The resource island hypothesis predicts that soil resources such as nitrogen, phosphorus, and water will be distributed evenly in grasslands but have a patchy distribution focused around plants in shrublands. This hypothesis predicts that microorganism numbers will follow resources and be (i) evenly distributed in grasslands, (ii) concentrated around individual plants in shrublands, and (iii) higher where resources are higher when comparing the same vegetation type. This study enumerated total heterotrophic bacteria and a subset of these, the nitrogen-efficient guild (NEG), in three shrublands (playa fringe mesquite, tar bush, and creosote) and two grasslands (playa and bajada). Both heterotrophs and NEG members followed the distribution pattern predicted by the resource island hypothesis. There were no significant differences in heterotroph or NEG numbers comparing at-plant and interplant samples for both the playa and bajada grasslands. Furthermore, populations were generally higher in nutrient-rich playa grasslands than nutrient-poor bajada grasslands. In contrast, both heterotroph and NEG numbers were higher at shrubs than between shrubs in all three shrub sites. These results suggest that resource abundance in resource islands predicts the distribution of heterotrophic bacterial numbers in desert soils.