HONEY BEE LEARNING AND POLLINATION ECOLOGY

Learning in the honey bee (Apis mellifera) was investigated using feeding dishes with peppermint-scented sucrose solution as reinforcement. Results are discussed in reference to coevolution of foraging strategies of insects and pollination strategies of plants.; Resistance to extinction following various numbers of learning trials was measured by counting responses to extinction dishes containing water. Resistance to extinction increased as practice increased from 1 to 10 learning trials, but decreased slightly from 10 to 100 trials. The relatively high resistance to extinction after little practice is probably beneficial to plants with low nectar production, and the relatively low resistance to extinction after little practice is probably beneficial to plants with low nectar production, and the relatively low resistance to extinction after much practice is probably beneficial to bees when nectar production ceases. Although bees oriented equally to peppermint-scented and unscented extinction dishes, they seldom landed on unscented dishes. This result stresses the importance of floral odor as a landing cue for bees. Interference from other bees simultaneously undergoing extinction with the test bee depressed responding of the test bee after 1 learning trial, but only when the extinction dish was unscented. This result is interpreted as disruption of short-term visual memory but not of olfactory memory. A hypothesis is presented that memory disruption due to interference at crowded foraging sites may act as a mechanism to terminate recruiting by honey bees.; Honey bees receiving partial reinforcement exhibited higher response rates to feeding dishes and higher resistances to extinction than bees receiving continuous reinforcement. Bees receiving an intermediate level of partial reinforcement recruited more new workers than bees receiving higher or lower levels. These findings are used to support the hypothesis that flowers use partial reinforcement to secure the most efficient pollination through the stimulation of optimum foraging behavior by bees.; A reversal experiment involving a two-choice spatial discrimination of left (L) vs. right (R) positions was conducted. Correct position choices were reinforced while incorrect choices (errors) yielded water. Error rate was initially low, then increased to a peak following the first reversal of sucrose and water positions, and later decreased progressively with successive reversals. Control experiements using the same two-choice discrimination but in which bees had no opportunity to learn either sucrose position or the reversal concept reproduced essentially the concave-downward error curve obtained in the reversal experiement. The hypothesis is advanced that, in the interest of distinguishing rewarding from unrewarding flowers, honey bees orient successively to different cues resulting in a temporary increase in errors when no distinguishing cues are readily apparent. In one control experiment in which sucrose position was constant, errors decreased linearly with practice, probably due to position learning. The decrease in errors in all experiments is partially credited to learning to recognize sucrose solution before physical contact. Both sucrose-solution recognition and position learning are probably used during flower foraging in the forms of nectar recognition and site constancy respectively. Possible bee odor accumulating at the sucrose dishes had little effect on responding. Bees preferred the downwind position when steady winds were blowing and their R position when no winds were present.