Honey bees, Apis mellifera, play an important role in many ecosystems, pollinating a wide variety of native, agricultural, and exotic plants. The recent decline in the number of feral and managed honey bee colonies in North America, as well as the arrival of Africanized honey bees, have caused concern about adequate pollination for agricultural crops and natural plant communities. However, little is known about feral colonies, and the feral population is the source for Africanized honey bees as they spread and infiltrate managed populations. The goal of my dissertation was to examine the ecology of feral honey bee colonies, adding the spatial context necessary to understand the population ecology and patterns of resource use by feral honey bees on the Welder Wildlife Refuge. I defined the functional heterogeneity of feral honey bee habitat by identifying the suitability of different habitats for feral colonies based on the distribution and abundance of important resources (cavities, nectar, and pollen). I evaluated the distribution and abundance of feral colonies by examining nest site characteristics, population trends, and spatial and temporal patterns in cavity use. Lastly, I examined resource use by evaluating patterns in pollen collection and identifying where and when honey bees searched for resources. Overall, the Welder Wildlife Refuge provided excellent habitat for feral honey bees, supporting a high density of feral colonies. The dense live oak habitat was the best overall source for cavities, nectar, and pollen. Nectar and pollen were abundant throughout the year, with the exception of December and January, when a large number of honey bees searched for resources. Cavities did not appear to vary in their suitability for feral colonies based on measured structural and environmental attributes, since no cavity attributes were correlated with indices of cavity quality. However, the cavity quality indices varied between cavities, suggesting some cavities were more suitable for feral honey bees than others. Colonies were aggregated within the study area, probably due to the distribution of resources. The invasion of Africanized honey bees appeared to fragment the existing European population, with Africanized colonies aggregated in distribution and European colonies random in distribution.