My lab’s primary research is focused on the process of species divergence, which is characterized by the emergence of discontinuous phenotypic and genetic variation in natural populations. The biology of speciation has been the subject of intense scrutiny and debate since at least the time of Darwin, but many unanswered questions remain. For example, how do traits that contribute to reproductive isolation emerge, and what role does adaptation play in generating these traits (either indirectly or directly)? How do we quantify the strength of reproductive barriers, and are some barriers more important than others? How does the genetic and developmental context of the organism impact the process?
I take an empirical approach to answer these questions, seeking ultimately to: identify the traits involved in divergence, understand the evolutionary processes that lead to changes in these traits, estimate the impact they have on overall reproductive isolation, and characterize their genetic basis and genomic impact. Further, to reveal generalities in the process, I strive to examine repeated instances of divergence in a comparative setting. My primary study organisms are species in the wildflower genus Mimulus. This genus is considered an emerging model organism in evolutionary biology, with spectacular species-level diversity, a rich history of ecological and genetic work, emerging genomic and transgenic resources, and a highly active and supportive community of fellow researchers. Below, I briefly outline the specific research projects I am actively pursuing in this dynamic system.
Habitat choice traits in limestone endemics
At the earliest stages of divergence, adaptation to different ecological conditions is likely to take a central role in initiating the process of speciation. Indeed, habitat isolation is widely accepted as a potentially important barrier, but there have been surprisingly few attempts to measure its impact. A key factor in moving from habitat specialization through local adaptation to speciation may be the emergence of traits that result in habitat choice. When organisms actively choose to reside (and reproduce) in the habitat to which they are best adapted, gene flow can be greatly reduced between taxa choosing alternate habitats.
Mimulus limestone endemics provide a unique opportunity to examine the impact of habitat choice on speciation. The species M. norrisii is a limestone specialist found in a very limited number of outcrops in the southern Sierra Nevada. In comparing M. norrisii to its closest relative (and presumed progenitor) M. floribundus, my lab has characterized two traits that allow the endemic to choose limestone substrates: pH sensitive seed dormancy and reduced dispersability. In the former, M. norrisii seeds will only germinate when they experience the high pH conditions consistent with limestone substrates. In the latter, M. norrisii exhibits a remarkable developmental shift in phototropic activity, shifting stem tissue from positive to negative phototropism as fruits develop. This shift results in the precise delivery of M. norrisii seeds to the habitat from which they came, ensuring that offspring experience the same limestone conditions that the maternal plant did. We are using a combination of QTL mapping and RNA-Seq approaches to identify the genetic mechanisms responsible for these traits. Further, at least four other limestone specialist species in the genus also exhibit habitat choice. We will therefore ask whether this trait evolves through parallelism or convergence by examining the genetic basis of this traits in repeated instances of its emergence.
Seasonal timing traits and life history variation in M. ringens
Clinal variation is often considered a strong indicator of local adaptation, and traits involved in seasonal timing are particularly likely to exhibit geographically structured phenotypes that are the outcomes of recurrent and strong selection. For example, flowering time in plants can vary tremendously across latitude, and the response to seasonal cues must often be finely tuned to ensure growth and reproduction occur under conditions that will maximize fitness.
In M. ringens, my lab has begun work on a latitudinal series of populations ranging from New York state in the north to Georgia in the south. These perennial plants typically require a period of cold each year to stimulate subsequent flowering, i.e. vernalization. In common garden settings, we have found that once vernalized, populations exhibit strong differences in flowering time, with northern populations flowering almost three weeks faster than southern. Surprisingly however, populations from across this range tend to require a similar duration of vernalization to stimulate flowering, despite very different winter conditions. We are therefore currently interested in determining whether southern populations respond to additional cues that allow them to circumvent vernalization requirements in the event of a warm winter.
Further, in examining vernalization requirements, we have discovered several populations that do not require vernalization at all. These populations were collected from habitats where seasonal water availability becomes very low, and they appear to represent a shift from a perennial to a more annual-like life history. Using a population genomic approach, we have found that these ‘annuals’ are very highly differentiated from the more common perennials, even for populations in very close geographic proximity. These results suggest the discovery of a cryptic species, and we are therefore interested in determining which reproductive barriers are preventing gene flow between life history types in this system. We have recently collected a much larger set of populations, with substantial numbers of both ‘annuals’ and perennials, and we aim to understand how discontinuous variation in life history arises despite presumably continuous variation in ecological conditions.
Sources of genomic variation during divergence in M. aurantiacus
In collaborative work with Matt Streisfeld at the University of Oregon, my lab has been involved in examining the divergence process in two ecotypes of M. aurantiacus. This species exhibits two discrete floral morphs, a red-flowered ecotype that occurs in coastal regions of southern California, and a yellow-flowered ecotype that occurs further inland. These two ecotypes are primarily isolated by pollinators, with hummingbirds serving as the primary pollinator for the red ecotype, and hawkmoths primarily visiting the yellow ecotype. However, pollinator fidelity is not complete, and a narrow hybrid zone is formed where their ranges meet. These ecotypes experience only weak genetic incompatibilities, but habitat conditions and ecophysiological traits vary considerably between red and yellow populations. We are therefore interested in how habitat and pollinator isolation interact to produce nearly complete ecological isolation in this system. Further, genomic variation between red and yellow ecotypes is highly heterogeneous, and recent work has shown a surprising correlation between genomic features and divergence in several pairs of closely related taxa in this group. We are therefore planning to continue this collaboration by developing a project that will use field experiments and experimental hybridizations across a wider range of taxa to test for the relative impacts of ecologically based divergent selection, genetic incompatibilities, and background selection in generating genomic differentiation.
The emergence of hybrid seed lethality across the genus Mimulus
Ecologically-based divergence in isolating traits may be important in driving the early stages of speciation in many cases. However, post-mating genetic incompatibilities are also thought to be crucial because these barriers may help complete the process, resulting in a permanent elimination of gene flow between groups. In preliminary work, I have found that several closely related species pairs within the genus Mimulus exhibit strong crossing barriers, often in the form of hybrid seed dysfunction. In a recently funded collaboration with John Willis at Duke University, Andrea Sweigart at University of Georgia, and Bob Franks at North Carolina State University, we aim to expand upon these results by investigating the frequency, strength, and relative importance of hybrid seed lethality across the entire genus. We will use crossing experiments for multiple pairs of taxa coupled with transcriptome-wide measures of genetic distance to determine the rate at which hybrid seed lethality evolves, and compare this to other forms of reproductive isolation such as male sterility. In pairs of species that produce hybrid seed lethality, we will employ genetic mapping studies and characterize the developmental mechanisms responsible using a range of approaches. By examining the genetic basis of hybrid lethality in multiple pairs of species, we aim to determine if predictable patterns emerge that will reveal general features of the speciation process in this group of plants. Ultimately, especially in taxa where ecological barriers have been well-studied, we expect to provide some of the most complete mechanistic pictures of the speciation process to date.