Anole Annals

Anolis sagrei mating. Image from Bob Cox’s lab website 

Although the Evolution meetings are coming to a close, we get to go out on a high note. Christian Cox gave one of the last talks of the day discussing the hormonal basis for gender differences in sexual size dimorphism in anoles. Sexual size dimorphism (SSD), or the tendency for the sexes to differ in the size of different traits, has been widely documented in nature. Usually the male exhibits comparatively larger features, such as bigger body size or larger ornaments. Anoles are an intriguing case of SSD, as the traits that can exhibit dimorphism can vary widely among species. Some species, such as Anolis carolinensis, exhibit SSD in multiple traits, including body size, head shape, and dewlap size. In contrast, other species exhibit minimal SSD. As an example, A. distichus from the Caribbean island of Hispaniola tends to show no SSD in body size or head shape, but has strong SSD in dewlap size.

Christian Cox and his collaborators posit that one mechanism underlying SSD may be a pleiotropic regulator that can couple and decouple dimorphism in different phenotypes and their candidate for this study was testosterone. They conducted experiments manipulating levels of testosterone in adult males and females of Anolis sagrei and assessed how body size, head shape, and dewlap traits changed. Anolis sagrei is a particularly good system for assessing the role of SSD in anoles. Male A. sagrei can be up to 50% larger and three times more massive than females.

To conduct the study, they took three year-old male and female lizards and gave them either testosterone or blank subdermal implants. They maintained lizards under laboratory conditions for two months and then gathered information on morphological dimensions and dewlap characteristics. Under testosterone treatment, males and females grew similarly, whereas males grew faster than females in the control group. This merits restating – they were able to make females grow like males just by applying testosterone! Clearly testosterone has strong effects on male-specific growth patterns.

To determine if testosterone affects metabolism, they measured metabolic rate using stop-flow respirometry. They found that testosterone treatment increased metabolic rate for males and females. Correspondingly, they found that visceral fat bodies were lower in testosterone treated animals, suggesting that increased growth is caused by shunting energy towards growth and away from storage metabolism. They further determined that testosterone treatment increased the size of the humerus and femur, but had no significant effect on jaw length and head width. Because this species exhibits little SSD with respect to head dimensions, perhaps this finding is not surprising, but I would be curious to know whether testosterone influences skull growth in species with SSD in head dimensions, such as A. carolinensis.

Finally, the authors found that testosterone led to increased dewlap size in both males and females. In fact, the dewlaps of testosterone-treated females were comparable in size to those of control males and eroded the sex differences that otherwise existed between them. Testosterone treatment decreased the saturation and brightness in the dewlap, leading the authors to suggest that it accelerates its development, as they posit that this color is representative of the fully developed dewlap in the wild.

Thus, they find strong evidence that testosterone plays a large role in modulating SSD in anoles. In particular, it abolishes differences in growth in various traits except for skull shape. And it can create male-like females as well as forge super-males. It would be interesting to see if, in addition to acquiring a male-like morphology, the females would tend to act like males, as well. Their next step is to conduct testosterone manipulation experiments in A. distichus, a species that has low SSD in body size and head shape, but strong SSD in dewlap size, to determine if the effects of testosterone are repeatable in a system exhibiting a pattern of SSD that is different from A. sagrei.

*****************

Extreme sex differences in the development of body size and sexual signals are mediated by hormonal pleiotropy in a dimorphic lizard. Authors: Cox, Christian L.; Hanninen, Amanda F; Cox, Robert M.

Gamble and Zarkower (2012) Current Biology

Tony Gamble, a postdoctoral researcher working with Dave Zarkower at the University of Minnesota, presented his work on uncovering sex-specific markers in geckos and anoles. Recent years have seen a large impetus to understand how sex chromosomes evolve. Sex chromosomes can be involved in sex-specific adaptation, genetic conflict, and other important modes of evolution. This line of research is particularly imperative in reptiles because not only do we have comparatively little information about sex chromosomes in this group, but different types of sex determining mechanisms have evolved multiple times and so there are likely multiple sex-specific mechanisms and multiple evolutionary transitions are at play (see Figure above).

