Pumas Remodeled - Comparing Traditional Capture-Recapture to Mark-Resight and Spatial Mark-Resight Models for Estimating Puma Densities Via Camera Traps

RICH, L; KELLY, M. J.; SOLLMANN, R.; NOSS, A. J.; DI BITETTI, M. S.; MAFFEI, L.; ARISPE, R. L.; PAVIOLO, A.; DE ANGELO, C. D.; DI BLANCO, Y. E.

Milwaukee

Conferencia; The Wildife Society 20th Annual Conference; 2013

The Wildlife Society

Camera trap surveys, in combination with traditional or spatially explicit capture-recapture (C-R) techniques, have been used to estimate the abundance and density of individually identifiable carnivores.  Mark-resight models provide a viable alternative when only a portion of the population is uniquely identifiable because they include data from marked (e.g., by scars) and unmarked (i.e., identifiable to species level) individuals.  Our goal was to estimate the density of pumas (Puma concolor) at study sites in Belize, Argentina, and Bolivia using camera trap surveys in combination with traditional C-R, mark-resight, and newly developed spatial mark-resight models.  Seventeen to 42 camera stations were deployed at the study sites and were active for 56 to 96 days.  We labeled photographic captures of pumas with either the individual?s ID or as unmarked.  We estimated the abundance of pumas using program CAPTURE and using the mark-resight model in program MARK; we estimated density by traditional methods of adding a buffer area around the trap array.  We then directly estimated density using spatial mark-resight models which adopt a Bayesian framework.  There were 48, 65, and 35 puma capture events in Belize, Argentina, and Bolivia, respectively.  Using traditional C-R techniques we estimated densities of 2.4, 0.5, and 12.4 pumas per 100km2 in Belize, Argentina, and Bolivia, respectively. Using mark-resight techniques, our estimates were lower at 0.62, 0.32, and 5.61 pumas per 100km2, respectively.  Density estimates had increased precision when using mark-resight vs. traditional C-R techniques.  We anticipate estimates from spatial mark-resight models will be lower than estimates from C-R and mark-resight techniques because ad hoc estimation of survey area often inflates density estimates.  Our research supports that camera trapping, in combination with mark-resight techniques, offers a feasible method for obtaining robust demographic estimates for carnivores when only a subset of the population is uniquely identifiable.