Thermophilous molluscs on Svalbard during the Holocene and their paleoclimatic implications O n 0 SALVIGSEN, STEVEN L. FORMAN A N D GIFFORD H . MILLER Salvigsen, 0.. Forman, S . L. & Miller, G . H. 1992: Thennophilous molluscs on Svalbard during the Holocene and their paleoclimatic implications. Polar Research 11(1), 1-10. Five species of guide fossils from the Holocene warm period in Svalbard arc considered: Mytilus edulis, Modiolus modiolus. Arctica islandica, Littorina littorea and Zirphaea crispata. Thcsc arc now extinct in Svalbard; Zirphaea crisputa, especially. requires considerable higher water temperatures than occur there today. Known radiocarbon dates on Mytilus, Modiolus and Zirphata are givcn. Thirty-four dates on Mytilus edulis show that it lived in Svalbard from before 9500 BP to about 3500 BP, and probably again around loo0 BP. Five dates on Modiolus and Zirphaea indicate a climatic optimum in Svalbard from about 8700 BP to 7700 BP. The most favourable places then had conditions similar to the northeastern coast of Finnmark. northernmost Norway, today. Mytilus edulis is considered a good climate indicator. and a future warming of the marine climate in Svalbard could be indicated by its cventual re-immigration into thc area. Otto Saloigsen, Norsk Polurinstitutt, P . O . Box 158. N-1330 Oslo Lufihaun, Norway; Steven L. Forman. Byrd Polar Research Center, Ohio State Unioersity. I03 Mendenhull Laboratory. 125 South Oaal Mull, Columbus, Ohio 43210-1308. USA: Gifford H. Miller, Center for Geochronological Research, I N S T A A R . Unioersity of Colorado, Boulder. Colorado 80309-0450, USA. An extensive sampling of the mollusc fauna of Svalbard (Fig. 1) took place during the second half of the last century (Feyling-Hanssen 1955 for references). Molluscs (especially Mytilus edulis) which are now extinct in Svalbard were found in raised beach deposits, and the term “Mytilus time” was introduced and discussed among geol- ogists and biologists (Nathorst 1884; Hogbom 1913). Feyling-Hanssen & J ~ r s t a d (1950) and Feyling- Hanssen (1955) described the Holocene mollusc fauna of inner Isfjorden in detail and also reviewed earlier observations of mollusc species in Svalbard. Only a few additional observations have been published since then, and Feyling- Hanssen’s work still provides the best general view of the Holocene mollusc fauna in Svalbard. From the abundance of species on different levels, he divided the Holocene in four climatic periods: Late-glacial cold period, Post-glacial temperate period, Post-glacial warm period and Sub-recent period. Without radiocarbon dates it was only possible to obtain tentative ages for these periods. However, some dates from these zones appeared in a later publication. (Feyling-Hanssen & Olsson 1960), Semevskij (1965) and Semevskij & Shatov (1980) have described the mollusc fauna of raised Holocene sediments in Van Mijenfjorden. The total number of mollusc species known from Holocene deposits in Svalbard is about 100. The majority are still living along the coast and in the fjords of Spitsbergen. Nine species, however, are now extinct here (Feyling-Hanssen 1955). This points to changing sea surface tem- peratures during the Holocene. In an archipelago like Svalbard the temperature of the seawater has a decisive influence on the terrestrial climate. The former and present distribution variances of molluscs have been used in several paleoclimatic studies of Arctic areas (e.g. Andrews 1972; Blake 1973; Hjort & Funder 1974; Funder & Weidick 1991). The species most used in variance studies have been Mytilus edulis and Chlarnys islandica which in Greenland and the Canadian Arctic had a much wider distribution during the Holocene climatic optimum than at present. Peacock (1989) has made a thorough review of marine molluscs and their potential for environmental studies, especially emphasising Arctica islandica. Feyling- Hanssen (1955) and Peacock (1989) are the key references for the discussion of problems dealt with in this paper. The purpose of this paper is to present the earlier observations together with more recent ones of some indicator species in the raised beach deposits of Svalbard. The main new contribution 2 Otto Salvigsen, Steven L . Forman and Gifford