EXCESS FLOWER PRODUCTION AND SELECTIVE FRUIT ABORTION - A MODEL OF POTENTIAL BENEFITS

Flowering plants often produce more flowers than fruits. An initial '' excess'' of flowers, although making no numerical contribution to frui t set, may indirectly increase female reproductive success by allowing selective maturation of fruits of superior quality. I use a framework based on order statistics to assess the potential fitness benefit fro m this ''wider choice'' mechanism. The analysis shows that a floral su rplus with subsequent selective abortion can generate large increases in mean female fitness. However, marginal fitness returns always dimin ished as the floral surplus increased (i.e., the fitness gain curve wa s always saturating), and imperfect selectivity of abortion could seve rely mute the advantages of surplus flowers. If the mating environment creates low variance in quality among developing fruits, then little benefit is derived from surplus flowers, while a high variance allows large fitness gains but with rapidly saturating benefits. The results imply that selection on flower number due to wider choice could be ver y strong in some circumstances, but that selection through this mechan ism may often favor only a modest number of excess Rowers.