Determining how pollinators visit plants versus how they carry and transfer pollen is an ongoing project in pollination ecology. The current tools for identifying the pollens that bees carry have different strengths and weaknesses when used for ecological inference. In this study we use three methods to better understand a system of congeneric, co-flowering plants in the genus Clarkia and their bee pollinators: observations of plant-pollinator contact in the field, and two different molecular methods to estimate the relative abundance of each Clarkia pollen in samples collected from pollinators. We use these methods to investigate if observations of plant-pollinator contact in the field correspond to the pollen bees carry; if individual bees carry Clarkia pollens in predictable ways, based on previous knowledge of their foraging behaviors; and how the three approaches differ for understanding plant-pollinator interactions. We find that observations of plant-pollinator contact are generally predictive of the pollens that bees carry while foraging, and network topologies using the three different methods are statistically indistinguishable from each other. Results from molecular pollen analysis also show that while bees can carry multiple species of Clarkia at the same time, they often carry one species of pollen. Our work contributes to the growing body of literature aimed at resolving how pollinators use floral resources. We suggest our novel relative amplicon quantification method as another tool in the developing molecular ecology and pollination biology toolbox.