Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe

by Guthrie, R. Dale

  Several fragments of information point to this interpretation. One is the presence of saiga antelope (Saiga tatarica) on the north slope of Alaska and in the Yukon Territory around 37,000 years ago, about the same time that Blue Babe lived in the Fairbanks region (Harington 1981). Today, these areas have over twice as much snow as saigas can tolerate. And in summer our ubiquitous tussock meadows would be totally unnegotiable. Nelson (1982) identified an assortment of insect fossils from interstadial deposits (Boutellier Interval) on the north slope. He found a number of species that do not occur there today and are found only in very dry conditions elsewhere. Yet these insects from arid habitats occur in deposits with other species still found in the north slope that are now limited to moist habitats. Pollen from this same Boutellier Interval found in Siberia and parts of eastern Beringia suggests that although spruce, larch, birch, cottonwood, and aspen were present, they grew much more sparsely than today (Hopkins et al. 1982). Unfortunately we have no pollen cores from interior Alaskan lakes which reach so far back into the past.

  As in Alaska and Siberia, Europe was without trees during the full glacial episodes. However, the European interstadial (isotope stage 3, in the deep-sea cores) was not characterized by a reinvasion of oak hardwood forests as occurred in the early Holocene. European interstadial pollen continues a pattern of herbaceous dominance with some pine (Bowen 1981). In that regard Europe was probably more like interior Alaska during the interstadial, as indeed it was similar to Alaska during the peak glacial.

  How did these glacial and interstadial changes affect steppe bison? Does Blue Babe, our interstadial (Boutellier) bison, differ from his glacial counterparts? We know that the Holocene profoundly affected bison all across the Holarctic, markedly reducing body size and changing horn size and shape (Sher 1971; Wilson 1975, 1978; Guthrie 1970, 1980). Did a similar thing occur during the last interstadial? Unfortunately our data are not good enough to make a detailed comparison, but we can gather some ideas from Blue Babe.

  Natural Trap Cave in northern Wyoming contains many bones that provide a long record of large mammals during the late Pleistocene. Gilbert and Martin (1984), who have worked with these fossils, found no reduction in body size during the last interglacial among sheep (Ovis), pronghorns (Antilocapra), bison (Bison), wolves (Canis), and wolverines (Gulo). They argue that this is a general pattern and that the dramatic dwarfing at the beginning of the Holocene is unique to that time. I agree with Gilbert and Martin and have argued (Guthrie 1984b) that Holocene changes indeed appear unique when viewed from biotic evidence, even though physical climatic data are not striking. From this we would expect the interstadial bison to be as large as a full glacial bison, but was it?

  Blue Babe and other interstadial bison remains collected from Pearl Creek are much larger horned than living bison; one would never mistake these large thick horns for those of a modern bison. However, the interstadial bison are somewhat smaller than most full glacial bison (see comparisons in figs. 7.1 and 7.6, and table 7.2). Blue Babe and the other interstadial bison from Pearl Creek fall within the lower range of Skinner and Kaisen’s (1947) size distributions of Alaskan Pleistocene bison (fig. 7.7).

  Fig. 7.6. Male skulls. The horn size of Alaskan Pleistocene steppe bison is quite variable. Intrapopulational differences probably account for much of this. Blue Babe’s horns are smaller than most steppe bison but larger than horns of living bison.

  Table 7.2 Size Comparisons of Blue Babe with Fossil Male Bison from Interior Alaska

  Sources: Skinner and Kaisen 1947.

  Note: Measured in mm.

  I have discussed body size among Alaskan Pleistocene large mammals in detail elsewhere (Guthrie 1984a). In that context, contrary to Gilbert and Martin’s (1984) finding, the same processes that caused Holocene dwarfing in Beringian bison were at work to a lesser extent during the interstadial. During the Holocene, the grassland environment was reduced and bison range in Alaska was limited to small islands of habitat. Usually this habitat was adjacent to mountains where windswept river outwash plains provided early succession herbs. Winter winds kept these herbs exposed. These would have been survival habitats but not dependably rich ranges with generous peaks of protein availability. Judging from the smaller size of interstadial bison, invading trees and shrubs as well as changes in the herbaceous community, reduced range quality, and poorer quality range led to smaller bison.

  We can make other comparisons between Blue Babe and extant bison. We can measure hoof phalanges to see if foot structure changed commensurate with the more moist, soft substrate we now have in the north. Today ungulates that use interior Alaskan lowlands (e.g., moose and caribou) have specially adapted feet to spread their weight, thus decreasing weight loading (kg/cm2). Sinclair (1977) observed that water buffalo (Bubalus) have broad hoofs which allow them to traverse swampy country, whereas bison have comparatively small feet for their size, being adapted to dry grasslands. Vereshchagin and Baryshnikov (1982) have measured the hooves of Beringian bison and concluded that they were much the same size as living bison. The hooves of Blue Babe were also similar to those of living bison, so the interstadial seems not to have affected hoof size significantly.

  Fig. 7.7. Alaskan Pleistocene male bison metacarpi. In this plot of metacarpal size of Alaskan bison, Blue Babe is near the lower end of the range. Any bone could be used in such a comparison, but metacarpi are particularly good because weight loading more or less determines the size of this particular bone.

