The island rule is an ecogeographical rule, which states that island populations and species get smaller or bigger than their mainland counterparts. These processes of dwarfism and gigantism are thought to depend on the resources available in the island. This rule was originally proposed by Van Valen in 1973, after the results of the works on island mammals developed by Foster in the previous decade (hence it is also called Foster’s rule). Since its proposal, many studies, reviews and theoretical models have tried to find some generalities from the observed patters of body size change (see e.g. Meiri et al Procs Roy Soc B 2008 or our recent works on Homo floresiensis or Jersey red deer), although the evidence for body size changes being the rule rather than the exception is limited (see Meiri et al J Biogeogr 2011).
In a short commentary, we criticize a recent approach to model the island rule developed by Biddick and Burns in a recent Ecology Letters paper. Briefly, their model makes three clearly unrealistic assumptions: it limits a priori the body size that the island populations can reach; assumes all species will show the same variance in size regardless of their biological and ecological differentes; and they assume that the time needed for evolutionary drift is constant for all populations. None of these assumptions is corroborated by the literature, which shows that there are no obvious bounds for body size changes, but importantly that all island species, and populations from the species in different islands, show different trends of body size variation. We also show how more realistic assumptions also recover the observed patterns of change in body size in a more meaningful way.
You can find this new paper at https://jhortal.com/project/2342/.
The picture shows a Cozumel raccoon (Procyon pygmaeus) photographed by Camazine on the northern end of Cozumel island, Yucatán, Mexico in 2010. Cozumel pygmy raccoons are a critically endangered species, c. 20% smaller and 50% ligther than their mainland counterparts.