Reproductive investment in a cleistogamous morph of <Emphasis Type="Italic">Polygonum jucundum</Emphasis> (Polygonaceae)

Cleistogamy, a breeding system with permanently closed and self-pollinated flowers, is expected to assure reproductive success at a lower cost. Previous studies have inferred the occurrence of cleistogamous flowers in Polygonum, but there are no detailed studies on their reproductive investment compared with that of the chasmogamous flowers in this genus. Here, we studied a cleistogamous morph of P. jucundum to investigate the investment in pollen number, tepal and nectary size. The number of pollen grains per flower was counted with a light microscope. Nectaries and perianths were observed via scanning electron microscope and light microscopy, photographed and measured via ImageJ. The perianths of the cleistogamous flowers, as well as the pollen numbers and nectary sizes, were significantly smaller than those of the chasmogamous flowers. The pollen numbers of the CL flowers were seven times lower than those in the CH flowers. The tepal areas of the CL flowers were, on average, approximately 38% those of the CH flowers. The nectary areas of the CH flowers were almost twice those of the CL flowers. In addition, the nectaries of the cleistogamous flowers were degenerated and inconspicuous, in distinct contrast with the well-developed and conspicuous nectaries of the chasmogamous flowers. Self-fertilization was completed and produced seeds. The cleistogamous P. jucundum, compared with chasmogamous individuals, exhibited lower costs in male function, pollinator attraction and reward structure investment. This cleistogamy appears to be favourable for the plant reproduction under suboptimal conditions.     

Evolution of floral dimorphism in a cleistogamous annual, <Emphasis Type="Italic">Impatiens noli-tangere</Emphasis> L. occurring under different environmental conditions

The resource allocation models for the evolution of a mixed production of cleistogamous and chasmogamous flowers were tested in four natural populations of Impatiens noli-tangere L. When we incorporated the estimated parameter values into our basic model, under the assumption that the environmental conditions did not change through the reproductive season, the model predicted the obligate chasmogamy in all four populations. However, including the fluctuation of the fertility of chasmogamous flowers in the model, the seasonal switching of chasmogamous and cleistogamous flower production becomes an evolutionary stable strategy, which explains the reproductive behavior of the individuals in two populations. Only after the relationship between the geitonogamy of chasmogamous flowers and the number of chasmogamous flower production was considered in the model did the simultaneous production of both of the dimorphic flowers become an evolutionary stable strategy. Further, the complete model clarifies the reproductive behavior of the individuals in the other two populations.     

Consequences of a mixed reproductive system in the hog peanut,Amphicarpaea bracteata, (Fabaceae)

Summary The forest annual, Amphicarpaea bracteata L. can reproduce via aerial chasmogamous, aerial cleistogamous, and subterranean cleistogamous flowers. Both plant size and light intensity influenced the utilization of the three modes of reproduction. chasmogamous and aerial cleistogamous flower number and the ratio of chasmogamous flowers to the total number of aerial flowers increased with plant size. The latter demonstrated a shift to xenogamy and outbreeding in larger plants. Light intensity indirectly influenced reproductive modes through its infuence on plant size. Seed set by both types of aerial flowers was low and unrelated to plant size. Subterranean seed number and the total dry weight of subterranean seeds per plant increased with size. The subterranean seeds of Amphicarpaea bracteata are thirty-four times larger than the aerial seeds (fresh weight). Under field conditions, subterranean seeds had greater germination after one year than acrial seeds. The plants arising from subterranean seeds were significantly larger and more fecund than those from aerial seeds. Seeds produced by aerial cleistogamous, hand selfpollinated chasmogamous, and naturally pollinated chasmogamous flowers had equivalent germination rates and produced plants of equal size and fecundity. This suggests that the outbred progeny from chasmogamous flowers have no advantage over the inbred progeny from aerial cleistogamous flowers.     

