Indexed on: 01 Apr '93Published on: 01 Apr '93Published in: The Botanical review; interpreting botanical progress
We obtained information on dichogamy and other aspects of the biology of over 4200 species of angiosperms from several hundred published and unpublished sources. We used this information to describe patterns of occurrence of dichogamy and to test specific hypotheses relating dichogamy to other characteristics of plants or their environments.Protandry was more common than protogyny at the intrafloral level, but the reverse was true at the interfloral level. Patterns of dichogamy varied significantly among major taxa, with protogyny more common among monocotyledons and primitive dicotyledons, and protandry expecially common in the Asteridae. Arctic species tended to be less dichogamous and more protogynous than temperate and tropical species. Aquatic and alpine species were especially protogynous. Patterns of dichogamy varied among sexual systems, with gynomonoecious and gynodioecious species especially protandrous, and monoecious species highly protogynous. Autogamous and self-compatible species were disproportionately protogynous. Flowers of intraflorally dichogamous species were slightly larger than those of adichogamous species, owing to the presence of many autogamous species in the latter group. Species with interfloral protogyny bore much smaller flowers than did species with interfloral protandry. Early-blooming species in north-temperate and polar regions were disproportionately protogynous. Sexual structures that abscised, shriveled or moved after completion of their function tended to be presented first, and those that facilitated the other sexual function were presented second. A negative association existed between type of intrafloral and interfloral dichogamy in diclinous species. Most animal-pollinated flowers were protandrous, except beetle-pollinated and refuge and trap blossoms. Wind pollination was markedly associated with protogyny. Vertical inflorescences visited by upwardly-moving vectors were protandrous.Our results suggest that three primary factors may be involved in promoting dichogamy: selection for avoidance of pollen-pistil interference, selection for avoidance of self-fertilization, and selection for synchrony of pollen discharge and stigma receptivity in the different flower types of diclinous species. In contrast to many earlier workers we reject the thesis that avoidance of self-fertilization is the universal or even the most important force in the evolution of most forms of dichogamy.We attribute the prevalence of intrafloral protandry to selection for avoiding interference between pollen export and pollen receipt. Intrafloral protogyny was associated with imprecise pollen transfer, where other means of avoiding pollen-pistil interference (e.g., herkogamy) are likely to be of limited value. The prevalence of interfloral protogyny seems to reflect the smaller size of unisexual flowers than bisexual flowers, the absence of intrafloral pollen-pistil interference in diclinous species, and selection for synchrony of pollen discharge from one flower type with stigma receptivity in the other.