Evaluating the Effectiveness of Passive Acoustic Monitoring at Detecting Anurans in Caribbean Tropical Forests

Passive acoustic monitoring (PAM) has emerged as an innovative technique that can be used to monitor biodiversity. Yet, the global distribution of its application is uneven, and few studies have evaluated its effectiveness for amphibians. Given the pressing need to improve conservation actions for this taxon, recommending efficient ways to monitor species is urgently needed. We aimed to evaluate the effectiveness of PAM by comparing the number of species, community composition, and species accumulation detected to those detected with the traditional method of visual encounter surveys (VES) in Caribbean tropical forests. We set out to use our findings to develop much-needed recommendations to guide monitoring and inform management for amphibians in the region. From five fixed tropical forest sites sampled across Trinidad and Tobago, we found that PAM detected 16 species, more than twice as many species as VES at six species. Using the Chao2 species richness estimator, PAM (16.4, SE 0.95) was also more comprehensive than VES (9.2, SE 4.31). Using a pattern-matching algorithm, no new species were detected from PAM after the first week of sampling in July within our fixed sample sites, which suggests this could be a key survey period to focus monitoring efforts. Effort in terms of time collecting data was similar between the two methods, totaling 1,500 minutes evaluated across all five sites; PAM involved more analysis time in processing the sound files, yet required fewer field visits and resulted in greater community completeness. Where resources permit, we recommend that PAM be used to survey amphibians in Caribbean forests, and that sampling should be concentrated during the first two months of the rainy season to monitor communities optimally.