Pollen dynamics of bumble-bee visitation on Echium vulgare

1. We quantified pollen deposition on the stigma, pollen removal from the anthers and pollen losses in Echium vulgare, visited by workers of Bombus terrestris under controlled conditions. We used dye as a pollen analogue. Bumble-bees were trained to visit a sequence of non-emasculated flowers to estimate pollen carryover and to visit individual flowers to estimate pollen loss. 2. Carryover of pollen grains and dye particles between flowers was similar, which justifies using dye as a pollen analogue. On average 93.8% of the dye particles on the bee were carried over to the next flower. Only a small fraction of the pollen grains was deposited on the stigma (0.15%). A much larger fraction (6.1%) was lost in another way: passively during flight, through grooming or on floral parts other than the stigma. The bees removed 44% of the pollen grains from a fresh flower and 50.3% of this removed pollen adhered to the bee. 3. We predict that, using the parameters mentioned above, during a single visit to a newly opened flower, a bee collects an amount of pollen grains which will bring about 60% geitonogamous self-pollination in the next flower visited. The expected percentage of self-pollination is considerably less if bees visit flowers that have been visited before.