Document: Linking beta diversity, environmental variation, and biodiversity assessment

View document

Faith DP and Ferrier S (2002; Linking beta diversity, environmental variation, and biodiversity assessment, Science online) provided an analysis in which only 15.6% of variation in Panamanian species dissimilarity (among sites) was not accounted for by measured environmental/spatial variables. (compared to Duivenvoorden et al.'s high value of 59% not accounted for). Faith and Ferrier provided the detailed analysis as a supplement (ref 17 to an Australian Museum web page). This document replaces that broken link.

File type:
File size:
133 kB


Effective biodiversity assessment makes best-possible use of available data to provide surrogate information for "all" of biodiversity and to allow comparisons among "all" areas. Environmental data is typically available for all areas but may be a weak biodiversity surrogate. Species records may be a good surrogate for other species, but typically are not available for all areas. These two forms of information can be combined by 1) summarising biotic information as compositional dissimilarities among areas, 2) linking observed dissimilarities to observed environmental differences, and then 3) using calculated environmental differences for new areas to assess biotic turnover. Approaches to this problem include ordination and dissimilarity/distance regression. Examples of both, using biodiversity data from Panama, highlight the advantages of using realistic models linking dissimilarity to environmental variation. While measured environmental variation explains Panama dissimilarities well, the ordination approach suggests that there are "missing" environmental variables that would increase explanation of biotic turnover among areas.

see pdf for complete description.

citation: Faith DP and Ferrier S (2002) Linking beta diversity, environmental variation, and biodiversity assessment, Science online, Published 22 July 2002. Available at:

Last Updated: