Search for a command to run...
Detecting individuals of amphibian and reptile species can be a daunting task. Detection can be hindered by various factors such as cryptic behavior, color patterns, or observer experience. These factors complicate the estimation of state variables of interest (e.g., abundance, occupancy, species richness) as well as the vital rates that induce changes in these state variables (e.g., survival probabilities for abundance; extinction probabilities for occupancy). Although ad hoc methods (e.g., counts uncorrected for detection, return rates) typically perform poorly in the face of no detection, they continue to be used extensively in various fields, including herpetology. However, formal approaches that estimate and account for the probability of detection, such as capture-mark-recapture (CMR) methods and distance sampling, are available. In this paper, we present classical approaches and recent advances in methods accounting for detectability that are particularly pertinent for herpetological data sets. Through examples, we illustrate the use of several methods, discuss their performance compared to that of ad hoc methods, and we suggest available software to perform these analyses. The methods we discuss control for imperfect detection and reduce bias in estimates of demographic parameters such as population size, survival, or, at other levels of biological organization, species occurrence. Among these methods, recently developed approaches that no longer require marked or resighted individuals should be particularly of interest to field herpetologists. We hope that our effort will encourage practitioners to implement some of the estimation methods presented herein instead of relying on ad hoc methods that make more limiting assumptions.