Bee populations are declining but bees are critically important for pollination. Through resource provisioning, landscape context impacts bees: bees with access to food and habitat are healthier. But landscape context also has epidemiological importance for bees. As a bee forages across landscapes for food, it acquires flower-associated microorganisms. These can be beneficial or pathogenic. Variation in human-managed landscapes may therefore influence bee health. This dissertation addresses whether and how resource availability and landscape composition in urban agricultural systems influence disease dynamics and microbiome composition in three species of domesticated bees with wild counterparts: orchard bees, bumble bees, and honey bees.

The research was conducted in 18-25 urban gardens along the central coast of California. Differences at these gardens in terms of local and landscape features (such crop diversity and natural cover) allowed me to ask how anthropogenic processes such as urbanization impact bee health. In the 1st and 2nd chapter, I examine how garden management influences parasite and pathogen prevalence. In the 3rd chapter, I compare microbiome composition between orchard bees. In the 4th chapter, I use qualitative methods to describe the social factors shaping sustainable beekeeping practices.

I found that floral resources in urban gardens are positively associated with the prevalence of Apicystis and A. borealis in honey bees and with the prevalence of Deformed Wing Virus and Acute Bee Paralysis Virus in Bombus vosnesenskii. While these findings suggest that floral resources in urban contexts may amplify disease risk, I also found that nesting site availability (bare soil) negatively predicts the prevalence of some parasites and pathogens in bumble bees. I suggest more research on the tradeoffs between resource provisioning and parasite and pathogen transmission. Furthermore, while floral resources were associated with disease transmission, they were also associated with the abundance of bacterial groups beneficial to bee health. In Osmia lignaria, floral abundance was correlated with Lactobacillus, which was associated with reduced Crithidia prevalence. These studies highlight complex interactions between environmental context, bee diversity, and bee-associated microbes. I contextualize these findings in a qualitative beekeeping study, suggesting that the unique features of the beekeeping industry can influence the outcomes of resource provisioning on bee health.

Ecological complexity supports biodiversity, robust community structure, and resilient ecosystem functions. In agroecosystems, adding ecological complexity by diversifying crops and agricultural landscapes can help foster complex webs of relationships between pests and their natural enemies that leads to enhanced biological pest control. Urban community gardens provide a unique opportunity to investigate how differences in human management and urban landscapes affect the insect biodiversity, species interactions, and ecosystem services that support food production in urban settings. In my dissertation, I investigate various forms and scales of ecological complexity in community gardens of the California central coast to understand the social and ecological processes that give rise to this complexity as well as its implications for biological pest control in urban agriculture. First, I quantify ecological complexity at the local scale of individual gardener plots by measuring the density, diversity, and connectivity of plot vegetation and its impacts on garden herbivores and natural enemies (Chapter 1). Next, I use ecological networks of interactions between pests and parasitoid wasps to measure changes in pest–parasitoid interactions along gradients of local garden management and landscape composition (Chapter 2). Additionally, I examine the metacommunity structure of herbivores and natural enemies inhabiting urban gardens in the study region and identify garden habitat and landscape characteristics that influence arthropod metacommunity structure (Chapter 3). Finally, I examine how differences in gardener aesthetic norms and management priorities lead to variation in ecological complexity in community gardens (Chapter 4). In line with agroecological theory, this work finds that higher amounts of ecological complexity support greater natural enemy diversity and higher rates of pest parasitism. While common preferences for “tidiness” may limit ecological complexity in some garden plots, gardeners who prioritize “wildness” in their plots present viable alternatives for garden management that help to diversify the ecological habitat and resources provided by community gardens.

Soil microbial communities play critical roles as decomposers, plant pathogens, and plant mutualists in terrestrial ecosystems. Understanding how these communities will respond to a changing environment requires well-replicated studies across diverse ecological contexts and a careful assessment of the methodological dependencies of large-scale microbiological research. In my dissertation, I develop methods for the study of soil microbial communities across three registers: informatic, ecological, and anthropological. First, I present a software package to facilitate access to an open-access, continental-scale dataset of soil microbial DNA metabarcoding sequences for ecological research (Chapter 1). Second, I use this dataset to predict the spatial distributions of fungal taxa across the U.S. and Canada and quantify the sensitivity of soil fungal community composition to ongoing climate change (Chapter 2). Finally, taking an ethnographic approach, I observe how soilborne plant disease comes to be known through a variety of scientific practices that stabilize distinct objects of study (Chapter 3). Together, this work illustrates that while soil microbial communities everywhere may be shifting, these shifts are not the same everywhere. The North American boreal forest occupies a particularly precarious climate in which even slight warming can create major shifts in the composition of its soil fungal communities. Additionally, the enrollment of soil microbial communities in the transition to post-fumigant agriculture contributes to a complex political terrain wherein the possibilities for knowing soil microbial communities are entangled with the possibilities for managing soilborne plant disease. By showcasing diverse methods for studying soil microbial communities, this dissertation puts forth an ontological approach to scientific inquiry that does not presuppose objects as a given but rather enacts them anew through methodological innovations.