The relation between pattern and landmark vision of the honeybee (Apis mellifera)

Abstract

This is an exploratory study which takes advantage of a new method to study the difference between discrimination of patterns and the use of landmarks by honeybees. The experiments examine the transfer of black and white cues between two-dimensional patterns that are seen ahead by the bees, and lateral landmarks on the walls of tunnels through which they fly, and vice versa. The Y-choice apparatus, in which freely Hying bees choose one of two targets from a fixed distance, was modified by adding, in each arm, a white paper tunnel followed by a transparent baffle, which controls the range at which the bees' choices are made. The results reveal that: (a) Bees can use both the distribution of black areas and the orientation of edges that they see on the side walls as they pass. So far as they were tested, the same visual features are used in landmarks as in targets. (b) The bees can immediately transfer the cues learned on tunnels to corresponding patterns on the targets, i.e. transfer forwards along their flight path, but not in the reverse direction. (c) New landmarks are learned in a few visits, sometimes even in one visit, but orientation cues on tunnels and targets take about 2 h to learn. (d) When trained on targets, they will not at first use the same cues on unfamiliar landmarks, but they rapidly learn to insert a new landmark into their flight path. The conclusion is that learning landmark locations and discriminating 2D patterns subtending < 45° in forward vision have different mechanisms.