Research Activities: Sampling Methods Evaluation

Background

       The goal of ARMI is to monitor amphibian populations using a probabilistic approach that will allow use of proportion area occupied statistical methods which incorporate detection probabilities. Thus each species is analyzed separately across repeated visits to multiple localities within each season, and across seasons.

       Because of the vast scope of amphibian diversity and life-history in the Southeastern US, SEARMI is challenged to develop a monitoring protocol that will include as many species as possible.  Previous ARMI inventory and monitoring included terrestrial sampling (drift fences, coverboards, PVC pipes, time constrained litter searches) and wetland sampling (dipnets, frogloggers, crayfish traps, minnow traps).  Because most amphibian species in the Coastal Plain region spend at least some portion of their life cycle in wetland habitats, focusing sampling on these habitats is likely to yield the most information about amphibian populations.

       In order to design our sampling protocol, we must understand the capabilities of our various sampling methods, including which species are likely to be captured, and whether or not we can collect accurate count data. The objective of this study was to compare the primary seven aquatic sampling techniques used by SEARMI in terms of the number and identity of species detected and abundance of individuals detected.

Methods

       We sampled 10 wetlands at the Katharine Ordway Preserve-Swisher Memorial Sanctuary from 2-11 May 2005 using seven aquatic sampling methods. The methods included 10 samples at each site with four methods which allow counts of individuals, including box trap (0.5 m² aluminum quadrat), metal dipnet, D-frame wooden dipnet, and crayfish trap. Three additional methods provided presence/absence data: visual encounter (animals observed during daytime sampling), froglogger (automated audio recording device), and aural (calls heard during daytime sampling).

Results

       The results of this study are currently being analyzed and prepared for publication in a peer-reviewed scientific journal.  Preliminary analysis reveals that more amphibian species are detected with frogloggers than the other six methods.

# Species  Time/trap Count of amphibians (± SE) Method Anurans Caudates (min) Box trap 7 2 14 50 ± 15 Crayfish trap 5 2 2.1 - D-frame dipnet 5 1 3.5 10 ± 5 Metal dipnet 6 1 3.5 15 ± 6 Froglogger 10 - - - Visual encounter 5 0 - - Aural 8 - - -

       The results of this study will help refine SEARMI's aquatic sampling protocol. Southeast ARMI is in the process of transitioning to a revised wetland sampling protocol.  Previous wetland sampling typically consisted of 1-day, 2-day, or 3-day or visits to each site; 1-3 visits/site/season; sampling methods: froglogger, crayfish traps (4-5), time constrained dipnetting, visual encounter, aural, and covariate (water chemistry, fish, invertebrates) sampling at each visit.  However, not all methods were used each day during a visit, covariates were not measured each day, and count data for each sampling method was not always consistently recorded.  This sampling protocol does not allow inclusion of sufficient sites (at least 25) at each refuge to use PAO models. Our proposed revised sampling protocol is shorter visits to sites (2-day visits) in which repeat observers are counted as repeat visits, with the goal of achieving at least two replicate visits each year.  Methods used will be frogloggers, crayfish traps (4-5), sweep constrained dipnetting, visual encounter, aural, and covariate sampling at each visit.  Count data will be recorded for crayfish traps and dipnetting.

next / back / table of contents