Norsk Geografisk Tidsskrift – Norwegian Journal of Geography Vol. 56, 56–61. Oslo. ISSN 0029-1951

Radiocarbon-dated Mytilus edulis and Modiolus modiolus from northern Svalbard: climatic implications

OTTO SALVIGSEN

Salvigsen, O. 2002. Radiocarbon-date d Mytilus edulis and Modiolus modiolus from northern Svalbard: climatic implications. Norsk GeograŽ sk Tidsskrift–Norwegian Journal of Geography Vol. 56, 56–61. Oslo. ISSN 0029-1951. Eighteen new age determinations on Mytilus edulis shells from northern Svalbard and two new age determinations on Modiolus modiolus shells are presented. Mytilus edulis appeared on the northern coast of Spitsbergen c. 9400 yr BP. Its maximum extension there seems to have been c. 7250 yr BP, which is also indicative of a climatic optimum. The Ž rst Ž nding of Modiolus modiolus shells from northern Spitsbergen is reported, and revealed the age of c. 8300 yr BP, which is also indicative of marine climatic optimum conditions . Mytilus lived at the head of Woodfjorden until at least 5300 yr BP, but it is not possible to date its Ž nal dis- appearance from the northern coast. Mytilus has been absent from large parts of the coasts of Svalbard during the entire Holocene, indicating limited in uence of warm Atlantic water there. Recent reports of living Mytilus edulis on Bjørnøya indicate a probable new occurrence there, and the question about a further immigration to Spitsbergen should be kept in mind. Such a faunistic change may provide data about a more general climatic warming in this part of the Arctic. Keywords: Mytilus, Modiolus, palaeoclimate , radiocarbon dates, Svalbard Otto Salvigsen, Department of Geography, University of Oslo, Postboks 1042 Blindern, NO-0316 Oslo, Norway. E-mail: ottosa@ geograŽ .uio.no

Introduction However, more detailed information from many localities is of course still of great interest. It should also be mentioned The occurrence of raised beach deposits with shells of that Mytilus edulis shells are known from interstadial Mytilus edulis (common mussel) in Svalbard has been known (Eemian) sediments at Kapp Ekholm, central Spitsbergen since 1861 (Blomstrand 1864, Chydnius 1865), when it was (Mangerud & Svendsen 1992, Funder 1993). They are, found in Isfjorden as well as in Sorgfjorden, northernmost however, beyond the scope of this paper. Spitsbergen. Since this is a thermophilous mollusc and not included in the present fauna of Svalbard, the palaeoclimatic implications of its former occurrence have been broadly discussed (see Feyling-Hanssen 1955, Salvigsen et al. 1992, Sampling and age determination Hjort et al. 1995). Most of the dated samples presented in Table 1 were After the introduction of the radiocarbon method for age collected during Ž eldwork in 1995 and 1997. The purpose determination, the number of dated samples of shells steadily of the sampling was to expand the area of dated Mytilus grew, and 34 dates on Mytilus edulis from Svalbard were edulis localities in northern Svalbard. The aim was primarily presented in Salvigsen et al. (1992). Most of them were from to sample the oldest and the youngest shells in a speciŽ c area, the west coast and the central fjord areas of Spitsbergen, respectively the highest lying and the lowest lying shells. showing that Mytilus edulis lived there from before 9500 yr Mytilus shells and shell fragments were found either in situ in BP to c. 3500 yr BP. Thereafter it probably brie y littoral cliff sections, or in raised beach gravel. In most cases, reimmigrated during the ‘Medieval warm period’ c. 1000 large samples were collected and whole shells or large yr BP. Only three dates, all >9000 yr BP, were from the fragments of good quality were sent to the dating laboratory northern coast of Svalbard. Salvigsen et al. (1992) also in Trondheim. The ages, determined