Conventional methods for enhancing connectivity in conservation planning do not always maintain gene flow

Protected area systems need to conserve species in places with suitable habitat that are configured to facilitate gene flow. Since genetic data require considerable resources to obtain, many proxy methods have been developed to generate plans for protected area systems (prioritizations) that facilitate gene flow without needing genetic data. However, the effectiveness of such methods—such as minimising fragmentation or enforcing contiguity among priority areas—remains largely untested. We investigated the ability of prioritizations to maintain gene flow when they are generated using conventional methods for promoting connectivity. Using existing environmental, genetic, and occurrence datasets, we created maps of habitat suitability and resistance to gene flow for nine alpine plant species. Next, we generated multispecies prioritizations that secured 10% of the suitable habitat for each species and attempted to maintain gene flow by (a) penalizing fragmentation, (b) representing species in contiguous areas of suitable habitat, and (c) representing species in contiguous areas with minimal resistance to gene flow as modelled from genetic data. We found that prioritizations generated using fragmentation penalties failed to represent seven of the nine species in areas that would maintain high levels of gene flow. Similarly, prioritizations that represented species in contiguous areas of suitable habitat were unable to maintain high levels of gene flow for six species—potentially because a few areas with high resistance can disrupt gene flow throughout an entire prioritization. Although prioritizations generated using genetic data successfully maintained gene flow, they also selected over three times more land than other prioritizations, suggesting that failing to account for gene flow when setting priorities may underestimate the scale of conservation action required. Synthesis and applications. We found that conventional methods for enhancing connectivity in conservation planning, such as spatially clustering priority areas or providing connected sections of suitable habitat, were generally unable to maintain high levels of gene flow. Our results suggest that conservation plans could be substantially improved by directly using genetic data, although whether this is a good choice for a particular situation will also depend on the costs of obtaining these data.