CONSEQUENCES OF LANDSCAPE MEDIATED CHANGES IN INSECT INTERACTIONS FOR WILD PLANT REPRODUCTION
The conversion of natural landscapes to agriculture significantly alters the community of both mutualist and antagonist insects persisting in the landscape. Given the importance of insects for food production through both pollination and herbivory, the effects of these landscape mediated changes in the insect community have been predominantly examined within crop systems. However, we currently lack a thorough understanding of the consequences of landscape simplification for wild plants co-existing in heavily modified landscapes. Wild plants in semi-natural habitats play an important role as a resource for pollinators, reservoir for natural enemies, and host for herbivores. Landscape mediated changes in the insect community may impose differential selection on wild plants, potentially creating a feedback loop such that plant adaptation reinforces traits in the insect community and patterns of insect availability. Here, I developed a predictive framework to build hypotheses for insect pollinator, herbivore, and natural enemy responses to landscape simplification and the subsequent effects on wild plant trait evolution (Chapter 1). This literature review highlights the significant role that context-dependence plays in mediating insect responses and the challenge in predicting plant responses to simultaneous shifts in all insect groups which may function as conflicting or reinforcing agents of selection on the same traits. To empirically evaluate how plants are responding to landscape simplification, I collected seeds from three plant species across the landscape gradient and grew them together in a greenhouse common garden (Chapter 2). I evaluated defensive and reproductive traits and found variable responses, with different land cover types best predicting trait variation for each species. For two of the three species, I found a reduction in floral display and herbivore leaf consumption in plants originating from more simplified landscapes. To evaluate if this trait variation is linked to changes in insect interactions associated with landscape simplification, I conducted a field experiment establishing phytometer plants across the landscape gradient and documenting insect herbivore and pollinator interactions (Chapter 3). Here, I found that the landscape indirectly affected plant reproduction through herbivore-mediated changes in floral display. Collectively, this work expands our understanding of wild plant responses to landscape simplification and demonstrates the complex interactions mediating outcomes at a landscape scale.