JOURNAL OF QUATERNARY SCIENCE (2003) 18(2) 151–168 Copyright 2003 John Wiley & Sons, Ltd. Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jqs.746
Multiproxy Devensian Late-glacial and Holocene environmental records at an Atlantic coastal site in Shetland GRAEME WHITTINGTON,1 * PAUL BUCKLAND,2 KEVIN. J. EDWARDS,3 MALCOLM GREENWOOD,4 ADRIAN M. HALL1 and MARIE ROBINSON1 1 School of Geography and Geosciences, Irvine Building, University of St Andrews, St Andrews, Fife KY16 9AL, UK 2 Department of Archaeology and Prehistory, University of Sheffield, Northgate House, West Street, Sheffield S1 4ET, UK 3 Department of Geography and Environment and Northern Studies Centre, University of Aberdeen, Elphinstone Road, Aberdeen AB24 3UF, UK 4 Department of Geography, University of Loughborough, Loughborough, Leicestershire LE11 3LU, UK Whittington, G., Buckland, P., Edwards, K. J., Greenwood, M., Hall, A. M. and Robinson, M. 2003. Multiproxy Devensian Late-glacial and Holocene environmental records at an Atlantic coastal site in Shetland. J. Quaternary Sci., Vol. 18 pp. 151–168. ISSN 0267-8179. Received 20 May 2002; Revised 8 November 2002; Accepted 17 December 2002
ABSTRACT: Marine erosion at Clettnadal, West Burra island off the west coast of Shetland, caused the drainage of a small water body at Clettnadal, exposing deposits of Late Devensian and Holocene age. Pollen, diatom and invertebrate analyses have provided variable records of environmental change during stratigraphical event GI-1. Event GS-1 is revealed by the non-pollen evidence, especially by Coleoptera, by sediment stratigraphy, and by radiocarbon dating. In contrast, the pollen evidence indicates that an arctic tundra flora, in which dwarf shrubs were prominent, persisted throughout the Late-glacial. The Holocene brought colonisation by tree birch, but by ca. Journal of Quaternary Science 9000 14 C yr BP the taxon had almost disappeared. This contrasts strongly with other Holocene pollen records for Shetland where both Betula and Corylus avellana-type survived longer—at some sites, for example, until ca. 2900 yr BP. The extreme westerly and exposed coastal situation of Clettnadal appears to be responsible both for a muted Late-glacial response in the pollen record of terrestrial vegetation and for the early replacement of woodland by a maritime grassland. The results provoke questions concerning biological stability at times of marked climatic change. Copyright 2003 John Wiley & Sons, Ltd. KEYWORDS:
Shetland; Scotland; invertebrates; diatoms; pollen; environmental change.
Introduction The Shetland Islands (59° 30 –60° 52 N; 0° 43 –2° 7 W) lie at the western extremity of the northwest European landmass. They are 150 km north of the Scottish mainland, 210 km west of Norway and at the same latitude as southern Greenland (Fig. 1A). Recent investigations have led to a greater understanding of their geomorphological history, especially for the Late Devensian period (Ross, 1996), whereas the palaeoenvironmental work has been concentrated mainly on the Holocene (Bennett et al., 1992, 1993; Edwards and Moss, 1993; Edwards et al., 1993; Whittington and Edwards, 1993; Hulme and Shirriffs, 1994). * Correspondence to: Graeme Whittington, School of Geography and Geosciences, Irvine Building, University of St Andrews, St Andrews, Fife KY16 9AL, UK. E-mail:
[email protected] Contract/grant sponsor: Carnegie Trust for the Universities of Scotland. Contract/grant sponsor: Natural Environment Research Council; Contract/grant number: GR9/1433 A.
