Eco-Evolutionary Ramifications of Herbicide-Driven Plant Modifications on Plant-Herbivore Interactions in Modern Agricultural Landscapes

Natural populations evolve in response to biotic and abiotic changes in their environment, which lead to wide variation in defense traits that can shape species interactions. Often, variation in defense is a result of complementary or conflicting strategies for dealing with external stressors. Agricultural systems can introduce novel stressors via herbicide exposure to non-crop plant systems surrounding fields. Of the various plant defense strategies, trichomes, or hair-like appendages on the outermost boundary, are known to play a critical role in how these sessile organisms interact with their environment. While herbicide drift is known to produce a variety of toxic effects in plants, little is known about its impact on plant-herbivore interactions and the evolution of plant defense strategies. Thus, my dissertation addresses the overarching question: Does herbicide drift impact plant-herbivore interactions and the evolution of plant trichomes? Using the annual invasive velvetleaf (Abutlion theophrasti) as the focal species, I conducted various field, greenhouse, and growth room experiments to explore the potential effects of herbicide drift. In the first data chapter (Chapter 2) of this dissertation, I investigate the impact of dicamba, a novel herbicide known to travel particularly far distances, on plant-herbivore interactions. The findings from this multi-year study revealed a significant increase in the phloem feeding silverleaf whitefly (Bermisia tabaci) abundance on plants exposed to herbicide. Here, I identified a significant phenotypic tradeoff between whitefly resistance and herbicide resistance. I also found evidence that dicamba drift significantly increased leaf chlorophyll content (mg/cm2) which positively affected whitefly abundance. In the second (Chapter 3) and third (Chapter 4) data chapters of this dissertation, I assessed whether glyphosate, a traditional herbicide found in “Roundup”, acts as an agent of selection on induced trichome traits (polymorphs and density) and whether variation observed in trichome traits represents dual or conflicting roles against herbicide and herbivory damage. I identified positive correlations between induced total trichome production and herbicide resistance as well as induced branched trichomes and herbicide resistance. The selection analysis revealed positive linear selection acting on induced trichome production and correlative selection favoring high induced trichome production and intermediate plant growth. In the third data chapter, I found that having a high proportion of branched trichomes not only had a positive effect on herbicide resistance, but it also had a positive effect on chewing herbivory resistance. In brief, my dissertation demonstrates that herbicide exposure can alter plant-herbivore interactions mediated by modified plant phenotypes. My thesis demonstrates that herbicide drift can create unintended reservoirs for agricultural pests like phloem feeding whiteflies that are attracted to chlorophyll content. It also provides novel evidence that herbicide is a selective agent on induced trichome production and demonstrates that branched trichomes can contribute synergistically to herbicide and chewing herbivory resistance, serving as a dual structural form of resistance reducing plant injury. While responses are expected to be species and herbicide dependent, collectively, this work reveals that herbicide exposure to non-target plant systems can significantly alter species interactions and community dynamics in habitats located at the agro-ecological interface.PhDEcology and Evolutionary BiologyUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/177892/1/nmjo_1.pd