MODERN VERTEBRATE TRACKS FROM LAKE MANYARA, TANZANIA AND THEIR PALEOBIOLOGICAL IMPLICATIONS

We studied mammal and bird track formation at the northern edge of Lak e Manyara, Tanzania, to develop models for interpreting fossil tracks and trackways. Lake Manyara is a closed-basin, alkaline lake in the Ea st African Rift System. The area has a high vertebrate diversity, allo wing us to investigate tracks in an environment similar to that of man y ancient track-bearing sequences. Three study sites, two on mud flats adjacent to the lake margin and a third on a delta floodplain, provid ed contrasting environments in which to assess the types of biological data that can potentially be extracted from fossil trackways. Our cen suses of mammals and their tracks revealed that most species that occu r within the study area leave a track record, and that common species leave abundant tracks, although numbers of trackways are not proportio nal to numbers of individuals. Logarithmic increases in track sampling area yield a linear increase in the proportion of both the medium and large-sized local mammals represented in a track record. Transect vs. area mapping methods produced different censusing results, probably b ecause of differences in monitoring periods and areal coverage. We dev eloped a model of expected track production rates that incorporates ac tivity budget and stride length data in addition to abundance data. By using these additional variables in a study of diurnal birds, we obta ined a much better estimator relating track abundance to trackmaker ab undance than that provided by census data alone. Proportions of differ ent types of tracks predicted by the model differ significantly from t he observed proportions, almost certainly because of microenvironmenta l differences between the censusing and track counting localities. Cen suses of fossil tracks will be biased toward greater numbers of deposi tional-environment generalists and away from habitat-specific species. Trackways of migratory animals were dominantly shoreline-parallel, wh ereas trackways of sedentary species were more variable. A strong shor eline-parallel environmental zonation at the Alkaline Flats site exert ed an influence on trackmaker distribution patterns, initial track for mation, and track preservation. Variations in habitat usage by differe nt species, as well as species abundance and directionality of movemen t, were all important in determining the number of preservable tracks a species produced within a given environmental zone. Fossil trackways are time-averaged, although over entirely different temporal scales t han are bones. Unlike bones, tracks are not space-averaged. Therefore, wherever possible, fossil track and bone studies should be used to co mplement each other, as they provide fundamentally different pictures of paleocommunities. Tracks provide ''snapshot'' views of localized as semblages of organisms useful in reconstructing autecological relation ships, whereas bones yield a broader image of a local fauna in which s easonal and microenvironmental variation are more commonly smoothed ou t.