General Research Interests
Adaptation, conservation, genomics, climate change, modeling, bioinformatics
Identifying adaptive variation and the drivers of population differences will aid in detecting the sensitivity, resiliency, or vulnerability of natural populations, as well as improve our understanding of how species will respond to environmental change. I am interested in the dynamics of hierarchically structured populations exhibiting extensive gene flow and the inferences available regarding evolutionary (adaptive and demographic) processes from genetic data. I am particularly interested in uncovering and describing the underlying genetic componants (i.e., the architecture) of traits important to adaptation, and using this knowledge to advance conservation, reforestation, and breeding. How do evolutionary processes influence the genetic and genomic architectures of complex traits? How does the genetic and genomic architecture influence resilience to change? How do these patterns and processes differ across closely related species/genera, or even across a species' range? How do adaptive responses within species influence population, community, and ecosystem processes? How do adaptive traits impact organization within ecological communities? What are the implications of these findings to conservation, industry, the economy, our overall understanding of evolution across taxa, and the future of populations in the face of climate change?
Current research at Northeastern
Methods using genomic information to forecast population maladaptation to climate change (i.e., genetic offset methods) are becoming increasingly common, yet there has been little exploration into the factors that could potentially affect performance of these methods, or mislead inference with regard to model accuracy during validation. These knowledge gaps pose serious hurdles toward the incorporation of such methods into management and policy. Dr. Lotterhos and I are using extensive sets of simulations to understand the circumstances in which these methods will be most useful. We are also extending machine learning algorithms for restoration purposes to identify the best climates for a given genotype.
Previous research
How useful is genomic data for predicting maladaptation to future climate? PDF
The pace of climate change threatens to discouple adaptive variation and environmental optima of many natural populations. Characterising the vulnerability of populations across a species range can be particularly useful to direct management priorities. Over the past decade, using genomic data to predict maladaptation of natural populations to future climate (i.e., genomic offset methods) have become increasingly widespread but with few instances where model predictions are validated with ground truth data. We use range-wide data across three conifer taxa to assess such methods. We validate these methods against two-year and 52-year growth and mortality measured in independent transplant experiments. Overall, we find that two common methods often better predict transplant performance than climatic or geographic distances. We also find that these models are surprisingly not improved using GEA candidates. Even with promising validation results, variation in model projections to future climates can make it difficult to identify the most maladapted populations using either method.
Evaluating genomic data for management of local adaptation in a changing climate PDF
Understanding how managed populations are adapted to their local environments will assist in both breeding and reforestation efforts by identifying environmental sources driving trait differentiation and pathogen susceptibility. To feasibly (weighing time, effort, and cost) disect these aspects of plant biology, there are a few common approaches that can be considered, but do they hold the same information content with regard to the environmental drivers of differentiation or the spatial scale of adaptation? To address this, we compared climate data, genomic data, and a long term provinance trial in lodgepole pine in this regard. We discuss relative strengths and weaknesses when considering the appropriate data for managing populations with broad geographic ranges. (see figure to the left showing range-wide clines in genomic regions underlying adaptation)
Evolutionary genomics of the invasive and defoliating gypsy moth PDF
Biological invasions from non-native species can have have large economic consequences and can fundamentally alter the ecology of the invaded system. As one such example, the European gypsy moth was introduced to the eastern United States around the late 19th century and has since spread north through Maine, west through Wisconsin, and south through the Carolinas. Consequently, the gypsy moth has become one of the most destructive hardwood pests in North America. To understand how this species has been so successful in a novel environment we sampled populations from the range edges by examining genomic variation underlying trait differentiation for three developmental traits. We described how the invasion was facilitated through adaptive variation found throughout the genome and that genomic patterns (ie linkage disequilibrium) among adaptive genetic componants were driving differences in the three traits among populations. (see map to the left showing the chronological spread of the gypsy moth across New England)
Effects of Forest Management on Fine-scale Gene Flow in a Fire-suppressed population of Pinus lambertiana Dougl. PDF
Effect of fire supression on forest crowding in the upper Yosemite Valley from Columbia Point (source unknown)
Climate change is exacerbating the effects of fire suppression in the west, and prescribed fire and forest thinning are two ways of mitigating these effects. Sugar pine is an important species in these forests, and was historically dominant before fire suppression. We looked at how these treatments affect mating patterns in this species and discuss how this may affect long-term forest dynamics. (see 3D stem map to the left; more about Teakettle Experimental Forest here).
Local adaptation of Pinus albicaulis Engelm. across fine spatial scales of the Lake Tahoe Basin - PDF
Whitebark pine is a foundational species in many high elevation forests and plays an important role in ecosystem function and services. Despite its influential role in these systems, whitebark pine is listed as a threatened species by IUCN. Therefore, understanding the spatial scale of adaptation in this species is vital for conservation efforts. Even so, much of our knowledge regarding adaptation in trees comes from broad geographic ranges, often beyond the scale at which most forest management is applied. Using a common garden we found that whitebark pine is adapted across the fine spatial scales of the Lake Tahoe Basin despite highly shared genetic variation across populations. We further showed that the adaptive signal driven by water availability is apparent across common garden and genomicw data, likely due to the rain shadow on the eastern Siera Nevada. (to the left see map of sampled populations from the Lake Tahoe Basin, and how quickly the rainshadow reduces annual preciptation by 50% in just under 15 miles)
Linkage maps and the genomic architecture of local adaptation
For non-model organisms, genomic resources needed to fully describe the genetic and genomic architectures of complex traits can face many hurdles. Because of their large genome size, conifers face many challenges in this regard, particularly with respect to the production of physical sequence maps with ordered genetic markers. Such resources can be used to test fundamental theory regarding the expected genetic architecture underlying adaptive traits under various demographic and evolutionary scenarios. To produce such resources, linkage maps are often used to both order and elongate blocks of sequence data. Across taxa, such maps were predominately made from well understood familial relationships or mating designs, but maps of single trees began increasing in numbers in the 1990s. In conifers, maternally derived haploid tissue (i.e., a single copy of the genome from megagametophye tissue) can be used to order markers within individual trees within a fairly straightforward framework. To begin to test fundamental theories of local adaptation in a non-model conifer, Pinus balfouriana, we constructed a dense linkage map using individuals collected from populations that historically exhibited gene flow but have subsequently become isolated from one another since the last glacial maximum. PDF
After completing this linkage map, we used niche modeling and indirect measures of WUE, the ratio of carbon isotopes fixed during photosynthesis (δ13C), as well as nutrient utilization, the ratio of nitrogen isotopes fixed during leaf-level resource utilization (μgN), to explore the importance of water availability as a determinant of geographical range of foxtail pine. Additionally, we explored the degree to which variation in δ13C and μgN is genetically based, differentiated among populations, and underlain by a genetic architecture with large effect loci. PDF