H . Miller I 1 20 1 80 24’ 30 / SVALBARD \ NORWEGIAN I I \ Thermophilous molluscs on Svalbard 3 Table 1 . Radiocarbon dates on Myrilus edulis, Modiolus modiolus and Zirphaea crirpara from Svalbard. The ages are quoted as first published or as reported from the laboratory. They are corrected to a 6'IC = -25C0 PDB. either using standard values or measured values for 6I3Cc. To correct for the reservoir effect, the apparent age for seawater in Svalbard has been subtracted. The actual dating laboratories and authors have used slightly different standard values for 6I1C as well as for the apparent age of seawater. This causes. however, only small differences which are of no significance for the conclusions in this paper. U 126 and U 173 have not been reservoir corrected, but when placed i n Fig. 2, 440 years have been subtracted. (Note that in 1967 The American Journal of Science issued Radiocarbon Measuremenfs: Comprehensive Index, 1950-1965 with corrections for change of Uppsala standard). Information about corrections of the datings of Samtleben (1985) is lacking. Field no. Zo-3 NlEB 2 0 - 3 PERL ( 2 0 - 2 MYT Sa85-28 Sa78-17 Sa8 1-47 3 K A F r c Sa85-79 B-47 LAN 83-108 Sa84-36 F84-329 Sa81 -80 Sa84-3 1 Sa87-40 S 1988-323 Sa82-04 3SVATH KWI F84-410 Sa84-34 Sa82-68 Sa81-48 B2 Sa84-29 No 350 b A2 Sa81-45 Sa81-20 3DAU4 B Sa81-78 S1988-643 No 343 b SL84-33 S 1988-733 H-E 209-83 Lab no. Location (Fig. 1) M a d . "C age Species Reference Gd-1255 Gd-1257 Gd-1258 T-6590 DIC-3076 T-3098 T-5029 GX-10038 U 173 T-5662 DIC-3054 T-8385 T-4411 T-6535 T-7666 T-8333 T-4944 DIC-2908 Beta-11291 DIC-3056 T-5214 T-5367 T-6534 Lu-2137 T-6283 U-130 Gd-1898 T-6591 T-6285 T-6284 T-7015 GX-10037 T-5030 T-8334 U 126 T-4628 Gd-2394 Kaffieyra (22) Kaffieyra (22) Kaffieyra (22) Sarseyra (23) Reinsdyrllya (25) GrBhuken (26) Dicksonfjorden (15) Kaffieyra (22) Blomstrandhalveya (24) Langgrunnodden (27) Kleivdalsbekken ( 5 ) Erdmannflya (17) Pricepynten (21) Kapp Ekholm (11) Dicksonfjorden (14) De Geerdalen (6) Erdmannflya (17) Derdalsbekken ( I ) Kapp Ekholm (10) Rcindalen (3) Svartfjellstranda (19) Kapp Wijk (13) St. Jonsfjorden (20) Erdmannflya (17) Kapp Morton (4) Bromclldalen (2) Dicksonfjorden (15) Bohcmanflya (16) Erdmannflya (17) Ekholmvika (10) Ebbadalen (11) Dicksonfjorden (14) GBsodden (8) Eidembukta (18) Ebbadalen ( I I ) Dicksonfjorden (14) Kapp Murdoch (7) Mytilusbekken (9) Nordfjorden (12) 6 6 6.5 0 9 8 30 7.5 4 9 13 16 3 30 7 26 6 2.3 25 17 13 1 3.5 5 10 16 5 10 8 17 15 2 18 4 5 4 7 6 3 9630 t 130 Myrilus (aragonite) 9410 t 110 Myrilus (calsitc) 9540 f I 10 MyriIus 9480 ? 100 Myrilus 9375 t 80 Myrilus 9360 t 110 Myrilus 9220 f 120 Myrilus 9195 f 270 Myrilus 9180 f 100 Myrilus 9070 f 190 Myrilus 9030 t 100 Myrilus 8970 t 110 Myrilus 8955 f 90 Myrilus 8930 t 70 Myrilus 8820 f 120 Modiolus 8730 t 170 Myrilus 8670 f 90 Modiolus 8620 f 120 Myrilus 8610 f 120 Zirphaea 8550 ? 60 Myrilus 8525 t 75 Myrilus 8415 t 110 Myfilus' 8265 ? 80 Modiolus* 8210 ? 90 Myrilus 8060 2 110 Myrilus 7870 ? 110 Myrilus 7850 k 130 Modiolus 7690 ? 80 Myrilus 7680 f 90 Myrilus 7595 ? 110 Myrilus 7440 f 60 Myfilus 7070 t 90 Myfilus 6440 ? 80 MyfiIus 6055 t 2(K) Myrilus 4780 t 90 Myrilus 4690 f 100 Myrilus 4200 ? 40 Myrilus 3810 ? 90 Mytilus 940 k 75 Mytilus Goslar & Pazdur 1985 Goslar & Pazdur 1985 Goslar & Pazdur 1985) Unpublished Forman 1990b Salvigsen & Osterholm 1982 Lauritzen et al. 1989 Forman 1990a Unpublished Blake 1961 Landvik et al. 1987 Salvigsen et al. 1990 Forman 1990a Mangerud & Svendsen unpubl. Unpublished Samtleben 1985 Salvigsen et al. 199U Unpublished Mangerud & Svendsen unpubl. Unpublished Forman 1986 Forman 1990b Forman 1989 Salvigsen et al. 1990 Salvigsen et al. 1990 Elgenma & Hellikscn 1986 Unpublished Salvigsen et al. 1990 Salvigsen et al. 1990 Feyling-Hanssen & Olsson I960 Klysz et al. 1988 Unpublished Salvigsen 1984 Forman 1990a Klysz et al. 1988 Unpublished Mangerud & Svendsen unpubl. Feyling-Hanssen & Olsson 1960 Samtleben 1985 'Dating performed on another species in the same bed. is, however, the presentation of radiocarbon dates (Table 1) which makes it possible t o obtain an absolute age for the Holocene marine climatic optimum in Svalbard. Physical setting A simplified picture of the present surface cur- rents in the Barents Sea area is shown in Fig. 2. 