Partitioning of foraging space and nesting substrates by birds was examined in a desert community in southern Arizona. The study site was dominated by low shrubs, with scattered thorn trees and stem succulents. Chi-square analysis demonstrated the importance of plant height, growth form, and patch location within the habitat in selection by birds of foraging and nesting substrates and, ultimately, in determination of bird community composition. Phainopeplas, ash-throated flycatchers, verdins, black-tailed gnatcatchers and gila woodpeckers favored thorn trees for feeding substrates, which they subdivided according to height, plant part and habitat patch. Most of these birds, as well as the common ground feeding species (cactus wrens, curve-billed thrashers, black-throated sparrows and Gambel's quail), also used thorn trees and succulents for nesting sites. The presence of thorn trees and stem succulents apparently enhances avian diversity in complex desert shrublands by providing protected nest sites and a structural framework that facilitates subdivision of foraging space. INTRODUCTION Most ecologists agree that vegetation exerts a major influence on the assemblage of breeding birds that occurs within a habitat. In addition to providing the architectural framework within which birds exist, vegetation affects the supply of both plant and animal foods, nest sites and song perches. MacArthur and MacArthur (1961) inspired a wealth of research into bird community structure when they found a predictive linear relationship between vegetation complexity and bird species diversity in forested habitats of the eastern U.S. Scholars subsequently studied bird communities in various temperate and tropical habitats in efforts to substantiate and refine MacArthur's predictive model (e.g, MacArthur et al., 1962; MacArthur et al., 1966; Recher, 1969; Karr, 1971). The arid portions of the southwestern U.S. are well-represented in this body of literature, with studies done by Tomoff (1974) and others (Austin, 1970; Carothers et al., 1974; Stamp, 1978). In keeping with the ecological attitudes of their day, most scholars studying desert bird communities during the 1970s were concerned with communities as units, which they often summarized with diversity indices. Little attention was paid to variation within communities, such as the use of different vegetation components by individual bird species. Indeed, Tomoff (1974) was one of the few students of arid lands of this era who sought underlying functional links between vegetation structure and birds by examining feeding and nesting habits of the community overall. More recently ecologists have probed beneath the superficial community generalizations that were an outgrowth of MacArthur's time to analyze both species interactions and spatial and temporal variation in community composition. Examples include recent studies of desert bird communities by Rosenberg et al. (1982) and Meents et al. (1983). Both studies are in-depth examinations of habitat characteristics and patterns of species occurrence, but, unlike Tomoffs work, do not attempt analysis of nesting patterns. The intent of this study is to examine patterns of vegetation use for feeding and nesting by common species breeding in a desert shrub habitat in southern Arizona. This analysis addresses two questions: (1) What plant cover characteristics separate the common bird species? (2) To what extent does partitioning of nest sites augment subdivision of foraging substrates in the determination of avian community composition in desert habitats?