100 -

Atriplex spp. (Oraches)

-

-

-

-

-

+

Calluna vulagaris (L.) Hull (Heather) shoot

-

-

-

+

-

-

Erica tetralix L. (Cross-leaved heath) charred leaf

-

-

-

+

-

-

Triglochin maritimus L. (Sea arrowgrass)

-

-

-

-

+

-

Juncus spp. (Rushes)

-

+

+

+

-

-

++

++

Chenopodium spp.(Goosefoots)

Monocot. Stem frags ++ ++ ++ ++ + = 1-5, ++ = 5-25, +++ = 25 - 50, ++++ = 50-100, +++++ = 100

67

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 5. Plant macro fossil remains from boreholes 8 and 9 Bore Hole Facies Depth

8 10 114119 100 -

8 8 140145 100 -

8 7 161166 100 -

8 7 168173 100 -

8 6 180185 100 -

8 4 305310 100 -

8 4 340345 100 -

8 4 385390 100 -

9 4 409414 100 -

9 3 439444 100 -

9 3 451456 100 +

9 3 462463 50 -

Chenopodium rubrum L. (Red goosefoot) Chenopodium spp. (Goosefoots) Atriplex spp. (Oraches) Rumex palustris Smith/R. maritimus L. (Marsh/golden dock) perianth

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

+

-

-

-

-

-

-

-

-

-

-

-

++

+

Ruppia maritima L./R. cirrhosa (Petagna) Grande (Beaked/spiral tasselweed)

-

-

-

-

-

-

-

-

+

-

-

+

Zannichellia palustris L. (Horned pondweed)

-

-

-

-

-

-

-

-

-

-

+

-

Juncus spp. (Rushes)

+

-

-

-

-

+

-

-

-

-

-

-

Vol/ml Ranunculus sceleratus L. (Celery-leaved buttercup)

68

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Bolboschoenus maritimus (L.) Palla/ Schoenoplectus sp. (Reichb.) Palla (Sea club-rush/Clubrushes) Charred frag.

-

-

+

-

-

-

-

-

-

-

-

-

Typha sp. (Bulrushes)

-

-

-

-

-

+

-

-

-

-

-

-

Phragmites australis (Cav.) Trin. Ex Steudal - stem frags

-

-

-

-

-

-

-

-

-

-

++

+++

++ -

++ -

++ -

++ -

+++ +

++++ +

-

-

-

-

-

-

++ + ++++ ++ ++ ++ Monocot. stem frags ++++ Charred monocot Charred dicot + + = 1-5, ++ = 5-25, +++ = 25 - 50, ++++ = 50-100, +++++ = 100+

69

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 6. Foram and ostracod assessment from borehole 1 ORGANIC REMAINS FACIES 9 Sample depth (bgl) 2.14m Sample depth (O.D.) +4.13m x x x x x

iron mineral (brown) plant debris + seeds molluscs (juveniles) foraminifera ostracods

Ecology

2.40m +3.87m x x

5 3.06m +3.21m x x x x

3.60m +2.67m x x x

Estuarine, intertidal mudflats; latterly with a little marine influence

tidal access

BRACKISH INDIGENOUS FORAMINIFERA FACIES 9 Sample depth (bgl) 2.14m Sample depth (O.D.) +4.13m xx Haynesina germanica Ammonia sp. x x Elphidium williamsoni o Jadammina macrescens

2.40m +3.87m x x

5 3.06m +3.21m xx xx x

3.60m +2.67m xx x

calcareous foraminifera of low-mid saltmarsh and tidal flats

agglutinating foraminifera of mid-high saltmarsh

70

Plot P9, Cabot Park, Avonmouth: post excavation analysis

OUTER ESTUARINE & MARINE FORAMINIFERA FACIES 9 5 Sample depth (bgl) 2.14m 2.40m 3.06m 3.60m Sample depth (O.D.) Miliolinella subrotunda Planorbulina mediterranensis lagenids Elphidium margaritaceum discorbids Ophtalmidium balkwilli

+4.13m +3.87m +3.21m +2.67m x x essentially marine species, but can penetrate outer estuaries x x o o o

