Intraspecific Resource Partitioning in the Bunble Bees Bombus Ternarius and B. Pennsylvanicus

I tested for an association between intraspecific size differences and differences in resource utilization in worker bumble bees (Bombus ternarius and B. pennsylvanicus). Bees were observed foraging in two fields in Minnesota where flower species with short corollas and species with long corollas occurred in both single—species and mixed—species stands. In mixed—species stands, foragers on the species with short corollas were found to have shorter proboscides than conspecific foragers on the species with long corollas. However, proboscis lengths of foragers on the species with a long corolla in single—species stands did not differ from those on the species with a short corolla in single—species stands. Thus, where a choice exists, bumble bee foragers select the species having a corolla most compatible to their proboscis length. I then compared conspecific foragers on a single flower species in one—species versus mixed—species stands. For foragers on flowering species with a short corolla, proboscis length was significantly shorter in the mixed—species than in the single—species stand, but this difference was not found for the species with a long corolla. Thus, presence of other flowering species can influenced the pollinator population of a species, relative to a single—species stand. Resource choice by foragers was analyzed using a probability function, calculated as the ratio of foragers with a given proboscis length on the species with a long corolla to the number with that proboscis length collected on both species. The shape of the probability function was fitted to a sigmoidal curve using probit analysis Regression of the linearized probabilities was significant. This procedure is a better measure of intra— or interspecies association than are measures such as niche overlap, because it quantifies a continuous distribution and accounts for variance in the species distribution.