Abstract
The Theory of Island Biogeography (TIB) promoted the idea that species richness within sites should depend on site connectivity, i.e. its connection with surrounding potential sources of immigrants. TIB has been extended to a wide array of fragmented ecosystems, beyond archipelagoes, surfing on the analogy between habitat patches and islands and the patch-matrix framework. However, patch connectivity often little contributes to explaining species richness in empirical studies. Before interpreting this trend as questioning the broad applicability of TIB principles, one first needs a clear identification of methods and contexts where strong effects of patch structural connectivity are likely to occur. Here, we use spatially explicit simulations of neutral metacommunities to show that patch connectivity effect on local species richness is maximized under a set of specific conditions: (i) patch delineation should be fine enough to prevent dispersal limitation within patches, (ii) patch connectivity indices should be scaled according to target organisms’ dispersal abilities and (iii) habitat amount and fragmentation should both lie in some intermediary range that still needs an empirically tractable definition. When those three criteria are met, the absence of effect of connectivity on species richness should be interpreted as contradicting TIB principles.
Footnotes
Thanks to the time and effort that PCI Ecology recommanders and reviewers spent reading and commenting our manuscript. We have fully revised it in the light of their remarks. (1) We refocused the manuscript on our main question (when should connectivity drive species richness) and streamlined the whole text accordingly. (2) We dropped unclear references to the habitat amount hypothesis, which would deserve another study. (3) We used quadratic relationships to study connectivity effects. (4) We dropped the combination of indices of various types, since a single index already led to very high R2 with the quadratic regression, making combination of indices prone to spurious effects. (5) We strengthened the use of bibliographical references.