Traditionally sex chromosomes were discovered by karyotyping, which is a method of separating and identifying the chromosomes. This is problematic in reptiles because the sex chromosomes of many species are homomorphic, meaning they are similarly shaped and, oftentimes, quite small. Gamble and Zarkower tried a different approach – RADseq – for identifying sex chromosomes. RADseq uses restriction enzymes to identify sex-specific markers. Their reasoning is that in XY systems (i.e., males are the heterogametic sex), you would expect males and females to exhibit X-specific markers and males to exhibit sex-specific markers unique to the Y (i.e., the non-recombining region). In ZW systems (i.e., females are the heterogametic sex), you would expect the opposite. In theory, this could prove a cheap and fast way to determine the sex chromosomes of different species and develop sex-specific markers.

The challenge for this study was to determine the sex chromosomes for the crested gecko and for the anole. Unlike the crested gecko, Anolis is genome-enabled and we have evidence that they are an XY system, and so they used anoles to pilot their method and confirm that it works before trying it on the crested gecko. However, anoles are not without their challenges. The sex chromosomes are not only homomorphic, but they are also micromorphic, meaning they are quite small. Furthermore, the Anolis genome was built using a female anole, making finding sex-specific markers on the non-recombining region (i.e., the Y chromosome) that much more challenging. Their RADseq approach worked quite well, however, as they were able to recover a male-specific marker in A. carolinensis, which they were able to confirm with PCR amplification. They repeated their results using more A. carolinensis (from a different clade), A. sagrei, and A. lineatopus, and were able to recover the same locus. When they conducted this method in the crested gecko, they found evidence for a ZW system and, correspondingly, recoverd two female-specific markers. Thus, they found that RADseq will work in a variety of taxa, even if they are not genome-enabled, and can successfully be used to uncover sex-specific markers. A neat application of this method is that, using their sex-specific primers, you can sequence an embryo to determine its sex, something that was not previously possible.

Figure from Nicholson et al. (2007) showing variation in dewlap color among various species of anoles.

As Nick Crawford, recent Ph.D. of Boston University, points out, the genomics era allows scientists unprecedented access to understanding the genetic basis of adaptation and, by extension, the genetics of speciation. For his doctoral thesis, Nick focused on understanding the genetics of colorful adaptation in Anolis lizards, which is genome-enabled. Adaptive radiations provide lots of variation among closely related organisms, making anoles a great system for studying the genetics of adaptation.

One feature of anoles that really stands out is how colorful they are. Just a casual glance at some of the color variation in dewlaps among species reveals that color is likely an important component of species diversification in anoles. Nick focused on Anolis marmoratus, a colorful anole from the Caribbean island of Guadeloupe. Anolis marmoratus is an excellent choice for studying the genetics underlying color. This species exhibits strong geographic variation in coloration and, as I discussed in my talk a few days ago, lacks a strong signal of genetic structure. In this case, searching for the genes underlying local adaptation can be conducted without the confounding effects of population structure.

One of Nick’s slides showing the ranges of A. m. marmoratus (orange color) and A. m. speciosus (blue) on the islands of Basse Terre (left) and Grande Terre (right) in Guadeloupe.

Nick focused on A. m. marmoratus, which has red marbling on its head, and A. m. speciosus, which has a blue head and, oftentimes, a blue body and tail. These two species are clinally distributed along the eastern side of Basse Terre and A. m. speciosus ranges into the nearby island of Grande Terre (see Figure 1). Rather than use RAD tags, Nick sequenced the genomes for 20 individuals (10 each per subspecies). For every 5 kb along the genome, Nick measured divergence using various metrics of structure and assessed sequence divergence.

Overall, Nick found that about 2% of the genome falls within divergent regions for these two subspecies. Importantly, he found divergence in two genes involved in carotenoid pigmentation and one gene involved in melanosome transport. Divergence in the two carotenoid genes could very well underlie the color divergence in A. m. marmoratus, which has distinct red marbling on its head. These genes fall in regions containing several fixed single nucleotide polymorphisms (SNPs) in a row. Nick suggests that these are likely single haplotypes that are being selected in different environments. Finally, he found no evidence of coding sequence changes, and so he posits that the modifications are probably cis-regulatory in nature. For many years we have been waiting to find out how divergence in coloration occurs in anoles. After seeing Nick’s work, it appears we are closer than ever before to understanding local color adaptation at a genomic level, so stayed tuned to his work for more to come.

