Late Quaternary glacial and environmental history of Kongsaya, Svalbard OLAFUR INGOLFSSON, FINNBOGI ROGNVALDSSON. HELENE BERGSTEN. LARS HEDENAS, GEOFFREY LEMDAHL. JUAN M. LIRIO and HANS PETTER SEJRUP Ing6lfsson, 0.. Rognvaldsson, F., Bergsten. H . . Hedenas, L., Lcmdahl, G . , Lirio, J . M . & Sejrup, H . P. 1995: Late Quaternary glacial and environmental history of Kongsoya, Svalbard. Polar Research 14(2), 123-139. On western Kongsoya. Svalbard. three coarsening-upwards sequences of marine to littoral sediments, separated by tills, are recognised in sections at ca 50-92 m above present sea level. These sequences show major glaciations in the northern Barents Sea, resulting in substantial glacioisostatic downpressing of Kongs0ya. Till fabrics indicate ice movements controlled by the local topography. while glaciotectonic deformations suggest that ice moved from an ice divide northeast of Kongscaya. independent of the local topography. The stratigraphical evidences show two pre-Holocene ice-free periods, when the climate was similar to or slightly warmer than at present. The age of these periods is not clear. It is suggested that the elder ice free interval is older than isotope stage 5c. The younger ice free interval could be of Eemian or Early Weichselian age. The uppermost succession of sublittoral-littoral sediments is of early Holocene age. It relates to the high ( r l O O m ) postglacial marine limit, dated to approximately 10,OOO BP. dlafur lngolfsson and Helene Bergsten, Earth Sciences Cenire, University of Goieborg, Guldhedsgaran 5a. S-413 81 Goteborg. Sweden: Finnbogi Rognualdsson, Fjolbrautarskoli Vesiurlands, Is-300 Akranes, Iceland; Lars Hedeniis, Deparimeni of Crypiogamic Botany, Swedish Museum of Naiural History, Box 50007. S-104 05 Slockholm, Sweden: Geoffrey Lemdahl, Deparimeni of Quaiernary Geology. Lund University. Tornauagen 13, S-223 63 Lund, Sweden; Juan M . Lirio, Instituio Anrrirrico Argeniino. Cerriio 1248. I010 Buenos Aires, Capiial, Argentina; Hans Petter Sejrup, Geological Institute, Uniuersiij of Rergen, Allegi. 4 1 , N-5007 Bergen, Norway. Introduction K o n g s ~ y a (191 km2) is the largest of a number of islands comprising Kong Karls Land in the easternmost part of the Svalbard archipelago (Fig. 1A). The island is mainly made up of Jurassic and Early Cretaceous strata, and the low ( ~ 3 0 0 m) mountains on eastern and western Kongsoya (Fig. 1B) are erosional remnants of a Cretaceous plateau, capped by basaltic lavas (Smith et al. 1976). Quaternary sediments mainly occur below 100 m a.s.1. (Salvigsen 1981). The purpose of the present investigation has been to locate and study pre-Holocene glacial and marine sediments on the western part of the island. Preliminary results were reported by Ing6lfsson et al. (1992), Rognvaldsson (1992) and Ingdfsson & Lirio (1993). Previous studies Geological observations were begun when Pike (1898) and Nathorst (1901) described raised marine beaches on the island and reported finds of driftwood at levels up to ca 40 m a.s.1. The idea of a glacier occupying the Barents Sea basin was first suggested by De Geer (1900). Nathorst (1901), who observed striated boulders on the plateau between Retziusfjellet and Tordenskjold- berget (Fig. lC), suggested that a Nordaustlandet glacier could have covered the Barents Shelf as far east as Franz Josef Land. Stromberg (1972) found glacial sculptures almost totally lacking on western Kongsaya. He discovered glacial striae at two localities, at high levels above cirques on Sjogrenfjellet and Tordenskjoldberget (Fig. 1C). Stromberg mapped the directions of glacial striae as indicating ice movements from SSW-SSE and concluded that most probably the striae were formed by former local glaciers of modest extent. Schytt et al. (1968) and Hoppe (1972) used the Kongscdya striae when reconstructing an extensive Weichselian Barents Ice Sheet, with an ice divide somewhere to the south-southeast of Kongs~ya. Knape (1971) reported driftwood from raised beaches up to 3 6 m a.s.1. and pre-Holocene marine shells from altitudes up to 89 m. The drift- wood samples were 14C dated and used for con- structingan uplift curve (Schytt et al. 1968; Hoppe 124 Olafur Ingolfsson et al. I SVALBARD 1W km 0 O- I WESTERN KONGSWA 0 5 k m F ; g . 1 . A : Svalbard location map. Localitieb mentroned in the t e x t : ( 1 ) Breggerhalveya; (2) Kapp Ekholm; (3) Linnedalen; ( 4 ) Ski1vika;Bellsund. B : The general phvsiography of Kongseya. C: Western Kongseya. The location of raised beaches is shown as an approximation of areas where beach ridges occur rather than location of individual ridges. 1972). indicating that beaches below 36 m were younger than 7000 BP Boulton (1979) suggested that there were two sets of beaches on K o n g s ~ y a , a lower Holocene set, and a higher >40,000 BP set extending to the marine limit at 100m. He suggested that Kong Karls Land had not been glaciated except by small glaciers during the Late Weichselian. Salvigsen ' (1981) published new radiocarbon dates from Kong Karls Land and presented an emergence curve for eastern Kongs- Qya. He showed that a shoreline displacement of more than 1 0 0 m had taken place during the Holocene, and his emergence curve is the strong- est evidence presented for a Late Weichselian Barents Ice Sheet over K o n g s ~ y a . Marine geo- logical data from the Barents Shelf (ElverhBi & Solheim 1983; Vorren et al. 1988; Elverhai et al. 1990; Solheim et al. 1990; Polyak & Solheim 1994) show glaciations several times during the Pleistocene. Geomorphic features on the sea floor in the northern Barents Sea, some 150 km south of Kongs@ya, are interpreted as fluted surfaces and De Geer moraines, thought to show that the Barents Shelf had been covered by a sliding temperate glacier, at least in the observation area (Solheim et al. 1990; Elverhai et al. 1990). Although the existence of a large Late Weich- selian ice sheet covering the northern and central Barents Sea area is now generally accepted, its extension, thickness and duration are still dis- cussed and debated (e.g. Mangerud et al. 1992; Elverhh et al. 1993; Forman et al. 1995; Lambeck 1995; Siegert & Dowdeswell 1995). Forman (1990) described the spatial variation in emer- Late Quaternary glacial and environmental history of K o n g s ~ y a , Svalbard 125 gence records from Svalbard, and concluded that the uplift pattern in Nordaustlandet, the islands in the Barents Sea and on central and western Spitsbergen, reflect glacial loading of the Barents Ice Sheet during the Late Weichselian. This con- clusion strongly indicates that the same ice dome or separated ice domes i n dynamic equilibrium have caused the isostatic depression observed. Thus, almost synchronous interglacial/inter- stadia1 events in the whole