Reproductive interactions between hybridizing irises: analyses of pollen-tube growth and fertilization success

Pollen-tube growth and seed siring ability were measured in crosses between the Louisiana iris species Iris fulva and Iris hexagona and their F₁ and F₂ hybrids. Flowers of the parental species were pollinated with self, outcross intraspecific, and interspecific pollen. Pollen-tube lengths were similar for all three pollen types in I. fulva, but in I. hexagona interspecific pollen tubes were longer than intraspecific pollen tubes. Pollen-tube lengths also differed for F₁ and F₂ flowers pollinated with I. fulva, I. hexagona, and hybrid pollen. For both hybrid classes I. fulva pollen tubes were the shortest while pollen tubes from I. hexagona and hybrids grew the furthest. Mixtures of genetically marked pollen were used to assess the seed siring ability of intra- and interspecific pollen in the parental species by varying the proportion of each pollen type in a replacement series design. For both species, the observed proportions of hybrid seeds were lower than the expected based on the frequency of each pollen type in the mixtures across all treatments. Flowers of I. fulva produced less than 10% hybrid progeny even when 75% of the pollen applied to stigmas was derived from interspecific flowers. The frequency of hybrid seed formation was somewhat greater in I. hexagona, but was still significantly lower than expected across all mixture treatments. Seed set per fruit remained constant across the mixture treatments for both species, but in I. fulva fruit set decreased with an increase in the proportion of interspecific pollen. The data indicate that both pre- and postfertilization processes contribute to discrimination against hybrid seed formation.     

POLLEN-TUBE COMPETITION,SIRING SUCCESS,AND CONSISTENT ASYMMETRIC HYBRIDIZATION IN LOUISIANA IRISES

Postpollination mechanisms can play an important role in limiting natural hybridization in plants. Reciprocal hand pollination experiments were performed to study these mechanisms in two species of Louisiana iris: Iris brevicaulis and I. fulva. Relative pollen-tube growth rates changed significantly through time, with I. fulva tubes increasingly outperforming I. brevicaulis tubes in both conspecific and heterospecific styles. However, this pattern of change in relative performance was a poor predictor of siring success: the majority of seeds sired by both maternal species was conspecific rather than hybrid. Experimental crosses and field studies show consistent asymmetric hybridization in Louisiana irises, with I. fulva being a more successful father and a more selective mother than both I. brevicaulis and a third species, I. hexagona. The cause of this pattern is not yet clear, but the pattern itself is unusual. Typically, short-styled species tend to be less successful in reciprocal crosses than long-styled relatives, but I. fulva has shorter styles than either I. brevicaulis or I. hexagona. The effects of pollen-tube competition, differential fertilization, and selective abortion in causing this pattern of asymmetric hybridization is discussed.     

CPDNA INHERITANCE IN INTERSPECIFIC CROSSES AND EVOLUTIONARY INFERENCE IN LOUISIANA IRISES

Inheritance of chloroplast DNA haplotypes was determined for progeny from interspecific crosses involving Iris fulva and Iris hexagona. Polymerase chain reaction amplification of chloroplast DNA followed by restriction fragment length analysis of the amplification products was used to identify the haplotypes of 213 experimental hybrids. This analysis allowed a test for maternal, paternal, and biparental inheritance in the hybrid offspring. Two of the hybrid progeny possessed haplotypes that were combinations of those present in the pollen and seed parents. One of the offspring possessed only the paternal haplotype. The remaining 210 plants had the haplotypes characteristic of the maternal plant. Chloroplast DNA variation in iris populations has previously been used to infer not only introgressive hybridization between I. fulva and I. hexagona, but also the greater role of direct pollen transfer relative to seed dispersal as the avenue for interspecific gene flow. We reexamined the previous conclusions concerning the mode of introgressive hybridization between I. fulva and I. hexagona in light of the results from the chloroplast DNA inheritance analysis. The low level of paternal and biparental inheritance detected in this analysis suggests that previous analyses using the chloroplast DNA as a seed-specific marker were robust. Furthermore, data concerning barriers to hybridization between I. fulva and I. hexagona suggest that the probability of chloroplast DNA introgression via pollen is low.     

