Stream amphibians as metrics of critical biological thresholds in the Pacific Northwest, U.S.A.: a response to Kroll et al.

Abstract

Summary1. Kroll, Hayes & MacCracken (in press) Concerns regarding the use of amphibians as metrics of critical biological thresholds: a comment on Welsh and Hodgson 2008. Freshwater Biology, criticised our paper [Welsh & Hodgson (2008) Amphibians as metrics of critical biological thresholds in forested headwater streams of the Pacific Northwest. Freshwater Biology, 53, 1470–1488] proposing the use of headwater stream amphibians as metrics of stream status in the Pacific Northwest (PNW). They argued that our analysis of previously published data reflected circular reasoning because we reached the same conclusions as the earlier studies. In fact, we conducted a meta‐analysis to address new questions about the optimum values and thresholds (based on animal densities) for abiotic stream attributes that were found to be important to these amphibians in earlier studies. This is analogous to determining blood pressure thresholds or fat‐to‐weight ratios that facilitate predicting human health based on meta‐analyses of earlier data from studies that found significant correlations between these variables and relative health.2. Kroll et al. argued that we should not make inference to environmental conditions across the PNW from data collected in California. We collected data from northern California and southern Oregon, the southern extent of the PNW. We made inference to the Klamath‐Siskiyou and North Coast bioregions, and argued that available research on these headwater species indicates that our results have the potential to be applied throughout the PNW with minimal regional adjustments.3. Kroll et al. contended that we need reproductive success, survival estimates and density estimates, corrected for detection probabilities, to establish relationships between animal density and stream attributes. Reproductive success and survival estimates are important for demographic modelling and life tables, but they are not necessary to demonstrate meaningful relationships with abiotic conditions. Both corrected occupancy estimates and individual detection probabilities are unnecessary, and take multiple sampling efforts per site, or onerous mark release and re‐capture studies, respectively, to determine accurately.4. Kroll et al. questioned the use of stream amphibians as a surrogate for measuring physical parameters, such as water temperature, claiming that measuring the physical parameters directly is more efficient. Here they misinterpreted the main point of our paper: stream organisms are integrators of what happens in a catchment, and carefully selected species can serve as surrogates for the biotic community and the relative condition of the network environment.5. Kroll et al. claimed that we demonstrated weak inferences regarding ecosystem processes. We argue that by relating densities of stream amphibians with changes along abiotic environmental gradients that are commonly affected by anthropogenic activities, we are establishing biological links to gradients that represent important ecosystem processes and identifying biometrics that can be used to quantify the status (health) of these gradients.