by the conventional presented three radiocarbon dates on shells of Modiolus radiocarbon method (except TUa-2191 which is an AMS modiolus (horse mussel) from the shores of Isfjorden, with date), are corrected for a reservoir effect of 440 years ages from c. 8700 yr BP to 7700 yr BP which were supposed (Mangerud & Gulliksen 1975). They have¡also been to indicate a marine climatic optimum in this period. corrected to a d13C = 25% PDB, in most cases by using a 13 ¡ Hjort et al. (1995) presented 11 age determinations on standard value of d C = 1.0 for shells. Mytilus edulis from Edgeøya and one such date from Only ages of radiocarbon years are presented in this study, Svenskøya, Kong Karls Land. They all reveal ages between mostly for practical reasons. The referred literature uses c. 8800 yr BP and 5000 yr BP and roughly delimit the marine radiocarbon years, and comparisons with former studies are climatic optimum period in eastern Svalbard. The coloniza- facilitated by sticking to radiocarbon years. tion of Mytilus in eastern Svalbard was supposed to be an effect of a maximum in ux of Atlantic water to the region. The purpose of this paper is to follow up the information given in Salvigsen et al. (1992) and Hjort et al. (1995) and Results discuss further the matters dealt with there. The paper shows Immigration of Mytilus edulis our present knowledge about the Holocene distribution of Mytilus edulis and Modiolus modiolus in northern Svalbard. Table 1 lists 22 radiocarbon dates on Mytilus edulis, of which

# 2002 Taylor & Francis NORSK GEOGRAFISK TIDSSKRIFT 56 (2002) Radiocarbon-date d Mytilus edulis and Modiolus modiolus from northern Svalbard 57

Table 1. Radiocarbon-date d shells of Mytilus edulis and Modiolus modiolus from northern Spitsbergen. Hannevigodde n is located in southern Spitsbergen. The ages show radiocarbon years with one sigma and corrected for a marine reservoir age of 440 years (Mangerud & Gulliksen 1975). ¡ Site Age Lab. no. Field no. m a.s.l. Comments

1. Reinsdyr ya 9375 80 DIC-3076 SL84-33 9 Salvigsen et al. 1992 2. Vogtvatnet 9360 § 110 T-3098 Sa 78-17 8 Salvigsen & O¨ sterholm 1982 3. Mushamna 9335 § 45 T-12236 Sa 95-95 5 Littoral section § 4. GraÊ huken 9295 70 HD15707-15601 SPE-22a 5 Bru¨ckner 1996 5. Sjøvernbukta 9180 § 65 T-12370 Sa 95-90 1 Fragments on modern beach 6. Wigdehlpynten 9135 § 125 T-13548 Sa 97-85 18 Fragments on raised beach 7. Langgrunnodden 9070 § 190 U 173 B-47 9 Blake 1961 8. Mosselbukta 9065 § 130 T-12840 Sa 79-09 9 Raised beach section 9. Kapp Kjeldsen 9060 § 75 T-12227 Sa 95-69 4 Raised beach 10. Mossellaguna 9015 § 105 T-13255 Sa 97-32 12 Raised beach section § 11. GraÊ huken 8985 60 T-12237 Sa 95-98 3 Littoral section 12. Worsleyhamna 8865 § 65 T-12371 Sa 95-04 4 Fragments on raised beach 13. Danielodden 8845 § 105 T-13251 Sa 97-17 9 Littoral section 14. Breiddholme n 8170 § 120 T-13257 Sa 97-49 5 Raised beach sediments 15. Mosselbukta 8165 § 70 T-13551 Sa 97-11 3.5 Fragments on raised beach 16. Karlsbreen 7700 § 75 T-12230 Sa 95-77 40 Fragments on frontal moraine § 17. BraÊ nevatnet 7260 105 T-13426 Sa 97-100 10 Raised beach 18. Skamdalen 6935 § 105 T-14497 1106-3 6 Littoral section 19. Halvdandalen 6620 § 80 T-12369 Sa 95-84 1 Fragments on modern beach 20. Karlsbreen 6255 § 45 T-12229 Sa 95-78 11 Fragments on frontal moraine 21. Forkdalen 5655 § 100 T-13260 Sa 97-74 5 Littoral section 22. Breiddholme n 5310 § 60 T-113258 Sa 97-51 1 Shells on modern beach 14. Breiddholme n 8285 § 95 T-13256 Sa 97-46 5 Modiolus on raised beach Hannevigodden 6555 § 70 TUa-2191 Sa 87-44 4 Modiolus on raised beach § four have been published previously. The locations of the and numerous and constitute real Mytilus beds in sea-facing sampling sites are shown in Figs. 1 and 2. sediment cliffs. The four oldest dates are