青藏高原东缘块茎堇菜鳞茎分配的个体大小依赖性

块茎堇菜(Viola tuberifera)为青藏高原特有两型闭锁花植物,属多年生草本,具独特的混和交配系统,既可通过早春开放花异花受精和夏季地上地下闭锁花自花受精有性繁殖,还可通过秋季新鳞茎无性繁殖产生后代。高山环境下,异花受精常因花粉限制而无法正常进行,自花受精和克隆繁殖成为保障植物种群正常繁衍的不二之选,而克隆繁殖更能在植株资源消耗最小的情况下保障子代的存活。该文以青藏高原东缘高寒草甸的混合繁育植物块茎堇菜为研究对象,探索其生长期内鳞茎分配的个体大小依赖性,以及植株如何权衡鳞茎的资源分配以适应个体大小的变化。结果表明：块茎堇菜生活史阶段的鳞茎分配具有个体大小依赖性,鳞茎分配与个体大小呈极显著负幂指数相关关系(P<0.01),个体越大,鳞茎分配越小;反之,个体越小,鳞茎分配越高。即块茎堇菜对鳞茎的资源投入受个体大小的制约,通过鳞茎分配比例的高低响应植株自身资源状况的变化,保障在高寒环境下植物种群的生存和繁衍。该研究结果为高山植物克隆繁殖的生活史进化提供了依据。     

Adapting Strategy of Biennial Endangered <Emphasis Type="Italic">Halenia corniculata</Emphasis> (L.) Cornaz to Environmental Stress and Its Implications for Conservation Plan

Under intensive environmental stress, reported that plants have complementary and positive interactions instead of competition among species. In order to examine whether there as a possible correlation between the performance of Halenia corniculata and the coverage/height of other plants with the same distribution, a total of 30 plots were established at two representative distributional areas. The height, total number of chasmogamous flowers and cleistogamous flowers, their overall plots coverage, and observed plants species on each plot were determined. Our results revealed that the number of chasmogamous and cleistogamous flowers of Halenia corniculata was increased when its height was increased. The performance of Halenia corniculata was found to have a significant correlation with the overall coverage of observed plants species in the plot as well as the height of observed plants in established plots. Our results suggest that the biennial plant Halenia corniculata might be able to choose a site inhabited by other perennial plants in the high-altitude region under environmental stress for appropriate growth and reproduction. One main reason that Halenia corniculata seedlings could be established in regions where other plants are inhabited might be due to the degree of light requirement by the seedling stage is different from that at the adult stage. The size of Halenia corniculata population in South Korea was found to be proportional to that of available sites that could support its life cycle, suggesting that available sites need to be secure in order to maintain its life cycle.     

Reproductive effort and sex allocation strategy in Commelina benghalensis L., a common monsoon weed

Commelina benghalensis L. exhibits variability in both foliar and floral features; every plant bears three types of branches and four types of flowers. The branches are negatively geotropic, positively geotropic and diageotropic. The flowers are uni- or bisexual, chasmogamous and cleistogamous. This variability influences the breeding system as well as resource allocation to male and female functions. The plants allocate c. 15% of their total resources to reproduction, the major part of which (68.9%) is devoted to production of aerial branches. The proportion of reproductive effort (RE) allocated to various branch systems is correlated with the availability of resources at the time of their differentiation. The pollen/ovule (P/O) ratio, female : male biomass ratio and reproductive output vary between different flower and branch types; variation is more pronounced in the latter. These variations notwithstanding, the results are in line with Charnov's sex allocation theory. The cleistogamous flowers of aerial branches are, however, an exception, being male- rather than female-biased. The reason behind the deviation is, in all probability, their recent evolution from chasmogamous flowers.  © 2002 The Linnean Society of London, Botanical Journal of the Linnean Society , 2002, 140 , 403−413.     

Why is cleistogamy a selected reproductive strategy in Impatiens capensis (Balsaminaceae)?

Cleistogamy (self-fertilization in closed flowers) differs from chasmogamy (open-pollinated fertilization) mainly in sustaining selfing. Why numerous species develop both of the reproductive modes on the same individuals has long puzzled biologists. In a novel hypothesis presented here, I propose that cleistogamy could be a means by which inbred lines are created and maintained in natural populations; these lines would continuously experience self-improvement via natural selection and via crosses among lines at the chasmogamous flowers to benefit the populations. Supporting evidence for the hypothesis was found in Impatiens capensis where cleistogamous ovules were fertilized proportionately less (56%) than chasmogamous ovules (67%) in natural populations, but crosses among cleistogamous progeny in the greenhouse led to a nearly 10% increase of fertilized chasmogamous ovules. I established a novel fitness model specific to the cleistogamous species to further examine how various aspects of the mating system affect plant performance. A low inbreeding depression (0.07) was consequently found for the surveyed natural populations of I. capensis , suggesting that the individual-level percentage of cleistogamy and the population-level selfing rate may have evolved in the direction of reducing the overall inbreeding depression. © 2002 The Linnean Society of London, Biological Journal of the Linnean Society , 2002, 75 , 543–553.     

