The ecologist in me has found it difficult to ignore the rich and intimate associations plants have with insects.These relationships are multifaceted, ranging from interdependent mutualisms to life-threatening antagonisms.The importance of one pollinator species to a plant, for example, can depend on the other pollinators present.In Asclepias incarnata, the worst pollinator with respect to wasting removed pollen can be the most important pollinator when it is abundant (Ivey et al., 2003).Such variability among pollinators in their effectiveness may be important for maintaining species boundaries among simultaneously flowering congeners, and I am currently exploring this question in milkweed communities.
    Variation among plants is also important to plant-insect interactions, and inbreeding, which is routine in nature, is one of the most important sources of genetic variation in plants.My work has shown that inbreeding in M. guttatus can change interactions with pollinators (Ivey and Carr, 2005) as well as herbivores (Ivey et al., 2004).These were among the first studies to demonstrate that inbreeding can have ecological consequences, and that these can extend beyond the inbred population.I also found wide variation among individual plants in the effects of inbreeding on interactions with insects (Ivey et al., 2004), which indicates that inbreeding may create opportunities for natural selection to act on traits associated with these interactions.These results also demonstrate that nonadditive genetic variation can be important for plant-insect interactions, and I am further exploring the genetic architecture of interactions between Mimulus guttatus and its most common herbivore in a diallel crossing experiment fully replicated across two herbivore environments.The results from this study will also provide insight into tradeoffs and genetic constraints associated with the evolution of plant defense.

    Perhaps the most astonishingly diverse feature of plants is their mating systems.This is particularly true for features that influence inbreeding.For example, traditional breeding studies (Wyatt et al., 1996; Wyatt et al., 1998) placed milkweed species into two categories: either self-infertile or fully self-fertile.My research with Asclepias incarnata, however, showed that this characterization was an oversimplification, and that individual plants and populations can vary in their expression of self-fertility (Ivey et al., 1999).More broadly, selfing rates (proportion of self-fertilized offspring) of many plant species are commonly observed to be intermediate, even though most models of mating system evolution predict only the extremes to be stable.I have argued that the importance of ecological interactions for explaining this discrepancy may not be fully appreciated (Ivey and Carr, 2005).Demonstrating that ecological interactions affect mating system variation in nature, however, can be challenging; for example, I found that the most widely used methods for estimating selfing rates of plants in nature are fraught with substantial imprecision (Ivey and Wyatt, 1999).Experimental approaches are likely to be more satisfying.
Although interactions between plants and herbivores are ubiquitous and among the most influential ecological interactions faced by plants, their effects on mating systems have scarcely been explored.My research, however, showed that herbivores alter the expression of inbreeding depression in plants in nature (Ivey et al., 2004), which provided compelling evidence that herbivory may affect mating system evolution.Inbreeding depression is a primary factor affecting the evolution of self-fertility.Furthermore, my experimental field studies showed that herbivores can change plant selfing rates (Ivey and Carr, 2005).Ecological interactions with herbivores may thereby contribute to maintenance of mating system variation.My ongoing research in this area involves demographic modeling, and preliminary results suggest that any environmental factor (such as herbivory) affecting selfing rates can maintain variation in selfing rates across a range of inbreeding depression levels.A second, ongoing collaborative study will discriminate among several hypotheses to identify mechanisms of herbivore impacts on selfing rates.These studies will provide new insight into the role of ecological interactions for mating system variation and evolution. 

    My undergraduate experience of producing formal floristic descriptions of Costa Rican palms taught me that biological insight can emerge from careful study of morphological variation.My later studies of the sources and consequences of morphological variation reflect this perspective.My research in Florida, for example, showed that the wide morphological variation of dominant Everglades macrophytes (Sagittaria lancifolia and Cladium jamaicense) could be explained by the structure and nutrient status of soils (Tarpey and Ivey, in press; Ivey et al., in prep).These results suggested that patterns of dominance, and even inferences about subspecific taxonomy, may reflect environmental variation.Further experiments revealed that S. lancifolia leaf morphology responded strongly to phosphorus treatments and thus can indicate phosphorus availability in this oligotrophic ecosystem (Richards and Ivey, 2004).Another project of mine provided an intuitive and novel response to the perennial question of the function of “drip-tips” in the humid tropics: they may be particularly important for reducing fungal pathogens (Ivey and De Silva, 2001).Elsewhere, I demonstrated that xylem-feeding herbivores could alter the morphology of Mimulus guttatus flowers, and that this change could affect pollinator behavior and traits associated with the mating system (Ivey and Carr 2005).Current projects are examining mechanisms involved in herbivore-induced changes in floral morphology and their consequences for the reproductive biology of M. guttatus.

One of the more frustrating challenges for evolutionary ecologists is the limited glimpse we have into the biological history of a population.As molecular techniques have become routinely accessible, however, our perspective has broadened considerably.For example, the dominant ecosystem component of the Florida Everglades, Cladium jamaicense, assumes broad monospecific stands in many areas, which inspired historical assumptions that these stands represent single asexually-reproduced genotypes.My work, however, demonstrated that genotypic diversity is common, even at small spatial scales, despite extensive clonal reproduction (Ivey and Richards, 2001B).I also showed that the extent of clonal reproduction impacts population genetic differentiation (Ivey and Richards, 2001A).I am currently compiling a review of molecular studies of clonal reproduction in plants to examine the impact of sampling methods and plant characteristics on estimates of asexual reproduction.In addition, I am working on a collaborative study of the population genetics of the widespread Amorpha fruitcosa and its narrowly restricted congener, A. nitens, which will guide taxonomic and management decisions.