  Fossil evidence indicates that the main characters of the Mammoth Steppe fauna (bison, horse, and mammoth) continued throughout interstadials. Yet at the end of the last glacial, most of the Mammoth Steppe fauna became extinct or severely reduced in their distribution and were supplanted by the boreal forest fauna of today—black bear, moose, and so forth. By this fact alone one would say that the interstadial large mammal community had a different climate than today exists in those same areas. Bison cannot live in Pearl Creek or the Fairbanks area now. In addition to forage insufficient for winter survival, forage that is there is frequently made inaccessible by deep snows a meter in depth undrifted, and occasionally well over a meter, far too deep for bison to survive the winter.

  The paleoenvironment of the last interstadial in interior Alaska is ripe for more study. It is an interesting time: a mixture of the Mammoth Steppe and the boreal forest. It must have been a rather unique period, with many internal variations over the several tens of thousands of years it lasted. Matthews (1982) proposes that the interstadial might have been the critical period when eastern and western steppe faunas and floras intermingled.

  Phylogenetic Kinship

  Some studies of bison phylogeny and biogeography propose multiple coexisting species across the north (Skinner and Kaisen 1947; Geist 1971b; McDonald 1981). Others see only a central line with considerable geographic and chronological diversity (Flerov 1967, 1977; Guthrie 1970, 1980; Wilson 1975, 1978). If the former is correct, bison are unique among artiodactyls in the far north in having had more than one species per genus at any one time and place in the last half of the Pleistocene. One characteristic of far northern artiodactyls (Ovibos, Alces, Rangifer, Ovis, and Saiga), proboscidians (Mammuthus), equids (Equus, caballids, and hemionids), rhinos (Coelodonata), or even most larger carnivores—brown bears (Ursus), wolves (Canis), and lions (Panthera)—is that they lived in a community with no other closely related species. This is understood if we see that in the continuous habitat of northern environments there is (and was) little home fidelity, which encourages high vagility and social systems that foster large-scale genetic mixing. Thus, I think it unlikely that two or more species of bison coexisted for any length of time. Although placed in the same genus, hemionids and true horses (caballids) are quite distantly related and may have occupied different habitats in Beringia (hemionids the uplands and horses the lowlands).

  The ungulate community in the north is comparatively simple and species are widespread, although individuals exhibit considerab
le variation in body size and social paraphernalia throughout their geographic ranges. Most if not all northern species show dramatic evolutionary changes when lineages are traced through time (Guthrie 1984a). Social paraphernalia can change rapidly, in either direction. Although Geist (1971b) implies a unidirectional colonization process and McDonald (1981) proposes a semipermanent “stasis” picture of artiodactyl populations moving through time and space, we can document rapid evolution in these populations even since the beginning of the Holocene. Wilson (1975, 1978) showed that Great Plains bison change from a large-horned form to the modern small-horned species in 9,000 years. Contrary to colonization or stasis models, most northern species have decreased in body size and in horn and antler size during the Holocene (Guthrie 1984a). Moose, Alces alces, for example, seem to have colonized the New World late in the last glacial or early Holocene, yet southern moose populations are now morphologically different from those that colonized the far north.

  Like other artiodactyl lines in the north, bison are a rather plastic and broadly adapted species, readily responding to selection pressures of local environments, both geographically or chronologically. In fact, bison seem especially sensitive in their evolutionary refinement to new environmental conditions and attendant changes in social organization. Wilson’s (1975) documentation of body size reductions in Holocene bison on expanding grasslands in the Great Plains is a case in point.

  Although Gentry (1967), and Groves (1980) link all the Bovini (cattle, bison, and buffalo) within a narrow evolutionary radiation (they wish to submerge bison within the genus Bos), the variety of social behaviors and anatomy within the group shows that similar environments can quickly override old phylogenetic ties, strongly modifying these characters. The use of rutting wallows by African buffalo, Syncerus, and American plains bison, B. bison, extreme sexual dimorphism in body size and horn size, and the clashing form of combat used by these two species are not an indication of their phylogenetic proximity but rather a product of a parallel behavioral adaptation to more open plains situations.

  The use of evolutionarily more plastic, external characters for predicting phylogenetic kinship has strengths and weaknesses. Its strength lies in the high variability of these traits, both between and within species, which make it rather easy to delineate and specify differences. The weakness lies in the swiftness with which these characters can change geographically or chronologically—and the ease with which they can reverse themselves. For example, American bison were widely dispersed in variety of habitats: from Mexico and Florida in the south to Alaska in the north, and there were a number of local forms with distinct external appearances. Most of these bison groups were killed before they were thoroughly described, but descriptions by early explorers portray animals quite different than the bison of the shortgrass plains. Judging from dramatic geographic variations in other North American ungulates, such as white-tailed deer (Odocoileus virginianus) and mule deer (Odocoilens hemionus), we would expect the same range in appearance among American bison. Although island and coastal populations of mule deer in the northwest are easily distinguished from their counterparts in the Rocky Mountains, phylogenetically both forms are undoubtedly quite close. The point is that differences between steppe bison of western Europe and of Alaska are within the range one would expect in a species of such wide distribution and different habitats. These differences are less than those of mule deer subspecies referred to above, and for that matter are in the same range as differences between wood bison (B. b. athabascae) and plains bison (B. b. bison) which were contiguously distributed and integrated on the Continent.