from the outer fjord area of Bockfjorden, the smallest branch of Woodfjorden, was Woodfjorden or from the western mouth of Wijdefjorden invaded by Mytilus edulis communities well before 9000 yr (site 2) and indicate that Mytilus edulis Ž rst settled here after BP, as shown by dated samples from Sjøvernbukta (5) and passing the NW corner of Svalbard on its way from more Kapp Kjeldsen (9). The former sample was from the modern southern areas. This immigration took place c. 9400 yr BP, beach, where shell fragments had been washed ashore after which seems to be only 200 to 300 years later than its Ž rst wave erosion of shallow sublittoral sediments. The latter appearance on the west coast (Salvigsen et al. 1992). In two sample, however, was picked from the highest-lying shell cases, Vogtvatnet (2) and Mushamna (3), the shells are large fragments on or near the surface of raised beach sediments. Mytilus edulis also lived in the inner part of Woodfjorden before 9000 yr BP. The highest-lying fragments at Wigdehl- pynten (6) show the minimum age for the Mytilus immigra- tion there.

Fig. 1. Location map of Svalbard showing sites where Mytilus edulis, black dots, and Modiolus modiolus, black triangles, have been radiocarbon-dated , illustrating the geographical distribution of shells from these thermophilous Fig. 2. Location map of parts of northern Svalbard showing sites from where molluscs in Svalbard based on Salvigsen et al. 1992, Hjort et al. 1995 and this Mytilus edulis and Modiolus modiolus have been radiocarbon-dated. study. Numbers refer to Table 1. 58 O. Salvigsen NORSK GEOGRAFISK TIDSSKRIFT 56 (2002)

The datings mentioned here indicate that Mytilus edulis Holocene than previously known (Salvigsen et al. 1992). Ž rst appeared on the northern coast c. 9400 yr BP, and during The shell was dated to 8285 95 yr BP (T-13256), which is the following 200 years it spread to Bockfjorden as well to the same age as nearby lying§Mytilus edulis shells from the inner Woodfjorden. During this period it also spread same level on Breiddholmen, 8170 120 yr BP (T-13257). eastwards to Langgrunnodde n (7) on Nordaustlandet. This age Ž ts well into the interval of§former dated Modiolus Three samples of the highest lying Mytilus shells and shell shells from Svalbard and supports the idea of the marine fragments from western Mosselhalvøya (8, 10 and 13) were climatic optimum, 8700–7700 yr BP (Salvigsen et al. 1992). assumed to represent an early appearance of Mytilus there. However, a new age determination on Modiolus shells from The oldest, from Mosselbukta (8), revealed the same age as southern Spitsbergen should also be mentioned. A few large that of the shells from Langgrunnodden, indicating a more or shell fragments of this mollusc were found in beach gravel on less contemporaneous settlement of Mytilus on these Hannevigodden, Wedel Jarlsberg Land. The largest fragment localities. To date, the southernmost known observation of has been dated and revealed the age 6555 70 yr BP (TUa- Mytilus in Wijdefjorden is from Skamdalen (18), but it is not 2191). This indicates that Modiolus modiolu§ s had a more known how far southwards it was able to penetrate in this extensive distribution southwards in Spitsbergen than pre- fjord (Fig. 2). viously known. It also lived in Spitsbergen waters consider- ably later than other datings of this mollusc indicate. Because of the few age determinations on Modiolus it cannot be stated if its presence has been sporadic or continuous from 8700 yr The marine climatic optimum BP to 6500 yr BP. The most conspicuous of the new dates is probably that from BraÊ nevatnet (17) at the head of Wahlenbergfjorden in Nordaustlandet. The occurrence of Mytilus edulis shells Extinction of Mytilus edulis in northern Svalbard there is, in many ways, remarkable. It