Research

My research program focuses on the evolution of complex phenotypes. I am particularly interested in the evolution of phenotypic integration, that is, how natural selection shapes groups of traits to produce well-adapted organisms. I use both empirical and theoretical approaches and incorporate ideas and techniques from diverse fields, including evolutionary genetics, behavioral ecology, and physiology.

Quantitative genetic integration
How does natural selection shape the genetic architecture of complex adaptations?

Genetic correlations between traits can be from two different perspectives in evolutionary theory. On the one hand, genetic correlations may be seen as a constraint because they can channel a population's evolutionary trajectory in a direction that may be less than optimal. On the other hand, genetic correlations may themselves evolve when selection favors certain groups of traits and thus may be seen as signatures of adaptation.

We know little about how patterns of genetic correlations (the "G matrix") evolves in nature. Caribbean Anolis lizards provide a unique opportunity to study the evolution of G because morphologically similar habitat specialists (or "ecomorphs") have evolved multiple times in different lineages, allowing us to separate the effects of shared history from shared ecology.

In collaboration with Butch Brodie and Jonathan Losos, I am measuring the G matrix in eight Anolis species that represent unique origins of four ecomorphs from three different lineages. We are also using this system to ask quantitative genetic questions about development and sexual dimorphism.

One of the major evolutionary mechanisms leading to integration is correlational selection, which occurs when groups of traits interact to affect fitness. Over time, correlational selection is expected to shape the G matrix, generating genetic correlations between functionally related traits. In earlier working on a songbird, the dark-eyed junco, I showed evidence that correlational sexual selection can lead to genetic integration of traits (body size and plumage) used by males in courtship and competition.

Genomic integration
How do evolutionarily changes fit in with the rest of the genome?

Integration is also an important consideration when considering evolution at a molecular level. Genes do not exist in a vacuum; they often interact with many other genes to produce complex phenotypes. Genomic and bioinformatic tools are now allowing us to ask ever more complex questions about the genetic basis such adaptations. For example, does selection tend to cause coordinated change in different parts of the genome? Do related genes evolve in predictable ways in response to a selection pressure?

Garter snakes in the genus Thamnophis are one of the best examples of evolution in action. These snakes consume one of most toxic animals in the world, the rough-skinned newt Taricha granulosa, which can produce enough tetrodotoxin (TTX) to kill most potential predators. The snake's resistance to TTX derives in part from evolutionary changes in voltage-gated sodium channels expressed in skeletal muscle(encoded by the gene SCN4a). TTX blocks these channels in other animals, but not in garter snakes.

SCN4a belongs to a family of genes that in tetrapods consists of nine different genes expressed in different tissues of the body. In order for a snake to survive when eating a toxic newt, it must maintain the functionality of other tissues with voltage-gated sodium channels. Thus, we expect other genes in this family, which are expressed in tissues such as peripheral nerves and brain, to exhibit parallel adaptation to TTX. I am currently using the garter snake bacterial artificial chromosome library and next-generation sequencing to find these genes and look for signatures of evolutionary change in response to toxic prey.

Physiology of integration
What mechanisms underlie integrated suites of traits?

Because of their ability to regulate the expression of many traits simultaneously, hormones are a major physiological factor underlying the integration of the phenotype. In songbirds, testosterone is particularly important for the expression of several traits related to obtaining mates. In collaboration with Ellen Ketterson, I have been exploring individual variation in testosterone-mediated phenotypes in dark-eyed juncos.

We found that many traits are associated with short-term fluctuations in testosterone levels mediated by the release of gonadotropin-releasing hormone (GnRH) in the brain. Such short-term testosterone elevations were associated with variation in territorial behavior, parental behavior, and plumage and seem to underlie the fundamental trade-off between mating effort and parental effort experienced by male juncos.

Testosterone is also intimately related to fitness. We found strong stabilizing selection on testosterone production via both survival and reproduction. This suggests that selection favors a stable combination of testosterone-mediated traits, which should maintain the current pattern of physiological integration.

Interaction and integration in social evolution
How does the social environment affect evolutionary change?

The theory of interacting phenotypes examines how social interactions affect the evolutionary process. In collaboration with Butch Brodie, Allen Moore, and Jason Wolf, I have been working on several extensions of this theory, including the development of statistical methods to measure model parameters.

Our work has led to generalized equations for predicting the evolution of social traits and an expansion of Hamilton's rule for the evolution of altruism that incorporates both IGEs and social selection. I am currently pursuing several lines of theoretical work, including modeling phenotypic interactions between species, elucidating the relationships between indirect genetic effects and game theory, and asking how social interactions affect quantitative genetic integration.

 

Collaborators