LB 11-148 - Climate change-informed conservation opportunity areas for sensitive species: modeling rare plants and animals into the future to detect refugia

Background/Question/Methods:

Integrating climate change into place-based conservation strategies is a pressing challenge for promoting future conservation success. In particular, the broad effects of climate change can make it difficult to prioritize specific actions in specific places. To address this, we have implemented a novel, place-based approach that incorporates current species conservation planning activities with climate change-informed species distribution modeling (SDM) and habitat vulnerability assessments with the goal of identifying a portfolio of conservation areas that, when combined with climate-adaptive conservation actions, will engender sensitive species persistence in the face of climate change. Working in New Mexico and using its State Wildlife Action Plan and Plant Conservation Strategy species lists, we identified 100 species of conservation concern from several taxonomic groups and that had sufficient observations for available modeling. As a first step, we computed ensemble SDMs using RandomForest, GLM, XGB, and MARS machine learning R algorithms on a subset of 10 plants and 10 animals that reflected a range of guilds and potential habitats. The modeling was based on 900 m grid cells and used a large suite of available biophysical predictors plus future distributions driven by WorldClim 2.1 CMIP6 GCM ensemble models projected out to 2060.

Results/Conclusions:

Outcomes were informative about the complexity of capturing areas of species overlap that would optimize conservation planning. That is, our test species occupied a wide range of habitats and had distinctly different projections into the future. But we were able to identify our first set of future potential conservation opportunity areas that lie outside the current reserve system for the state. Scale remains an issue: working at a 900 m grid cell size can be adequate for mobile animal species but becomes problematic for more stationary plant species, which also being relatively rare, may have significant habitat restrictions not captured at this coarser scale. Accordingly, our next steps will be to incorporate even finer downscaled climate data, add more species to modeling framework, and introduce models of habitat change and vulnerability based on vegetation to further corroborate operational distributions for species of conservation concern and identify priority areas where conservation activities that are most likely to preserve biodiversity.