Toward a unified view of diversity partitioning.

The debate on the properties and use of additive and multiplicative partitioning of species diversity exemplified in this forum reflects the broader challenge of quantifying and interpreting alpha, beta, and gamma diversity at multiple scales of sampling. As noted by Wilsey and Ricotta in their contributions, ecologists use a wide range of measures of beta diversity, partly because of differences in study questions of the investigators and partly because of the differences in their statistical properties. The lack of agreement on the numerous dimensionless measures of community turnover and dissimilarity is a longstanding problem in community ecology (Vellend 2001, Koleff et al. 2003, Magurran 2004, Ricotta 2007). So, too, there has been some disagreement on the statistical properties and applicability of the alpha and beta components of diversity partitioning. The popularity of Whittaker’s (1960) multiplicative partition of diversity, and subsequent additive partitions by MacArthur et al. (1966), Allan (1975), and Lande (1996), is that they provide a single set of values of alpha and beta diversity for a given sampling scale and therefore give a simple, intuitive measure of species diversity and composition. For this reason, we believe that partitioning methods are a powerful tool for quantifying spatial and temporal variation in biodiversity in a manner that is accessible to ecologists, managers, and non-scientists. The cost of simplicity is that partitioning methods discard information on sitespecific composition retained in pairwise dissimilarity and ordination that may be important to the underlying biophysical or land-use gradients that produce beta diversity (but see Hofer et al. 2008). Thus, as an Manuscript received 26 June 2009; accepted 29 June 2009. Corresponding Editor: A. M. Ellison. For reprints of this Forum, see footnote 1, p. 1962. 3 E-mail: [email protected] FORUM 1988 Ecology, Vol. 91, No. 7