the quaternary of east yorkshire and north lincolnshire

81 downloads 61 Views 2MB Size Report
(Radley & Simms 1967). Risby and Crosby Warrens and the upper sand unit ...... Kenward H. K. & Hal, A. R. 1995. Biological Evidence from 16-22 Coppergate.
THE QUATERNARY OF EAST YORKSHIRE AND NORTH LINCOLNSHIRE Field Guide

Edited by M.D. Bateman, P.C. Buckland, C.D. Frederick & N.J. Whitehouse

Contents iii Cover Photograph: Digital elevation map (DEM) of North Lincolnshire and East Yorkshire based on Ordnance survey 50 m resolution DEM. Reproduced by kind permission of Ordnance Survey. © Crown Copyright NC/01/476

Produced to accompany the Field Meeting to East Yorkshire and North Lincolnshire, based at University of Sheffield, 13th-16th September 2001. ISSN 0261 3611 ISBN 0 907780 54 7

© Quaternary Research Association, London, 2001

All rights reserved. No part of this book may be reprinted or utilised in any form or by any electronic, mechanical or other means, now known or hereafter invented, including photocopying and recording or in any information storage or retrieval system without permission in writing from the publishers.

Cartography, design and typesetting: Paul Coles & Graham Allsopp, Cartographic Services, Department of Geography, University of Sheffield, Winter Street, SHEFFIELD S10 2TN. Set in Giovanni Book and Optima using Quark XPress and Macromedia FreeHand on Macintosh computers.

Printing: Frontier Print & Design, Pickwick House, Chosen View Road, CHELTENHAM GL5 9LT.

QRA Publication Secretary: A.J.Howard, School of Geography, University of Leeds, LEEDS LS2 9JT

Recommended reference: Bateman, M.D., Buckland, P.C., Frederick, C.D. and Whitehouse, N.J. (eds), 2001. The Quaternary of East Yorkshire and North Lincolnshire. Field Guide. Quaternary Research Association, London.

Contents Contents Contributors Location map Acknowledgements Preface

iii v vii viii ix

Introduction to the Late Quaternary of East Yorkshire and North Lincolnshire The Glacial History of East Yorkshire D.J.A. Evans, S.A. Thomson & C.D. Clark

xx

Late Quaternary Record from Beyond the Icesheets M.D. Bateman & P.C. Buckland

xx

The Regional Fluvial Record A.J. Howard

xx

Regional Late Quaternary Marine and Perimarine Record J.R. Kirby

xx

Regional Vegetational History J.C. Tweddle

xx

Holocene Human-Landscape Interactions M.C. Lillie

xx

Day 1: The Holderness Coast Dimlington Cliff J.A. Catt

xx

Routh Quarry B.R. Gearey & M.C. Lillie

xx

Gembling S.A. Thomson & D.J.A. Evans

xx

Sewerby J.A. Catt

xx

Speeton Cliff G.D.Gaunt

xx

Fimber P.C. Buckland

xx

Day 2: Ice Marginal North Lincolnshire Armthorpe P.C. Buckland

xx

South Ferriby Cliff and Eastfield Farm C.D. Frederick, P.C. Buckland, M.D. Bateman & B. Owens

xx

Yarborough Quarry J.B. Murton, M.D. Bateman & M. Dinnin

xx

Contributors v

iv Contents Twigmoor Woods M.D. Bateman, C. Bristow & I. Livingstone

xx

Black Walk Nook M.D. Bateman

xx

Day 3: The Humberhead Levels Gringley-on-the-Hill P.C. Buckland

xx

Cove Farm Quarry M.D. Bateman, P.C. Buckland, R. Carpenter, S. Davies, C.D. Frederick, B.R. Gearey & N.J. Whitehouse

xx

Star Carr M. Dinnin & M. Welsh

xx

Thorne, Bradholme and Tudworth P.C. Buckland

xx

The Humberhead Peatlands N.J. Whitehouse, G. Boswijk & P.C. Buckland

xx

Thorne Moor G. Boswijk, N.J. Whitehouse & P.C. Buckland

xx

Hatfield Moor N.J. Whitehouse, P.C. Buckland, G. Boswijk & B.M. Smith

xx

Lindholme Island N.J. Whitehouse, P.C. Buckland, P. Wagner & B.M. Smith

xx

The Ontogeny of Thorne and Hatfield Moors N.J. Whitehouse, P.C. Buckland, G. Boswijk & B.M. Smith

xx

References

xx

The Quaternary of East Yorkshire and North Lincolnshire Editors Mark D. Bateman, Department of Geography, University of Sheffield, Winter Street, Sheffield S10 2TN. Paul C. Buckland, Archaeology & Prehistory Department, University of Sheffield, Northgate House, West Street, Sheffield S1 4ET. Charles D. Frederick, Archaeology & Prehistory Department, University of Sheffield, Northgate House, West Street, Sheffield S1 4ET. Nicki J. Whitehouse, School of Archaeology and Palaeoecology, Queens University of Belfast, Malone Road, Belfast BT7 1NN. List of Contributors

Gretel Boswijk, School of Environmental and Marine Sciences, University of Auckland, New Zealand. Charlie Bristow, School of Earth Sciences, Birkbeck College, University of London, Malet Street, London WC1E 7HX. Richard Carpenter, School of Geography & Archaeology, University of Exeter, Amory Building, Rennes Drive, Exeter EX4 4RJ. John A. Catt, Geography Department, University College London, 26 Bedford Way, London WClH OAP. Chris D. Clark, Department of Geography, University of Sheffield, Winter Street, Sheffield S10 2TN. Siwan Davies, Department of Geography, Royal Holloway, University of London, Egham, Surrey TW20 0EX. Mark Dinnin, School of Geography & Archaeology, University of Exeter, Amory Building, Rennes Drive, Exeter EX4 4RJ. David J.A. Evans, Department of Geography & Topographic Science, University of Glasgow, University Avenue, Glasgow G12 8QQ. Geoff D. Gaunt, 10 Foxhill Crescent, Weetwood, Leeds LS16 5PD. Ben R. Gearey, Wetland Archaeology and Environments Research Centre, University of Hull, Hull HU6 7RX. Andy J. Howard, School of Geography, University of Leeds, Leeds LS2 9JT. Jason R. Kirby, Department of Geographical Sciences, University of Plymouth, Drake Circus, Plymouth, Devon PL4 8AA. Malcolm C. Lillie, Wetland Archaeology and Environments Research Centre, University of Hull, Hull HU6 7RX. Ian Livingstone, School of Environmental Science, Nene College of Higher Education, Northampton NN2 7AL.

vi Contributors

Location map vii

BRIDLINGTON

N

Gembling

0

Routh

u Ri v e r H

de

us e

ol

ll

rn

n Do

er

Wolds

ent e r Tr

ire

Riv

s

sh

GAINSBOROUGH

GRIMSBY

ln

o lme

Day 3

Gingley-on-the-hill

Black Walk Nook

r

Star Carr

A nch

Day 1

co

Armthorpe

DONCASTER

Twigmoor Woods

Dimlington

in

Hatfield Moors

Cove Farm

Day 2

Yarborough SCUNTHORPE

Riv e r

R

Ferriby

L

Thorne Moors

be

Winteringham

er H

um

Riv

es

KINGSTONU P O N - HULL

iv

10 km

H

rO

s Wold

River Derwen t

YORK

R i ve

Speeton Sewerby Fimber

Yo r kshire

Julian B. Murton, School of Chemistry, Physics and Environmental Science, University of Sussex, Brighton BN1 9QJ. Bernard Owens, Centre for Palynology, University of Sheffield, Dainton Building, University of Sheffield, Brook Hill, Sheffield S3 7HF. Brian Smith, 139 Beccles Road, Oulton Broad, Lowestoft. Stephen A. Thomson, Department of Geography & Topographic Science, University of Glasgow, University Avenue, Glasgow G12 8QQ. John C. Tweddle, Royal Botanic Gardens, Kew, Wakehurst Place, Ardingly, West Sussex RH17 6TN. Matthew Welsh, 20 Leigh Drive, Wickham Bishops, Witham, Essex. Pat Wagner, Archaeology & Prehistory Department, University of Sheffield, Northgate House, West Street, Sheffield S1 4ET.

The region of this guide is covered at a scale of 1:50,000 by Ordnance Survey Landranger Series of Great Britain map sheets 101 (Scarborough), 105 (York), 106 (Market Weighton), 107 (Kingston upon Hull), 111 (Sheffield & Doncaster), 112 (Scunthorpe), and 113 (Grimsby). The region is also covered at 1:50,000 by British Geological Survey published sheets of 55/65 (Bridlington & Flamborough), 71 (Selby), 73 (Hornsea), 79 (Goole), 80 (Kington upon Hull), 81 & 82 (Patrington and Spurn Head), 88 (Doncaster), and 89 (Brigg).

viii Acknowledgements

Preface ix

Acknowledgements

Preface

The editors wish to thank all those who have assisted with the preparation for the field meeting and the production of this guide. Particular gratitude is extended to all the land-owners who have permitted access to their property. Financial Assistance for the work contained with this volume has come from a diverse range of sources including the Sheffield International Centre for Drylands Research, the English Heritage funded Humber Wetlands Survey project, NERC Scientific Services, The University of Sussex, and Edinburgh Oil and Gas. Thanks should be extended to both Paul Coles and Graham Allsopp, Cartographic Services, Department of Geography, University of Sheffield for drawing many of the diagrams and for the typesetting of this volume.

In many senses it seems proper at the start of the 21st Century to revisit the region in which, so we are reliably told, saw on a Yorkshire Geological Society field visit in 1962 the muting of the idea for a dedicated Quaternary group. Eventually from this sprang the Quaternary Field Studies Group and subsequently the Quaternary Research Association. Nearly thirty years have passed since the QRA last ran a joint East Yorkshire and North Lincolnshire field meeting and produced a fieldguide (Penny et.al. 1972). Since this time various parts of East Yorkshire have been revisited on a number of occasions by INQUA (Catt 1977) and the QRA (Ellis 1987; Bridgland et.al. 1999). It could be argued that this part of the UK has therefore been done to death, but we would contend this is not so. Not all sites in the aforementioned guides were visited as part of their respective field visits, and the ephemeral (and tantalising) nature of the exposures, their importance in terms of the Late Quaternary of Britain and the contentions associated with many of the sites warrant a further airing. As with all regions, the application of new techniques and the availability of new sites have allowed more information to be gathered, questions to be answered and new theories to be postulated. In addition, the rapidly eroding Holderness coast, lowering watertables and quarry activity mean that sites are being lost, are rapidly degrading, or are only open for a short period of time. As we write this, the 2001 Foot-and-Mouth outbreak is still disrupting site access, and has curtailed some of the research that we had hoped to present. Thus the exact itinerary of the field meeting may not exactly follow that outlined in the guide. The holistic approach of this guide presents research from both glacial and ice marginal sites, as well as significant sections of Lateglacial and Holocene sediments. Inevitably there are divergences of interpretation but we hope that this field guide and the field meeting will aid in participants critical evaluation of the data available from the region. Mark Bateman and Paul Buckland Sheffield, June 2001

x

The Glacial History of East Yorkshire 1

Introduction to the Late Quaternary of East Yorkshire and North Lincolnshire The Glacial History of East Yorkshire David J.A. Evans, Stephen A. Thomson & Chris D. Clark The Holderness region of East Yorkshire covers an area of approximately 750 km2, with the highest topographical point only 38 m OD at Dimlington High Land (TA 391217). Most of the topographical expression is the product of glacial deposition during the Last Glacial Maximum (Dimlington Stadial), with glacier ice moving into the region from ice divides in southern Scotland, Northumberland, and the Pennines. This topography comprises an undulating rolling landscape with prominent positive relief features around the villages of Tunstall, Holmpton, and Aldbrough in the south east of Holderness. A series of discontinuous, north-south trending ridges are evident on a simulated solar shaded digital elevation model (Figs. 1a, 1b) and aerial photography, although their subtlety makes them almost impossible to view from the ground, explaining earlier descriptions as ‘hummocky or chaotic’. North of Skipsea and inland towards the Yorkshire Wolds, the topography is dominated by a series of very subtle cross-cutting north-to-south and east-to-west ridges. These are often less than 10 m in elevation and as such are not evident to the west of the 10 m contour line on the 1:50,000 Landranger maps. Regional Glacial geomorphology The limits of Late Devensian glaciation and possible ice-marginal stillstands/readvances have been identified in East Yorkshire based upon a combination of glacial geomorphology and stratigraphy (Fig. 2). The erratics of the tills suggest that the lowlands of the east Yorkshire coast and the Vale of York were invaded by glacier ice flowing from the Tees drainage basin via Stainmore Gap, east Scotland, and the Pennine Dales. The ice probably extended on to the North Sea shelf, where it coalesced with Scandinavian ice (Carr 1999). To the west of Holderness and the Yorkshire Wolds, the regional ice advanced as a valley glacier to deposit three moraines/ice-marginal glaciofluvial landforms in the Vale of York, specifically the York moraine, the Escrick moraine and the Linton-Stutton gravels. The York and Escrick moraines continue onto the flanks of the Howardian Hills to the east and continue northwards to the Coxwold Gap where the Ampleforth (Escrick) moraine blocks the west end of the Vale of Pickering. Gaunt (1974, 1981) suggests that the Vale of York lobe initially advanced as a surge into Glacial Lake Humber to Wroot in Lincolnshire some time after 21.8 ka BP. The margin later stabilised at the York and Escrick moraines when the water level of Lake Humber dropped from 33 m to 8 m OD (See Bateman and Buckland, this volume). Interdigitation of lake sediments (the "25 foot drift") and the till and gravels of the Escrick Moraine indicate that the

2 David J.A. Evans, Stephen A. Thomson & Chris D. Clark

The Glacial History of East Yorkshire 3 00

90

10

30

20

40

?

