EMPTY FLOWERS

Bell (1986) predicted that plants should be able to grow a proportion of their flowers that do not produce any nectar, which thus escape the costs of nectar production, while gaining the benefits of insect pollination because most insect visitors either cannot discriminate, or would lose too much time attempting to do so. He worked out an ESS model predicting the proportion of 'cheating' flowers and discriminating insects: the proportion of cheaters should be D/H, where D is the discrimination time, and H the handling time of the insect visitors. We have tested Bell's hypothesis using the flowers of Cerinthe major L. (Boraginaceae) visited by Anthophora plumipes (Pallas) (Hymenoptera, Anthophoridae). From measurements of D and H, we predicted that 73% of flowers should be cheaters. Inspection of the ranked nectar production of individual flowers on individual days shows that between 75 and 85% are relatively low nectar producers. Reasons for this pattern are explored; in particular, the way in which nectar production varies with flower age may constitute a mechanism by which plants can play a mixed strategy.