Effects of light and nutrient availability on chasmogamy and cleistogamy in an understory tropical herb,Calathea micans (Marantaceae)

The effects of light and nutrient availability on chasmogamous and cleistogamous flower and fruit production were investigated in an understory tropical herb, Calathea micans (Marantaceae). I censused chasmogamous and cleistogamous flower and fruit production at five permanently marked demographic study sites in Costa Rica, characterized by different successional stages. Using a transplant experiment, I examined whether an increase in light and/or nutrients would increase chasmogamous and/or cleistogamous flower production. In natural populations as well as in the experiment, chasmogamous reproduction increased with light; nutrient availability significantly increased chasmogamous reproduction in the transplant experiment. Field observations indicated that very few plants reproduced and low investment in reproduction might have been due to poor environmental conditions. Larger plants had a higher probability of reproducing, but plant size did not have an effect on the mode of reproduction, chasmogamy vs. cleistogamy. However, the production of new vegetative shoots increased the probability of producing chasmogamous inflorescences. Cleistogamy may allow the plant to reproduce even in conditions of poor habitat quality and assures some seed set even in conditions unfavorable to plant growth.     

Cleistogamy in Scutellaria indica (Labiatae): effective mating system and population genetic structure

Scutellaria indica is a perennial herb with both chasmogamous (CH) and cleistogamous (CL) flowers on the same plant in some populations, and only CL flowers in other populations. Actual seed production by CH and CL flowers was investigated in populations of S. indica. The average seed set of CL flowers was 19 times higher than CH flowers, indicating much greater fertilization success. The CL seeds were also significantly heavier than the CH seeds. However, the resource cost of producing a CH flower was much higher than that of producing a CL flower. The CH flower was approximately seven times larger, and its pollen/ovule ratio was approximately five times higher than flowers. The level and pattern of genetic diversity at both allozyme and random amplified polymorphic DNA (RAPD) levels were consistent with a predominantly selfing system in the species. The average amount of within-population genetic variation was extremely low (A = 1.025, P = 2.36%, HO = 0.001 and HE = 0.008 based on allozyme data, and P = 8.94% and HE = 0.03 based on RAPD data). At the species level, the estimates of total gene diversity (HT) were 0. 101 based on allozyme data and 0.139 based on RAPD data. A very high level of genetic differentiation occurred between populations (allozyme GST = 0.92 and RAPD GST = 0.81). Genetic drift coupled with predominant cleistogamous selfing apparently played the major role in determining the population genetic structure in S. indica. Although the features associated with CH and CL flower and seed production seem to be sufficient for the evolution of complete cleistogamy in S. indica, random fixation of alternative alleles for dimorphic or complete cleistogamy in small populations could maintain the multiple strategy of chasmogamous and cleistogamous reproduction in the species.     

Herbivory and competition interact to affect reproductive traits and mating system expression in Impatiens capensis

As a step toward understanding how community context shapes mating system evolution, we investigated the combined role of two plant antagonisms, vegetative herbivory and intraspecific competition, for reproduction and mating system expression (relative production of selfing, cleistogamous and facultatively outcrossing, chasmogamous flowers and fruits) of Impatiens capensis. In a survey of I. capensis populations, we found that vegetative herbivory and intraspecific competition were positively correlated. In a greenhouse experiment where leaf damage and plant density were manipulated, multispecies interactions had dramatic effects on reproductive and mating system traits. Despite having additive effects on growth, herbivory and competition had nonadditive effects for mating system expression, chasmogamous fruit production, flower number and size, and cleistogamous flower production. Our results demonstrate that competitive interactions influence the effect of herbivory (and vice versa) on fitness components and mating system, and thus antagonisms may have unforeseen consequences for mating system evolution, population genetic diversity, and persistence.     

CLEISTOGAMY IN MICROLAENA POLYNODA (GRAMINEAE): AN EXAMINATION OF SOME MODEL PREDICTIONS

A population of the cleistogamous grass, Microlaena polynoda, was investigated to determine some of the factors responsible for adjusting the balance between reproduction by chasmogamous (CH) and cleistogamous (CL) flowers, the tissue costs associated with the two reproductive modes, the fate of progeny produced by each mode, and the genetic diversity of the progeny. Cleistogamous flower production begins earlier in the season than CH flower production. There is a distinct threshold of low light intensity below which mostly CL flowers are produced. Paternal expenditure per plant is an order of magnitude larger for the CH than the CL component. The opposite relation holds for maternal expenditure. Increased maternal expenditure in the CL component may be due to greater fertilization success and retrieval of paternal costs. Cleistogamous seeds are dispersed later than CH seeds. Following dispersal, the spikelet encloses the CL seed but not the CH seed, and is responsible for inducing dormancy. The ratio of seedlings arising from CH seeds to that from CL seeds in a natural habitat is significantly lower than the ratio of estimated numbers of CH to CL seeds produced. There were no detectable polymorphisms among ten presumptive enzyme loci assayed. Many of the features associated with CH and CL reproduction in M. polynoda are in accord with the theoretical requirements for the evolution of closed flower self-pollination and the maintenance of two distinct methods of reproduction.     

The maintenance of mixed mating by cleistogamy in the perennial violet Viola septemloba (Violaceae)

The production of both potentially outcrossed (chasmogamous) and obligately self-fertilized (cleistogamous) flowers presents a clear exception to the prediction that the only evolutionarily stable mating systems are complete selfing and complete outcrossing. Although cleistogamy has evolved repeatedly, the reason for its stability is not known for any species. We tested the hypothesis that the production of cleistogamous and chasmogamous flowers by a perennial violet constitutes adaptive phenotypic plasticity. We manipulated the season of flowering for each flower type and determined fruit set and the germination percentage of seeds produced by cleistogamous and chasmogamous flowers to test the hypothesis that adaptive plastic response to seasonal environmental variation makes mixed mating stable. Cleistogamous flowers had greater fruit set in all seasons and produced seeds with germination percentages as great as or greater than those from chasmogamous flowers. The consistent advantage of cleistogamous flowers is clearly not consistent with a role of adaptive plastic response to seasonal variation. The biomass cost of seed production by chasmogamous flowers was nearly three times that for cleistogamous flowers. Explaining why chasmogamous flower have not been eliminated by natural selection requires that this difference be balanced by an advantage to chasmogamous flowers that has not yet been identified.     

Cleistogamy as a bet‐hedging strategy in Oxalis acetosella, a perennial herb

1 Phenology and reproduction were studied in three populations of the cleistogamous herb Oxalis acetosella during three growing seasons, in order to see how the balance between chasmogamous (CH) and cleistogamous (CL) reproduction varies with temporal and spatial environmental variation and with plant size. The numbers of CH and CL flower buds, flowers, immature capsules and mature capsules were counted per ramet, ramet sizes were estimated, and mature capsules were collected for seed counts.
2 Both CH and CL flower numbers were positively correlated with ramet size expressed as leaf number, but the correlation was much stronger in the CL phase. CL flower production also varied significantly between years and sites. Thus, CL production in O. acetosella was not independent of resources and climatic variation.
3 When the effects of year and site were taken into account, the probability of an individual flower developing into a mature fruit was not higher in the CL phase than in the CH phase.
4 CL production was affected by fertilization success in the CH phase. Ramets with one or more CH flowers left unfertilized generally produced more CL buds than ramets with all their CH flowers fertilized. The former group also tended to have more seeds per CL capsule.
5 Since reproductive success in the two phases varies in different temporal and spatial patterns, cleistogamy in O. acetosella is considered to be a bet‐hedging strategy optimizing reproductive output in fluctuating environments.
6 The findings of this study are in conflict with the general view on cleistogamy as a fail‐safe mechanism for back‐up seed production, unaffected by variations in resource supply and environmental conditions.     

