Part of the confusion in thinking about temperature zones is that aridity is complex; it is not a simple matter of rainfall. There are two factors in the aridity-controlled component of plant communities—moisture input and moisture loss—and a lot of subfactors go into that equation: cloud cover, wind, fire, season of moisture input, season of moisture loss, soil permeability, ground cover, and, of course, seasonal patterns of temperature. Aridity is not so neat an equation as mean annual rainfall versus vegetation type, so it is more difficult to visualize. Degree days versus vegetation type is simple and straightforward, and today in the far north it makes a lot of sense. But this sense becomes misaligned in reconstructions of Pleistocene vegetation.
Patches of grasslands on some steep south slopes in interior Alaska are sometimes cited as an example of the temperature-controlled vegetation patterns. But it is a mistake to see these small patches of grass and sage as created by temperature; they are the result of aridity. Their upturned aspect makes them warmer, but it also means that they are always well drained, and these factors combine to produce aridity. Aridity, not warmth, excludes the surrounding boreal forest species.
The second thing that is confusing, especially to North American botanists, is that today we have very few grassy steppes in the far north to serve as an analogue to the Mammoth Steppe. Small grassy patches on south slopes and the distant steppes of central Asia are incomplete and remote images. Furthermore, the relatively rich Aleutian grasslands confuse the issue. Woody plants are unable to grow on the Aleutians because of cold, wet summers (degree days are inadequate for woody growth). Aleutian herbs, on the other hand, thrive in slightly above-freezing temperatures. These unusual grasslands are hard to classify and are sometimes referred to as tundra. Such misconceptions have provided little assistance in reconstructions of the past.
Missing Pica on the Mammoth Steppe
In addition to offering more for grazers to eat, Mammoth Steppe vegetation may also have provided more minerals and other nutrients for growth than is available today. We know that the higher rate of evapotranspiration (fig. 8.8) makes plants in arid regions usually higher in water-soluble minerals than their counterparts in moist conditions (Chapin 1980). Also, the present humic acid-rich soils of Alaska restrict cation cycling. Theoretically, at least, herbivore diets on the Mammoth Steppe would have been richer in minerals. Indirect evidence supports this proposition. Hiking in the Alaskan bush, one occasionally finds bones from an old kill and, even more frequently, shed antlers from moose and caribou. These remains usually show gnaw marks made by rodents (mice, squirrels, or porcupines), but in addition, antlers are eaten by the ungulates that once grew them. Sometimes one finds an entire antler that has been reduced to its main beam, with tines and palms completely eaten. Antler consumption is so frequent today that it is rare to find old antlers without gnaw marks.
Fig. 8.8. Evapotranspiration, mineral cycling, and pica. Terrestrial plants in arid regions have high evapotranspiration rates and hence generally have high concentrations of sodium in their tissues. Herbivores in these regions have no trouble getting sufficient minerals from plants. Herbivores eating plants with less minerals must supplement their diet, turning to mineral licks or “pica,” eating shed antlers or bones to obtain critical minerals. Pleistocene bones in Alaska only rarely show pica chew marks, whereas Holocene bones are regularly chewed.
This kind of antler chewing should not be confused with the bone gnawing of carnivores, as seen on Blue Babe’s bones. Herbivore-chewed bone looks different, and it is eaten for a different purpose. Carnivores are after fat in bone cavities and protein collagen among the mineral crystals. Carnivores are seldom mineral deficient.
Herbivores chew on antler and bone because they are after the minerals. This kind of bone chewing is well known and is caused by an ungulate being mineral deficient. Nutritionists call this sort of craving for unnatural food pica (a Latin term we also use for the magpie, Pica pica). But pica (chew) marks are rare on the tens of thousands of Pleistocene bones I have examined. They do occur, but very infrequently. I propose that the higher mineral content of Mammoth Steppe grasses, and probably soils as well, meant that bison and other Pleistocene grazers were seldom lacking essential minerals (fig. 8.8).
