ABSTRACT
Predominantly outcrossing plant species are expected to accumulate recessive deleterious mutations, which can be purged when in a homozygous state following selfing. Individuals may vary in their genetic load because of different selfing histories, which could lead to differences in inbreeding depression among families. Lineage-dependent inbreeding depression can appear in gynodioecious species if obligatory outcrossed females are more likely to produce female offspring and if partially selfing hermaphrodites are more likely to produce hermaphrodites. We investigated inbreeding depression at the zygote, seed, and germination stages in the gynomonoecious-gynodioecious Dianthus sylvestris, including pure-sexed plants and a mixed morph. We performed hand-pollinations on 56 plants, belonging to the three morphs, each receiving 2-3 cross treatments (out-, sib- and self-pollination) on multiple flowers. Effects of cross treatments varied among stages and influenced seed provisioning, with sibling competition mainly occurring within outcrossed fruits. We found significant inbreeding depression for seed mass and germination and cumulative early inbreeding depression varied greatly among families. Among sex morphs, we found that females and hermaphrodites differed in biparental inbreeding depression, whereas uniparental was similar for all. Significant inbreeding depression levels may play a role in female maintenance in this species, and individual variation in association with sex-lineages proclivity is discussed.
ABSTRACT
Female fecundity advantage in gynodioecious plants is required for the spread and maintenance of this reproductive system. However, not all reproductive characters show female advantage in all species. We used a meta-analysis to summarise differences between females and hermaphrodites reported from the literature for several reproductive traits. Further we tested three hypotheses, (1) that female plants of species with many ovules produce more seeds per fruit while those with few ovules produce heavier seeds, (2) that females are more pollen limited than hermaphrodites, and (3) that floral sexual size dimorphism is more pronounced in species with few ovules, either because female reproductive success is less limited by pollen availability in such species or because flowers with few ovules require a smaller floral structure to protect the carpels. Overall, females compared to hermaphrodites produced more but smaller flowers, had higher fruit set, higher total seed production, and produced heavier seeds that germinated better. Species with many versus few ovules differed in female advantage for flower size dimorphism, flower number, fruit set and total seed production. However seed size, seed set per fruit and seed germination differences between females and hermaphrodites did not differ significantly between species with few and many ovules. We also found no evidence for differential pollen limitation between females and hermaphrodites. Degree of floral sexual size dimorphism differed significantly between species with few and many ovules. Though pistillate flowers were generally smaller than those of hermaphrodites, species with many ovules showed less difference in flower size between the sexes, suggesting either that the protective role of the perianth constrains the evolution of sexual size dimorphism in species with many ovules or that selection for adequate pollination in species with many ovules impedes the reduction in flower size of females.
Subject(s)
Adaptation, Physiological , Fertility , Flowers/anatomy & histology , Plants , Fruit , Germination , Pollen , Seeds , Sex CharacteristicsABSTRACT
Some species described as gynodioecious are truly gynomonoecious-gynodioecious. Three distinct phenotypes may be found in their natural populations-female and hermaphrodite pure-sexed plants bearing either only pistillate or perfect flowers, respectively, and mixed plants bearing both types of flowers. In one such species, Dianthus sylvestris, we investigated mating system parameters using allozyme data. Outcrossing rates and correlations of outcrossed paternity were calculated for the three types of plants and separately for pistillate and perfect flowers on mixed plants. The mean outcrossing rate for the population was t(m) ± SD = 0.885 ± 0.032. Females were more outcrossed than hermaphrodites (0.987 ± 0.112 and 0.790 ± 0.076, respectively), whereas mixed plants were not significantly more or less outcrossed than hermaphrodites (0.840 ± 0.060). Within mixed plants, perfect flowers showed an intermediate outcrossing rate (0.898 ± 0.057), whereas pistillate flowers were as selfed as perfect flowers on hermaphrodite plants (0.782 ± 0.111). Family estimates of outcrossing rates were highly variable. Globally, no biparental inbreeding was detected in this species, and there was a mean of 61.5 ± 19.9% of full-sibs within families. Floral dimorphism between small pistillate and large perfect flowers together with pollinator preference for larger flowers could explain the observed patterns for both mating parameters. The advantages of gynomonoecy-gynodioecy are discussed. We conclude that mixed plants do not reduce selfing for all flowers on a plant, but perfect flowers on these plants seem to have an outcrossing advantage.
