Sheep, bison, caribou, and other ruminants on the Mammoth Steppe were giants (Guthrie 1984a). Ruminant species are adapted to take full advantage of high-quality and high-quantity resources in northern latitudes (fig. 9.14). Social paraphernalia such as horns and antlers of Mammoth Steppe ruminants were enormous, indicative of animals on superior summer range (Guthrie 1984b). Sheep, moose, caribou, bison, wapiti, and musk-oxen are now much smaller (fig. 9.15) than they were in the late Pleistocene, suggesting that range quality in the Holocene is not better, but worse.
Fig. 9.14. Phenology of same plant species in different habitats. Many northern plants are unusually high in protein because of their rapid growth during the long summer days. Lower sun angles result in a longer season of early growth vegetation (the most nutritious). Farther south, the supernutritious flush of early growth peaks rather abruptly. Northern animals can move about the landscape, feeding on highest quality forage for much of the summer.
Fig. 9.15. Schematic illustration of body-size changes in time and space. Northern large mammals enjoy higher quality forage for a longer growth season than their southern counterparts (fig. 9.14). This phenomenon occurs on a historical axis; peak range quality was even more sustained during the Pleistocene. I propose this explains the body-size trends along both axes.
Unlike ruminants, late Pleistocene Beringian caecalids were comparatively small. Mammoths, rhinos, and horses do not have rumens, and these caecalids (fig. 9.16) operate with different developmental and reproductive strategies as well. Caecalids—animals that compost their food in a large diverticulum in the hind gut—are geared to managing on low to modest nutrient levels spread across a long growing season (Guthrie 1984b). In contrast, ruminants have more rapid growth rates and can take advantage of nutrient peaks. Lengthening these peak periods selects for large body size in ruminants. Caecalids have more conservative growth rates and simply cannot fully utilize nutrient peaks; instead they rely on a long trajectory of moderate nutrient levels.
On the Mammoth Steppe, nutrient levels were very low during the winter, but nutrients of high quality were abundant during the critical growing season. Ruminants were able to make the most of this steep-sided arch of nutrients, and they grew to large size. Caecalids simply needed a longer growing season to reach their maximum potential size. This, combined with a different response to antiherbivory compounds (fig. 9.17), is the reason late Pleistocene mammoths, rhinos, and horses all across the north were small compared to earlier counterparts (Guthrie 1984a, 1984b). Although the issue of a species’ allocation of resources to reproduction, survival, and growth is complex, it ultimately relates to the quantity and quality of food available during the growth season. The limited northern growth season necessarily means that a ruminant, which is one of the largest representatives of its species, must be getting large quantities of high-quality forage. Growth-season pastures have to be optimal, not adequate. This logic must hold for Pleistocene Beringia as well (fig. 9.18).
Fig. 9.16. Ruminants and caecalids. There are two main digestive strategies among large mammal groups. Most cloven-hoofed, artiodactyl species are ruminants (bison, sheep, moose, etc.), while the perissodatyls (horses and rhinos) and proboscidians (mammoths) use their hind gut to ferment cellulose. Generally, monogastric hindgut fermenters (caecalids) are slower growing and reproduce more conservatively.
Our comparisons of bison size based on interstade (Boutelliere Interval) Blue Babe showed that interstadial bison are smaller than those from the peak glacial (Duvanny Yar). Partial reforestation during the Boutelliere Interval probably selected against large bison just as it did later, in the Holocene. This may be true for other species of the mammoth fauna, but as yet we lack detailed comparisons.
Fig. 9.17. Two main digestive strategies and antiherbivory defenses. The ruminants and caecalids portrayed in fig. 9.16 also differ in their response to plant defense compounds. In general, the rumen is a great detoxifier of chemical defenses, but has limitations in digesting high-fiber, poor-quality food. Caecalids can handle low-fiber food but have no rumen to detoxify compounds before they are absorbed. So caecalids and ruminants have their niches in which each excel, and though there is overlap, they are not quite the same.