4 Otto Saluigsen, Steven L . Forman and Gifford H . Miller 80' 78O 76' 74, 72' 70' 68 // -+ Atlantic water -+ Arctic water ...., Norwegian coastal current 11 I 5' 10' 15' 20' 25" 30' 35' 40' 45' 50' 55O Fig. 2. Surface currents in the Barents Sea, modified from Loeng 1985. Surface temperatures in this region are controlled mainly by the varying inflow of Arctic and Atlan- tic Water. This is illustrated by an example in Fig. 3 which shows the maximum sea surface temperatures during the summer of 1991. More representative measurements show that at the northernmost mainland of Norway the sum- mer surface temperature (SST) of Atlantic Water is about 10°C, and off western Svalbard it is 5°C or less (Gammelsr~d & Rudels 1981). The SST of the inner fjords may be higher because of the atmospheric heating in sheltered areas and the small inflow of cold water, but measurements are lacking. The limited knowledge of the sea surface temperatures during the summers makes it there- fore difficult to discuss present and former living conditions for the molluscs in detail. The overall sea ice conditions are given by Vinje (1985), but statistics for the duration of sea ice cover in the inner fjords are not available. Thermophilous molluscs on Svalbard 5 Fig. 3. Map with sea surface temperatures modified from Iskart nr. 33/91 published by the Norwegian Meteorological Institute. Ice-covered area (shaded) and temperature isolines are based on observations from ships and satellite measurements in the period August 15-19, 1991. This period showed the highest sea surface temperatures around Svalbard that year. Holocene distribution of thermophilous molluscs in Svalbard Five species of conspicuous guide fossils indi- cating deposits from the Holocene warm period in Svalbard will be considered: Mytilus edulis, Arctica islandica, Littorina littorea, Modiolus modiolus and Zirphaea crispata. These species are indicator species and can be correlated with open water summer surface temperatures (Pea- cock 1989). Apart from Mytilus they require con- siderable higher water temperatures than occur around Spitsbergen today. An indication of the size of the temperature demands may be afforded by Peacock’s (1989) comparison of modern dis- tributions along the Norwegian coasts with sea surface temperatures. From this it appears that Zirphaea crispata is distributed in areas with sum- 6 Otto Salvigsen, Steven L. Forman and Gifford H . Miller mer surface temperatures above 8”C, while Arc- tics islandica and Modiolus modiolus occur in areas with temperatures higher than 6°C in the Barents Sea (see also Funder & Weidick 1991). Chlamys islandica is usually regarded as a sub- arctic species, but it has recently been recorded alive from waters around the entire Svalbard archipelago ( B j ~ r n Gulliksen pers. comm. 1986). It occurs in raised beach deposits on SjuByane in the far north (Fig. l ) , as well as o n Kong Karls Land (Salvigsen observations 1978-1980). The wide distribution of Chlamys islandica in Svalbard demonstrates that this archipelago had and still has warmer SST than areas at the same latitude in Greenland and Arctic Canada (Funder & Weidic 1991). In contrast to what is the case in Svalbard, Chlamys islandica in East Greenland had a much wider distribution during the Holocene climatic optimum than at present (Hjort & Funder 1974). Mytilus edulis is now extinct in Svalbard waters but was widely distributed during the early Hol- ocene. Its shells have been found on the west coast from Nottinghambukta north of Hornsund (Dmoch 1977) to Blomstrandhalv~ya in Kongsfjorden. Mytilus edulis is abundant throughout the Isfjorden area, but not found in the inner parts of Van Mijenfjorden. In the north- western corner of the archipelago there has been no positive shoreline displacement after 9500 BP and Mytilus can thus only be expected in sub- marine deposits along the shore. On the northern coast it has been recorded from Raudfjorden in the west to Langgrunnodden on Nordaustlandet (Blake 1962). It has also lived in the Hin- lopenstretet area. O n the eastern coast it has been reported from southwestern E d g e ~ y a and from Svensk~ya, Kong Karls Land. On E d g e ~ y a it was only found at low levels