The Evolution of Bison bison: A View from the Southern Plains Patrick J. Lewis, Eileen Johnson, Briggs Buchanan, and Steven E. Churchill

ABSTRACT Morphological changes in late Quaternary bison generally have been defined by a rapid decline in body size and a reorientation of the horns. This morphological shift is used to mark the emergence of modern Bison bison, a transition conventionally thought to have originated on the Northern Plains after 5,000 B.P. Hypotheses regarding the impact of human hunting on this morphological transition vary from negligible to the driving force behind a speciation event. New data from the Southern Plains, however, define better the transition and the role that hunting may have played in driving this event. This research on Southern Plains bison indicates that fully modern bison appear on the Southern Plains much earlier (between 8,000 to 6,500 B.P.) than on the Northern Plains and that this morphological transition also involved a decrease in relative bone strength (robusticity). The morphological transition in bison does not correlate with changes in tool technology, hunting intensity, or hunting techniques. Therefore, the impacts of human hunting do not adequately account for the observed morphological changes in bison morphology. The change in bison morphology on the Southern Plains is correlated better with the spread of the C4 short grass ecosystem between 8,000 to 7,500 B.P. The properties of C4 grasses explain both the size reduction and the gracilization of late Quaternary bison.

INTRODUCTION The North American bison was a critical species for people inhabiting the Plains during the late Quaternary, serving as an important source of food and raw materials for many cultures. More broadly, however, bison impacted the lives of all Plainsdwelling peoples as the keystone species of the Plains ecosystem (Frison 1991; McHugh 1972). Any change in bison would have had a significant impact on the peoples of the Plains, and the evolutionary forces acting to change bison would have impacted all other species that were part of the Plains ecosystem, including humans. For these reasons, defining and understanding the changes in late Quaternary bison and the underlying causes for those changes are essential to the interpretation of prehistoric Plains cultures. Hypotheses proposed to explain the morphological changes in bison following the Wisconsin glaciation constitute some of the most compelling and contentious questions surrounding the evolution of Quaternary macrofaunas (McDonald 1981; Reeves 1973). Most hypotheses implicate either climatic changes (e.g., Wilson 1975) or the effects of

human hunting (e.g., Lott 2002; McDonald 1981) to explain changes in the bison lineage, and are based primarily on data from the Northern Plains. While Southern Plains bison have been studied from an archeological point of view (e.g., Bement 1999; Johnson 1987), hypotheses concerning the emergence of modern bison have yet to be explored fully in this region. New data, however, from a continuous and well-dated sequence of bison metapodials from the Lubbock Lake Landmark site, supplemented with specimens from two other Southern Plains localities, allow an improved understanding of the pattern of morphological change in the late Quaternary bison lineage. These new data have led to the formulation of a new hypothesis regarding the selective force guiding these changes in bison. Changes in the bison lineage during the terminal Pleistocene and early Holocene generally are described as a decrease in overall body size (ca. 20% reduction in linear measurements) and a reorientation of the horn cores (Dalquest and Schultz 1992; McDonald 1981; Wilson 1974a, 1975). These changes are suggested to have occurred gradually beginning in the late Pleistocene and continuing through the Holocene. Modern bison

Bulletin of the Texas Archeological Society 78 (2007)