Svalbard area could be expected, provided that this pattern of emer- gence is valid for earlier deglaciations. Mangerud et al. (1992) concluded that prior to the Late Weichselian glacial build-up, the Barents Ice Sheet was small or non-existent, and that glaciers on and around Svalbard were not much larger than today when the advance to the last glacial maximum started sometime after 25,000 BP. Methods The stratigraphic investigations were concen- trated at t w o main localities, the col of Hidden and in the Bogen cirque (Fig. 1C). The sites were spotted by shell fragments in scree, and then cleared by digging. Elevations of sections and of raised beaches were surveyed by repeated measurements by AIR-HB-1A electronic alti- meter and some were levelled by theodolite. Samples were collected for analysing grain-size, plant and insect macrofossil content, foraminifera and mollusc, species, Thermoluminiscence (TL) and 14C determinations and amino acid diagenesis. The I4C age determinations were performed at Lund University’s Radiocarbon Laboratory, and the TL-dating was carried out at the Nordic Laboratory for Lurniniscence Dating in Ris0, Denmark, using sand-sized potassium feldspars. Amino acid diagenesis was measured in the protein matrix of fossil molluscs at the Bergen Amino Acid Laboratory, following the procedure described by Miller et al. (1983). Glacial morphology Glacial sculpturing and erratics The landscape of Kongsoya (Fig. 1B) indicates that basal sliding and glacial erosion occurred when the island was completely covered with ice. On western Kongsoya the glacial erosion is con- centrated in shallow cols, valleys and cirques, leaving the plateaux relatively unmodified. The plateaux of Hirfagrehaugen, Retziusfjellet, Tor- denskjoldberget and Sjogrenfjellet, as well as basaltic bedrock outcrops at lower altitudes, have been searched for glacial sculpturing, striae and erratics. The basalt plateaux are deeply weath- ered, and no clear glacial sculpturing is evident. Usually the surface is covered with regolith and crude soils. I n a few localities bedrock outcrops show stoss-and-lee topography, but none were found to carry glacial striae. The only erratics found on the plateaux were remains of petrified Cretaceous tree trunks, probably eroded by the glacier from the local sedimentary bedrock and only transported a short distance. Raised beaches and the marine limit The marine limit on western Kongs~ya is at 94- 98 m a.s.1. (Fig. lC), which is in good agreement with Knape’s (1971) results. The highest beach ridges on H ~ g s l e t t a were levelled to ca 100m a.s.l., which is about 10 m lower than Salvigsen’s (1981) marine limit on eastern Kongsoya. Sal- vigsen & Nydal (1981) found that isostatic uplift has been greater in the eastern than in the western part of Kongsoya during the last 7000 IJC years, and they took this as an indication of a Late Weichselian ice centre east of Kongsoya. The material in raised beach ridges on western Kong- soya is primarily of local basaltic provenance, but crystalline rocks, granites and gneisses, are also found. These have probably been carried to the island by sea ice. The beach ridges at altitudes between ca 40 m and 60-70 m a.s.1. on H ~ g s l e t t a , formed between ca 9000 BP and 7000 BP (Salvigsen 1981), are composed of well-rounded cobbles and boulders, with little gravel or sand, while the finer grain sizes are more characteristic for beach ridges above and below. This could indicate a higher energy in the coastal environment and less annual sea-ice cover during the formation of intermediate altitude beach ridges. This is supported by the lack of driftwood on beaches above 35-40 m. Glacial stratigraphy of Hidalen Fluvial erosion in Hidalen (Fig. 1C) has exposed Late Quaternary sediments resting on Triassic and Lower Jurassic bedrock. Knape (1971) 126 Olafur Ingolfssorz et a / . m a.s.1. 90 85 80 75 70 65 60 55 50 48 15-31 Lithology 0 Sand Silt Gravel Massive diamicton Stratified El diamicton intraclasts m Sedimentary structures Lamination r;;l Load deformation Contacts Sharp I Y Erosional i Sheared Fossils I In silu shells P Paired shells w Whole shells I Shell fragments Directional features c3 Glaciotectonic orientation -. Fabric orientation 15-30A E ,/' I , "V k;" dl / I , ,,' / ..... , - - I 15-30 B ' I Fig. 2. Lithostratigraphy of the Hidalen sites. ,/' - ' A 1 Fig. 2. Lithostratigraphy of the Hidalen sites. reported shell samples of >38,000 and >Jo,000 BP from 89 and 67 m a d . from Scree on the northern side of Hidalen. We found shells. both whole and fragmented. in scree, at a number of localities. Four sites (numbered 15-30A. 15-30B, 15-30C and 15.31) were investigated in detail. and one additional site (14-03) was briefly surveyed and sampled. T h e lithostratigraphy of each site and correlations between sites are shown on Fig. 2 . A cross section of t h e Hidalen stratigraphy, with location of sites studied, and correlation between sites is shown in Fig. 3. Late Quaternary glacial and environmental history of KongsGya, Svalbard 127 Fig. 3. A cross section of the Hidalen stratigraphy, with location of sites studied. w E Unit A Unit A is the 'lowermost stratigraphical unit in Hidalen. It is only exposed in section 15-30C. It is a 0.4 m thick massive, compact, silty-sandy diamicton. Striated clasts were recognised in the diamicton. Its lower contact to the stratified, unli- thified Jurassic sand is sharp but deformed. Len- ses of the sand have been incorporated into the diamicton and smeared out, and wedges of dia- micton have been thrust into the substratum. The deforming force came from an easterly direction. The diamicton is interpreted to be a till. Unit B Unit B consists of a number of lithofacies, recog- nised in sections 15-30A, 15-30B and 15-30C, which are interpreted to have been deposited in a marine environment. In the lower part of the unit, laminated silts and sands and massive silts, with occasional outsized clasts, dominate. Con- tacts between lithofacies are usually conformable, sometimes loaded. The thickness of the low- ermost silty-sandy facies of the unit is 1-3 m. Mollusc shells, fragments and some whole, were sampled from a sandy-gravelly lag-horizon in sec- tion 15-30B. This part of the unit is interpreted as a stratified glaciomarine deposit, with dropstones. In the middle part of the unit, stratified dia- mictons and massive or laminated sands domin- ate. The diamictons carry numerous pebbles and cobbles, but a fabric studies did not reveal any preferred orientation of clasts (Fig. 4A). Occasional thin intrabeds of gravels occur, as well as pebble and cobble clasts. Ruptured and bent substratum below a clast, indicating drop, was noted at two sites. Shell fragments were sampled from one of the diamictons, but could not be identified to species. This part of the unit is inter- preted to be a sublittoral glaciomarine sediment, influenced by meltwater input and slumping. Its thickness in the sections is 1-2 m. In the upper part of the unit. massive diamic- ton, gravel and sand with many pebbles and cobbles, occur together with laminated sand and silt. The massive diamicton facies has a thickness of 0 . 