The use of isoenzyme markers to determine pollen flow and seed paternity mediated by Apis mellifera and Bombus spp. in Trifolium repens, a self-incompatible plant species

Five fixed isoenzyme selections, within the white clover (Trifolium repens) cultivar S100, were used as genetic markers in a study of the effectiveness of bees of different genera as pollen vectors. Five plants of each of the five selections were cloned to provide two identical plots of 25 plants with equal proportions of the five different pollen donor plants. Pollen flow within each plot, mediated by honey bees, Apis mellifera, and bumble bees, Bombus spp., was determined on an individual inflorescence basis in plots containing the different pollinator species. The movement between plants of the various pollen types was monitored by determining the paternity of the seedlings that resulted following pollination and fertilization using starch gel electrophoresis. Variation was found in the proportions of the different paternal donors to sire seed within pods. Following pollination by Apis, c. 50% of pods contained seed from ovules fertilized by pollen from two different selections, c. 25% with that from three or four selections and none with that from only one. Following pollination by Bombus, c. 30% of pods contained seed from ovules fertilized by pollen from two selections, c. 50% with that from three selections and c. 10% with that from one or four. Within an inflorescence, pollination by both Apis and Bombus resulted in c. 70% of seed having the same paternity. Within individual pods, one paternal selection dominated; in pods containing seeds of two, three or four paternities the dominant paternal selection accounted for 75%, 60% and 60%, respectively, following pollination by Apis, and 75%, 64% and 63%, respectively, following pollination by Bombus. The results are discussed in relation to the foraging behaviour of bees, pollen carryover and gene flow