Transfer along a continuum in the discriminative learning of honeybees

M. Bitterman,M. Walker,Younglim Lee

DOI: https://doi.org/10.1037/0735-7036.104.1.66

1990-02-16

Abstract:In each of 2 experiments on discriminativ e learning in free-flying honeybees (Apis mellifera), performance in a difficult problem was found to be facilitated by prior training in an easier problem. In Experiment 1, animals that were trained to detect a strong anomaly in the ambient geomagnetic field performed better when the intensity of the anomaly was reduced than did control animals that were trained from the outset with the weaker anomaly. In Experiment 2, animals that were trained to detect a 20-^1 drop of sucrose solution and then a 10-^1 drop of the same solution performed better when the size of the drop was reduced to 5 ii\ than did control animals trained from the outset with the 5-fil drop. These results are of interest because they add transfer along a continuum to a growing list of vertebrate learning phenomena found in honeybees and because of their bearing on a developing theory of discriminative learning in honeybees. Pavlov (1927) reported several instances in which discrimination between two closely similar stimuli seemed to be facilitated by pretraining with less similar stimuli varying in the same dimension. For example, a dog that had failed to discriminate between a white paper (reinforced stimulus; S+) and a light gray paper (unreinforced stimulus; S—) when trained with them from the outset did succeed in doing so after training in which the brightness of unreinforced stimulus was shifted in stepwise fashion from a much darker gray to the original light gray. Analogous results have been obtained repeatedly in work with rats—notably by Lawrence (1952), who provided the first formal demonstration of the phenomenon and named it transfer along a continuum—with pigeons (Marsh, 1969), with rabbits (Haberlandt, 1971), with fishes (Hemmings, 1965), even with octopuses (Sutherland, Mackintosh, & Mackintosh, 1963), and, as we report in this article, with honeybees. The implications of the new results are considered after the experiments have been described.