Abstract # 3119 Poster # 81:

Scheduled for Saturday, September 17, 2011 07:00 PM-09:00 PM: Session 14 (Salon G (Sixth Floor)) Poster Presentation


S. E. Johnson1, A. D. Gordon2, D. A. Raichlen3, S. Tecot3, S. M. Holmes1, C. Ingraldi1 and J. L. Verdolin4
1Department of Anthropology, University of Calgary, Calgary, AB T2N 1N4, Canada, 2Department of Anthropology, University at Albany, 3Department of Anthropology, University of Arizona, 4National Evolutionary Synthesis Center

A Lévy walk is a random walk with a distribution of step lengths (discrete movements) that is heavy-tailed and follows a power law. Lévy walks have been shown by some researchers to be an optimal search strategy for patchily distributed resources when animals possess limited knowledge of resource locations. Alternatively, with evenly distributed resources, animals may use step lengths with an exponential distribution (Brownian motion). We examined the fit of two Lévy walk (power law and truncated Pareto) and one exponential model to step length distributions in gray-headed lemurs (Eulemur cinereiceps), red-bellied lemurs (E. rubriventer), and black-and-white ruffed lemurs (Varecia variegata) in three rain forests in southeastern Madagascar (Agnalazaha, Ranomafana, and Kianjavato-Vatovavy, respectively). GPS locations were obtained during focal animal sampling on 2-3 social groups of each species (N=590-1271 total steps for each species). Because these species rely mainly on fruit (a patchy resource), we predicted that they would follow a Lévy walk pattern. However, step length distributions best fit exponential models for each species (AIC for gray-headed lemurs: power law=-2255.8, truncated Pareto=-2393.4, exponential=-2566.8; red-bellied lemurs: power law=-4078.6, truncated Pareto=-4617.7, exponential=-5441.3; black-and-white ruffed lemurs: power law=-2185.5/-1799.7, truncated Pareto=-2433.9/-2046.3, exponential=-2673.9/-2235.5). Identical patterns were observed across seasons. These results suggest several possible scenarios to be investigated further, including (1) these species use evenly distributed resources, and (2) they may exploit knowledge of resource locations for non-random searches.