Research

Evolution occurs in heterogeneous environments, over large geographic ranges, and in the presence of variable species assemblages. My research focuses on how genetic variation is organized both at the species level and in populations of each species, what
evolutionary processes regulate genetic variation, and how ecology can affect these evolutionary processes.genetic_variation_schematic

At the species level, discrete genetic units are maintained by reproductive isolation between species, and countered by hybridization and introgression. Populations within a species are subject to analogous processes – similarity is maintained by gene flow, but differentiation is promoted by barriers to dispersal, genetic drift, and local adaptation. Evolutionary processes promoting and inhibiting differentiation occur in the context of a heterogeneous environment and different ecological interactions, and I incorporate ecological attributes of species into studies of evolutionary process to gain a more complete understanding of interactions between species. My work can be broadly summarized as two focal question:

1. How consistent are outcomes of hybridization when related fish species come into secondary contact?

2. To what extent are fish species evolutionarily cohesive across their ranges, and what are the ecological and evolutionary consequences of variation in species cohesion?

A major finding of my research is that seemingly homogenous biological processes often produce variable outcomes. In studies of hybridization, I describe how an apparently consistent process – secondary contact and interbreeding between related Catostomus fish species – actually results in extremely variable outcomes, both across locations and in different hybrid crosses (Mandeville et al. 2015, Mandeville et al. in prep.). Similarly, my work on genetic variation within species has shown that magnitude and spatial arrangement of genetic variation vary across co-occurring closely related species, suggesting that species cohesion is variable (Mandeville & Buerkle, in prep.). Furthermore, ecological outcomes can be greatly affected by variation in evolutionary processes. For example, relative resource use by Catostomus hybrids and parental species varies across rivers, suggesting different relationships between species and hybrids across locations (Mandeville, Hall, & Buerkle, in prep.). These insights are enabled by sampling from many regions of the genome and sampling across many populations and multiple species. Working on a realistic spatial scale for each of these studies allowed me to identify the distribution of possible outcomes for each process, rather than assuming generality from a small number of instances of a process.

Hybridization between species

Hybridization between species tests reproductive isolation, and in some cases might lead to erosion of differences between species or facilitate major evolutionary or ecological shifts. As genomic datasets have become available and sampling multiple locations where hybridization occurs has become more feasible, it has become apparent that hybridization often varies across locations where species come into contact, potentially as a result of polymorphism in reproductive isolation.

Hybridization in Catostomus fishes

Using genomic data, I quantified variation in hybridization among six Catostomus species at >20 locations in the Upper Colorado River basin in Wyoming and Colorado. Hybridization was extremely variable. Contact between Catostomus species resulted in hybridization and extensive backcrossing in some rivers, and no hybridization in others, despite co-occurrence of parental species. Similarly, some rivers contained many different hybrid crosses, while other rivers contained a single cross or no hybrids.

DSC01756

A bluehead x longnose sucker hybrid.

Potential explanations for variability in hybridization include variation for genetic components of reproductive isolation and environmental variation. Another possiblity is that historical contingency or chance might influence outcomes hybridization. Regardless of what causes variation in outcomes of contact between related species, the existence of this variation suggests that we need to more carefully consider our understanding of interactions between a pair of species. In many instances of hybridization, it might be misleading to think about hybridization as a consistent outcome of an interaction between species; instead, we might be better served by thinking of hybridization as a variable evolutionary process with a range of possible outcomes. This complicates conservation efforts, while also providing a ray of hope: in some locations, no hybridization occurs, and genetically unaltered populations of native species persist.

Ecological relationships of Catostomus hybrids and parental species

To understand variation in genomic outcomes of hybridization, I compared ecological roles of hybrid individuals and parental species across rivers. I used stable isotope analysis of carbon and nitrogen to examine and compare resource use across three parental species and hybrid crosses between these species. Animals incorporate carbon and nitrogen isotopes into their tissues in proportion to the isotopic ratios in their food sources, so comparing isotope ratios can reveal whether individuals are likely to use similar or different resources. I found that native parental species do not overlap in isotope space. Non-native species and hybrids overlap in isotope space with both native parental species, but patterns of overlap vary among rivers. Additionally, in many rivers hybrids and non-native species occupy more isotope space, potentially suggesting greater variation in how these species and crosses use resources within a river. If more diverse isotope ratios correspond to more diverse resource use, it is possible that generalist feeding ecology could enable hybrids to thrive in different ways in different locations, while native parental species might be more narrowly adapted to a subset of locally available resources.

Evolutionary cohesion within species

Species are composed of relatively independent populations of individuals, and these populations are distributed across heterogeneous landscapes, where they encounter different biotic and abiotic conditions. We therefore expect that species will vary genetically across populations, both due to neutral processes like genetic drift, and because of local adaptation.

Evolutionary cohesion in six Catostomus species

I used genomic data to characterize the geographic distribution of genetic variation in six Catostomus species in the Upper Colorado River basin. I used common and rare genetic variants to learn about differentiation and connectivity among populations. I also compared genetic diversity within populations. Interestingly, different species had different spatial patterns of genetic differentiation across the same area, suggesting that variation among species can lead to differences in species cohesion. Variable cohesion across species is consistent with our expectations for how geographic structure and dispersal of individuals should affect population genetic structure, but these results do suggest caution in treating species as uniform, equivalent genetic units. Species harbor different amounts of variation, and genetic variation should be taken into account when planning for research or conservation.

dist_tree

Each Catostomus species is somewhat cohesive, as shown in this neighbor-joining tree of Ne’s genetic distance, but magnitude and spatial arrangement of genetic variation differs across species.

Population genetic structure in burbot (Lota lota)

To understand connectivity among populations of burbot (Lota lota) in the Wind River basin, I analyzed population genetic structure (Underwood, Mandeville, & Walters 2016).
This is of critical importance for conservation because burbot populations are decreasing in abundance throughout their range, and conservation strategies for burbot in Wyoming require an understanding of burbot movement among tributaries of the Wind River. Additionally, population genomic data has helped identify source populations of individual fish that become entrained in irrigation canals and other water diversion structures, and to identify source populations for introductions of burbot elsewhere in Wyoming.