Reproductive ecology of Viola mirabilis var. subglabra representing intermediate flowering characteristics between stemless and stemmed Viola

Viola (Violaceae) is one of the largest genera in angiosperms. This genus is essentially classified into stemless and stemmed groups based on growth morphology. However, Viola mirabilis var. subglabra is an exception in having intermediate flowering characteristics; cleistogamous (CL) flowers are formed in the axils of stem leaves, whereas chasmogamous (CH) flowers arise from basal rosettes (radical CH (CH(r)) flowers) and also in the axils of the stem (axially CH (CH(a)) flowers). To understand why the pattern of flower production varies in this Viola species, flower production was investigated in 10 Japanese populations from Hokkaido to the western part of Honshu in 2014 and 2015. Furthermore, flower characteristics were also compared between CH(r) and CH(a) flowers in Hokkaido. In this species, the production of CH flowers varied among individuals, and they were categorized into three groups, individuals that produced (i) only CH(r) flowers, (ii) only CH(a) flowers and (iii) both CH flowers. The frequency of these groups differed among populations, but some individuals changed the category between 2014 and 2015. Thus, the production of CH(r) and CH(a) flowers plastically changes depending on individual conditions and/or environmental factors. On the other hand, CH(r) and CH(a) flowers differed in flower size and flowering phenology. These results suggest that two types of CH flowers may play different roles in reproduction in each population, but fruit sets and seed sets did not differ between two types of CH flowers.     

Cleistogamy in the rare high Andean perennial herb Cryptantha capituliflora (Boraginaceae)

 Axial and apical flowers of Cryptantha capituliflora were analyzed with regard to morphology and pollen tube growth to assess the occurrence of cleistogamy. Although intermediate floral forms do occur, cleistogamous flowers were significantly smaller than chasmogamous flowers, had fewer anthers, and showed a distinctive stigmatic surface. Chasmogamous flowers can be cross-pollinated. Nevertheless, the growth of self-pollen tubes in few chasmogamous buds jointly with flower characters suggests that these flowers can probably produce fruits through autonomous selfing. The mean seed number per fruit did not differ between fruits from chasmogamous and cleistogamous flowers. Cleistogamous flowers were only observed in axial inflorescences, which are completely covered by the leaf. Other species of section Cryptantha also show the same trend, with cleistogamous flowers located in the lower half of the stems. This pattern is discussed in relation to dissimilarities in the outcrossing opportunities between flower types within the plant. Received May 22, 2002; accepted November 14, 2002 Published online: March 20, 2003     

INTRA-INFLORESCENCE VARIABILITY IN POLLEN/OVULE RATIOS IN THE CLEISTOGAMOUS SPECIES LAMIUM AMPLEXICAULE (LABIATAE)

A comparative study was made of the various floral forms in the main shoot inflorescence of the cleistogamous species Lamium amplexicaule L. (Labiatae). This study revealed significant differences in pollen/ovule (P/O) ratios between successively produced flowers. The variation in P/O ratio is due to a change in pollen count/flower and not in ovule number. The study included two plant populations, one grown in November when only closed flowers are produced, and the other grown in April when open flower production occurs at the upper nodes. The intra-inflorescence variation in P/O ratio among the spring population showed a correlation between P/O ratio and breeding system. The cleistogamous flowers had low P/O ratios of about 200 while the chasmogamous flowers had higher P/O ratios of about 600.     

Analysis of mating system and genetic structure in the endangered,amphicarpic plant,Lewton’s polygala (<Emphasis Type="Italic">Polygala lewtonii)</Emphasis>

Polygala lewtonii is a federally endangered, amphicarpic plant with a mixed mating system and three types of flowers: (1) aboveground, chasmogamous flowers (i.e., open-pollinated; CH), (2) aboveground, cleistogamous flowers (i.e., closed, selfing; CL) and (3) CL flowers on belowground stems (amphicarpy). Aboveground seeds are ant-dispersed, whereas belowground seeds are spaced across the length of the rhizome. Here, we collected individuals of P. lewtonii at both range-wide and fine geographic scales and genotyped them at 11 microsatellite loci. We analyzed patterns of genetic diversity and structure to understand: (1) the predominant mating system (selfing or outcrossing), (2) the movement of pollen and seeds across the landscape, and (3) the optimal strategy to conserve the full range of genetic variation. P. lewtonii reproduces predominantly by selfing or bi-parental inbreeding, but reproduction occurred through each of the three flower types. Some individuals produced by selfing/inbreeding were tightly clustered spatially, and were likely produced either by belowground flowers or by aboveground flowers with limited seed dispersal. Other selfed/inbred individuals were spatially separated (maximum of 15 m), and were likely produced by aboveground flowers followed by seed dispersal by ants. Fine-scale patterns of genetic structure indicate that some gene flow is occurring among aboveground CH flowers but both pollen and outcrossed seeds are moving limited distances (maximum of 0.5 km). Because genetic variation is structured at a fine spatial scale, protecting many populations is necessary to fully conserve the genetic variation in P. lewtonii. Conservation seed banking, if accompanied by research on seed germination requirements, may also contribute to the effective protection of genetic variation in P. lewtonii.     

