Burrowing animals, such as earthworms, crabs, prairie dogs, ground squirrels, ghost shrimps, and rodents, are often considered ecosystem engineers because of their influence on plant communities. For example, in coastal salt marshes, burrowing crabs can influence plant communities by modifying plant zonation, plant production, plant colonization, nutrient cycling, and erosion. Their effects on vegetation can vary across spatial and temporal gradients. In particular, the direction of burrower impacts (positive or negative) can depend on variation in the surrounding animal and plant community, as well as external abiotic conditions. Given that global climate change can alter such environmental conditions, such as salinity and nutrient availability, there is a need to predict when, where, and how burrowers will influence vegetation. Unfortunately, this ability is impeded because few studies have used comparative-experimental approaches to examine animal impacts on plants across space and time. In this dissertation, I explore the effects of burrowing crabs on California salt marsh plant communities using field manipulations and laboratory feeding assays. In my first chapter, I conducted a multi-site, multi-year field manipulation to examine burrowing crab impacts on plant communities in southern California salt marshes. I focused on the effects of burrowing crabs (Pachygrapsus crassipes [lined shore crab] and Uca crenulata [fiddler crab]) on the abundance of the two dominant marsh plants (Spartina foliosa [cordgrass] and Sarcocornia pacifica [pickleweed]), and I explored mechanisms underlying these effects by monitoring plant characteristics and sediment biogeochemistry. Crab impacts on plant community structure differed between each of our three sites. In contrast to our predictions, 1) plant-grazing crabs (lined shore crabs) had positive effects on cordgrass cover at one site and no effect on cordgrass production at a nearby site in the same marsh (Kendall-Frost marsh), and 2) detritivorous crabs (fiddler crabs) did not stimulate cordgrass production at another marsh (San Dieguito Lagoon). In fact, burrowing crabs suppressed cordgrass abundance at San Dieguito Lagoon, the site with the greatest detritivorous crab density (~10x the density of all other sites). Because crabs affected assemblage characteristics of cordgrass in the direction consistent with changes in cordgrass cover, we propose that marsh-specific crab effects on community structure were largely mediated through changes in cordgrass, as opposed to pickleweed. Importantly, crabs facilitated cordgrass during marsh-wide cordgrass loss, suggesting that crabs may mitigate environmental stress for this ecologically important plant. Cordgrass abundance can be a critical measure of marsh functioning and is often a restoration target, and thus maintaining healthy salt marsh functions should require monitoring and management of crab and cordgrass populations. In my second chapter, I combined these results with a similar field manipulation (e.g. multi-site and multi-year) in northern California to create a framework that could more broadly predict burrowing crab effects. While many studies have addressed burrowing crab impacts, few have sought to observe and predict these effects across multiple sites, multiple years, or both. This is especially true for salt marshes along the Pacific coast, where burrowing crab effects on plants have gone untested. In conjunction with our field experiments, we estimated total consumption of marsh plants by the dominant, herbivorous burrowing crab (the lined shore crab) by conducting laboratory feeding assays. Then, we used statistical models to predict crab effects using factors related to the crab community and soil conditions. By combining field, laboratory, and statistical modeling, my comparative-experimental approach allowed me to examine crab impacts across all site-year combinations. Crab effects varied from strongly positive to strongly negative, and depended upon our estimate of the total consumption pressure exerted by crabs an environmental conditions (i.e. salinity and ammonium). Crabs facilitated cordgrass at low total consumption pressure, extreme salinities, and intermediate levels of ammonium. Additionally, my models provided estimates of the threshold values of these environmental factors where the magnitude and direction of crab effects changed. Moving forward, we must seek to mechanistically understand how these key factors (grazing, salinity, and ammonium) drive inter- site variation in crab effects on cordgrass— as such information may be critical to the restoration and management of Pacific coast salt marshes. In my third chapter, I focused on the trophic mechanism by which burrowing crabs influence plants. Herbivores can have important impacts on plant communities, and palatability is among the important factors influencing herbivore consumption and herbivore impacts on plant communities. Here, I assessed the relative palatability of dominant marsh plants among three northern California salt marsh sites— all within one degree of latitude of each other. Although biogeographic approaches reveal that plant palatability to herbivores can vary across broad geographic scales, less is known about how the relative palatability of multiple plant species can vary across small scales. Such variation could be common given the species-specific responses of plants and herbivores to environmental conditions. To address this gap, I conducted multi-choice feeding assays with the lined shore crab - a consumer that has access to the leaves and roots of both cordgrass and pickleweed plants. I assessed the influence of plant species and tissue-types (roots and leaves) on crab feeding preference and the mechanisms underlying them. Surprisingly, the relative palatability of cordgrass and pickleweed switched between marshes within the study region. This shift may have been related to an increase in the palatability of pickleweed leaves at one of our sites. Because cordgrass palatability did not differ among these sites, the change in relative palatability did not appear related to changes in cordgrass. These patterns may be related to nutrient availability at our sites because plants at sites with high pickleweed palatability had lower C:N ratios than the other sites. However, the shift in the palatability of these two dominant plants does not appear to drive shifts in crab impacts on northern salt marsh plant communities, suggesting that crab impacts on plant communities is multi-faceted. The impact of the burrowing activity by crabs on plants may outweigh the consumptive effects of crabs on plants. We encourage future studies examining plant palatability to consider within-region variability in order to understand how such small-scale differences in plant palatability can alter local community structure and ecosystem function. Salt marshes filter water, buffer coastlines, bury atmospheric carbon, protect critical fisheries, and provide habitat for endangered and endemic species. These salt marshes, and the services they provide, are being critically impacted by human development and face a myriad of additional threats due to anthropogenic climate change. However, it will be impossible to predict how these threats will impact marshes if we lack an understanding of the basic species interactions that control the abundance of the foundation species (e.g. cordgrass) within these critical ecosystems. My dissertation highlights that burrowing animals can strongly control plant community composition and the production of cordgrass, and that these plant-animal interactions are highly variable across space and time. By taking a comparative-experimental approach, I was able to delve into this variation to identify possible biotic and abiotic drivers (e.g. grazing rates and soil conditions) that may help predict the magnitude and direction of burrowing crab effects on cordgrass in California marshes. Additionally, recognizing how these drivers may change with environmental stress, especially the effects of severe climate events (e.g. drought, storms, and heat spells), may help develop adaptive management strategies to buffer salt marshes from climate change— preserving the critical ecosystems services which they provide.