in the southwestern part of the island (Knipowitsch 1903), and not in situ, hence age estimates are not possible. During the 1991 fieldwork of the European Science Foun- dation project, “Polar North Atlantic Margins, Late Ceneozoic Evolution” (PONAM) in eastern Svalbard, several observations of Mytilus were recorded from E d g e ~ y a , but it was not observed on Barentscbya (different reports in Landvik 1991). Two small fragments from S v e n s k ~ y a have been reported from 25 m a . s. 1. (Hagg 1950). In 1979 Salvigsen visited the same locality but was unable t o find any traces of Mytilus, nor was it found in other raised beach deposits in Kong Karls Land. Its occurrence in these islands seems therefore to have been very limited. Mytilusedulis has been found in raised sediments as far east and north as in Frans Josef Land (Nansen 1902) and should therefore be expected to have occurred throughout Svalbard. However, in Nordaustlan- det Mytilus edulis has only been reported from the western area, but it should be kept in mind that the northern coast of Nordaustlandet has very few reports on shell faunas. Elsewhere, the summer surface temperature of the water has probably been too low for Mytilus during the Holocene. Salvigsen looked for Mytilus shells without success on SjuByane in 1978 and 1980. Arctica islandica, Littorina littorea, Modiolus modiolus and Zirphaea crispata require higher temperatures than Mytilus edulis, especially Zir- phaea crispata (Peacock 1989). However, all three species, are well known from Holocene sediments in the inner Isfjorden area. Arctica islandica has also been reported from Prim Karls Forland and Lomfjorden. Its occurrence in Lomfjorden, by Hinlopenstretet. is most remarkable and may indicate a former wider distribution than indi- cated by the reports. Littorina littorea seems to have had a similar distribution as Arctica islan- dica. It is found at many places in Isfjorden, and in the Woodfjorden area on the north coast. In addition t o finds referred to by Feyling-Hanssen (1955), a single find from southern Spitsbergen, north of Hornsund, has been made by Dmoch (1977). It occurs in Mytilus beds on Erdmannflya in Isfjorden (Salvigsen et al. 1990). Littorina lit- torea probably had a wider distribution than hith- erto reported, because it is not so conspicuous and easy to recognise in the field as the larger valves, e.g. those of Arctica islandica and Mytilus edrilis. Modiolus modiolus had until recently only been reported from Holocene sediments in inner Isfjorden, but in 1984 it was found on Erd- mannflya in outer Isfjorden, where a few frag- ments occurred among Mytilus edulis shells (Salvigsen et al. 1990). It has also been reported from sublittortal sediments at the head of St. Jonsfjorden (Forman 1986). Zirphaea crispata has only been reported from inner Isfjorden and seems to have had the most limited distribution in Svalbard of the here mentioned indicator species. Dating of the thermophilous molluscs Known dates on thermophilous molluscs from Svalbard are listed in Table 1: thirty-four dates Thermophilous molluscs on Svalbard 7 is lacking in the terraces of the “Late-glacial cold period” and the “Sub-recent period”. The most likely conclusion seems therefore to be that the climate was tolerable for Mytilus edulis during a continuous period of about 6000 years. Dates from the “empty” period could probably be obtained by a systematic dating of Myrilus shells from inner Isfjorden. The Mytilus terrace of Feyl- ing-Hanssen (1955) seems also t o be 5000 years or younger, and optimal conditions for Mytilus edulis seem to have occurred between 5000 and 4000 BP in inner Isfjorden. A surprising dating of Mytilus from Spitsbergen was presented by C. Samtleben (1985 and pers. com.) who studied microstructures of Mytilus shells from Svalbard and other areas. Shells on the present beach of Nordfjorden in inner Isfjorden were radiocarbon dated t o 940 f 75 years. This is the first and only Mytilus date from Svalbard younger than 3500 BP, and it indicates that Mytilus lived in Svalbard about 1000 years ago during what is called the Little Climatic Optimum. A confirmation of this single date would be desir- able, but probably difficult to obtain. The isostatic uplift