Date of Degree


Document Type


Degree Name





Ana Carolina Carnaval

Committee Members

Michael J. Hickerson

Frank Burbrink

Jason Munshi-South

Christopher Raxworthy

Ana Carolina Carnaval

Subject Categories

Biodiversity | Ecology and Evolutionary Biology | Evolution


phylogeography, Anolis, Dactyloidae, Amazonia, Atlantic Forest, population genetics


Shifts in the geographic distribution of habitats over time can promote dispersal and vicariance, thereby influencing large-scale biogeographic patterns and ecological processes. The establishment of corridors of suitable habitat across previously disjunct yet ecologically similar regions has been widely linked to climate change over time. Such climate-mediated habitat changes may have played a key role in the assembly of tropical biotas, including those of Amazonia and the coastal Atlantic Forest in South America, two of the most diverse ecosystems on Earth. The history of biogeographic associations between tropical rainforest habitats in response to former climatic shifts, as well as their consequences for the distribution and genetic composition of forest taxa, is the theme of this doctoral dissertation. To study the historical biogeography of these forest blocks, the four chapters presented here collect and analyze genetic data from South American species of true anoles (Anolis, Dactyloidae) and, to a lesser extent, bush anoles (Polychrus, Polychrotidae).

To improve inferences about historical biogeography and the drivers of diversification in South America, my dissertation begins with a revisionary phylogenetic study of poorly-known rainforest anole species. Chapter 1 provides an updated molecular phylogeny ofa highly diverse clade within the genus Anolis known as Dactyloa. My phylogenetic analyses find A. phyllorhinus to be nested within the punctatus group, suggesting independent evolution of a rostral proboscis in Anolis. Moreover, A. philopunctatus is recovered as nested within A. punctatus, with limited genetic divergence between distinct dewlap phenotypes. The most recent common ancestor of the Dactyloa and Norops clades of Anolis dates back to the Eocene. Most Amazonian species within these two clades diverged in the Miocene, while Amazonian species of Polychrus diverged in the Pliocene. Importantly, the temporal framework provided is this study yielded much-needed estimates of divergence times and mutation rates for my subsequent phylogeographic and population genetic analyses of historical biogeographic connections between Neotropical rainforests based on anoles.

Focusing on the southern montane region of the Atlantic Rainforest, where former biogeographic connections with other South American regions have been invoked to explain the affinities of a number of endemic taxa, Chapter 2 investigates the phylogenetic relationships of Anolis nasofrontalis and A. pseudotigrinus, known from few specimens collected more than 40 years ago. A comprehensive phylogenetic analysis recovers six main clades within the Dactyloa clade of Anolis, five of which were previously referred to as species series (aequatorialis, heterodermus, latifrons, punctatus, roquet). A sixth clade clustered A. nasofrontalis and A. pseudotigrinus with A. dissimilis from western Amazonia, A. calimae from the Andes, A. neblininus from the Guiana Shield, and two undescribed Andean taxa. I therefore define a sixth species series within Dactyloa: the neblininus series. Close phylogenetic relationships between highly disjunct, narrowly-distributed anoles suggest that patches of suitable habitat connected the southern Atlantic Forest to western South America during the Miocene, in agreement with the age of former connections between the central Andes and the Brazilian Shield as a result of Andean orogeny. The data also support the view of recurrent evolution (or loss) of a twig anole-like phenotype in mainland anoles, in apparent association with the occurrence in montane settings.

In Chapter 3, I employ sequence data at multiple loci from three co-distributed arboreal lizards (Anolis punctatus, A. ortonii, and Polychrus marmoratus) to infer historical relationships among disjunct Amazonian and Atlantic Forest populations and to test alternative historical demographic scenarios of colonization and vicariance using coalescent simulations and Approximate Bayesian Computation (ABC). Data from the better-sampled Anolis species support colonization of the Atlantic Forestfrom eastern Amazonia, while hierarchical ABC suggests that the three species colonized the Atlantic Forest synchronously during the mid-Pleistocene. I find support of population bottlenecks associated to founder events in the two Anolis, but not in P. marmoratus, consistently with their distinct ecological tolerances. These findings support that climatic fluctuations provided key opportunities for dispersal and forest colonization in eastern South America through the cessation of environmental barriers. Evidence of species-specific histories strengthens assertions that biological attributes play a role in responses to shared environmental change.

In Chapter 4, I assess whether range expansions across ecological gradients in South America are associated with the selection of genotypes that confer physiological adaptation. For that, I carry a comparative population genomic study based on restriction enzyme associated DNA markers from the anoles Anolis ortonii and A. punctatus, which colonized the Atlantic Rainforest from Amazonia. Clustering analyses indicate strong genetic differentiation between Atlantic Forest and Amazonian samples in both anole species, while historical demographic inference recovers large effective population sizes and Pleistocene divergences between populations in the two forest blocks. Low levels of gene flow indicate very limited post-divergence exchange of migrants between Amazonian and Atlantic forests, in spite of the presumed opportunities of connectivity provided by cyclic periods of increased humidity and forest expansions. Genome-environment association analyses recover eight protein-coding loci in A. ortonii and 14 in A. punctatus whose frequency is significantly correlated to environmental gradients, consistent with a scenario of local adaptation. These loci are involved with metabolic and regulation processes, performing molecule binding, transcription, hydrolase, and transporter activities. My results suggest a role of functional genomic differentiation and local adaptation associated with range expansions across heterogeneous landscapes in widely distributed species. The establishment of allele frequency differences in response to climatic regimes may have been favored by large effective population sizes and low migration rates across regions, consistently with proposed interactions between natural selection and the demographic trajectory of populations during the process of adaptation.