A common problem in the conservation of endangered species is that critical management decisions must be made without complete demographic information. When demographic data do exist they are often piecemeal, with different field methods used in different years, data that have been collected intermittently, small sample sizes, and information on only a subset of important parameters. New quantitative approaches, such as bayesian state-space models, meet many of these limitations by allowing multiple data types to be “integrated” into a single demographic analysis. The benefits of such models are that they generate more precise parameter estimates, standardize the error structure across data types, mechanistically link vital rate information to population abundance data, and estimate “hidden” parameters that are not necessarily measured in the field but can be extracted from the data. We applied this bayesian state-space approach to data on endangered Sierra Nevada bighorn sheep, the rarest subspecies of mountain sheep in
Results/Conclusions
We demonstrate the utility of this approach for maximizing information obtained from piecemeal data indicative of many endangered populations. Our bayesian state-space models, integrating minimum count, mark-resight, and telemetry data, increased precision in demographic parameter estimates over those obtained from any single data type alone. In addition, the models derived parameter values for years with missing data, standardized the error structure across data types, accurately estimated trends in key parameters through time, and fit covariates explaining population change. We found that the importance of weather, predation, and density on bighorn sheep demography varied among our subpopulations, a result with important implications for the conservation and management of this subspecies.