BRACKISH INDIGENOUS OSTRACODS FACIES 9 5 Sample depth (bgl) 2.14m 2.40m 3.06m 3.60m Sample depth (O.D.) +4.13m +3.87m +3.21m +2.67m x brackish ostracods of tidal flats and creeks Leptocythere psammophila o xx x Leptocythere lacertosa x Leptocythere castanea xx xx Leptocythere porcellanea OUTER ESTUARINE & MARINE OSTRACODS FACIES 9 5 Sample depth (bgl) 2.14m 2.40m 3.06m 3.60m Sample depth (O.D.) +4.13m +3.87m +3.21m +2.67m x essentially marine species, but can penetrate outer estuaries Leptocythere tenera o Hemicythere villosa o Bonnyannella robertsoni o Cytheropteron nodosum

71

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Paradoxostoma spp. Semicytherura striata

o o

Organic remains are recorded on a presence (x)/ absence basis only Foraminifera and ostracods are recorded: o - one specimen; x - several specimens; xx - common

72

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 7. Foram and ostracod assessment from borehole 9 FACIES

11

10

8

7

6

4

3

2

1

Sample depth (bgl) 0.94m

1.14m

1.22m

1.58m

1.63m

1.70m

1.82m

2.30m

2.38m

4.26m

4.32m

4.58m

5.04m

5.20m

5.78m

6.48m

Sample depth (O.D.) +5.36m

+5.16m

+5.08m

+4.72m

+4.67m

+4.60m

+4.42m

+4.00m

+3.92m

+2.04m

+1.92m

+1.72m

+1.26m

+1.10m

+0.53m

-0.18m

x x

x x

x x

x

x

x

x

x x x

x x x x

x x x x x

x x x x

x x x x

x x x x

iron mineral (brown) iron mineral (black) plant debris + seeds molluscs foraminifera ostracods iron tubes/rhizoliths

x x

x x

Semiterrestrial, but liable to coastal flooding; weathered

Ecology

x x x x x

x x x x

x

Estuarine, intertidal; becoming quite peripheral with declining marine influence

?

Initial infill of large estuarine embayment accompanying rapid sealevel rise; marine and some Pleistocene reworking

tidal access

?

tidal access

4

3

2

BRACKISH INDIGENOUS FORAMINIFERA

FACIES

11

10

8

7

6

1

Sample depth (bgl) 0.94m

1.14m

1.22m

1.58m

1.63m

1.70m

1.82m

2.30m

2.38m

4.26m

4.32m

4.58m

5.04m

5.20m

5.78m

6.48m

Sample depth (O.D.) +5.36m

+5.16m

+5.08m

+4.72m

+4.67m

+4.60m

+4.42m

+4.00m

+3.92m

+2.04m

+1.92m

+1.72m

+1.26m

+1.10m

+0.53m

-0.18m

x x x

xx x xx

x x x o x

x x x

x x x o o

x x

x x x

x x

x

Ammonia sp. Elphidium williamsoni Haynesina germanica Elphidium waddense Jadammina macrescens Trochammina inflata

o xx

o

calcareous foraminifera of low-mid saltmarsh and tidal flats

agglutinating foraminifera of mid-high saltmarsh

OUTER ESTUARINE & MARINE FORAMINIFERA

FACIES

11

10

8

7

6

4

3

2

1

Sample depth (bgl) 0.94m

1.14m

1.22m

1.58m

1.63m

1.70m

1.82m

2.30m

2.38m

4.26m

4.32m

4.58m

5.04m

5.20m

5.78m

6.48m

Sample depth (O.D.) +5.36m

+5.16m

+5.08m

+4.72m

+4.67m

+4.60m

+4.42m

+4.00m

+3.92m

+2.04m

+1.92m

+1.72m

+1.26m

+1.10m

+0.53m

-0.18m

x x x o o o

x x x x

x

x x

x

x

xx

x o

x

xx x x

x o

x

x

xx

x

o

Miliolinella subrotunda Elphidium margaritaceum Ammonia batavus lagenids Cibicides lobatulus Ophtalmidium balkwilli Planorbulina mediterranensis Cyclogyra involvens bolivinids discorbids Elphidium excavatum Patellina corrugata Trifarina angulosa

x o

x o

73

x o x o

x x x x x

essentially marine species, but can penetrate outer estuaries

Plot P9, Cabot Park, Avonmouth: post excavation analysis

"EXOTIC" FORAMINIFERA

FACIES

11

Sample depth (bgl) 0.94m Sample depth (O.D.)