198

Texas Archeological Society

are reportedly present on the Northern Plains after ca. 5,000 B.P. (uncalibrated radiocarbon age) (McDonald 1981; Wilson 1975). Changes in horn core morphology are vague, however, due to their highly variable nature (McDonald 1981; Wilson 1975) and plasticity (Guthrie 1966), such that attempts to use horn cores in taxonomic and phylogenetic analyses find them to be of limited utility. As such, horn cores, accepted as being fully modern in morphology on the Southern Plains by 10,000 B.P. (Dalquest and Schultz 1992), are not addressed further. Both human hunting and climatic change have been posited as the principle force behind the decrease in bison body size. Human hunting hypotheses generally suggest that hunting selected either for smaller body sizes or against larger body sizes (Lott 2002; McDonald 1981). Climatic hypotheses associate the warming temperatures of the Holocene with decreases in body size, following Bergmann’s rule (Bergmann 1847) and the modern size cline (a decrease in body size from north to south) seen in today’s bison (Wilson 1975). Most of the evidence given in support of these hypotheses is derived from Northern Plains populations from multiple localities, and is based principally on cranial characteristics known to be highly variable. To define better the rate and pattern of change in bison populations, a morphologically conservative element should be used and the sample broadened to include populations beyond the Northern Plains. Correlations between changing bison morphology and parameters associated with human hunting and climatic change hypotheses then can be examined. This study, therefore, focuses on Southern Plains bison metapodials to determine if the timing and pattern of change is similar to that defined on the Northern Plains and which of the competing hypotheses is best supported.

Kaisan 1947; Wilson 1975; Wyckoff and Dalquest 1997), this study focuses on the metapodials. Metapodials are functionally important weight-bearing elements for bison (Bedord 1974) and, more broadly, for the family Bovidae (Scott 1985). Metapodials are conservative morphologically, showing little variation in shape across the bovids (Scott 1985). Any significant changes in metapodial morphology, therefore, likely represent a functional, behavioral, or genotypic shift in bison relevant to hypotheses about the forces that were acting to shape the bison skeleton. Metapodials also are common elements in late Quaternary faunal assemblages (Kreutzer 1992), allowing for robust statistical analyses. Metapodial size, shape, and robusticity (relative cortical bone thickness) data were collected from osteometric and radiographic examination of a temporal series of 463 late Quaternary bison metapodials from three sites on the Southern Plains (Figure 1). Of those metapodials, 254 specimens were from Lubbock Lake (Johnson 1987) in western Texas. Dates were assigned to the Lubbock Lake specimens using multiple radiocarbon ages associated with archeological features and geologic strata (Holliday et al. 1983, 1985). In addition, 75 metapodials from Cooper (Bement 1999) and 134 from Certain (Buehler 1997), both sites located in western Oklahoma, were examined. These metapodials were dated according to radiocarbon assays and known ages of diagnostic stone tools (Bement 1999; Buehler 1997).

METHODS While much of the historic research regarding bison phylogeny has relied heavily on cranial characteristics (Guthrie 1990; McDonald 1981; Skinner and

Figure 1. Southern Plains and archeological sites where the fossil bison metapodials in this analysis were recovered.

Lewis, et al.—The Evolution of Bison bison 199 Proximal and distal anteroposterior and mediolateral width measurements were collected at the extremes on complete metapodials at the proximal and distal ends using sliding calipers. Unfused metapodials were not included in the study. Overall element length was taken at the extreme using an osteometric measuring board. Shaft diameters in the anteroposterior view and the mediolateral view also were collected from X-rays using sliding calipers. Robusticity (relative cortical bone thickness) was calculated using a published formula incorporating medullary cavity area, cortical bone thickness, and element size adjusted for estimated body mass (Trinkaus et al. 1994). Data were tested using ANOVA and Pairwise post-hoc Bonferroni/Dunn tests. A confidence level of 0.05 was set for all statistical tests.

only significant difference between the bison populations is between the population dating to 8,000 B.P. and the population dating to 6,500 B.P. A slight shift in metapodial shape appears to accompany this size decrease (Figure 3). Metapodials of similar length vary significantly by width (at the midshaft) when specimens from the group dating to 8,000 years and older are compared to specimens dating 6,500 years old and more recent (see Figure 3). A decrease in bone robusticity also is found, using a standard robusticity index adjusted for body size referred to as J/stand (Trinkaus et al. 1994). Again, this decrease in bone robusticity, or gracilization, of the metapodials occurs between 8,000 B.P. and 6,500 B.P. in concert with the change in body size and metapodial shape (Figure 4).

