Survival analysis: Informing recovery of Sierra Nevada bighorn sheep

Abstract

Listen

Translate article

ABSTRACTSurvival in small populations (e.g., Sierra Nevada bighorn sheep or Sierra bighorn [Ovis canadensis sierrae]) is often highly variable. External selective pressures vary in the degree to which they regulate survival by sex and age class. Understanding the important factors and risks for different demographic classes helps managers design strategies that enhance the recovery of endangered species, including Sierra bighorn. Our goal was to determine what population‐level factors (e.g., climate, habitat, population size, predation) affect survival and whether there are interactions between these factors by age and sex, and then apply our findings to recovery strategies. To this end, we conducted a known‐fate survival analysis for female and male Sierra bighorn with data collected over 12 years, and used model selection to evaluate models with spatial, environmental, and other population‐level factors hypothesized to be related to survival. Survival of adult Sierra bighorn declined continuously with age for both sexes; survival was generally higher for females than males, and there were no interactions between age and any environmental or population‐level factors. The top model for both sexes included the date of peak value of normalized difference vegetation index (NDVI) from the previous summer; NDVI had a similar positive relationship with survival for both sexes, which indicates that the later the growing season persists into the summer, the better survival the subsequent year. For females, survival also was negatively related to an index of abundance for mountain lions (Puma concolor), whereas the relationship was less apparent for males. Instead, top models for males indicated elevated survival during warm wet years, but years with late peaks in NDVI the previous year ameliorated the effect of a cold, dry winter. Finally, competitive models for males and females included a variable representing avalanche risk, indicating reduced survival in areas with increased avalanche risk. From a recovery management perspective, the lack of any interaction between age and other covariates suggests that although we may still select younger female Sierra bighorn for translocations (an essential recovery action) because they have higher reproductive value than old females, there were no additional negative synergies between age and other factors to consider. All variables are of value in guiding expectations for newly established populations and established source populations and some may help fine tune the selection of translocation areas. In addition, including predation, weather covariates, and catastrophic effects, such as avalanche risk, in projection models is important for realistic estimation of the time required to meet recovery goals and predicting population trajectories under likely climate change scenarios. Our approach is generalizable to other systems; we demonstrated how survival analyses can inform endangered species recovery management by indicating ideal areas for translocations and provided realistic estimates of time‐to‐recovery or other recovery metrics. © 2018 The Wildlife Society.