Figure 3. Effects of JH on aversive learning (A) and memory recall (B) in 2-day old worker bees raised with QMP.

Learning performance in control bees injected with vehicle alone (n = 42) was compared with learning in bees treated with JH (n = 42). Bees were conditioned using differential conditioning of the sting extension reflex. Each bee received 12 pseudo-randomised conditioning trials. In six trials, eugenol was paired with an electric shock (CS+) in the remaining trials 2-hexanol was presented without reinforcement (CS−). (A) The percentage of control bees displaying a sting extension response (SER) to CS+ did not change significantly across successive trials (GLMM, P = 0.7355, z = −0.338). However, the percentage of bees displaying sting extension in response to the CS− odour significantly decreased (GLMM, P = 0.0258, z = −2.228). The slopes of the CS+ and CS− response curves of the control bees were significantly different (GLMM, P≤0.05). In contrast to controls, JH-treated bees showed clear evidence of learning with a significant increase in the % of bees displaying SER over successive conditioning trials with CS+ (GLMM, P≤0.001, z = 3.784). These bees also displayed a significant decrease in responses to CS− (GLMM, P≤0.001, z = −3.964). The slopes of the CS+ and CS− response curves of the JH treated bees were also significantly different (GLMM, P≤0.001). Significant differences in responses to CS+ and CS− are indicated by plots that do not share a letter (GLM, P≤0.01). (B) The percentage of bees responding to CS+ with sting extension during the retention test was significantly higher in JH-treated bees than in controls, as indicated by an asterisk (P≤0.01). (C) Percentage of bees responding to electric shock at intensities ranging between 0.25 and 8 volts. Responses of control bees injected with vehicle alone did not differ from those of JH treated bees at any of the voltages tested. (D) Neither group of bees responded significantly with sting extension in response to placement alone.