during the last 1000 years has been insig- nificant, and many shells on the present shore have been washed out from older sediments, e.g. on Mytilus, four dates on Modiolus and one on Zirphaea. The original purpose for most of the dating has been either dating of old, mostly iso- statically elevated shorelines, or the dating of sediments. The highest lying and oldest shells have therefore often been chosen for dating. Along much of the western coast of Spitsbergen it has been possible t o date only the oldest Mytilus shells because younger shorelines are below present sea level (Forman 1990a). But never- theless, when seen together these dates give a great deal of information about the former occur- rence of thermophilous molluscs in Svalbard, even if they are not statistically representative for the climatic optimum. The dating results (Table 1) indicate that Mytilus edulis was well established in Svalbard waters before 9500 BP. The distribution of ages (Fig. 4) indicates its presence in Svalbard until at least about 3500 BP, even though no datings from between 6500 BP (T-4628) and 4700 BP (T-5030) have been obtained. Observations in the field, however, indicate that Mytilus edulis lived in Sval- bard also during this period. Feyling-Hanssen (1955) shows that Mytilus occurs in all terraces in the “Post-glacial temperate Period” and the “Post-glacial warm period” in Billefjorden, but it Zirphaea crispata Modiolus modiolus y Mytilus edulis 10 9 8 7 6 5 4 3 2 1 0 Radiocarbon Ages (ka) Fig. 4. Frequency distribution of radiocarbon dated thermophilous molluscs from Svalbard. 8 Otto Saluigsen, Steven L. Forman and Gifford H . Miller 7000 year-old shells from the surface of the mod- ern beach ridge in Dicksonfjorden, inner Isfjor- den (T-7015). Arctica islandica from Svalbard has not been radiocarbon dated, but the altitudes where it is found give some indications about minimum ages for its first occurrence. In Billefjorden it has been found at altitudes which indicate an immigration well before 8000 BP (Feyling-Hanssen 1955; Sal- vigsen 1984). Its highest reported level in Lomfjorden, 21 m (Kulling & Ahlmann 1936), indicates that it lived there before 7500 BP. Besides Mytilus, Arctica tlandica is the most abundant of the indicator species in Dickson- fjorden. Nor has Littorina littorea from Svalbard been radiocarbon dated, but it is found on levels from 31 m to 2 m in inner Isfjorden (Feyling-Hanssen 1955). This indicates a longlasting presence start- ing before 8000 BP. Dicksonfjorden probably had the most numerous occurrences in Svalbard of Modiolus modiolus. It was found in situ in sublittoral sedi- ments in a section at Kapp Nathorst. The Modiolus bearing layer was only about 1 m thick, indicating a relatively short presence. One valve with a weight of more than 50g yielded the radiocarbon age 8820 2 120 years BP (T-4411). A large fragment from another deposit in the inner part of the fjord, at Raudkollen, had the age 7850 k 90 years BP (T-6534). It was found in a stream cut through a marine terrace, and no in situ shells were observed there. A radiocarbon date on one fragment of Modiolus modiolus from Erdmannflya in outer Isfjorden gave the age 8670 * 130 years BP (T-6535). These three Modiolus dates, combined with field observa- tions, indicate that it lived in Isfjorden for more than 1000 years. Forman (1986, 1989) dated paired Mya truncata shells from sublittoral sands in St. Jonsfjorden which among other species included Modiolus modiolw, Mytilus edulis and Littorina littorea. The age 8265 5 80 (DIC-3056) indicates that the SST in the inner part of that fjord was considerably warmer before 8OOO BP than today. The only known date of the Low Arctic species Zirphaea crispata from Svalbard is on shells from Kapp Ekholm in Billefjorden. The age was 8610 k 120 years BP (T-8333) (Mangerud & Svendsen unpubl.). The dates on Modiolus and Zirphaea lead t o the conclusion that Svalbard had an early Holocene marine climatic optimum with a culmination between 9000 and 7500 BP. Most of the dates are close to 8600 BP, the most probable time for the maximum SST. Potential for the detection of climatic changes Changes of the marine climate in Svalbard can be deduced from the composition of the past marine macro