Past glaciation and sea levels on BjGrnGya, Svalbard OTTO SALVlGSEN and 0YSTEIN SLETTEMARK Salvigsen, 0. & Slettemark. 0. 1995: Past glaciation and sea levels o n Bjmnaya, Svalbard. Polar Research B j 0 r n ~ y a has a very thin cover of unconsolidated Quaternary sediments. Glacial erratics of local origin are spread throughout the lowland areas, and glacial striae indicate glacial movement which was centred in the middle of the island. No traces of the Barents Sea ice sheet have been observed on B j ~ m 0 y a . nor has there been any postglacial emergence of the island. Lake cores date the deglaciation of the lowlands to ca 10,ooO BP, and peat deposits on high mountains show that these were deglaciated before 8700 BP. Otto Salvigsen and 0 y s l e i n Slettemark, No& Polarinstiturt, Postboks 5072 Majorshta. N-0301 Oslo, Norway. 14(2), 245-251. Introduction The reconstruction of the Late Weichselian ice sheet in the Barents Sea area is still largely based on indirect evidence, such as the pattern of iso- static rebound, undated moraine ridges, general lithostratigraphy and the distribution of over- compacted diamicton (Elverhei et al. 1993; Lam- beck 1995). The objective for the present study was to gather new information on glacial features and former sea levels from B j ~ r n ~ y a , which with its position on the western edge of the Barents shelf may provide key information in the Late Weichselian discussion. Bj0rnoya is the southernmost island in the Sval- bard archipelago, 250 km south of Spitsbergen and 440 km north of the Norwegian mainland (Fig. 1). The island is 20 km from south to north, 15 km from west to east, and has a total area of 178 km2. It is situated on a submarine plateau with depths of less than 100 m. The predominantly soft rocks of Bjomoya are highly affected by wave abrasion, and the coasts therefore consist almost entirely of steep cliffs (Hogvard & Sollid 1988). The island can be divided into three geo- morphologic regions (Fig. 1): the northern plain containing about 700 shallow lakes, the southern mountains which attain an altitude of 440 m a.s.l., and the eastern mountain, Miseryfjellet. with its highest peak at 536 m a.s.1. The northern plain is a typical strandflat (Nan- sen 1922), most extensively developed at about 35 m a.s.1. Most of the lakes are small and only a few metres deep, and they are the result of selec- tive glacial erosion. The southern parts of the island consist of dolomites, limestones and shales, while the northern plain and Miseryfjellet are mainly built up of coal-bearing sandstones and limestones (Horn & Orvin 1928; Worsley & Edwards 1976). Karst phenomena are common in the limestone areas. Large scale features such as dolines, sub- terranean drainage, caves and surface collapse features occur in several places. U-shaped valleys, such as Ymerdalen (Fig. 2) in the southern part of the island, demonstrate former glacial activity. The permafrost is 60-70 m thick, and the upper 0.75 m thaws during the summer (Horn & Orvin 1928). This paper is based on one month's fieldwork in 1993. Transport in the field was exclusively by foot, and the entire island was surveyed. Exact positions of observations were obtained by a GPS receiver. Glacial sediments and erratics The island is covered by a thin and discontinuous layer of Quaternary sediments, and large auto- chthonous block fields characterise much of the sandstone areas of central Bjornoya (Fig. 3 ) . A thin (0.5-1.0 m) diamicton is locally exposed in sections along the coastal cliffs. Due to its com- position and in many cases direct presence above 216 0. Saluigsen & 0. Slettemark I 74: F-i,q. I . Bprnoya. Distribution of o h s e n e d glacial striae on bedrock with indication of the direction of movement (toward the observation point. wild black circle). The dotted line on lnset map I S the 500m contour h e which indicates the p o w i o n of the shelf edge Past glaciation and sea levels on B j ~ r n @ y a , Soalbard 247 Fig. 2. Bj0rncaya seen from the south. The partly peat- covered mountain Hambergfjellet in the foreground, and the U - shaped valley Ymerdalen to thc right of it. glacier sculptured rocks, i t is interpreted as a till, often disturbed by subsequent cryoturbation. Numerous cobbles and boulders, some with glacial striae (Fig. 4), are found throughout the