One cannot dismiss these radiocarbon dates as unreliable. Virtually all are from bone collagen or mummified soft tissue from frozen sediments. Alaskan Pleistocene bones generally retain over 90% of original bone collagen, and many still have some marrow. (Bone carbon from the mineral, or apatite, fraction is much less reliable for dating.) Likewise one cannot argue that there was a taphonomic bias to the preservation of bones of glacial age. Taphonomic bias, if it exists, probably runs against preservation during full glacial because this was a time of less moisture, and hence of less slope-wash redeposition of silt to cover bones.
Proposition C
The taxa of glacial (Duvanny Yar) pollen show that species from southern steppes were not present; glacial taxa were restricted to tundra species found throughout the north today. Thus, glacial vegetation was not similar to a steppe but was barren tundra, like that which now exists in the very far north.
Ritchie and Cwynar (1982) argue that the Mammoth Steppe did not exist and that the term steppe or grassland is inappropriate because “so far no taxon has been recorded in any late Pleistocene steppe element . . . or to the tundra steppes mapped . . . in western Greenland and in the Kolyma and Indigirka River watersheds of northern Chukotka” (p. 121). My argument that there was a grassland in the north during much of the Pleistocene does not imply that the same grassland community we now see to the south was then located in the far north. There is evidence of steppe species in interior Alaska during the full glacial (Bombin 1984), for example, Stipa, needlegrass, the main indicator species of extant steppe communities in North America and Eurasia. Needlegrass has retreated southward since the Duvanny Yar and no longer grows in Alaska. It was found in a ground squirrel nest from the Fairbanks area, dated at 18,230 +/− 410 B.P. (QC-668). On that important identification alone we can reject Proposition C, but there are additional data.
Indeed, there are numerous cases of northern Pleistocene sites containing faunal and flora elements that are now found only in grasslands farther to the south. This has provoked some interesting interpretations to avoid the conclusion that such species were once native to the far north. When Tikhomirov (1958) found graminoids that now grow very far to the south associated with Siberian mammoth mummies, he could only explain this by having the mammoth migrate far into the barren north with a stomach full of southern grasses. But this is, of course, impossible as proboscidians have a very short gut transit-time, on the order of twelve hours (Laws, Parker, and Johnstone 1975).
Failing to see how the far north could sustain mammoths and resorting instead to long-range migrations sounds familiar; indeed that is what Churcher (1980) and Colinvaux and West (1984) recently have proposed. However, the several-thousand-mile winter migrations of mammoth are not very probable. Mammoths had heavy graviportal, distally muscled legs, and they required much more energy to walk than other mammals. Robert White (pers. comm.) has calculated that, at lean weight, it takes a proboscidian, such as a mammoth, twice the energy to walk as to stand. In a more gracile ungulate, such as a caribou, however, walking requires only a 20% increase in energy over standing. It would have been energetically more efficient for mammoths to avoid long travel over a short period. Thick subcutaneous fat deposits on mammoth mummies suggest that, like other Alaskan ungulates, they did not undertake thousands-of-mile migrations to warmer climates farther south for the winter. Nomadic ranging is not the same as a mass exodus in a long migration. Besides, during glacial maxima, mammoths would have to go all the way from Alaska to Montana over several thousand miles of glaciers. Churcher theorized (1980) that Great Plains mammoths migrated from Alberta and Montana south to the warm Gulf Coast.
In any case, the “migrating mammoth” theories are internally contradictory, for if the north’s summer environment were as barren as proposed, mammoths probably would not have undergone an expensive seasonal migration to get there. Again we do not have to argue solely from theory; numerous Alaskan fossil mammoth teeth (bison and horse as well) show an annulus constriction at the end of the root, indicating a winter death (Guthrie, pers. obs.).
Among modern northern ungulates the more sedentary species generally put on more fat than do the long-distance migrators. The former store fat; the latter adopt a strategy of traveling light to reach a range where winter resources are available. Northern winter ranges are well below nutrient and caloric maintenance requirements for extant ungulates (this was probably true in the Pleistocene as well), so a common strategy is to decrease activity, where possible. Most northern ungulates actually enter a dormant state in the winter. By not trying to grow and by expending as little energy as possible they can survive on below-maintenance rations for a long time. As winter proceeds they lose body fat and their condition slowly declines; most enter spring with deficits, but alive, and quickly recover on rich early summer greenery.
