Lizards in an Evolutionary Tree: Ecology and Adaptive Radiation of Anoles

7. Evolution of Ecomorphological Diversity

113 7 EVOLUTION OF ECOMORPHOLOGICAL DIVERSITY Fortunately, the uncertainties about the biogeographic history of Anolis discussed in the previous chapter have little bearing on understanding of patterns of ecomorphological radiation and diversification, at least within the West Indies. Phylogenetic information indicates that for the most part anoles have radiated independently on each island of the Greater Antilles, regardless of how or when they got there. EVOLUTION OF THE ECOMORPHS Phylogenetic analysis is critical to investigating patterns of ecomorph evolution. The presence of members of the same ecomorph class on multiple islands could be explained in two ways. On one hand, each ecomorph class might have evolved a single time; in this scenario, the presence of each ecomorph class on multiple islands would be the result of overwater dispersal or vicariance. Given six ecomorph classes, this hypothesis would posit five evolutionary transitions from one ecomorph type to another (assuming one ecomorph class is ancestral). Alternatively, each ecomorph class could have arisen independently on each island upon which it occurs. This scenario would require at least 17 evolutionary transitions.147 The phylogeny (Fig. 7.1) clearly favors the latter hypothesis. In only one case are members of the same ecomorph class on different islands sister taxa: the twig anole 147. Recall that four ecomorph types occur on all four islands in the Greater Antilles, one occurs on three islands, and one occurs on two islands (Chapter 3). losos_ch07.qxd 4/11/09 8:16 AM Page 113 F I G U R E 7.1 Ecomorph evolution in the Greater Antilles. This is the same phylogeny as in Figures 5.6 and 6.1, but species other than those on Cuba (C), Hispaniola (H), Jamaica (J), and Puerto Rico (P) have been pruned out. To visualize the parsimony reconstruction of ecomorph evolution, assume that red (twig anole) is the ancestral state. Ignore unique anoles (brown) and count the number of evolutionary changes in color. Note that in Puerto Rico, two clades of trunk-ground and two clades of grassbush anoles form a clade, but there have only been three evolutionary transitions , rather than four, because the ancestor of the larger clade containing these four clades was either a grass-bush or a trunkground anole. On Jamaica, the ancestor of the trunk-crown/crown-giant clade was probably a trunkcrown anole, thus requiring only two evolutionary transitions. Overall, this approach yields an estimate of 19 evolutionary transistions. P P P J C C C C C C C H H H H H H H Crown-giant Grass-bush Trunk Trunk-crown Trunk-ground Twig Unique losos_ch07.qxd 4/14/09 12:35 PM Page 114 clade A. sheplani + A. placidus on Hispaniola is the sister taxon to the Cuban twig anoles; this clade likely originated in Cuba because it is nested within a larger clade of Cuban species.148 Consequently, in the vast majority of cases, the existence of the same ecomorph on different islands is the result of convergent evolution. The phenomenon of convergence of entire communities has often been suggested (e.g., Orians and Paine, 1983; Wiens, 1989), but this is probably the best documented example in terms both of phylogenetic evidence and quantitative measures of morphology and ecology (a topic to which I return in Chapter 17). Overall, a minimum of 19 evolutionary transitions in ecomorph are required by the phylogeny.149 The reason that this number is greater than the minimum required for independent evolution of each ecomorph type on each island (17) is that several ecomorphs have evolved multiple times on a single island: grass-bush anoles twice on Cuba and Hispaniola150 and either grass-bush or trunk-ground anoles twice on Puerto Rico.151 Two clades of twig anoles occur on both Hispaniola and Cuba,152 but in the most parsimonious reconstruction of ecomorphs, twig anoles are the ancestral state and thus are not convergent. The separation of the Cuban twig anoles into two clades is not strongly supported by the data and seems unlikely, given that the morphological differences among them are slight (e.g., Estrada and Hedges, 1995; Díaz et al., 1996). Although the DNA data clearly indicate the convergent nature of the ecomorphs across islands, this result was not surprising; Williams’ (1972) initial presentation of the ecomorphs, grounded in a phylogenetic understanding of Anolis based primarily on morphology, recognized their convergent nature.153 Although Williams was misled in a few...