island, except in the upper mountain areas. The highest observations of erratics are from Ant- arcticfjellet, about 360 m a.s.1. (Fig. 5). Horn & Orvin (1928) also reported erratics on the north- ern plain, for example, a six-metre high heap of sandstones named Knorten south of Kapp Duner, probably the frost-shattered remains of a very large sandstone boulder deposited there by glacial ice. All erratics on B j ~ r n o y a seem t o be of local origin and have been transported only a short distance. Traces of marine material on land should be expected if ice from a Barents Sea ice sheet had overridden the island during the last glaciation. During o u r extensive survey we thoroughly searched for shell fragments and other marine material, but found nothing of the kind. Fig. 3. Blockfield area from central BjBrneya. 238 0. Snliiigsen & 0. Slettentnrk N o wb-till sediments o r other unconsolidated sediment5 suspected t o he older than the last glaciation were tound Glacial striae The distribution ot observed glacial striae shown in Fig. 1 give5 a true image of the quantitative and spatial occurrence of glacial striae on Bjarnava. Glacial striae were probably fir5t observed h! A . G . Nathorst i n 1870 (Nathorst 1899). on the westernmosr cape3 of the island. O n e of these F i g . 4 Striated IxxhAx from the intcrior of Bjomoya. C o r n p a s for scale. capes still has the pame Rifleodden (Riiffeludden) which means "striation cape". The map of Horn & Orvin (1928) shows the direction of glacial striae observed at about ten different sites. Satersmoen & Hovden (1984) made observations on the Quaternary geology of B j ~ r n ~ y a , and their unpublished report contains another forty obser- vations of glacial striae, mainly from the middle and western parts of the island. W e searched the entire island for glacial striae and made about one hundred new observations (Salvigsen & Slette- mark 1993. 1994). T h e glacial striae on Bj@rn@ya are best pre- Fig. 5 . Erratic block o n Antarcticfjellet Past glaciation and sea levels on Bj@rn@ya, Svalbard 249 Fig. 6. Striatcd sandstone surface near Miketjerna, easternmost Bjomoya. Glacier movement from the W (from the left on the photo). Compass for scale. served in the sandstone areas, whereas the lime- stone surfaces seem t o have n o potential for long time preservation. Most of the striae were found after digging, brushing and washing bedrock sur- faces (Figs. 6 and 7), and only a few sites with exposed striations were found o n sandstone sur- faces north of Miseryfjellet. Ice movement directions were in most cases determined from small scale stoss and lee mor- phology, but crescentic gouges and large scale features such as roche moutonnkes were also used. N o glacial striae occur in the mountainous areas Fig. 7. Crossing glacial striae from southwestern Bjerneya, near Ellasjeen. Glacier movement from ENE and the SE (from the left on the photo). The pencil shows direction of the oldest striae, from the SE. and the compass shows the direction of the younger straie. of southern and western B j ~ r n ~ y a , due mainly t o the high degree of weathering on rock surfaces there. T h e northwestern part of the island, with its limestones and calcareous sandstones, is mostly covered by swamps, and the potential for finding striae there is very limited. O n e of the main objectives of our work was to determine whether or not B j ~ r n ~ y a had been covered by or infringed upon by the last Barents Sea ice sheet. With respect to glacial striae. the coast from Miseryfjellet t o Nordkapp was most interesting in this context, because any striae from the Barents Sea glacier in this area would be easy 250 0. Salvigseri & 0. Slettemark to distinguish from those formed by local glaciers. There a r e numerous roche moutonnees here with more or less well-preserved striae and other small scale glacial features. Crossing striae also occur in many places. Nearly all the glacial striae o n the eastern part of the island indicate ice movement from the island toward the E and N E , but a more detailed study of age relationship shows that also older striae from an ice movement toward the N a r e preserved in some places. They occur in lee positions and reveal a systematic pattern of development in this area. Further west, striae towards NW and S-SW are found, but they a