The size, connectivity, and quality of habitat patches can have multifaceted impacts on species and communities. In this dissertation, I combined a multi-year field survey, manipulative field experiments, and a literature review to investigate how spatial structure influences species and their trophic interactions in fragmented habitats. For all empirical work, I used as a study system the arthropod assemblage found on patches of the salt marsh grass Spartina patens.

In Chapter 1, I conducted seven surveys of habitat patches over three years to examine the effects of patch size, connectivity, and local environmental conditions on a guild of specialist sap-feeding herbivorous insects and their natural enemies. I found striking differences among species in the effects of both patch size and connectivity, which led to differences in species' relative abundances and trophic structure among these patches. In Chapter 2, I manipulated host plant quality and predator density to experimentally examine mechanisms that might structure this arthropod community. I found that positive responses of herbivores to experimentally-elevated patch quality were limited by dispersal constraints and that predation by abundant generalist spiders may constrain the spatial distribution of certain species.

Investigating systems beyond the marsh, I conducted a literature review and analysis in Chapter 3 wherein I examined whether the spatial structure of habitats generally influences trophic interactions. From the literature, I identified 171 studies of trophic interactions in fragmented habitats and found that the influence of fragmentation and related variables on the occurrence or strength of trophic interactions was largely predictable based on the habitat affinity of interacting species. With this dataset, I also identified key gaps in the fragmentation literature, including a heavy bias towards the study of two-species interactions. Therefore, in Chapter 4 I took advantage of my data from the salt marsh to identify how, in addition to the two-species interactions of parasitism and egg predation, more complex food web interactions might depend on variation in the size of habitat patches.

Overall, my findings show that variation in patch size can have varied, but predictable, effects on patch occupancy, population density, and interactions between species in fragmented habitats.