Van de Noort

Ch opter 24

Bronze Age perceptions of wetlonds: recent orchoeologicol work on the estuory

Robert Von de Noort

Introd uction This paper addresses the perception of different ing sea level, while the more extensive wetland wetlands in and around the Humber estuary in development commences from about 8000 cal BC, the Bronze Age. Combining past and current when marine base-levels reach current levels, that research, it will be argued that the perception of is around Ordnance Datum tODl (Gaunt and intertidal wetlands was nearly diametrically Tooley 1974; Berridge and pattison 19g4. Shennan opposed to the perception of riverine floodplains. and Andrews 2000). A wide variety of wetlands This contrasting perception is reflected in the develop, including the raised mires of rhorne and material culture of the Bronze Age, and may be Hatfield Moors and extensive river floodplain explained through the particular manner in which mires in the many rivers in the lowlands around landscapes changed following marine transgres_ the Humber estuary (Buckland and Sadler 19g5; sions. This work was largely undertaken within Gaunt 1994; Dinnin and Lillie 1995; Dinnin l9g7: the framework of the Humber Wetlands Survey, an Long et a|.1998; Neumann 199g;Lillie lggg). integrated archaeological and palaeoenvironmen- tal research programme funded by English The process of wetland deveropment in the river Heritage since Igg2. floodplains has been described in detail elsewhere (see for example Long et al.1g9g; Neumann 199g). and can be summarised as foilows. Marine trans- The Humber estuory gression (during periods of sea-level rise) results in The modern Humber estuary comprises about an impeded run-off of the tributaries of the 30,000 ha of water, sandbanks, mudflats, some Humber as the hydraulic gradient is reduced. This islands and an ever decreasing amount of salt- impeded run-off results in more frequent and marsh. The River Humber commences at the con_ increased longevity of overbank floodings and ris- fluence of the rivers ouse and Trent in the west, ing groundwater tables. This in turn favours the and flows into the between point Spurn growth of species that tolerate a high groundwater and Donna Hook in the east (Fig. 1). Through its table, especially Alnus (alder), in areas which pre_ many tributaries, including the rivers Ancholme viously were typically meadows (within the river and Hull, about 20 per cent of the landmass of floodplains) or deciduous forests (on higher is drained through the Humber (pethick ground). The high groundwater table and frequent 1e90). floods inhibit the humification of plant material, resulting in the development of floodplain peats or The Holocene development of the Humber as an mires. estuary commences only after c.g000 cal BC, with the first evidence for a tidal regime developing in where the floodplains are unconstrained, as is the the lower estuary. From c.6000 cal BC, the run_off case for many of the rivers in the Humber wet_ of this catchment is impeded by the continuing ris- lands including the rivers Trent, lower Ouse,

271 Seandala{ocht2

VALE

OF

YORK

a Howden Metton:::Nocth o ',::,l':i:lF.i!,b Whitlon rio.ngurnbold€ Foxfte et ffi,EHT,H Trent Winteringhorn :i::aBorrow - upon - Falls Humber Sunk r Skeffti ng lsla nd

imm'n9nom'-$- ii:'iiii.

