A sequential approach to minimise threats within selected areas for conservation

Core-periphery models allow predictions of persistence to be made with relatively little data. The rationale is that populations in the core of their geographical or ecological ranges occupy suitable habitats and exhibit higher and less variable densities. Populations along the peripheries tend to be more fragmented and therefore less likely to receive immigrants from other populations. A population's probability of persistence is expected to correlate positively with habitat suitability and immigration rate and to correlate negatively with demographic variability. These predictions may be invalidated by the effect of threats, which may cause some peripheral populations to persist rather than populations in the core. We expect that predictions of persistence from core-periphery models will be improved by incorporating information on threats, and illustrate one way in which threat could be integrated within quantitative area-selection methods. We illustrate this for Europe by showing that important areas for biodiversity, selected with presence data, have consistently more people than expected by chance, but that incorporating human density as a constraint to area-selection can reduce substantially this level of pressure. We also show that areas selected using simple core-periphery models have fewer people than areas selected with presence data only. These results support the idea that there are opportunities to identify important areas for the persistence of species that are located in areas with low human density.