繁殖保障和延迟自交的研究进展

尽管植物在进化过程中面临不利自花授粉的选择,但许多植物仍维持混合的授粉机制。繁殖保障假说是解释自交进化的最重要因子之一,一直是植物生殖生态学和进化生物学关注的焦点之一。概述了近年来的主要研究热点及其进展,包括自交进化的遗传和生态机制及理论模型探讨、繁殖保障假说的提出、验证自交能否提供繁殖保障的例证、延迟自交的类型及延迟自交能否提供繁殖保障的例证等方面。介绍了我国在繁殖保障和延迟自交方面研究的现状和不足之处,结合国际上研究繁殖保障假说的发展趋势已由单季节、单种群、单因子的研究阶段过渡到多季节、多种群、多因子（自交方式及其所占比例、花粉折损、种子折损、自交率和近交衰退）的综合研究阶段,及由传统的、经典的研究方法过渡到应用现代实验手段（如SSR、SNP等分子标记）和先进仪器设备的研究阶段,提出今后研究中应注意的问题。有必要借用多学科（植物学、生态学和分子生物学）的方法及手段进行不同物种的对比和综合细致的研究。     

Evolution of the Selfing Rate and Resource Allocation Models

Abstract First, evolutionary theories of selfing of terrestrial plants are reviewed briefly. The evolution of the selfing rate is controlled mainly by (1) the benefit of enhanced genetic relatedness to seeds and (2) the cost of lowered fitness of selfed offspring (inbreeding depression), being modified by (3) fertility assurance under pollen limitation, (4) reduced performance as pollen donor, (5) reduced expenditure to male function, and (6) lowered genetic recombination. Models of the joint evolution of selfing and inbreeding depression predict either strong outcrossing or predominant selfing. Although wind-pollinated plants fit the prediction, some animal-pollinated species have intermediate selfing rates, refuting the theory.
Second, three resource allocation models are analyzed, in which an individual plant optimally allocates limited resources to outcrossed seeds, selfed seeds, and to energy reserves for the next year. The first model explains how the number of outcrossed and selfed offspring change with plant size when they differ in dispersal distance. The second model predicts that, in a disturbed habitat, the plant is likely to be annual and to produce both selfed and outcrossed seeds; in contrast, in a stable habitat, the plant tends to be perennial and to abort selfed seeds selectively. Hand pollination may increase seed production for perennials but not for annuals. The third model explains the observed difference between animal and wind pollinated plants in the out-crossing rate pattern by the difference in the way pollen acquisition increases with investment.     

Relationship between reproductive assurance and mixed mating in perennial Kosteletzkya virginica

Recent works have shown that mixed mating systems often evolve despite strong inbreeding depression and reproductive assurance, which is one of the widely accepted explanations for the evolution of selfing. However, there have been few empirical studies on the relationship between mixed mating and reproductive assurance in perennial plants. In the herbaceous perennial, Kosteletzkya virginica, delayed selfing induced from context-dependent style curvature offers reproductive assurance, and adverse weather conditions significantly reduce pollinator visitation rates. In this study, our goals were (i) to experimentally evaluate pollinator failure rate, reproductive assurance, selfing rate and the relationships between them, and (ii) to measure inbreeding depression across multiple growth seasons. Results indicate that both population selfing rates and reproductive assurance are significantly and positively correlated with field estimates of pollinator failure rates, and there is a strong relationship between selfing rates and reproductive assurance. Inbreeding depression across multiple growth seasons ranged from 0.621 to 0.665, and there were no significant differences among different seasons. Our data demonstrates that a mixed mating system is beneficial because frequent pollinator failure has allowed reproductive assurance to evolve through delayed selfing which minimizes the risk of seed discounting and is still advantageous despite high inbreeding depression.     

Evolutionary Assurance vs. Mixed Mating

Reproductive assurance is a widely accepted explanation for the evolution of selfing, although theory suggests that an evolutionarily stable mixed mating strategy does not maximize seed production. We present a correlation analysis involving 28 species representing 23 families showing that selfing can evolve independently of inbreeding depression. We discuss the cost-benefit trade-off of selfing, in particular the incongruence of whether delayed selfing provides reproductive assurance in 22 species representing 14 families, in which pollen and seed discounting are minimized when pollinators or mates are scarce. Reproductive assurance, in response to frequent pollinator failure, can be reconciled with an evolutionarily stable mixed mating system contributed to by delayed selfing, which is still advantageous even if there is strong inbreeding depression.     

Patterns of self compatibility, inbreeding depression, outcrossing, and sex allocation in a marine bryozoan suggest the predominating influence of sperm competition

Sex allocation by simultaneous hermaphrodites is theoretically influenced by selfing rate, which is in turn influenced by the benefits of enhanced genomic transmission and reproductive assurance relative to the cost of inbreeding depression. The experimental investigation of these influences in seed plants has a rich pedigree, yet although such an approach is equally relevant to colonial invertebrates, which globally dominate subtidal communities on firm substrata, such studies have been scarce. We reared self‐compatible genets of the marine bryozoan Celleporella hyalina s.l. in the presence and absence of allosperm, and used molecular genetic markers for paternity analysis of progeny to test theoretical predictions that: (1) genets from focal populations with high selfing rates show less inbreeding depression than from focal populations with low selfing rates; (2) genets whose selfed progeny show inbreeding depression prefer outcross sperm (allosperm); and (3) genets bias sex allocation toward female function when reared in reproductive isolation. Offspring survivorship and paternity analysis were used to estimate levels of inbreeding depression and preference for outcrossing or selfing. Sex allocation was assessed by counting male and female zooids. As predicted, inbreeding depression was severe in selfed progeny of genets derived from the populations with low self‐compatibility rates, but, with one exception, was not detected in selfed progeny of genets derived from the populations with higher self‐compatibility rates. Also, as predicted, genets whose selfed progeny showed inbreeding depression preferred outcrossing, and a genet whose selfed progeny did not show inbreeding depression preferred selfing. Contrary to prediction, sex allocation in the majority of genets was not influenced by reproductive isolation. Lack of economy of male function may reflect the over‐riding influence of allosperm‐competition in typically dense breeding populations offering good opportunity for outcrossing. We suggest that hermaphroditism may be a plesiomorphic character of the crown group Bryozoa, prevented by phylogenetic constraint from being replaced by gonochorism and therefore not necessarily adaptive in all extant clades. © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98 , 519–531.     

Consequences of life history for inbreeding depression and mating system evolution in plants

Many plants are perennial, but most studies of inbreeding depression and mating system evolution focus on annuals. This paper extends a population genetic model of inbreeding depression due to recessive deleterious mutations to perennials. The model incorporates life history and mating system variation, and multiplicative selection across many genetic loci. In the absence of substantial mitotic mutation, perennials have higher mean fitness and lower, or even negative, inbreeding depression than annuals with the same mating system. As in annuals, self fertilization exposes deleterious recessive mutations to selection, increasing mean fitness and decreasing inbreeding depression. Including mitotic mutation decreases mean fitness while increasing inbreeding depression. Perenniality introduces a kind of selective sieve, such that strongly recessive mutations contribute disproportionately to mean fitness and inbreeding depression. In the presence of high mitotic mutation, this selective sieve may provide a mechanistic basis for high inbreeding depression observed in some long lived perennials. Without substantial mitotic mutation, it is difficult to reconcile genetically based models of inbreeding depression with the empirical generalization that perennials outcross while related annuals self fertilize.     

Context-dependent style curvature in Kosteletzkya virginica (Malvaceae) offers reproductive assurance under unpredictable pollinator environments

Delayed selfing is a reproductive strategy that can increase mating opportunities and provide reproductive assurance in the absence of pollinators. When external pollination fails to occur in the herbaceous perennial, Kosteletzkya virginica, the styles curve downward to bring anthers into contact with the stigmal lobes, resulting in delayed selfing. Our goal was to evaluate experimentally the effects of weather conditions on pollinator activity, style curvature, and plant fitness. We discovered that adverse weather conditions significantly reduced pollinator visitation rates, and increased the incidence of style curvature, which resulted in delayed selfing. Furthermore, elimination of delayed selfing by experimental manipulation reduced fruit and seed set compared to non-emasculated flowers, even under adverse weather conditions. Offspring produced through selfing were similar to outcrossed progeny in the majority of fitness traits measured, and inbreeding depression was limited. Our study demonstrates that context-dependent style curvature promotes delayed selfing and provides reproductive assurance in the absence of pollinators, with minimal loss of progeny performance through inbreeding depression.     

Selective seed abortion induced by nectar robbing in the selfing plant Comastoma pulmonarium

? Self-pollination often provides plants with the benefit of reproductive assurance; thus, it is generally assumed that species' interactions that alter floral attractiveness or rewards, such as nectar robbing, will have little effect on the seed production of selfing species. We challenge this view with experimental data from Comastoma pulmonarium, a selfing annual experiencing a high ratio of nectar robbing in the Qinghai-Tibet Plateau. ? We manipulated robbing (robbed or netted) and pollination mode (hand-selfed or hand-outcrossed) in a factorial design and measured the number of developing ovules and mature seeds, together with seed weight and seed germination, in each treatment. ? Robbing decreased the number of mature seeds, but not the number of developing ovules, suggesting a negative influence of robbers through indirect effects via selective seed abortion. We found no evidence for early-acting inbreeding depression, but found later-acting inbreeding depression. Our data also suggested that later-acting inbreeding depression of progeny from robbed flowers could be reduced in comparison with that from unrobbed flowers. ? We suggest that nectar robbing can have both negative and positive effects on the quantity and quality, respectively, of progeny produced in selfing plants, and challenge the view that robbing has no effect on selfing species.     

Does selfing provide reproductive assurance in the perennial herb Bulbine vagans (Asphodelaceae)?

Selection for selfing to provide reproductive assurance depends on the balance between increased reproductive output when pollinators or potential mates are scarce and the extent that inbreeding depression erodes such fertility gains. We use glasshouse and field experiments to examine the benefits of autonomous and facilitated selfing in Bulbine vagans . Autonomous selfing was delayed until after opportunities for outcrossing and reproductive output was 0.67 relative to manual selfing and open pollination. Values less than one probably reflected insufficient autonomous deposition of self pollen. In the field, reproductive output of emasculated flowers was 0.50 relative to intact flowers that could both outcross and self, indicating that outcross pollen was limited and that selfing boosted reproductive output. Because all pollen was removed from anthers before intact flowers closed, facilitated selfing rather than autonomous selfing occurred. In the glasshouse, inbreeding depression was 0.45, but under natural conditions would probably exceed 0.5. Values greater than 0.5 negate the automatic gene transmission advantage afforded by selfing and increasingly erode the benefits of reproductive assurance. We conclude that in B. vagans delayed and facilitated selfing can confer reproductive assurance, providing the latter does not usurp ovules that could be outcrossed.     

Allee effect and self-fertilization in hermaphrodites: reproductive assurance in a structured metapopulation

Reproductive assurance through selfing during colonization events or when population densities are low has often been put forward as a mechanism selecting for the evolution of self-fertilization. Such arguments emphasize on the role of both local demography and metapopulation processes. We developed a model for the evolution of self-fertilization in a structured metapopulation in which local densities are not steady because of population growth. Reproduction by selfing is density-independent (reproductive assurance) but selfed seeds endure inbreeding depression, whereas reproduction by outcrossing is density-dependent (Allee effect). First, we derived an analytical criterion for metapopulation viability as a function of the selfing rate and metapopulation parameters. We show that outcrossers can develop a viable metapopulation when they produce a high amount of dispersal seeds that counterbalances their incapacity to found new populations from low densities. Second, the model shows there is a positive feedback between demography and outcrossing rates, leading to either complete outcrossing or selfing. Specifically, we illustrate that inbreeding depression can paradoxically favor the evolution of selfing because of its negative effect on density. Also, complete outcrossing can be selected despite pollen limitation, although it does not provide a full seed set. This model underlines the influence of the mating system both on demography and gene dynamics in a metapopulation context.     

Life-history plasticity and inbreeding depression under mate limitation and predation risk: cumulative lifetime fitness dissected with a life table response experiment

Environmental effects on the evolution of mating systems are increasingly discussed, but we lack many examples of how environmental conditions affect the expression and consequences of alternative mating systems. Variation in mate availability sets up a trade-off between reproductive assurance and inbreeding depression, but the consequences of both mate limitation and inbreeding may depend on other environmental conditions. Predation risk is common under natural conditions, and known to affect allocation to reproduction, but we know little about the effects of isolation and inbreeding under predation risk. We reared selfed and outcrossed hermaphroditic freshwater snails (Physa acuta) in four environments (predator cues present or absent crossed with mating partners available or not) and quantified life-history traits and cumulative lifetime fitness. Our results confirm that isolation from mates can increase longevity and growth, resulting in higher lifetime fecundity. Thus, we observed no evidence for mate limitation of reproduction. However, reproduction under isolation (i.e., selfing) resulted in inbreeding depression, which should counteract the benefits of selfing. Inbreeding depression in fitness occurred in both predator and no-predator environments, but there was no overall change in inbreeding depression with predator cues. This represents, to our knowledge, the first empirical estimate of the effect of predation risk on inbreeding depression in an animal. Cumulative fitness was most influenced by early survival and especially early fecundity. As predation risk and inbreeding (both ancestral and due to a lack of mates) reduced early fecundity, these effect are predicted to have important contributions to population growth under natural conditions. Therefore life-history plasticity (e.g., delayed reproduction) is likely to be very important to overall fitness.     

Self-pollination and inbreeding depression in Acacia dealbata: Can selfing promote invasion in trees?

The ability to self-fertilise may promote invasiveness in plants by assuring reproduction when mate and pollinator availabilities are inadequate, provided that the benefit of increased fecundity via selfing is not outweighed by inbreeding depression. However, knowledge of breeding systems and inbreeding depression has been lacking for most introduced plants. In this study of the invasive Australian tree Acacia dealbata in its introduced range in South Africa, controlled pollination experiments indicated that the study population was at least partially self-compatible and had a high capacity for autonomous self-pollination. However, we found substantial inbreeding depression, with seeds per fruit, progeny survival and progeny growth being lower after self- than after cross-pollination. Progeny arising from self-pollination also had a higher frequency of certain traits – yellow colouration of leaves and pink or white colouration of stems – which were associated with lower rates of survival. High inbreeding depression in A. dealbata must detract from the reproductive assurance benefit of self-fertilisation, casting doubt on the hypothesis that self-fertilisation contributes to invasiveness in this species. As high inbreeding depression has also been reported in other self-compatible trees, future studies should elucidate whether self-fertilisation contributes to invasiveness of trees by assessing both reproductive assurance benefits and inbreeding depression costs over the lifetime of progeny.     

The cost of fluctuating inbreeding depression

We present a phenotypic model for the evolution of self-fertilization in an infinite population of annual hermaphrodites for the case in which fitness and inbreeding depression vary among generations (e.g., due to fluctuations in the environment from year to year). Conditions for the evolution of selfing, mixed mating, and outcrossing are derived and are compared with results from numerical calculations that assume a normal distribution of inbreeding depression. In contrast to the situation in which inbreeding depression does not vary, when inbreeding depression fluctuates in a stochastic manner among generations with a mean less than 0.5, selfing is not necessarily selected. Thus, fluctuating inbreeding depression can be viewed as an additional cost of selfing that may stabilize mixed mating systems. These results emphasize the need to take into account fluctuating inbreeding depression in empirical studies aimed at understanding mating system evolution in annuals.     

Better than nothing: Evolution of autonomous selfing under strong inbreeding depression in an alpine annual from the Qinghai–Tibet Plateau

In plants with infrequent pollinator services, the benefits of reproductive assurance could be eroded by severe ovule discounting and inbreeding depression (ID). However, it remains unclear how selfing evolves under complete pollinator failure and strong ID. We examined the mating system and ID under netting and robbing conditions in Comastoma pulmonarium (Turcz.) Toyok. (Gentianaceae), an alpine annual experiencing a high ratio of nectar robbery on the Qinghai–Tibet Plateau.Comastoma pulmonarium produced seeds via selfing at the study site. No pollinator was observed and thus the nectar was consumed exclusively by robbers. Inbreeding depression occurred in the life stages of seed mass and germination, and the cumulative ID was much higher than 0.5 under netting and robbing conditions. Overall, in comparison with netting, the magnitude of ID under robbing conditions tended to decrease. Our results suggested that selfing could assure reproduction for plants under complete pollinator failure and strong ID, supporting the “better than nothing” role of selfing and providing one of the few cases of the evolution of selfing under strong ID.     