N

?

70 Sewerby Bridlington

Kilham

0

5

10

Burton Agnes

km

N o r t h S e a

Great Kelk Barmston Driffield

60

Gembling

Foston on the Wolds Beeford

50

Hempholme Hornsea Brandesbuton Sigglesthorne

Leconfield

Routh

40

Beverley

Wawne

Ringbrough

Sutton

Bilton

30

Willerby Ridgmont Kingston Upon Hull

Holmpton

Ryhill

Keyingham

Patrington

Dimlington

20

Ridges mapped from DTM Skipsea till Withernsea till Buried Cliff Places mentioned in text

10

Figure 1b: Ice-transverse ridges mapped from the DEM. Western limits of Skipsea and Withernsea tills and location of Ipswichian interglacial cliffline also shown. Figure 1a: Simulated solar shaded DEM of Holderness, showing north-south orientated ridges. The subtle topography of the area is enhanced by simulated solar shading from the northeast. Based on Ordnance Survey 50m resolution DEM (reproduced by kind permission of Ordnance Survey. © Crown Copyright NC/01/476).

moraine was deposited at an ice-contact lake margin (Gaunt 1970). On the East Yorkshire coast of Holderness, a series of ice margins have been reconstructed by various researchers based upon till sheets and moraine assemblages. The oldest glacial deposit recognised in this area is the Basement Till, which is overlain by the Skipsea ("Drab") and Withernsea ("Purple") tills

along an extensive stretch of the Holderness coastline (Madgett 1975; Madgett & Catt 1978). Radiocarbon dates of 18,500 ± 400 14C yrs. BP and 18,240 ± 250 14 C yrs. BP obtained by Penny et al. (1969) on plant remains between the Basement and Skipsea Tills provide a maximum age for the onset of the Dimlington Stadial in the region (Rose 1985). An additional date for the onset of the Stadial of 17,500 ± 1,600 calendar years BP was obtained by thermoluminescence techniques from beneath the Skipsea Till on the Wolds dip slope (Wintle and Catt 1985). Peacock (1997) has argued that the Skipsea till represents a readvance by

4 David J.A. Evans, Stephen A. Thomson & Chris D. Clark

The Glacial History of East Yorkshire 5

N

0

20 km

Ri

ve

r

Hu

mb

er

Meltwater channels Esker Moraine Glacial lake Glacial lake deposit Figure 2: Map of glacial depositional features used in the reconstruction of the last glaciation of East Yorkshire (from British Ice Sheet GIS in preparation).

North Sea ice as late as 15-14ka radiocarbon years BP. McCabe et al. (1998) however proposed a ‘probable’ depositional time-frame for the Late Devensian tills of Holderness at 13,785 ± 115 to 13,955 ± 115 14C yrs. BP (Killard Point Stadial). A radiocarbon date of 13,045 ± 270 14C yrs. BP (Birm.317) on organics from a kettle hole at Roos provides a minimum for deglaciation (Beckett 1981). Rose (1985) and Wintle & Catt (1985) question this date due to the possibility of hard water errors. A variety of evidence, poorly constrained by absolute dating, has been used to propose limits of an early Devensian glacial advance and two later readvances. The westernmost limit of the drift on Holderness has traditionally been used as the limit of a Devensian glaciation. An early Devensian age for this has been supported by the relatively more subdued drifts west of the River Hull (e.g. Straw 1979), even though only two post-Basement tills have been identified in the area. Valentin (1957) traced a line connecting the outcrops of the Kelsey Hill Gravels at Burton Agnes, Brandesburton, Bilton and Patrington out to the North Sea, which he regarded as a product of the "second readvance"; this line also encloses the extent of the Withernsea (Purple) Till. Outside of this limit the margins of the "first readvance" are marked by subdued morainic ridges at Routh, Wawne and Sutton (Straw 1979).

To the north of Holderness, a prominent hummocky moraine on Flamborough Head was identified by Farrington and Mitchell (1951) and interpreted as an end moraine. This structure widens and stretches northwards to Speeton and then to Cayton where it has been called the Speeton moraine by Valentin (1957), who allocated it to a "second readvance" on Holderness. Specifically, the prominent morainic relief was the result of the compression of ridges on the Chalk scarp and the overriding of "first readvance" deposits. This was later termed the "Cayton-Speeton Stage" by Straw (1979). The Flamborough Moraine to the south was allocated to the "Seamer-Flamborough Stage" by Penny and Rawson (1969) who referred to the linearity in the moraine at Reighton as possible drumlins. At the east end of the Vale of Pickering, the preglacial drainage of the ancestral River Derwent was reversed once hummocky drift plugged the valley mouth south of Scarborough. A drift limit at 183 m just west of Scarborough (Straw 1979) provides an estimate of glacier thickness at the southern flanks of the North Yorkshire Moors/Tabular Hills. During glacier occupancy of the coast, glacial lake Pickering (first suggested by Kendall 1902) filled the Vale of Pickering and produced shorelines at 70m and 45m (Straw 1979). The 70m level is associated with a kame terrace between West Ayton and Wykeham, referred to as the Wykeham moraine and representative of the ice margin during the "Wykeham Stage" (Penny and Rawson 1969). An outwash fan at Seamer, fed by the Mere Valley, graded to the lower Lake Pickering level of 45m when ice then stood at the Cayton-Speeton Stage limit (Penny and Rawson 1969; Straw 1979). The readvances of Holderness and the east end of the Vale of Pickering were not recognised by Catt (1987a) or more recent workers because the limits were based largely upon freshness of hummocky topography (e.g. hummocks on east Holderness) and were not verified by stratigraphy or by dating evidence. Moreover, the smoother topography of the tills in west Holderness is a result of Holocene flooding and draping of lower altitude hummocks rather than an age difference compared to the slightly more hummocky drift east of the River Hull. The westernmost limits of the North Sea ice in Holderness during the LGM are presently drawn at the ‘feather edge of the Skipsea Till’ (previously mapped as Hessle Till; e.g. Suggate & West 1959). Additionally, in the Vale of Pickering a more extensive western ice limit was suggested by Foster (1985, 1987) based on the distribution of the Sherburn Sands, which represent an outwash train stretching from Flotmanby to Hovingham. Although readvances cannot be supported unequivocally based upon the evidence outlined above, the depositional landforms such as Valentin’s line (based upon outcrops of the Kelsey Hill gravels), the Wykeham moraine and the Cayton-Speeton moraine must represent a significant response by the glacier margin to either climatic signals or changes in regional ice sheet dynamics. With respect to the latter, Eyles et al. (1994) employed the evidence of the Holderness hummocky moraine and the feather edge of Withernsea Till in conjunction with the Kelsey Hill gravels to support their reconstruction of a surging ice sheet margin in East Yorkshire. The surging concept was first entertained by Lamplugh and later supported by

6 David J.A. Evans, Stephen A. Thomson & Chris D. Clark Boulton et al. (1977). Additionally, the low ice profile of 1:750, as represented by the east coast drift limit from North Norfolk to the North Yorkshire Moors, was regarded as possible evidence of a surging lobe by Straw (1979). Glacial stratigraphy of Holderness More than a century of research on the Quaternary stratigraphy of Holderness has resulted in a complex nomenclature (Table 1). Details on the sediments exposed at Sewerby, Dimlington, Skipsea and Kelsey Hill and environs provide an overview of the depositional sequence relating to glaciation of the Holderness region. The oldest firmly dated sediment in Holderness is the Ipswichian Interglacial beach gravel at Sewerby (Catt, this volume). This is overlain by sediments which document the drop in sea level and decline in temperatures at the beginning of the last (Devensian) glacial cycle and then the advance of glacier ice into the area during the Dimlington Stadial (Catt and Penny 1966; Catt 1987a,b). Of further importance with respect to long term sea level changes, is the buried Chalk cliff and associated marine platform (Fig. 1). This has been traced from Sewerby, where the modern cliff intersects it, inland to Driffield and then southwards to the Humber estuary (Crofts 1906; Crofts and Kendall in Kendall and Wroot 1924; Newton 1925; Valentin 1957). Some uncertainty surrounds the stratigraphic relationship between the interglacial beach deposits at Sewerby and the Basement till (Catt, this volume). Examination of over 600 borehole records indicates that the Basement till does not extend as far inland as the overlying Skipsea till (Fig. 3) as is >30m thick at Hornsea but thins towards the Wolds (Catt and Penny 1966). The limit of the Basement till lies along the Ordnance Survey 510000m line, although some outcrops extend towards the OS 505000m line around Cottingham and Leconfield. The Scottish and Scandinavian erratics with inclusions of ‘glauconitic sand’ (Carruthers 1948), most likely derived from the Bridlington Crag ‘erratic mass’ (Catt & Penny 1966; Catt 1987a) coupled with Chalk stringers, folding and shearing (Eyles et al. 1994) and clast lags in the upper 0.5 m of the massive clay matrix suggest a subsole deformation was responsible for the Basement till (Eyles et al. 1994), reworking marine sediments and incorporating rafts of older material. The rapidly eroding cliffs at Dimlington contain valuable stratigraphic evidence for the dating of the last glaciation of the British Isles, particularly the attainment of glacial maximum positions in lowland England. The stratigraphy briefly comprises Basement Till overlain by the Dimlington Silts with enclosed organic remains and two Late Devensian tills named the Skipsea and Withernsea tills (Catt, this volume; Catt and Penny 1966; Penny et al. 1969; Catt 1987c; Skipsea and Withernsea Members of Lewis 1999). Radiocarbon dates on the organic remains in the Dimlington Silts provide a maximum age of 18.2-18.5ka BP for the overriding of the site by the Late Devensian North Sea lobe of the British Ice Sheet, which deposited the Skipsea and Withernsea tills. This has resulted in the adoption of Dimlington as the climatostratigraphic type site and the use of the term Dimlington Stadial for the Late Devensian glacial episode in

The Glacial History of East Yorkshire 7 Guide

Wood & Rome (1868)

A

Freshwater marl; sand and gravel (+ chalk)

Sands, water and gravel

Hessle (boulder) Clay

Hessle (red clay + blue facies)

B C

Lamplugh (1879)

Dakyns (1879)

Reid (1885)

Alluvial

Raistrick (1929)

Lamplugh (1891) Post glacial + Alliuvial

Hessle

Red boulder clay

Hessle

Hessle clay

Hessle sand (or gravel) Patches of gravel

Sewerby gravels

Stratified gravels

Sewerby gravels

Purple Clay (without chalk

Brown Till

Purple boulder clay

Upper purple

Upper till

Gravels

Sand and gravels

Red band gravels

Intermediate stratified series (of Flamborough Head) + interior

Lower purple

Lower purple

Lower purple

Laminated clay of Bridlington

Sand (attributed to Basement)

Basement clay

Upper Basement Lower Basement

D E

F

G

Purple Clay (with chalk) & Bridlington Crag

Greenish-purple boulder clay

Purple bolder clay (lower)

H

Sands & Gravel (+ occasional clays)

Gravels, sands & silt (Laminated beds)

Sand

I

Bridlington Crag (sands and sandy clay) in Basement till

J

Basement clays of Holderness

Greenish-blue clay with shells

Gutta Percha clay ? Green boulder clay

Basement

Purple

Chalky rubble Blown sand kland wash Raised beach

K

L Bisat (1932)

A

Upper Hessle

Hessle (inland)

Lower Hessle (coast)

Upper Purple Clays (coast) - 2 beds

Weathered soil

B

Gravels (red band)

Late Glacial raised Sewerby gravels beach

C

Bisat (1939; 1940)

Mitchell et al. (1973) Edwards (1981) Eyles et al. (1994) Evans et al. (1995)

Guide

Flamborough Member Hornsea Member

D E

Lewis (1999) Fenland Formation

Purple

F Drab & Grey G

Lower Purple clays - 3 beds

Upper till

Upper till series

Withernsea till

Sand, silt and gravels - thin drab clays

Gravels

Gravels

Silts, sands and gravels

Canals (or) pipe fills

Withernsea Member Mill Hill Member

Upper Drab; Middle Drab; Kelsey Hill gravels Kirmington Silts Chalk rafts Lower Drab Clay Sub-Drab clay; Basement Drab clay

Lower till

Lower till series

Skipsea till

Skipsea till Formation

Lower purple

Sand (attributed to Basement)

Dimlington silts

H

Dimlington Bed

I J

Basement

Basement Sub Basement

Chalk rubble Basement till

Chalk rubble Basement till Chalk rubble

Basement till

Buried cliff beds (Sewerby)

K L

Basement till

Speeton shell bed

Bridlington Member Sewerby Member

Speeton shell bed

Table 1: Nomenclature for the Quaternary Stratigraphy of Holderness.