+5.36m

10

8

7

1.14m

1.22m

1.58m

1.63m

+5.16m

+5.08m

+4.72m

+4.67m

6 1.70m

1.82m

4 2.30m

2.38m

3 4.26m

4.32m

4.58m

2 5.04m

5.20m

1 5.78m 6.48m

+4.60m +4.42m +4.00m +3.92m +2.04m +1.92m +1.72m +1.26m +1.10m +0.53m -0.18m

x x

Elphidium crispum Elphidium fichtellianum

warm "southern" marine species

BRACKISH INDIGENOUS OSTRACODS

FACIES

11

10

8

7

6 1.82m

4 2.30m

1.58m

1.63m

1.70m

Sample depth (O.D.) +5.36m

+5.16m

+5.08m

+4.72m

+4.67m

+4.60m +4.42m +4.00m +3.92m +2.04m +1.92m +1.72m +1.26m +1.10m +0.53m -0.18m

x

4.32m

4.58m

x

5.04m

x

x

x x

x

4.26m

4.32m

5.20m

1

1.22m

x o

4.26m

2

1.14m

Leptocythere lacertosa Loxoconcha elliptica Leptocythere psammophila Cyprideis torosa Xestoleberis nitida

2.38m

3

Sample depth (bgl) 0.94m

5.78m 6.48m

x o x o

x x x

x x x xx o

brackish ostracods of tidal flats and creeks

OUTER ESTUARINE & MARINE OSTRACODS

FACIES

11

10

8

7

6 1.82m

4 2.30m

1.58m

1.63m

1.70m

Sample depth (O.D.) +5.36m

+5.16m

+5.08m

+4.72m

+4.67m

+4.60m +4.42m +4.00m +3.92m +2.04m +1.92m +1.72m +1.26m +1.10m +0.53m -0.18m

o x x

5.04m

5.20m

1

1.22m

o

4.58m

2

1.14m

Hemicytherura cellulosa Leptocythere tenera Semicytherura striata Cytheropteron nodosum Paradoxostoma spp. Semicytherura sella Hemicythere villosa Leptocythere pellucida Palmoconcha laevata Bonnyannella robertsoni

2.38m

3

Sample depth (bgl) 0.94m

5.78m 6.48m

essentially marine species, but can penetrate outer estuaries x x o x

o o x o

x x o o "EXOTIC" OSTRACODS

FACIES

11

10

8

7

6 1.82m

4 2.30m

2.38m

3 4.26m

4.32m

4.58m

2 5.04m

5.20m

1

Sample depth (bgl) 0.94m

1.14m

1.22m

1.58m

1.63m

1.70m

5.78m 6.48m

Sample depth (O.D.) +5.36m

+5.16m

+5.08m

+4.72m

+4.67m

+4.60m +4.42m +4.00m +3.92m +2.04m +1.92m +1.72m +1.26m +1.10m +0.53m -0.18m

o

Hemicytherura clathrata Finmarchinella angulata Finmarchinella finmarchica Robertsonites tuberculatus Aurila convexa

Organic remains are recorded on a presence (x)/ absence basis only Foraminifera and ostracods are recorded: o - one specimen; x - several specimens; xx - common

74

x x o o o

cold "northern" marine species

warm "southern" marine species

Plot P9, Cabot Park, Avonmouth: post excavation analysis

SPA Table 8. Mollusc from the Cabot Park 9 borehole samples.

Facies Borehole

5 1

5 1

4 9

Depth (cm)

298-303

310-315

409-414

Sample Vol/ml

4 100 -

5 100 -

15 100 6

2

16

-

2

-

-

-

1

-

6

18

6

Ventrosia (Hydrobia) ventrosa (Montagu) Peringia (Hydrobia) ulvae (Pennant) Myosotella (Ovatella) myosotis (Draparnaud) Galba (Lymnaea) truncatula (Müller) Totals

