You are here

Early Career Scientist Seminars

The J.F. Crow Institute for the Study of Evolution at the University of Wisconsin-Madison is inviting early-career evolutionary biologists from outside UW-Madison to apply to participate in an early-career scientist seminar series in spring 2019. Please come share your science with our community! 

The 3-5 speakers selected for the series will be invited to visit UW-Madison. The speaker will present a 50-minute seminar, ideally aimed at evolutionary biologists with a broad range of backgrounds. The speaker will also participate in a 45-minute discussion after the seminar with undergraduate evolution majors. For the day of the seminar we will schedule meetings with faculty and students working in evolutionary biology. The speaker would be responsible for their own travel to Madison, but would receive a $100 honorarium to offset travel costs. If an overnight stay is required, arrangements could be made to stay with a member of the Crow Institute. If any other special accomodations are required, we are happy to work with speakers to make sure their needs are met. 

Eligibility: Non-UW-Madison graduate students and postdoctoral fellows who received a Ph.D. no longer than 5 years ago. 

If you have any questions please contact Megan Frayer (, Jered Stratton ( or Tiago Ribeiro ( 

The application for a Spring 2019 seminar is now closed. Please check back here for future opportunities. 

Evolutionary dynamics of influenza across spatiotemporal scales

Katherine Xue

Graduate Student

University of Washington (Advisors: Josh Akey and Jesse Bloom)

Seminar Date: January 31, 2019

Influenza viruses evolve rapidly from year to year around the globe, with important consequences for human health. Recent advances in genome sequencing make it possible to track this evolution at higher resolution and to uncover evolutionary dynamics that operate at small scales of space and time. I present two case studies of unusual evolutionary dynamics in influenza. First, I will discuss an example of viral cooperation. RNA viruses like influenza mutate rapidly, generating a cloud of genetically diverse variants that have long been known to facilitate adaptation. Recent studies have suggested that cooperation between variants might also increase population fitness. We characterize two viral variants that differ by a single nonsynonymous mutation and show that mixed populations grow better than either variant alone in cell culture. Next, I will discuss the evolutionary dynamics of influenza within individual patient infections. Viral evolution on a global scale begins with de novo mutations within single infected hosts, but it has only recently become possible to systematically capture within-host genetic diversity. We deep-sequence longitudinal samples of influenza from long-term infections in immunocompromised patients and observe complex evolutionary patterns marked by extensive clonal interference. We find remarkable parallelism in evolution within and between hosts, suggesting concordance in evolutionary dynamics across multiple spatiotemporal scales. I end by discussing how the within-host dynamics of influenza contribute to the virus's global evolution.


Thrive with additional sets of genome: widespread paleopolyploidization buffers plants through Eocene global climatic upheaval

Liming Cai

Graduate Student

Harvard University (Advisor: Charles Davis)  

Seminar Date: February 28, 2019

Ancient whole genome duplications (WGDs) are important in eukaryotic genome evolution, and are especially prominent in plants. Recent genomic studies from large vascular plant clades, including ferns, gymnosperms, and angiosperms suggest that WGDs may represent a crucial mode of speciation. Moreover, numerous WGDs have been dated to events coinciding with major episodes of global and climatic upheaval, including the mass extinction at the KT boundary (~65 Ma) and during more recent intervals of global aridification in the Miocene (~10-5 Ma). These findings have led to the hypothesis that polyploidization may buffer lineages against the negative consequences of such disruptions. My recent work explores WGDs in the largely tropical flowering plant clade Malpighiales using a combination of newly sequenced transcriptomes and complete genomes from 42 species. We conservatively identify 22 ancient WGDs, widely distributed across Malpighiales subclades. Importantly, these events are clustered around the Eocene-Paleocene Transition (~54 Ma), during which time the planet was warmer and wetter than any period in the Cenozoic. These results establish that the Eocene Climate Optimum represents another, previously unrecognized, period of prolific WGDs in plants, and lends support to the hypothesis that polyploidization promotes adaptation and enhances plant survival during major episodes of global change. Malpighiales, in particular, may have been particularly influenced by these events given their predominance in the tropics where Eocene warming likely had profound impacts owing to the relatively tight thermal tolerances of tropical organisms.