Yesterday, Matt McElroy presented a phylogenetic analysis of the Puerto Rican radiation of anoles. The work was focused around the “stages of radiation” hypothesis that states that divergence occurs along different niche axes at different points in time. In the case of anoles, it has long been argued that the last stage in radiation is divergence of ecomorphologically similar species into different climatic niches.

McElroy constructed a phylogeny for 180 individuals of eight species, encompassing the geographic distribution of these species (most of which occur island-wide). Ten genes were sequenced, nine nuclear and one mitochondrial. The resulting phylogeny was well-resolved and in agreement with previous phylogenetic hypotheses, indicating that ecomorphs evolved relatively early in the radiation and that closely related sister taxa pairs are usually members of the same ecomorph, but differ in climate–the one exception–which always has struck me as odd, but apparently is correct–is the sister taxon relationship between the deep rainforest trunk ground species A. gundlachi and the xeric grass-bush species, A. poncensis.

The time of divergence was estimated for each of the four sister taxa pairs, indicating that there were three phases of radiation. The deepest split, pegged at 15 mya, was between the two trunk-crown species, A. evermanni and A. stratulus. At 10 mya, two pairs split simultaneously, the aforementioned one above and the two trunk-ground species, A. cooki and A. cristatellus. Both of these pairs include one species that occurs in the xeric southwestern portion of Puerto Rico, perhaps not a coincidence? Finally, 5 mya, the two grass-bush species, A. pulchellus and A. krugi diverged. These latter two species have recently been shown to hybridize, and McElroy’s data confirms that this is the only one of the four pairs in which hybridization occurs, perhaps due to their recency of divergence?

This is a fabulous example of detailed phylogenetic work spanning both interspecific comparisons and including the extensive degree of phylogeographic divergence that occurs within many anole species. More work of this sort is needed on anoles on the other three islands of the Greater Antilles. The monophyletic Jamaican radiation would be a good starting point.

The study of geographic variation has long been a foundation of evolutionary biology. As Martha Muñoz explained, in recent years attention has focused on whether geographic variation is the first stage in the process of speciation resulting from divergent selective pressures. Making the case for such ecological speciation requires demonstrating the occurrence of divergent selective pressures correlated with environmental differences and that the resulting  phenotypic differences lead to a reduction in genetic exchange among populations in different populations.

Martha reported on a study of the highly variable Guadeloupean anole, Anolis marmoratus. Detailed population level sampling confirmed that differences in body color correlate with environmental differences, with bluer populations in wetter habitats. These results strongly suggest the role of divergent natural selection as the cause, particularly because the differences were replicated across two different transects. Moreover, the fact that the color differences occur only in males suggests a role for divergent sexual selection.

Coalescent analyses of molecular genetic differences suggested that these differences occurred in the presence of gene flow, i.e., a parapatric model of divergence, rather than in allopatry. However, estimates of ongoing gene flow find no evidence of reduced gene flow across environmental  borders. In other words, even though selection is driving phenotypic differences, these differences are not leading to a reduction in genetic exchange–selection does not appear to be leading to speciation, contrary to the ecological speciation hypothesis.

These results are in agreement with studies on several other Lesser Antillean anoles. Martha pointed out that in situ cladogenetic speciation only occurs in anoles on the large islands in the Caribbean, suggesting that the lack of opportunity for allopatry on small islands precludes speciation from occurring, even in the presence of strong divergent selection.

These results were recently published in Molecular Ecology and are further discussed in a previous post.

In part II of the day’s Bahamian Anolis sagrei talks, Bob Cox addressed the question: How do males and females evolve different phenotypes despite sharing the same genome? And what better organism with which to study that question than anoles—lots of variation in dimorphism occurs not only among species, but also among populations within species.  This study focused on two populations of Bahamian A. sagrei in which the extent of male-biased dimorphism in body sizes varies—on Exuma, males are 33% bigger than females, whereas on Eleuthera, they are only 22% bigger. The difference is entirely the result of differences in male body size.

Cox asked three questions:

1. Do populations differ in sex-specific natural selection on body size?

2. Are sex-specific growth trajectories that give rise to sexual size dimorphism (SSD)?

3. Are differences indicative of differences in genetic correlations between sexes?

To address the first question, animals were caught at the start of breeding season; measured, marked and released; and then recaptured three months later to test for selection on body size. They found that selection is stabilizing on size in females, whereas there is strong directional  size for large size in males. However, selection doesn’t seem to differ among populations, so differences in selection would not seem to account for the differences in SSD.

