A note on the partitioning of allelic diversity

Gene diversity and allelic diversity are both recognized as useful criteria for establishing priorities of conservation. Similarly to gene diversity partitioning, based on variation of gene frequencies both within and between subpopulations, allelic diversity can be partitioned using variation in allele numbers. Allelic diversity components have to meet a number of desirable properties, among which orthogonality (i.e. independence of the within-subpopulation and between-subpopulation components), and concavity (i.e. no within-subpopulation diversity larger than total diversity). In this note it is shown how the partitioning recommended in species diversity studies should be transposed to allelic diversity. Attention is drawn to some methods of allelic diversity partitioning currently proposed, essentially the widely used rarefaction method of Petit et al. (Conserv Biol 12:844–855, 1998) and a more recently proposed method by Caballero and Rodriguez-Ramilo (Conserv Genet 11:2219–2229, 2010). Their contrasting properties and the potentially contrasting conclusions to be expected from their application in conservation are emphasized. Some simple examples are used to show how an improper measure of allelic diversity may lead to misleading conclusions when defining priorities of conservation.