Macroevolution, biogeography, and conservation in subterranean ecosystems: A tail of two salamanders

John Gerhardt Phillips


University of Idaho 

Seminar Date: March 28, 2019

Recent studies have begun to shed light on genetic patterns in cave species, which frequently display high levels of endemism, small range sizes, and habitat specialization. The combination of these factors limit dispersal and predispose subterranean organisms to higher rates of extinction. I investigate these ideas by testing how habitat specialization can constrain patterns of dispersal, as some species vary in mobility across ontogeny. Understanding such patterns provides insight into the influence of ontogenetic shifts on patterns of biogeography. Therefore, I use two plethodontid salamanders with different life histories using genome-scale data and a method newly applied to salamanders. I test for molecular variance among populations and associations between genetic distance and geographic features. In one species, divergence time estimates suggest rapid formation of the three major lineages in the Middle Miocene (relatively old for cave-adapted lineages). Gene flow among these populations appears negligible, indicating that most populations across the distribution are effectively isolated from one another. Paleodrainages, contemporary drainages, and geologic division all explain a proportion of genetic variation within one species, but none of these factors appear important in phylogeographic structure in the other (Georgia Blind Salamanders) where most sampled localities are non-monophyletic with a high frequency of haplotype sharing, even among localities in separate drainages. These results may reflect historical fragmentation and secondary admixture within the UFA. These results are important in understanding genetic structure and evolutionary patterns, and also are needed for conservation and management.


Inference and analysis of population-specific recombination maps

Jeffrey P. Spence

Graduate Student

University of California, Berkeley (Advisor: Yun Song) 

Seminar Date: ­­­April 11, 2019

Meiotic recombination—the shuffling of genetic material in sexually reproducing populations—plays an important role in evolution by decoupling the evolutionary fate of physically linked alleles. Recombination is also crucial for proper segregation of chromosomes during meiosis, and changes in the recombination machinery have been suggested as a possible mechanism for speciation.  Meanwhile, the rate of recombination along the genome is highly non-homogeneous: in many species recombination tends to occur in small regions called “hotspots” that correspond to locations in the genome where a protein, PRDM9, binds. PRDM9 is one of the fastest evolving proteins and as such there is rapid turnover in the location of these hotspots.  While the evolutionary dynamics of recombination hotspots have been studied from a theoretical perspective, attempts to link the theory to data have been stymied by the confounding effect of differences in the demographic histories of different populations.  I will present a new method capable of disentangling these demographic effects that I have applied to the 26 diverse human populations in the 1000 Genomes Project.  The resulting population-specific fine-scale recombination maps allow us to directly test theoretical predictions about the rate at which hotspots become cold and explore modulators of fine-scale recombination rate beyond PRDM9 binding.


Evolution of Annualism in Killifishes: An Eco-Evo-Devo Approach

Andrew W. Thompson


Michigan State University (Advisor: Ingo Braasch)

Seminar Date: May 2, 2019

Annual killifishes inhabit seasonal pools that desiccate, resulting in the death of the adult population. Unique adaptations including specialized egg structures, desiccation resistance, and up to three diapause stages slowing developmental and metabolic rates enable the embryonic population to survive annual dry seasons. When the habitat floods, annual killifish terminate their third and final diapause (DIII), hatch, and begin a new lifecycle. Here we explore the genomics of embryonic DIII in annual killifishes. We use scanning electron microscopy, comparative transcriptomics, phylogenetic methods, and model rates of gene evolution to investigate the genetics of killifish annualism, diapause, and environmentally-cued hatching. We discover hundreds of candidate genes involved in diapause and delayed hatching in three killifish species. Specifically, we find 10 differentially expressed killifish transcripts with homologs also differentially expressed in the same direction during dormancy in other animals. These 10 transcripts illustrate the conserved roles of these homologs during delayed development in metazoans. Additionally, tight linkage of diapause and hatching with the expression of a complex family of hatching enzymes leads us to analyze regulatory mechanisms associated with environmentally-cued hatching. Lastly, we show that diapause has up to 7 origins in killifishes and detect over 160 genes with increased rates of molecular evolution in annual compared to non-annual killifishes. Our integrative framework combining development, genomics, evolution, and ecology provides important insights into the mechanisms of diapause and the diversity of vertebrate hatching strategies as well as candidate genes associated with stress tolerance in the face of changing environments.


