A bee-vectoring method has been recently developed against greenhouse insect pests. In this study, we conducted experiments to determine if bumble bees, bombus terrestris L. could vector Bacillus thuringiensis subsp. aizawai from hives equipped with a pathogen dispenser to leaf and flower of sweet pepper. First, in laboratory, insecticidal activity of B. t. subsp. aizawai NT0423 WP (Tobagi, Dongbu Hitek Co.) against 2nd instar larva of Spodoptera litura Fabricius was evaluated to determine if the amount of B. thuringiensis deposited on the leaf and flower would be sufficient to control the tobacco cutworm. The B. t concentration of the product in this experiment was estimated as 2.8-3.1 x 109 CFU/ml. The B. t. concentration on the body of bumble bees that was captured as they exited hives with filled pathogen dispensers was 4.6 x 109 cfu/ml. At the laboratory bioassay, the mortality of tobacco cutworm was serially measured depending on the pathogen concentrations. At the cage trials, the B. t. delivered by bumble bees was detected in bees, flowers and leaves. Through the several trials, the control effect on the tobacco cutworm has been investigating. The side effect on bumble bee from the first trial was not found in this experiment.

To understand geographic genetic variation of the species and relationships among populations of the bumble bee, Bombus ardens, that is utilized as green house pollinator we expanded our investigation by sequencing somewhat longer mitochondrial DNA (mtDNA) fragment, covering some uninvestigated regions within the species distribution, and analyzing the sequence data in terms of population genetic structure. For the purpose of study, a portion of mitochondrial COI gene, corresponding to "DNA Barcode" region (658 bp) was sequenced from 160 individuals of B. ardens collected over 15 localities in Korea. The sequence data revealed overall relatively low genetic diversity within species, with a maximum sequence divergence of 0.3%. Geographically, one haplotype (BARBA01) was found in all localities surveyed, with the frequency of 91% (145 among 160 individuals), whereas other haplotypes were found in a locality mostly as a single individual, suggesting that haplotype distribution can be summarized as coexistence of widespread, one dominant haplotype and regionally restricted, other haplotypes. Overall, very high rate of per generation female migration (Nm = 4.6 ~ infinite) and very low level of geographic substitution (FST = 0 ~ 0.099) among localities were characteristic. Although some populations were genetically subdivided from the remaining localities in the hierarchical analysis, there was regional polarity on this subdivision. Taken together with gene flow estimates, the nature of genetic divergence of the bumble bee populations is characterized as one that possessing low genetic diversity, high gene flow, and wide spread of one dominant haplotype, consistent with the previous finding. To have further detailed information of this valuable genetic resource, further longer and variable molecular portion is under investigating.

To understand geographic genetic variation of the species and relationships among populations of the bumble bee, Bombus ardens, we sequenced a portion of mitochondrial COI gene, which corresponds to "DNA Barcode" region (658 bp) from 101 B. ardens individuals collected over 11 localities in Korea. The sequence data were used to investigate genetic diversity within populations and species, geographic variation within species, phylogeographic relationship among populations, and phylogenetic relationship among haplotypes. A total of nine haplortpes were found, but they very close to each other (a maximum sequence divergence of 0. 304%). Summarized, overall moderate to low genetic diversity within populations and species was characteristic, concordant with the high potential to disperse of B. ardens in Korea. There was no clear regional subdivision was observed and relatively high rate of gene flow among localities and low FST was characteristic.