THE IDENTITY OF THE PREVIOUS VISITOR INFLUENCES FLOWER REJECTION BY NECTAR-COLLECTING BEES

In the field, recently probed flowers of borage, Borago officinalis, t ypically contained little or no nectar (and hence were relatively unre warding), whether probed by a bumblebee, Bombus spp., worker or a hone ybee, Apis mellifera. However, a nectar-collecting bee was likely to r eject a recently probed flower only if the previous visitor was a cons pecific (honeybees) or congener (bumblebees); the effect Was especiall y marked in honeybees. Honeybees rejected more than 80% of flowers pro bed by conspecifics less than 20 s previously, but less than 20% of fl owers probed by bumblebees less than 20 s previously. Only if the prev ious bee was a conspecific or congener did the probability of a bee pr obing a flower increase with the time since the last probing visit. Ot herwise, the probability of a bee probing was independent of the time elapsed since the last visit. Bees' reactions to flowers whose nectar content had been manipulated independently of prior visits suggested t hat bees were repelled from flowers by species- or genus-specific chem ical cues deposited by previous bees. Laboratory studies elsewhere hav e reported that honeybees are repelled from artificial feeders by vola tile bee-deposited chemicals. My results constitute strong evidence th at such cues are used by nectar-collecting honeybees in the field, and also suggest that bumblebees respond to similar cues. Calculations sh ow that the ability to detect recently visited flowers may help bees t o make a foraging profit; especially when bee densities are high. Thus , bee-deposited chemicals may confer information and economic advantag es to foraging alongside conspecifics or congeners. (C) 1998 The Assoc iation for the Study of Animal Behaviour.