Conflicting nest survival thresholds across a wetland network alter management benchmarks for an endangered bird

Abstract

Small changes in the environment can have large impacts, leading to ‘tipping points’ or ‘threshold’ effects on populations. Identifying such effects can provide practitioners benchmarks to both minimize negative impacts and foster positive benefits for populations. Nonetheless, identifying thresholds can be challenging as they can vary spatially, temporally, and in their form (i.e., the shape of the threshold). Such variation, if gone undetected, may stymie the identification of reliable conservation benchmarks. We developed Bayesian change-point models that estimate thresholds of environmental effects that can vary in the number, magnitude and form in space and/or time. We applied these models to estimate thresholds in the effects of hydrology on nest survival of an endangered, wetland-dependent bird, the snail kite (Rostrhamus sociabilis plumbeus). Using data from 2790 nests at nine wetlands spanning the entire U.S. breeding range over a 21-year time period, we contrasted different threshold models for hydrology effects. Overall, both low water and high water, and high rates of change (declining or increasing water levels, recession and ascension, respectively), had negative, threshold effects on nest survival. Change-point models emphasized that the form, extent, and magnitude of thresholds vary across the breeding range. These benchmarks for water management illustrate that nest survival can be potentially increased over three-fold when thresholds are avoided. Furthermore, our results caution that applying incorrect threshold models to environmental problems can lead to inappropriate targets for conservation. Our modeling framework provides a generalizable way to address this issue and reliably estimate the effect of complex change points and their uncertainty on populations.