Since the earlier work of Birnie (1981) and her re-examination of the site at Aith (Birnie, 2000), there have been no published, detailed palaeoenvironmental examinations of the Late-glacial period, except for a consideration of Late-glacial pollen and charcoal/fire relationships (Edwards et al., 2000). Exploratory pollen analysis and rangefinder radiocarbon dating (Hall, 1993) showed that a site discovered in 1991 at Clettnadal (National Grid Reference HU 358299) on West Burra, an island off Shetland’s mainland west coast, possessed Late-glacial as well as early Holocene deposits. As they occurred in an exposed face, they provided the opportunity for detailed palaeoenvironmental investigations. In addition to AMS radiocarbon dating, sedimentological, pollen, limited macrofossil, diatom and invertebrate analyses have been undertaken, with the latter providing the first such data for the Shetland Islands. Of special interest was the opportunity that the site provided to compare the relative sensitivity through time of proxy indicators for a site located on the Atlantic edge of northwest Europe, an area of major climatic change during the late Quaternary period.
152
JOURNAL OF QUATERNARY SCIENCE 20° W
A
20°E
0°
B
70 70°N
LLAA NND D
LA
Unst
A Y Circle
ND
N
O
ICE
Arctic
R W
GR EEE NN
AT L
S h e tl a n d AN
TIC
OCE
Yell
AN
Gunnister Gunnister Water Water
Dallicam Water
0°
C
Grunna Water
West Burra
ATLANTIC
Fetlar
60°N
Mainland Murraster
es
t
Vo
e
Bressay
Loch of Brunatwatt
West Burra
Clettnadal
Clettnadal
Langlochs Lang LochsMire Mire Aith
Virda Vatn Virda Vatn
30
50
40
20 20
10 10
Fugla Stack
W
OCEAN
Contours at 10 m intervals 0
300 m
© Crown Copyright. OS. An EDINA Digimap/JISC supplied service.
0
20 km
Figure 1 The location of Clettnadal in Shetland and the North Atlantic
The site and lithology The place name Clettnadal refers to a deep, cliffed inlet (geo) on the west coast of West Burra adjacent to the prominent feature of Fugla Stack (Fig. 1C). The solid geology of the area belongs to the Whiteness Division of metamorphic rocks, with West Burra lying mainly in the psammite group, although at Clettnadal the surface outcrop is of micaceous schist (Mykura, 1976). The site at Clettnadal (Fig. 2) lies ca. 20 m above sea-level and to the rear of a rock platform, which terminates to the west in a near-vertical sea cliff. The site formerly supported a small water body, which may have drained westwards into
the Atlantic. It would have shown strong similarities to an extant and probably larger lochan, Virda Vatn, which lies to the south (Fig. 1C). Unlike Virda Vatn, the former water body at Clettnadal has been extinguished by cliff recession, and exposure of the lacustrine sediments has resulted from extreme storm-waves partially stripping unconsolidated materials from the surface of the rock platform. This rock platform is striated to 304° and overlain by varying thicknesses of deposits. At the point where samples were taken for micropalaeontological and sedimentological analyses, the lacustrine deposits are 1.00 m in thickness, reaching 3.20 m in depth from the ground surface (Fig. 3). Their lithology consists principally of silts and sands and they are overlain by diamicton and breccia, as described in Table 1 and shown in columns beside the palaeontological diagrams. Of note are strata rich in the leaves of the dwarf shrubs Salix herbacea and Betula nana (plant nomenclature follows Stace (1997)) and a stratigraphical division that occurs at 1.00 m where a lag of schist boulders, up to 40 cm in length, is present. This upper breccia comprises crudely horizontally stratified rubble and pebbly grit. Clasts in the rubble bands dip at 5° and reach a maximum A-axis length of 15 cm, although 1–4 cm is common. The breccia is representative of a group of cliff-top deposits found at many points around the coast of Shetland. The blocks are mobilised by extreme waves during major storms and are moved inland to form low storm ridges or a scatter of blocks. The location of the blocks at Clettnadal indicates the action of massive waves at a height in excess of 19 m OD.
