Ardea
Official journal of the Netherlands Ornithologists' Union

login


[close window] [previous abstract] [next abstract]

Ploeger P.L. (1968) Geographical differentiation in arctic Anatidae as a result of isolation during the last glacial. ARDEA 56 (1-2): 1-159
Of the non-marine arctic species of birds 45 percent show some form of geographical variation. The geographical forms of these species very frequently have their main breeding ranges in or in the vicinity of areas which were ice-free during the Last Glacial. Disjunctions in their present circumpolar breeding ranges often chiefly coincide with areas that were certainly covered by ice-sheets during the Last Glacial. These patterns of distribution and geographical variation suggest the influence of the Last Glacial For a detailed analysis of distribution, geographical variation, ecological preference, life habits and glacial history I selected the group of arctic swans, geese and ducks (Anatidae) with 18 species occurring in 38 geographical forms. During the Last Glacial time glacial conditions seriously affected the whole arctic habitat and so there is the problem as to where these birds had their breeding grounds during this period (refuge areas). Over 38% of the species studied show some form of geographical variation. Therefore, there arises a second problem: when and where the geographical isolation preceding present geographical variation may have taken place. The first problem is dealt with in the Physical-Geographical Part, the second in the Zoogeographical Part of this study. A. Physical Geographical Part Since at present nearly all arctic Anatidae breed in low tundra the problem may be put as follows: where did low tundra refuge areas occur during the maximum of the Last Glacial? To answer this question the Last-Glacial distribution of ice-sheets and heavy local mountain glaciations in the present arctic regions have been studied and mapped (Fig. la). The main lines of the extension of these ice sheets and glaciers are fairly well known. Opinions differ widely as to the questions whether in some places peripheral zones of the ice-sheets remained unglaciated and harboured an arctic flora and fauna. The most disputed of these peripheral zones are in Greenland, Iceland, northwest and north Norway, the northwest Canadian Arctic Archipelago and along the southern border of the Cordilleran and Laurentide ice-sheets in North America. In Chapters 2, 5, 7, 11 and 12 attempts are made to answer these questions from a physical-geographical point of view, using as evidence geological, geomorphological, palynological, phytogeographical and zoogeographical data. For coastal areas of Iceland and northwest and north Norway outlet-glaciers have been reconstructed (Chapters 5 and 7); admirality charts and present geomorphological and glaciological data have been used as a basis. Areas known to have had extensive ice-free areas with arctic tundra or subarctic woodland are briefly mentioned in Chapters 8 (North Sea area), 9 (Northern Eurasia) and 10 (Bering Sea area). Areas that probably did not offer ice-free refuge areas or about which only very fragmentary data are available are only briefly mentioned: Labrador (Chapter 3), Newfoundland Bank (Chapter 4), Spitsbergen and Spitsbergen Bank (Chapter 6). Greenland There is geological and phytogeographical evidence for ice-free areas in Peary Land. Other ice-free refuge areas probably existed, especially on the shelf along the west coast of Greenland south of 71o N lat. (Fig. 3) (Chapter 2). It is unlikely that there were ice-free refuge areas along the coasts of Labrador (Chapter 3). Newfoundland Bank probably remained ice-free (Chapter 4, Fig. 4). In Chapter 5 (Iceland) much attention has been paid to the problem of dispersal of plants and insects from W Europe to Iceland by migrating birds. The reconstruction of outlet-glaciers has shown that fairly large emerged shelf areas off west Iceland probably remained ice-free (Fig. 5). Spitsbergen Bank (Chapter 6) may have been partly ice-free (Fig. 6). Climatological circumstances were probably too bad to offer a refuge area for plants and animals, but this is not certain. In Chapter 7 (Norway) a summary is given of the discussion pro and contra ice-free areas in Norway. Field-observations made by the author and the results of laboratory research convinced the author that ice-free areas existed in And÷y (Vesterslen) and S÷r÷y (West Finnmark). Reconstruction of over 50 outlet-glaciers has shown that large emerged shelf areas may have remained ice-free off northwest and north Norway (7.2.1, Fig. 14 and 15). There is a fair agreement between the geological, phytogeographical and zoogeographical evidence for the localization of ice-free refuge areas in northwest and north Norway (Paragraph 7.4). For the North Sea area (Chapter 8) all data fairly consistently point to predominantly subarctic to arctic conditions during the cold phases of the Last Glacial. In Eurasia the total tundra area was much larger than it is at present (Fig. la). Possibly ice-free shelf areas existed in the Barents and Kara Seas (Fig. 16). As a result of the eustatic lowering of the sea-level the north