Britain (Rose 1985). The time period 26,000 to 13,000 years BP is similarly referred to as the Dimlington Chronozone. Extensive deposits of silts, sands and gravels occur beneath the Skipsea till at Sigglesthorne, Kilham, Hull, Beverley, Foston-on-the-Wolds and Driffield (Fig. 3). It is conceivable that the pitted depressions in the Basement till could contain similar material to the Dimlington Silts, thus inferring that these fine to coarse sediments are regionally extensive and not limited to Dimlington. The Skipsea Till and overlying Withernsea Till are clearly differentiated based

8 David J.A. Evans, Stephen A. Thomson & Chris D. Clark

05

70 00

10

15

20

25

The Glacial History of East Yorkshire 9

30

N Bridlington 65

60

Driffield 55

Scale 1km

50

Hornsea

Log vertical axis scale

1km

45

40

5m 10m 15m 20m 25m 30m

35

30

Dimlington 00

05

10

15

25

30

Silt, sand & gravel Withernsea Till Skipsea Till Basement Till

Figure 3: Fence diagram of Quaternary sediments on Holderness, compiled from BGS borehole records.

upon colour, erratic content, grain size, and mineralogy (Madgett and Catt 1978; Catt, this volume). They are also separated by a sharp erosional contact along which numerous discontinuous lenses of sands and silts and occasionally gravels lie in shallow scours on the surface of the Skipsea Till. Variable degrees of glaciotectonic disturbance are evident within this intrabeds. The spatial distribution of the Withernsea Till is far more restricted (Valentin 1957; Madgett 1975; Straw 1979; Catt 1991), although it is thought to extend some distance offshore (Cameron et al. 1987), forming part of the Bolders Bank Formation (BBF; Balson & Jeffery 1991; Cameron et al. 1992). Donovan (1973) and Cameron et al. (1987) state that the till offshore is a mixture of Skipsea, Withernsea, Basement and Hunstanton till. Catt (1991) suggests that no one till is dominant while Carr (1999) correlates the BBF with the Skipsea Till. The Withernsea till extends a maximum distance of 15 km inland from the present coast but its landward margin trends offshore at Easington in the south and at Mappleton in the north (Catt and Penny 1966; Madgett 1975). Pebble fabrics taken from the Skipsea and Withernsea tills by Penny and Catt (1967) reveal a consistent northeast to southwest ice flow direction. The textural and structural complexities of the Skipsea and Withernsea tills led Bisat (1939, 1940) to

propose their detailed stratigraphic subdivision. More recent research has settled for a simple two-tier stratigraphy (Catt and Penny 1966; Madgett and Catt 1978; Catt 1987a; Eyles et al. 1994; Evans et al. 1995) and deposition by 'two-tier' glacier (Catt and Penny 1968; Madgett and Catt 1978; Edwards 1981). In this, the Skipsea Till was deposited by ice originating in Northumberland and southern Scotland and flowing southwards along the Yorkshire coast. The Withernsea Till was deposited by a Tees valley ice stream which overrode the Skipsea Till ice and was then carried piggyback style southwards to the Holderness area. An alternative to this "piggyback" or simultaneous deposition model has been postulated by Eyles et al. (1994) based on the stratified sediment bodies that occur between and within the Skipsea and Withernsea tills. They interpreted these as marine formed between onshore surges by the North Sea glacier lobe. Thus the Basement, Skipsea and Withernsea tills are all regarded as the products of repeated onshore surging or regular (steady state) subglacial deformation during the Devensian glaciation. (Eyles et al. 1994; Evans et al. 1995). The sand and gravel lenses/intrabeds between the Skipsea and Withernsea till have been related to efficient drainage along an extensive subglacial linked cavity system which created a complex canal network that was infilled with massive, planar and cross bedded sands and gravels (e.g. Evans et al. 1995; Benn & Evans 1996). Borehole records indicate that the intrabeds/canal fills pinch and swell over the space of tens to hundreds of metres but did not supply the glaciofluvial sediments covering large areas of interior Holderness. The intrabeds (perhaps more correctly interbeds) at the junction of the Skipsea and the Withernsea till probably supplied the sediment for the Kelsey Hill gravels, but cannot account for the production of the proglacial outwash deposits north of Sproatley. The intrabeds observed within the subglacial diamictons are interpreted as the product of canal sedimentation during active glacier advance (Evans et al. 1995), and as such were not active during the Dimlington Stadial retreat. The proglacial outwash deposits, as discussed more specifically in the Gembling section (Thomson & Evans, this volume), must have been supplied by meltwater flowing at the ice-bed interface, either as sheet flow or along Nye or river channels before being evacuated from ice marginal portals into icemarginal lakes. The "Kelsey Hill Gravels" (Mill Hill Member, Lewis 1999) constitute an important Quaternary sediment/landform assemblage deposited at the margin of the Dimlington Stadial ice sheet in East Yorkshire. The gravels form a low sinuous ridge up to 15m high, widening southwards and linking the villages of Ridgmont, Thorngumbald, Ryhill and Keyingham. Eyles et al. (1994) have suggested that the Kelsey Hill Gravels actually form a series of north-south trending parallel ridges, and that Kelsey Hill constitutes the westernmost of these ridges. The rich fauna contained within the gravels (Prestwich 1861; Penny 1963) has been at the centre of considerable controversy over the origin of the landform-sediment assemblage (cf Geikie 1877; Reid 1885; Sheppard 1895; Sheppard and Stather 1907; Bisat 1940; Carruthers 1948; Baden-Powell 1956).

10 David J.A. Evans, Stephen A. Thomson & Chris D. Clark The stratigraphic position of the Kelsey Hill Gravels, between the Dimlington Stadial age Skipsea and Withernsea tills, suggests that sedimentation was a product of glacier marginal fluctuations. However, interpretations of the exact depositional environment range from marine to glaciofluvial (Lamplugh 1925; Penny 1963; Catt and Penny 1966; Eyles et al. 1994). On the eastern flank of Kelsey Hill, the Withernsea Till possesses a sharp (erosional) contact with the Kelsey Hill Gravels (Fig. 2), the latter having been glaciotectonised and incorporated as small lenses by shearing into the base of the till. According to Eyles et al. (1994), this stratigraphic relationship between the Kelsey Hill Gravels and the Skipsea and Withernsea tills is evident also in exposures through sinuous ridges to the north at Brandesburton and Hempholme (Phemister 1922). Moreover, it is similar to the relationship between tills and intervening stratified sediments along the Holderness coast (see also Evans et al. 1995; Benn and Evans 1996). Catt and Penny (1966) used the sinuous nature of the northsouth trending ridge containing the Kelsey Hill Gravels to suggest that it is an esker, an interpretation supported by the north-south trending palaeocurrents recorded in the sediments. They further suggest that the widening and lowering of the esker form to the south represents the subaerial continuation of the esker where a marginal outwash fan emanated from the subglacial tunnel mouth. Sometime after its deposition the esker was covered by the Withernsea Till, explained by Catt and Penny (1966) as a product of the later stages of drainage of a tiered ice sheet within which the Withernsea Till was the last to melt-out. Early interpretations of the Kelsey Hill Gravels were considerably influenced by the dominance of temperate fossils, resulting in proposals for an interglacial status for the deposits. However, Catt and Penny (1966) used the diversity of the fossil assemblage, in addition to the water-worn nature of the vertebrate remains, in the Kelsey Hill Gravels to suggest that they were derived from preexisting deposits. This is verified to some extent by the fact that the deposit lies between two tills, but Devensian age tills on Holderness are rarely shelly, as would be the case if glaciers had reworked pre-existing shell-rich sediments. The only possible source for shells is the Bridlington Crag that contains an arctic assemblage rather than a temperate one. Furthermore, the well-preserved nature of the fossils is difficult to reconcile with the long travel distances normally associated with esker production. A further alternative origin is the glacial transport of rafts of Ipswichian sediments from the Humberside embayment (Berridge and Pattison 1994), an interpretation that explains the lobate arrangement of the Kelsey Hill Gravel outliers and similar glacifluvial gravel mounds to the south of the Humber. Similarly, Valentin (1957) interpreted the elongate ridge containing the Kelsey Hill Gravels to the north of the Humber as an ice margin, possibly documenting a readvance of the North Sea lobe during its recession from the Dimlington Stadial maximum. Overlying the Late Devensian tills is a sequence of silts, sands and gravels, termed the ‘Sewerby gravels’ by Dakyns (1879; 1880; Lamplugh 1884; 1887; Catt & Penny 1966) with respect to deposits found at Sewerby cliff. Although texturally and lithologically similar to the intrabed gravels, the Sewerby gravels

The Glacial History of East Yorkshire 11 have been traditionally interpreted as glaciofluvial outwash deposited during the retreat of the Dimlington Stadial glacier (Lamplugh 1884; Catt 1987b). Postglacial modification of the Sewerby gravels and similar sheet gravels at Mill Hill, Catwick and Barmston strongly suggest ice-wedge formation. Most recently, Eyles et al. (1994) proposed a marine raised beach origin for the Kelsey Hill Gravels and Sewerby Gravels at Bridlington, the latter originally regarded as proglacial outwash (Lamplugh 1884; 1887; Catt 1987a). The imbrication and openwork nature of the gravels together with their flat-lying nature are used to support a near shore, shallow beach face environment. Further support for this interpretation is: a) the north-south orientation of the ridges containing the Kelsey Hill Gravels, parallel to the trend of the buried interglacial chalk cliffline and the modern coast; and b) the prolific mixed marine and freshwater fossil assemblage. Eyles et al. (1994) go on to suggest that the North Sea Lobe surged into a shallow marine embayment, deforming and incorporating parts of the Kelsey Hill Gravels in the base of the Withernsea Till. However, the high sea level required for the marine flooding of Holderness during the last glaciation is difficult to reconcile with glacioeustatic and glacioisostatic trends indicating that the North Sea was dry land until approximately 6,500 yrs BP, well after ice had disappeared from the British landscape (Funnell and Pearson 1989; Brew et al. 1992; Funnell 1995; Lambeck 1995). In addition, the Ipswichian affinities of the marine mollusca and freshwater gastropods indicate that they could not have been living on the Holderness coast during the Devensian glaciation and therefore were probably derived. Nevertheless, the ice-damming of proglacial lake water on Holderness, into which the North Sea Lobe could have surged, is a palaeogeographic scenario that requires further testing. Thick rhythmite sequences with dropstones such as those at Barmston (Evans et al. 1995) clearly document the existence of at least small lakes on the freshly deglaciated terrain. Summary • The oldest glacial deposit on Holderness is the Basement Till, dating to a preIpswichian glacier advance that reached a line drawn approximately from Willerby to Great Kelk. • During the Late Devensian Dimlington Stadial, beginning some time after 18.2-18.5ka BP based upon sub-till organics at Dimlington, a North Sea lobe of the British Ice Sheet advanced beyond the Basement Till limit to a line immediately west of the Ipswichian interglacial cliff line. The ice margin also penetrated up the Humber to block the Humber Gap and dam the regional drainage to produce the high level (33 m OD) Glacial Lake Humber. The Vale of York glacier lobe initially advanced as far south as Wroot, perhaps during a surge, but later stabilised at the York and Escrick moraines which where deposited in contact with the waters of the later (8 m OD) phase of Lake Humber. Further north the ice occupied the Flamborough/Speeton moraine and penetrated to Flotmanby and Hovingham at the west end of the Vale of Pickering.

12 David J.A. Evans, Stephen A. Thomson & Chris D. Clark •







Numerous moraines and other ice-marginal deposits have been cited as evidence of glacier readvances or stillstands during recession from the Dimlington Stadial maximum. Although some have been questioned based upon apparent freshness of topography, former glacier margins are clearly recorded by features like the West Ayton kame terrace/Wykeham moraine. The former limit of the Dimlington Stadial North Sea glacier lobe on Holderness is demarcated by the edge of the Skipsea Till, derived from Northumberland and southern Scotland. Inside this limit the deglaciated terrain comprises numerous ice-transverse ridges many of which contain cross-stratified sands and gravels. The less extensive Withernsea Till records incursion by Tees Valley ice onto the southernmost tip of Holderness. The dual till sequence of southeast Holderness (Skipsea Till overlain by Withernsea Till) has been interpreted as the product of deposition by a twotiered glacier but there is no reason to discount an explanation involving repeated onshore surging by different ice streams within the North Sea lobe. The internal structures of the tills strongly suggest a deforming bed origin. Till intrabeds, interpreted as canal fills, suggest that periods of deformation till accretion were punctuated by phases of subglacial meltwater drainage over a soft, erodible substrate. Where the subglacial drainage networks exited from the glacier margin they prograded thick sequences of subaqeuous fan sediments (Gembling, Thomson & Evans, this volume). These fans coalesced along individual ice margins to produce "ice-transverse" ridges. The Kelsey Hill Gravels represent one such subglacial-proglacial meltwater system that was active between the deposition of the Skipsea and Withernsea tills. Although there is evidence of localized ponding on Holderness, for example the rhythmite sequences at Barmston and the subaqueous fans at Gembling and Kelsey Hill, there is no unequivocal evidence of glaciomarine or nearshore sedimentation in association with glacier recession in the region.