Field methods and sample storage The field section was cut by spade and sampling took place from a freshly cleaned face. Three overlapping monoliths of the lacustrine deposits, which begin at 2.10 m from the ground surface, were taken for all subsampling apart from that required for invertebrates. Field sampling for the latter involved
Figure 2 A view to the Atlantic across Clettnadal. The figure in the middle ground is standing on the floor of the former basin. Storm-wave damage is visible in the scar on the left. The clasts in the foreground have been mobilised and emplaced by storm-waves Copyright 2003 John Wiley & Sons, Ltd.
J. Quaternary Sci., Vol. 18(2) 151–168 (2003)
SHETLAND LATE-GLACIAL AND HOLOCENE ENVIRONMENTS
153
Figure 3 The floor of the former water body at Clettnadal, the remnants of the Late-glacial and Holocene limnic deposits and their diamict and breccia capping Table 1
Sediment lithology of the deposits in the Clettnadal basin
2.10–0 m
2.21–2.10 m
2.49–2.21 m
2.71–2.49 m 2.89–2.71 m 2.96–2.89 m 3.03–2.96 m 3.06–3.03 m 3.10–3.06 m 3.20–3.10 m >3.20 m
Schistose breccia in which several iron pans have developed (most notably one at 2.10–2.04 m). 2.10–1.00 m coarse, openwork breccia with partial infill of sand and sparse shell debris and individual schist clasts penetrate the diamicton to a depth of 25 cm. 0.50–0 m consists of an arcuate storm-ridge, cut into by recent wave action Grey-brown, sandy, silty diamict. Matrix-supported angular schist clasts, granule- to pebble-size present throughout. Basal contact irregular, with lobes protruding into underlying silty sand to a depth of 12 cm Yellow silty sand. Upper surface markedly undulating (up to 12 cm) and occasional schist pebbles above 2.31 m. LOI ca. 15% Dark brown sandy gyttja. Faint laminations. Betula nana leaves. LOI 40% Sand and silt. Salix herbacea and Betula nana leaves. LOI decreases to ca. 10–17% Blue-grey silt and sand. Salix herbacea and Betula nana leaves. LOI ca. 20–25% Blue-grey silt and sand. Salix herbacea leaves. LOI ca. 20–25% Poorly laminated brown silt and sand. LOI ca. 6–11% Blue-grey, coarse, plastic, laminated, micaceous schist silt. LOI 5% Glacial erratic lag Mica schist bedrock, striated
the extraction of 16 samples 5 cm thick, each measuring approximately 10 l, although the thickness was varied where necessary to take account of stratigraphical boundaries. The top of the uppermost invertebrate sample was coincident with the base of the schist breccia and sampling continued down to the base of the blue-grey silt overlying schist bedrock (Table 1). All samples were stored at 4 ° C until subsampling was undertaken. Copyright 2003 John Wiley & Sons, Ltd.
Laboratory methods Laboratory methods are presented below. Summary aspects of sedimentological results are shown in Table 1, and Fig. 4 portrays the stratigraphical relationships between samples.
Sedimentology Sedimentological analyses (Fig. 5) consisted of loss-onignition (LOI) measurements (4 h at 550 ° C), particle size determinations on LOI residues by sieving for material >53 µm (63–250 µm, fine sand; 250–500 µm, medium sand; >500 µm, coarse sand) and Coulter Counter measurements for material in the silt- and clay-size range (2% of total diatoms (TD). The taxa are grouped on the profile according to trophic status, pH and salinity preferences. Taxa were assigned to classes using autecological data (e.g. Denys, 1991/92). The classes indicate preference for freshwater, nutrient-poor water (class 1) through mesotrophic to eutrophic (classes 2 and 3), to halophilous (class 4) and mesohalobous to euhalobous (class 5) conditions. Class 6 represents unclassified taxa. One of the major taxa found at Clettnadal is Stauroforma (Flower et al., 1996). Following discussion with S.J.M. Droop (personal communication, 1996) an attempt was made during counting at this site to distinguish ‘types’, using striation densities, which are designated Stauroforma A, Stauroforma B and Stauroforma A/B (intermediate). The malform ‘type’ included clearly asymmetric and abnormal forms. Table 3 is a summary listing of the major diatom taxa and TD concentration by zone from near the base of the lacustrine section.