High propagule production and reproductive fitness homeostasis contribute to the invasiveness of Lespedeza cuneata (Fabaceae)

Comparative studies of congeneric native and exotic species have proved fruitful in understanding plant traits that foster invasion. Using this approach, we investigate the complex reproductive system of the invasive Lespedeza cuneata (Dum.-Cours.) G. Don in relation to three native congeners in the variable environment of the North American tallgrass prairie. Lespedeza species produce both chasmogamous (CH) and cleistogamous (CL) flowers, and propagate clonally via vegetative buds. Utilizing multiple natural populations over 2 years, we investigated reproductive modes of individuals from bagged and unbagged treatments of each species. We found that L. cuneata produced a mean of five times as many seeds and a significantly greater number of vegetative buds than any native studied, and over twenty times as many CH flowers. Insect visitation significantly affected seed set in CH flowers, though some autonomous CH selfing occurred in all species. The invasive relied relatively less on selfing than the natives and exhibited less variation in reproductive output from both modes of reproduction. We conclude that the diverse reproductive biology and wide regeneration niche of L. cuneata in relation to its native congeners confer a fitness homeostasis that facilitates the successful spread of this invasive under a wide range of conditions.     

VARIATION IN THE DEGREE OF CLEISTOGAMY WITHIN AND AMONG SPECIES OF THE GRASS DANTHONIA

North American species of the grass Danthonia bear distinct chasmogamous and cleistogamous flowers. Four taxa occurring in North Carolina (D. compressa, D. epilis, D. sericea, and D. spicata) were examined for variation in the degree of cleistogamy. The flowering culms bear a single terminal panicle consisting of 4 to 12 spikelets of chasmogamous flowers. In each leaf axil, at the nodes of the flowering culm, is a single spikelet of cleistogamous flowers completely surrounded by the leaf sheath. The percentage of cleistogamous flowers produced on any culm depends on the number of chasmogamous and cleistogamous spikelets and the number of flowers in each type of spikelet. All four characters vary among the taxa examined. Danthonia compressa produces, on the average, 50% cleistogamous flowers; D. spicata, 25% cleistogamous flowers; D. sericea and D. epilis, 5% cleistogamous flowers. The degree of cleistogamy in D. spicata is associated with certain habitat features. Populations from mountain sites, disturbed sites, and non-woodland sites produced higher percentages of cleistogamous flowers than did populations from piedmont, undisturbed, and woodland sites, respectively. Grazing may favor increased cleistogamy because cleistogamous flowers are produced lower on the plant. In North Carolina, the most frequently grazed Danthonia taxa also produce the highest percentage of cleistogamous flowers.     

SPATIAL AND TEMPORAL VARIATION IN CHASMOGAMY AND CLEISTOGAMY IN OXALIS MONTANA (OXALIDACEAE)

Chasmogamous (CH) and cleistogamous (CL) flower production was investigated in natural populations of the perennial herb Oxalis montana in southern Quebec, Canada. Every 10–12 days during two flowering seasons, we recorded the reproductive output of approximately 2,000 plants distributed among five forest sites. The percentage of plants flowering, proportion of flowering plants producing CH and CL flowers, CH and CL flower number per flowering plant, and the proportion of flowers that are CL differed significantly among sites and corresponded to site differences in forest type and habitat quality. Unlike patterns in most CL species, however, reproduction by cleistogamy increased in sites with habitat conditions favorable to plant growth and sexual reproduction, and decreased in less productive sites. Chasmogamous and CL flower production increased with increasing plant size but plant size explained a significantly greater proportion of the variation in CL flower numbers. The percentage of flowering plants producing CH flowers decreased between years while the proportion of CL flowers increased at all sites during the second flowering season. The somewhat unusual aspects of reproduction in Oxalis may stem from its perennial habit which allows use of stored resources in adjusting the balance of cleistogamy versus chasmogamy in different environmental regimes.