fauna. Radiocarbon dating of new appear- ances and extinctions of indicator species provides a chronology of climatic changes during the Hol- ocene. The molluscs dealt with here are expected to respond quickly to SST changes. They can spread to new areas as their larvae are carried by currents. Small, fresh species of Mytilus edulis have also been found on seaweed at Spitsbergen (Heintz 1926; Feyling-Hanssen 1955) where they probably have drifted from southern areas. The spread and the settlement of the planktonic larvae could have been assisted by northward fluc- tuations of the warm Atlantic Water. Adult mol- luscs with their greater tolerance could then survive intervals with colder marine climate, and successful spawnings could take place in the most favourable summers (Peacock 1989). The climatic changes can also be seen in Hol- ocene deposits with micro- and macrofossils of plants (Hyvarinen 1970; Gottlich & Hornburg 1982; Serebarannyy et al. 1984; Birks 1991), but for the development of the flora there will always be a time lag compared with the actual climatic changes. The exact limiting temperatures for the indi- cator species are not known, and it is therefore difficult to quantify temperatures for Svalbard during the Holocene. Feyling-Hanssen (1955, p. 33) discussed paleoclimatic evidence of the fossil fauna. He concluded that during the “Post-glacial warm period” the climate was “no worse than that of the low-arctic subregion of to-day, any way, no better than that of the high-boreal”. H e also suggested that the climatic conditions were more favourable in the beginning of this period than later in the same period. The abundance of Mytilus edulis on low levels in Isfjorden indicates the latest slight improvement of the climate before the final decline of the warm period. Samtleben (1985) observed microstructures in Mytilus shells from inner Isfjorden resembling those in living Thermophilous molluscs on Svalbard 9 mussels from northeastern Norway, indicating similar climatic conditions. The microstructure of the shells shows strong seasonality with appar- ently short but favourable summer periods and relatively long winter periods. Samtleben also constructed a temperature curve from values of Mytilus shells from Isfjorden (suggesting the same salinity as that of present time summer waters in the fjord); the curve indicated summer surface temperatures up to 10°C. The mollusc faunas show that the marine cli- mate in Svalbard during most of the Holocene was warmer than that of today. During the climatic optimum, in the most favourable places in Spits- bergen, it was probably similar t o the present climate on the eastern coast of Finnmark, north- ernmost Norway. Our conclusions regarding the rise in SST during the Holocene are that Mytilus edulis suggests a rise of about 1°C in SST and Zirphaea crispata at least a 3°C rise in SST. This is in agreement with the conclusions of Funder & Weidick (1991) for Greenland during the Holo- cene. Funder & Weidick also emphasise that the immigration of boreal molluscs to Greenland was determined by oceanic rather than climatic change. Birks (1991), in her studies of the veg- etational history of west Spitsbergen, indicates a mean July temperature for the early Holocene about 2°C higher than today. It is obvious that the molluscs will play an important role i n further studies of the mid Hol- ocene climatic optimum in Svalbard and other Arctic areas. More remarkable, however, is the fact that the Little Climatic Optimum about 1000 years ago also seems to have had favourable con- ditions for Mytilus edulis in inner Isfjorden. Molluscs in Svalbard reflect warmer SST, and the mollusc fauna of western Svalbard may be instrumental in detecting a possible larger supply of warm Atlantic Water to Svalbard. Molluscs respond mostly to increasing summer tempera- tures. They therefore can only supply limited information about the rest of the year and the mean annual temperature. This leads to the conclusion that if a future climatic optimum develops in the Arctic, it may be detected in the mollusc fauna of western Sval- bard if its character is analogous to the Holocene climatic optimum. Most greenhouse climate models show an asymmetric warming in the winter with