Effectiveness of Environmental Surrogates for the Selection of Conservation Area Networks

Rapid biodiversity assessment and conservation planning require the use of easily quantified and estimated surrogates for biodiversity. Using data sets from Québec and Queensland, we applied four methods to assess the extent to which environmental surrogates can represent biodiversity components: (1) surrogacy graphs; (2) marginal representation plots; (3) Hamming distance function; and (4) Syrjala statistical test for spatial congruence. For Québec we used 719 faunal and floral species as biodiversity components, and for Queensland we used 2348 plant species. We used four climatic parameter types (annual mean temperature, minimum temperature during the coldest quarter, maximum temperature during the hottest quarter, and annual precipitation), along with slope, elevation, aspect, and soil types, as environmental surrogates. To study the effect of scale, we analyzed the data at seven spatial scales ranging from 0.01◦ to 0.10◦ longitude and latitude. At targeted representations of 10% for environmental surrogates and biodiversity components, all four methods indicated that using a full set of environmental surrogates systematically provided better results than selecting areas at random, usually ensuring that ≥90% of the biodiversity components achieved the 10% targets at scales coarser than 0.02◦. The performance of surrogates improved with coarser spatial resolutions. Thus, environmental surrogate sets are useful tools for biodiversity conservation planning. A recommended protocol for the use of such surrogates consists of randomly selecting a set of areas for which distributional data are available, identifying an optimal surrogate set based on these areas, and subsequently prioritizing places for conservation based on the optimal surrogate set.