Bumble bee nest abundance, foraging distance, and host-plant reproduction: implications for management and...

Recent reports of global declines in pollinator species imply an urgent need to assess the abundance of native pollinators and density-dependent benefits for linked plants. In this study, we investigated (1) pollinator nest distributions and estimated colony abundances, (2) the relationship between abundances of foraging workers and the number of nests they represent, (3) pollinator foraging ranges, and (4) the relationship between pollinator abundance and plant reproduction. We examined these questions in an alpine ecosystem in the Colorado Rocky Mountains, focusing on four alpine bumble bee species (Bombus balteatus, B. flavifrons, B. bifarius, and B. sylvicola), and two host plants that differ in their degrees of pollinator specialization (Trifolium dasyphyllum and T. parryi). Using microsatellites, we found that estimated colony abundances among Bombus species ranged from ~18 to 78 colonies/0.01 km2. The long-tongued species B. balteatus was most common, especially high above treeline, but the subalpine species B. bifarius was unexpectedly abundant for this elevation range. Nests detected among sampled foragers of each species were correlated with the number of foragers caught. Foraging ranges were smaller than expected for all Bombus species, ranging from 25 to 110 m. Fruit set for the specialized plant, Trifolium parryi, was positively related to the abundance of its Bombus pollinator. In contrast, fruit set for the generalized plant, T. dasyphyllum, was related to abundance of all Bombus species. Because forager abundance was related to nest abundance of each Bombus species and was an equally effective predictor of plant fecundity, forager inventories are probably suitable for assessing the health of outcrossing plant populations. However, nest abundance, rather than forager abundance, better reflects demographic and genetic health in populations of eusocial pollinators such as bumble bees. Development of models incorporating the parameters we have measured here (nest abundance, forager abundance, and foraging distance) could increase the usefulness of foraging worker inventories in nionitoring, managing, and conserving pollinator populations.