7 m , and consists of a silty-sandy matrix carrying rounded t o sub-angular pebbles. It has a sharp to erosional lower contact to the laminated facies below. The massive gravel and sand facies is about 0.25m thick. It grades upwards from granules and coarse sand to coarse and medium sand. Its lower contact is erosional. The massive bed grades into interstratified, laminated sand and silt, with occasional granules and pebbles. Together, the lithofacies indicate decelerating flow during deposition from a meltwater current entering the sublittoral environment. The upper- most lithofacies in unit B, found only in section 15-30B, is a thin bed ( ~ 2 0 c m ) of well-sorted, planar cross-bedded sand. It has an erosional lower contact and carries shell fragments. It is interpreted as a littoral deposit. Unit B in section 15-30A is heavily gla- ciotectonised, and a tight, recumbent syncline has developed (Fig. 5) below a thrust plane. The fold 128 (ilnfur Ingciltnon et al. A N Unit B, site 15-30A n: 25 V l : 295°130S1: 0.557 B N Unit C, site 15-31 n: 27 V l : 111°17"S1: 0.808 C N Unit E, site 15-30A n: 28 vi: i 3 2 ° ~ 1 0 S i : 0.752 Fig 4 Fabric analvre.; from western Kongsaya. A . fabric from a diamicton tacies in unit B. s ~ k 15-30A: B : fabric from unit C a t site 15-31. C. tnhric from unit E at site 15-30A. axis trend NW-SE. and the fold reflects over- folding towards the southwest. with ice move- ments from t h e northeast. Molluscs and foraminifera. - T h e mollusc fauna of unit B is p o o r , with only three species recognised (Table 1), and it can only be used to confirm a marine environment. All shells and fragments a r e of large, robust individuals. Fifteen samples from section 15-30A were prepared for foraminifera content, but none were found. Six samples pre- pared from section 15-30C, yielded three speci- mens of benthic and o n e specimen of planktonic foraminifera (Table 2). T h e specimens were worn and fragmented, and could not b e identified t o species. T h e almost total absence of foraminifera as well as juvenile and smaller molluscs in t h e sediments could be the result of dissolution. Plant macrofossils. - O n e sample from unit B, section 15-30A. was analysed for plant mac- rofossils a n d was found to contain o n e fruit of pondweed (Pontarnogeton), together with a num- ber of unspecified leaves, as well as bryophytes (Table 3). Wetland a n d moist heath bryophytes are well represented i n this sample, comprising Aitlacomniicrn turgidum, Drepanocladus (s. str.) s p . , Philonotis fontana/tomentalla, Pohlia wah- lenbergii a n d Tayloria sp. Drepanocladus species usually occur in nutrient-enriched environments. Philonotis species and P . wahlenbergii grow in moving water. Brachythecium reflexurn, Cer- arodon purpureus, Hylocomiurn cf. splendens and Pogonatum occur in mineral-poor places. Brachythecium reflexurn and Hylocomium indi- cate stable ground, whereas most of the other taxa indicate m o r e unstable conditions, with a sparse cover of vascular plants. Sanionia ortho- thecioides is today a coastal plant in northern E u r o p e and Svalbard, reaching southwards t o southern Finland, Sweden, Norway and Scotland (Hedenas 1989; Long 1992, 1993). It is typically found in shore meadows and rock crevices. T h e species recorded from unit B still occur on Svalbard. T h e few remains of vascular plants and the moss record show a poor vegetation of phan- erogams during the period when unit B was deposited. T h e wetland bryophytes and those of drier habitats indicate that a large part of the environment was mineral-poor, in contrast with the mineral-rich conditions usually found in peri- glacial environments or in areas which have recently been deglaciated (Miller 1987; Hedenas Late Quaternary glacial and environmental history of Kongs@ya, Soalbard 129 Fig. 5 . Site 15-30A in Hidalen. The section is orientated E (lefl) 10 W (right). 1992, 1994). Thus, the bryophytes suggest an environment where the minerals have been lea- ched out, which could suggest that the ground had been ice free for a considerable period of time. Unit C Unit C is a massive, silty-sandy diamicton, with abundant rounded to angular pebbles and cobbles and some shell fragments, recognised in sections 15-30C and 15-31. Its thickness is 0.6-1 m in the sections. Striated clasts occur, and a fabric study at site 15-31 (Fig. 4B) showed preferred orien- tation of clasts with dip towards ESE. The lower contact of the unit is erosive and sheared, and unit B below is at places sheared and folded below the contact. In a small, overturned anticline at the contact between units B and C in section 15- 31 the axial plane dipped towards northeast and the fold axis trended NW-SE, indicating that the deforming push came from the northeast. Unit C Table I . Subfossil molluscs from western Kongs0ya. Identified by S. Funder Site Unit 15-30B 15-30C 15-31 14-03 23-01 B D D D BIVALVIA Chlamys blandrca Serripes groenlandicw Macoma calcarea 1 Hiatella arcrica f Mya truncata f ECHINODERMATA Srrongylocentrorur droebachensis CIRRIPEDIA Balanw balanw C f f C 1 1 1 f 2 f f f S f C 5 2 f frequent (>20 valves/fragments); c: common (11-20); s: scarce (4-10); 1-3: rare (number of valves/fragments). 130 Olafur Ingolfsson et al. Table 2. Foraminifera1 species recorded in samples from units B and D in Hidalen. The upper number of each species refers to the actual number of counted specimens in the sample while the lower number refers to number of individuals when related to lOOg of dry sediment. Identified by H . Bergsten. Sample number 95-427 95-407 95-409 95-422 95-424 95-430 95-434 95-437 Site 15-30C 15-30C 15-30C 15-30C 15-30C 15-3oC 15-31 15-31 Unit B B B D D D D D Species Asrrononron gallowav I Loeblich & Tappan Buccella hannai arctica Voloshinova Carsrdulrna renrforme Norvang Elphidrum asklundr Brotzen Elp hidrum exca uatum (Terquem) Havnesrna orbiculare (Brady) Islandrella helenae Feyling-Hanssen & Buzas Reophax (p Sprroplectammina bifornilr (Parker & Jones) Verneulrna urcIicu Hoglund Vdrid mlcareou\ , p r ~ i e s Varia. agglutinated species Planktonic foraminifera No of benthic foraminifera in the sample N o of specimens related to 100 g 5edirnent (dried) 1 2 1 1 1.9 22 22.9 1 1 164 170 7 1 1 . 1 21 5 4 3 21 9 58.2 3 8 1 8 . 3 1.1 1 1 1 1 1 . 1 1 1 - 1 2 217 56 5 0 . 7 1 2 216 60.3 5 4 154 4 . 8 171.6 2 2 . 2 21 2 3 . 