Estimating puma densities from camera trapping across three study sites: Bolivia, Argentina, Belize

KELLY, MARCELLA; NOSS, ANDREW; DI BITETTI, MARIO; MAFFEI, LEONARDO; ARISPE, ROSARIO; PAVIOLO, AGUSTÍN; DE ANGELO, CARLOS; DI BLANCO, YAMIL

JOURNAL OF MAMMALOGY

ALLIANCE COMMUNICATIONS GROUP DIVISION ALLEN PRESS

Año: 2008 vol. 89 p. 408 - 418

0022-2372

Abundance estimates are extremely valuable for species conservation yet determining abundance for elusive, wide-ranging, carnivores such as the puma is difficult. We estimated puma density using remote cameras across study sites in Bolivia Argentina, and Belize. We used obvious and subtle markings to identify individual pumas in photographs and conducted double-blind identifications to examine the degree of agreement among investigators. Average identification agreement between pair-wise investigators was high at nearly 80.0% and 3-way agreement was lower at 72.9%. Disagreement between investigators was uncommon and was much lower than the opposite of agreement averaging 7.8% (or 0.69% for 3-way disagreement). Puma abundance varied more from site to site than between investigators. Bolivian pumas moved the shortest distances between camera stations and Argentinean pumas the longest, but distances among cameras and survey area covered varied among sites. Puma densities varied significantly and consistently from site to site regardless of investigator. Averaged across investigator, Bolivia had significantly more pumas per 100 km2 at 16.4 (± 3.6) than Belize at 3.42 (± 1.3) or Argentina at 0.67 (± 0.2). The high puma numbers in Bolivia may be influenced by close camera spacing and smaller survey area. Argentina’s puma numbers match those of low-density North American sites, and those for Belize are consistent with the Pantanal and high-density North American sites. Puma densities can be reliably estimated with remote cameras, and our work presents the first density estimates for Central America and for forested environments in South America.