Why Plants can be too Attractive -- A Discussion of Measures to Estimate Male Fitness

A direct measurement of paternity (e.g. with genetic markers; Devlin & Ellstrand 1990) is time consuming and only feasible in small populations. Male fitness is, therefore, nearly always 'measured' indirectly as the number of pollinator visits or as the amount of pollen grains removed from the anthers although in a few cases pollen export to other plants was estimated with pollen analogues (Waser & Price 1982; for a recent discussion on estimating male fitness in plants see Stanton et al. 1992). Pollen grain removal is assumed to be closely related to pollen export to other plants and hence to male fitness. The step from pollen removal to pollen export is, however, problematic. We will show that the assumption that pollen grain removal is closely related to pollen export to other plants may not be true, especially if plants with different attractiveness to pollinators are considered Pyke's (1991) statement in a recent letter to Nature reflects the general view of the relationship between a plant's attractiveness to pollinators and its fitness: 'All else being equal, a plant that produces more nectar will have a higher average nectar-standing crop per flower when visited by a pollinator, and will therefore have on average more flowers probed during each visit by a pollinator, receive and transmit more pollen, and produce more seeds (its own and those of other plants that it has fathered).' Because seed number per plant (female fitness) is often not limited by pollination, this view is particularly relevant to the number of seeds a plant has fathered (male fitness). In contrast to general belief, we shall show that plants that provide a higher reward and have flowers that are most frequently visited are not necessarily the most fit: plants can be too attractive even if no energetic costs are involved in nectar production. Increasing attractiveness to pollinators leads to more visits per flower and consequently to more pollen removal from the anthers (e.g. Pyke 1981; Waddington 1981; Zimmerman 1983) but, because of increased geitonogamy, this increased pollen removal will not always lead to higher pollen export to other plants (Charlesworth & Charlesworth 1987). Modelling the effects of attractiveness to pollinators on pollen export