DISCUSSION RESULTS In general, the pattern of morphological change Data from the Southern Plains indicate a difon the Southern Plains is quite different than that ferent pattern of size change than that seen on the reported on the Northern Plains. The Southern Plains Northern Plains. The Southern Plains pattern is one pattern is best summarized as relative stasis before of relative stasis in body size before 8,000 B.P. and 8,000 B.P., a rapid change in size, shape, and following 6,500 B.P. (Figure 2). A rapid decrease robusticity between 8,000 B.P. and 6,500 B.P., folin size occurs between 8,000 and 6,500 B.P. This lowed by stasis after 6,500 B.P. Bison reach their pattern is found in both sexes and for both metatarmodern form on the Southern Plains, then, somesals and metacarpals. No statistically significant time in the 1,500 year period between 8,000 B.P. difference exists between the populations 8,000 and 6,500 B.P., well before the appearance of modyears old and older, nor does a significant differern morphology in Northern Plains bison at ca. ence occur between those populations 6,500 years 5,000 B.P. (McDonald 1981; Wilson 1975). The old and more recent. Pairwise tests find that the extremely rapid shift in Southern Plains metapodial morphology also contrasts starkly with the gradual pattern of change reported for Northern Plains bison (Wilson 1974a). With the pattern of morphological change on the Southern Plains established, variables associated with both the human hunting hypothesis and the climatic change hypothesis can be examined. The period of rapid change in bison did not correlate with an increase in population pressure (Haynes 2002; Meltzer 2000). Bison were hunted intensively on the Southern Plains both before and after this change without any sigFigure 2. Change in metacarpal length for male and female bison for the five nificant change in bison morpholLubbock Lake populations. ogy. Also, no selection against

200

Texas Archeological Society

dramatic change occur in tool technology (Frison et al. 1976; Johnson 1997). While it may be argued that the change from Paleoindian to Early Archaic tools constitutes a major change in technology, when undeniable paradigmatic shifts in hunting strategy and tool type do occur (e.g., the introduction of the bow and arrow, guns, and horses) no change in bison morphology is found. Likewise, increased predation from humans should be associated with increased herd movements. Such increases in movement would increase the Figure 3. Change in metacarpal shape as based on log length (lnL) to log bending forces and, therefore, be mid-shaft width (lnW) measurements for ancient bison (Ba=8,000 B.P. and associated with more robust bones. older) and modern bison (Bb=6,500 B.P. and more recent). Just the opposite is found, as bone robusticity decreases sharply with the decrease in body size. No large animals occurred during the period of change, obvious mechanism is apparent that would connect as kills continue to have females, juveniles, and changes in bison morphology with human hunting males, and exhibit similar standard deviation levels based on the variables examined here. for metric variables. Selection against large indiIn general, the changes in bison morphology viduals also would not explain why a change in appear to be explained better by the changing clishape and robusticity of the metapodials occurs. mate. Body size does decrease with warming temMcDonald (1981) suggests that the change in peratures, as would be expected (Wilson 1975). A size was not driven by selection of large animals, correlation exists between metapodial shape and but that smaller bison would have had an advantage vegetation density in modern bovids, whereby over larger bison. Smaller bison supposedly would be better able to escape and to breed more quickly. At issue with this hypothesis, however, is the fact that humans were not the primary predator of bison (Fuller 1959; Lott 2002; McHugh 1972). While humans were hunting bison during the period of rapid morphological change in the lineage, wolves remained their primary predator. It does not follow from a natural selection point of view for bison to become smaller in order to escape their secondary predator if such a change increased the likelihood of their being killed by their primary predator. No major shift in hunting strategy precedes the change in Figure 4. Boxplot of log J/standard (J/stand) for male metatarsals, estimating bone robusticity. Median, 75th percentile, and 90th percentile are depicted. bison morphology, nor does a

