Local Adaptation in Mainland Anole Lizards: Integrating Population History and Genome-Environment Associations

Figure 1. Anolis punctatus, South America’s coolest lowland anole – literally. Picture by Renato Recoder.

In ectothermic organisms, environmental factors such as temperature and water availability constrain physiological and behavioral performance. Therefore, the occurrence of species in varied environments may be associated with local adaptation. On the other hand, experimental studies have shown that physiological function can be highly conserved within species over broad environmental gradients, which may be associated with the homogenizing effects of population gene flow. In a recently published study, we focus on widespread South American anoles to investigate whether the occurrence of species in distinct environments is linked to local adaptation and whether population structure and history have constrained adaptive differentiation.

Based on molecular data, my collaborators and I have previously found that arboreal lizard species have independently colonized the Atlantic Forest from Amazonia, subsequently expanding southward towards subtropical regions. This is the case of Anolis ortonii and Anolis punctatus (Fig. 1), whose ranges now encompass a climatic gradient from warm and wet conditions in Amazonia to cooler and less rainy settings in the Atlantic Forest. Our new study investigates whether species establishment in distinct climates is associated with potentially adaptive genetic differentiation between populations. To this purpose, we implement genome-environment association analyses on the basis of thousands of restriction site-associated DNA markers. Moreover, to estimate levels of gene flow – a force that could oppose adaptive differentiation – we perform historical demographic inference under a genetic coalescent framework. Lastly, to characterize the climatic gradients presently occupied by A. ortonii and A. punctatus, we estimate climatic space occupancy over their ranges.

Analyses of genetic structure inferred distinct populations in Amazonia and the Atlantic Forest in both anole species (Fig. 2), suggesting that separation of these forests following a period of contact in the past has favored genetic divergence. In the two species, historical demographic analyses inferred large effective population sizes, mid-Pleistocene colonizations of the Atlantic Forest from Amazonia, and post-divergence population gene flow (Fig. 3). These results support the hypothesis of recurrent rainforest expansions that connected presently disjunct biomes in northern South America.

Figure 2. Genetic clustering based on all SNPs from Anolis ortonii (A) as well as all SNPs (B) and candidate SNPs only (C) from A. punctatus. Proportions in pie charts on maps correspond to ancestry coefficients estimated by genetic clustering analyses. Grey areas on map indicate South American rainforests. Red arrows indicate A. punctatus sample MTR 20798 from Pacaraima on the Brazil-Venezuela border in the Guiana Shield region, a locality that is climatically similar to Atlantic Forest sites (see Fig. 4); this sample is genetically more similar to eastern Amazonian samples based in the entire SNP dataset, yet more similar to Atlantic Forest samples based on the candidate SNPs only.

Figure 3. Population history (from SNAPP) and historical demographic parameters (from G-PhoCS) inferred for Anolis ortonii (A) and A. punctatus (B). Parameters are the time of coalescence between populations (in millions of years, Mya), effective population sizes (in millions of individuals, M), and migration rates (in migrants per generation, m/g). Colors of terminals correspond to genetic clusters in Fig. 2.

Genome-environment association analyses found allele frequencies of 86 SNPs in 39 loci to be significantly associated with climatic gradients in A. punctatus. Among the candidate loci, eleven uniquely mapped to known protein-coding genes in the reference genome of Anolis carolinensis; two mapped non-specifically to more than four genes; and the remaining mapped against non-coding regions, which may correspond to regions that regulate gene expression or that are physically linked to genes that underwent selection. In the case of A. ortonii, no SNPs were associated with temperature and precipitation variation across space. Constraints related to population structure and history do not seem sufficient to explain discrepant signatures of adaptation between the two anole species; instead, this discrepancy may be related to species differences in climatic space occupancy over their ranges (Fig. 4).

Figure 4. Environmental space occupancy along latitude based on climatic PC1 for Anolis ortonii and A. punctatus. Samples used in genetic analyses are indicated with a black dot. Higher PC scores correspond to drier and colder sites. Dashed line indicates the approximate region of a pronounced north-south climatic turnover in the Atlantic Forest identified by previous studies. Red arrow indicates A. punctatus sample MTR 20798 from Pacaraima, a mid-elevation site (820 m above sea level) in the Guiana Shield region that overlaps climatically with Atlantic Forest sites (horizontal axis). Note that the two species experience largely similar climates in Amazonia and the northern Atlantic Forest, yet A. punctatus occupies cooler and less humid localities that are not occupied by A. ortonii in the southern Atlantic Forest.

The candidate genes identified in A. punctatus play essential roles in energy metabolism, immunity, development, and cell signaling, providing insights about the physiological processes that may have experienced selection in response to climatic regimes. Similar to our study, other investigations of anole lizards found differences in the frequency of alleles that underlie ecologically relevant physiological processes between populations that inhabit contrasting habitats. These examples support the hypothesis that adaptation to colder climates has played an essential role in range expansions across anole taxa, including mainland and Caribbean forms that span altitudinal and latitudinal gradients.

Anolis ortonii. Spotting this cryptically colored rainforest anole is quite challenging indeed. Picture by Miguel T. Rodrigues.

This investigation illustrates how studies of adaptation on the basis of genome-environment association analyses can benefit from knowledge about the history of landscape occupation by the species under investigation. Data on population structure and history can provide insight into how gene flow and natural selection interact and shape population genetic differentiation. Moreover, information about the direction and routes of colonization of new habitats can support spatial sampling design, help to characterize landscape gradients, and support the formulation of hypotheses about how organisms have responded to environmental variation in space.

To know more:

Prates I., Penna A., Rodrigues M. T., Carnaval A. C. (2018). Local adaptation in mainland anole lizards: Integrating population history and genome-environment associations. Ecology and Evolution, early view online.

Author
Recent Posts

Ivan Prates

Peter Buck Postdoctoral Fellow at Smithsonian's National Museum of Natural History. I study mainland anole evolution, biogeography, and comparative genomics.