Riding the wave: testing how rapid evolution shapes the dynamics of range expansions

Christopher Weiss-Lehman

University of Minnesota (Advisor: Allison Shaw) 

Seminar Date: February 22, 2018

Although range expansions have long been modeled as purely ecological phenomena, recent evidence shows that evolutionary changes can shape the expansion process. Multiple possible selection pressures can cause evolution during range expansion, some of them direct outcomes of expansion dynamics. However, neutral processes also can play a significant role in driving evolutionary change during expansion, particularly in the form of gene surfing. The patterns in allele frequency changes produced by adaptive and neutral mechanisms in a single realization of the expansion process and the resulting impacts on expansion dynamics from different mechanisms can be difficult to distinguish. I present results from a highly replicated microcosm experiment with the red flour beetle (Tribolium castaneum) assessing the impact of neutral and adaptive evolutionary changes on range expansion dynamics. Using data from pooled, whole genome sequencing of beetles from the experiment with demographic data on spread rates and trait values, my collaborators and I quantified the relative contributions of adaptive and neutral mechanisms in driving evolutionary changes during expansion. In particular, we demonstrated an important role for gene surfing, as a neutral evolutionary process, in driving increased variance in expansion speeds among replicate populations. I explore this result further by pairing data from the red flour beetle system with a theoretical model exploring the interaction of habitat heterogeneity and rapid evolution in driving variance in expansion patterns. The results I present in this talk demonstrate a clear need to incorporate evolutionary mechanisms when making predictions for range expansions, even over relatively short timescales.


Disentangling the effects of mutation and selection on the evolution of gene expression and regulation

Brian Metzger

University of Chicago (Advisor: Joe Thornton) 

Seminar Date: March 15, 2018

New mutations are the ultimate source of heritable variation. Unfortunately, the action of natural selection alters the types and frequencies of phenotypes observed in natural populations. As a consequence, the relative roles mutation and selection play in shaping patterns of phenotypic variation in natural populations are often difficult to disentangle. This is particularly true for changes in gene expression as the effects of mutations altering gene regulation are difficult to predict. To address this problem, I measured the effects of hundreds of new cis- and trans-regulatory mutations in the absence of natural selection on the expression of the Saccharomyces cerevisiae TDH3 gene. These measurements revealed differences in the effects of cis- and trans-regulatory mutations, indicating that the molecular mechanism by which a mutation alters gene expression plays an important role in the likely phenotypic consequences of that mutation. To determine how natural selection acts on these differences, I compared the effects of these new mutations to the effects of cis- and trans-regulatory polymorphisms segregating amongst over 60 S. cerevisiae strains isolated from a variety of natural environments. This comparison revealed differences in the action of natural selection between cis- and trans-regulatory changes due to differences in the effects of new mutations, including differences in the action of natural selection on the variability in gene expression amongst genetically identical cells, i.e. gene expression noise. These results show how the mutational process helps shape patterns of natural variation by producing biases in the types and frequencies of phenotypic variation.