SPA Table 9. Insect assessment of the samples from Borehole 2

Borehole 2 Depth (m) 1.76-1.82 2.28-2.34 2.90-2.96 3.16-3.30 3.50-3.62 3.82-3.94

Flot

Sediment

X X X X X X

X X X X X X

SPA Table 10. Insect assessment of the samples from Borehole 7

Borehole 7 Depth (m) 3.22 3.74 4.34 4.46 4.52 4.72 4.8

Flot X X X X X  X

Sediment X X X X X X X

75

Comment

Cercyon sp. Philonthus sp

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Bibliography Anderson, R 2005 An annotated list of the non-marine Mollusca of Britain and Ireland. Journal of Conchology 38 607-637 Andrew, R 1980 A Practical Guide to Pollen Analysis. Technical Guide 1, Quaternary Research Association, Cambridge Battarbee, RW 1986 Diatom analysis. In Berglund BE (ed) Handbook of Holocene palaeoecology and palaeohydrology Chichester: Wiley Behre, KE (ed) 1986 Anthropogenic Indicators in Pollen Diagrams Balkema, Rotterdam Clark, AJ 1996 Seeing beneath the soil (2nd edition) Batsford, London Crowther, J 2003 ‘Potential magnetic susceptibility and fractional conversion studies of archaeological soils and sediments’ Archaeometry 45 685- 701 Crowther, J and Barker, P 1995 ‘Magnetic susceptibility: distinguishing anthropogenic effects from the natural’ Archaeological Prospection 2 207-215 Denys, L And De Wolff, H 1999 Diatoms as indicators of coastal palaeoenvironments and relative sea level change. In, Stoermer, E F And Smol, J P (eds), 1999. The Diatoms: Applications for the Environmental and Earth Sciences Cambridge: Cambridge University Press Graham, IDG and Scollar, I 1976 ‘Limitations on magnetic prospection in archaeology imposed by soil properties’ Archaeo-Physika 6 1-124 Gardiner, JP, Allen, MJ, Hamilton-Dyer, S, Laidlaw, M and Scaife, RG 2002 ‘Making the Most of it, late prehistoric pastoralism in the Avon Levels’ Proceedings of the Prehistoric Society 68 1-40 Geological Survey of Canada 2008 Canadian Diatom Database. Available from: http://gsc.nrcan.gc.ca/paleo/diatoms/taxon_e.php. Accessed 20/01/2009. Hartley, B 1996 An atlas of British diatoms - based on illustrations by. Barber, H G and. Carter, J R. Edited by P. A Sims. Bristol: Biopress Hustedt, F 1953 ‘Die systematic der diatomeen in Ihren Bezeihungen zur Geologie und Ökologie nebst einer revision des Halobien-Systems’ Svensk Botanisk Tedskrift 47 509-519 Janus, H 1979 The Young Specialist Looks at Land and Freshwater Molluscs. London: Burke Books Jones, J, Tinsley, H, McDonnell, R, Cameron, N, Haslett, S and Smith, D N 2004 ‘Palaeoenvironmental Analyses from intertidal deposits on the foreshore at Minehead, Somerset’ Archaeology in the Severn Estuary 15 Kenward, H K, Hall, A R, and Jones, A K G 1980 ‘A Tested Set of Techniques for the Extraction of Plant and Animal Macrofossils from Waterlogged Archaeological Deposits’. Scientific Archaeology 22 3-15 Kenward, H K, Engleman, C, Robertson, A and Large, F 1985 ‘Rapid Scanning of Urban Archaeological Deposits for Insect Remains’ Circaea 3 163-72 Kerney, M P 1999 Atlas of the land and freshwater molluscs of Britain and Ireland Colchester: Harley Books Kolbe, R W 1927 ‘Sur Ökologie, Morphologie und Systematik der Brackwasser Diatomeen. Die Kieselalgen des Sperenberger Salzgebietes’ Pflanzenforschung 7 1-146 Le Borgne 1955 ‘Susceptibilité magnétique anormale du sol superficiel. Annales de Geophysique’ 11 399-419 Macan, T T 1977 A key to the British Fresh- and brackish-water gastropods Freshwater Biological Association. (4th edn.)