Non Glacial and Post Glacial History 13

Non Glacial and Post Glacial History Mark D. Bateman & Paul C. Buckland This contribution seeks to look at areas beyond the limits of the Late Devensian British icesheet and give an overview of post glacial geomorphological and sedimentological changes for the region as a whole. Periglacial Phase During the early and middle Devensian intensive periglacial activity took place forming widespread cryoturbation structures, e.g. the involuted Older River Gravel at Armthorpe (Buckland, this volume), and ice-wedge casts beneath Devensian lacustrine deposits near Darfield (Gaunt 1976, p.373). Ventifacts and remnants of cold-climate desert pavements from this period have also been reported from Aldborough (Gaunt 1970), east of Knaresborough (Gaunt 1976), in the Aire and Calder valleys (Edwards 1936; Bisat 1946) and at York (Gaunt 1970). They also can be collected from the surface of the Older River Gravel around Hatfield. Lake Humber There is little evidence for Devensian sediment deposition until Late Devensian ice blocked the Humber Gap from the North Sea (Gaunt 1994). In doing so it impounded the glacial outwash and water from the rivers Trent, Idle, Torne, Don, Aire, Ouse, Derwent and Ancholme creating a large pro-glacial Lake, termed Lake Humber (Gaunt et al. 1971; 1974; Kent 1980; Evans et al., this volume Fig.**2; Fig. **4). This seems to have overflowed at least intermittently via the Witham Gap at Lincoln. Two stages of Lake Humber have been proposed one at ca 30 m OD and another later stage at ca 8 m OD. Lake Humber rose initially to above 30 m OD (the '100 foot strandline' of Edwards 1937), an interpretation based on patchy reworked sediments, interpreted as littoral, and stoneless clay, as that reported at Kirmington Airport, (Gaunt et.al. 1992). The timing of this high lake stage is not well known, although it overlies the periglacial surfaces described above. The patchy nature of the deposits led Gaunt et al. (1992) to conclude that this high level lake stage was short lived. A bone fragment, found not in situ but thought to have come from the base or within lacustrine deposits near Brantingham, has been radiocarbon dated to 21,835 ± 1660 14C yrs BP (>24 cal 14C yrs BP)(Gaunt 1974). However, the only available date for when ice could have been blocking the Humber Gap is that obtain from the Dimlington silts of 18,240 ± 250 14C yrs. BP (ca 21,900 cal 14C yrs BP) (Penny et al. 1969). Recently a TL date of 22.67 ± 1.4 ka was obtained on sediment interpreted as high level Lake Humber from the Ancholme Valley near Caistor (Bateman et.al. 2000). Additionally a TL age from above, in sediments which had clearly undergone periglacial processes, indicates the Lake had receded by at least 17.67 ± 1.2 ka (ibid). High stage Lake Humber would therefore seem to have existed for a maximum of 4000 years

14 Mark D. Bateman & Paul C. Buckland between ca 22 - 18 ka, but in reality was probably far shorter lived. The sands and gravel outcrops at Thorne, Bradholme, Tudworth, Lindholme, Wroot and High Burnham in North Lincolnshire form a marked topographic feature sweeping round from NNE to SSE to East – West (Gaunt 1976b; 1987, Evans et al., this volume Fig. **1). These deposits contain an erratic assemblage dominated by Carboniferous sandstones and Magnesian Limestone, and their distribution and altitude led Gaunt (1976b) to interpret them as having been laid down by ice as it surged transiently into Lake Humber and then rapidly melted. When this occurred is not exactly known. Their stratigraphic position is above the periglacial surface on the Older River Gravel, and concordant with the high stage Lake Humber marginal sands and gravels. At both Tudworth and Lindholme (Whitehouse et.al., this volume), Lake Humber clay-silts appear to partially overlie the feature. The second stage of Lake Humber rose to a shoreline at ca 8 m OD, and is synonymous with the 25 foot Drift deposits (Gaunt 1994). It includes thick (up to 20 m) deposits of bluish grey to reddish brown finely laminated silt and clays which, in terms of volume is much the largest Quaternary deposit in the Vale of York (Gaunt 1987; Gaunt et.al. 1992). Its upper surface is completely flat and devoid of features. It is however slightly inclined, rising from 6 m OD SE of Doncaster to 14 m OD around York, where it partly laps against and partly underlies the Escrick and York Moraines (Gaunt 1987; Gaunt et.al. 1992). This reflects isostatic depression at the time of deposition. This lower lake stage appears not to have been wholly ice constrained, but to have continued due to morainic material blocking the Humber Gap (Frederick et.al., this volume). Lake Humber eventually disappeared, apparently as a result of sediment infilling rather than breach of the moraine in the Humber Gap (Gaunt 1981; 1994). The thickness and continuity of sediments suggest a much longer duration for this low Lake Humber stage (Gaunt 1987). Between the two lake levels there is a limited amount of evidence suggesting the lake emptied down to -4 m OD before partially refilling to ca 8 m OD (Gaunt 1987). Sections in gravel working between Finningley and Austerfield (SE 675976) show interbedded sands and clay-silts, probably littoral to both phases of the lake, containing two cryoturbated phases overlying Older River Gravel and beneath Lake Humber clay-silts. These would support Gaunt’s (1994) argument for a subaerial phase between the two lakes. A palaeosol, which developed on the lake deposits at West Moor, Armthorpe, north-east of Doncaster, has been dated to ca 11,100 ± 200 14C yrs. BP (13,010 ± 200 cal 14C yrs. BP; N-810), and provides a minimum age for the final disappearance of this lake (Gaunt et.al. 1971), although insect faunas from Armthorpe (Buckland, this volume) and Sandtoft (Buckland, unpubl.) indicate that drainage had taken place before 12,500 14C yrs. BP. At the latter site, a block of coal, 30 cm across was incorporated in the lake silts, which were overlain by sands and peat containing an insect fauna indicating conditions at least as warm as present day.

Non Glacial and Post Glacial History 15 Rivers During the early and middle Devensian Glacial Stage the rivers Aire, Don, Idle and Trent transversing the district incised wide valleys down to -19 m OD, in some cases through bedrock towards the Humber Gap (Gaunt 1987; 1994). After Lake Humber disappeared, rivers initiated courses across the abandoned plain, depositing levées adjacent to their braided channels (Gaunt et.al. 1971). These deposits are generally not more than 2.5 m thick, forming low ridges and mounds in a linear distribution, running eastwards across the northern part of Hatfield Chase (Gaunt 1994). Not infrequently thin Lateglacial peats with insect faunas indicating extensive Carex marsh occur beneath and within the sands. New evidence from Cove Farm near Haxey, north Lincolnshire (Whitehouse et.al., this volume) suggests that prior to this, there was a period of rapid fluvial incision and equally rapid infilling on the proto Idle before the establishment of single channel meandering rivers (Howard, this volume). The Cove Farm evidence shows that braided stream sedimentation continued at least until late in Pollen Zone II (Whitehouse et.al., this volume). However, by the beginning of the Holocene or possibly during the Loch Lomond Stadial, till and other deposits impeding drainage through the Humber Gap moraine had been breached. This allowed rivers to once again deeply incised their courses down to nearly -20 m below OD in response to the continuing low sea levels (Gaunt 1994). Subsequently sea levels in the Humber Estuary rose again (Kirby, this volume) and between 7,000 to 6,000 14C yrs. BP rapid sea level rise caused river channel aggradation to occur. Associated with this, the rising water table caused much of the low-lying river floodplains and clay soils became increasingly waterlogged and peat deposits began to form in the more low-lying areas (Gaunt 1994). In the former bed of the Hampole Beck at Sutton Common, north of Doncaster, deposition had begun by ca 6300-5200 BP (Lillie 1997), although dates from elsewhere in the Levels would appear to indicate that the main phase of wetland initiation, for example, on Thorne and Hatfield Moors, occurred at about 5,500-5,300 cal BP (Whitehouse et al., this volume). By about ca 3,500 14C yrs. BP, sea-level was around OD, and the river channels had been largely infilled; floodwaters spread beyond the channel confines, causing overbank alluviation and the landscape developed into a complex of wetlands ranging from riverine fens to raised mires (Gaunt 1981; 1994; Dinnin 1997a-d). Buckland and Sadler (1985) note a phase of late Roman alluviation at Sandtoft, which appears to relate to changes in land use, leading to extensive loss of topsoil into the river valleys, and similar evidence from the Trent (Riley et al. 1995) suggests that these changes were taking place from the late second century onwards. Coversand The superficial sediments of much of North Lincolnshire and the southern part of the Vale of York comprise of cold-climate aeolian sands termed coversands. They form two of the four main coversand deposits in Britain, which

Non Glacial and Post Glacial History 17

16 Mark D. Bateman & Paul C. Buckland Hull

Cuesta scarp Coversand









k

▲ L OW ER





▲ ▲

▲ ▲ ▲ ▲







ho









▲ ▲

▲ ▲



e R i v e r Tr n t













e

O

S

To r n

C

IRE W O LD

nc er A lme

HAXEY

N

SH

BRIGG Riv

v

LI

LN

CO DGE O LN I NC NSSHH ER E LLI N EDEG E I RI

▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲

S

Ri

er

E



LI A

SCUNTHORPE

Isle of Axholme

TH

▲ ▲









▲ ▲ ▲ ▲

▲ ▲ ▲ ▲

Bec

on

▲ W

i nt

e rt

▲ ▲ ▲ ▲ ▲ ▲

N



5 km

SANDTOFT

m

BARTON



Maximum limit of Lake Humber

0

Hu

River



Maximum extent of late Devensian Ice

ber



▲▲▲

CAISTOR

Figure **4: Distribution of coversand deposits in Northern Lincolnshire (After Bateman 1998).

collectively lie at the north-western margin of the European sand sheet (Catt 1977; Bateman 1995a). The latter stretches east from northern France through to Poland and into Russia and Ukraine (Zeeberg 1998) as a spatially continuous flat to undulating sandsheet with a notable paucity of dunes (Koster 1982). Although ventifacts have been reported around Wakefield and Knottingley (Edwards 1936) and at Aldborough (Gaunt, 1970), these relate to an earlier aeolian phase. Beyond some work carried out by Matthews (1969; 1970) around Sutton on-the-Forest, very little has been reported about the coversands of Yorkshire. In contrast, North Lincolnshire arguably has the best exposures of British coversand due to past and present quarry activity around Scunthorpe. The Lincolnshire coversands extend from the Humber in the north to beyond Gainsborough in the south and from the Isle of Axholme in the west across to the Lincolnshire Wolds, an area of approximately 400 km2 (Fig. **4). Only localised patches of coversand have been reported from on top of the Wolds, e.g. Fonaby Top (Straw 1963), and except for around Brandesburton, coversands are entirely absent on the low-lying Holderness plain to the east of the Yorkshire Wolds. The fragmented surface distribution of coversands belies the fact that they probably formed as a near continuous sheet into which the rivers Trent and Ancholme have subsequently incised their courses and capped with alluvium. Coversand thickness varies from up to 7 m on the dip slope of the Lincolnshire Limestone to generally less than 2 m (Gaunt et.al. 1992). The sands

are usually pale yellow to brown (10YR 6/2 to 7/4) with a mean grain size of 2.09 phi (~217 µm), moderately well sorted (σ =0.67) with sub-round to round grains, few clasts and little silt or clay. Bedding is most visible on weathered sections, and varies from the thin (41,000 ± 200 BP (NPL-88) and >42,000 ± 200 BP (Birm. 217) respectively. Whilst the Ipswichian sites at Langham and Westfield Farm are still potentially accessible, the site at Austerfield is now deeply buried beneath domestic waste, although the deposit is known to extend further north (Gaunt 1994; Gaunt pers. comm.). A fourth Ipswichian site on the River Aire in the Wortley District of Leeds, was identified by the discovery in 1852 of Hippopotamus amphibius Linné remains in a brick pit (Harkness et al. 1977). Sadly, this site is now also lost.

The Regional Fluvial Record 23

Exposures LOIS Sites

1. Beilsbeck Farm

6. Wortley

11. Catterick

16. Tadcaster

2. Scrooby

7. Oxbow

12. Kettlewell

17. Beal

3. Austerfield

8. Cawood

13. Myton on Swale

18. Bole Ings

4. Langham

9. Tadcaster

14. Kirk Hammerton

5. Westfield Farm

10. Tiln

15. York

Figure **6: Main rivers of North Lincolnshire and Yorkshire

During the Late Devensian Dimlington Stadial, the regions rivers where highly disrupted by the North Sea ice lobe blocking the Humber Estuary resulting in the creation of proglacial Lake Humber (for details see Bateman and Buckland, this volume). The 25 Foot Drift, a sequence comprising sands overlain by silts and clays, were laid down in the low lake level stage. Its demise was probably through silting up rather than by drainage as the upper sands of the 25 Foot Drift form low discontinuous ridges and mounds. These are interpreted as levées deposited by the re-establishment of meandering rivers flowing across this newly exposed land surface (Gaunt et.al. 1971).

