In search of evidence for the experience of pain in honeybees: A self-administration study

Despite their common use as model organisms in scientific experiments, pain and suffering in insects remains controversial and poorly understood. Here we explore potential pain experience in honeybees (Apis mellifera) by testing the self-administration of an analgesic drug. Foragers were subjected to two different types of injuries: (i) a clip that applied continuous pressure to one leg and (ii) amputation of one tarsus. The bees were given a choice between two feeders, one offering pure sucrose solution, the other sucrose solution plus morphine. We found that sustained pinching had no effect on the amount of morphine consumed, and hence is unlikely to be experienced as painful. The amputated bees did not shift their relative preference towards the analgesic either, but consumed more morphine and more solution in total compared to intact controls. While our data do not provide evidence for the self-administration of morphine in response to pain, they suggest that injured bees increase their overall food intake, presumably to meet the increased energy requirements for an immune response caused by wounding. We conclude that further experiments are required to gain insights into potential pain-like states in honeybees and other insects.

. Pie charts comparing the mortality rate of the control and the injured bees (top row; n=number of bees), and the percentage of cages in which bees died (bottom row; n=number of cages) in Experiment 1. Mortality was equally high in the injured and control groups (see Supplementary Table S2).

Figure S2.
Pie charts comparing the mortality rate of the control and the injured bees (top row; n=number of bees), and the percentage of cages in which bees died (bottom row; n=number of cages) in Experiment 2. Deaths occurred significantly more often in the amputated group than in the corresponding control (see Supplementary Table S2).

Consumption and Feeder Colour
To examine whether the bees' solution intake was influenced by the colour of the feeder, we tested if the amount of morphine and sucrose consumed varied between the blue and the yellow feeders that were used in the experiments.
For the control group of Experiment 1 (Clip), no significant difference in consumption was found between the blue and the yellow feeder ( Supplementary Fig. S3). The injured bees, however, drank more sucrose solution when it was offered in the yellow feeder (blue: 0.39g, yellow: 0.49g, W=193, P=0.034; Supplementary Fig. S4).
In the Amputation Experiment, the control bees showed an increased consumption when the yellow feeder contained sucrose solution (blue: 0.38g, yellow: 0.53g, W=170, P=0.001; Supplementary Fig. S5). The injured bees in this experiment displayed no colour-dependent sucrose consumption, but they drank more morphine solution when it was provided in a blue feeder (blue: 0.30g, yellow: 0.20g, W=492, P=0.004; Supplementary Fig. S6).
Taken together, this analysis revealed no consistent colour preference across experiments and groups. Most importantly, the bees showed a preference for sucrose over morphine irrespective of the feeder colour. The morphine to sucrose ratio stayed approximately the same, and this pattern was not overruled when the feeder colours were swapped. Figure S3. Absolute consumption (in g, per bee, over four days) of morphine solution (M) and pure sucrose solution (S) of the control group of Experiment 1 (Clip) in relation to the feeder colour. Consumption is shown separately for the blue (b) (n=25) and the yellow (y) (n=26) feeders. Mean and median values are represented by red dots and horizontal black lines, respectively. No significant difference in consumption was found between the blue and the yellow feeders, either for the morphine solution or the sucrose solution (see Supplementary Table S2). Figure S4. Absolute consumption (in g, per bee, over four days) of morphine solution (M) and pure sucrose solution (S) of the injured group of Experiment 1 (Clip) in relation to the feeder colour. Consumption is shown separately for the blue (b) (n=26) and the yellow (y) (n=23) feeders. Mean and median values are represented by red dots and horizontal black lines, respectively. Bees consumed significantly more sucrose solution when it was offered in the yellow feeder. The morphine consumption did not differ significantly between the blue and the yellow feeders (see Supplementary Table S2). Figure S5. Absolute consumption (in g, per bee, over four days) of morphine solution (M) and pure sucrose solution (S) of the control group of Experiment 2 (Amputation) in relation to the feeder colour. Consumption is shown separately for the blue (b) (n=27) and the yellow (y) (n=27) feeders. Mean and median values are represented by red dots and horizontal black lines, respectively. Bees consumed significantly more sucrose solution when it was offered in the yellow feeder. The morphine consumption did not differ significantly between the blue and the yellow feeders (see Supplementary Table S2). Figure S6. Absolute consumption (in g, per bee, over four days) of morphine solution (M) and pure sucrose solution (S) of the injured group of Experiment 2 (Amputation) in relation to the feeder colour. Consumption is shown separately for the blue (b) (n=25) and the yellow (y) (n=27) feeders. Mean and median values are represented by red dots and horizontal black lines, respectively. Bees consumed significantly more morphine solution when it was offered in the blue feeder. The sucrose consumption did not differ significantly between the blue and the yellow feeders (see Supplementary Table S2).