4 2 2 . 2 1 1 . 2 2 2 . 2 34 86 40.5 9 5 . 9 1 1.1 6 11.1 6 5 7.1 5 6 4 1 4 . 7 11 5 i n 54 284 6 4 . 3 316.5 is interpreted to be a lodgement till, deposited by a glacier flowing from the east through the Hidalen col. The till fabric indicates that the deforming push came from the southeast. while the sub-till fold geometry shows an ice movement from the northeast. We suggest that the deform- ation took place under a thick ice flowing across K o n g s ~ y a from the northeast and that the till fabric indicates ice push through the Hidalen col and relates t o a later deposition phase. with thin- ner ice and topographically controlled iceflow. Unit D Unit D is a facies association of laminated silts and sands. overlain by cross-stratified sands and gravels and cobble-gravels, reflecting a shal- loaing-upwards, marine t o littoral sediments T h e thickness of the unit is 5-8 m in sections 15-30C and 15-31, respectively. T h e lowermost 2-3 m are a succession of laminated sand and laminated to massive silt. Contacts between lithofacies a r e usually conformable, s h a r p or graded. Thin intra- beds of well sorted sand and lags of granules and pebble gravels, with erosive lower contact, occur scattered in the unit. T h e laminates sand and silt facies carry mollusc shells (Table l), paired as well as single a n d fragments, and single shells and fragments also occur in t h e sandy intrabeds. This part of the unit is interpreted t o be a sublittoral sediment. Lag horizons and intrabeds of sand were caused by current activity and small scale slumping. T h e sublittoral sediments a r e discordantly over- lain by a 1.5-4 m thick deposit of cross-stratified gravels and sands. T h e gravels a r e alternatively Late Quaternary glacial and environmental history of Kongseya, Soalbard 131 clast- and matrix-supported, and pebbles are usually well rounded. The unit grades upwards from pebble gravels at the base to cobble gravels at the top. Shell fragments occur throughout the unit. A 1.5 m high section was excavated in the mouth of the Hidalen valley, at 72 m a.s.1. (section 14-03), consisting mainly of angular to well- rounded clast-contact cobbles and boulders, with a silty-sandy matrix. The matrix contains abun- dant fossil molluscs with many large, paired indi- viduals of Chlamys islandica (Table 1). The sediments are similar to those in the present day beach ridges, where paired valves of C. idandica are also frequent. It is concluded that this part of unit D was deposited as a beach foreset, under similar conditions as the present, with seasonal open water and substantial wave action. Molluscs and foraminifera. - The mollusc fauna of unit D is poor, with five bivalve and one echino- derm species and fragments of Balanus balanus (Table 1). As with the fauna from the underlying unit B, the shells are biased towards large and thick individuals and the species composition does not define any clear mollusc community. The presence of Chlamys islandica does. however, show that the marine environment was influenced by Atlantic water. C. islandica is present around Kongsoya today, but does not enter high-arctic waters. Foraminifera1 specimens were found in five samples from unit D (Table 2). Both calcareous and agglutinated species were noted. Many speci- mens were fragmented or showed surface wear which could be the result of dissolution and dia- genetic effects. The most abundant species are lslandiella helenae, Verneuilina arctica, Astronion gallowayi, Buccella hannai arctica and Cassidulina reniforme. When comparing the foraminifera1 faunas of unit D with those previously reported from Svalbard (Feyling-Hansen & Ulleberg 1984; Miller et al. 1989; Nagy 1984; Lycke et al. 1992), the unit D sediments contain a comparatively poor fauna. However, the species in unit D occur at many other Svalbard sites, and are also found in the recent fauna (Hansen & Knudsen 1992). The faunas of unit D indicate seasonally open water, with marine conditions resembling those of today. Plant and insect rnacrofossils. - Four samples from unit D in Hidalen were analysed for plant and insect macrofossil content (Table 3). Remains of bryophytes dominate the record, just as in unit B . The samples from unit D were slightly richer in remains of vascular plants than unit B, and insect remains were also found. Among the vas- cular plants, Potamogeton jiliformis is an aquatic growing in eutrophic but clear shaliow water on sandy or gravely bottoms (Mossberg et al. 1992). It is widespread with an almost circumpolar dis- tribution, but it is not found growing on Svalbard today (HultCn & Fries 1986; R ~ n n i n g 1979). Selaginella selaginoides grows on wet calcareous ground such as spring mires and along the borders of brooks and lakes. It is circumpolar with a northern distribution limit along the Arctic coast of Scandinavia, Russia, Siberia, and Alaska, where it is found up to the low arctic zone. It is not found today on Svalbard (HultCn & Fries 1986; Mossberg et al. 1992; R ~ n n i n g 1979). Among the four taxa of insect, rove beetles (Staphylinidae) of the genus Stenus are usually found along the margins of water or other wet places (Harde 1984). The larvae of caddis flies (Trichoptera) are mainly found in fast moving streams, rivers and ponds. The larvae of midges (Chironomidae, Diptera) are also mainly aquatic. and many of them live in the muddy bottoms of lakes (Fitter & Manuel 1986). The diversity of the present insect fauna on Svalbard is very low, and neither beetles nor caddis flies have been recorded on Kongsoya (Danielsson pers. commun. 1994). Two of the vascular plants as well as two of the identified insect taxa have a more southerly distribution today and are not found on Svalbard, which suggests that the cli- mate was slightly warmer than today when unit D sediments were deposited. The bryophytes were badly preserved and few fragments could be identified to species. Among those, wetland plants were very. sparse, com- prising Aulacomnium turgidum, Plagiomnium ellipticum, Pseudocalliergon brevifolium , Scor- pidium cossoni and Warnsto@a sarmentosa (Table 3). Pseudocalliergon brevifolium and S . cossoni are found mainly in mineral-rich to cal- careous wetland environments, A . turgidum grows mainly in intermediately mineral-rich or sometimes rather mineral-rich places. Warnsrorfia sarmentosa is a species of intermediately mineral- rich habitats, in springs or in spring-influenced environments, whereas P. ellipticum is found in different wet and not too mineral-poor environ- ments. Most of the identified taxa indicate unstable, open habitats with a sparse cover of 132 O h f u r lngolfsson et al. Table 3. Vascular plants. insects and Bryphytes found in samples from Hidalen and Bogen. western Kongsaya. Vascular