However, their recapture studies allowed them to measure growth rates, and they found that males grow significantly faster on the island with higher SSD (Exuma). Animals were then raised in a lab common garden to see if the same differences in growth occur. Preliminary results show that males from Exuma grow faster in the common garden, suggesting either genetic differences or something early in development lead to growth differences (these are wild caught animals).  To further test this hypothesis, lab-raised juveniles were tested and early results indicate no differences in growth rates, which suggests that differences in SSD may not be genetically based, but these results are very preliminary. Despite lack of evidence for sex-specific growth trajectories, there is evidence for sex differences in genetic correlations between the sexes for body size (i.e., are growth rates correlated in opposite sex siblings). These correlations are much weaker in the high SSD population than in the low SSD population—these results, too, are preliminary.

“How do ectotherms evolve in response to changes in their thermal environment?” asked Mike Logan of Dartmouth University. Logan and colleagues studied adaptive evolution in the thermal performance curve—what is the optimal temperature for performance? How does selection work on components of the curve—i.e., optimal temperature, performance breadth (range at which organism can perform at 80% of maximum) and maximal performance capability.

Logan made four predictions:

1. Optimal temperature should be coadapted with mean body temperature, which may be related to mean environmental temperature;

2. Performance breadth should be coadapted with variance in body temperature;

3. Specialist-generalist temperature. Individuals faced with broad range of temperatures can’t specialize as well to particular temperatures;

4. thermodynamic effect—the “hotter is better” hypothesis, i.e., that a positive correlation will exist between maximum performance and optimal performance temperature

The study focused on two populations of A. sagrei, a natural population on the island of Great Exuma and an experimental transplanted population on Eleuthera. The transplant involved moving lizards from a shady site similar to the natural one to a more exposed, warmer site. The researchers measured running speed at five temperatures and calculated a performance curve for each individual, then marked animals and let them go, recapturing them at end of season to quantify selection on performance curve characteristics. They also measured operative temperature at the sites. Natural population and the source location for the transplant were similar and cool, with a mean environmental temperature of about 29. The warm transplant site was about 2-3 degrees higher, also with higher variance.

Positive directional selection for optimal performance temperature was detected in the transplanted population—lizards with higher optimal temperature survived better. No such selection occurred in baseline population. Moreover, selection on performance breadth was stabilizing in the natural population and directional in transplanted population.

This fabulous study has important implications, as Logan noted. Anolis sagrei is known to thermoregulate strongly, but these data suggest that behavior won’t inure individuals from strong selection in a novel, warmer habitat. Moreover, the study has important implications for the ability of populations to adapt to changing climates.

Ian Wang kicked off the anole portion of this year’s Evolution meetings by presenting his Young Investigators Prize lecture on the role of geographic distance and environmental differences as causes of genetic differentiation among populations. Teasing out the effects of these two variables is difficult because they tend to be correlated–nearby populations tend to share similar environments, whereas more distant populations are more likely to occur in different environments.

Ian reported on a new program he has developed, entitled MMRR (pronounced “merrrrr”–I can’t remember what it stands for) to statistically disentangle the two effects. He presented case studies on Yosemite toads and strawberry poison frogs where application of this new method revealed a previously unappreciated effect of environment in determining genetic differentiation. He then reported on a comparative analysis of 17 Caribbean Anolis species in which, as a generality, geographic distance (“isolation-by-distance”) accounted for twice as much of the variation in genetic differentiation as did environmental differences (“isolation-by-environment”).  Interestingly, and inextricably, the major exception was the three species on Jamaica, for which IBE accounted for very little variation. These results were recently published in Ecology Letters and the subject of a previous post.

Ian then presented new results examining geographic variation in morphology in two co-occurring Puerto Rican species, Anolis cristatellus and A. stratulus. In an extension of the structural equation modelling approach used in the anole work, Ian investigated the extent to which morphological variation among populations could be accounted for by environmental variation, controlling for geographic and genetic differences among populations. The results indicated that body size variation in body species was correlated with environment, with larger lizards in hotter/drier areas  (see photo above). In addition, in A. cristatellus, longer-limbed lizards also occur in hotter/drier areas. This is an exciting approach that opens new doors to the study of geographic variation in morphology, and I anticipate that it will be widely emulated.