Selecting areas for species persistence using occurrence data Original Research Article

Quantitative area selection methods seek to maximise the amount of biodiversity represented in networks of areas for conservation. However, because criteria for persistence are usually not incorporated, there is no guarantee against choosing areas where species have low probabilities of persistence. Here, we propose one framework for introducing criteria for persistence into quantitative area-selection methods when dealing with atlas data for large numbers of species. The framework includes three steps: (1) fit models explaining current occurrence of species; (2) transform current probabilities of occurrence into estimates of persistence using available information on expected threats and species’ vulnerability; and (3) select complementary areas to ensure high estimates of persistence for each species. This paper provides an example using coarse-scale data for European trees, without threat data. Three approaches for modelling species probabilities of occurrence are compared: first, by considering occurrence in relation to environmental variation; second, by considering occurrence in relation to patterns of geographical aggregation or contagion among records; and third, by combining these two components. The third model fits the original data most closely, but field assessments of persistence estimates are needed. As expected, introducing additional constraints into area selection reduces the flexibility (fewer alternative sets of areas) and increases the cost (more areas needed to achieve the goal). However, the proposed method increases the overall expected probability of persistence for the species. This benefit is greatest among the species with the most restricted ranges, which are the species of greatest conservation concern.