To understand the dynamic interplay between the evolution and demography of rare and invasive species, my research examines two complementary themes: how evolution and genetic variation shape demographic patterns, and in turn, how ecological interactions, such as competition, climatic stress, and pollination, drive evolution within populations.

From an evolutionary perspective, I am investigating two related problems. First, the wide diversity of chemical and morphological defenses found in plants are believed to be adaptations against herbivores, but despite numerous studies that document natural selection by herbivores on plants, and macroevolutionary comparative evidence consistent with a history of coevolution between plants and herbivores, there is little convincing evidence that selection by herbivores actually results in an evolutionary response of defensive traits within plant populations. In collaboration with colleagues from several institutions I am conducting a long-term selection experiment in the field, to test whether selection by a community of over 60 herbivores species leads to predictable changes in the frequency of clonal genotypes within populations, where genotypes differ in chemical and morphological plant traits related to resistance against herbivores. This work integrates ecological, molecular, and phytochemical techniques (read more).

First, gene flow, between crops and their wild relatives has the potential to alter the evolutionary trajectory of weed populations. Our research in Michigan and California examines the ecological contexts under which hybrid genotypes succeed and the evolutionary consequences of this hybridization for weed life history in annual weeds, Raphanus raphanistrum and R. sativus. We have shown that the success of annual weeds may be stimulated by hybridization. Within this framework, we have published several noteworthy results. Advanced-generation hybrids are more fecund than weedy relatives in novel settings and introgression rates are environmentally dependent. (Campbell et al., 2006). Competition enhances hybrid fecundity relative to wild parents, reduces life history differences, and promotes crop-allele introgression (Campbell & Snow, 2007). Crop ferality is unlikely to evolve without the aid of gene flow from wild, weedy relatives in Raphanus sativus (Snow & Campbell, 2005, submitted). We are now beginning a long term research program determine the relationship between individual fecundity or weediness and population growth (λ) or invasiveness. Recent literature on hybridization has suggested that the process may lead to increased invasiveness. However, this has yet to be empirically tested and our work will provide the framework for consequent studies linking the success of individual, weedy hybrid plants to hybrid population dynamics. Further, we will be making several key contributions to theoretical plant demography. For instance, we will be comparing the success of matrix modeling approaches with individual based modelling approaches in accurately predicting population dynamics. Also, we will test the validity of elasticity analysis predictions with manipulative experiments that alter key vital rates. Finally, we will experimentally explore the impact of variation in environmental conditions and genotype on population growth rates

The second problem I am addressing focuses on the fact that a paucity of data hinders managers’ abilities to set scientifically defensible recovery goals and criteria for all but a few species. Therefore, recovery plans commonly recommend a laundry list of research needs and recovery actions that differ little among species. Such blanket recommendations hinder prioritizing efforts for individual species and make prioritizing among species nearly impossible. Yet, we know a lot more than we realize about at-risk species as a group. We have created a broadly applicable eco-informatic approach to recovery planning of endangered species (plants, animals, fish, invertebrates). The comparative, eco-informatic approaches we will develop, which are broadly relevant to both plants and animals, is intended to be implemented for all species at risk in the USA. We are specifically interested in the genetic risks associated with small population size, declining population size, loss of populations, increased isolation and hybridization frequency. Within this framework, a sample of our noteworthy discoveries that we are currently writing up for publication include the following: US federal recovery plans recommended genetic data collection for taxa with lower abundances, and declining population sizes, population numbers, and ranges (P<0.001). In contrast, research published by academic researchers focused on listed taxa with larger population sizes (P<0.001) and did not prioritize taxa with declining number of populations (P=0.31) or range size (P=0.41). Therefore, research priorities of agency personnel writing recovery plans more objectively reflected the relative extinction risk of taxa with small and declining populations than did the publication record of conservation biologists. Finally, although habitat fragmentation affects population differentiation in animal populations in a clear and theoretically predicted manner, this is not the case for plant populations. Instead, plant population differentiation is not predictable based on scientifically interpreted habitat fragmentation patterns. We will be developing new definitions of habitat fragmentation in order to better identify genetic risks associated with plant populations.

From an ecological perspective, I am trying to understand the ecological, demographic, and evolutionary consequences of global change (e.g., climate change, habitat fragmentation) on functional genetic diversity (e.g., controlling flowering phenology and mate compatibility) within plant populations and for plant mating systems.  This work is aimed at estimating the consequences of diversity (genetic, phenotypic and individual mating strategies) on plant population dynamics and ultimately the maintenance of species diversity.  In part, it involves ecological synthesis of existing datasets. Accompanying these eco-informatic queries, I will also employ empirical, experimental manipulations of diversity in both greenhouse and field manipulations of drought- and heat-stress in order to test the predictions created by the above eco-informatic analyses. These and other planned projects will help improve our understanding the biodiversity consequences of global change.