Assessment of site-occupancy modeling as a technique to monitor Hochstetter ’ s frog populations

Abstract

Hochstetter’s frog is a small, cryptic, semi-aquatic species. Isolated populations are found in forested catchments in the northern half of the North Island of New Zealand. The current status of these populations is unknown due to the lack of a rigorous monitoring technique. This study assessed the potential of site-occupancy modeling as a tool for monitoring Hochstetter’s frog populations. The primary aim of the study was to obtain estimates of detection probability for Hochstetter’s frog, to determine whether the amount of sampling effort required for a fully implemented monitoring programme is likely to be prohibitive. Multiple sites were established in each of three different stream habitats, and the presence / absence of frogs was recorded for each site during repeat surveys in winter / spring 2003 and summer / autumn 2004. The estimates of detection probability obtained indicate that the models have the potential to generate unbiased estimates of occupancy based on low sampling effort (four to six repeat surveys per site), and that in many instances it may be possible to separately monitor juvenile and sub-adult / adult frogs within a population. No major difficulties were encountered when implementing the technique in the field. It is recommended that anyone considering monitoring Hochstetter’s frog populations trial this technique, to determine its applicability in specific locations. Results indicate that site- occupancy modeling also may be applicable to monitoring other species in situations where more traditional rigorous methods (e.g. capture-recapture or distance sampling) cannot be applied.