Plant macrofossils The leaves of Betula and Salix herbacea were easily detectable by eye. They were identified where necessary by reference to modern type materials and their presence was noted (Fig. 4). No further search for plant macrofossils took place.
Invertebrates The compressed nature of the sediments made processing for invertebrates difficult and each sample was soaked in strong washing soda for several days before disaggregation Copyright 2003 John Wiley & Sons, Ltd.
over a 300 µm sieve and paraffin flotation (Coope and Osborne, 1968) could be achieved. The macroscopic organic component of most samples consisted of only a few tens of grams in the most gyttja-like material and less than 5 g in the more inorganic samples. Preservation in most samples was poor, with much of the material being both heavily fragmented and testaceous, leading to problems in identification. Sufficient material remained, however, to provide some indication of the nature of both the local environment and climate. Insect data are presented in Table 4, with taxonomy after Barnard (1985) and Lucht (1987).
Radiocarbon dating Eleven AMS radiocarbon dates were obtained on bulk sediment matrix from the site. They are displayed on Figs 5, 7 and 8. Calibration has been undertaken (Table 5) and the values are based on the 2 σ range of Method B of CALIB 4.3 (Stuiver and Reimer, 2002) using the highest probability value. Dating estimates are based on straight-line extrapolations, with the dates rounded to the nearest 10 yr. The date of 12 070 ± 85 yr BP is out of chronological sequence with the overlying date of 11 180 ± 80 yr BP, although there is a slight overlap after calibration. The determination of an acceptable age–depth curve is difficult in such circumstances (cf. Edwards and Whittington, 2000) and the one chosen (by-passing date AA-33286) represents one possible interpretation (Fig. 10). Above the date of 7070 ± 30 yr BP problems in the interpretation of the sedimentation history discourage the extension of the age–depth curve. This is discussed later. Dates in the text are expressed as uncorrected radiocarbon years BP unless specified otherwise. J. Quaternary Sci., Vol. 18(2) 151–168 (2003)
Silty sand
310
305
300
295
290
285
280
275
270
265
260
255
250
245
240
s ar
BP
20
7070±30
310
305
300
295
290
285
280
275
270
265
260
255
250
245
240
235
20
40
-ty
pe
Ju n
ip
er
gr
Em
tr
pe
lla
e
av
20
di
ff. 40
60
80
yp C
m
un
is
na
lix
Sa
20
pe -ty
e yp
20
20
40
m ru flo pe -t ni ty ype s s r u s t la te liu ra pe cri isel ne al go fo ui flo la -ty a str le si pe os m ni ty tu et ng a se lu us lus alu ab l u l l i c b m a f a y u i a a o s cu cu p ill lla zi tru d nt ph nt ex ex ex re un un tha alic pen er io de te li yr tto an an al ni up um um um R R C Th Fi Po Li M U H R R R
us
un e
um
Blue-grey silt
ea ac Po
ni
um
20
er
ac
co m us
Sand and silt
20
e 40
ae
ff. I
20
s
te
e so
40
hi
ec
ra
po
s no
I
20
s
te
e so O
ph
200 400 600 800
um di o p ly Po
m su os gl
io
ris
st
cu
la
20
er Pt
Figure 7 Percentage pollen diagram for selected taxa from Clettnadal. Each cross represents one pollen grain where percentages are 7 × 107
ca. 3 × 107
Fragilaria pseudoconstruens 53; Stauroforma B 35; F. construens var. venter 9; Staurosirella pinnata 7 Stauroforma A 25; Navicula schmassmannii 15 Fragilaria pseudoconstruens 36; Psammothidium grischumum 16; F. construens var. venter 12; Staurosirella pinnata 8
>7 × 107
