Bumble bee queen production in semi-field studies: assessment of endpoints and challenges

Bumble bees (Bombus terrestris L; Hymenoptera, Apidae) provide important pollination services and are commercially used, e.g. in greenhouse cultures. Consequently, the impacts of pesticides on bumble bees were already tested in the past. In the light of the newest EFSA guidance document on the risk assessment of plant protection products for pollinators standardized higher tier studies for pollinators are needed (EFSA 2013). For that reason a ringtest protocol for a bumble bee semi-field study design was developed in the ICPPR Non-Apis working group starting in 2015 to date. The central endpoint in a higher tier bumble bee study is the colony reproduction success (production of young queens, Cabrera et al. 2016). The endpoint is chosen because at the end of the annual life cycle of a bumble bee colony all workers die and only young queens overwinter. Queens that survive establish a new colony in the following year. However, assessing queen reproduction is challenging. Many variables can influence the number of produced queens, such as the right timing for the termination of the study or the condition of the colony at study start. Furthermore, young queen weights are measured. Weight is used as indicator of diapause survival. Literature values of average weight needed for survival before overwintering state 0.8 g for a young queen for successful overwintering (Beekman et al. 1998). Based on data from ring tests of 2016 and 2017 we tried to answer several open questions concerning queen reproduction, i.e. how can the experimental set-up influence queen weights and how high is the natural variation in queen numbers and queen weight/size?