SHADE TOLERANCE AND ITS EFFECT ON THE SEGREGATION OF TWO SPECIES OF LOUISIANA IRIS AND THEIR HYBRIDS

Iris fulva Ker. Gawler and Iris hexagona Walter have overlapping geographic ranges in Louisiana. In areas of overlap hybrids are fairly common. Iris hexagona occupies the borders of freshwater marshes of southern Louisiana while I. fulva can be found farther north along edges of natural levees, canals and swamps. Where the natural levee penetrates the marsh, natural hybridization can occur between I. hexagona and I. fulva. It has been suggested that one principal explanation for the segregation of the two species is that I. fulva grows best in semishade and I. hexagona grows best in full sun. A greenhouse study was conducted using rhizomes collected from the field to test this hypothesis and determine the relative shade tolerance of two natural hybrid types. Iris fulva, I. hexagona, and the two hybrid taxa were grown under 0% (control), 50% (medium shade), and 80% (high shade) reduction of sunlight for 6 months and then harvested. Iris fulva was found to be more tolerant of shading than I. hexagona and the two hybrids. Further, I. fulva was found to grow as well in control as in medium shade. Both hybrid taxa were more shade tolerant than I. hexagona. Iris hexagona was greatly affected by all levels of shade. In general, the results suggest that these hybrids are intermediate to the parental taxa in terms of shade tolerance.     

THE EFFECT OF HABITAT ON PARENTAL AND HYBRID FITNESS: TRANSPLANT EXPERIMENTS WITH LOUISIANA IRISES

We performed transplant experiments with Louisiana irises to test the assumptions of three models of hybrid zone structure: the bounded hybrid superiority model, the mosaic model, and the tension zone model. Rhizomes of Iris fulva, I. hexagona, and F₁ and F₂ hybrids were planted at four sites in southeastern Louisiana in 1994. Wild irises grew at all four sites, but differed in genotypic composition among sites. The sites were characterized by (1) pure I. fulva plants; (2) I. fulva-like hybrids; (3) I. hexagona-like hybrids; and (4) pure I. hexagona plants. The sites differed significantly in light availability, soil moisture and chemical composition, and vegetation. Survival of transplants was high in all sites and did not differ significantly among plant classes. Iris hexagona produced significantly more leaf material than I. fulva at the I. hexagona and I. hexagona hybrid sites. The two species did not differ in leaf production at the I. fulva and I. fulva hybrid sites. Leaf production by both classes of hybrid was as great as, or significantly greater than, both parental classes in all sites. Iris hexagona rhizomes gained mass in the I. hexagona and I. hexagona hybrid sites, but lost mass in the I. fulva and I. fulva hybrid sites. Iris fulva rhizomes lost mass in all sites. There were no significant differences in rhizome growth among classes at the I. fulva site. At all other sites, F₁ rhizomes grew significantly more than all other classes except for I. hexagona at the I. hexagona hybrid site. There were no significant differences among classes in the production of new ramets. Overall blooming frequencies were 30% for I. fulva, 10% for F₁s, 3% for F₂s, and 0.7% for I. hexagona. Blooming frequency did not differ among sites for I. fulva, but significantly more F₁s bloomed at the I. hexagona site than at the I. fulva site. These results are inconsistent with all three models of hybrid zone structure. They suggest that once rhizomes become established, hybrids can reproduce by clonal growth as successfully as parents in all habitats, and can outperform them in some habitats. Clonal reproduction may ensure the long-term survival of early generation hybrids and allow the establishment of introgressed populations, despite the fact that F₁ hybrids are rarely produced in nature.     