Even for the more mesic interstadial (Boutellier Interval) there are extralimital steppe forms in the far north. Plant macrofossils (e.g., Ukraintseva 1981) and insects (Kislev and Nazarov 1985), which now live only on dry steppes far to the south, are commonly identified at northern Siberian sites. The same is true for Alaska and the Yukon Territory (e.g., Matthews 1979; Nelson 1982). Plant species and beetles most characteristic of the Mammoth Steppe seem to have remained in the area or somewhat south in “relict” habitats, if we define that term liberally to include habitats more similar to dry Pleistocene conditions and do not necessarily imply that particular plant species have remained on a site as continuous occupants. Kislev and Nazarov (1985) point to the steppic character of many insects in deposits from the last glacial on the Kolyma Lowland in northwest Asia. The genera Conioclenonus and Stephanochelonus are very frequent in those deposits, whereas today they occur primarily in southern steppes with a few isolated populations in relict habitats in the north. One insect group more abundant in the Pleistocene than at present was the dung beetle, indicating that greater quantities of large-mammal dung were present than today (Matthews 1982).
Fig. 9.7. Two steppe mammals now extinct in Alaska. American badgers (Taxidae taxus) and black-footed ferrets (Mustela eversmanni ~ M. nigripes) were present in Alaska and the Yukon Territory during the late Pleistocene. Their presence is strong corroborating evidence of a steppe environment. The steppe ferret, M. eversmanni, is the Asiatic remnant of this Alaskan relative, as is M. nigripes on the American steppes. Their disappearance in the far north supports the theory that ecological change rather than human overkill led to the large mammal extinctions in Beringia. Humans would not likely have hunted these two small carnivores to extinction.
Large- and medium-sized mammals provide the most striking examples of species found in Alaskan late Pleistocene deposits but which now live much farther south. In addition to the bison, horses, hemionids, and saigas already mentioned are American badgers (Taxidae taxus) and Eurasian steppe ferrets (Mustela eversmanni) (fig. 9.7), whose counterparts are today restricted to more arid mid-latitude grasslands (fig. 9.8). We do not have direct dates on these two species as their bones are too small for conventional radiocarbon dating, but we do know that their main prey, ground squirrels (Spermophilus), lived in the same localities near Fairbanks only during the full glacial (marine isotope stage 2). Ground squirrel nests offer abundant and suitable material to date. Dates we have for ground squirrel nests are listed in table 9.4.
Like the fossil dates discussed earlier, the ground squirrel materials were collected in an unbiased manner from late Pleistocene sections. It is unlikely that badgers and ferrets were present without ground squirrels, which is not to say that where there are ground squirrels there are badgers and ferrets. Ground squirrels now live in alpine and tundra areas of Alaska, which do not meet other requirements of badgers and ferrets.
Colinvaux and West (1984) argue that the failure of small mammals from southern grasslands to colonize the north shows that there was no northern grassland and that large mammal remains have given us a false picture. Small mammal fossils, indeed, do not present the s
ame picture as those of large mammals, but environmental limits for small mammals are not the same as limits for large mammals.
Fig. 9.8. Regional extinction of badger and ferret in the far north. During the Holocene, badger (Taxidae taxus) and black-footed ferret (Mustela nigripes) have been confined to the midcontinent grasslands. But in late Pleistocene times both species lived as far north as Alaska and the Yukon Territory.
Table 9.4 Chronological Distribution of the Ground Squirrel
There are several reasons why small mammals are not good indicator taxa for a Mammoth Steppe. Small mammals are more limited by cold (fig. 9.9). Most large mammals have a broad thermoneutral zone and do not have to increase metabolic rate until many degrees below zero. Large mammals are more sensitive than small mammals to the availability of appropriate foods and snow depth. Thin snow cover on the Mammoth Steppe aided large mammals but had the reverse impact on mouse-sized small mammals which require considerable snow cover as insulation. Such small mammals do not now overwinter in areas blown free of snow.
