The International Research Fellowship Program enables U.S. scientists and engineers to conduct nine to twenty-four months of research abroad. The program's awards provide opportunities for joint research, and the use of unique or complementary facilities, expertise and experimental conditions abroad.
This award will support a twenty-four-month research fellowship by Dr. Sean D. Schoville to work with Dr. Olivier Francois at Grenoble Institute of Technology in France.
Glacial cycles have caused rapid, regional reorganization of alpine ecosystems at repeated intervals in the last 2.7 million years. These ecosystems provide a naturally replicated model system to study how recent spatial and temporal environmental change can influence processes of species formation and community assembly. At present, the utilization of genetic data to make inferences about the spatial and temporal history of populations is largely restricted to a limited set of analytical models. Approximate Bayesian computation (ABC) is a methodological approach that can be used to expand the set of population models considered and provide quantitative criteria to statistically discriminate between these alternative models. This research project utilizes ABC methodology to examine patterns of demographic change and the timing of divergence in alpine populations, across many co-occurring species in three geographically independent alpine ecosystems. This comparison examines whether spatial, temporal and demographic change in alpine species follows predicable patterns across the Northern Hemisphere.
Training objectives of this research include statistical analysis of population genetic data, bioinformatics, and computational biology. On the applied side, this research provides an important advance in the field of biogeography by implementing objective, statistical criteria when discriminating between alternative evolutionary hypotheses and allowing broad scale comparisons to be built on quantitative methods. Climate change and landscape modification are the primary environmental agents that are shared in the evolutionary history of alpine species. Demonstrating the importance of these environmental factors in shaping the evolution of co-distributed species will have implications for the effects of ongoing climate changes and the strategies needed to manage threatened alpine species both for short-term ecological viability and long-term evolutionary potential. This research develops international collaborative projects on alpine biodiversity across the Northern Hemisphere, focusing on evolutionary history, ecosystem organization and conservation.
The objective of our project was to develop novel strategies for inferring the evolutionary processes causing genetic differences among populations. The intellectual merits of this project include the integration of statistical methods and computational tools to develop bioinformatics approaches that are broadly applicable to plant and animal populations. One primary goal was to develop and utilize statistical modeling approaches to infer changes in past population size in alpine species. Alpine species are often considered at risk, due to their small geographic ranges and sensitivity to environmental change. I used Bayesian modeling approaches to examine the population history of several alpine species, including the butterfly Colias behrii and the salamander Hydromantes platycephalus, both rare species found in the Sierra Nevada Mountains of California. We found that population size changed dramatically in response to past climate change, where the butterflies in particular had evidence of population collapse during a cold climatic period ~500 years ago. This loss of genetic diversity leaves these butterflies at risk of inbreeding and limits their ability to adapt to ongoing environmental changes. A second goal in our project was to develop statistical modeling approaches to detect patterns of genetic adaptation to the environment. Drawing on statistical approaches developed in machine learning, we designed a powerful new algorithm, as well as software (LFMM), to infer genetic adaptation from genomic data. Additionally, this led to a related statistical method and software (SPFA) that identifies distinct populations in species inhabiting a landscape, after adjusting for the correlations that arise in spatial samples of genetic variation. These approaches have been used in two applied projects, one involving a butterfly population inhabiting an urban-rural gradient in Marseille, France, and another focused on alpine plants inhabiting the European Alps. Finally, this research led to a review of recent developments in detecting adaptive variation in plant and animal populations. These project outcomes not only advance knowledge within the field of evolutionary biology, but also create linkages with other research fields, including statistics and computational biology. The broader impacts of this project include the development of international partnerships with prominent statisticians in France, as well as collaborations with international working groups studying evolutionary biology, landscape genetics, climate change and genetic adaptation. Additionally, this research impacted graduate student training in France through collaborative projects with three PhD students, and benefitted high school students during guest lectures at the Urban school in San Francisco and Beijing High School in China.