GRI MSBY

FigureI Locotionof the Humberestuory

Foulness, HuIl, and Ancholme, a landward expan- particularly significant. The North Ferriby inter- sion of the floodplain mires is observed during tidal foreshore includes up to 2rn of post-glacial marine transgression. On the side of the river or deposits overlying glacial till, first noted in 1930 estuary, reed swamps, saltmarsh and mudbanks and published by Bisat (1932). Frequent visits to may deveop, resulting in peat and clastic sedi- the foreshore by Christopher and Edward Wright, ments overlying the basal peats. In periods of initially for the study of faunal remains, led to the marine regressioh, & seaward expansion of the identification of archaeological structures, includ- floodplain mire can be observed, resulting in an ing hurdles, of which one example was excavated intercalated or upper peat. The maximum lateral at a later date and dated to the Middle Bronze extent of floodplain mire development is reached Age (McGrail 1998 and pers. comm.). In 1937, the at c.1000cal BC. remains of a prehistoric boat, Ferriby-l, were found and in the following decades, remains of The existence of a basal peat in the floodplains two additional Bronze Age and one Iron Age boats throughout much of the Humber wetlands outside were discovered (cf. Wright 1990). the river channels has formed the basis for much work there, including the development of a simple The remains of Ferriby-l comprise a near-com- mathematical model in which altitude, position in plete bottom structure, including remains of two the estuary and date of basal peat are linked keel planks, outer bottom planks and part of the 'inserted (Shennan and Andrews 2000). Using this model, it lowest side-strake, together with an is possible to determine whether floodplain mire piece' in one of the outer bottom planks, repre- development had commenced at a certain time in senting a major repair in antiquity. All planks a certain place. This model can be used to deter- were fashioned from Quercus (oak). The overall mine the full extent of wetland development in dimensions of the in situ boat remains were the Humber wetlands at any point in time from 13.32m by 1.67m. The main characteristics 'sewing', c.5000 years BC. include the fastening of the planks by using withies of twisted ye*, at an interval of 0.2 to 0.3m. The seams of the planks were grooved, BronzeAge remoinson the Humber and made watertight with moss. A second charac- estuory teristic was the use of transverse timbers which The archaeological resource of the Humber estu- were passed through integral cleats which had ary has long been recognised as of great impor- been carved out of the solid planks, giving the hull tance, with the finds from North Ferriby being the required stiffness. Recent radiocarbon assay

272 Van tle Noort

aimed at providing more accurate dates for che Ferriby boats, which included the complete removal of contaminants including waxes thal had been used in conservation, provides a date of 1gB0- 1690 cal BC for Ferriby 1 (Wrrght et at.200I).

The design of Ferriby-2 and trerriby-3, as can be ascertained from the fragmentary remains, closely resembled that of Ferriby-l. The recent research redated Ferriby 2 to lg20-r770 cal BC and Ferriby 3 to 2020-1780 cal BC (Wright et a\.2001). Most Figure3 BronzeAge trockwoy on the Humberfore_ shore,c.l40O-c.900 col BC.photo graph copyright interestingly, the fragments of Ferriby_B were Hurnber Weflonds projecf, lJniversityof Hult found resting on a platform of Alnus (alder) posts and a worn Quercus (oak) plank, suggesting that sive survey of the intertidal wetlands of the the boat-planks had been put aside, presumably Humber foreshore. During lgg7-gg, over 40 for some sort of repair. This find, and the concen- archaeological sites were discovered, the majority tration of boat remains, form the basis for the sug- of which were dated to the Bronze Age, the Roman gestion that the North Ferriby foreshore was a period and the post-medieval period. sites of the Bronze Age boatyard. BronzeAge and Roman period relate to landscapes which were subsequently buried by sediments fol- Parts of the estuary were surveyed between rgg4 lowing marine transgressions, but the post_ and 1997 as part of the Humber Wetlands Survey, medieval sites have never been fully buried and but the acquisition of a rigid inflatable boat in timber structures and alignments remain visible 1997 provided opportunities for the more exten_ at low tide.

Figure2Ihe surveyof o BronzeAge trockwoy (c.14OO cc BC)on the Humberforeshore. phofogroph copy_ righf Humber WeftandsProjecf, Universifyof Hutt

