An evolutionary slant on species-area curves.

Uncertainty about the past has made it difficult to evaluate the relative roles of ecological and historical processes in generating species diversity. Although diversity is predicted to be higher on larger islands compared with that on smaller islands because of the effect of area on immigration and extinction rates, ecological theories do not fully account for positive species–area relationships if the source of new species is intra-island speciation, rather than immigration. In such cases, it becomes an evolutionary question of why the diversities differ between islands. Speciation rates are expected to be higher in larger areas, but until now, there had been no quantitative test of this theory. By reconstructing the number of immigration and speciation events on a phylogeny of Caribbean island Anolis lizards and comparing these quantities with island area, Losos and Schluter1 test the evolutionary contribution to species–area relationships. Their study showed that, above a threshold island size, speciation surpasses immigration as a source of new species. On large islands, species are derived in situ from ancestral lineages, but on small islands, the species are always distantly related, and thus must have immigrated. As predicted, speciation rates also scaled positively with island area. Using computer simulations, the authors show that neither the number of ancestral lineages, nor differences in the time islands harbored species accounted for the increased number of speciation events on larger islands. Although a declining extinction rate with increasing island area could produce a positive species–area relationship, Losos and Schluter demonstrate that such an explanation is improbable. Estimates derived from fitting the phylogeny to a birth–death model An evolutionary slant on species–area curves There is now a plethora of studies showing that owners of a resource are advantaged in pairwise contests for that resource. There are also many explanations: the use of ownership as an arbitrary cue for contest settlement, the accumulation of competitively superior individuals as owners, the presence of mechanistic advantages for owners, and owners perceiving a higher value for the contested resource. Several studies have attempted to unravel these influences using manipulative experimentation. In a recent study on Tasmanian snow skinks (Niveoscincus microlepidotus)1, Olsson and Shine take a complementary approach. By observing naturally occurring contests between skinks of known identity, they contrasted the importance of ownership and physical attributes to contest outcome. Male snow skinks establish loosely defined and partially overlapping ‘home ranges’. When two neighbouring skinks meet, the ensuing contest is usually resolved by display, but vigorous – and often injurious – fighting is also common. Contests are terminated when the loser flees the disputed area. In all 32 pairwise incidence of escalated snow skink fights fails to support this explanation, as conventional settlement is predicted to evolve to minimize injuries (some contestants lost entire limbs). There was no size difference between intruders and residents, so the ownership advantage cannot be explained by an accumulation of large males as residents (body size determines competitive ability in many lizard species). However, as Olsson and Shine only explored size-related physical characteristics, other less obvious decisive asymmetries might have been present. Snow skink home ranges usually contain sun-basking sites and (basking) residents could have higher body temperatures, and perhaps fighting ability, than do intruders (as in the speckled wood butterfly, Pararge aegeria). Asymmetries in resource value could also apply; home ranges might be more valuable to residents because a displaced resident will become a nonowner, whereas a vanquished intruder can retreat to its own home range. Although Olsson and Shine’s study presents valuable field data, the explanation of owner advantage remains presently elusive, and demands closer examination. On a practical note, we recommend that future analyses of multivariate contest data employ a single analysis (e.g. logistic analysis of covariance) rather than a series of separate tests. We also propose that the treatment of ‘ownership’ as a purely dichotomous variable can gloss over important ecological and contextual details. In this study, all intruding skinks were owners elsewhere, whereas in other taxa intruders are often floaters. Intruders in some contest systems are able to partially exploit a territory, and possibly acquire a degree of residency status, before contests occur. The resource ownership phenomenon is clearly complex, and a more refined consideration might be required to fully elucidate its general advantage.