was Ž rst reported by Sandford (1929) from morainic boulder clay at the northern A crucial question is how long Mytilus edulis molluscs were shore of the fjord about 55 m a.s.l. Another site with Mytilus able to live in northern Svalbard at the limits of their edulis shell is located near the outlet of the small lake northernmost extent. Shoreline displacement curves (Salvig- BraÊ nevatnet (17). Wahlenbergfjorden is located between the sen & O¨ sterholm 1982, Forman 1990) show that only two major ice caps of Nordaustlandet, and the local climate is beaches from the Ž rst part of the Holocene are raised above presently much in uenced by the adjacent glaciers. The the present sea level. It is possible to sample the oldest extent of the glaciers during the Holocene warm period is Mytilus edulis shells on land, while deposits with younger unknown, but they were probably much reduced in size shells are still below sea level and not accessible for normal compared to their present position. Some shells for dating sampling. The raised beaches in the area show increasing were sampled by the Russian geologist A. Tebenkov, and uplift from north to south along the fjords (see isobase map in revealed the age 7260 105 yr BP (T-13426), which is Bondevik et al. 1995), and the youngest shells should identical with the age §of the dated shell fragments from therefore be accessible at the head of the fjords. In some Svenskøya, 7260 100 yr BP (AA-R853, Hjort et al. 1995). places, shell fragments from sublittoral deposits are washed However, the livin§g conditions of the molluscs from the two ashore and can be found on the present beach. localities seem to have been quite different. As previously Shell fragments from two such deposits in Woodfjorden mentioned, two localities with abundant Mytilus shells are have been dated, Halvdandalen (19) and Breiddholmen (22), known from Wahlenbergfjorden contrary to Svenskøya, indicating that Mytilus edulis lived in the fjord until at least where it has not been possible to Ž nd more than a few small 5300 yr BP. Breiddholmen is an islet at the head of fragments of Mytilus (Salvigsen et al. 1992). These two dates Woodfjorden mainly consisting of littoral gravel with so lead to the conclusion that Mytilus edulis seems to have been many whole, blue shells and shell fragments that part of it most widespread in Svalbard c. 7250 yr BP. could be characterized as a Mytilus bank. Its maximum Modiolus modiolus has previously been associated with height is c. 6 m above the present sea level. the Holocene climatic optimum in Svalbard (Feyling- The large frontal moraine of Karlsbreen (20) in the valley Hanssen 1955, Salvigsen et al. 1992). Three age determina- of Bockfjorden contains mainly marine sediments thrusted tions on shells from this thermophilous mollusc in Isfjorden, and pushed to higher levels by the last advance of central Spitsbergen, indicate that it lived there from c. 8700 Karlsbreen, probably during the ‘Little Ice Age’. Shells of yr BP to 7800 yr BP. It was then assumed that this period Mytilus edulis occur in these sediments, and the youngest represented the maximum of the marine climatic optimum in dated shells revealed an age of 6255 45 yr BP (T-12229). Svalbard (Salvigsen et al. 1992). The youngest shells from the west sid§e of Wijdefjorden are One, almost complete, valve of Modiolus modiolus was from a low sea-facing cliff in Forkdalen (21) with an age of c. found at the head of Woodfjorden on the islet Breiddholmen 5650 yr BP. The radiocarbon age determinations only show (14). It was visible on the surface of the raised shore gravel, the minimum duration of the Mytilus period in this northern and was found together with many Mytilus shells. It was area, mainly because eventually younger shells are still cracked in