r e also compatible with a local ice sheet centred in the middle of the island. The glacial striae shown o n the m a p are of course not all synchronous. but the age dif- ferences are probably not large. The striae in the periphery of the island may be somewhat older than those in the middle of the island. The oldest striae seem to be from a glacier centred on the northern slope of Miseryfjellet. Most probably a successive displacement of the ice-divide to the middle of the island, northwest of Miseryfjellet . then followed. The most numerous traces of glaciation on Bjciirnciiya are from glaciers centred there. T h e greatest thickness of ice then probably occurred near the geographical centre of the island. Most of the glacial features on Bjsrnciiya seem to be from rather rigid ice. and plastically sculp- tured forms and features d o not occur. T h e glacial features indicate a modest thickness of the ice cover. and the highest mountains were not necess- arily covered by this local ice cap. The main conclusion, however. is that there are no traces of a Barents Sea ice sheet on Bjernciiya. T h e many observed glacial features only indicate activity from local glaciers. T h e presence of such a regional ice sheet o n the island at an earlier stage cannot be ruled o u t , but if it existed, the activity of local glaciers has erased all traces of it. Age of the last deglaciation W e know little about the age of the glacial striae, but most probably they represent the latest de- glaciation phase o n the island. T h e best minimum ages for the deglaciation of Bjsrneya are from datings o n lake sediments. The first lake corings o n Bjsrn0ya were carried out by Hyvarinen (1968). who cored eighteen lakes and obtained five conventional radiocarbon dates between 8900 and 11,200 yrs BP. However, because of possible contamination with isotopically "dead" carbon from t h e local rocks, the accuracy of these dates could be questioned; Hyvarinen concluded that they at least suggested that the sediments in the B j ~ r n ~ y a lakes covered the last 10.000 years. During the PONAM 1Y93 expedition, new cores were taken from some of the previously cored lakes. a n d AMS radiocarbon dates on plant macro fossils (Bryophytes) gave ages u p to 9800 years BP (Wohlfarth et a]. 1995). confirming the conclusion of Hyvarinen (1968) that the sedi- mentation in the lakes started about 10.000 years ago. O n the summit plateau of Hambergfjellet (440 m a.s.1.) (Fig. 2). there is an extensive area of 1-3 m high. peat-covered and ice-cored hillocks (Engelskjan 1987). Similar mounds also occur on t o p of Fuglefjellet (410m a.s.1.). Dark brown peat. as well as underlying ice, a r e exposed by erosion in many places, but near-surface perma- frost prevented further investigation by digging. T h e frozen peat or the ice lenses seem to rest directly on coarse limestone gravel. Samples of peat and moss remnants have been dated, and the bottom sample from a section on Fuglefjellet was dated t o 8705 t 120 BP (T-11227). This gives a minimum age for the first peat formation, and thus also for the deglaciation of the higher parts of B j o r n ~ y a . Sea levels T h e first geologists who visited B j e r n ~ y a found n o traces of marine action higher than the present beach zone (Anderson 1900). Horn & Orvin (1928), however, advocated a contrasting view and concluded (p. 142). "The last movement of the land in post-glacial times consisted in a sinking to about 35 metres above the present level, followed by an uplifting to t h e present level." However, their observations of rounded pebbles interpreted to have been formed by beach pro- cesses a r e highly disputable and have never been confirmed by others. O u r investigations on Bjciirn~yn lead us to the conclusion that there are no traces of raised beaches o n t h e island. In some places, especially along the coast north of Miseryfjellet, till-covered Pust glaciation and sea levels on Bj@m@yu, Soulburd 251 bedrock surfaces slope modestly and evenly towards the sea until they reach the present beach, revealing that after the deglaciation no sea levels occurred higher than that of the present. The same conclusion can b e made from other sites o n the northern and western coasts, where low land meets the sea. T h e area near the outlet of the small river named Benda, on