Many populations exhibit collective behavior, where interactions among nearby individuals scale up to cause emergent patterns in the behavior of groups, as in the coordinated movement of a flock of birds or a school of fish. Populations influenced by collective behavior violate the assumption of mass action that underlies most ecological models, in which individuals are viewed as statistically independent. However, the ecological significance of collective behavior is not well understood, because studies have been limited to populations where high throughput ethological data is available, such as in the laboratory or in computer simulations. This dissertation tests for the signal of collective behavior in ecological data-data on the distribution patterns of organisms collected on a coarser spatial and temporal scale than the underlying processes-and examines the influence of collective behavior on population dynamics. Data on the locations of migratory caribou (collected every five days by satellite tracking collars) are shown to be generated by two distinct processes. The first process creates broad-scale spatiotemporal order in movement patterns, and is likely driven by seasonally and spatially fluctuating environmental and physiological cues. The second process creates finer-scale order that is likely due to behavioral interactions among nearby individuals. The strength of alignment in the velocities of nearby individuals varies systematically with time of year, suggesting that collective behavior can be a dynamic property of migratory populations. The dissertation then considers collective mobility patterns in humans, analyzing census data on the commuting patterns of workers in Canadian cities. The level of order in commuting patterns varies systematically among cities. In particular, in some cities a disproportionate number of workers travel to work in a few focal locations. Simulations of the spread of a respiratory infection in each city predict differences among cities in the risk of an epidemic, due to systematic variation in the level of order in the commuting patterns of workers. In particular, larger cities tend to be more highly organized and, as a result, have a disproportionately higher probability of sparking an epidemic. The dissertation then explores the role of large cities in supporting the emergence of a new strain of influenza in dogs. The analysis combines demographic data on animal shelters in the United States, molecular data from the pathogen and seroprevalence estimates from the literature to show that large animal shelters in major metropolitan areas function as endemic reservoirs for the virus, facilitating sporadic outbreaks in the wider population. In sum the dissertation research shows that collective behavior can sometimes be detected and characterized in ecological data without recourse to fine-scale behavioral observations, and that collective behavior can significantly alter population dynamics at broad spatial and temporal scales.