24 Andy J. Howard Holocene The Holocene fluvial record of the region is much more complete. During the 1990s, the geomorphology, geochemistry and hydrology of the rivers of the Yorkshire Ouse basin, particularly those draining from the Pennine uplands (the Swale, Ure, Nidd, Wharfe and Aire), were intensively studied as part of the Natural Environment Research Council’s ‘Land Ocean Interaction Study’ (Leeks and Walling 1999). The data from this study provide arguably the most complete record of basin-scale fluvial evolution in the UK, as well as providing datasets relating to more recent aspects of sedimentation and erosion (Walling et al. 1998; 1999; Lawler et al. 1999). By the early Holocene, high gradient upland and piedmont river systems were progressively incising their valley floors in response to declining sediment supply and isostatic recovery, leaving a series of river terraces above the contemporary channel (Howard et al. 1999; Macklin et al. 2000; Taylor et al. 2000). The vertical and lateral instability of rivers within the uplands and at the upland margin contrasts with that of the lowland and perimarine river landscape, where low gradient multi-channelled anastomosed systems were characterised by low rates of lateral movement and relatively high rates of finegrained sedimentation (Macklin et al. 2000; Taylor et al. 2000). Documentary and cartographic evidence indicates that many of these anastomosed rivers survived in the southern part of the County until large-scale drainage and land reclamation in the early 17th Century AD (Dinnin 1997a). Episodes of river incision and sedimentation, principally variations in flood frequency and magnitude have been correlated with periods of climate change, and particularly in the last 1000 years with human impact (Howard et al. 1999; Macklin et al. 2000). Since the Bronze Age, progressive deforestation, associated with pastoral and arable farming has probably enhanced both runoff and the supply of fine-grained sediments to river valley floors. Since the 11th Century AD and during the last 250 years, the base-metal mining industry of the Yorkshire Dales has been a major source of fine-grained sediments for rivers of the Ouse basin (Hudson-Edwards et al. 1999a). In the lower Trent Valley, the most important palaeoenvironmental record is from an abandoned channel of the Trent at Bole Ings (Dinnin 1997b; Dinnin and Brayshay 1999). Spanning the period ca 8300-2700 BP, pollen and insect remains indicate a heavily wooded floodplain dissected by laterally shifting multiple channels during the early Holocene. Through time, the river became stabilised and developed as a single channel or anastomosed system. The impact of human activity, particularly clearance increased through time, although the proxy record also hints at the possibility of sea level change affecting floodplain environments and processes.

Regional Late-Quaternary Marine and Perimarine Records 25

Regional Late-Quaternary Marine and Perimarine Records Jason R. Kirby Over the last 10 years or so there has been a proliferation of research investigating Holocene marine and estuarine environments in East Yorkshire and North Lincolnshire, and their interaction with the fluvial and catchment systems. The purpose of this paper is to summarise and integrate this new information into the framework established by earlier workers. This will be presented in two parts, with the marine record separate from the perimarine sequences. All radiocarbon ages have been calibrated using the CALIB 3 program of Stuiver & Reimer (1993) and are quoted as calibrated years before present (cal yrs BP). The marine record in the Humber Estuary The first study aimed at reconstructing regional sea-level change within the Humber was by Gaunt & Tooley (1974), who described a pattern of estuarine sedimentation spanning the last ca 7000 cal yrs BP, with the chronology based on material collected from a range of palaeoenvironments (including wood and herbaceous peat and intertidal molluscs). The oldest data come from basal peats at Market Place, Hull and Union Dock, Grimsby (Fig **7), and it is clear that sea level (RSL) fell during the late Devensian with fluvial incision through at least 22 m of sediment to a depth of ca -16.5 m OD (Gaunt and Tooley 1974). This paper provided the model for the estuary for many years, until further detailed sea-level studies were carried out (Long et al. 1998). Although no systematic sealevel studies were undertaken in the intervening period, various archaeological projects provided new information regarding sea-level change in the estuary, from Brigg in the Ancholme valley and Hasholme in the Foulness valley (Smith et al. 1981; Millet & McGrail 1987; McGrail 1990). Dinnin and Lillie (1995) and Long et al. (1998) provide comprehensive reviews of this data and present new information from sites in Holderness (Thirtle Bridge; Roos Drain; Halsham Carr; Kilnsea Warren), North Lincolnshire (Union Dock, Grimsby; Barrow Haven; Newton Marsh, Tetney), and the southern part of the Vale of York (Sandholme Lodge; East Clough). This shows a RSL rise from a minimum level of ca -11 m OD at 7500 cal yrs BP to present levels. Gaunt & Tooley (1974), Dinnin & Lillie (1995) and Long et al. (1998) suggest that by ca 5500 cal yrs BP, mean sea level (MSL) had risen to within a few metres of OD. Long et al. (1998) calculate that the rate of RSL rise decreased from 3.9 mm cal yr1 between ca 7500 and 4000 cal yrs BP, to an average of 1 mm cal yr-1 during the last 4000 cal yrs BP. Although the spatial and temporal coverage of sea-level index points was insufficient to allow estuary-wide patterns of coastal development to be clearly determined, the general pattern of Holocene coastal evolution was established. The most recent LOEPS data has removed many of the spatial and temporal gaps evident in the RSL record with wide coverage of sea-level index points

26 Jason R. Kirby

Regional Late-Quaternary Marine and Perimarine Records 27

Ri v e r

R i v e r De

Sw a

nt rwe

le

N

0

R

R ive F r

R r O ive

ou

u

2

Ousefleet Ousefle et

Don R iv e

HULL 9 13

16

14 15

6

8 4

Garthorpe

17 Brigg

r

lley

Torn e er le R i v iver Id R

me Va hol Anc

ent River W

DONCASTER DONCASTER

s

1 Goole

3

km

10 ln es

11

se

20

r Hull ive

YORK

Misterton

Waddingham

Hu 7

mb e

GRIMSBY 2

rE stu ar y

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Hook Lane Eskamhorn Farm Hirst Courtney Kilnsea Warren Newton Marsh, Tetney Thirtle Bridge Union Dock, Grimsby Barrow Haven Market Place, Hull Hasholme Sandholme Lodge Bole Ings East Clough (Melton) Roos Drain Halsham Carr Thorne Waterside Crowle

0

-5

-10

Inner - basal index points Inner - intercalated index points Outer - basal index points Outer - intercalated index points RSL model Limiting dates

-15 LOIS core sites (Shenan & Andrews 2000)

Thorne & Hatfield Moors

12

-20 0

Figure **7: Location map of sites showing evidence of marine or perimarine environments.

throughout the estuary and adjacent offshore areas (Metcalfe et al. 2000; Shennan et al. 2000a; 2000b). These new data show that by ca 8000 cal yrs BP, estuarine conditions existed in the outer estuary with sea level about 17 m below present (see Fig. **8). This is in agreement with predictions for North Sea palaeogeography by Shennan et al. (2000b). Relative sea level rose quickly between ca 8000 and 6000 cal yrs BP with many sites in the outer and middle estuary recording marine conditions (Gaunt & Tooley 1974; Long et al. 1998; Metcalfe et al. 2000). These changes had a knock on effect on the hydrology within the inner estuary, with impeded freshwater drainage ‘backing-up’ causing ponding, waterlogging and paludification. After ca 6000 cal yrs BP, the marine transgression progressed up into the lower valleys of the inner estuary and the rate of rise began to slow (Kirby 1999; Metcalfe et al. 2000). Important changes in palaeotidal regime also occurred during the early- to mid-Holocene in the Humber and east coast area, with spring tidal range only 63% of its present magnitude ca 8000 cal yrs BP (Plater et al. 2000; Shennan et al. 2000a; 2000b). Model predictions show that tidal range increased by 60 cm between ca 8000 and 6000 cal yrs BP, in response to variations in tidal prism and estuary configuration, with only minor changes occurring since (Shennan et al. 2000b). The spread of data points in Fig. **8 is large due to a combination of local and regional factors, with significant errors arising from problems of sediment compaction – especially with the intercalated data points. This means that small-scale fluctuations are less than the amplitude of the combined error margins and cannot be discerned within the sea-level envelope. However,

2

4

6

8

10

12

cal ka BP Figure **8 Holocene relative sea-level curve for the Humber estuary (Metcalfe et al., 2000; Shennan et al., 2000a). Error bars reflect vertical (e.g. indicative range; sediment compaction etc.) and horizontal uncertainties (2 sigma calibrated age range). Data are from the radiocarbon database of UK sea-level index points, provided by the Sea-Level Research Unit, Department of Geography, University of Durham, with kind permission of Professor Ian Shennan.

coverage of data (both stratigraphic and archaeological) for the certain periods is sufficient to indicate significant changes in estuary configuration, in particular between ca 4000 and 1900 cal yrs BP (Long et al. 1998). During this time, a period of positive sea-level tendency (increase in marine conditions) is followed by a phase of negative sea-level tendency (removal of marine conditions), which coincides with a period of global climate change and RSL fall in the Netherlands (van Geel et al. 1996). It is apparent that from ca 4000 cal yrs BP many sites were inundated as estuarine conditions expand to their maximum Holocene extent in the Humber estuary (Smith, 1958a; 1958b; Fletcher 1981, Smith et al. 1981; Long et al. 1998; Neumann 1998; Kirby 1999; Metcalfe et al. 2000). Boats and trackways of Bronze Age date has been found buried within this estuarine alluvium (Wright & Churchill 1965; Crowther 1987; Buckland et al. 1990; Long et al. 1998; Fletcher et al. 1999). The pattern of coastal development was first proposed by Smith (1958a; 1958b) in north Lincolnshire, based on archaeological evidence from the Ancholme valley dating from the late Bronze Age and early Iron Age. Subsequent work in the Ancholme valley (Fletcher 1981; Smith et al. 1981; Switsur 1981; Neumann 1998) provided radiocarbon dates to support this observation.

28 Jason R. Kirby Fletcher (1981) suggested that marine conditions reached their maximum extent at ca 2700 cal yrs BP in the Ancholme valley when estuarine sediments were accumulating 22 km inland at Waddingham Holmes. This trend is upheld by sea-level work from elsewhere in the estuary (Gaunt & Tooley 1974; Tooley 1978; Tooley 1990; Millet & McGrail 1987; Long et al. 1998; Kirby 1999; Metcalfe et al. 2000). In the Aire valley, the Holocene marine limit has been located to close to the Hirst Courtney, 15 km upstream from the confluence with the river Ouse. Evidence for a retraction of intertidal conditions and a period of shoreline advance dates from after ca 3000 cal yrs BP (Long et al. 1998; Kirby 1999; Metcalfe et al. 2000) with a number of coastal sites in the Humber (e.g. Hasholme and Brigg) recording a replacement by freshwater conditions (Smith et al. 1981; Jordan 1987). The development of surface peats in the Ancholme and Hull valley, and elsewhere in the estuary (e.g. Barrow Haven) are further evidence for estuarine contraction (Smith 1958a; Fletcher 1981; Dinnin & Lillie 1995; Long et al. 1998). This sequence of events is also supported by data from the perimarine lower river valley areas in the inner estuary, although stratigraphic data are limited due to the proximity of upper peats to the present day surface. This renders them most susceptible to drainage and damage from agricultural activities. Only isolated remains of this regressive peat are now evident in the Ancholme (Smith 1958a; Neumann 1998) and Aire valleys (Kirby 1999). Discerning a causal mechanism for these changes is difficult at present for several reasons. Any vertical fall in sea-level during this time cannot have been large, since no deviation from the average trend of late-Holocene RSL rise is apparent on the curve. Also, there are problems in linking these estuarine changes with climate signals due to the plateau in the radiocarbon calibration curve during this time (Pilcher 1991; van Geel et al. 1996), and the likely time lag between RSL change and the response of the sedimentary system (Long et al. 1998). Local factors (e.g. sediment supply, geomorphic thresholds, site exposure) also complicate any attempts at climate/sea level correlation. Local processes may be invoked to explain the contrasting nature and timing of coastal response in some Humber river valleys (Dinnin & Lillie 1995) and regional differences with the East Anglian Fenland, which shows an opposite pattern of coastal development at this time (Waller 1994a; Long et al. 1998). Lastly, these changes occur at a time when human activity within the Humber lowlands was intensifying. The impact of extensive catchment changes (and associated increased sediment yield) on patterns of sedimentation within the coastal and estuarine system due to anthropogenic causes has been discussed (Buckland & Sadler 1985), but requires further quantitative investigation. In particular, it is necessary to determine sediment provenance and develop mineral dating techniques to establish a link between terrestrial and marine sedimentation in the Humber estuary (Long et al. 1998).

