Along the Santa Clara River in California, populations of the federally and state listed least Bell’s vireo (Vireo bellii pusillus) are recovering from near extirpation. This recovery owes largely to habitat protection and restoration, as well as to controlling cowbird nest parasitism. We developed a decision support tool to aid in ongoing planning and coordination between conservation partners in the region. A major challenge we encountered in the development of this tool is a common problem in many conservation planning projects, namely uncertainty. There are many sources of uncertainty in this system including those arising from under-measured, but known, parameters such as regional species vital rates. The river system is also dynamic and periodically impacted by hard-to-predict stochastic events such as drought, flooding, and invasion by non-native species. We used a global sensitivity analysis approach that generates a set of models to sample across the full range of parameter and model structure uncertainty. We then replicated the set of models under a range of simulated management scenarios representing a variety of cowbird and habitat management strategies and intensities. This approach allows for a full accounting of parameter uncertainty while illustrating the effects of different management scenarios.
Results/Conclusions
The simulated vireo populations spanned a range of model outcomes which can be generalized in two modes. In the first mode, total population size largely tracked with the amount of suitable habitat available in the system. The other mode resulted in population sizes remaining much lower than the carrying capacity. These two modes represent the effects of different management scenarios. When cowbird control is inadequate, reduced fecundity becomes the primary limiting factor. Whereas, when cowbird impacts are sufficiently controlled, habitat protection and restoration become the primary drivers. We developed an interactive visualization tool that allows the user to explore model results by setting management goals and seeing which management scenarios have a higher probability of meeting those goals under uncertainty. The interface also allows the user to adjust uncertainty bounds on individual parameters and explore the impact on predicted model outcomes. Our hope is that the tool will aid in adaptive management of least Bell’s vireo in the region based on ongoing monitoring of fecundity and habitat conditions. This approach can be adapted to other projects to systematically identify and account for uncertainty, provide an intuitive visualization tool to investigate different management scenarios, and allow projects to move forward despite uncertainty.