Anoles are models for studying evolution in the wild. Not only do anoles have a history of repeatedly diversifying to specialize in the same types of microhabitats in the same ways across the Greater Antilles, these lizards also have a tendency to adapt on rapid timescales to environmental change — be it the addition or subtraction of a predator or competitor, a polar vortex, a change to the structural environment, or a hurricane.

Anoles are also models for urban evolution. Why? Anoles are found abundantly across the Caribbean in urban and forest environments where they specialize in divergent microenvironments characterized by shifts in climate and physical structure. Urban habitats tend to be warmer, drier, more open, and dominated by buildings and impervious surfaces instead of vegetation — providing the perfect opportunity for repeated adaptation to a novel combination of environmental conditions. In other words, Caribbean cities provide a replicated natural laboratory to study adaptation as it happens when these lizards colonize and thrive in urbanizing areas. And there is no shortage of urban-tolerant and urbanophilic anole species to choose from!

Species of Anolis lizards are found in urban environments across the Caribbean (photos CC-BY K. Winchell; Earth at night by NASA).

I have focused a lot on one species of anole, the Puerto Rican Crested Anole (Anolis cristatellus). Previously, I found that these lizards exhibit rapid parallel morphological changes in limb length and toepad structure, and that these changes appear to help urban lizards navigate the structural environment, dominated by buildings spread far apart and smooth surfaces like metal, more effectively. And although we found these differences were maintained in common garden rearing (supporting a genetic basis), we didn’t know what genes underlie these phenotypic changes or if the same genes were responsible for the parallel morphological changes across populations. Moreover, we previously dug into another key phenotype — thermal tolerance — finding parallel increases in thermal maximum in urban lizards and evidence of selection at the genetic level, but we knew little of the myriad other ways these lizards might be responding to urbanization at both the phenotypic and genomic levels.

In our study published in the Proceedings of the National Academy of Sciences (PNAS) this week, along with co-authors Shane Campbell-Staton, Jonathan Losos, Liam Revell, Brian Verrelli, and Anthony Geneva, we set out to understand the genetic basis of these phenotypic shifts and to explore how else lizards might be adapting to urban environments.

Integrating environmental, phenotypic, and genomic data

Our study started with the idea that it is essential to connect environmental, phenotypic, and genomic changes to understand how evolutionary processes shape adaptations in novel environments. Without all three of these, we have an incomplete picture. For example, if we don’t know how the environment differs, we can only guess what phenotypic changes mean; similarly if we don’t know how phenotypes differ, we can only guess what genomic changes mean. Thus we first showed that the urban environments we studied exhibit parallel changes before digging into the genomics. Focusing on three pairs of urban-forest populations across the island of Puerto Rico, we showed that urban habitats diverge in parallel from their forest counterparts in multiple environmental dimensions.

Urban environments diverge in parallel from forest environments across the three regions sampled. Figure 1 (A&B) from Winchell et al. (2023, PNAS); satellite imagery Google Earth and Maxar Technologies 2001.

This parallel environmental divergence sets the stage for parallel adaptive phenotypic and genomic divergence as a consequence of parallel selection pressures. To test this hypothesis, we sampled 96 lizards from these paired urban and forest sites, took digital xrays and high-resolution images, sampled a small piece of the tail (which grows back in this group of lizards), and released the lizards back into their habitats to live our their lives. Back in the lab, we then measured six key morphological traits (below: limb length, toepad area, and number of specialized scales known as lamellae) and extracted DNA from the tissue samples. We sequenced a subset of the DNA that codes for genes (the exome).

Morphology of Anolis lizards: limb length, toepad area, and lamellae number on toepads. From Winchell et al. (2023, PNAS) Figure 3A.

An assumption of our genomic analyses was that the urban populations we sampled were in fact independent genetic replicates. This might not be the case, for example, if there were widespread gene flow across the island (e.g., all populations were highly connected) or if there were substantial translocations between cities. If this were the case, then differences detected at the genetic level could be the result of repeated natural selection in each urban population, but it would also not be possible to rule out a single selection event in one population followed by the spread of beneficial alleles across highly connected populations. Although a subtle difference, these two scenarios tell us something different about how evolution is proceeding. To address this, we looked at the underlying population structure and relatedness among individuals to confirm that we did have three replicated population pairs in which individuals within a region were more closely related to each other than to individuals in other regions. In other words, the urban populations arose repeatedly and independently across the island.

Urban-associated genomic divergence

We incorporated this underlying population structure information into our analyses. We tested first for urban-associated genomic divergence (i.e., what regions of DNA differ between urban and forest populations) using complementary methods (genotype-environment association test – GEA, and principal components of genomic variation – PCA). We identified 33 genes that changed between urban and forest populations across all three pairs, suggesting a set of genes that may underlie urban adaptations. These genes related to immune function, wound healing, inflammatory responses, neural function, metabolism, and skin development and pigmentation. These exciting findings open the door for many new areas of exploration to better understand how these lizards are evolutionarily responding to urban environments. However, we caution that without clear links to phenotypic changes, we can’t be sure what these urban-associated changes actually mean for urban lizards at this time.

