Bison Behavior and Development

To understand the annual cycle of past bison herds, analogues are sought in the modern herds. Relating modern herd dynamics to prehistoric behavior has problems often related to confinement and management practices affecting modern herds. But many of the actions of animals on large preserves provide models to be applied to prehistoric settings. Modern bison behavior is generally considered an appropriate analog for Bison bison. Hunters pursuing Bison antiquus or Bison occidentalis may have had to deal with behavioral characteristics somewhat different from the modern species. However, the redundant use of similar landscape features to kill modern and ancient bison species suggests that bison behavior related to handling has not changed significantly in the last 11,000 years. Bison kills of all ages can be found in arroyo traps, dune fields, at the base of cliffs, and possibly constructed corrals. Other aspects of ancient bison annual movements, herd sizes, and behavior are less well extrapolated from an archeological context.

Stable Isotope Analysis

Studies designed to compile data related to the movements of prehistoric Bison bison, let alone Bison antiquus, have met with little success. One such study of Bison bison utilizes stable carbon isotopes found in the bones to track bison movement from one grassland assemblage to another. Underlying this study is the metabolic relationship between the plant foods selected by bison and the carbon isotope ratio incorporated into the bone. The stable carbon isotopes contained in the bone collagen is directly relatable (linked) to the type of grasses consumed. If the bison move between two habitats with different grass composition, then their bone isotope ratios would be at variance with the isotope composition of both grasslands. Of course, this assumes that the composition of the grassland is known. This is not always easy to do when dealing with prehistoric areas where plant pollen, phytolith, or macrofossil remains are not preserved or where major shifts in plant communities are suspected. In areas where plant communities are stable and contain either an extensive mosaic of species or a single suite of species, the movement of bison from one area to another will not be detectable in the bone isotopes.

The three kills at the Cooper site provide the opportunity to track changes in stable carbon and nitrogen isotopes in roughly contemporaneous herds. Unfortunately, no pollen and very few phytoliths were preserved in the Cooper site deposits. Thus, it is impossible to reconstruct the grassland composition near the site.

The presence of three closely timed kill deposits, however, provides a means to overcome this problem. Comparing the isotopes from individuals in one bonebed with those from the others should indicate if any of the herds are from distinctly different grassland areas. If samples from each of the three kills show no differences, then it can be assumed that the animals followed similar feeding/wandering cycles that may or may not include migratory patterns. If, however, differences are evident, then this could support the possibility that the herds followed different patterns. Ultimately, differences could indicate the presence of resident and migratory herds.

Bone preservation is another concern in these studies. Bone that has undergone extensive chemical weathering will contain diagenetically altered carbon isotopes. Such samples do not render acceptable results. Fortunately, relationships between carbon and nitrogen stable isotopes are known to indicate that diagenesis has occurred.

In the Cooper samples, the ratio of stable carbon isotopes between the collagen and apatite fractions of the bone is similar to that found in modern bone. A similar situation is seen in the ratio of carbon to nitrogen. Low preservation of carbon, however, suggests some diagenetic alteration has occurred.

The stable isotopes, then, of six samples, two from each kill, are far from conclusive. A cross plot of the stable carbon isotope ratios from the collagen against the apatite form two clusters. The first cluster on the left side of the graph contains two points. The second cluster contains four samples. When identified to bonebed and age of individual it is seen that the first cluster consists of a two year old from the Upper Kill and a two year old from the Lower Kill. The second cluster is composed of a five year old from the Upper Kill, two four year olds from the Middle Kill, and a four year old from the Lower Kill.

The reason for the separation of the two juveniles from the older individuals is not currently understood. A similar relationship between young and old is evident in a plot of Bison bison remains from the Texas panhandle. The segregation based on age may be tied to fractionation of carbon from a diet of mother’s milk. Further research is needed in this area.

Another pattern is the trend for the two Upper Kill samples to lie to the left of the other bone samples. The five-year old from the Upper Kill is removed from the two Middle Kill samples by approximately the same distance as the Middle Kill samples are from the Lower Kill. The Lower Kill animal consumed a higher proportion of C4 grasses than did the two Middle Kill animals. The Upper Kill animal consumed even less C4 grasses than did the Middle Kill animals. Although not conclusive, this suggests that individuals from each of the three kill herds consumed grasses with different stable carbon ratios. This suggests that each herd made use of different grassland areas or ranges. Additional samples are needed to confirm this trend.

Skeletal Growth as a Clue to Bison Herd Movement

Another study utilizes a ratio of the length of the femur to length of metatarsal by age to indicate growth differences related to forage quality rather then genetics. This ratio was devised to explain differences in body size among ungulate herds of identical genetic composition inhabiting similar geographical regions. In its application to deer populations on adjacent Alaskan islands where breeding populations intermingled, it was found that genetic differences were insufficient to account for the observed somatic differences in the two island populations. The application of this ratio was altered to use hind foot length (hoof to proximal metatarsal) rather than just metatarsal length to allow easier field measurements. However, the femur/metatarsal ratio was the preferred ratio when available. The deer plot by sex showed better separation between males of the two herds than the smaller females. However, the differences were also exhibited in the female groups.

Applying this technique to bison populations is founded in studies such as the analysis of the Badlands National Park bison in South Dakota. In this study, two bulls and three bison cows from a herd in Colorado were introduced to a herd in South Dakota in an attempt to provide greater genetic diversity to the breeding population. The Colorado animals were noticeably smaller and of different pelage characteristics than the individuals in the host herd which was of Nebraska descent. Analysis of mtDNA revealed no genetic differences between the two bison populations, and the observed differences were attributed to habitat inequalities.

Initial plots of appendicular element lengths indicate the female animals from the Upper Kill are larger than females in either of the other two kill levels at Cooper. Radii length from female skeletons older than 2 average 353.3 mm, 340.8 mm, and 338.0 mm for the Upper, Middle, and Lower kills, respectively. The Upper Kill radii are on average 15 mm longer than radii from either of the other kills. Similar size differences are seen in the other long bones. The source of this variation is yet to be determined. Genetic analysis awaits funding. The possibility that environmental factors and not genetic factors are at work have been partially explored. The stable isotope analysis provides some indication that the herds killed at Cooper migrated from areas of the plains where grass composition was slightly different.

Applying the analytical technique to the bison remains from the Upper and Middle kills at Cooper provide interesting, and somewhat unexpected, results. The Lower kill was omitted from this analysis due to its small sample size. Plots of the ratio of femur length to metatarsal length against age yielded differing regression lines for the Upper and Middle kill animals. The regression line for the Middle Kill animals is considerably steeper than that for the Upper Kill population. This suggests that the Middle Kill population subsisted in a differnet nutritional niche than did the Upper Kill animals. The level of significance or robusticity of this analyis is difficult to evaluate until similar analyses are conducted on larger numbers of animals--preferably from living bison groups. These results are only presented to suggest that the animals from these herds were subjected to different nutritional and environmental conditions during growth.

The results of this analysis and the stable isotope analysis, combined with the near-contemporaneity of the three herds, as indicated by trampled bone fracture morphology and geomorphological analysis leads to the conclusion that the Cooper herds came from different grazing pastures. Whether all herds were migratory or there is a combination of resident and migratory herds in the Cooper deposits can not be ascertained at this time.

skullbut.jpg (5711 bytes)