several pieces, but the shape of the shell could still below the present sea level. Mytilus edulis lived in Isfjorden be recognized. This is the Ž rst time Modiolus shells have until c. 3500 yr BP (Salvigsen et al. 1992) and in SW been reported from northern Spitsbergen, indicating that this Edgeøya until at least 5000 yr BP (Hjort et al. 1995). Mytilus mollusc had a more extensive distribution during the appeared in northern Spitsbergen earlier than in the waters NORSK GEOGRAFISK TIDSSKRIFT 56 (2002) Radiocarbon-date d Mytilus edulis and Modiolus modiolus from northern Svalbard 59 around Edgeøya. It seems therefore likely that it lived in the similar to that of the fjords in central Spitsbergen (Salvigsen north also after 5000 yr BP, but the time of its extinction et al. 1992). It should also be mentioned that the gastropod there is yet unknown. Littorina littorea, now extinct in Svalbard, is also fairly frequent in shore deposits in the Woodfjorden area. Hjort & Funder (1974) showed that Mytilus edulis lived in Areas without Mytilus edulis shells central East Greenland between 8000 yr BP and 5500 yr BP. It was suggested that the area was settled by larvae transport Our present knowledge about Mytilus edulis shells on raised from Svalbard by the West Spitsbergen current, implying beaches in Svalbard now makes it possible to point out areas greater in ux of Atlantic water along the East Greenland where this mollusc seems to have been totally absent during coast than at present. Water  owing along the west coast of the Holocene, as illustrated in Fig. 1. In Nordaustlandet, it Greenland is partly derived from the East Greenland current, has not been found on either the north coast east of and the surface temperature there is mainly controlled by Langgrunnodde n (7) or on the islands north of Nordaustlan- varying in ow of Atlantic and polar water (Funder & det (Feyling-Hanssen 1955, Salvigsen et al. 1992). Weidick 1991). Conditions in Svalbard and Greenland are It has not been possible to Ž nd Mytilus shells on/in the obviously related, as suggested by the agreement in dating raised beaches along the west coast of Spitsbergen south of results of the thermophilous molluscs in the actual areas. Hornsund, and it has not been reported from the east coast of Spitsbergen. It was not found on Barentsøya during extensive Ž eldwork in 1991 (Hjort et al. 1995), and it seems also to be absent in southern Hinlopenstretet, including Svartknaus ya, Discussion of living conditions southern Nordaustlandet. The many small islands south of The occurrence of Mytilus and Modiolus is usually correlated Edgeøya, Tusenøyane, are poorly investigated, and no Ž nds with summer surface temperatures (SST), but the use of these of Mytilus edulis shells have been reported from this area. species for exact palaeotemperature determination is limited Neither are Mytilus shells found on the island Hopen, SE because biological factors may also in uence the distribution Svalbard. (Navaro & Thompson 1996). However, the varying Holo- cene occurrence of these molluscs indicates distinct relative temperature changes in some areas of Svalbard, especially In uence of Atlantic water along the west and north coast of Spitsbergen, but also on Edgeøya. In other areas, most of Nordaustlandet and east The climatic conditions in the Barents Sea are mainly coast of Spitsbergen, thermophilous molluscs were absent determined by the in ux of Atlantic water (AÊ dlandsvik & during the entire Holocene, indicating that these areas were Loeng 1991, Loeng 1991). The heat transport of the Atlantic less in uenced by warm Atlantic water but dominated by the water has shown considerable variations during the Holocene cold Arctic waters also during the Holocene climatic (Hald et al. 1996, 1999, Duplessy et al. 2001). Climatic