SPECIES RICHNESS WITHIN FAMILIES OF FLOWERING PLANTS

Variation in species and genus richness among families of flowering pl ants was examined with respect to four classification variables: geogr aphical distribution, growth form, pollination mode, and dispersal mod e. Previous studies have estimated rates of species proliferation from age and contemporary diversity. Here we found that the earliest appea rances in the fossil record are correlated with contemporary familial species richness, abundance in the fossil record, and the independent variables considered in this analysis. Thus, we believe that the fossi l record does not provide reasonable estimates of the ages of families and that the rate of species proliferation cannot be calculated from such data without bias. Accordingly, our subsequent analyses were base d on contemporary species richness of families. Although the classific ation variables were interrelated, each made largely independent contr ibutions to familial species richness. Cosmopolitan families were 5.6 times more species-rich than strictly tropical families and 35 times m ore species-rich than strictly temperate families. Families including both herbaceous and woody growth forms were 5.7 and 14 times more spec ies-rich than families with either growth form alone. Although animal pollination was significantly associated with elevated familial specie s richness, the effect was statistically weak. The most prominent effe ct was that families with both abiotic and biotic dispersal had more t han 10 times as many species as families with either dispersal mode al one. Our analyses also revealed that families having both dispersal mo des were more likely to have several growth forms, suggesting that evo lutionary flexibility of morphology may be generalized over diverse as pects of life history. These results do not support the idea that poll ination and dispersal by animals were primarily responsible for the tr emendous proliferation of angiosperm species, either by producing popu lation structures conducive to speciation or by applying selection for diversification. Instead, the importance of varied dispersal mode, gr owth form, and climate zone in predicting high familial species richne ss suggests that a capacity to diversify morphologically and physiolog ically may have been primarily responsible for high rates of species p roliferation in the flowering plants.