Regional Late-Quaternary Marine and Perimarine Records 29 The perimarine record in the Inner Humber Estuary Tooley (1986) highlighted the largely untapped potential of sediments deposited in the perimarine zone for addressing a wide range of palaeoenvironmental research themes. Palaeoecological studies from floodplain sites in the lower reaches of tidal river systems can be used to address questions of regional importance such as sea-level change since the high groundwater levels required to maintain eutrophic fen carr vegetation systems that flourish in such coastal margins are controlled by sea level. During the early Holocene, the lower reaches of rivers such as the Ouse and Aire were entirely fluvial systems with relatively dry valley bottoms. Peat formation was uncommon in these areas at this time – presumably reflecting the lack of waterlogging due to low base levels. However, local site conditions were suitable for preservation of organic material within some of the rivers in the Humberhead Levels area (e.g. Went, Idle and Torne) during the early-Holocene (Dinnin 1997b; Lillie 1997b; Lillie & Gearey 1999). These deposits are likely to represent peat formation due to local conditions (e.g. waterlogging in isolated depressions in the pre-Holocene subsurface), as they were accumulating too early to be directly related to the regional influence of sea-level rise. Howard (this volume) suggests that during the early Holocene, river channels in the perimarine zone were characterised by multi-channelled anastomosed systems (Dinnin & Brayshay 1999; Macklin et al. 2000; Taylor et al. 2000). There is tentative lithostratigraphic evidence from a long profile borehole transect in the lower Aire valley (between Hook Lane and Eskamhorn (Kirby 1999) to support this conclusion (see Fig. **9). Minerogenic fluvial sediments are recorded at the base of many of the cores suggesting that the pre-peat environment was a valley bottom consisting of several channel systems. The majority of peat formation however, was time-transgressive up the river valleys and attributable to base-level change. As marine conditions encroached into the Humber estuary, groundwater levels began to rise in the lower valley areas and floodplain aggradation of predominantly organic material ensued. Radiocarbon dates from the base of these deposits show that paludification had begun as early as ca 9200 cal BP at Bole Ings in the lower Trent (Brayshay & Dinnin 1999) and the lower Ouse at Ousefleet (Metcalfe et al. 2000). Peat formation then spread elsewhere in the lower valley areas with dates of ca 8000 cal yrs BP in the lower Trent at Garthorpe (Ridgway et al. 2000), ca 7800 cal yrs BP at Hook Lane the lower Aire valley (Kirby 1999) and ca 7000 cal yrs BP in the Ancholme at Brigg (Neumann 1998). The variation in ages is probably due to depth of fluvial incision, and other local factors. Hence, later ages come from valley bottom peats at a higher position relative to OD (Dinnin 1997a). At Hasholme in the Foulness valley, peat was forming by ca 6500 cal yrs BP in response to raised water levels (Millet & McGrail 1987). Evidence from the Humberhead Levels indicates base levels had risen sufficiently to cause peat formation within river floodplains of the Don at Thorne Waterside and Crowle, and the river Idle at Misterton by ca 4500 cal BP (Buckland & Dolby 1973; Buckland & Sadler 1985; Smith 1985). Fen peat deposits also began to

30 Jason R. Kirby

Regional Late-Quaternary Marine and Perimarine Records 31

W

E

SE 3734 2318 RC 2 4

SE 7294 2581 6

8

10

R. 12 Aire

14 R. Aire

16

18

20

22

24

26

28

30

32

34

36

38

40

42

44

46

R. 48 Aire

50

52 R. Aire

54

NOT SAMPLED

-4

NOT SAMPLED

0 -2

NOT SAMPLED

2

NOT SAMPLED

Altitude (M.O.D.)

4

-6 -8 -10

0

400 m

?

Top soil / disturbed ground Orange - brown - grey silty clay Blue or medium - grey organic clay Brown - grey clayey wood peat

?

?

?

?

Grey laminated humic clays Grey strongly laminated sands, silts and clays Stiff blue - grey - pink sandy clay or clayey sand

Figure **9: Schematic lithostratigraphic of Lower Aire long profile borehole transect between Hook Lane and Eskamhorn (after Kirby 1999)

accumulate at Thorne and Hatfield Moors during this time prior to the onset of ombrotrophic conditions (Buckland 1979; Smith 1985; Buckland & Dinnin 1997; Whitehouse et al. this volume). These ages provide evidence for the first influence of Holocene sea-level rise in the inner estuary with impeded water draining seawards leading to waterlogging and organic floodplain aggradation. These freshwater peats rise in altitude and become younger in an upstream direction (Neumann 1998; Kirby 1999; Lillie & Gearey 1999). Such stratigraphic data can be used as limiting sealevel index points (see Fig. **9), because although they are freshwater samples, paludification was controlled by sea-level and as such, they constrain maximum altitudes for the highest tide levels at that time (Neumann 1998; Kirby 1999; Metcalfe et al. 2000; Shennan et al. 2000a). The lower Aire valley A database of over 100 borehole records has been used to reconstruct the floodplain depositional environments in the lower Aire valley (Kirby 1999), and pollen, wood macrofossil and diatom analysis from 3 sites (Hook Lane, Eskamhorn and Hirst Courtney - see Fig. **7) enable the nature and timing of floodplain vegetation change to be determined. The longitudinal profile shown in Fig. **9 characterises the range of sediment types comprising the river Aire perimarine valley fill deposits. Generally, a clayey wood peat is found overlying the pre-Holocene subsurface, which grades into clays of estuarine origin. The pattern of floodplain development in the lower Aire valley is summarised in Fig. **10. From the date of peat inception in the lowest valley site at Hook Lane, it is clear that peat inception due to waterlogging, led to the spread of fen carr wetland vegetation communities (dominated by Alnus glutinosa (alder)) on the valley bottom from ca 8000 cal yrs BP. Pollen analysis of the basal sediments and overlying peat show that the valley margins prior to paludification comprised a mixture of dryland tree taxa

dominated by Tilia (lime), with Quercus (oak), Ulmus (elm) and Corylus (hazel) important components. Apparent declines in Tilia pollen at the base of these sequences are a reflection of the initial waterlogging of the valley bottoms and subsequent marginalisation of dryland tree taxa, dominated by lime, due the expansion of wetland area and not due to anthropogenic causes (Waller 1994b). Diatom evidence indicates a continued encroachment of estuarine conditions into the upper reaches of the Humber system subsequent to the initial paludification. This resulted in the development of large tidal lagoons, which flooded the alder carr vegetation on the floodplain in the vicinity of Goole some time after ca 7500 cal yrs BP (boreholes RC 47-52 in Fig. **9). Creation of these floodbasins was probably the result of a combination of factors, including the relatively rapid rate of sea-level rise coupled with the backing-up of freshwater drainage – possibly impeded also by the constriction of riverine outflow due to widespread accumulation of peat in the lower valley areas at this time. Palaeoecological analysis shows these lagoons (termed ‘fluviolagoons’ by van der Woude 1983) to be similar in nature to comparable depositional environments studied from the Dutch lowlands during the early to mid Holocene (e.g. van der Woude 1983; Vos & van Heeringen 1997). In the vicinity of Goole in the lower Aire valley, a permanently submerged brackish tidal lagoon was surrounded by reedswamp communities, in areas of shallow water, with fen carr communities towards the valley margins (Kirby 1999). The fluviolagoon silted-up, probably due to an increase in tidal asymmetry, and range (Shennan et al. 2000b), which resulted in a net surplus of sediment into the floodbasins. After ca 6000 cal yrs BP, the tidal lagoon conditions were replaced by alder carr, which dominated the mid-Holocene period. Indeed, throughout the Humber lowlands in general, alder fen carr vegetation communities were extant in all the river valleys during this time (Smith 1958a; Long et al. 1998; Lillie 1997a; 1997b; 1998; Neumann 1998; Brayshay & Dinnin 1999; Kirby 1999; Lillie & Gearey 1999; 2000; Metcalfe et al. 2000). This was a period of relative ecological stability within riverine contexts, as thick uninterrupted wood peats accumulated – keeping pace with the rate of RSL rise. Dense fen carr vegetation dominated the backswamp floodplain areas, with pollen from of Alnus glutinosa, Salix and Corylus avellana-type particularly prominent. Lithostratigraphic investigations show these peats to contain a high and variable clay content, reflecting continued inwashing of tidally influenced river water and also perhaps high levels of oxidation of organic material within the carr environment. During the mid- to late Holocene period, the influence of estuarine expansion was again registered in the lower Aire valley, with all sites recording an increase in water levels associated with rising tide levels seawards from ca 5000 cal BP to ca 2800 cal BP (Fig. **10). Deposition of estuarine alluvium was widespread during this time with invasive tidal creek networks probably responsible for erosion of parts of the main floodplain peat (e.g. between RC 24 and 29 in Fig. **9). The effect of salinity is important in dictating the nature of the vegetation change up the valley in response to water level rise during this

Regional Late-Quaternary Marine and Perimarine Records 33

32 Jason R. Kirby HIRST COURTNEY

Silty clay

y ph ra tig rt a os Bi ?

Freshwater river marsh & mudflat Fen & reed swamp Open carr & fen meadow

Rising water table

Radiocarbon yrs. BP (uncorrected)

3000 Clayey wood peat

ESKAMHORN

t Li

st ho

y ph ra tig ra

HOOK LANE y ph ra tig rt a

os Bi

? Organic clay

2000

y ph ra tig rt a

? Saltmarsh & mudflat

os ith

L

?

Fen meadow

Alnus/Salix carr

Salix carr Rising water table

4000 4200

Organic clay

Alnus carr Clayey wood peat

Ulmus decline

Clayey wood peat

5400

7000

Supratidal saltmarsh

Saltmarsh, mudflat & expanding freshwater communities locally Saltmarsh & Rising mudflat water estuarine flooding table

3000

4000 4400 4700 5000

Alnus/Salix carr & saltmarsh nearby

5000

6000

y ph ra tig ra t os Bi ?

Alnus carr with some openings 6000

Radiocarbon CAL. yrs. BP

y ph ra tig tr a s o th Li ?

Ulmus decline Alnus carr

Alnus carr Humic clay with leaf laminations

Fluviolagoon with brackish influence

Wood peat

Alnus carr

7000 7500 7800 8000

Figure **10: Summary of the timing and nature of floodplain development and vegetation changes in the Lower Aire valley (Kirkby 1999).

time. At the lowest sites at Hook Lane and Eskamhorn, as the alder carr flooded and was pushed to higher parts of the valley sides, willow was able to codominate the increasingly wet local conditions before saltmarsh plants become evident, along with diatoms indicative of brackish intertidal environments (see Fig. **10). At Hirst Courtney, the uppermost site in the study reach, the effect of a rising water table is registered later and is manifested by encroaching tidal freshwater. Here, the fen carr community shifts in composition from alder to willow dominated and freshwater fen meadow conditions are abundant with Filipendula (meadowsweet) the main taxon. Freshwater reedswamp conditions replaced fen meadow as sediment laden waters flooded the site and freshwater intertidal (cf Palustrine, sensu Fletcher et al. 1993) conditions ensued. This marks approximately the Holocene marine limit in the lower Aire valley (Kirby 1999). Diatom evidence indicates a freshening and perhaps drying of the local environment at several sites after ca 2800 cal yrs BP. Indeed, there is lithological evidence for the remains of an upper peat unit in several cores in the valley (e.g. RC 25, 36 and 45, Fig. **9). These findings support the evidence for a contraction of estuarine conditions during the Bronze Age/Iron Age transition, with negative sea-level tendencies recorded at many sites throughout the estuary at this time (Long et al. 1998; Metcalfe et al. 2000). Publication of the details of these palaeoenvironmental results from the lower Aire valley is currently under preparation (Kirby in prep.)

Summary The pattern and extent of sea-level change in the Holocene period is now well known, but for certain time periods (e.g. pre-8000 cal yrs BP and post-2000 cal yrs BP) data are lacking. In addition to temporal gaps, there remains spatial variability in the data, with more information concerning the sea-level history of the middle and outer estuary. Consequently, the pattern of water level change in the inner estuary is less well understood. Sites within the lower reaches of river valley systems are complex because both fluvial and estuarine processes influence deposition and the magnitude of the early- to mid-Holocene palaeotidal changes were greater in the inner estuarine areas (Shennan et al. 2000a). The way in which floodplain vegetation responds to regional processes such as water level change in the marginal freshwater tidal river valleys is shown below. Much of the following section concerns the results of an investigation into floodplain dynamics and wetland development in the lower Aire valley (Kirby 1999), which sheds light on the sensitivity of wetland vegetation to estuarine/catchment change.