Fig. 9.18. Body size and available protein. Maximum body size among ungulates depends on (1) nutritional quality (primarily protein) sufficient for growth and (2) the seasonal duration of that protein level. There is a nutritional level below which an animal cannot grow and another which exceeds the animal’s capacity. The annual nutrient curve greatly affects body size. Under a high and broad curve animals not only become developmentally large, but they are selected to increase their maximum growth potential—they evolve a larger body size.
One could argue that bison reintroduced are doing well because they are so large bodied. A number of bison from the Delta herd rank high in trophy record books. These seem to be on excellent range and at the maximum of their genetic potential, but we must remember that these bison use farmers’ barley. Large mammals can be pushed to maximum genetic size by supplemental feeding during the growth season and thus can exceed the body and horn size of wild populations from which they were taken (Guthrie 1984a).
Faunal Coherence of the Fauna and Continuity of the Mammoth Steppe
The initial impetus for bringing the vast conglomeration of arid grass-dominated communities together under the name of Mammoth Steppe was the recurrent faunal melody over this area. The more common large herbivores (Chaline 1972; Stuart 1982) in the Pleistocene of western Europe during colder and drier episodes (and to some degree during warmer and wetter periods) are horses, steppe bison, woolly mammoths, reindeer, red deer, and woolly rhinos. These are almost exactly the species we find across eastern Europe (Kowalski 1967a), across Asia (Vereshchagin and Baryshnikov 1982) and northern china (Liu and Li 1984) in the south, to the polar areas on the north (Sher 1974). For the most part, these species also occur in Alaska and the Yukon Territory (Guthrie 1968, 1982; Harington 1978). This continuity is not fortuitous. It indicates an enormous expanse of similar communities, which can legitimately and usefully be characterized as a larger unit, despite obvious local and regional variants.
The present boreal forest is analogous in scale and biotic continuity to the Mammoth Steppe. No one would be purist enough to argue against the concept of a taiga, or boreal forest, because it can be shown to have different regional communities and facies of species composition due to aspect, drainage, altitude, latitude, history, and so forth, or major physiognomic forms in its different stages of succession. Locally and regionally it has very different subcommunities, with considerable geographic variation in species composition, but wherever one goes in the boreal forests, from Siberia to the Canadian northwoods, there is a commonality of character we understand in terms of “the boreal forest.”
The Mammoth Steppe Concept
I think ecologists have difficulty imagining the Mammoth Steppe because it was a biotic zone that no longer exists. We meet the obverse difficulty in realizing that today’s biotic zones are not quite the primeval structures we once believed them to be. We have only recently come to see biotic zones as more dynamic units, some of which had a quite different character during the Pleistocene. Today, deciduous and boreal forests are rather cleanly separated; this was not true during the last glacial. Likewise, steppes and tundra are now widely separated zones, yet they were not so distinct during the last glacial. Pleistocene mixtures may look strange to us, but we must not assume too much. Holocene biotic zones may be more exceptional than assemblages we see in the fossil record.
Not long ago, when most of today’s senior scientists received their training, it was thought intuitively obvious that vegetational zones in North America and Europe were squeezed farther south during the glacials. Preliminary evidence, such as the boreal forests being pushed south, and the space between conifers and the ice sheet being occupied by tundra, seemed to corroborate that squashed sandwich image (Budel 1951). But more recen
t evidence shows, contrary to intuitive sense, that the sandwich theory does not hold.
Most glacial-aged biotic communities have no modern analogue (see Matthews 1979 and Guthrie 1984a as examples of literature reviews). In fact, most of Europe was a steppe during the last glaciation, even southern Europe (e.g., Kowalski 1967a; Beug 1968; Brunnacker 1974, 1980; Woillard 1978; Cassoli 1972). This glacial steppe (Kaltsteppe in German) was not, however, identical to the Holocene Eurasian steppe, although it shared many xeric steppe plant and animal species. Fossil data are in agreement that these northern steppes were strange mixtures of species living in the same communities—mixtures without a modern analogue. But this noanalogue steppe image is still a new way of thinking in Europe.