plants and insects identified by G . Lemdahl. Bryophytes identified by L. Hedenas. The nomenclature of Mossberg et al. (1992) is used for the vascular plants. whereas the nomenclature for the bryophytes generally follows Soderstrom et al. (1992). Site' Unit 15-30A 15-30C 15-31 23-01 Hidalen Hidalen Hidalen Bogen 5 D D TAXON Salis sp Selaginella selaginoides Poramogeron filiformis Carex sp Cerasrium sp Papaoer sp. Poramogeron sp. unspecified leaves unspecified wood Srenui sp. Staphvlinidae indet. Tnchoptera indet. Ambl~sregraceae sp. Aulacomnrum polwrre Aitlacomnrum rurgidimi Barrramia irhyphglla Brach,vlhecium reflexurn Brachytltecium trachvpodium Brvoenrhrophdlum recuriirosrrum B r w m cf pseudotriquerrum Bryunr sp C'erarodon purpurew Dicranella 5p. Dicranum cf. m a j w Dicranum cf. scoparium Dicranaceae sp Disrichium s p . Dirrirhum flexicaule Drepanocladrcs s . str. cf. Grimmio sp. Hylocomrirm splendens (incl. H. alaskanum) Hylocomr~tm cf. splendens (incl. H. alaskanum) Hypnum hambergeri Hypnum rei~olurum cf. Lophozia hyperborealpoerkei Lophozia sp,. Oncophorw uirens Philonoris fonranolromenrella Plagiomnium ellipricum Pogonarum dentarum Pogonarum dentarumlirrnigerum Pogonarum/Polyrrichum sp. Pohlia wahlen bergii Pohlia cf. wahlenhergii Pohlrn sp Polyrrichaceae sp. Polytrichasrrum alpinum PoI.vtrichashum sexangulare Polytrichastrum/Polynichum sp. Polytrichum commune (incl. P. jensenii) Polyrrvchum hyperboreum Polytrichum junipcrinumlsrricium Polvrnchum s p . Potfiaceae s p . Pseudocallrergon hreuifolium Racomirrium sp. 1F + 3s l s + 1L 1 s 3 s I S 1 s + 2L 1 s 1 s 11MS IF 1s 1 Mts + Lc IS + 2L 1s + 2L 1 s IS LIS IS 8S 12s +TW.W 5 MS 3m5 1F 15 ++ + 1 As +Lc 21 85 25 31 11 15 h35 4s+ 1L 5s + 2L IS 15 2 s + 3L I S + 1L 15 15 1 s 25 1 s + I B ZS/B 15 15 15 1s 11 2L 2 s 1s 1 s ZS I L 1 s 1 s i 2L 7s + 5L 1 s + 1L 4L 3L l s + I L 1S/B 2s 11 20S/3L* 25 35 1s+ 1L ZL 35 15 15 15 1s + 1L 11 35 15 Late Quaternary glacial and enuironmentul history of K o n g s ~ y u , Svulbard 133 Table 3. (Continued) Site 1 5 - 3 0 4 15-30C 15-31 23-01 Hid a I e n Hidalen Hidalen Bogen Unit B D D Sanionia urthothecioides Saniunia cf. orfhotheciuides Sanionia uncinata cf. Sanionia uncinafa Sanionia sp. Scurpidium cossoni Sframinergon srramineum Tayloria sp. Torfula sp. Warnsturfia exannulata Warnstorfia sarmenfusa Indet. akrokarp Indet. pleurokarp Indet. perigonium Indet. 1s 1s/B 1s 1B 1S/B 3s 1s 3 s 4s 2s 1L 6 s + 1L" lS/B 11s 13s 2 s I t 1 s 5S/B + 1L 2s 1s 1L S = shoot; B = branch: L = leaf; MS = megaspore; F = fruit; TW = twigs; W = wood; S = seed; Mts = metasternum; Lc = larval case; Th =thorax; Relative frequence, + = 1-10, ++ = 10-100. tWith antheridia and paraphyses. *These 2 0 stems and leaves were not checked for species identity (no transverse sections made). **Most of the shoots are of the same phenotype as is usually found in the large late snow-beds in the Scandinavian mountain range today. vascular plants. Some taxa, such as .Bryoer- ythrophyllum recuruirostrum, Distichium sp., Ditrichum flexicaule, Hypnum bambergeri and H . reuolutum are typical of mineral-rich to calcareous environments, whereas Bartramia ithyphylla. Ceratodon purpureus, Pogonatum dentatum, Polytrichastrum alpinum, the Poly- trichum species and Racomitrium sp. are either found in poorer habitats or have a rather wide range as regards the mineral status of the habitat. The moss species found in the present study still grow on Svalbard, except for Pseudocalliergon breuifolium, which has its closest locality in the Scandinavian mountain range. Unit E Unit E discordantly overlies unit D. It is a 0.5 m- thick, sometimes matrix supported and some- times clast supported, massive diamicton. It car- ries striated, angular clasts and shell fragments. At places it is a massive deposit of clast-contact, angular boulders, with silty-sandy matrix. A fab- ric analysis showed preferred orientation of clasts in a southeast-northwest direction, with dip towards the southeast (Fig. 4C). It was observed in section 15-31 and on top of section 15-30A. It is tight and compact and conducts groundwater on its upper surface. It could be followed as a moisture-bearing horizon in the walls of the Hidalen ravine from section 15-31 towards and over section 14-03. It is interpreted to be a till from a glacier that has moved through the Hidalen col. from east towards west. Unit F Unit F is a 0.5 m thick deposit of stratified fine to coarse sand with silt laminae, interpreted as sublittoral sediments, which are in turn dis- cordantly overlain by a 8-10 rn thick deposit of cross-stratified sands and gravels. The base of the unit is at 79 m a.s.1. The unit is upward coarsening, from sand and pebble gravels at the base of the cross-stratified facies, to cobble gravels at the top, at ca 92 m a.s.1. The uppermost 2-3 m are cobbles and boulders which can be followed OR the surface above the Hidalen ravine to the coarse-grained high-energy beach ridges that mark the marine limit at ca 100 m a.s.1. in the mouth of Hidalen. It is interpreted to be a succession of sublittoral t o littoral sediments. deposited when the sea stood at the postglacial marine limit. The glacial stratigraphy of Bogen Section 25-01 (Fig. 6) is located about 1.5 km from the coast. in a shallow side ravine to the main Bogen gully (Fig. 1C). inside the Bogen cirque. The top of the section is at 82 m a.s.1. The cirque mouth is transversed by a 94-08 m high boulder ridge. marking the marine limit. The section is composed of two interstratified lithofacies. laminated sand and laminated silt. with minor intrabeds of massive sand and massive silt. T h e thickness of the strata as exposed at site 23-01 is about 7 m . T h e clayey-sandy sediments carry in situ shells and abundant shell fragments. The deposition has been continuous. with mod- erate sediment input, and minor current rework- ing during deposition. probably in a shallow- water. low-energ), marine environment. Drop- stones occur, mainly as occasional granules and pebbles. T h e section is capped by thin colluvium. The whole section has been glaciotectonically deformed, and pushed up-slope from the floor of the Bogen valley. In the northern part of the section (left on Fig. 6 ) . a gently inclined, tight anticlinal fold has developed, and in the southern part of the section the sequence is transected by series of tightly spaced low angle reverse faults. The fold axis has a SE-NW direction, and the deforming thrust has come from the northeast. The tightly spaced low angle reverse faults have an imbricated pattern, slightly concave with apparent dip towards the north. T h e faulting has occurred subsequent t o the folding in a NE-SW orientated stress field. T h e direction of gla- ciotectonic deformations in section 23-Oi shows that t