EVOLUTION OF THE MAGNITUDE AND TIMING OF INBREEDING DEPRESSION IN PLANTS

Estimates of inbreeding depression obtained from the literature were used to evaluate the association between inbreeding depression and the degree of self-fertilization in natural plant populations. Theoretical models predict that the magnitude of inbreeding depression will decrease with inbreeding as deleterious recessive alleles are expressed and purged through selection. If selection acts differentially among life history stages and deleterious effects are uncorrelated among stages, then the timing of inbreeding depression may also evolve with inbreeding. Estimates of cumulative inbreeding depression and stage-specific inbreeding depression (four stages: seed production of parent, germination, juvenile survival, and growth/reproduction) were compiled for 79 populations (using means of replicates, N = 62) comprising 54 species from 23 families of vascular plants. Where available, data on the mating system also were collected and used as a measure of inbreeding history. A significant negative correlation was found between cumulative inbreeding depression and the primary selfing rate for the combined sample of angiosperms (N = 35) and gymnosperms (N = 9); the correlation was significant for angiosperms but not gymnosperms examined separately. The average inbreeding depression in predominantly selfing species (δ = 0.23) was significantly less (43%) than that in predominantly outcrossing species (δ = 0.53). These results support the theoretical prediction that selfing reduces the magnitude of inbreeding depression. Most self-fertilizing species expressed the majority of their inbreeding depression late in the life cycle, at the stage of growth/reproduction (14 of 18 species), whereas outcrossing species expressed much of their inbreeding depression either early, at seed production (17 of 40 species), or late (19 species). For species with four life stages examined, selfing and outcrossing species differed in the magnitude of inbreeding depression at the stage of seed production (selfing δ = 0.05, N = 11; outcrossing δ = 0.32, N = 31), germination (selfing δ = 0.02, outcrossing δ = 0.12), and survival to reproduction (selfing δ = 0.04, outcrossing δ = 0.15), but not at growth and reproduction (selfing δ = 0.21, outcrossing δ = 0.27); inbreeding depression in selfers relative to outcrossers increased from early to late life stages. These results support the hypothesis that most early acting inbreeding depression is due to recessive lethals and can be purged through inbreeding, whereas much of the late-acting inbreeding depression is due to weakly deleterious mutations and is very difficult to purge, even under extreme inbreeding.     

Conservation implications of the reproductive ecology of Agalinis acuta (Scrophulariaceae)

Reproductive ecology of Agalinis acuta was investigated by examining potential for self-fertilization before and at anthesis, reproductive output from outcrossed vs. selfed matings, and effects of browsing, plant size, and conspecific plant density on seed and fruit production. These features of a plant species can provide indirect information pertinent to conservation such as patterns and maintenance of genetic diversity, risk associated with inbreeding depression, and changes in pollinator abundance or effectiveness. The species is self-compatible, with 97% of selfed flowers setting fruit; pollinators were not required for reproduction. However, seed set in self-pollinated fruits averaged 17-20% less than that in open-pollinated fruits. Geitonogamous and facilitated selfing are possible throughout anthesis and autonomous selfing is possible late in anthesis as corollas abscise. Delaying self-pollination until after outcrossing opportunities likely limits selfing rates and thus reduces risks associated with inbreeding but allows reproduction in absence of pollinators. Supplementing pollen on open-pollinated flowers yielded no additional seed set over controls. Neither early-season browsing of primary stems nor conspecific plant density had significant effects on number of fruits per plant, on fruit size, or on number of seeds from open-pollinated flowers. Currently, reproduction appears to be high (about 2400 seeds/plant), and future risks due to lack of genetic diversity are likely low.     

Delayed selfing and resource reallocations in relation to mate availability in the freshwater snail Physa acuta

We study the influence of mate availability on the mating behavior of the self-fertile, preferentially outcrossing freshwater snail Physa acuta. Previous optimization theory indicated that mating system interacts with life-history traits to influence the age at first reproduction, providing three testable predictions. First, isolated individuals should reproduce later than individuals with available mates in the expectancy of finding a partner and avoiding the cost of inbreeding. Second, resource reallocation to future fecundity is needed for such reproductive delays to evolve. Third, the reproductive delay can be optimized with respect to life-history traits (e.g., survival, growth) and the mating system (inbreeding depression). Our results largely validate these predictions. First, reproduction is significantly delayed in isolated individuals ("selfers") as compared with individuals frequently exposed to mates ("outcrossers"). Second, delayed reproduction is associated with reallocation to future growth, survival, and fecundity, although fecundity is also affected by the mating system (selfing vs. outcrossing). Third, the reproductive delay found (approximately 2 wk) is consistent with quantitative predictions from optimization models. The delay is largely heritable, which might be partly explained by among-family differences in the amount of inbreeding depression (mating system) but not growth or survival.     

