Biodiversity viability analysis (BVA) and the ED surrogates strategy: combining alpha and beta diversity models to fill our biodiversity knowledge gaps

‘Biodiversity’ refers to the variety of all life, yet biodiversity assessments typically focus only on a small subset of all species. Biodiversity “surrogates” strategies can help us to fill biodiversity knowledge gaps. The challenge is to use the information we do have to make inferences about the vast amount of biodiversity that is still unknown to science. The “ED” surrogates approach generates hypothetical species and their distributions among sites or communities in a region. This means that we can perform many useful assessments as if we had information for all species in the region.

The “ED” surrogates methods (1-4) integrate available environmental and biotic information, and effectively combine models of α diversity (richness) and ß diversity (compositional dissimilarity among communities). The ED approach uses these models to generate hypothetical species in environmental space, and their distributions among sites in the region. The turnover of the hypothetical species in the environmental space reflects both richness variation and the magnitude of the dissimilarities or distances between sites in the space (1-3).

Various calculations based on the ED method are based on the idea of counting-up numbers of hypothetical species. An important calculation is the complementarity value of a site – how many hypothetical species it adds to given set of (e.g. already-protected) sites. Because ED estimates "complementarity values at the level of species counts" (3), over all species in the region, it has been used both for conservation planning and for survey design. One well-known ED-based method, “survey gap analysis” (SGA, 5), selects the best sites for future ecological surveys to discover new species. These are the places where ED indicates lots of un-sampled hypothetical species.

Faith et al (3) extended this approach to map the distributions in geographic space of the hypothetical species. This “biodiversity viability analysis” (BVA) maps the hypothetical species in geographic space and then uses this spatial information for each species for various biodiversity assessments. Faith et al (3) introduced the link from ED’s hypothetical species to BVA as follows:

“We will adopt an ordination approach as our basic pattern-model to make assertions about persistence of species. If an ordination, combining available biotic and environmental data, provides surrogate information for all species, then these species will have unimodal responses to this environmental space and contribute to an overall picture of species “turnover.” Actual species help infer the ordination pattern, but the ordination pattern, as a biodiversity surrogate, allows us to make inferences about all other species. We will use the term “inferred species” to refer to any hypothetical species inferred to have a general unimodal response in some part of ordination space .... Our strategy is to then translate information about any inferred species from ordination space to its implied distribution in geographic space (taking advantage of the link that environmental data for all areas provides from ordination space to geographic space). It is possible as a consequence to make assertions about the viability of these species within given geographic areas. For example, we may regard any species that has only a small fragmentary range in an area as having only a small probability of persistence there, even under full protective land-uses. These contributions to the complementarity values for areas may reflect a broad range of viability considerations, including gains in connectivity, patchiness of fragments, proximity to threatening processes, and other factors.”

Faith et al (3) also presented an example study, based on an environmental space derived using museum collections data for skinks in NSW. They mapped both real and hypothetical species in geographic space, by taking each geographic area in turn, finding its corresponding position in environmental space, and then recording the area as a geographic “presence” for the species in question if the area’s position in environmental space fell within the species’ range boundaries. As an assessment example, they calculated the current total area coverage for each hypothetical species in the NSW national parks, illustrating how one might assess representativeness and persistence of overall biodiversity in protected areas. Faith et al (3) concluded that:
“These preliminary results for NSW, using museum collection records for a single taxonomic group, suggest that combining such data with existing environmental data provides complementarity values that not only incorporate standard representation gains but also gains in persistence of biodiversity. These complementarity values, when re-calculated on different taxonomic groups yield approximately the same ordination space and sets of inferred species for the region.”

This BVA / ED approach clearly allows a wide range of biodiversity assessments. Faith et al (6) recently outlined the applicability of the approach to climate change impacts assessments, and extensions to counting general hypothetical “features” rather than just species.

A recent study (Mokany, K., Harwood, T. D., Overton, J. M., Barker, G. M. and Ferrier, S. 2011. Ecology Letters, 14: 1043–1051) provides a method (“FOAM”) that mimics the ED/BVA generation of “hypothetical species” that are allocated to sites so as to match predicted richness patterns and dissimilarities in composition. While no reference or comparison was made to the existing ED/BVA methods, it is easy to show that their approach approximates the typical BVA/ED generation of hypothetical species based on alpha and beta diversity models. On the other hand, their method loses some useful information that BVA/ED derives from the environmental space. I explore this issue elsewhere.


1. Faith, D.P. & Walker, P.A. (1996) Environmental diversity: on the best-possible use of surrogate data for assessing the relative biodiversity of sets of areas. Biodiversity and Conservation, 5, 399–415.

2. Faith, D.P., Ferrier, S., Walker, P.A. (2004) The ED strategy: how species-level surrogates indicate general biodiversity patterns through an ‘environmental diversity’ perspective. J. Biogeogr. 31, 1207–1217.

3. Faith D.P., G. Carter, G. Cassis, S. Ferrier, L. Wilkie (2003) Complementarity, biodiversity viability analysis, and policy-based algorithms for conservation. Environmental Science & Policy 6: 311–328.

4. Faith, D.P. (2011) Attempted tests of the surrogacy value of the ED environmental diversity measures highlight the need for corroboration assessment of surrogacy hypotheses. Ecological Indicators 11, 745-748.

5. Funk, V.A., Richardson, K.S., Ferrier, S., (2005) Survey-gap analysis in expeditionary research: where do we go from here? Biol. J. Linn. Soc. 85, 549–567.

6. Faith D.P., Lozupone C.A., Nipperess D., Knight R. (2009) The Cladistic Basis for the Phylogenetic Diversity (PD) Measure Links Evolutionary Features to Environmental Gradients and Supports Broad Applications of Microbial Ecology’s “Phylogenetic Beta Diversity” Framework. International Journal of Molecular Sciences 10:4723-4741


Dr Dan Faith , Senior Principal Research Scientist email:danfaith8[at]
Last Updated: