Transgenic GFP as a molecular marker for approaches to quantify pollination mechanism and gene flow in Arabidopsis thaliana

Arabidopsis thaliana is commonly regarded as a self-pollinated plant. We observed that the stigma in each flower of A. thaliana cannot be pollinated by its own pollen in the early phases of the flowering process, when the anthers had dehisced but the filaments were still too short for the pollen to be deposited on the stigma. In the later stages, after elongation of the filaments, self-pollination can occur. After artificial pollination of the flower of a wild plant with GFP transgenic pollen grains in earlier stages of flowering, GFP expressed within epidermal cells was detected in some of the offspring (26.1-57.1 %). Wind-mediated pollen dispersal was poor but is likely to exist in natural habitats, while insects were observed visiting flowers of A. thaliana in natural and experimental populations. We constructed an experimental population consisting of 28 GFP transgenic plants and 240 wild plants and examined gene flow in the population. The result was that the distance of gene flow was limited to 0.5 m. 22 offspring with expressed GFP were found in 28,299 filial individuals examined, which suggested a relatively low outcrossing rate (0.74%). We conclude that outcrossing in populations of A. thaliana is mainly due to insect pollination. The data on gene flow could be useful to assess the ecological hazards of experimental transgene combinations.     

Subdivision and haplotype structure in natural populations of Arabidopsis lyrata

High levels of inbreeding are expected to cause a strong reduction in levels of genetic variability, effective recombination rates and in adaptation compared with related outcrossing populations. We examined patterns of DNA polymorphism at five nuclear loci and one chloroplast locus within and between four populations of the outcrossing plant Arabidopsis lyrata, a close relative of the highly self-fertilizing model species A. thaliana. The observed patterns are compared with species-wide polymorphism at orthologous loci, as well as within- and between-population patterns at other studied loci in A. thaliana. In addition to evidence for much higher average within-population diversity, species-wide levels of silent polymorphism are generally higher in A. lyrata than in A. thaliana, unlike the results from a previous study of the ADH locus. However, polymorphism is also low in the North American A. lyrata subspecies lyrata compared with the European subspecies petraea, suggesting either a population bottleneck in North American populations or recent admixture involving diverged European populations. Differentiation between the two subspecies is strong, although there are few fixed differences, suggesting that their isolation is recent. Estimates of intralocus recombination rates and analysis of haplotype structure in European A. lyrata populations indicate lower recombination than predicted based on the variability together with physical recombination rates estimated from A. thaliana. This may be due to strong population subdivision, or to recent departures from demographic equilibrium such as a bottleneck or population admixture. Alternatively, there may be consistently lower recombination rates in the outcrossing species. In contrast, estimates of recombination rates from species-wide samples of A. thaliana are close to the values expected assuming a high rate of self-fertilization. Complex population histories in both A. thaliana and A. lyrata complicate theoretical predictions and empirical tests of the effects of inbreeding on polymorphism and molecular evolution.     

Rates and patterns of molecular evolution in inbred and outbred Arabidopsis

The evolution of self-fertilization is associated with a large reduction in the effective rate of recombination and a corresponding decline in effective population size. If many spontaneous mutations are slightly deleterious, this shift in the breeding system is expected to lead to a reduced efficacy of natural selection and genome-wide changes in the rates of molecular evolution. Here, we investigate the effects of the breeding system on molecular evolution in the highly self-fertilizing plant Arabidopsis thaliana by comparing its coding and noncoding genomic regions with those of its close outcrossing relative, the self-incompatible A. lyrata. More distantly related species in the Brassicaceae are used as outgroups to polarize the substitutions along each lineage. In contrast to expectations, no significant difference in the rates of protein evolution is observed between selfing and outcrossing Arabidopsis species. Similarly, no consistent overall difference in codon bias is observed between the species, although for low-biased genes A. lyrata shows significantly higher major codon usage. There is also evidence of intron size evolution in A. thaliana, which has consistently smaller introns than its outcrossing congener, potentially reflecting directional selection on intron size. The results are discussed in the context of heterogeneity in selection coefficients across loci and the effects of life history and population structure on rates of molecular evolution. Using estimates of substitution rates in coding regions and approximate estimates of divergence and generation times, the genomic deleterious mutation rate (U) for amino acid substitutions in Arabidopsis is estimated to be approximately 0.2-0.6 per generation.     

Cross-species microsatellite markers for elucidating population genetic structure in Arabidopsis and Arabis (Brassicaeae)

Species closely related to model organisms present the opportunity to efficiently apply molecular and functional tools developed by a large research community to taxa with different ecological and evolutionary histories. We complied 42 microsatellite loci that amplify under common conditions in four closely related Arabidopsis: A. thaliana; A. halleri; A. lyrata ssp. lyrata; and A. lyrata ssp. petraea, as well as in one more distantly related crucifer; Arabis drummondii. Variation at these loci is amenable to a diversity of applications including population genetics, phylogeographical analyses, mapping of inter and intraspecific crosses, and recombination mapping. Our analysis of microsatellite variation illustrates significant differences in population genetic parameters among three Arabidopsis species. A population of A. thaliana, an inbreeding annual plant associated with disturbed habitats, was highly monomorphic (P = 8% percent polymorphic loci) and only 0.2% heterozygous for 648 locus-by-individual combinations. A population of the self-incompatible perennial herb, A. halleri, was more genetically variable (P = 71%) and had an excess of heterozygosity that may reflect a recent population bottleneck associated with human-mediated founder events. A population of the self-incompatible perennial herb, A. lyrata ssp. petraea, was even more genetically variable (P = 86%) and appeared to be at mutation-drift equilibrium. Population structure estimated from neutrally evolving loci provides an empirical expectation against which hypotheses of adaptive evolution at functional loci can be tested.     

Deleterious mutation in related species of the plant genus Amsinckia with contrasting mating systems

Theory for the evolution of modifiers of the rate of mutation suggests that a lower rate of mutation may evolve after the breakdown of mechanisms that enforce outcrossing. Mutation accumulation (MA) experiments were conducted to compare deleterious mutation parameters in two closely related species of the plant genus Amsinckia, a group that exhibits wide variation in the mating system. One of the two species studied (A. douglasiana) is predominantly outcrossed in natural populations, where as the other species (A. gloriosa) is predominantly self-pollinated. Progeny assays of flower number per plant from generation 1 lines (control) and generation 11 lines (MA treatment) were conducted in both species. Dry weight measurements of progeny from the control and MA treatment in A. douglasiana also were made. Estimation of mutation parameters was conducted using maximum likelihood under the assumption of a gamma distribution of mutational effects. The two species exhibited similar rates and effects of deleterious mutation affecting flower number. Estimates of mutation rate for dry weight in A. douglasiana are close to those for flower number. Overall, the estimates of mutation parameters observed in these species are intermediate within the range reported for fitness components in other eukaryotes. The results are discussed within the context of evolutionary change in deleterious mutation accompanying mating system evolution and with respect to previous estimates of mutation parameters based on assays of inbreeding depression and the assumption of mutation-selection equilibrium.     

Impact of mating systems on patterns of sequence polymorphism in flowering plants

A fundamental challenge in population genetics and molecular evolution is to understand the forces shaping the patterns of genetic diversity within and among species. Among them, mating systems are thought to have important influences on molecular diversity and genome evolution. Selfing is expected to reduce effective population size, Ne, and effective recombination rates, directly leading to reduced polymorphism and increased linkage disequilibrium compared with outcrossing. Increased isolation between populations also results directly from selfing or indirectly from evolutionary changes, such as small flowers and low pollen output, leading to greater differentiation of molecular markers than under outcrossing. The lower effective recombination rate increases the likelihood of hitch-hiking, further reducing within-deme diversity of selfers and thus increasing their genetic differentiation. There are also indirect effects on molecular evolutionary processes. Low Ne reduces the efficacy of selection; in selfers, selection should thus be less efficient in removing deleterious mutations. The rarity of heterozygous sites in selfers leads to infrequent action of biased conversion towards GC, which tends to increase sequences' GC content in the most highly recombining genome regions of outcrossers. To test these predictions in plants, we used a newly developed sequence polymorphism database to investigate the effects of mating system differences on sequence polymorphism and genome evolution in a wide set of plant species. We also took into account other life-history traits, including life form (whether annual or perennial herbs, and woody perennial) and the modes of pollination and seed dispersal, which are known to affect enzyme and DNA marker polymorphism. We show that among various life-history traits, mating systems have the greatest influence on patterns of polymorphism.     

Unexpectedly high outcrossing rate in both dense and sparse patches in self-compatible Pedicularis rex (Orobanchaceae)

Pedicularis rex, a self-compatible alpine herb, depends exclusively on pollinators for reproduction. Previous study has revealed that 74.7 % of genetic variability occurs among populations in Yunnan, indicating a mixed mating system for the species. The large floral display also indicates high potential geitonogamy. Therefore, plant density and display size may be important factors influencing the realized mating system of P. rex. In the study reported here we explored the effects of plant density and floral display size on pollinator movements and thus outcrossing rate in two patches of in P. rex at low and high plant densities. We examined foraging behavior of bumblebees, measured reproductive outputs under hand and natural pollination, and estimated the outcrossing rates using RAPD markers. Floral display size was significantly larger and bumblebees visited significantly more flowers in sequence on individual plants in the sparse patch than in the dense patch. We found a significant positive correlation between floral display size and successive visits in the sparse patch, but not in the dense patch. Regression analysis revealed a significant inverse correlation between fruit production and flower production per plant with a theoretical maximum of 63 fruits per plant, but no relationship between seed production and flower production per plant. Contrary to previous studies on genetic structure, however, we found an unexpectedly high outcrossing rate in both the sparse and the dense patch (t _m = 1.151 and 0.924, respectively). Resource limitation may intensify seed competition and result in selected seed abortion, which may largely explain the unexpectedly high outcrossing rate.     

How multilocus genotypic pattern helps to understand the history of selfing populations: a case study in Medicago truncatula

The occurrence of populations exhibiting high genetic diversity in predominantly selfing species remains a puzzling question, since under regular selfing genetic diversity is expected to be depleted at a faster rate than under outcrossing. Fine-scale population genetics approaches may help to answer this question. Here we study a natural population of the legume Medicago truncatula in which both the fine-scale spatial structure and the selfing rate are characterized using three different methods. Selfing rate estimates were very high ( approximately 99%) irrespective of the method used. A clear pattern of isolation by distance reflecting small seed dispersal distances was detected. Combining genotypic data over loci, we could define 34 multilocus genotypes. Among those, six highly inbred genotypes (lines) represented more than 75% of the individuals studied and harboured all the allelic variation present in the population. We also detected a large set of multilocus genotypes resembling recombinant inbred lines between the most frequent lines occurring in the population. This finding illustrates the importance of rare recombination in redistributing available allelic diversity into new genotypic combinations. This study shows how multilocus and fine-scale spatial analyses may help to understand the population history of self-fertilizing species, especially to make inferences about the relative role of foundation/migration and recombination events in such populations.     

Fine-scale genetic structure of life history stages in the food-deceptive orchid Orchis purpurea

In natural plant populations, fine-scale spatial genetic structure can result from limited gene flow, selection pressures or historical events, but the role of each factor is in general hard to discern. One way to investigate the origination of spatial genetic structure within a plant population consists of comparing spatial genetic structure among different life history stages. In this study, spatial genetic structure of the food-deceptive orchid Orchis purpurea was determined across life history stages in two populations that were regenerating after many years of population decline. Based on demographic analyses (2001-2004), we distinguished between recruits and adult plants. For both sites, there was no difference in the proportion of polymorphic loci and expected heterozygosity between life history stages. However, spatial autocorrelation analyses showed that spatial genetic structure increased in magnitude with life history stage. Weak or no spatial genetic structure was observed for recruits, whereas adult plants showed a pattern that is consistent with that found in other species with a predominantly outcrossing mating system. The observed differences between seedlings and adults are probably a consequence of changes in management of the two study sites and associated demographic changes in both populations. Our results illustrate that recurrent population crashes and recovery may strongly affect genetic diversity and fine-scale spatial genetic structure of plant populations.     

Effects of self-pollination and outcrossing with cultivated plants in small natural populations of American ginseng, Panax quinquefolius (Araliaceae)

For rare plants, self-pollination and inbreeding can increase in small populations, while unusual levels of outcrossing can occur through restoration efforts. To study both inbreeding and outcrossing, we performed experimental pollinations using Panax quinquefolius (American ginseng), a wild-harvested plant with a mixed mating system. For inbreeding, plants were either cross-pollinated within the population or self-pollinated, which resulted in a higher proportion of seeds from self-pollinated flowers. For outcrossing, wild plants were either cross-pollinated within the population or with cultivated plants from West Virginia or Wisconsin. Offspring of all crosses were followed for 4 yr. Two-yr-old seedlings from self-pollination had 45% smaller leaf areas and 33% smaller heights relative to those from cross-pollination. Leaf area is a positive predictor of longer-term survival in wild populations. Our results suggest inbreeding depression, which is unexpected in this self-fertile species. Seedlings from crosses with cultivated plants had 127% greater leaf area and 165% greater root biomass relative to outcrosses within the population. The accelerated growth suggests genetic differences between wild and cultivated populations, but outbreeding depression may not appear until later generations. Assessment of the ultimate fitness consequences of introducing cultivated genotypes requires monitoring over longer time periods.     

Multilocus genomics of outcrossing plant populations

The structure and organization of natural plant populations can be understood by estimating the genetic parameters related to mating behavior, recombination frequency, and gene associations with DNA-based markers typed throughout the genome. We developed a statistical and computational model for estimating and testing these parameters from multilocus data collected in a natural population. This model, constructed by a maximum likelihood approach and implemented within the EM algorithm, is shown to be robust for simultaneously estimating the outcrossing rate, recombination frequencies and linkage disequilibria. The algorithm built with three or more markers allows the characterization of crossover interference in meiosis and high-order disequilibria among different genes, thus providing a powerful tool for illustrating a detailed picture of genetic diversity and organization in natural populations. Computer simulations demonstrate the statistical properties of the proposed model. This multilocus model will be useful for studying the pattern and amount of genetic variation within and among populations to further infer the evolutionary history of a plant species.     

Genetic effects of habitat fragmentation in Viola pubescens (Violaceae), a perennial herb with chasmogamous and cleistogamous flowers

The reproductive biology of a plant species is important in the response of populations to habitat fragmentation, especially if plant-pollinator interactions are disrupted. The genetic effects of forest fragmentation were examined in the common understorey herb Viola pubescens, a species that produces self-pollinated cleistogamous (CL) flowers and potentially outcrossing chasmogamous (CH) flowers. Using allozymes, we measured genetic variation in different sized populations. These were located in woodlots of various sizes (0.5-40.5 ha) and distances from one another (0.3-46 km) within the agricultural landscape of central Ohio in the Midwestern United States. Changes in forest cover of each woodlot within the past 180 years were determined from historical sources and aerial photographs. Woodlot and population sizes were significantly and positively correlated with measures of genetic variation (A, P, HO and HE), with variation highest in populations in the largest woodlot population and lowest in the smallest woodlot population. Most large woodlots resulted from fluctuations in forest cover over the past 60 years, while smaller fragments remained the same size. Overall, populations in Crawford County were genetically differentiated from one another (theta = 0.34), but there was no relationship between genetic and geographical distance. Preliminary evidence for a single year indicated a high rate of outcrossing in most populations. Despite the CH/CL reproductive advantage and apparent outcrossing, populations of V. pubescens in small woodlots remain susceptible to potentially detrimental effects of fragmentation such as genetic drift and reduced levels of genetic variation.     

Effective population size and tests of neutrality at cytoplasmic genes in Arabidopsis.

Cytoplasmic genomes typically lack recombination, implying that genetic hitch-hiking could be a predominant force structuring nucleotide polymorphism in the chloroplast and mitochondria. We test this hypothesis by analysing nucleotide polymorphism data at 28 loci across the chloroplast and mitochondria of the outcrossing plant Arabidopsis lyrata, and compare patterns with multiple nuclear loci, and the highly selfing Arabidopsis thaliana. The maximum likelihood estimate of the ratio of effective population size at cytoplasmic relative to nuclear genes in A. lyrata does not depart from the neutral expectation of 0.5. Similarly, the ratio of effective size in A. thaliana is close to unity, the neutral expectation for a highly selfing species. The results are thus consistent with neutral organelle polymorphism in these species or with comparable effects of hitch-hiking in both cytoplasmic and nuclear genes, in contrast to the results of recent studies on gynodioecious taxa. The four-gamete test and composite likelihood estimation provide evidence for very low levels of recombination in the organelles of A. lyrata, although permutation tests do not suggest that adjacent polymorphic sites are more closely linked than more distant sites across the two genomes, suggesting that mutation hotspots or very low rates of gene conversion could explain the data.     

DETERMINANTS OF OUTCROSSING RATE IN A PREDOMINANTLY SELF-FERTILIZING WEED,DATURA STRAMONIUM (SOLANACEAE)

We measured outcrossing rates of several North Carolina populations of the annual weed Datura stramonium including both natural populations and experimental populations in which we manipulated plant spatial arrangement. Because capsules of D. stramonium typically produce hundreds of seeds and we used an easily scored genetic marker for flower and hypocotyl color, we could measure outcrossing rates accurately for both individual plants and single flowers. The population-wide estimates of outcrossing rates were surprisingly low for a species with showy, entomophilous flowers and ranged from 1.9% in an experimental population with a “clumped” spatial arrangement to 8.5% in an experimental population with a “dispersed” arrangement. These low values were not produced by pollinator discrimination among flower color morphs, as determined by outcrossing measurements on test plants of different colors and by direct observations of pollinator behavior. For individual plants and single flowers in the experimental populations, variation in outcrossing rates was significantly affected by such population-wide characteristics as plant spatial arrangement and nightly fluctuations in total floral abundance. However, by far the most important factor was stigma position. Flowers with stigmas above the anthers had significantly higher outcrossing rates than did flowers with overlapping stigma and anthers. The strong effect of floral morphology suggests that the very low population-wide levels of outcrossing in D. stramonium may represent a persistent mixing mating system rather than a transition to complete selfing.     

Constructing a linkage–linkage disequilibrium map using dominant-segregating markers

The relationship between linkage disequilibrium (LD) and recombination fraction can be used to infer the pattern of genetic variation and evolutionary process in humans and other systems. We described a computational framework to construct a linkage–LD map from commonly used biallelic, single-nucleotide polymorphism (SNP) markers for outcrossing plants by which the decline of LD is visualized with genetic distance. The framework was derived from an open-pollinated (OP) design composed of plants randomly sampled from a natural population and seeds from each sampled plant, enabling simultaneous estimation of the LD in the natural population and recombination fraction due to allelic co-segregation during meiosis. We modified the framework to infer evolutionary pasts of natural populations using those marker types that are segregating in a dominant manner, given their role in creating and maintaining population genetic diversity. A sophisticated two-level EM algorithm was implemented to estimate and retrieve the missing information of segregation characterized by dominant-segregating markers such as single methylation polymorphisms. The model was applied to study the relationship between linkage and LD for a non-model outcrossing species, a gymnosperm species, Torreya grandis, naturally distributed in mountains of the southeastern China. The linkage–LD map constructed from various types of molecular markers opens a powerful gateway for studying the history of plant evolution.     

Population genetic structure in the dioecious pioneer plant species Hippophae rhamnoides investigated by random amplified polymorphic DNA (RAPD) markers

Hippophae rhamnoides is an outcrossing pioneer plant species with a severely fragmented distribution. Random amplified polymorphic DNA (RAPD) marker variation was analysed in 10 populations of ssp. rhamnoides and in one population of ssp. mongolica to estimate the amount and distribution of genetic variability. No less than 89.7% of the scorable markers were polymorphic, but few of these were fixed and populations consequently differed mainly by frequency variation of individual markers. Within-population gene diversity was somewhat low for an outcrossing plant species: 0.192 or 0.159 for ssp. rhamnoides , depending on whether it was based on all 156 polymorphic RAPDs or on only those 63 RAPDs that fulfilled the 3/ N criterion. Analysis of molecular variance applied to the ssp. rhamnoides showed only 15% between-population variability, indicating a relatively restricted population differentiation as expected in outcrossing species and shown in several other AMOVA studies. The tendency for island populations to be somewhat more differentiated, and to have less within-population diversity than mainland populations, may indicate an effect of population fragmentation. Genetic distance estimates among populations, obtained with and without pruning of RAPD loci on the basis of the 3/ N criterion, were generally in very good agreement. Cluster analyses and principal coordinate analyses showed populations of ssp. rhamnoides to be rather close, but quite isolated from the single ssp. mongolica population. Genetic and geographical distances between the ssp. rhamnoides populations were not associated, indicating that large-scale geographical and ecotypic differentiation was not reflected in the RAPD profiles.     

The role of breeding system and inbreeding depression in the maintenance of an outcrossing mating strategy in Silene virginica (Caryophyllaceae)

The goal of this study was to understand the interaction among breeding system, mating system, and expression of inbreeding depression in the hermaphroditic, primarily hummingbird-pollinated, iteroparous, short-lived perennial Silene virginica. We performed hand-selfed and hand-outcrossed pollinations in the field, conducted detailed floral observations within individual flowers and plants, and assayed adult tissue from flowering plants for a genetic estimate of population outcrossing rate. We quantified the opportunity for geitonogamy as the proportion of days each plant exhibited simultaneous male and female function, i.e., asynchronous expression of male- and female-phased flowers. Expression of cumulative inbreeding depression based on germination rate and total flower production in the glasshouse was ～40% and was congruent with the estimated high outcrossing rate of 0.89. Floral observations demonstrated strong temporal protandry within each flower (dichogamy) as well as complete spatial separation between male and female function within each flower (herkogamy). On average, 29% of the time there were both male- and female-phased flowers present on an individual plant. We conclude that our estimate of inbreeding depression is compatible with a largely outcrossing mating system and the amount of selfing observed, likely results from geitonogamy. This study illustrates the utility of examining both the causes and the consequences of inbreeding via selfing to provide additional insights into the evolution of plant mating systems.     

RAPD variation within and among natural populations of outcrossing willow-leaved foxglove (Digitalis obscura L.)

 Random amplified polymorphic DNA (RAPD) markers were used to assess levels and patterns of genetic diversity in Digitalis obscura L. (Scrophulariaceae), an outcrossing cardenolide-producing medicinal plant species. A total of 50 plants from six natural populations on the Iberian Peninsula were analysed by six arbitrarily chosen decamer primers resulting in 96 highly reproducible RAPD bands. To avoid bias in parameter estimation, analyses of population genetic structure were restricted to bands (35 of 96) whose observed frequencies were less than 1–3/n in each population. The analysis of molecular variance (AMOVA) with distances among individuals corrected for the dominant nature of RAPDs (genotypic analysis) showed that most of the variation (84.8%) occurred among individuals within populations, which is expected for an outcrossing organism. Of the remaining variance, 9.7% was attributed to differences between regions, and 5.5% for differences among populations within regions. Estimates of the Wright, Weir and Cockerham and Lynch and Milligan F_ST from null-allele frequencies corroborated AMOVA partitioning and provided significant evidence for population differentiation in D. obscura. A non-parametric test for the homogeneity of molecular variance (HOMOVA) also showed significant differences in the amount of genetic variability present in the six populations. UPGMA cluster analyses, based on Apostol genetic distance, revealed grouping of some geographically proximate populations. Nevertheless, a Mantel test did not give a significant correlation between geographic and genetic distances. This is the first report of the partitioning of genetic variability within and between populations of D. obscura and provides important baseline data for optimising sampling strategies and for conserving the genetic resources of this medicinal species. Received: 7 September 1998 / Accepted: 28 November 1998     

Impact of insect pollinator group and floral display size on outcrossing rate

Despite the strong influence of pollination ecology on the evolution of selfing, we have little information on how distinct groups of insect pollinators influence outcrossing rate. However, differences in behavior between pollinator groups could easily influence how each group affects outcrossing rate. We examined the influence of distinct insect pollinator groups on outcrossing rate in the rocky mountain columbine, Aquilegia coerulea. The impact of population size, plant density, size of floral display, and herkogamy (spatial separation between anthers and stigmas) on outcrossing rate was also considered as these variables were previously found to affect outcrossing rate in some plant species. We quantified correlations between all independent variables and used simple and two-factor regressions to determine direct and indirect impact of each independent variable on outcrossing rate. Outcrossing rate increased significantly with hawkmoth abundance but not with the abundance of any of the other groups of floral visitors, which included bumblebees, solitary bees, syrphid flies, and muscidae. Outcrossing rate was also significantly affected by floral display size and together, hawkmoth abundance and floral display size explained 87% of the variation in outcrossing rate. None of the other independent variables directly affected the outcrossing rate. This is the first report of a significant impact of pollinator type on outcrossing rate. Hawkmoths did not visit fewer flowers per plant relative to other pollinator groups but preferred visiting female-phase flowers first on a plant. Both the behavior of pollinators and floral display size affected outcrossing rate via their impact on the level of geitonogamous (among flower) selfing. Given that geitonogamous selfing is never advantageous, the variation in outcrossing rate and maintenance of mixed mating systems in populations of A. coerulea may not require an adaptive explanation.     

Contrasting variation within and covariation between gender-related traits in autogamous versus outcrossing species: Alternative evolutionary predictions

We present several predictions concerning the expression of genetic variation in, and covariation among, gender-related traits in perfect-flowered plant taxa with different breeding systems. We start with the inference that the pollen:ovule (P/O) ratio in obligately autogamous species (in which the ovules in a flower are fertilized only by the pollen it produces) should be under much stronger stabilizing selection than in outcrossing taxa. Consequently, we predict that obligately autogamous taxa should exhibit lower genetic coefficients of variation in the P/O ratio. Nevertheless, genetic variation in both pollen and ovule production per flower might persist within autogamous as well as outcrossing populations. In autogamous taxa, genotypes with relatively few pollen grains and ovules per flower (but producing relatively high numbers of flowers) and genotypes with comparatively high numbers of gametes per flower (but producing relatively few flowers) could co-exist if lifetime flower production is selectively neutral. In contrast, in outcrossers, the maintenance of genetic variation in ovule and pollen production per flower might result predominantly from their ability to maintain variation in phenotypic and functional gender. Given genetic variation in primary sexual traits, we predict that the genetic correlation between investment in male and female gametes per flower should qualitatively differ between selfers and outcrossers. We predict a positive genetic correlation between pollen and ovule production per flower in obligately autogamous taxa, primarily because strong stabilizing selection on the P/O ratio should select against the gender specialists that would be necessary to effect a negative genetic correlation between mean pollen and ovule production per flower. Moreover, the fact that autogamous individuals are 50% female and 50% male means that gender-biased phenotypes cannot be functionally gender-biased, preventing selection from favouring phenotypic extremes. In contrast, in outcrossing taxa, in which functionally male- and female-biased genotypes may co-exist, the maintenance of contrasting genders could contribute to the expression of negative genetic correlations between pollen and ovule production per flower. We discuss these and a number of corollary predictions, and we provide a preliminary empirical test of the first prediction. This revised version was published online in July 2006 with corrections to the Cover Date.