the southernmost plain by the western coast, also demonstrates the non-marine character of the landscape above the present beach zone. A t this site small, modern driftwood logs and other flotsam have been thrown by the waves up to 10 m above sea level, which as far as we know is the highest swell activity observed in Svalbard. Above that there a r e n o traces of former sea levels. It can therefore be concluded that since the Late Weichselian deglaciation t h e eustatic sea level rise must have exceeded the isostatic recovery of B j ~ r n ~ y a . In 1864 Nordenskiold placed an iron bolt in a rock outcrop in Russehamna (Fig. 1) and measured the contemporaneous sea level (see Horn & Orvin 1928). Accurate measurements and calculations were thereafter carried o u t in 1922 (Horn & Orvin 1928), and n o change of sea level was found. At the same time, elevations of a number of triangulation points were determined in relation to the new sea level measurements. These altitudes have recently been compared with the results of tidal gauge measurements through- out o n e year, 1989-1990. The conclusion was the same as in 1922, that no measurable sea level changes have taken place on Bjornoya in the intervening period (Trond Eiken pers. cornm. 1995). Conclusions Only striae from local glaciers occur on B j B r n ~ y a . All displaced rocks found are from the island itself, and n o sediments or rocks from the sur- rounding sea bed were transported onto the island during t h e last glaciation. T h e Late Weichselian glaciation of B j ~ r n o y a has been a local one. and the regional Barents Sea glaciation did not affect the island. T h e likely age of deglaciation on B j ~ r n ~ y a is No raised beaches or other traces of sea levels about 10,000 yrs BP. higher than the present occur o n B j ~ r n ~ y a . Acknowledgements. - We are thankful to the people who were involved in our transport to and from Bjerneya by respectively the Norwegian Coast Guard vessel GRIMSHOLM and the Norwegian Polar Institute vessel LANCE. The staff of Bjerndya Radio were most helpful in allowing us to use their huts on the island and the facilities of their station. Thanks also to C. Hjort, J . Y . Landvik and K. Malmberg Persson for valuable comments to preliminary versions of the paper. References Anderson, J . G. 1900: Den svenska expeditionen till Beeren Eiland sommaren 1899. Ymer 20, 42-454. Elverhei, A . , Fjeldskaar, W . . Solheim. A . Nyland-Berg. M . & Russwurm, L. 1993: The Barents Sea ice-sheet: A model of its growth and decay during the last ice maximum. Q . Sci. Reu. 12, 86S973. Engelskjen, T. 1987: Eco-geographical relations of the Bjern- Oya vascular flora, Svalbard. Polar Res. 5 n . s.. 79-127. Horn, G. & Orvin, A. K . 1928: Geology of Bear Island. Norsk Polarinst. Skr. 15. 152 p p . , 9 pl.. 1 geol. map. Hyvarinen, H . 1968: Late-Quaternary sediment cores from lakes o n Bjerneya. Geogr. Ann. 50A. 235-245. Hegvard, K. 6i Sollid, J . L . 1988: Kystkart Svalbard B 5 Sdrkapp 1 :200,000. Norsk Polarinst. Temakart 10. Lambeck, K . 1995: Constraints on the Late Weichseliian ice sheet over the Barents Sea from observations of raised shore- lines. Quat. Sci. Reo. 14, 1-16. Nansen, F. 1922: The strandflat and isostasy. K . Norske Vidensk. Selsk. Skr. 1 . 1 1 . Kristiania. Nathorst, A. G. 1898: Om 1898 Brs svenska polarekspedition. Ymer 18, 321-348. Nathorst, A . G . 1899: Nigra upplysningar till den nya kartan over Beeren Eiland. Ymrr 19, 171-185. Salvigsen. 0. & Slettemark, 0. 1993: Glacial geology of Bjern- eya, Svalbard. Pp. 42-45 in Landvik, J . (ed.): PONAM field work in Svalbard 1993. Preliminary Report. Salvigsen, 0. & Slettemark. 0. 1994: Nedisningsspor, spesielt isskuringsstriper. p i Bjerneya, Svalbard. Abstracfs 21: N o r - diska Geologiska Vintermotet. Lulei. P. 177. Sztersmoen. B. H . & Hovden, 0. 1984: Rapporrfra oppliold pd Bj@rn@ya sommereii 1984. Report to Norsk Polarinstitutt and University of Bergen. 10 pp, 1 map. Unpublished. Wohlfarth. B . , Lenidahl, G . , Olsson. S . , Persson. T . , Snowball, I . , Ising, J . & Jones, V. 1995: Early Holocene environment on Bj0rneya (Svalbard) inferred from multidisciplinary lake sediment studies. Polar Res. 1 4 ( 2 ) , 253-275. Worsley, D. & Edwards, M . B. 1976: The Upper Palaeozoic succession of Bjerneya. Norsk Polarinsf. Arbok 1974, 17-34.