Organismal form and function arises through complex interactions among traits ranging from growth and reproduction to behavioral and hormonal responses to environmental stimuli. In this time of global climate change and unprecedented ecosystem disturbance, accurately predicting how organisms respond over the long-term to environmental change is essential. The degree to which traits are integrated, how strongly they covary, ultimately dictates a population's response to environmental pressures. The investigators capitalize on the reproductive system of mangrove rivulus, a fish that inhabits imperiled mangrove ecosystems and is capable of producing offspring genetically identical to the parent and all siblings, to determine with exceptional resolution how life history (e.g., growth, reproduction), behavioral (e.g., aggression, risk-taking), and morphological (e.g., shape, size) traits covary at the phenotypic and genetic levels. They expect differences among populations in the composition and interconnectedness of trait networks, which would provide evidence that the physical, community, and social environments exert strong influences on trait integration. The investigators also seek to understand the mechanisms underlying trait integration. Steroids like the androgen testosterone and the stress hormone cortisol influence the expression of multiple traits simultaneously. Thus, the investigators expect to uncover linkages between hormone profiles and patterns of trait integration. This study combines an integral knowledge of the mangrove rivulus' genetics with behavioral, ecological, and physiological techniques, to explore central questions in biology, how are complex phenotypes organized and how can we predict phenotypic responses to environmental change? By studying organisms as integrated wholes we can derive more accurate predictions about how ecosystem disturbances impact native flora and fauna. The investigators will participate in outreach activities to educate the public about mangrove ecosystem inhabitants, threats to the ecosystem, and how knowledge gained from basic biological exploration can transform our everyday activities and drive positive changes at the individual, community, and national levels.
Data will be maintained and available at (http://datadryad.org).
When we observe any organism, be it fungus, plants, or animals, we see a complex amalgamation of phenotypic traits that ultimately dictate the organisms' form and function. Often, the various aspects of an organism's phenotype are studied in isolation, where the focus is on understanding, for example, aggressive behavior, rates of reproduction or growth, or specific morphological features. These traits, however, can be mediated by the same underlying mechanism. Steroid hormones like testosterone and estrogen can affect the expression of behavior and certain morphological characteristics, and can regulate reproductive output - that is, steroid hormones can have pleiotropic effects on the phenotype. In this awarded study, we sought to examine the potential for hormones to mediate suites of phenotypic traits, and to determine whether correlations among traits were consistent or widely divergent within and among populations. We used a remarkable fish species, the mangrove rivulus (Kryptolebias marmoratus), to address these issues. This animal is one of only two self-fertilizing hermaphroditic vertebrates capable of producting offspring that are essentially clones of the parent and all siblings. Thus, we were able to generate replicates of the same genotype (clones), examine multiple genotypes per population, and explore whether phenotypes differed among eight geographically distinct populations. We raised rivulus in a common laboratory environment for approximately one year and quantified an arsenal of phenotypic traits including body size growth rates, fecundity, egg size, and reproductive rates as well as the areas of the caudal, anal, and dorsal fins, aggressive behavior, risk-taking behavior, and exploratory behavior. Because rivulus inhabit notoriously variable (and sometimes inhospitable) mangrove swamps and because rivulus populations experience incredibly different types of mangrove environment (e.g., variation in temperature, tidal flux, the types of heterospecific competitors, and predation threat), we hypothesized that: 1) phenotypic traits and their correlations would vary among populations, indicating that each population adapted to its local conditions; 2) a diversity of phenotypic traits (and correlation structures) would be maintained in each population because tremendous spatial and temporal variation in the mangroves coupled with known microhabitat usage by rivulus would support many phenotypes with equal fitness; and 3) a significant proportion of variation in how phenotypic traits are correlated would be explained by baseline levels of four hormones (cortisol [stress hormone], testosterone, estradiol, and 11-ketotestosterone [a fish-specific androgenic steroid]). Our data demonstrated that exploratory behavior, risk-averse behavior, titers of testosterone and the stress hormone cortisol, age at maturity, reproductive investment (egg size), body size (mass and standard length), and growth rates varied considerably among populations despite the animals being raised in a common laboratory environment. This provides strong support for the hypothesis that rivulus populations have experienced divergent selection pressures that favored phenotypic characteristics best adapted for local conditions. In addition, our data demonstrated considerable within-population phenotypic variation, providing strong support for the possibility that mangrove habitats, and the abundant microhabitats therein, maintain a diverse contingent of phenotypes perhaps via some form of balancing selection (e.g., frequency-dependent selection, fitness variation at any number of spatiotemporal scales). Preliminary analyses indicate potentially interesting variation in the correlation structure among phenotypic traits from population to population. For example, in some populations, there existed the classic trade-off between growth and reproduction (individuals that grew faster matured more slowly from a reproductive standpoint); in other populations, this trade-off disappeared, with some genotypes showing simultaneous investment in growth and reproduction. There were very few correlations among behavioral and life history traits and, contrary to our hypotheses, we found no convincing evidence that steroid hormones mediate linkages between phenotypic traits. With that said, we are now engaged in a productive collaboration with researchers in the UK to investigate, using a powerful multivariate statistical framework, the extent of variation in the "integrated" phenotype among populations and the potential role for hormones as mediators of phenotypic integration. In addition, because we can control, in remarkable fashion, the genetic composition of our experiments, we will be able to estimate the amount of variance-covariance among phenotypic traits that has a genetic basis; with these data in hand, we will be able to predict the evolutionary trajectories of the different populations when particular types of selection are applied. As part of our broader impacts, we have been active in outreach in state parks and preserves throughout Florida, where we engage the public while conducting field work and where we have forged productive relationships that will allow us to lead programs for children and adults about mangrove ecosystem biodiversity and about the peculiar reproductive strategy of the mangrove rivulus.
