Seed Dispersal by Animals: a Role in Angiosperm Diversification?

The astounding diversity of angiosperms, and their prevalence in most presentday terrestrial habitats, has often attracted the attention of evolutionary biologists, and a number of factors have been proposed to explain the success and extraordinary radiation of this group, particularly in relation to gymnosperms (see, e.g., stebbins 1974, 1981; Raven 1977; Regal 1977; Doyle 1978; Mulcahy 1979; Burger 1981; Queller 1983; Crepet 1984; Kubitzki and Gottlieb 1984; Doyle and Donoghue 1986). Seed dispersal by animals has been sometimes considered one of the factors helping to explain angiosperm diversification (Baker 1963; Regal 1977; Burger 1981; Crepet 1984; Tiffney 19841, although this hypothesis has gone essentially untested to date. If seed dispersal by animals contributed significantly to angiosperm diversification, one should expect to find (1) that this seed-dispersal method was either a significant innovation of angiosperms (a feature unique to this group) or that it occurs proportionally more often in angiosperms than in gymnosperms; (2) that extant groups exhibiting this feature tend to be taxonomically more diverse than sister groups lacking it; and (3) that biotically dispersed groups were of greater proportional significance during the early, critical periods of angiosperm diversification. The objective of this paper is to test these simple predictions. Seed dispersal by animals (biotic dispersal) takes place through a variety of mechanisms (van der Pijl 19821, including endozoochory (dispersal by vertebrate guts; Janzen 19831, epizoochory (dispersal by adhesion; Sorensen 1986), and dyszoochory (dispersal through scatter hoarding and related behaviors; Vander Wall and Balda 1977; Bossema 1979). I consider only endozoochory, because it seems the most genuine and widespread method of biotic dispersal among extant plants (Ridley 1930; van der Pijl 1982; Howe 1986) and among plants from the more distant past, since "[fruit] flesh has long been the primary attractant of biotic dispersal agents" (Tiffney 1986b, p. 300). Furthermore, endozoochory has been the method of biotic dispersal generally implied in the explanations of angiosperm diversification quoted above. Apart from these considerations, methodological reasons also suggest confining the analyses to endozoochory. The assignment of taxa with imperfectly known dispersal ecologies to seed-dispersal categories must rely on consideration of diaspore morphology. It is for endozoochory that the most reliable inferences