Body Mass Prediction from Dental and Postcranial Measurements in Primates and Their Nearest Relatives

To evaluate alternative hypotheses for the role of mass and muscle-induced forces in joint construction, the allometric scaling relationships of the articular facets of the talus were estimated with phylogenetic regressions. Many articular surfaces scale with significant positive allometry, suggesting that mass-induced forces are an important influence for the bony architecture of synovial joints.

Using a large sample of primates and their nearest living relatives, body mass prediction equations were generated from the articular facet areas of the talus and calcaneus. Those facets that scaled with positive allometry were both accurate and precise. Compared to previously published prediction equations, the novel equations developed for this study were substantially more reliable.

Several methodological debates for body mass prediction were also evaluated. Prediction equations had their highest correlations when species with greater than a 20% difference between sexes are represented by both males and females. Using dental measurements from cercopithecoids housed at the National Museum of Natural History, predictive accuracy was maximized when body mass was predicted using a mean value estimated from a robust sample. Even when only a single individual was represented, tests of predictive accuracy using primates with associated body masses from several localities (Hacienda La Pacifica, Costa Rica; Gombe Stream National Park, Tanzania; Amboseli Reserve, Kenya; and the Duke Lemur Center) demonstrated that prediction equations provide more accurate predictions of species mean values than individual-specific values.

The importance of longitudinal change in body mass was evaluated by comparing coefficients of variation for individual and mean body mass of the populations at La Pacifica, Gombe, and the Duke Lemur Center. Individual coefficients of variation were significantly greater than the population coefficients of variation, which suggests that mean body masses are more stable “targets” of prediction.

Finally, the novel prediction equations were applied to a sample of sympatric primates with associated dental and postcranial elements from the middle Eocene of Wyoming, including Notharctus tenebrosus, Smilodectes gracilis, Omomys carteri, and Hemiacodon gracilis. New body mass predictions suggest two pairs of similarly sized primates: N. tenebrosus and S. gracilis (~2500g), and O. carteri and H. gracilis (~400g). Thus, niche partitioning between closely related taxa was probably achieved through differences in diets, rather than differences in body mass.