Species-area and species-sampling effort relationships: disentangling the effects

The effects of area and SE were both significant, but the former one was noticeably weaker. The SSERs were roughly linear in log-log space, whereas the SARs demonstrated scale-dependent behavior with a noticeable threshold (slope breakpoint). Species richness was almost area-independent below this threshold (the small area effect, SAE) but followed typical power-law SAR beyond the threshold. This effect was similar to the small island effect but occurred for arbitrarily delineated areas within continuous habitats. Parameters of the SAR curves depended on organism size. The upper limit of the SAE increased from microorganisms to meiofauna to macrofauna. Also, SAR curves for unicellular groups had significantly lower slopes. SAE is supposed to indicate a spatial range of statistical homogeneity in species composition. Its upper limit corresponds to the characteristic size of a local community (a single habitat occupied by a common species pool). Interpretations of SAR and SSER parameters in terms of alpha- and beta-diversity are proposed. Both SAR and SSER slopes obtained from univariate regressions are overestimated. This upward bias depends on sampling design, decreasing for SAR but increasing for SSER with more unequally spaced samples. Both spatial extent and sampling effort should be taken into account to disentangle properly their effects on diversity.