1. Genetic polymorphisms of flowering plants can influence pollinator foraging but it is not known whether heritable foraging polymorphisms of pollinators influence their pollination efficacies. Honey bees Apis mellifera L. visit cranberry flowers for nectar but rarely for pollen when alternative preferred flowers grow nearby.
2. Cranberry flowers visited once by pollen-foraging honey bees received four-fold more stigmatic pollen than flowers visited by mere nectar-foragers (excluding nectar thieves). Manual greenhouse pollinations with fixed numbers of pollen tetrads (0, 2, 4, 8, 16, 32) achieved maximal fruit set with just eight pollen tetrads. Pollen-foraging honey bees yielded a calculated 63% more berries than equal numbers of non-thieving nectar-foragers, even though both classes of forager made stigmatic contact.
3. Colonies headed by queens of a pollen-hoarding genotype fielded significantly more pollen-foraging trips than standard commercial genotypes, as did hives fitted with permanently engaged pollen traps or colonies containing more larvae. Pollen-hoarding colonies together brought back twice as many cranberry pollen loads as control colonies, which was marginally significant despite marked daily variation in the proportion of collected pollen that was cranberry.
4. Caloric supplementation of matched, paired colonies failed to enhance pollen foraging despite the meagre nectar yields of individual cranberry flowers.
5. Heritable behavioural polymorphisms of the honey bee, such as pollen-hoarding, can enhance fruit and seed set by a floral host (e.g. cranberry), but only if more preferred pollen hosts are absent or rare. Otherwise, honey bees' broad polylecty, flight range, and daily idiosyncrasies in floral fidelity will obscure specific pollen-foraging differences at a given floral host, even among paired colonies in a seemingly uniform agricultural setting.