How reliable are amphibian population metrics? A response to Kroll et al.

Kroll et al. [Kroll, A.J., Runge, J.P., MacCracken, J.G., 2009. Unreliable amphibian population metrics may obfuscate more than they reveal. Biological Conservation] criticized our recent advocacy for combining readily attainable metrics of population status to gain insight about relationships between terrestrial plethodontid salamanders and forest succession [Welsh, Jr., H.H., Pope, K.L., Wheeler, C.A., 2008. Using multiple metrics to assess the effects of forest succession on population status: a comparative study of two terrestrial salamanders in the US Pacific Northwest. Biological Conservation 141, 1149–1160]. They argue that each of our metrics has the potential to be biased or flawed and, therefore, any inferences so derived are likely to be uncertain and misleading. One of our main goals was to highlight the value of combining multiple quantitative approaches with logic, knowledge of species’ biology and parsimony when interpreting findings, as a means to counter uncertainty that might occur with any single metric. We applied a sampling design that minimizes the chance of bias in detections and counts, included multiple statistical methods to test relationships, and incorporated past research and species’ biology when interpreting findings. We disagree with Kroll et al. [Kroll, A.J., Runge, J.P., MacCracken, J.G., 2009. Unreliable amphibian population metrics may obfuscate more than they reveal. Biological Conservation] that the potential biases associated with the individual metrics mean that the inferences from the study are too uncertain to be useful by managers. We examined metrics that are more informative than simple site occupancy, including relative abundance, life stage distributions, and body condition, to inform our knowledge of population status. Combining metrics allows us to use readily attainable data to yield deeper insights into population structure and related spatial aspects. We agree on the value of having the most detailed and accurate data possible, but do not believe this standard should limit one to conducting only intensive experimental studies to provide detailed information on vital rates of terrestrial salamanders. The controlled experiment approach is extremely difficult and would be considerably less feasible than using population metrics to assess the status of plethodontid salamanders on forest landscapes at scales relevant to land management in the Pacific Northwest.