Correlations and Depletion Patterns of Marrow Fat in Caribou Bones

Abstract

We used linear regression to examine the relationship between marrow fat and marrow fat of the other 5 long leg bones (tibia, metatarsus, humerus, radius, and metacarpus) and mandibles using paired bones from individual caribou (Rangifer tarandus granti) in Alaska. Correlations were high in all regressions (range of r = 0.90-0.98) and suggested that all bones were useful as gross indices to relative body condition. However, when interpreting fat content of bones other than femurs and making between-bone comparisons, there is merit in indexing fat values to a femur standard. The percent marrow fat in all leg bones of individual caribou was more uniform than in moose (Alces alces). The sequence of fat depletion in caribou long bones was closer to that of moose than other wild ungulates. Fat depletion in both moose and caribou advances more rapidly in proximal long bones than in more distal ones. Mandible marrow fat can be used as a relative condition index if limitations are recognized. J. WILDL. MANAGE. 51(2):365-371 Field biologists are frequently faced with assessing the cause of death in large mammals when only skeletal parts remain. This situation is particularly common in predator-prey studies. Inferences about the physical condition of prey selected by predators have major implications, particularly when the condition of the prey population is known. Mech and DelGiudice (1985) have discussed potential limitations of using marrow fat in ungulate long bones as an index of relative body condition. For a number of reasons (Cheatum 1949) marrow fat in the has emerged as the relative condition index of choice to estimate relative body condition at death in many ungulate species. The substantial interest in marrow fat as a relative condition index has stimulated much investigation and refinement of techniques for marrow fat determinations since Harris (1945) reported that advancing malnutrition changed the bone marrow fat content in white-tailed (Odocoileus virginianus) and mule (0. hemionus) deer. Femur marrow fat has been estimated by ?6 general approaches: (1) ether extraction (AOAC method in Horwitz [1965:346], cited by Greer [1968]), (2) visual estimation (Harris 1945, Cheatum 1949), (3) compression (Greer 1968), (4) oven drying (Neiland 1970), (5) reagent-dry assay (Verme and Holland 1973), and (6) freeze drying (Hunt 1979). Ether extraction techniques yield reliable quantitative data but are tedious, expensive, and potentially dangerous. Consequently, they are used infrequently in wildlife studies. Subjectivity of the visual estimation technique limits its usefulness. The compression method is only quasiquantitative. The several drying techniques are relatively fast, inexpensive, and sufficiently accurate for most wildlife work (Hunt 1979). A logical extension of the relative condition index has been the use of marrow from other leg bones (Ballard et al. 1981, Fong 1981, Peterson et al. 1982) and from mandibles (Baker and Leuth 1966, Purol et al. 1977, Snider 1980, Ballard et al. 1981). Use of these additional bones as relative condition indices is appealing for many reasons. On frozen, intact carcasses femurs are much more difficult to extract than other bones. Large carnivores often consume all or portions of the femurs but frequently leave distal bones and mandibles intact. Mandibles, in particular, are often available in large numbers from hunterkilled animals. While investigating natural mortality in Alaska's western Arctic caribou herd, we found many caribou carcasses that were missing femurs, but contained other long bones or mandibles. To determine if these other bones were useful as indices of relative body condition, we tested the correlation between paired marrow values for femurs vs. other leg bones and mandibles. Those correlations and fat depletion patterns of those bones are the subject of this paper. Many Alaska Dep. Fish and Game employees