Lewis, et al.—The Evolution of Bison bison 201 species associated with denser vegetation typifrom Southern Plains localities (Fredlund 2002; cally have relatively wider metapodials than those Fredlund et al. 2002, 2003) offer a possible explain areas where vegetation is sparse (Scott 1985). nation for both the decrease in body size and in As the amount of brushy vegetation on the Southrobusticity. A dramatic and rapid shift from C3 to C4 grass occurs on the Southern Plains beginning ern Plains decreased, bison metapodials become at ca. 10,000 B.P. and completed by 8,000 B.P., thinner. Lastly, given the earlier date of modern just when changes in bison morphology begin to morphology on the Southern Plains than on the appear in the fossil record. C4 short grasses have Northern Plains, the pattern of morphological an advantage over C3 long grasses under warm change spreads south to north along with the Holocene conditions, as C3 grasses are less tolertemperature gradient. ant of warm, dry conditions. The decrease in biWhen changes in temperature and rainfall are son body size begins at approximately the same examined with the changes in bison morphology, time that tall grass (C3) ecosystems begin to give however, several shortcomings of the climatic hyway to short grass (C4) ecosystems on the Southpothesis become evident. Based on both floral and ern High Plains in the early Holocene. Additionfaunal changes (see Johnson 1987), the late Pleisally, both C4 grasses and smaller bison spread tocene and early Holocene are characterized by a northward during the Holocene. Once C4 grasses warming and drying trend on the Southern Plains. become dominant on the Southern High Plains ca. This trend peaks at ca. 7,000 to 5,000 B.P., during 8,000 to 7,500 B.P., C3 grasses never rebound the deposition of stratum 3 at Lubbock Lake (Figdespite the cooling climate of the late Holocene ure 5), and is followed by a return to cooler, moister (Fredlund 2002; Fredlund et al. 2002, 2003). conditions. Bison morphological change, therefore, The timing and pattern of the short grass ecooccurs as the climate reaches its warmest and driest system’s spread follows that of all morphological conditions. But, if bison morphology were tracking changes in bison metapodials on the Southern closely with temperature and rainfall, then the dePlains. The productivity and nutrition of C4 grasses crease in body size would have begun much earlier may explain why the transition in grass types when the climate first began to change. While the would cause such a suite of changes in bison morwarming and drying trends appear to explain the phology. C4 grasses are more productive per acre progression toward relatively smaller bison in the than are C3 grasses (Howe 2000). With greater late Quaternary, it does not explain the period of productivity, bison would not be required to range rapid change between 8,000 and 6,500 B.P. as far to find similar amounts of food. Such a Likewise, if the morphological changes during this period were strictly phenotypic responses to changing temperatures, then they should have reversed when the conditions cooled after ca. 4,500 B.P. (stratum 4 at Lubbock Lake). Also, the decrease in metapodial robusticity during the late Quaternary is inconsistent with a drying, degrading habitat where bison would have been forced to range further to find adequate food. This increased movement would have increased bending forces on the skeletal elements of the limbs and led to more robust metapodials, not less robust Figure 5. Changes in temperature (T, grey points) and precipitation (P, black metapodials. The climatic model, points) for the 5 strata at Lubbock Lake. Stratum 1 dates from ca. 11,500as it stands, fails to address these 11,000 B.P., stratum 2 dates from ca. 11,000-8,000 B.P., stratum 3 dates from ca. 8,000-5,000 B.P., stratum 4 dates from ca. 5,000-1,000 B.P., and inconsistencies. stratum 5 dates from ca. 1,000 B.P. to today (Johnson 1987). Recent phytolith analyses

202

Texas Archeological Society

change in grass types may have led, therefore, to a decrease in herd movements and a subsequent reduction in range during the late Quaternary. This reduction in movement patterns would have led to decreased bending forces on the skeletal elements of the limbs and ultimately to less robust metapodials. The pattern of robusticity change would be a rapid shift during the period of transition from C3 to C4 grasses, followed and preceded by relative morphological stasis. Despite higher productivity, C4 grasses possess a larger cell wall and less protein than their C3 counterparts, and are thus less nutritious (Howe 2000). Such a change in nutritional quality may have selected for bison with smaller body sizes, as smaller mammals generally require absolutely fewer calories and have reduced protein demands (Calder 1984). Larger mammals have higher total rates of metabolism than do smaller species, and the increase in metabolic rate is not proportional to mass (Calder 1984). A doubling of mass necessitates a 64% increase in basal metabolic rate, for example, while decreasing the mass by half reduces the basal metabolic rate by only 39% (McNab 1990). A mammal, therefore, can respond to decreases in available energy by either reducing mass or reducing the metabolic rate at a fixed mass (McNab 1990). Thus, change in bison body size would have been rapid during the transition from C4 to C3 grasses (and may have occurred passively, as bison fed on protein-poor C4 grasses), and been followed and preceded by general morphological stasis.

