The species-area relationship, self-similarity, and the true meaning of the z-value

The power model, S= cA(z) (where S is number of species, A is area, and c and z are fitted constants), is the model most commonly fitted to species-area data assessing species diversity. We use the self-similarity properties of this model to reveal patterns implicated by the z parameter. We present the basic arithmetic leading both to the fraction of new species added when two areas are combined and to species overlap between two areas of the same size, given a continuous sampling scheme. The fraction of new species resulting from expansion of an area can be expressed as alpha(z)-1, where alpha is the expansion factor. Consequently, z-values can be converted to a scale-invariant species overlap between two equally sized areas, since the proportion of species in common between the two areas is 2-2(z). Calculating overlap when adding areas of the same size reveals the intrinsic effect of distance assumed by the bisectional scheme. We use overlap area relationships from empirical data sets to illustrate how answers to the single large or several small reserves (SLOSS) question vary between data sets and with scale. We conclude that species overlap and the effect of distance between sample areas or isolates should be addressed when discussing species area relationships, and lack of fit to the power model can be caused by its assumption of a scale-invariant overlap relationship.