34

Regional Vegetational History 35

Regional Vegetational History John C. Tweddle This paper provides a summary of the principal aspects and main knowledge gaps of the vegetational histories of eastern Yorkshire and northern Lincolnshire particularly during the Late Devensian and Holocene (Fig. 11). Pre-Devensian to Mid-Devensian A limited interglacial sequence obtained from a thin peat band contained within estuarine silts close to Kirmington in northeast Lincolnshire (Watts 1959) provides the only published pre-Devensian palynological record from the region. The low number of samples analysed limit interpretation, but it appears that the local vegetation was dominated by reedswamp and saltmarsh communities. Species-rich woodland containing abundant Quercus, Pinus and Alnus, and lower frequencies of Picea, Ilex, Ulmus and Corylus occurred. Although absolute dates are lacking, comparison of the pollen spectra with data from other sites in Britain led the author to suggest a Hoxnian (OIS 9?) age for the sequence (ibid). The vegetational history of the area during the Early-Mid Devensian is largely unknown, with no published pollen data available. It seems likely though that localities free of ice (e.g. the Lincolnshire and Yorkshire Wolds) would have supported limited areas of tundra vegetation (Flenley 1990). The Late Devensian More information is available concerning vegetational development during the Late-Devensian and in particular the Lateglacial period. Insect and plant macrofossil evidence from the immediately pre-Late Devensian ice advance deposits at Dimlington show a cold tundra landscape, which perhaps did not lie far from the contemporary ice front (Penny et al. 1969). The earlier part of the Lateglacial (the Windermere Interstadial) has been dated in Holderness as spanning ca 13,045 ± 14C yrs.270 BP (Birm-317) and 11,220 ± 220 14C yrs.BP (Birm-406; Beckett 1975). Despite the availability of several pollen records relating to this period, spatial coverage is far from complete with no profiles from the Lincolnshire Marsh, north of Butterbump (Suggate & West 1959), the Humberhead Levels, or the upland areas of the Yorkshire and Lincolnshire Wolds, and only one from the Vale of York (Bartley 1962; Gearey this volume). A greater number of records have been published from Holderness, most notably from The Bog at Roos (Beckett 1975; 1981; see Fig. 2), Gransmoor (Walker et al. 1993; see Fig. 13), and Skipsea Withow Mere (Hunt et al. 1984a), with the result that the Lateglacial vegetational and environmental history of the region is comparatively well known. Throughout the Lateglacial and early Holocene, low sea levels resulted in the exposure of the North Sea Basin, and Holderness and the Lincolnshire Marsh formed part of a low-lying plain linked directly to the continental mainland

36 John C. Tweddle

Regional Vegetational History 37

10

Yorkshire Wolds

0

16

10 km

12

rO

Hu

R i ve

River

River Derwen t

YORK

6

N

3 4

us e

ll

11

5

2

Ho lde

HULL

s

be

es

um

rn

er H Riv

r

13 1 Do

n

SCUNTHORPE

DONCASTER

Ri

A nch

ve

17

7 Riv e r

rT ren t

v

GRIMSBY

14

er

Ri

15

o lme

9

Time period

Lincolnshire Wolds

Pre-Devensian Late glacial Holocene

8 LINCOLN

Land over 60m 1 2 3 4 5

- Kirmington - The Bog at Roos - Gransmoor - Skipsea Withow Mere - Sproatley Bog

6 7 8 9 10

- Tadcaster - Messingham - Askham Bog - Bole Ings - Willow Garth

11 12 13 14 15

- Gilderson Marr - Cess Dell - Thorne Moors - Hatfield Moors - Brigg

16 - Routh Quarry 17 - Rossington

Figure **11: map showing the locations of the geographical areas and palaeoenvironmental sites referred to in the text.

(Jelgersma 1979; Coles 1998). During the earlier part of the Windermere Interstadial, an open sedge and grass-based community dominated the vegetation of Holderness, with herbs suggestive of unleached and calcium-rich soils frequent (e.g. Helianthemum), and woody taxa limited to isolated patches of Betula or Betula-Salix scrub (Beckett 1975; 1981; Walker et al. 1993). Aquatic and damp ground taxa, including Potamogeton, Myriophyllum and Equisetum, were frequent. Although the landscape remained relatively open during the midInterstadial, an increase in the area of Juniperus communis scrub occurred at both Gransmoor (from ca 12,400 - 12,500 BP) and Skipsea Withow Mere.

Interestingly, at The Bog at Roos Hippophaë rhamnoides increased rather than Juniperus (subzone RB-1b), possibly in response to a decrease in precipitation at the time (Beckett 1981). The next major vegetational change came with the local expansion of Betula, dated at Gransmoor to ca 12,000-11,800 BP (Walker et al., 1993). The coincident drops in Juniperus and/or Hippöphae representation in all three records suggest the direct replacement of juniper/sea buckthorn scrub with birch woodland. Although Betula dominated the vegetation of Holderness during the Mid-Late Interstadial, a marked, but apparently short-lived reduction in woodland extent occurred around ca 11,500 BP (Walker et al. 1993), with open sedge-rich grassland becoming frequent once more. A similar feature is seen in the Lateglacial Interstadial record from a number of other sites in northern England (e.g. Bartley 1962; Jones 1976; Dark 1998; Innes 1999), and is likely to reflect a response to a period of climatic cooling. There is a marked similarity between the records from Holderness and from Tadcaster on the northwest margin of the Vale of York (Bartley 1962), although at the latter, H. rhamnoides persisted throughout the Interstadial at low frequencies. Although data are lacking, it is reasonable to suggest that similar floral assemblages occurred on till-based soils throughout the region. The vegetational development of the higher ground of the Yorkshire and Lincolnshire Wolds is far less clear, but it is probable that a greater proportion of open ground persisted. More records are available for the Loch Lomond Stadial, although the Holderness data continue to provide the greatest level of detail. A number of marked changes are evident within the pollen record at the transition to the Loch Lomond Stadial (ca 11,220 ± 220 14C yrs.BP [Birm-406] Beckett 1975). The vegetational records are similar from all sites (e.g. Figs. 12 and 13), with Cyperaceae and Poaceae dominant, disturbed ground herbs (e.g. Artemisia-type and Rumex) common, and Betula and Salix present at low levels only. Pinus sylvestris pollen is also frequent at The Bog at Roos (up to 30% TLP; Beckett 1975), and macrofossil evidence is provided at Gransmoor, where a Late Upper Palaeolithic uniserial harpoon point is embedded in a piece of ?Sorbus sp. (Sheldrick et al. 1997). Low frequencies of Empetrum nigrum attest to the presence of heathland elements. Although birches were infrequent, it is probable that both dwarf and tree birches were present, a suggestion that is supported by the macrofossil record from Skipsea Withow Mere (Hunt et al. 1984a), and pollen grain measurements from central Holderness (Beckett 1981; Tweddle 2000). Sedge-dominated steppe vegetation occurred in the Vale of York (Bartley 1962), and the limited evidence from Messingham, northwest Lincolnshire shows a similar pattern (Buckland 1982). Collectively the pollen, plant macrofossil and Coleopteran data indicate a harsh environment with open sedge-based fen dominant and some dwarf willow (Salix herbacea). More sheltered localities may have supported birch or tree willow scrub (ibid). The deteriorating climatic conditions show equally clearly in the sedimentological record, with the high frequencies of (reworked) PreQuaternary microfossils in the records from Gransmoor (Walker et al. 1993),

38 John C. Tweddle Sproatley Bog, Gilderson Marr and Cess Dell (Tweddle 2000). The only available microscopic charcoal record covering the Lateglacial is for the site of Sproatley Bog, central Holderness (Tweddle 2000), and this only covers the Mid-Late Loch Lomond Stadial. Extremely high concentrations of fragments not readily separable from charcoal occur throughout, and whilst it is probable that some of the input reflects burning of the steppe-type vegetation, inwashed coal fragments eroded from the underlying till probably account for a large proportion of the input. At Messingham, there is a loose association between a Late Upper Palaeolithic end scraper and macroscopic Salix charcoal (Buckland 1984). The Early/Mid Holocene (ca 10,000-5,000 BP) Greater detail is available for the Holocene than for the Lateglacial, with records available for at least part of the time period from all areas (Fig. **11). Although there is a long history of palynological investigation within Holderness (e.g. Clark & Godwin 1956; Beckett 1975; 1981; Blackham & Flenley 1984), until recently the majority of the published records have been of low temporal resolution and predominately undated (Dinnin 1995). Recent palynological investigations into the vegetational histories of four small (60 spp of foraminifera and smaller numbers of scaphopods, cirripedes, fish teeth and vertebrae and echinoid spines. It includes mainly littoral and shallow-water boreal species, such as the molluscs Arctica islandica, Astarte borealis, Chlamys islandica, Hiatella arctica,

55

Serripes groenlandicum, Amauropsis islandica and Boreoscala groenlandicum, the ostracods Krithe glacialis, Rabimilis mirabilis and Acanthocythereis dunelmensis and the foraminifera Lagena pseudocatenulata, L. apiopleura, Globigerina bulloides var. borealis, Elphidium subarcticum, Elphidiella arctica, Dentalina frobisherensis, D. pauperata, D. baggi, and Trichohyalus bartletti. Reid (1885) pointed out that it includes no species with an exclusively southern range, and that the assemblage is more markedly boreal than that of any other British deposit. Based on the proportion of extinct molluscs, Wood (1870) and BadenPowell (1956) suggested that the assemblage is ‘early glacial’ (i.e. Anglian) in age, but pollen and dinoflagellate cysts from Bridlington suggest a Pastonian age (Reid and Downie, 1973), and Lewis (1999) reported amino acid ratios for Mya truncata shells indicating an Early Pleistocene age. The original marine sediment removed from the floor of the North Sea could have been deposited at various times. The Bridlington Crag (bluish-grey clay) at Dimlington also contains erratic cobbles, mainly igneous and metamorphic rocks from Scotland or Scandinavia, which were probably dropped from floating ice. At both Dimlington and Bridlington it therefore seems to form rafts of mainly cold-water (perhaps partly glaciomarine) sediment torn from the floor of the North Sea and incorporated into the Basement Till. Incorporation could have occurred either when the Basement glacier grounded on the floor of the North Sea (or advanced over it when it was dry because of a eustatic fall of sea-level), or when the ice that deposited the Skipsea Till subsequently disturbed and remobilised the surface of the Basement (Berridge and Pattison, 1994, p. 43). At Dimlington a few of the typical Bridlington Crag fossils, notably Macoma balthica, can be found in the Basement Till remote from any of the rafts. They also occur, but much more rarely, in the Skipsea Till. Small amounts of seafloor sediment have therefore been assimilated into the matrices of these tills but have lost their characteristic colours by admixture with other glacial detritus, so both have some characteristics of assimilation or deformation till (Boulton and Hindmarsh 1987). The long axes of clasts in the Basement Till show two clear directions of preferred orientation: NNE-SSW and WNW-ESE (Penny and Catt 1967). Because of the inclusions of marine sediment, the former is more likely to represent the true direction of ice movement. The strongest WNW-ESE maxima occur in the upper 2-3 m of the till, and are associated with isoclinal overfolds, whose axial planes trend in the same direction. The folding is sometimes picked out by lines of chalk debris (Eyles et al. 1994). The folds and axial planar orientation of clasts could have resulted either from subglacial deformation during deposition of the Basement Till or from subsequent disturbance by the Late Devensian glacier (depositing the Skipsea Till), as this also invaded the East Yorkshire coast from a NNE direction. Locally the uppermost 20-80 cm of the Basement Till in the Dimlington cliff section is slightly oxidised and incompletely decalcified (Fig. **16), indicating a period of subaerial weathering before deposition of the overlying Skipsea Till.

56 John A. Catt

Dimlington Cliff

57

Figure **16. Skipsea Till overlying a weakly weathered surface of Basement Till, Dimlington.

Figure **17. Glaciotectonic basin containing Late Devensian lacustrine silts and sands on the deformed upper surface of Basement Till and overlain by Skipsea Till, Dimlington.

Till resembling the Basement in colour, particle size distribution and erratic suite (often including Scandinavian rocks and occasional arctic shells) has been recorded at scattered sites in north-east England, including Welton-le-Wold in Lincolnshire (the Welton Till of Alabaster and Straw 1976), Reighton in Filey Bay (Lamplugh 1879) and Warren House Gill in Co. Durham (the ‘Scandinavian Drift’ of Trechmann 1931) as well as at Dimlington, Bridlington and Sewerby. At Sewerby its surface was reddened by weathering before deposition of the Sewerby (Ipswichian) raised beach (Catt and Penny 1966), and at Warren House Gill large erratics in the uppermost layers of the Scandinavian Drift were weathered before deposition of the Northern or Cheviot Drift, equivalent to the Skipsea Till (Trechmann 1952). b) The Dimlington Silts and Sands (= Dimlington Bed of Lewis, 1999): these have been recorded in a series of enclosed basins in the surface of the Basement Till at various places along the Dimlington cliff section. Those seen between 1941 (Bisat and Dell 1941) and 1964 (Catt and Penn, 1966) are shown in black on Fig. **15. The basins are 1-4 m deep and 5-50+ m wide. Within the deeper basins the succession is dark brown stratified silts at the base passing up by alternation into yellow sand; smaller basins may contain only the basal silts. Reid (1885, p. 44) recorded similar deposits lying within hollows in the surface of the Basement Till at Bridlington. The bedding within the basins at Dimlington is not horizontal, but is deformed synclinally, parallel to the underlying surface of the Basement Till (Fig. **17), and the basins often occur

above the general surface of the Basement Till. This suggests that the individual basins were not formed by erosion, nor are they hollows in the original surface of the Basement Till, but are glaciotectonic in origin, i.e. formed by folding and disturbance of the upper layer of the Basement Till together with an extensive cover of silts and sands, which therefore could have accumulated in a single large shallow lake. This disturbance probably occurred during advance of the glacier that deposited the overlying Skipsea Till. In places upper layers of the interstratified silts and sands immediately beneath the Skipsea Till have also been deformed into isoclinal drag folds (Fig. **18) indicating NNE-SSW compression by the overriding Late Devensian glacier. The same compression also explains sets of conjugate shears sometimes seen within the silts. The basal silty layers often contain black strands of moss (Pohlia wahlenbergii var. glacialis), which were radio carbon-dated to 18,500 ± 400 years BP (I-3372) and 18,240 ± 250 BP (Birm. 108) (Penny et al. 1969). This species is typical of cold water habitats such as glacial meltwater streams. The silts have also yielded 5 species of boreal freshwater ostracods (Catt and Penny, 1966), an impoverished fauna of 13 Coleoptera (Penny et al. 1969), larval heads of Diptera (as yet unidentified), a single bird bone (unidentified), rare seeds of aquatic plants (Daphnia ephippia, Eleocharis palustris, Menyanthes trifoliata, Potamogeton alpinus and Potamogeton filiformis) and sparse (probably fartravelled) pollen of Pinus and Betula (Catt 1987, Table 7). The Coleoptera provide the most specific palaeo-environmental evidence, and indicate deposition in a shallow lake with little aquatic vegetation surrounded by