During full glacials, rigorous physical barriers separated Alaska from environments farther south, barring small-mammal colonization. Glacial connections to northern Eurasia were open for colonization, and, despite difficulties with reduced snow insulation, characteristic steppe species of small mammals moved into the Eurasian portion of the Mammoth Steppe. Vangengeim (1975) pointed out that the steppe vole, Lagurus, was much farther north than it is today. Likewise, the Asian ground squirrel, Spermophilus major, steppe hamster, Cricetus cricetus, and steppe pika, Ochotona pusilla, moved north and far west of their present range, reaching England. The jerboa, Allactaga jaculus, an occupant of the Asian steppes, also moved into central Europe during the last glacial (Kahlke 1975).
Instead of southern small mammals moving into Alaska during the glacials, small mammals with northern cold and snow adaptations moved south. The brown lemming, Lemmus, and the collared lemming, Dicrostonyx, moved into the central part of North America, outcompeting local microtines (fig. 9.10). In addition to the barriers formed by continental glaciers, the combination of cold and thin snow cover probably made it difficult for southern small mammals to colonize the far north, despite the fact that some of those species might have been well adapted to northern vegetation. The same happened in the Old World; Lemmus and Dicrostonyx are common Pleistocene fossils in Europe all the way to the Mediterranean. Today these arctic lemmings dominate small-mammal communities only in the far north.
Colinvaux and West (1984) are correct in observing that the Alaskan Pleistocene small-mammal community had many of the same species as today, but the conclusion they draw from this, that the vegetation was much the same, does not necessarily follow. Alaskan small mammals (Dicrostonyx, Lemmus, and northern Microtus species) are also found in the caves of southern France among most of the same Mammoth Steppe fauna (bison, horse, mammoth, reindeer, lion, and even saiga) that occur in Beringia, and certainly no one would want to label the Perigord, the type site of the Magdalenians, as polar desert on the basis of small mammals. As in the Würm IV of the Perigord, fossil small mammals from glacial (Duvanny Yar) Alaska are not as good an indicator of climate and vegetation as are large mammals.
Fig. 9.9. The effect of snow depth on large and small mammals. Deep snow is generally detrimental to large mammals—it restricts their mobility and ability to obtain food, especially for species that feed on plants under the snow. Small mammals are usually more sensitive to winter temperatures. Deep snow insulates small mammals from extreme cold, creating a special environment called the subnivian space at ground level.
At this point I turn from testing propositions of the palynologists to my list of arguments for the existence of a Mammoth Steppe. I have been discussing the Mammoth Steppe climate at its most extreme (during full glacials), but we know aridity and grasslands could interfinger with trees as they did during the interstadial when Blue Babe was alive, because bison, horses, mammoths, and even saigas, main elements of the mammoth fauna, continued right through this period (Matthews 1982). In fact grasslands mingle with boreal forest today in both North America and northern Asia, but this mixing of grassland and taiga is more apparent along the southern border of the boreal forest than along its northern border (fig 9.11). This suggests that the aridity controlling the southern limit of the boreal forest was a more important force limiting trees during the Pleistocene than was cold, which today sets northern tree limit.
Fig. 9.10. Pleistocene range extension of collared lemmings. A number of small mammal species now found in the far north extended their range southward during the last glacial. Collared lemmings, Dicrostonyx, are divided into two species in North America (here shown in hatch marks of different directions). Both of these species are found as late Pleistocene fossils far south of their present ranges. Dicrostonyx has a number of cold adaptations which allow it to tolerate thin snow cover and severe cold. Because of low snow cover, species of small mammals adapted to southern steppes did not move northward during this cold, dry episode.
Fig. 9.11. Grassland—boreal forest ecotone. This southern margin of boreal forest seems more directly limited by aridity than temperature. Here aspen trees (Populus tremuloides) are able to grow on the more moist, north face of these grassy hills in central Alberta. Aridity may also have been a critical factor eliminating trees from interior Alaska during the late Pleistocene.
Pleistocene Paleoecology of Beringia as the Northeast Wing of the Eurasian Mammoth Steppe
I have argued (Guthrie 1980, 1982, 1984a, 1984b) that the character of Pleistocene Beringia was a special case of a larger, more inclusive biotic zone, the Mammoth Steppe. The concept of Beringia tends to confine our vision of the larger unit, as Beringia was only the eastern wing of the Mammoth Steppe (fig. 9.12).
