Spatial characteristics of species distributions as drivers in conservation prioritization

Spatial prioritization, based on the biogeographical identification of priority areas for conservation actions, is an important aspect of conservation planning. Although the influence of factors such as costs, threats or use of surrogates on the resulting priorities has been studied extensively, relatively little is known about how the spatial characteristics of species distributions drive the spatial pattern of priorities in multi-species conservation plans. Using datasets from Australia and Finland, we explore how excluding or including a given species changes spatial priorities in a multi-species prioritization. We develop three metrics to quantify changes in priorities, and explore how these changes depend on the total number of species used in the prioritization, the spatial characteristics of the given species distribution, and how species share their space with other species used in the prioritization. We randomly selected 12 set of 10 species from each dataset, and explore the influence of each of these species in prioritizations done for a total of 10, 20, 50 or 100 species. We show that spatial priorities become increasingly stable as the number of species is increased, and that the stability of highest and lowest priority areas behave differently. When less than 50 species were used in a prioritization, intermediately rare species that occupy mostly species-poor habitats tend to have the greatest influence on priorities, whereas very rare and common species that co-occur with many other species tend to have a small influence. Our results present a systematic method to explore the stability of spatial priorities to changes in the species pool used for a conservation plan. Although the analysed two datasets differed in data type, location, scale and species composition, they both showed how using a small number of species leads to unstable spatial solutions, where the choice to include or exclude an individual species can strongly influence the conservation outcome. Our results also show that conservation planners should carefully assess the use of spatial prioritizations for identifying least important areas (e.g. for development) as these can be particularly unstable when the prioritization is based on a small number of species.