Joel McGlothlin received the 2011 Dobzhansky Prize, given to an outstanding young scientist in the field of evolutionary biology. Successfully fending off technical AV difficulties, Joel gave a fascinating talk in which he examined the idea that evolution should occur along lines of least genetic resistance, which are determined by the genetic correlations among traits. This hypothesis predicts that as traits diverge, they should diverge in ways that mirror the trait genetic correlations. Moreover, one would expect that this effect would attenuate over time, so that more distantly related species would show less evidence of diverging along these lines.

But this hypothesis is based on the assumption that genetic correlations remain constant through time, which may not be true. Indeed, such correlations may be evolutionarily malleable and may be shaped by natural selection, thus serving as evidence of selection, rather than necessarily functioning as a constraint on evolution.

To distinguish among these possibilities, Joel examined patterns of genetic correlations (called the G-matrix) across the Greater Antillean anole radiations. If G evolves slowly and under the influence of genetic drift, then a relationship should exist between degree of phylogenetic similarity and G—more closely related species should be more similar. Alternatively, if natural selection drives genetic correlations, then species that are ecomorphologically similar—members of the same ecomorph category—might be expected to be most similar in G matrix, regardless of how closely related they are.

To test this hypothesis, McGlothlin estimated G matrices for seven anole species. This is no mean feat: the G-matrix for each species was determined by breeding 100+ individuals and raising 2845 juveniles reared to the age of six months, during which time they were x-rayed four times, for nearly 10,000 x-rays, from which a number of traits were measured—a huge amount of work! Indeed, this study far surpasses in numbers of species any previous comparative study of G-matrix evolution.

The results have a little bit for everyone. All species are moderately to highly correlated in their genetic matrices. Overall, a moderate effect of phylogeny is apparent: more closely related species tend to be more similar, giving evidence for evidence for evolution along lines of least genetic resistance. But there is one outlier: the two distantly-related trunk-ground anoles, A. sagrei and A. cristatellus, are extremely similar in G-matrix, evidence for a role of convergent natural selection. Also, the most distinctive species in the
study, the only grass-bush anole, A. pulchellus, is very different in genetic structure from close relatives.

Examining morphology on a trait-by-trait basis indicates that the largest variation in
genetic correlation among species involves the correlations between limb vs. head elements (as opposed to correlations among head elements, or among limb elements).  McGlothlin suggested that these inter-element differences probably arise because the ecomorphs are distinguished by differences among shape modules; for example, trunk-ground anoles have much longer hindlimbs relative to other body elements. Selection to turn a trunk-ground anole into a trunk-crown, or vice-versa, needs to remodel the relationship among body parts.

In sum, these results reveal support for both constraining influences of genetic correlations, in which are embedded a strong role of natural selection in breaking these constraints, at least sometimes. McGlothlin concluded: adaptive radiation doesn’t seem to be constrained by G; natural selection shapes its own constraints.

Final quote of the meeting (this being the last talk in the meeting): “The best adaptive radiation is the Anolis radiation.”

Jonathan Losos
Latest posts by Jonathan Losos (see all)