modest or no changes during the summer (Mitchell et al. 1990). The Holocene climatic optimum thus seems at our present state of knowl- edge to be a bad analog for greenhouse warming in the Arctic. Conclusions During the Holocene, Isfjorden had the largest occurrence in Svalbard of molluscs which can be correlated with higher open water summer temperatures, indicating that this area had and still has the warmest marine climate in the archi- pelago. Mytilus edulis lived in Svalbard from at least 9500 to 3500 BP and probably again around 1000 BP. The species which require the highest tem- peratures, Zirphaea crispata and Modiolus modiolus, indicate optimal conditions around 8700 to 7800 BP. The distribution of thermophilous molluscs during the Holocene climatic optimum is related to the present pattern of summer surface tem- peratures, although temperatures then were 1- 3°C higher. Molluscs may immigrate t o Svalbard via the warm West Spitsbergen Current and are therefore able to respond quickly t o climatic changes. If a global warming results in higher summer surface temperatures, Mytilus edulis may soon live again along the shores of Svalbard. It is considered as a reliable climate indicator and is thus useful for detection of a possible warming of the marine climate in the Arctic. Acknowledgements. - We would like to thank Jan Mangerud and John Inge Svendsen for providing unpublished radiocarbon dates. References Anderson, G . 1910: Die jetzige und fossile Quartarflora Spits- bergens als Zeugnis von Klimaanderungen. Posfglariale Kli- maueranderungen, 409-417. Stockholm. Andrews, J. T. 1972: Recent and fossil growth rates of marine bivalves, Canadian Arctic, and Late-Quaternary Arctic mar- ine environments. Paleogeogr. Paleoclimatol., Paleoecol., Birks, H . H. 1991: Holocene vegetational history and climatic change in west Spitsbergen - plant macrofossils from Skardtj~ma, an Arctic lake. The Holocene 1 , 209-218. Blake, W . , Jr. 1%1: Radiocarbon dating of raised beaches in Nordaustlandet, Spitsbergen. Pp. 133-145 in Raasch, G . 0. (ed.): Geology o f f h e Arcfic. Univ. of Toronto Press, Toronto. Blake, W . , Jr. 1%2: Geomorphology and glacial geology in Nordaustlandef, Spitsbergen. Ph. D . Thesis Ohio State Univ. 470 pp. 157-176. 10 Otto Salvigsen, Steven L . Forman and Gifford H . Miller Blake, W., Jr. 1973: Former occurrence of Mytilus edulis on Coburg island, Arctic archipelago. Naturalkte Canadian 100, Dmoch, 1. 1977: Characteristic of a fauna from a 4 m terrace at Nottinghambukta (West Spitsbergen). Acta Universitatis Nikolai Copernici (AUNC). Geografia 13, Torun, 181-197. Elgersma, A. & Helliksen, D. 1986: Kvartargeologiske under- s@kelser i Van Mijenfjordomrddet, Spitsbergen. Cand. scient. thesis, Univ. of Bergen. 219 pp. + figs. Feyling-Hanssen. R. W. 1955: Stratigraphy of the marine Late- Pleistocene of Billefjorden, Vestspitsbergen. Norsk Polar- inst. Skrifer 107. 186 pp. Feyling-Hanssen, R. W. & Jarstad, F. A. 1950: Quaternary Fossils from the Sassen-area in Isfjorden, West-Spitsbergen (The Marine Mollusc Fauna). Norsk Polarinst. Skrifier 94. Feyling-Hanssen, R. W. & Olsson, I. 1960: Five radiocarbon datings of post glacial shorelines in central Spitsbergen. Norsk Geogr. Tidrskr. 17, 122-131. Forman, S . L. 1986: Quaternary glacial, marine, and soil devel- opmental history of the Forlandsund area, Western Spits- bergen, Svalbard. Ph.D. thesis, Univ. of Colorado, 330 pp. Forman, S. L. 1989: Late Weichselian glaciation and deg- laciation of the Forlandsundet area, western Spitsbergen. Boreas 18, 51-61. Forman, S . L. 1990a: Post-glacial relative sea-level history of northwestern Spitsbergen, Svalbard. Geol. Soc. of Am. Bull. 102, 1580-1590. Forman. S. L. (comp.) 1990b: Svalbard Radiocarbon Date List 1. Occasional Paper No. 47 INSTAAR, Boulder. 48 pp. Funder, S. & Weidick, A. 1991: Holocene boreal molluscs in Greenland - paleooceanographic implications. Paleogr. Paleoclimatol. Paleoecol85, 123-135. Gammelsrad, T. & Rudels, B. 1981: Hydrographic and currrent measurements in the Fram Strait, August 1981. Polar Research I n . s . , 115-126. Goslar, T. & Pazdur, M. F. 1985: Datowanie muszli mieczakow metoda I4C. Kwartalnik Geologiczny 29, 459-412. Gottlich, K. & Hornburg, P. 1982: Ein Zeuge warmzeitlicher Moore im Adventdalen auf Spitsbergen. Telma 12, 253- 260. Heintz, A. 1926: Bliskjell PA Spitsbergen. Norsk Geogr. Tidrskr. 