Axelrod (1958) suggested that the late Tertiary shift in regional climate -- the elimination of summer rains -- had a profound influence on the evolution of biotic provinces in the American southwest. In particular, the taxa endemic to biotic provinces characterized by summer drought, e.g., the Mojave Desert, should be derived from ancestors that likely inhabited regions of summer rain, e.g., the Chihuahuan Desert. Further, the derived features of summer-drought taxa should be related to water stress. I examined Axelrod's thesis, using a combination of phylogenetic systematics, physiological ecology, and vicariance biogeography. The first chapter is a cladistic study of the pinyon pines, 13 taxa of small trees that range from the summer-wet regions of Mexico to the summer drought regions of Nevada and California. A parsimony analysis using twenty morphological characters showed that the most recently derived pinyons are from regions of summer drought. The "summer-drought" taxa are characterized by relatively few needles per fascicle. Because fewer needles per fascicle results in a reduction in the needle surface-to-volume ratio, Haller (1965) hypothesized that fewer needles in pines is an adaptation to reduce transpirational water loss. The second chapter reports on a two year study of the xylem pressure potentials of single- and double-needled fascicles of hybrid pinyons in central Arizona. The results showed no significant differences between single- and double-needles. I concluded that either needle morphology does not effect water relations, or that the relatively high precipitation during the study did not allow significant water stress to occur. The third chapter uses the methods of vicariance biogeography to search for a common pattern of relationship between southwestern biotic provinces, as indicated by the relationships of their endemic taxa. Using a biogeographic parsimony analysis, I compared the area cladograms of six taxa -- junipers, pinyon pines, the composite Palafoxia, hedgehog cactus, desert tortoises, and gecko lizards. The most parsimonious area cladogram supports Axelrod's (1958) hypothesis, but also shows that some taxa, notably the junipers, support other patterns of area relationships, e.g., summer-drought primitive. I suggest that there is no single pattern of area relationships because of the effects of the Pleistocene (including dispersal and extinction) and vicariance events other than the Tertiary climatic change, e.g., the separation of the Baja peninsula from mainland Mexico during the Miocene.