The 82nd Annual Meeting of the American Association of Physical Anthropologists (2013)

Applying Extreme Value Analysis in assessing the material properties of the most challenging foods consumed by primates


1Department of Anatomy, Kansas City University of Medicine and Biosciences, 2Department of Anatomy and Neurobiology, Northeast Ohio Medical University, 3Institute of Population Genetics, University of Veterinary Medicine Vienna, 4Department of Anthropology, Rutgers, The State University of New Jersey

Thursday Afternoon, 200DE Add to calendar

The average dietary toughness for closely-related species that exhibit distinct craniodental features for obdurate and non-obdurate feeding (e.g. Sapajus apella versus Cebus olivaceus) can be statistically similar. What often distinguishes food material property (FMP) distributions between these species is a right-tailed skew marking the inclusion of a few exceedingly tough and/or stiff foods in the obdurate feeder’s diet. This observation suggests that maximum FMP values may better differentiate obdurate from non-obdurate feeders and may indicate a propensity for fallback feeding in species with higher maximum FMPs. Here, we consider methods for comparing maximum values within the framework of Extreme Value Analysis – a branch of statistics focused on extreme deviations. We consider randomization approaches, point over threshold (POT) analyses and fitting FMPs to generalized extreme value distributions in pairwise comparisons of obdurate versus non-obdurate feeding capuchin, orangutan and lemur species. As an initial case study, randomization approaches suggest observed difference in maximum dietary toughness between S.apella and C.olivaceus exceeds that expected by chance with significantly greater maxima in the S.apella diet. Preliminary POT analysis similarly suggests S.apella consumes a greater number of tougher foods over 2,000 J/m2. The negative shape parameter in the resulting fit to a Generalized Pareto distribution suggests an upper limit to maximum toughness in S.apella. This upper bound may reflect masticatory apparatus limitations and underlie a switch to manual food processing. We hope these approaches add to our toolkit for studying primate diets and advance the theoretical importance of considering variance in FMPs in obdurate feeders.

comments powered by Disqus