  Strangely enough the confusing variety of fossil bison names is mainly an American phenomenon. Eurasians have placed most of the large, middle-to-late Pleistocene bison into a catchall species called the steppe bison (Bison priscus), although they have acknowledged that it varied chronologically and geographically. European systematists prefer a broad identity, much the same as in the brown bear, Ursus arctos, or reindeer, Rangifer tarandus, which shows analogous temporal and spatial variations. They have given the various forms of B. priscus subspecific status, for example, B. priscus dimenutus, B. priscus longicornis, and so forth.

  There are several reasons for Eurasian conservatism in naming late Pleistocene bison; one is philosophical, but another must be the comparative paucity of bison in the Old World. Although bison are not an uncommon fossil in the Old World, they occur in far greater quantities in the American Great Plains sites. It is relatively safe to say that herds of millions of bison were unique to the Holocene Great Plains.

  Unfortunately, in North America there is still no unanimous agreement as to which bison species were present when and where, nor the place of their origins. However, a general pattern is emerging of the overall features of the zoogeography of various “forms.” Their exact systematic placement is yet to be agreed on, but that is not my main concern at present. My choice of nomenclature does not necessarily imply agreement with a particular taxonomic status, but is rather a reference to a commonly recognized form.

  Bison origins remain obscure. There are a number of bison or bisonlike skulls from the late Pliocene–early Pleistocene of southern Asia—a relatively important center of bovine evolution and diversification. At first they were quite cattlelike, with none of the swollen frontals or orbital protrusion of later bison. The horns began even this early to swing laterally instead of forward as in most cattle groups. Probison dehmi and Bison sivalensis have been described from India, and B. paleosinensis from China. (See Wilson 1975 and McDonald 1981 for further discussions involving the controversy of bison origins.) Very early in the Quaternary these small, somewhat lightly built bison spread throughout Eurasia and were seen in Europe as forms variously called B. tamanensis, B. voigtstedtensis, B. langenocornis, and B. schotensacki. There may be problems with too fragmentary remains and sexual identification; however, this is currently a focus of interest to several paleontologists and a clearer picture may soon emerge.

  Whatever the exact lineage, the generalization of a more cattlelike ancestor seems to hold, as does a stem form that was small bodied and small horned in comparison to some later bison. This early diversification probably still centered around a woodland-parkland environment through temperate Eurasia. However, early in the Pleistocene another form arose that was considerably larger and had long stout horns—the steppe bison, B. priscus; it persisted throughout the rest of the Pleistocene, disappearing only at the beginning of the Holocene. B. priscus existed as far eastward as England and westward into North America. It was even found northward as far as Novaya Zemlya, 75° N latitude, and southward into Spain and the Caucasus, 40° N latitude. B. priscus was the dominant form throughout Eurasia during most of the Pleistocene, and it seems to have been virtually ubiquitous in semiopen and open country. During the glacials it can be found everywhere in the Mammoth Steppe (the Arctic grasslands spreading across northern Eurasia and Beringia) and during the interglacials along the central belt of temperate Eurasian steppes. Interestingly, B. priscus, although common, seldom dominates any Eurasian Pleistocene faunal assemblage.

  Because of the chronological persistence and ubiquity of B. priscus, an understanding of its paleobiology is perhaps a key to an understanding of bison evolution. This is particularly true as it becomes apparent that B. priscus probably gave rise to the North American lineages.

  The colonization of central North America by bison is also a controversial issue. However, debate has narrowed the time of first arrival from late Illinoian to early Sangamon. (It is central North America which is the area critical to our interest, as Alaska and the Yukon Territory are zoogeographically linked more closely to northeastern Asia.)

  Whatever the exact date of colonization, it is agreed that these already large Old World bison, B. priscus, became even larger in central North America. In fact, they became enormous, with horn cores reaching 2 m from tip to tip; this New World giant has been given a separate species category, B. lat
ifrons. Their bones are found throughout much of unglaciated North America from California to Florida, but the greatest concentrations seem to be along a line from Alberta to Texas, just east of the Rocky Mountains, and in the intermontane basins just to the west.

  As with its origins, the disappearance of B. latifrons is also controversial. It seems to be the Sangamon bison and probably decreased in size at the onset of the Wisconsinan or soon after, throughout all or most of its range.

  There is a smaller form of bison occurring after B. latifrons (the chronology of which is again disputed). The main point is that there are enough transitional forms to conclude that the gigantic B. latifrons was not a “terminal” line but graded into a comparatively smaller (but still quite large) Wisconsinan form, called B. antiquus. This line continued to decrease somewhat in body size throughout the Wisconsinan. Like B. latifrons its distributional abundance falls along the line from Alberta to Texas, although it too has a scattered occurrence from Florida to California.