It has always been assumed (e.g., Sutcliffe 1970) that bone-eating herbivores are deficient in phosphorus, but more recent data suggest that the critical mineral is sodium (and perhaps calcium and magnesium as well). The flush of broad-leaf growth in the herbivores’ spring diet has high ratios of potassium to other minerals and seems to cause sodium, calcium, and magnesium ions to be bumped from their systems and excreted in the urine. Plants have a much lower concentration of minerals than do animals, but potassium is an exception. Plants have over twice as much potassium as animals (Batzli and Juna 1980), which means that herbivores must work physiologically to excrete it from their systems (fig. 8.9). But with the spring flush of high potassium, herbivores on lower sodium diets are thrown into a pathological sodium deficiency. Whatever its cause, spring is a time during which Alaskan ungulates become mineral deficient and sometimes travel great distances to mineral licks. This also seems to be the time when bone and antler chewing occurs. Since the new plant growth is rich in potassium, I suspect the comparatively high sodium, calcium, and magnesium in animal tissue (including antler and bone) are within the taste range of a salt-deficient herbivore.
Fig. 8.9. Relative concentrations of minerals in animal and plant tissues. Because of their drastically different physiologies, animals have more minerals in their tissue than do plants. The one major exception is potassium (K).
Fig. 8.10. Areas of lick use by bison. Within the contiguous United States, recent bison living in the West seemed to have made little use of licks, whereas bison in the eastern deciduous forest relied heavily on licks (horizontal stripes).
Jones and Hansen (1985) compared eastern parkland bison of the United States with those in the Great Plains, showing how soil minerals affect bison distribution and movement. Eastern bison living on more arid soils were restricted in their movements because they were tied to mineral licks (fig. 8.10). The freedom of movement on the Great Plains was made possible by the base-rich soils. Bison on shortgrass prairies did not use licks, an indication to these authors of high levels of calcium, magnesium, sulphur, and sodium in the shortgrass diet. Likewise, as pointed out earlier, northern cervids on Alaskan acid soils readily chew on shed antlers, but this behavior is rare among cervids living on base-rich Texas soils (Krausman and Bissonette 1977).
There is an ongoing controversy as to whether bone-pica and mineral-lick use is due to magnesium deficiency (Jones and Hansen 1985) or sodium deficiency (Weeks and Kirkpatrick 1976). Most researchers in Alaska have concluded that sodium loss is mainly responsible for these behaviors (Tankersley 1981). The comparatively low incidence of bone and antler chewing by herbivores during the Pleistocene (fig. 8.8) indicates sufficient plant mineral quality and lack of seasonal mineral deficiencies among herbivores. Recently Ostercamp (pers. comm.) has found low sodium content in Holocene soils and exceptionally high sodium content in frozen Pleistocene sediments.
Bright Days and Cold Clear Nights
The different albedo of glacial ice and the exposed outer continental shelf seems to have created stable high pressure areas over much of the northern landscape. In interior Alaska these produced low rainfall and low snowfall, mainly clear bright skies in the summer, and cold night skies in the winter. In addition to moisture loss by evaporation, sublimation would have increased. In most habitats, resulting aridity would have favored vegetation with incomplete ground cover, exposing raw soil.
Because the Mammoth Steppe was so far north, it was influenced, either directly or indirectly, by periglacial activity throughout much of its area. Continental ice lowered sea levels, changed upper atmospheric circulation, produced windier conditions, increased erosion, and created loess-forming situations, dramatically changing soil and vegetation. Of course the M
ammoth Steppe vegetation occurred in areas quite distant from actual glaciers, but it was always indirectly affected by glaciers.
This high seasonality, new soils, and inhospitable conditions for woody plants favored grasses and grasslike plants. Today there are many cold- and dry-adapted tundra plants that could and did flourish in such an environment. This low sward, predominantly of herbs, created a varied habitat, with both xeric facies and better-watered facies near streams and downslope from persistent snowdrifts. It would have looked similar to shortgrass regions, or steppes as they are often called in Asia, which today occur at a variety of latitudes (figs. 8.11 and 8.12). Arid late Pleistocene conditions in the north combined xeric grassland forms with xeric tundra species, probably in mixes both within and between local communities.