273 Seanddla{ocht2

Among the Bronze Age finds was a fragment of a Both trackways were built in a saltmarsh environ- plank-built boat, which was found on the North ment, and were found to have been constructed Sea coast at Kilnsea, but which must be attributed across tidal creeks. Although such trackways may to the Kilnsea Fleet, a relatively small tributary of have been used in fishing and water fowling, it is the Humber (Van de Noort et al. 1999). Although more probable that they were primarily built to only a single plank was recovered, it shows the maximize or maintain the exploitation of the salt- characteristic integral cleat, through which trans- marsh as pasture for cattle, and possibly sheep. verse timbers were put to provide stiffness to the The rationale for this interpretation comes partly hull. The fragment was dated to 1870-1670 cal BC, from the location and design of the trackways. The or Early Bronze Age. It has been argued that the trackways did not provide access to the waterfront, sewn or plank-built boats of the Bronze Age were but rather were positioned alongside the river, and used for communication with , south the method by which the hurdles were secured into 'heavy of the Humber, and continental Europe, thereby their place suggests some sort of traffic'. playing a role in exchange and the maintenance of Furthermore, the remains of aurochs have been elite networks which were well established by this found on the North Ferriby foreshore, while the date (Van de Noort et a|.1999). practice of saltmarsh exploitation as pasture for cattle and sheep is well known from the medieval Sites dated to the period c.1800 cal BC to 400 cal and post-medieval periods (see for example Rippon BC include individual stakes, a platform and stake 1997). It may be postulated that the trackways alignments. The majority of these sites survive were constructed during a marine transgressive only as the bottom ends of the stakes driven into phase, during which tidal creeks typically extend the peat, with the contemporaneous surface levels landwards. This would evidently have restricted desiccated and eroded. On the basis of correspond- the exploitation of the saltmarsh, unless access to i.tg structures from the Essex coast and the saltmarsh across tidal creeks could have been pro- Solent, the majority of these sites are thought to vided. have been fish traps and weirs, hides and plat- forms for the hunting of wildfowl (cf. Wilkinson In short, the Bronze Age archaeological resource of and Murphy 1985; Loader et a|.1997). the Humber includes nautical finds, making the estuary a gateway for communication and Furthermore, a number of trackways of Bronze Age exchange, and also a large number of structures date were identifred and two were partly excavated. illustrating the high biological productivity of the The excavated trackways display a number of com- estuary. Its exploitable resource potential was high mon characteristics. The woven panels were con- and included fish, waterfowl and rich pasture land, structed on dryland from mainly coppiced hazel, in although a certain inventiveness and determina- the same way wattle fencing is made, before being tion was required from anyone who wished to transported to the estuary. The hurdles of the earli- exploit this resource. er of the two trackways, which has been dated by radiocarbon assay to c.1400 cal BC, were laid direct- ly on the saltmarsh, and held in place by stakes dri- Bronze Age remoins from the river ven or hammered into the muds (Fletcher et al. flood ploins 1999; Fig. 2). The second of the two trackways, While the wetlands in the Humber estuary played which has been dated to c.1400-c.900 cal BC, was an important role in Bronze Age subsistence and also made of hurdles, but its construction lacked the exchange, the wetlands which developed in the finesse of the other trackway. The hurdles were laid tributaries of the Humber. such as the rivers upon roundwood timbers, with one end resting on a Trent, Ancholme, Hull and Foulness, offer a very cross-piece which in turn was supported by the ver- different picture. The systematic survey of selected ticals (Fig. 3). The hurdles themselves were again areas of these rivers shows a particular close spa- held in place by stakes driven into the saltmarsh. tial association between the late Mesolithic and