h e glacier which overran the site had advanced across t h e 300 m high Retziusfjellet pla- teau, into the cirque. Thus the glacier has been a part of a larger ice sheet covering Kongsarya and probably this part of the Barents Sea. Molhscs. - Only four species of bivalves and o n e echinoderm were recognised (Table 1). T h e shells were biased towards large individuals, and no juveniles were recognised. This, together with the total absence of foraminifera, indicates that carbonates may have dissolved in the sediments. Plant macrofossils. - Only o n e taxon of vascular plants. Selaginella selaginoides, was recognised. together with fragments of leaves, stems and wood (Table 3 ) . T h e bryophyte assemblage is similar t o t h e flora in unit B in Hidalen, with well- Fig. 6 . Site 13-01 in Bogen. North to the left of the photo. The anticlinal fold shows a glacial push from northeast and that the sediments have been pushed up-slope from the floor of the Bogen cirque. Late Quaternary glacial and enuironmental history of Kongsgya, Soalbard 135 represented wetland species. This assemblage has more species indicative of drier habitats than unit B in Hidalen. but taken together the flora shows poor vegetation of vascular plants and bryophytes in mineral-poor, leached out soils. The presence of Selaginella selaginoides suggests that the cli- mate was slightly warmer on Kongscdya during the deposition of these sediments than today. Age and correlations Radiocarbon and T L ages Four 14C age determinations were made on samples from units B and D in Hidalen and the Bogen strata. All gave infinite ages (Table 4). One TL dating from the Bogen strata yielded an age of 148 5 15 ka. Amino acid ratios The degree of isoleucine epimerization. expressed as aIle/Ile ratios, were measured in 21 samples from five different sites in Hidalen and from the Bogen site. The samples were prepared b y the method described by (Miller et al. 1983) and the aIle/Ile ratios in the total (bound + free) and free fraction are presented in Table 5 . Most of the shells were of the species Hiatella arctica and Table 4 . Radiocarbon dated pelecypod samples from western Kongs0ya. Site Sample No. Lab. No. Age BP hC-13 ( % ) ~ Hidalen 15-30C 95-405 Lu-3823 >38.000 -1.9 Hidalen 15-31 95-438 Lu-3821 > 40,000 2.5 Hidalen 14-03 95-439 Lu-3822 > 40 .000 3.1 Bogen 23-01 95-440 Lu-3824 >37.000 3.2 All samples are of in situ Mya truncata except sample 95-439 from site 14-03 where a reworked. paired specimen of Chla~nys islandica was dated. Table 5. Amino acid aIle/Ile ratios from western K o n g s ~ y a . Svalbard ~ Site BAL- Unit number 15-30c 15-31 14-03 15-30A 15-30B 1530A 23-01 D 2529 D 2612 D 2613 D 3073 D 2575 D 2610 D 261 1 D 3074 D 3072 E 2693 E 2694 E 2695 B 2530 B 2531 B 2574 E 2528 2524 2525 2527 2576 Species Mya truncata Hiatella arctica Mva truncata Mya truncata Hiatella arctica Hiatella arctica Hiatella arcricu Mya truncata Mya truncata Hiatella sp. Hiatella sp. Hiatella sp. Mean: Mya truncara Hiatella arctica Mya sp. Hiatella sp. Mva truncata Mva truncata Mya truncata Mya cruncata Mean: Hydrolysed aIle/Ile ratio 0.039 ? 0.006 0.049 f 0.001 0.050 f 0.004 0.029 i 0.004 0.056 i 0.005 0.035 f 0.000 0.044 t 0.001 0.034 t 0.001 0.043 t- 0.003 0.046 2 0.000 0.054 f 0.001 0.046 * 0.001 0.043 t- 0.008 0.085 f 0.008 0.074 f 0.006 0.080 ? 0.003 0.071 f 0.016 0.095 ? 0.002 0.085 ? 0.002 0.095 t 0.002 0.082 f 0.005 0.083 * 0.009 Free aIle/Ile ratio 0.243 ? 0.072 0.233 f 0.009 0.199 t 0.014 0.229 f 0.028 0.207 f 0.069 0.192 t 0.019 0.188 f 0.015 0.194 t 0.006 0.241 2 0.016 0.258 ? 0.013 0.290 t 0.020 0.231 f 0.038 0.372 ? 0.025 0.425 f 0.008 0.354 f 0.001 0.301 t- 0.086 0.303 f 0.022 0.433 ? 0.016 0.401 f 0.016 0.381 t 0.003 0.373 ? 0.013 0.380 * 0.042 Comments from scree from scree from scree in situ from scrcc from scree from scree in situ not in situ fragment fragment fragment not in situ not in situ fragment fragment in situ in situ in situ in situ The BAL-number is the sample number at the Bergen amino acid laboratory. 136 C 0 0 .- e !? P 5 e Q, e 0, C in 0 .- .- e e Q, Q, m - - . - - 0 Mya truncata - Hiatella arctica Fig. 7. Amino acid aIle/Ile 0,02 0,04 0,06 0 , o a 0 , l O 0 , 1 2 ratios in all measured allellle ratios in the hydrolysed fraction samplcs from weatcm Kongsdya. M y a triiiicata, but some fragments could not be identified. O n the basis of amino acid ratios (method described by A n d r e w et al. 1985) they were identified as Hiatella sp. and My" sp. The sample material falls into two groups of different age (Fig. 7 ) . The older group. with an average of 0.083 I 0.00Y in the hydrolysed fraction for 8 samples, of shells from unit B in section 15-30B in Hidalen and from section 23-01 in Bogen. as well as a fragment from unit E in section 15- 30A. The younger group. with an average of 0.043 5 0.009 for the hydrolysed fraction for 12 samples. consists of shells from unit D at sites 15- 3OC. 15-.71 and 14-03. together with fragments sampled from unit E in section 15-30A. T h e relatively large variations in ratios within the groups might reflect different temperature history as a result of burial depth as some of the samples were found i n scree material at or close to the surface. S I X samples. coming from unit D but collected from surface scree. yielded an average ratio of 0.045 5 0.007 for the hydrolysed fraction and 0.234 5 0.038 for the free fraction. Two samples collected in situ from the same deposit. near the base of the active layer at 0.1- 0.6 m in the sediments, yielded an average ratio of 0.031 i- 0.003 for the hydrolysed fraction and I ) . 1Y3 _t 0.007 for the free fraction. We consider the mean aIle/Ile ratio from in situ shells as more reliable than the total mean for the younger group. The amino acid ratios confirm the strati- graphical evidence of two ice-free intervals, dur- ing deposition of units B and D. in Hidalen. T h e amino acid ratios suggests that the marine beds in Bogen correlate with unit B in Hidalen. Correlating the Kongsgya stratigraphy to Spitsbergen Correlations between different sites on Spits- bergen is difficult because of lack of reliable absol- ute dating methods. Therefore such correlations of possible Early Weichselian and Eemian sites have relied heavily on amino acid ratios (Miller et al. 1989; Sejrup bt Larsen 1991, Mangerud bt Svendsen 1992; Landvik et al. 1992). Even if other factors than diagenetic temperature may influence on the o n the degree of epimerisation, the temperature is the most crucial (Sejrup bt Haugen 1994). Therefore, in order to use the amino acid ratios for regional stratigraphic cor- relations and establishing relative chronologies, the temperature history has to be comparable from site t o site (Sejrup 1990). As demonstrated with the data from