76

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Moore, P D, Webb, J A & Collinson, M D 1991 Pollen Analysis (2nd ed) Blackwell, Oxford Nayling, N and Caseldine, A E 1997 ‘Excavations at Caldicott, Gwent. Bronze Age Palaeochannels in the Lower Nedern Valley’ CBA Research Report 108 348pp Paddock, E A 2003 ‘A Sub-fossil insect assemblage from the ‘Fourth Peat’, Redwick, Gwent c. 4910 – 2930 BP’ Archaeology in the Severn Estuary 13 47-52 Scollar, I, Tabbagh, A, Hesse, A and Herzog, I 1999 Archaeological prospecting and remote sensing Cambridge University Press Smith, D N, Osborne, P J & Barrett, J 1997 ‘Preliminary Palaeoentomological research at the Iron Age site at Goldcliff 1991-93’ in Ashworth, A C, Buckland, P C & Sadler, J D (eds.) An Inordinate Fondness For Insects: Quaternary Proceedings No. 5. 255-267. Chichester: J. Wiley and Sons Smith, D N, Osborne, P J & Barrett, J 2000 ‘Beetles as indicators of past environments and human activity at Goldcliff’ In Bell, M, Caseldine, A & Neumann, H (eds.) ‘Prehistoric Intertidal Archaeology in the Welsh Severn Estuary’ CBA Research Report 120 245-260. York Stace, C 1997 A New Flora of the British Isles. Cambridge University Press, Cambridge Tetlow, E A 2005 The Palaeoentomology of the Coastal Woodlands and Salt marshes of the Severn Estuary and Bristol Channel. Unpublished PhD Thesis, University of Birmingham. Tetlow, E A 2007 ‘Insect assemblages from the lower and upper peats’ In Bell, M G (ed) Prehistoric Coastal Communities: The Mesolithic in western Britain CBA Research Report 149 201-208. York Tite, M S 1972 ‘The influence of geology on the magnetic susceptibility of soils on archaeological sites’ Archaeometry 14 229-236 Tite, M S and Mullins, C 1971 ‘Enhancement of magnetic susceptibility of soils on archaeological sites’ Archaeometry 13 209-19 Tuck, M 2006 Cabot Park, Strategic Drainage Proposals; Second pond at P11 (Pond 6) and pond to the north of the former site of Poplar Farm (Pond 7), Avonmouth, Bristol: archaeological field evaluation GGAT Report 2006/092 Walker, M J C, Bell, M, Caseldine, A E and Crowther, J 1997 Palaeoecological investigations at Barland’s Farm, Magor, South Wales Report prepared for GGAT Walker, M J C, Caseldine, A E, Barrow, K, James, J H, Cameron, N G and Kreiser, A 1998a Palaeobotanical assessment of Late Flandrian sediments from Rockingham Farm, Avonmouth, near Bristol Report prepared for the Glamorgan-Gwent Archaeological Trust Ltd 20pp Walker, M J C, Bell, M, Caseldine, A E, Cameron, N G, Hunter, K L, James, J H, Johnson, S and Smith, D N 1998b ‘Palaeoecological investigations of middle and late Flandrian buried peats on the Caldicot Levels, Severn Estuary, Wales’ Proceedings of the Geological Association 109 51-78 Walker, M J C, Caseldine, A E, Crowther, J, Johnson, S, James, J H and Macphail, R I 1999a Palaeoenvironmental assessment of Late Flandrian sediments from Kites Corner, Avonmouth, near Bristol. Report prepared for the GlamorganGwent Archaeological Trust Ltd 34pp Walker, M J C, Caseldine, A E, Barrow, K, Cameron, N G, Dobinson, S J, James, J H, Kreiser, A and McPhail, R I 1999b Palaeoenvironmental assessment of Late Flandrian sediments from Cabot Park, near Avonmouth, Bristol Report prepared for the Glamorgan-Gwent Archaeological Trust Ltd 33pp

77

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Whittaker, J E 2007 Plot 5000, Avonmouth Western Approaches Distribution Park. Part 1, Microfossil Assessment Report of Borehole 11. The Natural History Museum, for Oxford Archaeology 4pp Witkowski, A, Lange-Bertalot, H, & Metzeltin, D 2000 ‘Diatom flora of marine coasts’ In Lange- Bertalot, H, (ed.) Iconographia diatomologica 7 Yates, A M 2001a Cabot Park Phase 2: Plot P10 Avonmouth, Bristol, Archaeological Field evaluation. GGAT Report 2001/055 Yates, A M 2001b Cabot Park Phase 2: Plots P5/P6 Avonmouth, Bristol, Archaeological Field evaluation. GGAT Report 2001/077

78

Plot P9, Cabot Park, Avonmouth: post excavation analysis

Appendix II: Notes on the Archive The site archive (including artefacts and ecofacts subject to the agreement of the site owners) will be deposited with an appropriate receiving institution (Bristol Museum), in accordance with the UKIC and IFA Guidelines (Archaeological Archives: a guide to best practice in creation, compilation, transfer and curation (2007)), and also subject to the agreement of the legal landowner. A copy of the report and archive index will be deposited with the National Monuments Record, RCAHMW, Aberystwyth, and the Regional Historic Environment Record, curated by the Glamorgan-Gwent Archaeological Trust.

79