) 7,4 Van de Noort

early Neolithic wetlands but the distance However, Bronze Age activity in the river flood- between the material culture of the late plains of the Humber wetlands is represented by Neolithic and Bronze Age periods to the nearest metal objects, especially axeheads, weapons and river is considerably greater. In areas where the ornaments, which have been recorded since the survey was undertaken unhindered by modern eighteenth century. About B0 locations of Bronze land use, for example in the Humberhead Levels Ag" gold or bronze finds from the region are or the Ancholme valleSr, the mean distance of recorded, concentrated in the floodplains of the sites to the nearest contemp orary river was rivers Trent, Ancholme, Hull and Foulness (Dudley c.100m for the late Mesolithic and early 1949; Davey and Knowles Ig72; Davey lgTB; Neolithic periods, but nearing c.b00m in the late Loughlin and Miller lgTg; Van de Noort and Neolithic and Bronze Age periods (Van de Noort Fenwick et a\.1998). At several locations, for exam_ et al. 1997, Van de Noort and Neumann et al. ple at Burringham on the Trent floodplain, at 1998). No spatial correlation between sites of Appleby on the Ancholme floodplain and at Leven Iron Age date with their nearest rivers was on the Hull floodplain, concentrations or hoards of established during these surveys. metalwork were found, but other finds remain iso_ lated (cf.van de Noort and Davies 1998). There can The growing distance between archaeological be little doubt that these were votive deposits, and material and wetlands has been mainly attrib_ the practice has been described for much of mod- uted to the increasing importance of agricultur_ ern, western and central Europe as metal deposi_ al activities, and a decreasing role of hunting, tion in'wet places' (see for example Bradley 1990). fishing and gathering within the subsistence strategies. Evidence for small-scale forest clear- The antiquarian dates of the majority of these dis- ance and limited agricultural activity from coveries precludes detailed information on stratig- pollen analysis has been dated to c.2500 cal BC raph5r, but where such information was noted, the (cf. Neumann 1998), but evidence for larger_ context of the gold and bronze objects is invariably scale agricultural activity is not recorded in the on top of the clay but beneath the peat, pertaining region until the later Bronze Ag", around 1000 to the early stages of mire development in the find cal BC (see for example Taylor 1gg5). The pollen Iocation (see for example Davey and Knowles analysis from the region suggests a pastoral_ 1972). The use of the mathematical model in which dominant subsistence between c.2500 and 1000 altitude, position in the estuary, and date of basal cal BC, and the widespread introduction of peat are linked, confirms a broad correlation arable agriculture only after 1000 cal BC. between the dated objects and the contemporane- Nevertheless, settlement locations from c.2800 ous onset of wetland development at the find spot. cal BC are considered to have been selected pri_ Furthermore, the type of wetland was always a marily for their proximity to the arable produc_ floodplain mire, particularly alder-dominated fen. tion units, and this consideration was growing In contrast, only a single bronze object of Bronze in importance over time. Age date has been found on the foreshore of the Humber itself. This is a knife from the North Thus, few Bronze Age archaeological sites are Ferriby foreshore (Wright 1gg0). known from the wetlands in the river floodplains of the Humber wetlands. Settlements appear to be concentrated on the drylands or on the wet_ BronzeAge perceptions of weilonds land margins. The only recorded trackway from a This illustrates the varying perceptions of wet_ river floodplain, at in the Ancholme valley, lands in the Bronze Age in the Humber area. On provides access across, rather than access into, one hand, the intertidal wetlands of the Humber ,raft, the floodplain mires, and the Brigg has sim_ foreshore can be characterrzed. by their high bio_ ilarly been described as a ferry for transport logical productivity and exploitable resource, with across the Ancholme valley (McGrail lggg). abundance of fish and waterfowl, and extensive

275 Seandala[ocht2 opportunities for pasture and exchange. This is erosion and desiccation (Van de Noort and Ellis reflected in the archapological record - boats, 2000). Nevertheless, although Neolithic-period trackways, fishweirs, fishtraps and a platform wetland landscapes remain overwhelmingly which may have been used for wildfowling. On the buried by deep sediments of later date, the spatial other hand, the expanding alder-dominated fen distribution of polished stone axeheads on the mires in the river valleys have a lower biological floodplain of the River Ancholme appears closely productivity, and a more limited exploitable associated with the lateral wetland expansion in resource. Most significa4t in the pastoral economy the Neolithic period (cf. Van de Noort and Davies of the Bronze Ag", the replacement of former 1993). This distribution has been interpreted as meadowland by fen-type mires, which are unsuit- representing a very considerable amount of forest able for use as pasture, must have been a signifi- clearance (May I976),but this does not explain the cant landscape change from a human perspective. occurrence of axeheads on the floodplain, which Here, the archaeological record is one of metal would have been unsuitable for agriculture. deposition and sites which offer a route across Landscapes of Iron Age date have suffered exten- wetlands, such as the trackway and boat from sively from desiccation and erosion, and few metal Brigg. finds of this period are recorded from the Humber wetlands, despite the fact that throughout western If we consider the role of these votive deposits as and central Europe the frequency of metal deposi- 'grfts to the gods', and an association of human tion appears to increase in Hallstatt D (from c.600 burials with these votive deposits, which has been cal BC). observed outside the Humber wetlands in the Thames estuary and in the East Anglian Fens (see for example Bradley and Gordon 1988; Bradley Conclusions 1990: HalI and Coles 1994). a fundamental divi- Few attempts have been made to understand the sion can be noted in the perception of these differ- environmental context of the phenomenon of 'wet ent wetland types. The intertidal wetlands of the votive deposits in places'. However, on the Humber estuary were perceived as living wet- basis of closely integrated research by archaeolo- lands, central to subsistence and exchange activi- gists and palaeoenvironmentalists within the ties, and devoid of religious or ideological activity. framework of the Humber Wetlands Survey, where The mires of the river floodplains were perceived we have been able to compare the perception of dif- 'lifeless' as wetlands, peripheral to subsistence ferent types of wetlands, it was concluded that pos- activities but central to religious activity. sibly as early as the Neolithic, and certainly by the Bronze Ag", mire-type wetlands were perceived as Although votive depositions in'wet places' outside radically different from the intertidal wetlands. the Humber wetlands are dated from the early The intertidal wetlands of the Humber estuary Neolithic through to the late Iron Age (see for were perceived as living wetlands, central to sub- example Bradley 1990), a bias in the wetland sistence and exchange activities, and devoid of reli- archaeological record in the Humber wetlands pre- gious or ideological activity. The mires of the river 'lifeless' vents detailed investigations of perceptions of wet- floodplains were perceived as wetlands, lands for these periods. This bias is the result of peripheral to subsistence activities but central to marine transgtessions and regressions and recent religous activity. Van de Noort

