Environmental information in a recent bone assemblage: roles of taphonomic processes and ecological change

Taphonomic processes have the potential to obscure or enhance original associations between paleoenvironments and organisms through selective destruction, post-mortem transport, and time-averaging. Ecological changes through time in habitat utilization by living organisms also can blur distinctions between faunas or floras associated with particular habitats. If taphonomic processes affect habitat specificity in bone assemblages, then this should be revealed in time sequences of remains from particular habitats. We analyzed Recent surface bone in six habitats of Amboseli National Park, Kenya, to determine whether taxonomic composition varied with the degree of taphonomic alteration (as indicated by bone weathering stage). The dataset consisted of 1362 separate bone occurrences classified according to their state of weathering. These bone occurrences represented 27 mammalian taxa from bush, openwood, densewood, lakebed, plains, and swamp habitats. The distinctiveness of the bone assemblages was measured by the quantified Dice coefficient, which accounts for both the presence or absence of taxa and their relative proportions. The differences among the weathering stage subsamples were compared with known environmental changes at Amboseli over the estimated time period of bone accumulation (15–25 years). Habitat distinctiveness based on the bone assemblages was lower in subsets with higher weathering stages, although not all habitats showed this to the same degree, and open habitats (e.g., plains, lakebed) generally remained distinct from closed habitats (densewood, bush). The greatest change in the similarity among habitats occurred between weathering stages 2 and 3, which corresponds to deaths between 1969 and 1973, a period of woodland retreat and expansion of open habitats in the park. We attribute the loss of environmental resolution in the weathered bone subsamples primarily to time-averaging and the shifting of habitat boundaries. The time period represented by these bone assemblages is small compared to the time-averaged paleosol or land surface assemblages in the fossil record, suggesting that ecological fluctuations can compromise interpretations of habitat heterogeneity in fossil assemblages. Our study indicates that it is probably the magnitude of ecological changes rather than time per se that limits resolution. If distinct paleohabitats can be supported based on time-averaged fossil evidence, however, this implies that ecological fluctuations were not sufficient to homogenize these distinctions over the period of time-averaging.