EcoRest: an R package for rapid assessment habitat modeling to inform ecosystem management and restoration

Background/Question/Methods: Ecosystem management and restoration projects often require decision-makers to evaluate and choose among potential management alternatives. Ecological modeling informs the decision-making process by simplifying complex ecological questions into discrete cause-and-effect relationships, forecasting outcomes of potential management actions, and testing actions against alternative future scenarios such as land use or climate change. Myriad analytical tools exist for outcomes such as spatial prioritization (Marxan), population modeling (popbio package in R), and countless forms of statistical model development and application. However, tools and models are often also developed ad hoc to support project-specific management objectives. To streamline ecological modeling, a flexible, reproducible model was developed for habitat suitability analyses, which can be applied across ecosystem types and taxa. The model was programmed in the R Statistical Software Language and is available as a package for future users. The EcoRest package contains functions allowing for model visualization, habitat unit computation, and sensitivity analyses of model parameters. The package includes over 300 taxa-specific habitat models (previously published by the US Fish and Wildlife Service) as well as a generic data structure easily generated for ad hoc modeling problems. The package also contains cost-effectiveness and incremental cost analysis (CEICA) functions, commonly applied in restoration decision-making.

Results/Conclusions: We demonstrate the utility of the EcoRest package for informing a freshwater restoration project, with emphasis on alternatives comparison and sensitivity analysis. The Huron Island habitat rehabilitation and enhancement project is developing large river restoration actions in Pool 18 of the Upper Mississippi River. This case study demonstrates how the EcoRest package could be applied to forecast the effects of management alternatives on four taxa (bluegill, walleye, channel catfish, and smallmouth bass) and compare benefits with restoration costs using CEICA. Furthermore, the EcoRest package was also applied to examine model sensitivity of these four habitat models, which informed project planning decisions. Ultimately, this presentation describes the development, utility, and application of an integrated habitat-decision model, which is available for broader use in the R modeling environment as the EcoRest package.