References

Berridge, N.G. and Pattison, J. 1994. Geolngy of the country around, Grimsby and patrington. Memoir for 1:50,000 sheets 90 and 91 and 81 and 82 (England and Wales). London: British Geological Suwey, HMSO.

Bisat, W.S. 1932. Glacial and post-glacial sections on the Humber shore at North Ferribv. Ilans Hull Geolngical Soc.?, 83-95.

Bradley, R. 1990' The passage of anns: an archaeological analysis of prehistoric hoards and uotiue deposits. Cambridge: Cambridge University Press.

Bradley, R. and Gordon, K. 1988. Human skuls from the river Thames.Anfi quity g2, 50L-9.

Buckland, PC. and Sadler,J. 1985.The nature oflate Flandrian alluviation in the Humberhead Levels. East Midlands Geographer 8, 2J9-El.

'Bronze Davey,P.J. 1973. Age metalwork liom Lincolnshire,. Arci aeotogia 1:O4,E!_I2T . 'The Davey, P.J. and Knowles, G.c. 1972. Appleby Hoard'. Arclr aeological Journal rz}, rs4-6l. Dinnin, M. 1997.The palaeoenvironmental survey ofthe Rivers Idle, Torn and Old River Don. In R. van de Noort and Ellis, S. (eds), Wetland heritage of the Humberhead. Leuels: an archaeological surwey. ;gttll: Humber Wetlands Project, University of HulI, 4Z-28. Dinnin, M. and Lillie, M. 1995. The palaeoenvironmental survey of southern Holderness and evidence for sea-level charrge. In R. van de Noort. and s. Ellis (eds), wetland heritage of Hold.erness: an archaeological suruey. HuIl: Humber Wetlands project, University of HuIl, gZ_120-

Dudley, H. 1949-Early days in north-west Lincolnshire- scunthorpe: w.H. anrl c.H. caldicott.

Fletcher, w., Head, R-, chapman, H., Fenwick, H., van de Noort, R., and Lillie, M. 1999.The archaeological

survey of the Humber estuarJz.In R. Van de Noort and S. Ellis (forthcoming) Wetlan6 heritage of the Va-le ofYork: an archaeological survey. Hull: Humber Wetlands Project, University of Hull.

Gannt, G.D. 1994. C,eology of the country ciround Goole, Doncaster and the IsIe of Athnlme. Memoir for one- inch sheets 79 end 88 (England. and Wales). London: HMSO.

Gaunt, G'D. and Tooley,M.J.1974. Evidence for Flandrian sea-levelchanges in th'e Humber estuary and adjacent areas.Bulletin of the GeologicalSociety of Great Britain g,2b_41.

Hall' D. and coles, J.M. !994. Fenland,survql An essayin land,scapeand persistence.London: English Heritage.

277 Seanddlqiocht2

Lillie, M. 1999.The palaeoenvironmental suwey of the Humber estuary. In R. Van de Noort and S. Ellis (eds),Wetland heritage of the VaIe ofYorh: an archaeologicalsuruey. HulI: Humber Wetlands Project, University of Hull.

