Estima de una población de corzo mediante modelos de capturarecaptura clásicos y espacialmente explícitos

Abstract

Modern wildlife conservation and management requires accurate methods for population census and monitoring. In this work, we estimated the roe deer population (Capreolus capreolus) of Quintos de Mora (Toledo, Spain) in two close areas: one open and other fenced. For roe deer population estimates we used capture-recapture models and photo-identification of males by camera traps on feeding points by summer 2012. We used both a classic non-spatially explicit and a spatially explicit (SECR) capture-recapture models. Maximum-likelihood estimations MLE and Bayesian approaches were used in both cases. In the open area, deer estimates varied between 0.50 ± 0.14 (SECR approach) and 1.04 ± 0.14 (classic Bayesian approach) individuals per square kilometer, whereas in the fenced area between 7.58 ± 0.34 (classical Bayesian approach) and 11.25 ± 3.05 (SECR MLE approach) individuals per square kilometer. Finally we compared our estimations with direct counts. Although our estimates resulted consistent we recommend the use of SECR approaches for avoiding sampling area misspecifications.