optimum in Svalbard. conditions in and around Svalbard are a minor but important Bjørnøya, the southernmost Svalbard island, has no raised detail in the large climatic system of the Barents Sea and beach deposits (Salvigsen & Slettemark 1995), and the adjoining Arctic areas. Today, the western and northern occurrence of Mytilus around this island during the Holocene coasts are the areas on Svalbard most affected by Atlantic is therefore unknown. water, and this general pattern also prevailed previously Until recently, Modiolus and Mytilus have not been found during the Holocene. living in Svalbard waters. They are considered reliable The occurrence of thermophilous molluscs is usually climate indicators and should therefore be useful for the assumed to be connected with an in ow of Atlantic water detection of a possible recent warming of the marine climate warmer than that of today around Svalbard (Peacock 1989, in the Arctic. In the Svalbard area especially the distribution Salvigsen et al. 1992, Hjort et al. 1995). It has been shown pattern of Mytilus edulis should be sensitive for an increased that Mytilus edulis lived in Svalbard continuously from c. in ux of warmer Atlantic water. 9600 yr BP to 3500 yr BP, and probably also for a short Mytilus edulis is mainly an intertidal to shallow-water period c. 1000 yr BP (Salvigsen et al. 1992). The northern bivalve which spreads with pelagic larvae and byssus coast of Spitsbergen was strongly in uenced by the in ow of drifting, and should therefore be subject to a rapid immigra- warm Atlantic water during the Ž rst half of the Holocene, tion of new areas when drifting conditions are favourable. A with a summer sea surface temperature probably 1–3° C Swedish expedition in 1861 found small Mytilus edulis on warmer than present conditions (Peacock 1989, Salvigsen et seaweed in Spitsbergen (Heer 1870) and considered it to al. 1992). On Nordaustlandet, only the western part seems to belong to the recent fauna of Spitsbergen. None of the later have been in uenced by the Atlantic water, which probably investigations of the recent mollusc fauna of Svalbard have penetrated southwards through Hinlopenstretet at least as far recorded Mytilus edulis living there (see Feyling-Hanssen as to Wahlenbergfjorden. The timing of the maximum of the 1955, Weslawski et al. 1993). Possible living animals Holocene climatic optimum could be indicated by the attached to seaweed or other  otsam, e.g. plastic Ž sh crates, occurrence of Modiolus modiolus at the head of Woodfjorden have been reported from several localities in Spitsbergen (see c. 8300 yr BP and by the occurrence of Mytilus edulis at the references in Hjort et al. 1995). During the summer of 1995, head of Wahlenbergfjorden about 1000 years later. The fresh-looking species of Mytilus edulis occurred fairly marine climate in Woodfjorden seems to have been much frequently on the modern shores of Liefdefjorden (western 60 O. Salvigsen NORSK GEOGRAFISK TIDSSKRIFT 56 (2002) branch of Woodfjorden), often attached to different kinds of References  otsam and jetsam but also with many in a free position in the beach zone. Dispersal may also be achieved by juveniles AÊ dlandsvik, B. & Loeng, H. 1991. A study of the climatic system in the and small adults drifting in the water column (Martel & Chia Barents Sea. Polar Research 10, 45–49. 1991). Bayne (1964) demonstrated how juvenile Mytilus Bayne, B. L. 1964. Primary and secondary settlement in Mytilus edulis L. (Mollusca). Journal of Animal Ecology 33, 513–523. edulis can disperse and colonize habitats where larval Blake, W. Jr. 1961. Radiocarbon dating of raised beaches in Nordaustlandet, settlement had been low or absent. Thus, the spreading Spitsbergen. Raasch, G. O. (ed.) Geology of the Arctic, 133–145. conditions for Mytilus edulis from more southern areas, as University of Toronto Press, Toronto. the mainland of Norway, seem excellent with the main  ow Blomstrand, C. V. 1864. Geognostiska iakttagelser under en resa till of Atlantic water of the West Spitsbergen Current (Loeng Spetsbergen aÊ r 1861. Kungliga Svenska Vetenskapsakademien . Handlin- gar. N.F. 4. Stockholm. 