Loader, R. and Tomalin, D. IggT - Time and tide: An archaeologicalsuruey of the Wootton-Quarrcoasr. IsIe of Wight: IOW Council.

Long, A.J., Innes, J.B., Kirby, J.R., Lloyd, J.M., Rutherford, M.M., Shennan, I. and Tooley,M.J. 1998.Holocene

sea-levelchange and coastal evolution in the Humber estuary, eastern England: an assessmentof rapid coastal change.The HoloceneI (2),229-47.

Loughlin, N. and Miller, K.R. 1979.A suruey of archaeological sitesin Humberside. : Humberside Libraries and Amenities.

May, J.1976. Prehistoric Lincolnshire - Lincoln: History of Lincolnshire Committee.

McGrail, S. 1998. Humber wrecks. In J.P. Delgado (ed.)Ezcyclopaedia of underwater and maritime archaeology.London: British Museum Press, 199-200.

Neumann, H. 1998.The palaeoenvironmental survey ofthe Ancholme valley. In R. Van de Noort and S. Eilis

(eds),Wetland heritage of the Ancholme and lower T'rentvalleys: d.narchaeological suruey. Hull: Humber Wetlands Project, University of Hull, 75-101.

Pethick, J.S. 1990. The Humber estuary. In S. Ellis and D.R. Crowther (eds),Humber perspectiues:a region through the ages.HulI: University Press, 54-70.

Rippon, S. L997. Seuernestuary. Landscape euolution and wetldnd reclamqtion. Lercester:Leicester University Press.

Shennan, I. And Andrews, J.E.2000. HoloceneLand-Ocean Interdction and Enuironmental Change around the North Seo. London: GeologicalSociety, Special Publications 166.

Switsur, V.R. and Wright, E.V 1989.Radiocarbon ages and calibrated dates for the boats from North Feniby, Humberside - a reappraisal. Archaeological Journal 148,58-67.

Taylor, D. 1995.New Polien data from the Keyingham valley, southern Holderness.In R. Van de Noort and S. Ellis (eds.)Wetland heritage of Holderness:an archaeologicalsui"uey. Hull: Humber Wetlands Project, University of Hull, 121-8.

Van de Noort, R. and Davies, P. 1993. Wetland heritage, an archaeological d,ssessmentof the Humber

Wetlands. HulI: Humber Wetlands Project, IJniversity of Hull. Uande Noort

van de Noort, R. and Ellis, s. 2000. The Hu.b", u.tuury' managing the archaeologicalresource ln a dynamic environment. In K. Pye and J.R.L. Alien (eds) coastal and,estuarine enuironments: sedimentolngy,geomorphology and geoarchaeolagy.London: GeoiogicalSociety Special publications 175,4L9-27.

van de Noort, R. Dinnin, M., Lillie, M., Head, R., Fenwick, H. anil chapman ,H. rgg7. conclusion. In R. van de Noort and S. Ellis (eds).Werland heritage of the Humberhead,Levels: an archaeologicalsurvey. H;ulll: Humber Wetlands Project, University of Hull, 453-61.

van de Noort, R., Fenwick, H., and Head, R. 1998. Introduction to the archaeologicalsurvey. In R. van 4e Noort and S. Ellis (eds).Wetland heritage of the Ancholme and. Iower Trent ualleys:an archaeological survey. Hu\l: Humber Wetlands Project, University of Hull, 123-40.

van de Noort, R., Neumann, H., Li]]ie, M., Head, R., Fenwick, H. and chapman, H. 199g. conclusion. In R. Van de Noort and S. Ellis (eds) Wetland heritage of the Ancholme and lower Trent ualleys:an archaeologicalsuruey - Hull: Humber WeUands project, University ofHull,2gg_96. 'Kilnsea van de Noort, R., Middleton, R., Foxon,A. and Bayliss, A. 1999.The boat', and some impiications from the discovery of England's oldest boat remains.Antiquity 78,1.BI_L4}.

Wilkinson, T.J' and Murphy, P. 1995. Archaeology of the Essex coast, Volume I: the Hullbridge suruey. East Anglian Archaeological Reports 71.

Wright, E.Y 1990. The Ferriby boats:seacraft of the BronzeAge. Lontlon: Rou edge.

I

I 279 I l