Monitoring multiple species: Estimating state variables and exploring the efficacy of a monitoring program

Abstract

Monitoring programs have the potential to identify population declines and differentiate among the possible cause(s) of these declines. Recent criticisms regarding the design of monitoring programs have highlighted a failure to clearly state objectives and to address detectability and spatial sampling issues. Here, we incorporate these criticisms to design an efficient monitoring program whose goals are to determine environmental factors which influence the current distribution and measure change in distributions over time for a suite of amphibians. In designing the study we (1) specified a priori factors that may relate to occupancy, extinction, and colonization probabilities and (2) used the data collected (incorporating detectability) to address our scientific questions and adjust our sampling protocols. Our results highlight the role of wetland hydroperiod and other local covariates in the probability of amphibian occupancy. There was a change in overall occupancy probabilities for most species over the first three years of monitoring. Most colonization and extinction estimates were constant over time (years) and space (among wetlands), with one notable exception: local extinction probabilities for Rana clamitans were lower for wetlands with longer hydroperiods. We used information from the target system to generate scenarios of population change and gauge the ability of the current sampling to meet monitoring goals. Our results highlight the limitations of the current sampling design, emphasizing the need for long-term efforts, with periodic re-evaluation of the program in a framework that can inform management decisions.