TRANSMISSION ADVANTAGE FAVORS SELFING ALLELE IN EXPERIMENTAL POPULATIONS OF SELF‐INCOMPATIBLE WITHERINGIA SOLANACEA (SOLANACEAE)

The evolution of self‐fertilization is one of the most commonly traversed transitions in flowering plants, with profound implications for population genetic structure and evolutionary potential. We investigated factors influencing this transition using Witheringia solanacea, a predominantly self‐incompatible (SI) species within which self‐compatible (SC) genotypes have been identified. We showed that self‐compatibility in this species segregates with variation at the S‐locus as inherited by plants in F1 and F2 generations. To examine reproductive assurance and the transmission advantage of selfing, we placed SC and SI genotypes in genetically replicated gardens and monitored male and female reproductive success, as well as selfing rates of SC plants. Self‐compatibility did not lead to increased fruit or seed set, even under conditions of pollinator scarcity, and the realized selfing rate of SC plants was less than 10%. SC plants had higher fruit abortion rates, consistent with previous evidence showing strong inbreeding depression at the embryonic stage. Although the selfing allele did not provide reproductive assurance under observed conditions, it also did not cause pollen discounting, so the transmission advantage of selfing should promote its spread. Given observed numbers of S‐alleles and selfing rates, self‐compatibility should spread even under conditions of exceedingly high initial inbreeding depression.     

Delayed selfing as an optimal mating strategy in preferentially outcrossing species: theoretical analysis of the optimal age at first reproduction in relation to mate availability

The evolution of mating systems and that of life history have usually been modeled separately. However, they may be to some extent coupled in natural situations because they rely on the same phenotypic traits. Here, we focus on one of these traits, the age at first reproduction, in a species able to self- and cross-fertilize. When inbreeding depression is strong, self-fertile species preferentially cross-fertilize. However, outcrossing is not always possible when the availability of sexual partners is limited. The optimal reproductive strategy in this case would be to wait for a sexual partner for a certain period of time (the waiting time) and then switch to selfing if no mates have been encountered (reproductive assurance strategy). We predict the evolution of an optimal waiting time depending on the efficiency of resource reallocation to late fecundity, on the inbreeding depression, and on the instantaneous probability of encountering a partner versus dying. As a consequence of reduced mate availability, intermediate selfing rates can be generated in preferentially outcrossing populations, but they are lowered by the existence of a waiting time. Our model may thus explain low selfing rates observed in natural populations of many self-fertile, preferentially outcrossing plants or animal species.     

EVOLUTION OF UNISEXUALITY IN THE HAWAIIAN FLORA: A TEST OF MICROEVOLUTIONARY THEORY

The evolution of separate sexes as a means of avoiding self-fertilization requires the controversial coexistence of large inbreeding depression and high selfing rate in the ancestral hermaphrodite population. Fitness components of adult females and hermaphrodites in nature, of their open-pollinated progeny, and of experimental selfs and outcrosses onto hermaphrodites were compared in endemic Hawaiian Bidens sandvicensis, all of whose known populations are gynodioecious, consisting of a mixture of females and hermaphrodites. Multilocus selfing rates of hermaphrodites were also estimated, and sex morph ratio monitored over four seasons in three populations of B. sandvicensis and one population of gynodioecious B. cervicata. Total mean inbreeding depression in seed set (in the glasshouse), germination rate (in an open-air nursery on Kauai), and first year survivorship and fecundity in the field were estimated as 0.94 (SE 0.04), and occurred primarily in drought months. Lower survivorship and fecundity of selfs were partially explained by their consistently smaller size. Open-pollinated seed of females had significantly lower germination rate, proportion flowering, and fecundity than outcrossed progeny of hermaphrodites, suggesting moderate biparental inbreeding in females and a lack of any non-outcrossing advantage to progeny of females. In all fitness components, open-pollinated progeny of hermaphrodites were inferior to those of females and to outcrosses, and in most components were superior to selfs. Total performance of open-pollinated progeny of females relative to those of hermaphrodites was calculated as 2.3 (SE = 0.4), but since inflorescences of females also set 20% to 50% more seed than those of hermaphrodites, their total relative ovule success was estimated as 3.2 (SE = 0.5). If inheritance of male sterility is nuclear, this superiority is sufficient to maintain females in frequencies over 20% in populations, whose actual frequencies ranged from 14% to 33%. In four populations, selfing rates of hermaphrodites, assayed in seedlings, were 0.50, 0.45, 0.25, and 0.30, but since substantial inbreeding depression occurred prior to germination, the mean selfing rate of hermaphrodite ovules exceeded 0.57. Female frequencies were significantly higher in the two populations with higher hermaphrodite selfing rate. These results suggest that inbreeding depression can exert a profound influence on the mating system of self-compatible plants on Hawaii and perhaps other oceanic islands, and can be sufficiently strong to electively favor the elimination of the male function.