Premium
An empirical evaluation of camera trapping and spatially explicit capture‐recapture models for estimating chimpanzee density
Author(s) -
DesprésEinspenner MarieLyne,
Howe Eric J.,
Drapeau Pierre,
Kühl Hjalmar S.
Publication year - 2017
Publication title -
american journal of primatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.988
H-Index - 81
eISSN - 1098-2345
pISSN - 0275-2565
DOI - 10.1002/ajp.22647
Subject(s) - mark and recapture , camera trap , density estimation , sampling (signal processing) , abundance (ecology) , statistics , estimator , computer science , population , ecology , mathematics , biology , habitat , computer vision , demography , filter (signal processing) , sociology
Empirical validations of survey methods for estimating animal densities are rare, despite the fact that only an application to a population of known density can demonstrate their reliability under field conditions and constraints. Here, we present a field validation of camera trapping in combination with spatially explicit capture‐recapture (SECR) methods for enumerating chimpanzee populations. We used 83 camera traps to sample a habituated community of western chimpanzees ( Pan troglodytes verus) of known community and territory size in Taï National Park, Ivory Coast, and estimated community size and density using spatially explicit capture‐recapture models. We aimed to: (1) validate camera trapping as a means to collect capture‐recapture data for chimpanzees; (2) validate SECR methods to estimate chimpanzee density from camera trap data; (3) compare the efficacy of targeting locations frequently visited by chimpanzees versus deploying cameras according to a systematic design; (4) evaluate the performance of SECR estimators with reduced sampling effort; and (5) identify sources of heterogeneity in detection probabilities. Ten months of camera trapping provided abundant capture‐recapture data. All weaned individuals were detected, most of them multiple times, at both an array of targeted locations, and a systematic grid of cameras positioned randomly within the study area, though detection probabilities were higher at targeted locations. SECR abundance estimates were accurate and precise, and analyses of subsets of the data indicated that the majority of individuals in a community could be detected with as few as five traps deployed within their territory. Our results highlight the potential of camera trapping for cost‐effective monitoring of chimpanzee populations.