1991). Bondevik, S., Mangerud, J., Ronnert., L. & Salvigsen, O. 1995. Postglacial The southernmost island of the Svalbard archipelago, sea-level history of Edgeøya and Barentsøya, eastern Svalbard. Polar Bjørnøya, is located about midway between Spitsbergen and Research 14(2), 153–180. the mainland of Norway. It should therefore be regarded as a Bru¨ckner, H. 1996. Studies of beach deposits in northern Spitsbergen. possible ‘stopover’ locality for a potential immigration of Ma¨usbacher R. & Schulte, A. (eds.) Beitrage zur Physiogeographie . Festschrift fu¨r Dietrich Barsch. Heidelberger Geographische Arbeiten Mytilus edulis to the Spitsbergen coasts. It is remarkable that 104, 375–388. this mollusc has recently been reported to live on Bjørnøya at Christensen, B. O. 1965. Notes of the littoral Fauna of Bear Island. Astarte Ri eodden, which is the easternmost point of the island 26, 1–15. (Weslawski et al. 1995, 1997). Previous investigations Chydenius, K. 1865. Svenska expeditione n till Spetsbergen aÊ r 1861 under (Christensen 1965, Gulliksen 1979) have not reported any ledning av Otto Torell. Norstedt & So¨ner, Stockholm. Duplessy, J.-C., Ivanova, E., Murdmaa, I., Paterne, M. & Labeyrie, L. 2001. presence of Mytilus around the shores of Bjørnøya. Thus, the Holocene paleoceanograph y of the northern Barents Sea and variations of Mytilus edulis on Bjørnøya is most probably a new, recent the northward heat transport by the Atlantic Ocean. Boreas 30, 2–16. immigration, which should indicate a more favourable Feyling-Hanssen, R. W. 1955. Stratigraphy of the marine Late-Pleistocene of marine climate with increasing sea surface temperatures Billefjorden, Vestspitsbergen. Norsk Polarinstitut t Skrifter 107. 186 pp. around Bjørnøya. The Holocene history of Mytilus edulis Forman, S. 1990. Post-glacial relative sea-level history of northwestern around Bjørnøya is unknown, but the question about an Spitsbergen, Svalbard. Geological Society of America Bulletin 102, 1580– 1590. ongoing warming of the Barents Sea waters should be Funder, S. 1993. Pre-Holocene faunas from Spitsbergen – Ž eld observations. tentatively discussed based on the new Ž nding. Landvik, J. (ed.) PONAM. Fieldwork in Svalbard 1993, 31–39. UNIS, Mytilus edulis is considered to be a reliable indicator of the Longyearbyen. marine climate, and a near future immigration to Spitsbergen Funder, S. & Weidick, A. 1991. Holocene boreal molluscs in Greenland – is not unlikely. An eventual discovery of living communities paleoceanographi c implications. Paleogeography , Paleoclimatology , Pa- leoecology 85, 123–135. of Mytilus edulis there would indicate a climate much like Gulliksen, B. 1979. Shallow water bentic fauna from Bear Island. Astarte 12, that c. 1000 yr BP during the so-called Medieval warm 5–12. period (Salvigsen et al. 1992). Hald, M., Dokken, T. & Hagen, S. 1996. Paleoceanography on the European arctic margin during the last deglaciation . Andrews J. T., Austen W. E., Bergsten, H. & Jennings, A. E. (eds.) Late Quaternary Paleoceanograph y of the North Atlantic Margins. Geological Society Special Publication Conclusions 111, 275–287. Hald, M., Kolstad, V., Polyak, L., Forman, S. L., Herlihy, F. A., Ivanov, G. & Mytilus edulis lived on the north coast of Svalbard from c. Nescheretov, A. 1999. Late-glacial and Holocene paleoceanograph y and 9400 yr BP until at least 5300 yr BP. Modiolus modiolus sedimentary environments in the St. Anna