There is growing consensus within the conservation community that the future of biodiversity and ecosystem services will be determined outside of reserve boundaries, in human-dominated landscapes. An accumulating body of work has shown that substantial proportions of biodiversity can exist in human-dominated landscapes. Two key biological questions, however, remain unanswered: 1) what happens to ecological interactions in human-altered habitats? and 2) what are the possibilities for the persistence of biotic communities in these areas? These questions are clearly intertwined, for in many cases persistence is dependent upon continued ecological interactions. Here we discuss what is known about the persistence of bee communities and the pollination interactions that they mediate in human-dominated landscapes. Bees provide an especially interesting case study, because while ongoing bee declines have been well documented in some taxa and locations, paradoxically other studies show long-term resilience of bees to land-use change, with diversity being maintained over decadal time scales. To further inform these trends and patterns, we also present novel results from ongoing studies assessing changes in bee foraging using stable isotopes of carbon and nitrogen, and concurrent work on bee population and landscape genetics in a human-altered landscape in southern Costa Rica. Preliminary results suggest that both landscape context (proximity to native forest) and biotic context (abundance of invasive Apis) are key factors in maintaining pollination interactions and bee and plant community persistence.