Should landscape variation in population status be assessed with individual‐ or population‐level indicators?

Abstract

Monitoring trends in wildlife populations is essential to effective management. Evaluating only a single metric can be cost‐effective; however, how a given metric responds to environmental variation may differ from other population‐ or individual‐level metrics. Consequently, conclusions about the effect of environmental variation on a species should consider the relationship between different metrics. We investigated how population‐level metrics (occupancy and relative abundance) and an individual‐level metric (body condition) responded to landscape variation in 3 semi‐aquatic salamander species. In the summer of 2015 and 2016, we conducted repeated count surveys at 71 streams on the Cumberland Plateau, Tennessee, USA, actively counting and measuring salamanders. We estimated occupancy and abundance of larval and post‐metamorphic animals using community‐occupancy and community‐N mixture models, respectively. We estimated body condition of the most observed life stage for each species using a standardized mass index. All models included the same set of covariates for comparison. Occupancy and abundance responses for each life stage‐by‐species combination were positively correlated for 4 of 6 combinations, with salamander abundance responding to a larger suite of covariates. Body condition was not linked to covariates in the same fashion as population‐level metrics, although consistency across species was detected for 1 covariate, watershed size. Our results indicate that inter‐specific and inter‐life stage variation in evaluated metrics is present within salamanders of this study system. Management for populations of salamanders in the region should consider monitoring multiple response metrics, to capture both population‐ and individual‐level responses, as these may not always be correlated.