robust.prioritizr: Robust Systematic Conservation Prioritization

Frankie H T Cho , Jeffrey Hanson

1. Climate change poses significant threats to biodiversity. To ensure the long-term persistence of species, protected areas must be established in locations that will safeguard suitable habitats in the future. Although statistical models can predict where such habitats may occur under different future scenarios, designing protected areas that can effectively protect these habitats across a wide range of futures remains challenging.
2. We present the robust.prioritizr R package as a decision support tool for systematic conservation planning. This tool is designed to identify priority areas for establishing protected areas that are robust against uncertainty. Using advances in robust optimization, this tool identifies priority areas that can achieve conservation objectives cost-effectively under a wide range of plausible future scenarios. Our novel approach a) allows users to flexibly specify their desired level of robustness to explore trade-offs between risk reduction and solution cost; b) does not make assumptions about the statistical distribution of uncertainty; and c) uses exact solvers for mixed-integer linear programming problems to guarantee solution optimality.
3. We examine a case study based in Victoria, Australia, to showcase the tool. This case study involved 872 native species, 12,988 candidate areas for selection, and four climate change scenarios across five time steps. Using this tool, we identified priority areas for cost-effectively meeting representation target thresholds for each species under various future scenarios in the region. Additionally, compared with prioritization generated using conventional approaches, prioritizations generated with the tool were better able to achieve conservation objectives across multiple future scenarios.
4. Synthesis and applications. Our study allows conservation scientists and practitioners to create conservation plans that are robust to uncertainties. The tool was developed as an open-source \texttt{R} package to enhance the \texttt{prioritizr R} package and is available on the Comprehensive R Archive Network (CRAN). By explicitly considering multiple future scenarios during priority setting, conservation plans can be made more resilient to the impacts of climate change on biodiversity.