Male Competition and Fitness Landscapes
I am working with Liliana Lettieri in Jenny Boughman's lab at Michigan State University on a project investigating the genomic basis of morphological traits and behaviors important for sexual isolation in benthic and limnetic stickleback fish from British Columbia (Gasterosteus species complex).
This work requires a "mapping population", which is a set of second generation hybrid fish with a mixture of both species' genomes. These hybrid fish have heterogenous genomes; this paired with markers developed from the original pure species can be used to locate segments of the genome that influence traits of interest.
An added, but underutilized, bonus of creating this population of hybrid fish is that we can get a detailed understanding of the phenoptype-fitness landscape. This is because this population includes individuals with unique trait combinations no longer seen in either pure species. In effect, we fill out the multidimensional phenotypic landscape.
I am currently focusing on the male competition fitness landscape. This landscape is of interest because male competition for territories and nests occurs before females shop for mates. Thus, male competition acts as a phenotypic filter that could have important impacts on which fish females are able to sample, and therefore, on how speciation proceeds. This work is now in press at Ecology Letters!
Gene Expression During Female Mate Discrimination
A female’s mate choice decision is critical for her fitness, and discrimination against heterospecifics is an important component of this process. I am working in Jenny Boughman's lab at Michigan State University and Hans Hofmann's lab at University of Texas - Austin on a BEACON funded project to examine changes in gene expression that occur in response to courtship by male conspecific and heterospecific sticklebacks. This work was done with the help of undergrads Jonatan Martinez and Benjamin Wurst and REU Katelyn Doolittle.
Cognition and Natural Selection:
I am taking advantage of the recent species radiation of sticklebacks in British Columbia (Gasterosteus species complex) to study cognitive evolution in a comparative framework. Selection is strongly implicated when consistent differences between limnetic and benthic fish are demonstrated across lakes. I use two very different, but complementary approaches:
1) First, I conduct behavioral experiments to assess specific cognitive abilities in each species. As one example, I have been quantifying social information use, specifically the ability to use the feeding behavior of others to locate a food patch. This work has been conducted in Jenny Boughman's lab at Michigan State University and has involved undergrads Whitley Lehto, Ross Minter, Ashley Baird, and Sarah LoPresto and REU Ashley Lindo. I am starting to expand this work to other types of learning, for example, quantifying the relative role of genes and environment in producing species differences in spatial learning (a project initiated by undergrad Jonatan Martinez with help from undergrad Benjamin Wurst and REU William Fetzner). I am also developing an automated operant conditioning device for large-scale quantification of discrimination and reversal learning (a project with undergrads Clayton Batko and Peter Vites and REU William Fetzner).
2) Second, I collect neuroanatomical data on specific brain structures. Here we present data from a magnetic resonance imaging (MRI) project. Limnetic and benthic fish differ in their feeding habits, sociality, habitat complexity, body size, shape, antipredator defenses, mating traits, and dependence on vision vs. olfaction, suggesting differences in cognitive abilities and neuroanatomical structures exist. This work is a collaboration between myself, Jenny Boughman at Michigan State University and Victoria Braithwaite and Thomas Neuberger at Penn State University, with help from REU Katelyn Doolittle.
Cognition and Sexual Selection:
In my doctoral dissertation research at the University of Maryland (advised by Gerald Borgia) I tested the cognitive performance hypothesis that males with better cognitive performance have higher reproductive success. Male display and female mate choice are highly complex processes in many species of animals and likely requires the need for complex and high-functioning brains. By presenting male satin bowerbirds (Ptilonorhynchus violaceus) problem solving tests, I have been able to show that males who are better problem solvers have higher reproductive success. To develop these problem solving tests, I took advantage of male aversion to red objects on their bowers and created an obstacle to removing the offending red things. You can read the publication that resulted from this work here: [pdf].
I also found that individuals who were on average better at a variety of cognitive tasks, including problem solving ability, were sexually preferred. Measures which integrate information from a large variety of cognitve tasks are strongly correlated with male mating success; in fact these are some of the best predictors of mating success in this species [pdf]. How do females choose these males with better cognitive ability? The answer to this question has turned out to be very complex, with individual display traits not being good indicators, but rather, males who are able to most effectively produce the most complete and complex displays are more likely to have better overall cognitive ability [free-access pdf].
Ecological Consequences of Habitat Modification:
My undergraduate work at William and Mary with my advisor Dan Cristol focused on whether created saltmarshes effectively replace natural saltmarshes that have been destroyed for development. During my first year on this project I was responsible for conducting vegetation transects and showed that created wetlands are primarily monculture stands of Spartina alterniflora and therefore the diversity of plant species is much lower in created wetlands than the natural wetlands they are meant to replace. This work was done as part of David Desrochers' masters thesis where he showed that avian communities differed between the two wetland types as well, with created wetlands having lower abundance and richness of breeding birds. You can read the publication that resulted from this work here: [pdf].
My second year on this project and my honors thesis involved a productivity study of red-winged blackbirds (Agelaius phoeniceus) in these same created and natural saltmarshes. While red-winged blackbirds are not of conservation concern in this habitat, they are ubiquitous and can serve as a model system for studying differences in reproductive output in these two types of wetland. In southeastern Virginia, wetlands are replaced at the same rate as they are destroyed, a requirement of the Clean Water Act. I developed a mathematical model that explored the consequences of replacing habitat of one quality with habitat of a lower quality (e.g. created saltmarshes had lower productivity due to higher egg predation). I showed that if the created habitat is a sink there is a "point of no return" where the population cannot be maintained at the current abundance if more natural habitat is destroyed. Also a one-for-one replacement strategy does not work if the created habitat is not of the same quality as the habitat being destroyed. [pdf]
Marine populations of sticklebacks have invaded multiple post-glacial lakes in British Columbia and independently speciated.
We have a large population of hybrid fish created for a QTL mapping population. These fish are very useful for understanding the phenotype-fitness map.
Experimental design for gene expression study.
Social information use experiment: delayed local enhancement
Images from MRI. (a) Limnetic fish, sagittal section. (b) Benthic fish, sagittal section.
Male solving barrier problem.
Red coverage experiment setup (above) with red tile that has been covered very little (left) versus a red tile that has been covered almost completely (right).
Natural marsh. Notice the variation in topography and the multiple species of plants.
Created marsh. A "wet soccer field".