A hypothesis is presented which states that flower production in hermaphroditic flowering plants is primarily controlled by male function. The male function hypothesis predicts a lower fruit—to—flower ratio for hermaphrodites as compared to monoecious or dioecious plants. The hypothesis also predicts that self—compatible hermophrodites should exhibit a higher percent fruit—set than self—incompatible hermaphrodites. These predictions are supported by fruit—set data complied from the literature. An alternative hypothesis relating fruit—set to the probability of self—fertilization also predicts low fruit—set for hermaphrodites as compared to monoecius or dioecious plants. The self—incompatibility hypothesis is tested and rejected on the basis of fruit—set patterns in self—incompatible andromonoecious, self—incompatible monoecious, and self—compatible monoecious species. The effect of the male function hypothesis on current ideas concerning low fruit—set in hermaphrodites is then examined.

I compared ten life history traits (vegetative reproduction, breeding system, compatibility, pollination system, shade tolerance, habitat, life span, life form, morphology, and toxicity) from two existing databases for the 19,960 plant species that occur in the USA. I used two-way tests of independence to determine if there were significant life history traits that distinguish weeds from non-weeds, exotic weeds from native weeds, and invasive exotic weeds from non-invasive exotic weeds. Life span was the most significant life history trait for weeds in general; weeds were more likely to be annuals and biennials and less likely to perennials than non-weeds. In addition, vegetative reproduction, breeding system, compatibility, shade tolerance, and life form were related to life span. Annual and biennial weeds (whether native, exotic, or exotic invasives) were more likely to be wetland adapted, armed, and toxic than annual or biennial non-weeds. Perennial weeds (whether native, exotic, or exotic invasives) were less likely to be forbs or subshrubs, and more likely to be wetland adapted, toxic, shade intolerant, grasses, vines and trees than perennial non-weeds. Exotic annual and perennial weeds were less likely to be wetland species than native weeds, but more likely to be wetland species than non-weeds. Invasive exotic weeds, in contrast, were less likely to be forbs and more likely to be perennial, monoecious, self-incompatible, and trees and than non-invasive exotics.

No abstract

Fruit-set values for 447 species of plants were examined for variations due to compatibility, breeding system, life form, latitude, type of fruit, and type of pollination. The results indicate significant differences between self-compatible and self-incompatible species in terms of 1) average fruit-set and 2) the effect of the independent variables. Breeding system, life form, and latitude were found to be the only significant independent variables for self-incompatible species, while fruit type and latitude were found to be significant for self-compatible plants. Although latitude was a significant variable for self-compatible and self-incompatible species, it was correlated with other variables and may actually have no direct effect on fruit-flower ratios in plants. Hypotheses explaining the effect of the significant independent variables on fruit-flower ratios were then examined and, when possible, predictions were tested. The results indicate that the selective-abortion and bet-hedging hypotheses may be important factors contributing to the low fruit-set in self-incompatible plants, while the male-function hypothesis may explain the lower fruit-flower ratios in hermaphrodites.