SELECTION ON A FLORAL COLOR POLYMORPHISM IN THE COMMON MORNING GLORY (IPOMOEA PURPUREA): THE EFFECTS OF OVERDOMINANCE IN SEED SIZE.

Abstract

Understanding the processes that maintain genetic variation in natural populations is a major focus of evolutionary biology. Theoretical investigations have shown that more than a dozen processes can explain the maintenance of genetic variation (see Hartl 1980 for a summary), whereas experiments with laboratory populations indicate that at least some of these processes can operate under some conditions (e.g., Dobzhansky 1948; Dobzhansky and Pavlovsky 1953; Ehrman 1967; Powell and Wistrand 1978; Jones and Probert 1980; Cavender and Clegg 1981). Nevertheless, detailed studies of the selective forces acting to maintain genetic variation in natural populations are rare. The study reported here is part of an ongoing examination of the selective forces acting to maintain genetic variation at a locus influencing floral pigment intensity in the common morning glory, Ipomoea purpurea. This locus (designated W) is polymorphic in natural populations throughout the southeastern United States. Plants homozygous for the w allele have white flowers with pigmented rays (i.e., whites). Heterozygous plants have lightly pigmented flowers with dark rays (i.e., lights), and homozygotes for the W allele have darkly pigmented flowers with even darker rays (i.e., darks) (Ennos and Clegg 1983; Epperson and Clegg 1988). In natural populations, frequencies of the white allele generally range from 0 to 0.4, with a mean of about 0.1 (Epperson and Clegg 1986). Previous work has shown that when white (ww) plants are in the minority, they are undervisited by bees (Brown and Clegg 1984; Epperson and Clegg 1986; Rausher et al. 1993). This undervisitation results in higher selfing rates for white plants, without an associated reduction in the contribution of whites to the outcross pollen pool (i.e., no pollen discounting, Rausher et al. 1993; Iwao 1995). Moreover, inbreeding depression is minimal or absent in North American populations that have been examined (Pear 1983; S. M. Chang, pers. comm. 1996). These observations indicate that the white allele should enjoy a transmission advantage, and thus increase in frequency, when rare (Fisher 1941) and may explain the protection of the white allele in natural populations. It is unclear, however, what forces operate to protect the dark allele. In the absence of any other selective forces acting on the W locus, a combination of genetic drift and active protection of the white allele is expected to result in the fixation of the white allele. The high frequency of the dark allele (0.6-1) in most natural populations therefore suggests the action of other selective forces favoring the dark allele.