Constraints and Competition in the Evolution of Flowering Phenology

Abstract

We tested whether the flowering times of animal—pollinated plants are influenced by phylogenetic membership and by life form (e.g., annual, perennial, etc.). We analyzed existing data sets on 2298 animal—pollinated angiosperms of North and South Carolina and 1575 animal—pollinated angiosperms of temperate Japan, and also analyzed a null model based upon the Carolina flora. Our analyses of the complete data sets and of subsets including only the largest families showed that: (1) the floras of the Carolinas and temperate Japan have bimodal distributions of flowering times, with peaks in spring and late summer; (2) within each of these floras, families differ very significantly in flowering time; (3) for most families, flowering times in the Carolinas and in Japan are statistically indistinguishable; (4) the sequences of flowering, ordered by family, are also statistically indistinguishable in the two floras; (5) within each flora, skewness of flowering time differs markedly among families; (6) for a typical family, the skewness of flowering time is the same in the two floras; (7) there is a significant negative correlation between skewness and mean family flowering time; and (8) life forms differ in flowering time, though less markedly than families. These results demonstrate that phylogenetic membership and life form strongly influence a species' flowering time. We argue that seasonal limitations of flowering times are caused by phylogenetic constraints, which may not have changed for millions of years. This study does not provide the degree of resolution needed to determine whether or not there is natural selection for alteration of flowering times in these floras, whether by competition or other mechanisms. However, our results do suggest that competition for pollination does not push species' flowering times beyond seasonal boundaries imposed by phylogenetic constraints. The effects of competition on flowering time are probably limited to small and temporary readjustments of the phenological relationships of competitors. Studies of the flowering times of animal—pollinated plants must consider the strong limits to seasonal displacement imposed by phylogeny and life form, and the probable existence of many alternative modes of escape from competition besides seasonal shifts. We consider the implications of recent paleobotanical studies that suggest that temperate plant communities may not have been sufficiently stable through time for plants to have achieved competitive equilibria. Phylogenetic and life history constraints are likely to influence the evolution of many community characters besides flowering time. The methods we have used and the conclusions we have drawn are relevant to a wide variety of ecological characters in plants and animals.