Urban-associated genomic divergence. Manhattan plot of genotype-environment association test (GEA) highlights several genomic regions diverging in parallel across three pairs of urban-forest populations. Dark gray points are SNPs identified as outliers by the GEA analysis and the red points are those that are also identified by the PCA analyses. From supplementary materials of Winchell et al. (2023, PNAS).

Genetic underpinnings of urban morphological adaptations

We next dug into the genomic basis of adaptive urban phenotypic changes: limb and toepad morphology. We confirmed that these populations exhibit the same parallel morphological shifts we previously observed: urban lizards had longer limbs, larger toepads, and more specialized toepad scales (lamellae) compared to their forest counterparts across all three population pairs. We conducted a series of GWAS (genome-wide association studies) to identify genomic regions associated with these traits across all of our populations. We identified a genome-wide signal of divergence associated with these traits, suggesting many genes are involved in shaping trait variation (not surprising for complex quantitative traits in wild populations).

We then leveraged the fact that we observe a strong and repeated divergence in these phenotypes between urban and forest populations to try to tease a signal from this noise. By focusing in on the subset of morphology-associated genes that were also identified by the urbanization-association test (i.e., the genes changing in urban lizards), we were able to hone in on a set of genes underlying the urban-specific morphological changes. We then further narrowed this set of genes to those that were associated with both anterior and posterior elements of each trait (i.e., are pleiotropic) to arrive at a set of 93 candidate genes. When we took a look at what some of these candidate genes are responsible for in other organisms, we were surprised to find that mutations in these genes are involved in bone formation, elongation, and pathology of limbs in humans and mice and are important for collagen, keratin, and skin development.

Genomic parallelism

Finally, we wanted to know if the same genetic changes underlie the parallel morphological changes that we observed. Using a couple of different approaches, we found that the same genomic targets of selection underlie urban-associated morphological divergence across populations. In other words, adaptive divergence in urban lizards is occurring via repeated selection on the same genetic regions independently across the three cities we looked at.

Figure 4 (D-F) in Winchell et al. (2023, PNAS) showing parallel divergence at the allele level for SNPs associated with urban morphological adaptations.

What do these genomic changes mean?

Our findings are significant for a number of reasons, but I’ll focus on ones that I think are most exciting.

First, these findings help us understand how rapid adaptive changes occur in complex phenotypes. Polygenic selection on standing genetic variation appears to be a key process underlying adaptation to urban environments and shifts in protein coding genomic regions (exons) can effectuate rapid changes in morphology. Intriguingly, many of the genes we identified as associated with the urban morphological adaptations in these lizards are implicated in disease states in humans and other model organisms. What does this mean? One interpretation is that changes in these genes in some circumstances are deleterious but in other organisms and under different selective landscapes may be exactly what is needed to effectuate rapid and drastic changes.

Second, these findings contribute to the ongoing debate in evolutionary biology on the relative importance of contingency versus determinism. In other words, does evolution proceed in the same way when organisms are challenged by the same selection pressures, or does idiosyncratic variation across populations and environments play a more important role? We found that parallel morphological shifts in urban lizards in response to parallel environmental changes are underlain by parallel changes in the DNA of these lizards. This parallelism at multiple levels of biological hierarchy suggests that evolution is proceeding in a predictable way, even down to the genetic level, in urban environments. What does this mean practically? This suggests (with some caveats) that as urbanization spreads and intensifies, these lizards have the genetic machinery to adaptively respond in the same ways and that we might be able to predict how other species will respond to urbanization just by looking at their DNA.

Third, although humans aren’t subject to the same whims of nature that lizards are, we still share these urban environments with them. That means that we are exposed to many of the same stresses, including pollutants, environmental stressors (e.g., light at night), and processed foods, which also have effects on human reproduction and survival (i.e., fitness). Identifying genomic regions that are changing in urban animals can help us understand the many ways that other organisms are affected by urban environments, including humans. Moreover, understanding how changes in the morphology genes that we observed can produce adaptive phenotypes in one organism and deleterious phenotypes in another may also open the door to better understanding diseases in humans.

In summary, by studying how these little lizards are adapting to urban environments we can better understand the mechanisms of evolution, answer big theoretical questions in biology, inform conservation in an urbanizing world, and shed light on organismal and human responses to the stresses of our modern world.

This post was also shared on Life in the City: Evolution in an Urbanizing World. 

To learn more about this study:

Read the paper: Winchell, K. M., Campbell-Staton, S. C., Losos, J. B., Revell, L. J., Verrelli, B. C., & Geneva, A. J. (2023). Genome-wide parallelism underlies contemporary adaptation in urban lizards. Proceedings of the National Academy of Sciences, 120(3), e2216789120.

Check out some of the press coverage: AP News, CBC radio, Primera Hora

Latest posts by Kristin Winchell (see all)