How does natural selection create and maintain discrete phenotypic variation?

Evolution by natural selection cannot occur without phenotypic and genetic variation, yet selection depletes this very stock of variation. How does variation persist despite chronic selection? Polymorphism offers insight into this question because phenotypic variation can be partitioned into discrete “morphs”, often corresponding to simple genetic differences at one or several loci. On the Great Bahamas Bank, female brown anoles exhibit one of three distinct dorsal patterns: Diamond (D), Bar (B), and intermediate Diamond-Bar (DB) morphs (Fig. 1). Ryan Calsbeek, Lauren Bonvini and I have begun to study this sex-limited polymorphism to elucidate its genetic basis and determine whether it is maintained by natural selection.

Related Papers

Calsbeek, R., L. Bonvini and R. M. Cox. 2009. Geographic variation, frequency-dependent selection, and the maintenance of a female-limited polymorphism. Evolution

Check out Ryan Calsbeek’s lab for other papers related to the evolution of polymorphism in anoles.

Figure 2. Left panels depict two primary study islands in The Bahamas. Center panels show the relative frequencies of Diamond (D), Diamond-Bar (DB), and Bar (B) morphs at 13 individual sites located on these two major islands and their offshore cays. Right panels show the mean frequency of each morph across sites on each major island. From Calsbeek et al. (2009)

Figure 1. Photos of female brown anoles, illustrating typical variation within and between the three dorsal pattern morphs. From Calsbeek et al. (2009).

Geographic variation. One striking aspect of this dorsal-pattern polymorphism is that the frequency of alternative morphs varies between the islands of Eleuthera and Great Exuma. Diamond-Bar females are common on each island, but Diamond females are rare on Eleuthera, whereas Bar females are rare on Great Exuma (Fig. 2). These patterns are highly consistent across separate populations on each island, with the notable exception of several barrier cays offshore from Great Exuma. On two of these cays, the “rare” Bar morph has been lost, but on a third cay, the “common” Diamond morph is almost entirely absent (Fig. 2). This geographic variation raises interesting questions about the ways in which natural selection, gene flow, and genetic drift interact to shape biological variation.

Genetic basis. Breeding studies reveal that dorsal pattern has a strong genetic basis, although the most parsimonious mechanism for inheritance involves at least two loci. One locus appears to control the expression of the Diamond and Bar phenotypes, while a second locus exerts an epistatic effect that overrides the production of these “pure” phenotypes to produce the Diamond-Bar pattern. Pinpointing the details of this genetic mechanism is complicated by the fact that males do not express the polymorphism.

Adaptive significance. One hypothesis for geographic variation in polymorphism is that natural selection favors different morphs on different islands, with inter-island dispersal potentially maintaing genetic variation despite selection. However, our results do not indicate any difference in morph-specific selection between Eleuthera and Great Exuma. In other species, polymorphism can be maintained by negative frequency-dependent selection (i.e., selection favors the rare morph). Our results suggest an alternative scenario:

Fitness (adult viability) is independent of relative frequency in Diamond and Bar morphs, but it increases dramatically with relative frequency in Diamond-Bar females (Fig. 3). Typically, this type of positive frequency-dependence should eliminate variation by driving one morph to fixation. However, in brown anoles, the genetics of the dorsal pattern may actually maintain the polymorphism. Because the epistatic Diamond-Bar locus overrides the “pure” Diamond and Bar phenotypes, selection favoring Diamon-Bar phenotypes should maintain both D and B alleles in the population. These D and B alleles, along with alternative alleles at the epistatic locus, may also be sheltered from selection in males, where they do not influence dorsal phenotype. We are currently investigating potential morph differences in other components of fitness, such as fecundity, and are planning to characterize the geographic mosaic of this polymorphism across a larger spatial scale throughout the entire Bahamas archipelago.

Figure 3. Relative fitness as a function of frequency for Diamond (black), Bar (gray), and Diamond-Bar (white) morphs. Points represents initial frequencies and subsequent survival for a given year of study.