Studies on year-round rearing of Japanese native bumblebees (Bombus spp.) for buzz-foraging crop pollination

Bumblebees (Bombus spp.) have evolved buzz foraging, a unique pollen-collection behavior that makes them an important pollinator in natural ecosystems and a useful crop pollinator in agriculture. Since the first successful mass production of the European bumblebee (Bombus terrestris) in Belgium in 1987, commercially available bumblebee species have been used widely for crop pollination of greenhouse tomatoes (Lycopersicon esculentum) and eggplants (Solanum metongena), neither of which produce nectar rewards for bees. Commercial packages of B. terrestris were imported into Japan for the first time in 1991. These imports save farmers labor, produce high-quality fruits, and reduce use of chemical pesticides. However, escape of these non-native bumblebees may harm the native biota. The possible ecological impacts are hybridization with native species, nest usurpation, spread of non-native parasites and diseases, etc. These are strong arguments favoring commercialization and use of Japanese native bumblebees for pollination instead of introduced B. terrestris. I investigated year-round mass production of Japanese native bumblebees and evaluated their pollination efficiency of greenhouse tomatoes. The results are as follows: 1. Laboratory reared queens of B. hypocrita and B. Ignitus survived artificial hibernation induced by chilling at 5 deg C for 4 months. B. hypocrita queens preferred dry vermiculite in place of soil, while B. ignitus queens preferred wet vermiculite or peat moss at 50%-55% moisture content. After chilling, stepped temperature increases of 10 deg C in 6 h resulted in higher survival rates for both species of queens. B. hypocrita queens younger than 10-days old gave better nesting results than older queens. 2. Worker and new-queen productivity of laboratory reared B. hypocrita and B. Ignitus are influenced by delays in oviposition and first worker emergence. In B. hypocrita, colonies with oviposition delays of 20 or less days produced significantly more female castes than colonies with oviposition delays of 21 or more days. In B. Ignitus, colonies with oviposition delays of 10 or less days produced significantly more female castes than colonies with oviposition delays of 11 or more days. In both species, colonies with long delays in first worker emergence produced fewer female castes than normal colonies in which first worker emergence occurred at 28 or less days. These two parameters permit early identification of normal colonies for commercial mass production. 3. Colony development in B. hypocrita and B. ignitus has four oviposition phases as already described in B. terrestris. 4. Worker productivity of B. Ignitus is higher than that of B. hypocrita. The duration of fertilized-egg oviposition (DFO) in B. ignitus is 47.5 +-5,3 days, about 10 days longer than in B. hypocrita (36.6 +-4.7 days). The foundress B. hypocrita queen produces progeny queens from fertilized eggs in Phase 2. However, progeny-queen production occurred only after Phase 3 in B. ignitus, suggesting that B. hypocrita has flexible production of the reproductive caste. Two field populations of B. hypocrita were collected from Yamanashi Prefecture (P-Y) and Shizuoka Prefecture (P-S) but no differences in DFO were detected. However, there was a significant difference in the DFO of the F1 generations of laboratory-reared populations. The DFO (F1) of the P-S progeny was significantly shorter than that of the P-Y progeny. The P-S colonies tended to produce progeny queens earlier than the P-Y colonies. 5. The ovary and corpora allata (CA) of older workers of B. hypocrita and B. ignitus developed even when the foundress queen was present in the colony. Competitive behavior between older workers and the foundress queen and younger workers resulted in higher mortality of younger workers during the competition phase. 6. Buzz foraging of greenhouse tomatoes (Lycopersicon esculentum) by four species of Japanese native bumblebees was examined. There was no difference in the pollination efficiency between the native bumblebees and imported bumblebees (B. terrestris). The fruiting rate was high (84%-100%) and there were almost no puffy fruits (0%-7%). 7. Tomato flowering is largely influenced by air temperature and tomato flowers exposed to cold conditions produce less pollen than flowers exposed to warm conditions. However, temperature has no effect on pollen viability. 8. The foraging behavior of Japanese native bumblebees was observed using artificial feeding dishes. Workers continued visiting for about 10 minutes even after foraging dishes became empty, suggesting that memory plays a role. Although foraging workers were attracted by volatile substances from pollen, they did not continue foraging in the absence of pollen. Laboratory experiments suggest that bumblebees recognize the presence of tomato pollen in a greenhouse. When tomato plants have no pollen, foraging workers seemed to target other flowers. Mass production of Japanese native bumblebees and their application in greenhouse pollination is discussed based on these findings.