The Salmonid Population Viability Project: Modeling Trout Viability in a Desert Landscape Under Current and Forecasted Conditions

In an ideal world, species conservation planning would be guided by formal population viability analyses (PVA) to determine the probability that each population will persist, with and without management actions.  Primarily due to a lack of accessible data-driven methods for PVA that can be used across broad spatial scales, most conservation planning instead relies on indirect surrogates of viability (e.g., habitat size). We developed a new statistical Spatio-Temporal Population Viability Model (ST-PVM) that combines fish sampling data with remotely-sensed data to deliver simultaneous estimates of carrying capacity, inter-annual variability, and viability for many populations across large areas.  Remotely-sensed spatial covariates describe habitat size and quality, while temporal variability is a function of temperature and flow.  The approach can leverage information from well-sampled areas to extrapolate to poorly sampled or even un-sampled populations, under current and future climates.  We conducted a pilot study of Lahontan cutthroat trout, a federally threatened trout subspecies native to the Great Basin Desert, to generate simultaneous estimates of extinction probability and carrying capacity for 38 populations.  We discuss on-going model refinement/expansion, as well as estimation under climate change and exploration of the efficacy of different management scenarios (e.g., barrier or non-native trout removal).