coast of east Siberia shifted 1,000 km to the north. In Middle Siberia the coalescent ice caps of Taimyr Peninsula and Putorana Mountains formed an important disjunction between the western and eastern tundra flora and fauna (Chapter 9). In the Bering Sea area Last-Glacial summer temperatures may have been only a few degrees lower than they are at present. The area was an important refuge area during the Last Glacial (Chapter 10, Fig. 17). The north-western Canadian Arctic Archipelago remained unglaciated during the Last Glacial. During the maximum of the Last Glacial summer temperature were perhaps not much lower than they are at present. There is some evidence that during Late-Glacial time summer temperatures were higher than they are at present. The most probable refuge area is western Banks Islands (Chapter 11, Fig. 18). There is evidence of the presence of coniferous boreal forest or woodland along the southern border of the Laurentide ice sheet during the Last Glacial. In some areas in the NE U.S.A. a treeless landscape existed during the cold phases of the Late Glacial. Traces of permafrost have been found in many places. In extensive areas traces of periglacial frost phenomena are lacking. It is concluded that a continuous tundra zone with permafrost and an arctic flora did not exist along the southern border of the Laurentide ice sheet during any time of the Last Glacial (Chapter 12). The period with the maximum extent of inland ice did not necessarily coincide with the lowest temperatures in the northernmost refuge areas (Chapter 13). During the development of inland ice-sheets the centres of ice outflow must have gradually shifted their positions. Therefore, though all large regions with inland ice-sheets may have been covered by ice during some time of the Last Glacial, this did not necessarily take place simultaneously. Consequently, refuge areas may have shifted their positions during the course of the Last Glacial. This shifting complicated the efforts made towards the reconstruction of their localization. On the other hand, however, through this shifting plants and animals must have had a much greater chance of surviving (Chapter 14). The warm ocean currents in the North Pacific and the North Atlantic may have prevented the ocean coasts from being choked with ice-bergs (Chapter 15). It is suggested that ice-free shelf areas off W Greenland, Iceland, Newfoundland, Spitsbergen Bank and Norway offered good breeding and feeding possibilities for arctic birds (Chapter 16). B. Zoogeographical Part. In the Zoogeographical Part the Last-Glacial breeding grounds of the geographical forms and species of arctic swans, geese and ducks (Anatidae) are placed in the potential refuge breeding grounds found in the Physical Geographical Part of this study. Last-Glacial refuge breeding grounds have been searched for in the vicinity of the present breeding ranges of the species and subspecies studied. Whenever this proved to be appropriate the Last-Glacial breeding grounds were searched for along the present migration route (e.g. Anser albifrons flavirostris). In some species alternative possibilities for Last-Glacial breeding ranges are discussed. In these cases present-day subspecific differentiation within the species has been used to indicate which alternative should be taken as more likely. Last-Glacial breeding ranges have been reconstructed and they are mapped alongside the present breeding ranges (Fig. 19-37). Each of the arctic species of Anatidae is discussed separately. The chapters contain paragraphs on present ecological preference, geographical forms, supposed Last-Glacial breeding grounds, reconstruction of Last-Glacial history, and the possible influences of the postglacial hypsithermal interval and of wintering grounds on the process of geographical differentiation. The hypothetical Last-Glacial refuge areas of arctic Anatidae are summarized in Chapter 36. To gain a first check of the results of this study the predominantly arctic genus Calidris (sandpipers) is used for the sake of comparison and it is studied along the same lines as those with arctic Anatidae. The results are summarized in Paragraph 37.2 and discussed in Paragraph 37.3. It is concluded that the forms of the genus Calidris generally show the same pattern of geographical variation, distribution, and Last-Glacial distributional history as Anatidae. The glacial history of predominantly boreal birds is assumed to have been different from that of arctic species. This assumption is confirmed by means of a comparison of the predominantly boreal genus Tringa (Greenshank and other large sandpipers and waders) with the Anatidae and Calidris. The results are summarized in Paragraph 38.2 and discussed in Paragraph 38.3. The conclusion is arrived at that the present distribution and the geographical variation in the arctic species of Anatidae should largely be ascribed to the physical-geographical situation during the Last Glacial. In some cases the postglacial hypsithermal interval seems to have advanced the geographical differentiation that existed or had started during the Last Glacial.


[close window] [previous abstract] [next abstract]