A parallel reconstruction of glacial steppe developed almost independently in eastern Europe and northern Asia. Beginning with an attempt to account for the widespread “mammoth fauna,” early researchers proposed a more temperate Pleistocene climate. Serebrovskij (1935) argued against this and proposed taiga and tundra vegetation similar to that of today. Later, paleobotanists and paleontologists came to yet another conclusion, with which there is now general agreement—that most of unglaciated Eurasia was more like steppe than tundra, but again, not identical to extant steppes (e.g., Giterman and Globeva 1967; Vangengeim 1967; Frenzel 1968; Sher 1968, 1974; Yurtsev 1974; Vereshchagin and Baryshnikov 1982).
I view the development of ideas about a Pleistocene steppe in Alaska as part of that same debate, even though ideas about an Alaskan steppe developed rather independently, based primarily on large-mammal evidence. It is a new and, for me, truly exciting paradigm to see these debates interconnected.
Some critics of the Mammoth Steppe concept have not argued for a barren tundra or polar desert but instead have attempted to merge the two camps into a hybrid. They contend that there was neither steppe nor tundra but a mosaic of the two (Hopkins et al. 1982; Schweger 1982; Anderson 1982, 1985; Bombin 1984). This seems to me a spurious compromise. I think most parties agree that Beringia had no exact modern analogue; it was not exactly like any extant biotope, steppe or tundra (figs. 9.19 and 9.20), but their use of the “mosaic” category seems to misunderstand the concept of a physiognomic unit. Certainly all grasslands, tundras, boreal coniferous forests, deciduous forests, and so on are mosaics of different communities and subcommunities regulated by aspect, slope, drainage, altitude, fire history, herbivory pressure, and almost an infinite number of other variables. Species combine and mix individualistically across the landscape to meet their specific needs and competitive edge; the result is a complex mix of communities and subcommunities. There is never a monoculture of sameness, nor was one implied in the concept of a Mammoth Steppe. The real issue is whether, taken as a whole, it was physiognomically more like the variety of things we lump under the heading of “tundra” or more like the variety of things we lump under the heading of “steppes.” It does not seem to have been a zone of separate card-shuffled steppe and tundra communities, as the “mosaic camp” proposes, but more likely was a very complex amalgam of what we see today in quite separate communities, both on small and large grain scales.
Fig. 9.19. Lowland tundra habitat. Northern tundra is composed primarily of hydric and mesic plants that consist of relatively slow-growing, rather toxic species well defended against herbivores.
Fig. 9.20. The boreal forest. As it spreads across Eurasia and North America, the boreal forest has a variety of local communities, but these share a similar physiognomy. Plants in this biome are well defended against large herbivores. There are two exceptions: successional growth after a fire and disturbed, nutrient-rich communities along streams.
Since the communities of the unglaciated north of Europe, Asia, and Beringia seem to have compromised an environment that had no adequate analogue, I did not want to confuse the issue by employing a label with a modern image, because it obviously was not floristically or faunistically like any existing grassland. Mosaic-mix terms such as tundra-steppe or steppe-tundra seem likewise inappropriate. The term Mammoth Steppe emphasizes that the concept includes both fauna and vegetation. What the Soviets call the “mammoth fauna” played a large part in our understanding of these northern glacial communities. The woolly mammoth prefix points to its northern character, approximating the distribution of woolly mammoths, while steppe suggests a predominance of low, thin sward. In Eurasia, and even in North America, we refer to arid grasslands of low-sward profile as steppes. Not that these must be homogenous to use that title; they are all made up of a mosaic of communities and subcommunities.
It is my conviction that the Mammoth Steppe environment was a dominant fact of life in the north in late Pleistocene times. Contrary to the views of Colinvaux, Cwynar, Ritchie, and West, the evidence supports this contention. The geographic vastness of this environmentally similar habitat commands a special appellation, and the double entendre of a Mammoth Steppe seems uniquely appropriate.
