The influence of model frameworks in spatial planning of regional climate-adaptive connectivity for conservation planning

Landscape connectivity improves species’ capacity to adapt to climate change. These models are increasingly needed and available for climate-change conservation planning. However, their relative strengths and weaknesses are unclear. We asked how well do the spatial outputs from four connectivity models intended to support climate change conservation agree? To understand the implications of selecting one or several approaches, we compared various combinations of four connectivity models for ecoregions in California, U.S.A. Two models are based on landscape structure, Land Facet Corridors and Omniscape, while two other models, Meta-Corridor Approach and Network Flow Analysis (NFA), use focal species’ range dynamics to determine connectivity. We also describe how each approach integrates climate-adaptive connectivity concepts. Variation in modeling methods, objectives, inputs, and landscape representations strongly affects the modeled connectivity patterns. For the region where all four models were run, almost three quarters of the landscape was selected by one or more models, but three or more agree for only 9.5% of the area, all of which is riparian. This emphasizes the importance of riparian areas for climate adaptation. We found NFA prioritized connections close to protected areas, while Meta-Corridor avoided higher cost agricultural and developed areas. The structural models agreed in areas with low human impact but Omniscape avoided areas of low topographic diversity and Land Facet Corridors avoided connections in areas with no protected areas. Connectivity models should be selected based on the conservation objectives, such as spatial scale to be implemented, and a combination of models may be best.