The crudest measures of connectivity are based on the distance from a given habitat patch to other occupied habitat patches, which might provide immigrants. This is an index of isolation, and is turned into an index of connectivity by taking its inverse. Indices of this type use only the distance to the nearest habitat patch, ignoring all other patches that might provide immigrants, and variation in the size of source populations in different patches. Meta-analyses and simulations both confirm that nearest-neighbor metrics are crude estimates of connectivity and that they have lower predictive ability than the more sophisticated measures discussed below. Measures could be converted from structural measures to potential connectivity by converting links from distances to binary presences or absences, or probabilities using known movement distances and a dispersal kernel (describing the probability of dispersal as a function of distance), respectively. Nearest-neighbor distances are the most widely used of the different kinds of connectivity metrics, despite their low predictive ability.
Was this article helpful?