Kongssya, burial depth during ice-free conditions can seriously affect the aIle/ Ile ratios o n samples of same age. A composite stratigraphical scheme for Kongs- Oya is shown in Fig. 8. T h e T L determination from Bogen indicates that the older ice free interval o n Kongscdva is older than isotope stage 5e. T h e Late Quaternary glacial and environmental history of Kongsgya, Svalbard 137 I -ocal Composite unit ! lithological log - C B shells I 1 shells Depositional event Sublittoral deposition Subglacial deposition t > 4 0 000 Littoral deposition C .r p m~ a : >38 000 Shallow-marine deposition >40 000 Subglacial deposition Littoral deposition Sublittoral deposition 148klC Shallow-marine deposition Subglacial deposition Mean allellle ratios - 0.031 i0 003 0.083 io.009 Voposed :hronostratigraphy Holocene Late Weichselian Early Weichselian or Eemian Pre-Eemian Fig. 8. Composite stratigraphy of western Kongsbya. Unit thicknesses are arhitrary. Lithological legend in Fig. 7 Table 6. Amino acid ratios in pelecypod shells from sites on Spitsbergen. For comparison. the amino acid values from Kongsaya and their proposed age. Isotope stage Chronostratigraphy Br~ggerhalvbya Kapp Ekholm Linntdalen Skilvika Kongsaya 1 Holocene 0.01s t 0.003 2 Late Weichselian 3 Middle Weichselian 0.026 k 0.004 0.019 t 0.003 4 Sa Early Weichselian 0.031 2 0.003 0.027 ? 0.003 0.031 t 0.003 5h s c 0.0h3 ? 0.013 0.037 t 0.00X Sd Se Eemian 0.044 * 0.004 0.069 t 0.008 Pre-Eemian 0.121 t 0.009 1).083 i- 0.009 Based on Miller et al. (198Y), Mangerud & Svendsen (1992). Lenne & Mangerud (1991) and Landvik et a l . (1992) mean amino acid ratios are considerably higher than those obtained from assumed Eemian austlandet (Blake 1989). deposits on Spitsbergen (Table 6). and we suggest that unit B pre-dates isotope stage 5e. Amino acid ratios in the same range as the older group on Kongsoya have been reported from Nord- If the glacier advance which deposited unit E , the uppermost till, correlates with a Late Weich- selian Barents Ice Sheet, the younger ice free interval. unit D. could possibly correlate with the Kapp Ekholm interstadial. Mangerud & Svend- sen (1Y97) assign i t the age of 1O.OOO-~o.OOO BP and argue that then the whole Barents Sea area was ice free. The aIle/'Ile ratios obtained from i n situ shells from unit D on Kongssya are slightly higher than mean ratios from the Kapp Ekholm interstadial deposits. Since the Kongssya samples come froni altitudes of between ca 55 and 75 m a.s.1.. as compared to 4 5 m a.s.1. for the Kapp Ekholm samples. and the annual mean tem- perature is lower on Kongssya than at Kapp Ekholm. one would expect that similar aIle/lle ratios v.ould indicate higher age for the Kongssya deposit\. We suggest that unit D sediments are either o f Early Weichselian or Eemian age. If our correlations are correct. our data support the conclusion of Sejrup 6r Larsen (1991) that a hio- stratigraphical distinction between interglacial and interstadial ciepo5its at such high latitudes may be difficult. Summary and conclusions The stratigraphic data from Kongssya show three coarsening-upwards successions of marine to lit- toral sediments. separated by tills. These show at least three Late Quaternary major glaciations of the northern Barents Sea. with ice thick enough to cause substantial glacioisostatic downpressing of western K o n g s ~ y a . T h e high (2100 m) post- glacial marine limit. dated to around 11).000 BP. suggests that this part of the Barents Sea was covered by a n ice sheet during the Late Weich- selian. and the uppermost till in the Kongssya stratigraphy is related to that glacial event. Glaciotectonic deformations suggest that ice at some time moved independently of the local topography from an ice-divide northeast of Kong- s0ya. The stratigraphical evidence shows two pre- Holocene ice-free periods. where the climate was similar to or slightly warmer than at present. The a_ge of these ice free intervals is not clear. We suggest that the older ice free period predates isotope stage 5e. T h e younger ice free period could either be of Eemian or Early Weichselian age. -Ickriu\s Itdgernerirs. - We are grateful to the Sorwegian auth- uritier for d l o n i n g u s to d o fieldwork in Kongsaya within the northed\f Svalbard nature reserve. as well as for logistic supporl. The Nor\\sgian Polar Institute supported us in various ways. A . Hansen looked for foraminifera in some of our older sample5 S . Funder analysed the mollusc species. 1. Karnefelt and R . Danielsson provided information on the present fauna and flora of Kongsova. The Swedish Natural Sciencc Research Council financed the re5earch of 0. Ing6lfsson. and has given generous grants for this project. The participation of F. Kognvaldsson and H . P . Sejrup was made possiblc by contributions from the Research Council of Norway, NorsL Hydro, Saga. Statoil and the Norwegian Petroleum Directorate. The Polar Research Secretariat ol Sweden gave various logistics support. The Department of Quaternary Geology. Lund Univcrsity. pro- vided funds to pay for two of the " C dates. Helicopter transport waq financed through a grant from the European Commission to ihc European Science Foundation. The paper bencfitcd from comments made by W. Blake. S . Funder and S . Forman. This is a contribution to thc European Science Foundation project: Polar North Atlantic Margins. Late Cenozoic Evolution ( P O N X h l l References Andrews. J . T.. Miller. G . H . , Davies. D. C. & D;ivie\. K . H 1985: Gcneric identification of fragmentary Quatcrnary molluxs by amino acid chromatography: a tool for Quat- crnar) and Paleuntological research. Grol. J . 2 0 . I-2(1 Blahe. W . 1989: Radiocarbon dating by Accelerator Mdsa Spec- tometr): a confribution to the chronology of Holocene cbcnts in Nordaustlandet, Svalhard. Georgr. Ann. 71A, 59-74. Boulton. G . S 197'): Glacial history of the Spitsbergcn archi- pelago and the problem of a Barents Shelf ice sheet. Boreus s. 31-57, De Geer. G . 1900: Om dstra Spitshcrgens glaciation under iatiden. Geol. Foreri. Srockholm Forharid. 22, 427-136 Eherhoi. A . 6: Solheim. A . 1983: The Barents Sea ice sheet - a sedimentdogical discussion. Polar lies. I , 23-42. El\erhoi. A , . Nyland-Berg. M . Russwurm. L . Rr Solheim, A . 19W: Late Weichselian ice recession i central Barents Sea. Pp. 29g-307 in Bleil. U . & Thiede. J . (eds.): Geological Hisrorv of the P o l a r ocean.^. Kluwer Academic Publishers. Dordrecht. E h r r h o i . A,. Fjeldskaar. W.. Solheim, A . . Nyland-Berg. M . 6: Russwmrm. L. 1993: The Barents Sea Ice Sheet - a model of its growth and decay during the last ice maximum Quar. S1-r. Reo. 1-7. 86.%874. Feyling-Hansaen, R . W. & Ulleberg. K . 1984: A Tertiary- Quatcrnar) section at Sarsbukta. Spitsbergen. Svalhard and its toraminitera P o l a r Re.s. 2. 