Date Published: March 29, 2017
Publisher: Public Library of Science
Author(s): Robert Brodschneider, Anika Libor, Vera Kupelwieser, Karl Crailsheim, Olav Rueppell.
We measured the distribution of sugar solution within groups of caged honey bees (Apis mellifera) under standard in vitro laboratory conditions using 14C polyethylene glycol as a radioactive marker to analyze ingestion by individual bees after group feeding. We studied the impact of different experimental setups by varying the number of bees, age of bees, origin of bees, duration of experiment, the amount of available diet, and the influence of the neurotoxic pesticide imidacloprid in the diet on the feeding and food sharing behavior (trophallaxis). Sugar solution was non-uniformly distributed in bees in 36 out of 135 cages. As a measure of the extent to which the sugar diet was equally distributed between caged bees, we calculated the (inner 80%) intake ratio by dividing the intake of the 90th percentile bee by the intake of the 10th percentile bee. This intake ratio ranged from 1.3 to 94.8 in 133 individual cages, further supporting a non-uniform distribution of food among caged bees. We can expect a cage with 10 or 30 bees containing one bee that ingests, on average, the 8.8-fold of the bee in the same cage ingesting the smallest quantity of food. Inner 80% intake ratios were lower in experiments with a permanent or chronic offering of labelled sugar solution compared to temporary or acute feedings. After pooling the data of replicates to achieve a higher statistical power we compared different experimental setups. We found that uniform food distribution is best approached with 10 newly emerged bees per cage, which originate from a brood comb from a single colony. We also investigated the trophallaxis between caged honey bees which originally consumed the diet and newly added bees. Color marked bees were starved and added to the cages in a ratio of 10:5 or 20:20 after the initial set of bees consumed all the labelled sugar solution. The distribution of the labelled sugar solution by trophallaxis within 48 hours to added bees was 25% (10:5) or 45% (20:20) of the initial sugar solution. Imidacloprid at its median lethal dose (LD50) in the sugar solution reduced this post-feeding food transmission to 27% (20:20). Our results show that differences in food intake exist within caged bees that may lead to differential exposure that can influence the interpretation of toxicity tests.
Exposure to insecticidal pesticides is among other stress factors like parasitic mites, pathogens, poor hive management, loss of foraging habitat, and/or poor nutrition thought to be associated with declines in honey bee health or outright losses of honey bee colonies [1–8]. Effects of pesticides on honey bees are studied in vivo at the colony level in field or semi-field studies, and are studied in vitro using small groups of bees or individual bees under laboratory conditions [9, 10]. For oral toxicity testing, individual bees may be exposed to liquid diets by a single feeding method  or a number of caged bees is fed simultaneously by the group feeding method  which is the most commonly used method for experiments with pesticides and pathogens. The latter test assumes uniform distribution of food among caged honey bees and that trophallaxis plays an important role in distributing food and hence the test substance.
Experiments were carried out from July to September 2014 using A. mellifera carnica from colonies of the institute of zoology. Since A. mellifera is not a protected species, no specific permission was required to perform the study. Experimental cages were modified 200 mL plastic cups . Due to possible radioactive contaminations during the experiments, all cages were used once and then discarded. For different research questions, we used 10 to 40 bees per cage. Before each experiment, bees were starved up to two hours as specified in the OECD guidelines for the testing of chemicals No. 213 . During starvation period dead bees were replaced before applying different experimental feeding regimes (temporary, permanent, acute or chronic) and caging conditions.
We evaluated the suitability of honey bee group feeding to assess toxicity and toward a better understanding of group feeding under in vitro conditions. We used a stable radioactive marker (14C-labelled PEG) to trace food sharing among caged honey bees since polyethylene glycol accumulates in the body and therefore reflects the cumulative consumption of a caged bee [25, 26]. Toxicological studies are based on the assumption that caged honey bees all consume the same share of an offered food; however high variability in toxicity thresholds generated in laboratory-based studies and discrepancies between field-realistic levels suggest that exposure may not be adequately assessed . These findings can be explained by either different sensitivities to test substances or simply by an unequal consumption [30, 31]. For the first time we quantified two characteristics of food distribution in caged honey bees. First, we established an intake ratio, which describes the gap between the bee ingesting the highest amount of food and the bee ingesting the lowest quantity of food in the cage. Second we tested the distribution of food in caged bees for a uniform distribution and compared different experimental setups.