- The size, shape, and stability of a species’ dietary niche can both influence and reflect a variety of biological patterns, including species interactions, extinction risk, and ecosystem function. This is particularly apparent when dietary changes manifest at ecosystem and clade scales to profoundly affect macroecological and macroevolutionary trajectories. However, many studies exploring interactions, extinction, and ecosystem function rarely take into account dietary breadth across broad temporal and spatial scales, despite the fact that many ecological processes unfold over temporal and spatial scales that are beyond the scope of traditional ecology. This dissertation addresses this gap by testing the hypothesized drivers of two macroecological and macroevolutionary patterns using dietary niche breadth reconstructed from historical and paleontological canid specimens. Canid predators represent a model system for exploring broad patterns of ecology and evolution given their strong interspecific interactions coupled with a historical legacy of human-driven exterminations resulting in novel community dynamics. Additionally, North American canids offer a rich and diverse fossil history complete with iterative patterns of extinction and radiation used to describe macroevolutionary theory. I quantified dietary niche breadth for western North American canids via stable isotope analysis (SIA) and dental microwear texture analysis (DMTA) and applied these techniques across the spatial and temporal extent of canid distributions. Across space, I explored how mesopredator release has altered the dietary niche breadth of coyotes following the historical extirpation of gray wolves from the Pacific Northwest. Through time, I utilized the 33 million years of evolutionary history preserved in the North American canid fossil record to test the hypothesized link between dietary breadth and the ultimate extinction of canid lineages. In order to make more complete use of fossil, historical, and modern museum collections, which can contain fragmentary specimens, I also developed a novel approach to DMTA, devised to increase sample sizes while not biasing reconstructed dietary behaviors. I demonstrated that multiple facets along the cheek teeth in canids yield comparable microwear signals regardless of molar type or bite force (chapter 2). Thus scans from multiple molars can be combined to increase sample sizes among taxa with limited material. Looking across a latitudinal gradient along Western North America, I found that coyotes sympatric with wolves have reduced dietary niche breadth compared with coyotes sympatric with wolves. Furthermore, DMTA and SIA independently suggest released coyotes increased dietary plasticity following a reduction in scavenging behavior, previously facilitated by wolves (chapter 3). Extinction risk has been hypothesized to be positively correlated with dietary specialization, known as the macroevolutionary ratchet. The fossil record of canids has revealed iterative ratchets as multiple clades evolved towards hypercarnivory followed by rapid lineage extinctions. Morphological traits previously used to describe the macroevolutionary ratchet in canid evolution, however, are unable to capture the dietary breadth of a species. I found morphological traits were inferior to DMTA parameters at explaining extinction risk. Counter to expectations, I observed a positive correlation between specialization and lineage duration and that specialization was not correlated with traditionally-used dietary categories, suggesting that overspecialization in diet alone did not drive iterative extinctions in canids. Outcomes of this dissertation offer direct hypotheses for management officials dealing with proliferating mesopredators and trophic restructuring today. Additionally, my temporal analysis advances our fundamental understanding of macroevolutionary ratchets, and will enable future community-level studies of how species interactions influenced past evolutionary trajectories.