58 John A. Catt

Fig. **18. Isoclinal drag fold in Late Devensian lacustrine silts and sands overlain by Skipsea Till, Dimlington.

expanses of almost bare sand and silt with a patchy covering of moss. Most are hardy boreo-montane species indicating very cold conditions. Together with the known habitat of the moss, this suggests the Late Devensian glacier was close by and probably reached Dimlington soon after 18,240 years BP. Eyles et al. (1994) reported mud drapes on ripples in the Dimlington Silts, and concluded that they indicated a shallow marine or tidal river environment. However, the very specific palaeoenvironmental interpretation of the fauna and flora clearly shows that this is incorrect. Also current knowledge of glacioisostatic and glacioeustatic changes during the Late Devensian (Funnell 1995; Lambeck 1995) indicates that Dimlington would have been well above sea level at the time the silts were deposited. c) Skipsea Till ( = Skipsea Member of Lewis, 1999; Skipsea Till Formation of Evans et al. 1995): at Dimlington this is 5-9 m thick and has a uniform dark greyish brown (10YR 3/2) matrix, containing less clay (22-38%) but more silt (34-44%) and generally more carbonate (6.1-21.7%) than the Basement Till (Madgett and Catt 1978). It is slightly less compact than the Basement Till (Bell and Forster 1991). The total sand (60-2000 µm) content (22-42%) is similar to that of the Basement Till, but the mineralogical composition of the fine sand (60-250 µm) distinguishes the two tills quite clearly: the Basement contains more epidote and hornblende but less pyrites, siderite, chlorite and biotite than the Skipsea. The coarse silt (20-60 µm) mineralogy also distinguishes the two tills in the same ways (Madgett and Catt 1978). In the 6-16 mm size range, Chalk,

Dimlington Cliff

59

flint and shale clasts are more abundant in the Skipsea Till than in the Basement. Chalk and flint increase progressively southwards from Flamborough Head to North Norfolk. Rarer erratic types include rhomb porphyries, larvikite, Cheviot porphyries, granites, dolerite from the Whin Sill, Carboniferous and Magnesian Limestones, greywackes, coal and metamorphic rocks. These suggest that the Skipsea Till was deposited by east coast ice, which invaded the North Sea basin mainly from southern Scotland and Northumberland. A till similar in colour to the Skipsea Till can in fact be traced northwards along the east coast through Co. Durham and Northumberland into SE Scotland. The pollen in the till is derived principally from Carboniferous, Jurassic and Lower-Middle Cretaceous deposits (Hunt et al. 1984), all of which occur along the east coast of Britain and on the floor of the North Sea. Pleistocene pollen accounts for only 11% of the total spectrum. NNE-SSW ice movement along most of the Yorkshire coast is indicated by macrofabric studies (Penny and Catt 1967) and by fold axes trending WNWESE (Evans et al. 1995). The Skipsea Till is more variable in colour and composition than the Basement or Withernsea Tills. In particular in contains grey, white and red streaks and bands (Fig. 19), which are composed almost entirely of crushed sheared Liassic mudstone (or sometimes Coal Measure shale), Chalk (+ flint) and Triassic marl and sandstone, respectively. Most of the streaks are 20-80 cm thick and only a few metres in lateral extent but some, such as the "Red Band" noted by Reid (1885) in the Hornsea-Cowden area, can be traced laterally for hundreds of metres or even kilometres. They often show contortions and steeply inclined reverse faults indicating NNE-SSW compression. Bisat (1939; 1940a; 1940b) used the more laterally extensive of these sheared and incompletely assimilated rafts, together with stratified sands/gravels and minor differences of matrix colour and erratics in the intervening layers of the till, to subdivide it into five beds (the Basement Drab, Sub-Drab, Lower Drab, Middle Drab and Upper Drab). The extent of each subdivision as seen in the 1930s and 1940s was hand-drawn by Bisat on a long section of the Holderness cliffs. He never published this, but it was redrawn after his death by Catt and Madgett (1981). Most of Bisat’s sequence was originally established in northern Holderness, and he had some difficulty correlating it with the till at Dimlington, as the two areas are separated by an 11.5 km gap (from Holmpton to Tunstall) where the Skipsea Till is below low-tide level and the coast exposures show only the overlying Withernsea Till. His later field notes suggested a five-fold subdivision of the Skipsea Till at Dimlington (see Catt and Madgett, 1981, Fig. 1, sections A-C), though he was still uncertain how to correlate this sequence with that north of the Holmpton-Tunstall gap. As Eyles et al. (1994) pointed out, the crude bedding of the Skipsea Till indicates extensive shearing in the parent ice and incomplete subglacial mixing. The lower contact of the Skipsea Till is sharp, whether it overlies Basement Till (Fig. **16) or the Dimlington Silts and Sands. However, in places there is a clast pavement or gravel stringer often only one clast thick between the two tills. d) Stratified sands, silty clays and fine chalky gravels: these extend for a

60 John A. Catt

Dimlington Cliff

61

Fig. **19. Sheared and contorted rafts of chalk and Coal Measure mudstone in Late Devensian Skipsea Till, Dimlington (photo: J.Barber).

Fig. **20. Ripple-bedded glaciofluvial sands with mud drapes, between the Skipsea and Withernsea Tills, Dimlington (photo: M.D. Bateman).

distance of approximately 1 km beneath the highest part of Dimlington Cliff, and reach a maximum thickness of about 4 m. Comparison of the cliff sections recorded by Bisat in the 1930s (Catt and Madgett 1981, Fig. 1B), Penny in the 1950s (Catt and Penny 1966, Plate 24, also shown in Fig. **15) and BGS staff in the 1980s (Berridge and Pattison, 1994, Fig. 23) suggests that to the east these deposits lay entirely within the Skipsea Till, but as coast erosion proceeded westwards they have risen initially to a level entirely between the Skipsea and Withernsea Tills, and are now partly between the two tills and partly (on the southern side of Dimlington High Land) within the Skipsea Till again. This transgressive character suggests that they were deposited by an englacial meltwater stream. The sands and interbedded silty clays are either planarbedded or show ripples with mud drapes (Fig. **20); they contain no indigenous fossils. The fine chalky gravels occur mainly at the base towards the southern end of the section, where they show small contorted intrusions into the top of the Skipsea Till. The gravels also contain many small mudstone clasts and are weakly impregnated with black iron and manganese oxides, giving a characteristic mottled (‘pepper and salt’) appearance. Other bodies of stratified silts, sands and gravels, again with no indigenous fossils, are often exposed between the Skipsea and Withernsea Tills and as intrabeds of various dimensions within the Skipsea Till on many parts of the Holderness coast. Reid (1885, p. 46) and most later workers have concluded that they were deposited in subglacial or englacial streams and lakes. As Evans et al.

(1995) pointed out, their lower surfaces are often concave downwards and their upper surfaces planar, suggesting deposition in subglacial or englacial drainage channels resembling the ‘mini-eskers’ of Alley (1991) or the ‘canals’ of Clark and Walder (1994). However, many have been contorted and faulted, presumably by disturbance during emplacement of overlying till layers. Lewis (1999) designated the main bed of stratified sands, silty clays and fine chalky gravels at Dimlington as the stratotype of the Mill Hill Member, probably because Eyles et al. (1994) had correlated it with coarser gravels occurring between the Skipsea and Withernsea Tills at Mill Hill and Kelsey Hill near Keyingham, approximately 15 km inland. The inland Kelsey Hill Gravels are richly fossiliferous, and were interpreted as a Late Devensian marine beach by Eyles et al. (1994). However, this seems unlikely as other knowledge of glacioeustatic and glacioisostatic changes show that the North Sea area was dry land throughout the Late Devensian and early Holocene (e.g. Funnell 1995; Lambeck 1995). Also the fauna at Keyingham is rather mixed, with temperate marine and freshwater molluscs (Wilson et al. 1954) and cold-tolerant and warm-loving vertebrates (Penny 1963), so at least some and perhaps all of the faunal components there were derived from earlier deposits. North of Keyingham the gravels form a sinuous N-S ridge with a thin impersistent capping of Withernsea Till, and Catt and Penny (1966) interpreted them as an esker composed of englacial outwash derived from an unusually fossiliferous part of the Late Devensian ice sheet. Because of their unfossiliferous nature and

62 John A. Catt transgressive position within the Late Devensian sequence, the Stratified sands, silty clays and fine chalky gravels at Dimlington are unsuitable as a stratotype for the Kelsey Hill Gravels or Mill Hill Member. e) Withernsea Till ( = Withernsea Member of Lewis, 1999): this reaches a maximum thickness of 24 m at Dimlington High Land, but the uppermost 5 m shows evidence of Holocene soil development (oxidation, gleying, decalcification and softening of stones). In unweathered form the till has a uniform dark brown (7.5YR 3/2) matrix, composed of 30-40% clay, 42-48% silt and 14-25% sand, and the carbonate content (8.5-10.4%) has a narrower range than either the Skipsea or Basement Tills. In the fine sand and coarse silt fractions it contains more chlorite, chamosite and biotite but less epidote and hornblende than the Skipsea Till, and in the 6-16 mm size-range it contains more Carboniferous, Triassic and Liassic shale, siltstone and limestones than the Skipsea (Madgett and Catt 1978). Igneous and metamorphic rocks are rarer than in the Skipsea Till, and the Chalk and flint contents increase northwards along the Holderness coast (cf Skipsea Till). Many of the igneous erratics (e.g. Shap Granite) are from the Lake District, suggesting that the Withernsea Till resulted at least partly from the Late Devensian ice stream, which crossed the Pennines via the Stainmore Gap and Lower Teesdale (Catt 1991, Fig. 118). As in the Skipsea Till, stone orientation measurements suggest a consistent NNE-SSW direction of ice flow (Penny and Catt 1967). The lower contact of the till on the Skipsea Till or on glaciofluvial gravels is sharp. Because of landslipping, the Withernsea Till is usually less well exposed than the Skipsea. Despite this, Bisat (1939; 1940a) managed to subdivide it into five beds (see Catt and Madgett, 1981 for further details). It has proved difficult to trace these in recent exposures, but subdivisions defined by intrabeds of stratified sand and gravel can be recognised at many localities on the Holderness coast. These stratified deposits show similar features to the stratified deposits within and above the Skipsea Till. From petrographic analyses of subsoil samples collected throughout Holderness, Madgett and Catt (1978) showed that the Withernsea Till occupies an arcuate area of SE Holderness, extending along the coast from just north of Easington to Hornsea and inland for up to 15 km (i.e. to the eastern side of Kelsey Hill). There are no known onshore equivalents south of the Humber Estuary, but north of Flamborough Head it reappears in the coastal sections as far as the Tees Estuary. On the floor of the North Sea it extends to about 1º 60' E and as far south as 53º 25' N (Donovan 1973), though in the immediate offshore area the Skipsea Till forms the sea floor. The outcrop of the Withernsea Till in SE Holderness is distinctly more hummocky than many areas where this till is absent. Eyles et al. (1994) attributed this to repeated advances of the Withernsea ice margin as it retreated, to subglacial squeezing of poorly consolidated Withernsea Till or to downwasting of marginal dead ice originally emplaced by surging. However, the geomorphological distinction is more apparent than real, as

Dimlington Cliff

63

the areas of more subdued topography outside the Withernsea Till margin are covered either by proglacial sandur and glaciolacustrine sediments deposited during deglaciation (Evans et al. 1995) or by Holocene marine, fluvial and peat deposits (Catt and Penny 1966). The earliest radiocarbon dates of organic Late-glacial sediments within kettle-holes on the surface of the Withernsea Till (Beckett, 1981) suggest that SE Holderness was ice-free by 13,000 BP. Both the Skipsea and Withernsea Tills were therefore deposited within a period of no more than 5000 radiocarbon years. However, from comparisons with Late Devensian advances in the northern North Sea and eastern Ireland, Peacock (1997) suggested that the advance into Holderness may have been as late as Heinrich Event H1 at approximately 14 ka BP, in which case the two Late Devensian tills were deposited in a period of only 1000 years. Age of the Basement Till At Dimlington the Basement Till underlies and is therefore older than the Dimlington Silts and Sands, which were deposited around 18,240-18,500 14C years BP (uncalibrated). The localised gleying and partial decalcification of its uppermost layers beneath the Skipsea Till suggest a brief period of subaerial exposure and pedogenesis (probably