Trough, Eurasian Arctic Ocean lived at the head of Woodfjorden c. 8300 yr BP. margin. Paleogeography , Paleoclimatology , Paleoecolog y 146, 229–249. The climatic optimum in this northern area is indicated by Heer, O. 1870. Die miocene Flora and Fauna Spitsbergens. Mit einen Anhang the latter age and the age of Mytilus from Wahlenbergfjor- u¨ber diluvialen Ablagrungen Spitsbergens. Kungliga Svenska Vetenskaps- akademien. Handlingar . N.F. 8. Stockholm. den, Nordaustlandet c. 7250 yr BP. Hjort, C. & Funder, S. 1974. The subfossil occurrence of Mytilus edulis L. in Surface temperatures around Svalbard are mainly deter- central East Greenland. Boreas 3, 23–33. mined by varying in ow of polar and Atlantic water, and the Hjort, C., Mangerud, J., Adrielsson, L., Bondevik, S., Landvik, J. Y. & occurrence of thermophilous molluscs indicates less in ow Salvigsen, O. 1995. Radiocarbon dated common mussels Mytilus edulis of polar water than at present and/or increased in ow of from eastern Svalbard and the Holocene marine climatic optimum. Polar warm Atlantic water. Research 14(2), 239–243. Loeng, H. 1991. Features of the physical oceanographi c conditions of the Bjørnøya presently constitutes the northern limit of living Barents Sea. Polar Research 10, 5–18. Mytilus edulis communities. Mangerud, J. & Gulliksen, S. 1975. Apparent radiocarbon age of recent marine shells from Norway, Spitsbergen and Arctic Canada. Quaternary Research 5, 263–273. Mangerud, J. & Svendsen, J. I. 1992. The last interglacial-glacia l period on Acknowledgements.–Shells from BraÊ nevatnet, Wahlenbergfjorden , were Spitsbergen, Svalbard. Quaternary Science Reviews 11, 633–664. provided by Aleksandr Tebenkov and shells from Skamdalen, Wijdefjorden, Martel, A. & Chia, F.-S. 1991. Drifting and dispersal of small bivalves and were provided by Vladimir Sharin. Kirsti Høgvard, Weston Blake, Jr. and gastropods with direct development . Journal of Experimental Marine Stian Soltvedt assisted during Ž eldwork. Jon Tolgensbakk made the Biology and Ecology 150, 131–147. illustrations. Atle Nesje and Rolf Sørensen provided valuable referee Navarro, J. M. & Thompson, R. J. 1996. Physiological energetics of the horse comments. mussel Modiolus modiolus in a cold ocean environment. Marine Ecology Progress Series 138, 135–148. Manuscript submitted 19 January 2002; accepted 14 February 2002 Peacock, J. D. 1989. Marine molluscs and Late Quaternary environmental NORSK GEOGRAFISK TIDSSKRIFT 56 (2002) Radiocarbon-date d Mytilus edulis and Modiolus modiolus from northern Svalbard 61

studies with particular references to the Late-glacial period in Northwest Sandford, K. S. 1929. The glacial conditions and the Quaternary history of Europe. A review. Quaternary Science Reviews 8, 179–192. North-East Land. Geographical Journal 74, 543–552. Salvigsen, O. & O¨ sterholm, H. 1982. Radiocarbon dated raised beaches and Weslawski, J. M., Wiktor, J. Zajaczkowski, M. & Swerpel, S. 1993. Intertidal glacial history of the northern coast of Spitsbergen, Svalbard. Polar zone of Svalbard. 1. Macroorganisms density and biomass. Polar Biology Research 1, 97–115. 13, 73–79. Weslawski, J. M., Wiktor, J., Zajaczkowski, M. & Duris, Z. 1995. Tidal Zone Salvigsen, O., Forman, S. L. & Miller, G. H. 1992. Thermophilous molluscs Project 1994. Bjørnøya. Report from study supported by Norsk Polar- on Svalbard during the Holocene and their paleoclimatic implications. institutt, Oslo, and Institute of Oceanology PAS, Sopot, Gdansk. Polar Research 11(1), 1–10. Weslawski, J. M., Zajaczkowski, M., Wiktor, J. & Szymelfenig, M. 1997. Salvigsen, O. & Slettemark, Ø. 1995. Past glaciations and sea levels on Intertidal zone of Svalbard. 3 Littoral of a subarctic, oceanic Island. Bjørnøya, Svalbard. Polar Research 14, 245–251. Bjørnøya. Polar Biology 15, 45–52.