77-106. Fitter. K . & Manuel. R . 1986: Field G u i d e to rhe Freshwater [.iff of Britarri and N o r t h W e s t Europe. Collins. London. 382 PP Forman. S . L . 1990: Pix-glacial relative sea-level history of north-western Spitsbergen, Svalbard. Geol. Soc. ofAm. Bull. 1 I).? , 1580- 1SW. Forman. S L . . Lubinski. D , Miller. H .Snyder. I . . Matishov. G . Korsun. S . & Myslivets. V . 1995: Postglacial cmrrgence and distribution of late Weichselian ice-sheet loads in the northern Barents and Kara Seas. Russia. Geol. 2 3 . 113-116. Hansen. A . & Knudsen. K. L. 1992: Recent foraminifera i n Freemansundet. eastern Svalbard I i i Moller. P.. lljort. C . 6: Ingolfsson. 0. (eds.). Lundqua Report 35. 177-179. Harde. R. W. 1984: A Field Guidein Colourro Beetles. Octopus B o o k a . London. 334 pp. Hedenas. L. 1989: The genus Sanionia (Musci) in Northwrstern Europe. a taxonomic revision. Ann. Bor. F m n i c i 20. 399- 419. Late Quaternury glacial and environmental history of K o n g s ~ y a , Svulbard 139 Hedenas. L. 1992: A subfossil find of Pseirdoculliergon bre- uifoliimm in northern Sweden. Liridbergia 16, 150-152. Hedenas. L. 1994: A subfossil occurrence of Loeskvpnum wickesii in northern Sweden. Liridbergia 18, 131-134. Hoppe, G . 1972: Ice sheets around the Norwegian Sea during the Wiirm glaciation. AMBlO Spec. Rep. 2: 25-29. Hultkn, E.. &Fries, M . 1986: Atlas of Norih European Vascular Plants. Vol 1-111. Koeltz Scientific Books, Konigstein. 1172 PP . Ingolfsson, 0.. Rognvaldsson, F. & Sejrup, H. P. 1992: The glacial history of western Kongseya. Svalbard. In Moller, P . , Hjort, C. & Ingolfsson. 0. (eds.), Lundqua Rep. 35,25-49. Ingolfsson, 0. & Lirio. J . M. 1993: The glacial geology of Kongseya revisited. Pp. 6 1 1 in Landvik, J. (ed.): P O N A M fieldwork in Sualbard 1YY3. Preliminary Report. Knape, P. 1971; C-14 dateringar av hojda strandlinjer, synkrona pimpstensnivser och iaktagelser av hogsta kustlinjen p i Sval- bard. Unpublished lic. dissertation, S t o c k h o l m Uniuersiiei, Naturgeografiska Institutionen: 142 pp. Lambeck, K . 1995: Constraints on the Late Weichselian ice sheet over the Barents Sea from observations of raised shore- lines. Qirar. Sci. Reu. 14, 1-16. Landvik, J . Y . , Bolstad, M . . Lycke, A. C., Mangerud, J . & Sejrup, H . P. 1992: Weichselian stratigraphy and palaeo- cnvironment at Bellsund. western Svalbard. Boreas 21, 335- 358. Long, D. G. 1992: Sanionia orthoihecioides (Lindb.) Loeske in Scotland, new to the British Isles. 1. Bryol. 17. 111-117. Long, D. G. 1993: Sanionia orthothecioides on the Scottish mainland. J . Bryol. 17, 513514. Lycke, A . K., Mangerud, J . & Sejrup. H. P. 1992: Late Quat- ernary foraminifera1 stratigraphy from western Svalbard. Boreas 21, 271-288. Lenne, I . Rr Mangerud, J. 1991: An Early or Middle Weich- sclian sequence of proglacial. shallow-marine sediments on western Svalbard. Boreas 20. 85-104. Mangerud, J . & Svendsen. J . I . 1992: The last interglacial- glacial period on Spitsbergen. Svalbard. Quar. Sci. Reu. 11, 633-664. Mangerud, J . , Bolstad, M.. Elgersma. A , . Helliksen, D.. Land- vik. J . Y . , L m n e , I . , Lycke, A . K . Salvigsen, 0.. Sandahl. T. & Svendsen, J . I . 1992: The last glacial maximum on Spitsbergen. Svalbard. Quai. Res. 38, 1-31. Millcr. N. G . 1987: Late quaternary fossil moss floras of eastern North America: evidence of major floristic changes during the late Pleistocene-early Holocene transition. Symp. Biol. Hungarica 3S, 34s360. Miller, G. l i . , Sejrup. H. P., Mangerud, J. & Andersen. B.. G. 1983: Amino acid ratios in Quaternary molluscs and foraminifera from western Norway: correlation, geochron- ology and palentemperature estimates. Boreas 12, 107-124. Miller, G . H . , Sejrup. H . P . . Lehman. S. J. & Forman, S . L. 1989: Glacial history and marine environmental change dur- ing the last interglacial-glacial cycle, western Spitsbergen. Boreas 18, 273-296. Mossberg, B . , Stenberg, L & Ericsson. S . 1992: Den nordiskcr floran. Wahlstriim & Widstrand, Stockholm. 696 pp. Nagy. J. 1984: Quaternary glaciomarine deposits and fora- minifera from Edgeoya. Svalbard. Boreas 13, 319-332. Nathorst. A . G . 1901: Bidrag till Kong Karls Lands geologi. Geol. F o r m . Stockholm Forhand. 23, 341-378. Pike, A . 1898. A Cruise on the East of Spitsbergen. Georgr. J . I , 365-371. Polyak, L. & Solheim. A. 1994: Late- and postglacial environ- ments in the northern Barents Sea. west of Franz Joscf Land. Polar Res. 13. 197-207. Rognvaldsson. F. 1992: A m i n o acid geochronology and pala- eorhermometrv of Quaternary sedimenrs on Jameson Lund, East Greenland and Kongsmya. Sualbard. Unpublished Cand. Scient. thesis, Department of Geology. University of Bergen. Norway. 89 pp. Renning, 0. 1. 1979: Sualbards flora. Polarhindbok nr. 1 . Norsk Polarinstitutt, Oslo. 128 pp. Salvigsen, 0. 1981: Radiocarbon dated raised beaches in Kong Karls Land, Svalbard. and their consequences for the glacial history of the Barents Sea. Geogr. Ann. 63A: 2KL291. Salvigsen. 0. Rr Nydal. R. 1981: The Weichselian glaciatmn In Svalbard before 15.000 B.P Boreas 10. 433-446. Schytt. V., Hoppe, G . , Blake, W . . J r . & Grosswald. M. G . 1968: The extent of the Wiirm glaciation in the European arctic. lnternaiional Association of Scientific Hydrologv. Gerr- era1 assemblv of Bern 1967. Publicailon 79, 207-216. Sejrup. H. P. 1990: Amino acid geochronology. I n Medd. Grgnl. Geosci. 22, 36-39. Sejrup, H . P. & Larsen. E. 1991: Eemian-Early Weichselian N-S temperature gradients; North Atlantic - NW Europe. Quar. lnr. 10-12. 161-166. Sejrup, H . P. & Haugen. J . E . 1994: Amino acid diagenesis in the marine bivalve Arctica islandica L i m e from NV-Euro- pean sites: only time and temperature? J . Quar. Sci. 9. 301- 310. Siegert, M. J. & Dowdeswell, J . A . 1995: Numerical modelling of the Late Weichselian Svalbard-Barents Sea ice sheet. @at. Res 43, 1-13. Smith, D . G . , Harland, W . B. Hughes. N. F. & Picton. C. 1976: The geology of Kong Karls Land. Svalbard. Geol. Mag. 113: 193-232. Solheim, A.. Russwurm, L.. Elverhei, A . & Nyland-Berg. M . 1990: Glacial geomorphic features i n the northern Barenth Sea: direct evidence for grounded ice and implication, f o r the pattern of deglaciation and late glacial sedimentation. Pp. 253268 in Dowdeswell. J. A. & Scourse. J . D. (eds.): Glacimaritie Enuironments: P r o c e u a and SedirneriiA, Grol. Soc. London. Spec. Publ. 53. Stromberg. B. 1972: Glacial striae in southern Hinlopcnstrcdct and Kong Karls Land. Svalbard. Geogr. A n n . 54A Soderstrom, L., Hedenas. L. & Hallingback, T . 1992. Chcck- lista over Sveriges mossor. Myrinia 2. 13-56, Vorren, T . 0.. Hald, M. & Lebesbye E . 1988: Late Cenozoic environments in the Barents Sea. Paleocenogr. 3 , 601-612.