Characterization of “pollen balls” in commercial populations of the alfalfa leafcutting bee, Megachile rotundata in Canada

Abstract

SummaryTo determine the quality of their leafcutting bees (Megachile rotundata), Canadian producers submit samples of loose cells to the Canadian Alfalfa Leafcutter Bee Cocoon Testing Centre (CCTC), Brooks, Alberta, for x-ray analysis. As part of the x-ray analysis report, the CCTC uses the category “pollen balls” to describe bee cells that have been provisioned with nectar and pollen, but in which the presence of larval development is not apparent. We determined the nature of this pollen ball category by comparing results obtained through x-ray analysis to those determined by visual examination of cut-open cells. A total of 35,241 cells were cut open and characterized as either: mouldy; broodless dry or wet provisions; collapsed egg; dead larvae (Instar 1, 2, 3 or 4); live prepupa; parasite; predator; or chalkbrood. Accuracy analysis indicated that the best agreement between the estimates of percent pollen balls obtained visually and using the x-ray method, occurred when wet and dry provision, collapsed egg, dead first or second larvae and all cell types characterized as mouldy were included in the visual categories used to define pollen balls. Overall, dry provisions constituted the highest proportion of pollen balls, followed by mouldy cells, collapsed eggs, wet provisions and finally dead first and second instar larvae. All correlation analyses between the categories making up pollen balls and the total number of pollen balls indicated that none of the single categories is consistently responsible for the overall increase in the percentage of pollen balls.