NEW HYPOTHESIS The new hypothesis proposed, therefore, is that the changing grasslands of the Southern Plains drove morphological changes in the late Quaternary bison lineage. This hypothesis, focused on the spread of an abundant but nutritionally poor primary food source for bison, explains both the decrease in bone robusticity and the decrease in body size. The timing of morphological changes and the earlier appearance of modern bison on the Southern Plains also are explained, as both modern bison morphology and C4 grasses continued to spread northward as the late Quaternary climate dried and warmed. The effect of changing vegetation on late Quaternary bison provides increasing evidence that the dramatic changes in the North American fauna following the

Wisconsin glaciation were driven primarily by the altering environment rather than hunting pressure from humans.

ACKNOWLEDGEMENTS The authors thank Dr. Leland Bement and Kent Buehler (Oklahoma Archeological Survey) for access to Cooper and Certain site materials, their help in collecting data, and their insights into bison evolution; and Dr. Glen Fredlund (University of Wisconsin-Milwaukee) for sharing data. Research was funded by the Museum of Texas Tech University. This manuscript represents part of the ongoing Lubbock Lake Landmark regional research into late Quaternary adaptations to ecological, climatic, landscape, and cultural changes on the Southern Plains.

REFERENCES CITED Bedord, J. N. 1974 Morphological Variation in Bison Metacarpals and Metatarsals. In The Casper Site: A Hell Gap Bison Kill on the High Plains, edited by G. C. Frison, pp. 199-240. Academic Press, New York. Bement, L. C. 1999 Bison Hunting at Cooper Site: Where Lightening Bolts Drew Thundering Herds. University of Oklahoma Press, Norman. Bergmann, C. 1847 Über die Verhältnisse der wärmeökonomie der Thiere zu ihrer Grösse. Göttinger Studien 1:595-708. Buehler, K. J. 1997 Where’s the Cliff? Late Archaic Bison Kills in the Southern Plains. In “Southern Plains Bison Procurement and Utilization from Paleoindian to Historic,” edited by L. C. Bement and K. J. Buehler, pp. 135-144. Plains Anthropologist 42 (159). Calder, W. A. 1984 Size, Function, and Life History. Harvard University Press, Cambridge. Dalquest, W. W. and G. E. Schultz 1992 Ice Age Mammals of Northwestern Texas. Midwestern University Press, Wichita Falls, Texas. Fredlund, G. G. 2002 Phytolith Evidence for Change in Late Pleistocene and Holocene Grassland Composition on the South-

