POSTDISPERSAL SIBLING COMPETITION AND THE EVOLUTION OF SINGLE-SEEDEDNESS IN CRYPTANTHA FLAVA.

Abstract

Listen

Translate article

The evolution of indehiscent (unopening) single-seeded fruits dehiscent, multiseeded fruits has apparently occurred many times among flowering plants (Stebbins 1967; Cronquist 1968; Casper and Wiens 1981; Uma Shaanker et al. 1988). In some taxa, single-seededness is achieved through a reduction in ovule number, whereas in others more than one ovule is present, but only one matures into a seed. The repeated evolution of single-seededness together with indehiscence suggests strong selection pressure for single-seeded dispersal units (Cronquist 1968; Casper et al. 1992). Single-seeded dispersal units could have several advantages. For wind-dispersed species in which the entire fruit serves as the diaspore, a reduction in seed number can increase dispersal distance (Augspurger and Hogan 1983; Augspurger 1986), which is important when the probability of postdispersal seed predation and/or seedling mortality increases with local seed densities or proximity to the parent plant (Janzen 1970; Connell 1971; Howe and Smallwood 1982). However, because the association of single-seededness and indehiscence also occurs among species dispersed by animals (Casper et al. 1992), additional, more-general fitness advantages are likely. Packaging only one seed per fruit could minimize predispersal seed predation (Herrera 1984) or compensate for increases in the number of fruits or seed size (Adams 1967; Marshall et al. 1985). Uma Shaanker et al. (1988) suggest that the abortion of all but one ovule results from competition among the developing offspring and may not be in the best interest of the maternal plant. This study tests another possibility-that sibling competition arising when more than one seed germinates in the same location is an important factor selecting for single-seeded dispersal units (Casper and Wiens 1981; Casper et al. 1992). Sibling competition is used here to mean a density-dependent reduction in performance among interacting siblings relative to the performance of siblings when they are not interacting, the definition favored by Cheplick (1992). Sibling competition has also been applied to the relative intensity of competition among genetically similar individuals compared with competition among individuals less closely related (Williams and Mitton 1973; Williams 1975; Maynard Smith 1978). Although both forms of sibling competition may be relevant to some evolutionary problems, the first usage more often is considered important in the evolution of some dispersal and seed dormancy traits (Hamilton and May 1977; Venable and Lawlor 1980; Schoen and Lloyd 1984; Ellner 1986; Quinn and Engel 1986), whereas the latter form is more often hypothesized to be a factor in the evolution of sexual breeding systems and outcrossing (Williams and Mitton 1973; Williams 1975; Maynard Smith 1978; Taylor 1979; Bulmer 1980). Empirical studies, which have addressed mostly the second hypothesis (e.g., Ellstrand and Antonovics 1985; Schmitt and Antonovics 1986; Willson et al. 1987; Kelley et al. 1988; McCall et al. 1989; Tonsor 1989), differ widely in their findings (Cheplick 1992). In this study, competition among seedlings from the same dispersal unit is examined in the semidesert, herbaceous perennial Cryptantha flava. Like most members of the Boraginaceae, C. flava has four ovules; although all four ovules often initiate embryos, usually only one (sometimes two) develops into a mature nutlet (seed). At least eight congeneric species exhibit this fixed reduction in seed number, and in all cases, the nutlet is persistent within a highly pubescent calyx, which serves as the dispersal unit and is functionally analogous to an indehiscent fruit. The unit is dispersed by wind, but oneand two-