Honeybee brood testing under semi-field and field conditions according to Oomen and OECD GD 75: is there a difference of the brood termination rate?
According to current European regulations on the risk assessment of plant protection products, the risk on honey bee larvae or honey bee brood has to be addressed. If the assessment indicates, that a potential risk cannot be excluded based on data derived from laboratory studies, two higher-tier options are given by the EFSA bee Guidance Document to refine this under more realistic conditions: the Oomen bee brood feeding test and brood studies performed according to the OECD Guidance Document 75. Both study types focus on the brood termination rate (BTR) as the key endpoint. While the Oomen brood test investigates the brood development after the acute or chronic administration of a test item spiked sugar solution to unconfined colonies, brood studies according to OECD GD 75 are performed under semi-field confined exposure conditions and examine potential effects on the bee brood after the overspray of a bee attractive flowering crop. However, the evaluation of historical data from semi-field studies according to OECD GD 75 showed a strong variability of the BTR of pre-imaginal stages developing from marked eggs (BTReggs) in the control. As an alternative, field studies according to EPPO 170 which comprise bee brood evaluations according to OECD GD 75 were considered to produce more reliable termination data. The statistical analysis of available control data shows that Oomen feeding studies and bee brood studies performed under field conditions lead to significantly lower BTReggs of = 20% compared to semi-field bee brood studies for which a mean BTR of about 30% is observed. Moreover, studies with unconfined colonies show a high proportion of control replicates with BTReggs =30% and =40% indicating a higher reliability compared to semifield studies. A comparison of the possibilities and limitations of the three methods shows the strength of each method. In Oomen studies, the exposure of the brood and of the hive bees only can be regarded as artificial. However, the test concentrations can be adjusted to specific needs and to different feeding durations of at least one (acute) or 9 days (chronic). Furthermore, the absence of ‘caging effects’, the low dependency on climatic or crop conditions, the potential to test also herbicides which control dicotyledonous plants (since no crop plant is adversely affected by its mode of action) and an exposure period of at least nine days in chronic Oomen studies are crucial advantages. In contrast, the exposure scenarios of the two other methods are much more realistic and especially for semi-field studies a worst-case situation. Moreover, they also include exposure via pollen and exposure levels and durations, which strongly depend on the application rate and the flowering period of the treated crop. Whereas a dilution of plant protection product residues cannot be excluded during the exposure period in studies with unconfined colonies due to the shift to untreated flowering plants in the surrounding, this is not given for semi-field studies.