Lewis, et al.—The Evolution of Bison bison 203 ern High Plains. Paper presented at the 60th Annual Plains Anthropological Conference, Oklahoma City. Fredlund, G. G., V. T. Holliday, and E. Johnson 2002 The Rise of the Short Grass Steppe Ecosystem on the Southern High Plains: Evidence from the San Jon Site, Eastern New Mexico. Paper presented at the 2nd International Past Grasslands Research Conference, St. Cloud, Minnesota. 2003 Vegetation Change on the Southern High Plains During the Pleistocene-Holocene Transition. Paper presented at the 16th INQUA Congress, Reno, Nevada. Frison, G. C. 1991 Prehistoric Hunters of the High Plains. Academic Press, San Diego. Frison, G. C., M. Wilson, and D. J. Wilson 1976 Fossil Bison and Artifacts from an Early Altithermal Period Arroyo Trap in Wyoming. American Antiquity 41:28-57. Fuller, W. A. 1959 Behavior and Social Organization of the Wild Bison of Wood Buffalo National Park, Canada. Arctic 13:2-19. Guthrie, R. D. 1966 Bison Horn Cores—Character Choice and Systematics. Journal of Paleontology 40:738-740. 1990 Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe. University of Chicago Press, Chicago. Haynes, G. 2002 The Early Settlement of North America: The Clovis Era. Cambridge University Press, Cambridge. Holliday, V. T., E. Johnson, H. Haas, and R. Stuckenrath 1983 Radiocarbon Ages from the Lubbock Lake Site, 1950-1980: Framework for Cultural and Ecological Change on the Southern High Plains. Plains Anthropologist 28:165-182. 1985 Radiocarbon Ages from the Lubbock Lake Site: 1981-1984. Plains Anthropologist 30:277-291. Howe, H. F. 2000 Grass Response to Seasonal Burns in Experimental Plantings. Journal of Range Management 53:437-441.

Lott, D. F. 2002 American Bison: A Natural History. University of California Press, Berkeley. Kreutzer, L. A. 1992 Bison and Deer Bone Mineral Densities: Comparisons and Implications for the Interpretation of Archaeological Faunas. Journal of Archaeological Science 19:271-294. McDonald, J. N. 1981 North American Bison: Their Classification and Evolution. University of California Press, Berkeley. McHugh, T. 1972 The Time of the Buffalo. Alfred A. Knopf, New York. McNab, B. K. 1990 The Physiological Significance of Body Size. In Body Size in Mammalian Paleobiology: Estimation and Biological Implications, edited by J. Dammuth and B. J. MacFadden, pp. 11-23. Cambridge University Press, Cambridge. Meltzer, D. J. 2000 Human Responses to Middle Holocene (Altithermal) Climates on the North American Great Plains. Quaternary Research 52:404-416. Reeves, B. O. K. 1973 The Concept of an Altithermal Cultural Hiatus in Northern Plains Prehistory. American Anthropologist 75:1221-1253. Scott, K. M. 1985 Allometric Trends and Locomotor Adaptations in the Bovidae. Bulletin of the American Museum of Natural History 179:197-288. Skinner, M. F. and O. C. Kaisen 1947 The Fossil Bison of Alaska and Preliminary Revision of the Genus. Bulletin of the American Museum of Natural History 89:123-256. Trinkaus, E., C. E. Churchill, and C. B. Ruff 1994 Postcranial Robusticity in Homo, II: Humeral Bilateral Asymmetry and Bone Plasticity. American Journal of Physical Anthropology 93:1-34.

Johnson, E. 1987 Lubbock Lake: Late Quaternary Studies on the Southern High Plains. Texas A&M University Press, College Station, Texas.

Wilson, M. V. 1974a The Casper Local Fauna and its Fossil Bison. In The Casper Site: A Hell Gap Bison Kill on the High Plains, edited by G. C. Frison, pp. 125-171. Academic Press, New York.

1997 Late Quaternary bison utilization at Lubbock Lake, Southern High Plains of Texas. In “Southern Plains Bison Procurement and Utilization from Paleoindian to Historic,” edited by L. C. Bement and K. J. Buehler, pp. 45-62. Plains Anthropologist 42(159).

1974b History of the Bison in Wyoming with Particular Reference to Early Holocene Forms. In Applied Geology and Archaeology: The Holocene History of Wyoming, edited by M. Wilson, pp. 91-99. Report of Investigations No. 10. The Geological Survey of Wyoming, Laramie.

204

Texas Archeological Society

1975 Holocene Fossil Bison from Wyoming and Adjacent Areas. Masters of Science Thesis, University of Wyoming, Laramie. Wyckoff, D. G. and W. W. Dalquest. 1997 From Whence They Came: The Paleontology of Southern Plains bison. In “Southern Plains Bison Procurement and Utilization from Paleoindian to Historic,” edited by L. C. Bement and K. J. Buehler, pp. 5-32. Plains Anthropologist 42(159).