Great Lakes Network amphibian monitoring: Initial synthesis report, 2013–2016

Casper, Gary; Nadeau, Stefanie

• The probability models for all amphibian species from 2013 through 2016 indicate that species’ occupancy of monitored wetlands was stable or increasing, with relatively low turnover among sites. We caution however that these are currently relatively short-term data, with only two years of data available for some parks. • The program meets its primary objective by achieving near perfect detection for most species in the Great Lakes Inventory and Monitoring Network. The high detection probabilities strongly suggest that there are very few false absences in the data. False presence rates by the automated recognition process varied across species but were overall very low, and were zero for site occupancy (due to manual proofing). Similarly, false positives for site occupancy are eliminated in the double-observer manual surveys. • Success of automated recognition varied by species, but was successful for most species and allows for the efficient analysis of very large data sets (>150 samples per site per year). • A protocol for manual surveys of spectrograms was developed that allows for efficient and accurate sampling of a 5-minute recording in approximately 2 minutes. This protocol provides robust data on 30 samples per site per year, including scoring of call index values (CIV) as a proxy for tracking population size. • There is potential to increase the program sensitivity by utilizing CIV ranks in proportional trend analyses. This will allow for discovery of site-specific ongoing declines (or increases), whereas occupancy trend analyses only detect declines when species have already disappeared from multiple sites. • An amphibian occupancy record for each monitored site is now established (Appendix C), and will continue to grow in value with annual additions. • The State Endangered Blanchard’s Cricket Frog was discovered at MISS and represents three new county records and an opportunity to better manage for this species. Interesting absences of Spring Peeper at MISS and Northern Leopard Frog at ISRO and PIRO are reported. Other species that appear rare or absent at specific parks are discussed in Appendix C. • We strongly recommend that ARS be deployed in autumn on delayed start programs for the following spring, in most parks where early spring visits are logistically difficult or impossible due to weather. This will eliminate a persistent problem of deployments too late to track early calling species. This should greatly improve several aspects of data analysis and better meet program goals. We also recommend that early morning sampling be implemented at four parks, (APIS, GRPO, ISRO, PIRO), to improve Mink Frog detection. This is simply a programming upgrade, with some extra effort for data analyses, but a pilot study suggests this will double Mink Frog detection rates.

Great Lakes Network amphibian monitoring: Initial synthesis report, 2013–2016

Series

Collections