10
BISON HUNTING ON THE MAMMOTH STEPPE
Bison and Human Colonization of the Far North
Early peoples lived in Eurasia for hundreds of thousands of years (fig. 10.1), but they did not colonize the Asian far north until the waning phases of the last glacial. What kept them from utilizing the large-mammal resources on the Mammoth Steppe?
The evidence available, as I see it, argues for colonization of Beringia, north of the 60° parallel, at the close of the last glaciation (Duvany Yar Interval), and at no time prior to that. This means that human access to North America first occurred 12,000–13,000 years before the present. In fact, northward colonization of Eurasians during the late glacial seems to have coincided with a general ecological change that reduced diversity and abundance of large mammals, that is, people were extending their range northward despite a reduction in large-mammal resources.
I suspect there was one factor that kept eastern Asians from capitalizing on the herds of northern bison prior to that time, and it can be applied at two different times. In a word, it is technology—technology inadequate to cope with the severity of northern climates, cold and all its correlates.
The far north is an inhospitable place for a tropical hominid. We who live above 60° latitude do so with an advanced technology of well-insulated housing, tailored clothing, efficient heat sources, and a suite of tools that reduce danger, frustration, and expended energy. All peoples who have occupied the far north have had these in one form or another. One can survive in a lean-to shelter facing an open spruce fire, but more is required to live a productive life—to cope with newborn babies and illness, lay up secure stores, move to seasonally optimal habitats, and so on.
During most of the last glacial there were no plentiful supplies of wood (if any at all) north of 60° latitude. There were no ericad berries or mast crops from woody plants. Even with a highly sophisticated technology, life without wood is difficult, if not impossible; I would argue that such a technology (if it even exists) is hundreds of years old, not tens of thousands. No one lived in the far north in either Eurasia or North America during the height of the last glacial event (fig. 10.2).
Fig. 10.1. The cold wall to colonization. Early Paleolithic sites scatter across Eurasia in an irregular pattern, but with a definite boundary, indicated here by the thick black line. The lower Paleolithic, of course, covers a long period of time, and this northern boundary must have shifted with varying Pleistocene climate. This boundary reflected the limits of techniques developed for living in an extremely rigorous climate and landscape.
However, during Blue Babe’s time (throughout the last interstadial) there was wood in the far north. That warmer and wetter saddle within the last glacial stage lasted until around 28,000–30,000 years ago. Although there were anatomically modern people in Eurasia at that time, the sophistication of their northern-adapted technology was inadequate for the far north. For example, there is no evidence that 30,000 years ago people had sewing needles with which to make weatherproof seamed clothing. The lithic t
echnology of most of the last interstade was crude. Lean and efficient stone projectile points came late during that time, and even the thrown spear itself, judging from projectile points, was probably fresh on the scene.
Fig. 10.2. Northern shift in the cold wall due to technological and climatic change. Middle and late Pleistocene sites scattered across Eurasia also show a northern boundary (black line). This boundary was not breached until around 15,000 years ago, and sites are not common north of that line until 2,000 or 3,000 years later. Some earlier sites have been described north of this line, but they are controversial.
The presence of wood alone was not enough to permit access; interstade people 35,000 years ago still faced a limit at 60° north. Yet the warm and wet interstade probably allowed people to move farther north in Eurasia than they had ever been, an expansion that was slowly driven southward during the last glacial (Duvanny Yar Interval).
Fig. 10.3. Three potential barriers to human colonization of the New World.
After the last glacial, trees again entered the north around 12,000–13,000 years ago, and people came north with the new trees; this time they were not stopped at 60° but pushed northeastward into the then exposed Bering-Chuckchi Platform and farther, into North America. We know from archaeological sites that these people were hunting the mammoth fauna, especially bison. Colinvaux and West (1984) unnecessarily assumed that the lack of archaeological evidence in the far north during the glacials was due to the lack of resident game to hunt. According to the fossil record, the game was there; something else, not large mammal resources, excluded humans.
Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe Page 27