9,14-16. Hjort, C. & Funder, S. 1974: The subfossil Occurrence of Mytilus edulis L. i n central East Greenland. Boreas 3,23-33. HyvBrinen, H. 1970: Flandrian pollen diagrams from Svalbard. Geogr. Ann. 52A, 213-222. Hi@, R. 1950: Kvartara marina fossil frAn Spetsbergen insam- lada av svenska expeditioner. Geol. Foren. Forhandl. 72, Hdgbom, B. 1913: Om Spetsbergens Mytilustid. Geol. Fiiren. Stockh. Forh. 35, 151-156. Klysz, P., Lindner, L., Makowska. A., Marks, L. & Wysok- inski, L. 1988: Late Quaternary glacial episodes and sea level changes in the northeastern Billefjord region, Central Spitsbergen. Acta Geologica Polonica 38, 105-123. Knipowitsch, N. 1903: 1V. Mollusca und Brachiopoda. (Zool- ogische Ergebnisse der russischen Expeditionen nach Spits- bergen. Nachtrag.) Extrait de I‘Annuaire du Musee Zoologique de I‘Academie des Sciences de St. Petersbourg. 8. Kulling, 0. & Ahlmann, H. W. 1936: Observations on raised 51-58. 85 PP. 331-347. 11 PP. beaches and their faunas. Surface markings. Geogr. Ann. 18, Landvik, J. Y., Mangerud, J. & Salvigsen, 0. 1987: The Late Weichselian and Holocene shoreline displacement on the west-central coast of Svalbard. Polar Research 5 n. s., 29-44. Landvik, J . Y. (Ed.) 1991: Fieldwork in Svalbard 1991. PONAM. Preliminary report. 59 pp. Lauritzen, O., Andresen, A., Salvigsen, 0. & Winsnes T. S. 1989: Geological map of Svalbard l:lW,ooO. Sheet C8G Eillefjorden. Map + 32 pp. text. Loeng, H. 1985: Ecological Features of the Barents Sea. The 6th International Conference of Comitk Arctique. Marine Living System of the Far North. Fairbanks. Alaska. 54 pp. Mitchell, J. F. B., Manabe, S., Meleshko, V. & Tokioka, T. 1990: Equilibrium Climate Change - and its implications for the Future pp. 131-172 in Houghton, J. T., Jenkins, G. J. & Ephraums, J. J. (eds.): Climate Change. The lPCC Scientific Assessment. Cambridge. Nansen, F. 1902: The oceanography of the North Polar Basin. The Norwegian North Polar Expedition, 1893- 1896. Scient. res. 3, Christiania. Nathorst, A. G. 1884: Redog6relse for den tiUsammans med G. de Geer i r 1882 foretagna geologiska expeditionen till Spetsbergen. Bihang ti1 K. Svenska Vetensk. Akad. Handl. 9. Stockholm. 79 pp. Peacock, J. D. 1989: Marine molluscs and Late Quaternary environmental studies with particular reference to the Late- glacial period in Northwestern Europe: A review. Quaternary Science Reviews 8, 119-192. Salvigsen, 0. 1984: Occurrence of pumice on raised beaches and Holocene shoreline displacement in the inner Isfjorden area, Svalbard. Polar Research 2 a s . , 107-113. Salvigsen, 0. & Osterholm, H. 1982: Radiocarbon dated raised beaches and glacial history of the northern coast of Spits- bergen. Polar Research 1,97-115. Salvigsen, 0 . Elgersma, A., Hjort, C., Lagerlund, E., Liestl, 0. & Svensson, N.-0. 1990: Glacial history and shoreline displacement on Erdmannflya and Bohemanflya, Spitsber- gen, Svalbard. Polar Research 8, 261-213. Samtleben, C. 1985: Climatic influence on shell microstructure in Mytilus edulis from Spitsbergen. Poster presentation on the “75. Jahrestagung der Geologischen Vereinung”, 1985 Kiel, FRG. Semevskij, D. V. 1965: Marine terraces of Van Mijenfjorden and Billefjorden and their Palaeontological characterization. Materials on the geology of Spitsbergen. NIIGA, Leningrad, 230-238. (Internal translation from Russian, Norsk Polar- institutt.) Semevskij, D. V. & Skatov, E. P. 1980: Quaternary fossil fauna of invertebrates of the archipelago of Svalbard. Pp. 12%133 in: Geology of the sedimentary platform cover of the archipelago of Svalbard. Collection of scientific papers. NIIGA, Leningrad. (Internal translation from Russian, Norsk Polarinstitutt.) Serebarannyy, L. P., Tishkov, A. A., Malaysova, Ye. S., Sol- omina, O.N. & ll’ves, E. 0. 1984: Reconstruction of the development of vegetation in Arctic high latitudes. Izvestiya AN SSR, seriya geografichenskaya, 6 . (Translation from Rus- sian: Polar Geography and Geology 9, 308-320.) Vmje, T. 1985: Drift, composition, morphology and distribution of the sea ice fields in the Barents Sea. Norsk Polarinst. Skr. 179 C , 26 pp. + map. 1-19.