It has recently been established that even first-generation hatchery fish can have much lower reproductive success in the wild when compared to their wild-born counterparts owing to rapid genetic adaptation to captivity (i.e., the hatchery). In fact, a single generation in captivity can result in a substantial response to selection on traits that are beneficial in captivity, but severely maladaptive in the wild. Using a combination of pedigree analyses and gene expression data, we are searching for the proximate and ultimate causes of fitness differences between hatchery and wild fish.
As this research continues to unfold we will examine the broader patterns of contemporary evolution. Are there characteristics unique to salmon that predispose them to rapidly adapting to novel environments? How common is contemporary evolution? What are the relevant interactions between epigenetics and genetic adaptation? Answers to those questions will have broad implications for the fields of ecology and evolution.
Predicting and understanding species’ dispersal patterns remains a fundamental question in ecology and evolution. For example in marine systems, the vast majority of species possess a pelagic larval stage, during which the larvae may disperse hundreds to thousands of kilometers. Yet it is difficult to directly track marine larvae, because of the vast size of the ocean and the miniscule size of larvae. Nevertheless, it is vitally important to know where and to what extent larvae disperse in order to: (1) effectively characterize metapopulations, (2) provide effective fisheries management, and (3) design and implement successful networks of marine protected areas (MPAs).
One further challenge to studying dispersal in marine systems is that marine species are often characterized by high variance in reproductive success. Thus, disentangling the relative contributions of genetic drift, gene-flow, and selection (e.g., local adaptation) is a primary research objective in marine ecology and evolution.
Parentage and Relatedness
Parentage and relatedness measures are powerful tools for establishing pedigrees, examining patterns of reproductive success, and for directly determining dispersal events. The goals of all parentage analyses are to minimize type I errors (the false classification of unrelated individuals as parent-offspring pairs), while maximizing power (correctly identifying as many real parent-offspring pairs within a data set). In many data sets it may be very difficult to detect any true parent-offspring pairs because the number of candidate parents sampled is low or simply unknown. Therefore, I have developed new Bayesian parentage methods and grandparentage methods that do not require any demographic estimates and that maximize power while minimizing false assignments. These methods should prove valuable to answer a wide variety of research questions.