Likewise, the relative lack of snow and warm middays in autumn would have extended the other end of the growing season. Autumns would have been characterized by striking diurnal temperature fluctuations: bright days and cold clear nights. Clear summer days would have raised summer soil temperatures near the surface, but at deeper levels the ground would have remained cool because bare or almost bare winter soil would have given its heat to winter skies to a much greater extent than the snow- and vegetation-insulated soils of today. But if summer thaws were fast and deep in these conditions, the opposite was true of winter. Heat was pulled from great depths, creating and maintaining permanently frozen ground below reach of the summer thaw. Winters must have been cold, as they are today, creating permafrost conditions that we now see in fossil form relatively far south of present-day permafrost activity. Winter winds would have disturbed the temperature inversions that create our most extreme winter temperatures today, so actual temperature readings might not have been so low, but the chill factor calculated by including winds would have made for very cold conditions.
Fig. 8.11. Asian steppes. The steppes in central Asia are the closest thing remaining to the Pleistocene Mammoth Steppe, and they may serve as analogue. Horses, hemiones, saiga antelope, ferrets, and lions, which became extinct in Alaska, continued to live and thrive in the Asian steppes until the latter part of the Holocene. (Photo by David Murray)
Fig. 8.12. Steppe grasses in central Asia. The grasses are stipa, a common grass of the steppes. (Photo by David Murray)
We have to consider the possibility that the Mammoth Steppe had the unusual combination of relatively deep summer thaws and an accumulation of deeply frozen ground (fig. 8.13). Today that combination seems odd because permafrost now exists where summer soil temperatures are low—north-facing slopes, high altitudes, and in the Arctic lowlands—making it easy to assume that the widespread creation of permafrost farther south had to involve colder summers. But permafrost is produced and maintained by a series of phenomena not distilled in mean summer temperatures. I propose that the Pleistocene pattern was a reversal of what we find today throughout the subarctic or even most of the Arctic. Today summer thaw is shallow, heat is no longer being withdrawn from deep ground, and there has been a Holocene reduction in the extent of permafrost (fig. 8.13).
Although there are some significant temperature (insolation) differences between those of today and those that prevailed during full glacial (isotope stage 2), I argue that, in the far north, insulation is also a significant force. Today’s soils are well insulated all year. Snow limits the amount of heat extracted from the ground in winter, and during summer the vegetation mat decreases the amount of heat gained. These insulators buffer soil temperatures.
If, as we have proposed, there was exposed soil within the open vegetation cover during the glacials, more heat would have been absorbed by the soil in summer. Less rain in summer meant clear summer skies, which meant drier surfaces, which altogether meant greater heat accumulation in the soil during summer—deeper thaws. (This is in the face of a reduced sun angle 18,000 years ago and hence less potential maximum insolation.) It is possible that summer soil temperatures were warmer during the full glacial and that the air temperatures were cooler than those of today. Clear summer skies would have allowed soil temperatures to exceed those of today, while termperatures a meter above ground may have been cooler. Dry soil surfaces would also have lost less heat from evaporation.
Fig. 8.13. Pleistocene and Holocene soil temperatures. Meager snow cover and soil litter during the Pleistocene promoted the removal of large amounts of heat creating permanently frozen ground (permafrost). Less litter also allowed deeper summer thaw. The Holocene situation is reversed.
On subarctic roads, where snow cover is removed, there is deep penetration of frost, and often considerable frost cracking as these cold-soaked areas contract. Contraction cracks appear down the road, like ice wedges and polygonal ground in the making; however, the dark pavement evidently absorbs so much heat in spring, summer, and autumn that it pays back this winter heat debt and no permanently frozen ground accumulates.
Permafrost specialists are familiar with this phenomenon of reduced insulative cover producing permafrost (Washburn 1980), but it has not been emphasized in reconstructions of Pleistocene soil heat dynamics.
