Most of my dissertation research focuses on the Bahamas mosquitofish (Gambusia hubbsi), a small, promiscuous live-bearing fish that inhabits blue holes across Andros Island, The Bahamas.
Blue holes are water-filled vertical caves that filled with water ~10,000 years ago, and have since been colonized by just a few fish species. Blue holes can be characterized as either high- or low-predation (from the perspective of G. hubbsi) based on the presence or absence of a predatory fish, the bigmouth sleeper (Gobiomorus dormitor).
No environmental factor systematically co-varies with predation regime, and these populations have been isolated for thousands of years, rendering this system a natural experiment to test predictions for how predation can affect evolutionary trajectories.
Genitalia evolve very rapidly—probably faster than any other trait possessed by internally fertilizing organisms—and exhibit diverse and often bizarre morphologies, but why? Consensus in the field is that rapid genital evolution is driven by sexual selection. Although it is commonly understood that the nature and strength of sexual selection can depend on ecological conditions, whether genital evolution may ultimately be driven by ecological variation has rarely been considered. Threat of predation can be a particularly potent force altering the context of sexual selection by promoting less conspicuous mating displays, reduced courtship, and faster copulation duration. Male G. hubbsi transfer sperm internally to females using a modified anal fin, the gonopodium.
The very distal tip of the gonopodium is highly differentiated, comprising hooks, spines, and serrae encapsulated in soft tissue. I am characterizing G. hubbsi gonopodial distal-tip shape and testing for evolutionary divergence between populations with and without predators. In a related system (three closely related Gambusia species inhabiting tidal creeks across Bahamian islands), I am investigating whether human-mediated changes to the environment (habitat fragmentation resulting in predator exclusion) has driven rapid (~50 years) divergence in male genital shape.
Honest sexual signals
Sexual signals, such as brightly colored traits, are often used to attract mates and provide information about individual “quality”. Body condition has long been used in sexual selection studies as a proxy for fitness, but it is becoming increasing clear that body condition does not predict whole-organism performance, which is ultimately the target of natural selection. In order to gain genetic benefits for their offspring, females should prefer males that are successful in their local environment. But, the most important type of whole-organism performance will depend on local selection pressures. Rapid-burst escape swimming speed is of utmost importance for Bahamas mosquitofish living with predatory fish, yet that performance is irrelevant in blue holes without predators. At the same time, although resource availability does not differ between high- and low-predation blue holes, low-predation populations have much greater population densities, so competition for resources and foraging efficiency will be more intense than in populations with predators. I am employing a combination of lab and field experiments to investigate the nature of honest sexual signals in Bahamas mosquitofish.
Biologists have long recognized that organismal phenotypes are ultimately complex networks of correlated traits (in other words, traits are rarely independent of each other), a concept known as phenotypic integration. These trait integration networks can have important evolutionary and ecological consequences by potentially constraining evolutionary trajectories, but also perhaps by adaptively enhancing functional performance via correlational selection. Natural and sexual selection may both act directly on trait correlations, but may favor different strengths of covariance, or even different networks. I am investigating the nature of phenotypic integration in Bahamas mosquitofish by characterizing the trait correlation structure of 12 different behavioral, morphological, and physiological traits on wild Bahamas mosquitofish, investigating whether trait integration networks have diverged between high- and low-predation populations, and testing predictions for functional ecological and sexual performance based on particular trait integration networks.