Revealing rhizobial fitness across symbiotic and free-living environments: identifying fitness tradeoffs and genomic variants

Liana Burghardt

University of Minnesota (Advisor: Peter Tiffin)

Seminar Date: March 22, 2018

When living in symbiosis with legume hosts, rhizobial bacteria convert atmospheric nitrogen into a plant-available form. In exchange, rhizobia receive carbon in the form of dicarboxylates and a favorable environment for reproduction. Rhizobia do not always live in symbiosis and the majority of the population at any given time is free-living in the soil. Little is known about how adaptation to the free-living environment affects the stability of the mutualism or the extent to which host genotype mediates competitive outcomes between closely related microbial strains. Here we present a variant of the ‘evolve and re-sequence’ approach that provides a robust method to estimate bacterial fitness and identify genetic variants responsible for fitness. We show that plant hosts impose strong selection on rhizobial populations, symbiotic fitness and free-living fitness are not correlated, and there is a trade-off between success in soil and in resource-rich liquid media. Selection in hosts caused strong shifts in allele frequency across the genome and these shifts were particularly strong for variants in genes associated with motility, transcriptional master regulators of nitrogen fixation, and genes involved in host/rhizobia signaling. Our work demonstrates the potential power of select and re-sequencing approaches to characterize fitness in microbial species as well as the naturally occurring allelic variants responsible for that variation. In the case of rhizobia, characterizing soil- and host-specific selection and identifying the causative rhizobial genes may allow for manipulation of this symbiosis to increase agricultural yields and provide insight into the biological processes that govern ecosystem productivity.


Variable hybridization outcomes reflect variable reproductive isolation

Elizabeth Mandeville 

University of Wyoming (Advisors: Katie Wagner and Annika Walters) 

Seminar Date: April 12, 2018

Interspecific hybridization can reveal mechanisms of reproductive isolation between related species, and also has long term implications for the evolution of hybridizing taxa. Although hybridizing species are often distributed across heterogeneous environments in the context of variable communities, variation in outcomes of hybridization is still incompletely understood. My recent work quantifies variation in hybridization across replicate instances of hybridization in fish taxa in the mountain west, specifically suckers and trout. Comparison of these two empirical instances of hybridization is informative about how hybridization varies across a large geographic area. My collaborators and I analyzed hybridization among six Catostomus species across the Upper Colorado River basin and found extreme variation in hybridization across locations. Different hybrid crosses were present in different locations, despite similar species assemblages. Within hybrid crosses, hybridization varied from only first generation hybrids to extensive hybridization with backcrossing. Similarly, our study of hybridization between introduced rainbow trout and native Yellowstone cutthroat trout revealed variation in extent of hybridization across 27 replicate tributaries, but patterns of variation were slightly different. Species always hybridized when sympatric, but the degree and direction of backcrossing varied. The next challenge in our work will be identifying precisely why hybridization outcomes vary. Variation might result from uneven fitness of hybrids across locations, polymorphism in genetic incompatibilities, chance, unidentified historical contingencies, or some combination thereof. We are currently investigating these possible mechanisms in both study systems. Our results suggest that one or a few instances of hybridization represent all interactions between the focal species.


Ecology, sexual selection, and phenotypic diversification

Justa Heinen-Kay

University of Minnesota (Advisor: Marlene Zuk)

Seminar Date: April 26, 2018 

Sexual traits are often the most flamboyant aspects of biodiversity and are critically important in generating and maintaining species boundaries. The classic paradigm is that natural and sexual selection act in opposite directions, with sexual selection favoring trait elaboration while natural selection provides a braking force. However, this often fails to reflect what is observed in nature because the evolutionary dynamics of sexual traits are rarely that simple. My research takes a novel, integrative approach to understanding the evolution of sexual traits that involves ecology, social dynamics, trait interactions, physiology, and plasticity. First, I will discuss some of my dissertation work with Bahamian mosquitofish that demonstrates divergence in conspicuous signals, behaviors, and male genitalia driven by population-level differences in predation due to natural variation over 1000’s of years and human impacts on the environment over 10’s of years. Next, I will discuss my postdoctoral research that takes the question of how natural and sexual selection interact a step further to investigate what happens when one evolutionary force overpowers the other. This work centers on Pacific field crickets in Hawaii that rapidly experienced the evolutionary loss of male song due to selection from a deadly parasitoid fly. My work has shown both plastic and pleiotropic influences on female reproductive investment and exploratory behaviors following rapid spread of the silencing mutation. Together, my research sheds light on how the interplay between natural and sexual selection influences diversification, and the causes and consequences of sexual trait evolution in the wild.