Snow considerably retards spring in the north today. At the vernal equinox, when we have the same length of day as the rest of the world, Fairbanks is still in the throes of winter. Snow takes weeks to sublimate and melt. Breakup occurs during seventeen hours of day-light in late April, a full month after equinox (fig. 8.14). Variations of this phenomenon occur all across the north. Most northern plants are very hardy to night frost and would begin to grow in the warmth of the spring sun much earlier if no snow were present. This delayed uncovering of plants, a dramatic boom of plant growth almost immediately after snowmelt, creates a sense that there is no spring in the far north. Indeed, it is less than two weeks between breakup and green-up. This is one of the advantages “Holocene” plants such as spruce and Labrador tea have over Pleistocene grasses. Spruce and Labrador tea are evergreens which can begin to photosynthesize early, losing little by the postponed spring. In Fairbanks, trees are green by the end of May, while the grass sward still wears its winter brown. Grasses have their biomass hidden below the frozen surface and must wait until the soil is thawed; hence they get off to a late start.
Fig. 8.14. Hours of sunshine received at 65° north (Fairbanks, Alaska). This diamond shows how little sunlight is available during the winter. The asymmetric relation of the summer solstice and the four-month green season is also clearly apparent. Long summer days provide considerable photosynthetic potential for plants that are adapted to use the extra hours of light.
Fig. 8.15. Proxy climatic curves summarized from Wijmstra and Van der Hammen 1974: (1) pollen (Fuquene); (2) pollen (Macedonia); (3–5) deep-sea cores of % fauna (V23-82, V23-83, and 180-73); (6 and 7) oxygen isotope ratios (V28-283, 280) from deep-sea cores. Blue Babe is shown in relation to these cold-warm episodes as falling between the two glacial peaks of the last major glacial event.
Although the points emphasized in this and the following chapter deal mainly with glacial episodes, they pertain almost as much to the interstadials, which differ from full glacials only by degree (fig. 8.15). Interglacial episodes on the other hand, are marked by more qualitative change.
Snow would have been minimal and redistributed by wind when it did occur, leaving much land exposed, increasing heat loss, and destroying woody plants, but making the standing dead sward available to ungulates. Moreover, exposed soils would have advanced spring thaw. Deep snows now consume the sun’s energy a month to two months past the spring equinox, but exposed ground absorbs the sun’s heat much faster than snow. Earlier thaw would have lengthened the time nutritious plants were available, thus increasing the growth season for large herbivores.
Pleistocene eolian activity made a dusty landscape: dust in the air, dirty snow, and hazy skies. Sunrises and sunsets must have been spectacular, as the northern sun rolled along the horizon. The dust would have diffused the light and given the north a richer tone.
9
A
RGUMENTS AND CONTROVERSIES ABOUT THE MAMMOTH STEPPE
For a number of years (Guthrie 1966b, 1968, 1980, 1982, 1984a, 1984b) I have argued that data from Pleistocene large mammals indicate an arid steppe existed across northern Eurasia and Alaska during the glacials. Mammalian fossil evidence suggests that during the Pleistocene, ice-free northern areas had both greater carrying capacity and diversity of large mammals than we see today. This was a vast area, with complex communities and wide geographic variations, yet it seemed, nevertheless, to have an integrity that justified my use of a general name: the Mammoth Steppe. Old World faunal evidence and faunal data from Alaska and the Yukon Territory have supported that portrayal (e.g., Matthews 1979; Harington 1978). Sketchy evidence provided by relics within modern vegetation has, at least, not been in conflict with the Mammoth Steppe concept (e.g., Yurtsev 1974; Young 1982).
Several palynologists (pollen specialists), however, have taken issue with the idea of a Mammoth Steppe, arguing on the basis of pollen data that a grass-dominated vegetation did not exist during glacial episodes, nor were large mammalian grazers associated with it (Cwynar and Ritchie 1980; Cwynar 1980; Ritchie and Cwynar 1982; Ritchie 1984; Colinvaux 1980, 1986). At first, I (and other paleoecologists) accepted their data, if not their interpretations, as presented (Guthrie 1982; Matthews 1982) and tried, rather unsuccessfully, to account for the apparent conflict between pollen data and other fossil evidence. During my reconstruction work with Blue Babe, yet another attack on the Mammoth Steppe concept (Colinvaux and West 1984) prompted me, for the first time, to take a critical look at the palynological data itself. I found, certainly to my surprise, that some palynologists have misread their own data. Rather than being at odds with the Mammoth Steppe concept, the pollen record strongly supports the idea, or so I propose in this chapter.
Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe Page 22
