Connecting the Dots: Assessing Landscape Connectivity Algorithms for Biodiversity Conservation

Context To understand and characterize the dispersal of organisms in a fragmented landscape, scientists often use connectivity matrices - tables that contain the probabilities of successful dispersal between different pairs of patches. Objective While mechanistic, individual-based correlated random walk (CRW) models are commonly used to estimate these probabilities, simpler, deterministic alternatives exist based on distance and patch size which are easier to develop, use and understand. However, the relative performance of these simpler algorithms compared to the CRW model is not well understood. We ask, how good are simplified algorithms in mimicking the CRW model? Method To address this gap, we compared the connectivity matrices of ten simple algorithms to those of a CRW model across 36 landscape-disperser combinations. Results Our results show that the frequently used exponential decay algorithm (EXP) did not perform well, with a mean R ² of 0.745 and a minimum R ² of 0.185 between the connectivities of the EXP model and the CRW model. On the other hand, the CRD-lim model - which uses a constant•radius/distance relation within a maximum inter-patch distance (d _max ) - performed best, with a mean R ² of 0.918 and a minimum R ² of 0.809. Conclusion Overall, our results show that the CRD-lim algorithm is a good alternative to random walk models when assessing connectivity matrices for a specific landscapes and species in case a full individual-based CRW is not feasible, for example because data are scarce or a multi-species perspective is taken.