The ubiquity of the mammoth fauna over the northern part of three continents suggests that their habitats—available grass, shallow snow cover, dry summers, and windy climate—were also wide-spread, with a diverse permutation due to local influences. Indeed, these vegetational and climatic patterns have been reconstructed elsewhere throughout the Mammoth Steppe (e.g., Kowalski 1967b; Sher 1974; Yurtsev 1974; Vereshchagin and Baryshnikov 1982). The general character of Pleistocene vegetation in Alaska and the Yukon Territory cannot be fully discussed in isolation from its Eurasian relationships.
Fig. 9.12. Beringia and the Mammoth Steppe. These terms have been used to describe the biotic integrity of the far north during the late Pleistocene. The earlier concept of Beringia is especially useful in describing Siberian-Alaskan connections. Beringia extends from the Lena River in the east to the McKenzie River in the west. However, most major plant and animal taxa in Alaska and Siberia during the late Pleistocene also had affiliations with the entirety of northern Eurasia and Alaska–Yukon Territory.
In that regard, Ritchie and Cwynar (1982) have implied that my proposed extension of the Mammoth Steppe westward from Beringia to Europe is incorrect and propose that Europe was also a polar desert. They use Woillard’s (1978) pollen evaluation of the famous Grand Pile core in northwestern France as typical, stating that, “she identifies nine stadials . . . interpreted as representing open, discontinuous tundra and polar desert” (p. 120). That is not how I read Woillard’s interpretation, nor do I find her using the terms tundra or polar desert; instead, she uses the pollen profile to interpret glacials as, “periods . . . characterized by the marked development of a Graminae-Artemisia steppe in which grow other steppe elements . . . and also by the presence of subarctic-alpine species” (p. 16). As in Beringia, the dominant pollen of the full glacial Grand Pile is grass and sedge.
A Gut-Level Argument for the Mammoth Steppe
The proposition has been put forth by many that most of the “mammoth fauna” were grazing specialists. Colinvaux and West (1984) and Ritchie (1984) have argued that bison, horses, and mammoths were not grazers and thus could get by on barren tundra or polar desert vegetation comprised of cushion plants, lichen, and miniature h
erbs and shrubs. Two lines of faunal evidence that could help resolve this issue are (a) some comparisons of glacial fauna with their nearest modern analogues and (b) information about anatomical specializations. Large-mammal groups have long evolutionary histories of partitioning dietary resources, and, given large enough samples, they can be good indicators of paleoenvironments (e.g., Vrba 1976).
Although we must be extremely cautious with arguments of analogy involving diets and relating diets to extinct species, there is least likelihood of error with bison, horse, and mammoth. These three species also comprised most of the large-mammal numbers and biomass (Guthrie 1968).
As I discussed in an earlier chapter, bison are adapted to open grasslands. Since the colonizing of central North America, bison south of Alaska have been concentrated in two areas: along the rain shadow of the Rocky Mountains where aridity created open plains; in the eastern part of the Gulf Coast in Florida, which was also apparently a more open savanna or grassland in late Pleistocene times. Bison in the east were restricted to more open parks in deciduous woodlands and were infrequent on the northwestern side of the Rocky Mountains. Remnant populations of European wisent, Bison bonasus, are found in atypical areas. Like the last feral horses and cattle confined to the Carmargue salt swamps in southern France, wild European bison now live in the last remaining fragments of wilderness—the forests of eastern Europe. Other areas have all been taken by agriculture; what once were bison steppes is now farmland. Bison lived on grasslands of those areas within historic times, including the valley parklands of Lithuania, the Carpathian uplands, the entire Caucasus, Transcaucasia, and the highlands of northern Iran (Flerov 1977). Peden et al. (1974) found that bison, in comparison to cattle, are more efficient digesters of low-quality grasses. Cattle and bison seem to have partitioned habitat: woodland and woodland parks for cattle; highland meadows, woodland parks, and open grasslands for bison (Guthrie 1980).
Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe Page 25