NATURAL HYBRIDIZATION BETWEEN IRIS FULVA AND IRIS HEXAGONA: PATTERN OF RIBOSOMAL DNA VARIATION

Genetic variation associated with the natural hybridization of Iris fulva and I. hexagona was investigated to test for the occurrence of introgression. These species have been viewed as a classic example of the process of introgressive hybridization (Anderson, 1949). However, more recent studies have concluded that there has not been an exchange of genetic material between I. fulva and I. hexagona (Randolph et al., 1967). Our analysis has involved the examination of both allopatric and parapatric populations of I. fulva and I. hexagona with reference to diagnostic ribosomal DNA markers. The pattern of variation in the parapatric population indicates the presence of the repeated backcrossing necessary to the process of introgressive hybridization. Indeed, in the region of parapatry, we suggest that localized introgression of ribosomal sequences has occurred into both I. fulva and I. hexagona. Significantly, we have also detected the presence of the diagnostic ribosomal markers from each species in allopatric populations of the alternate species. Our findings suggest that not only is introgressive hybridization presently occurring in parapatry between I. fulva and I. hexagona, but that past hybridization between these species has resulted in introgression into areas of allopatry.     

HYBRID FITNESS IN THE LOUISIANA IRISES: ANALYSIS OF PARENTAL AND F1 PERFORMANCE

The assumption of hybrid inferiority is central to the two models most widely applied to the prediction of hybrid zone evolution. Both the tension zone and mosaic models assume that natural selection acts against hybrids regardless of the environment in which they occur. To test this assumption, we investigated components of fitness in Iris fulva, I. hexagona and their reciprocal F₁ hybrids under greenhouse conditions. The four cross types were compared on the basis of seed germination, vegetative and clonal growth, and sexual reproduction. In all cases, the hybrids performed as well as, or significantly better than, both of their parents. These results suggest that F₁ hybrids between I. fulva and I. hexagona are at least as fit as their parents. The results of this study are therefore inconsistent with the assumptions of both the tension zone and mosaic models of hybrid zone evolution.     

NEUTRAL GENE FLOW ACROSS SINGLE LOCUS CLINES

We study the strength of the genetic barrier to neutral gene flow in a general one-locus, two-allele model that includes viability selection as well as fertility selection and premating isolation. We have separately considered adult migration and pollen migration. Our theoretical results suggest that selection against hybrid formation in the form of fertility selection or assortative mating is more effective in preventing introgression of neutral alleles than is reduction in hybrid viability. We argue that in experimental studies of introgression testing of F₁ hybrids is as important as testing of parental forms. To illustrate the utility of this approach, we estimate the strength of the barrier to neutral gene flow between Piriqueta caroliniana and P. viridis, between Iris hexagona and I. fulva, and between Chorthippus brunneus and C. biguttulus.     

Integrating Bayesian genomic cline analyses and association mapping of morphological and ecological traits to dissect reproductive isolation and introgression in a Louisiana Iris hybrid zone

Hybrid zones provide unique opportunities to examine reproductive isolation and introgression in nature. We utilized 45,384 single nucleotide polymorphism (SNP) loci to perform association mapping of 14 floral, vegetative and ecological traits that differ between Iris hexagona and Iris fulva, and to investigate, using a Bayesian genomic cline (BGC) framework, patterns of genomic introgression in a large and phenotypically diverse hybrid zone in southern Louisiana. Many loci of small effect size were consistently found to be associated with phenotypic variation across all traits, and several individual loci were revealed to influence phenotypic variation across multiple traits. Patterns of genomic introgression were quite heterogeneous throughout the Louisiana Iris genome, with I. hexagona alleles tending to be favoured over those of I. fulva. Loci that were found to have exceptional patterns of introgression were also found to be significantly associated with phenotypic variation in a small number of morphological traits. However, this was the exception rather than the rule, as most loci that were associated with morphological trait variation were not significantly associated with excess ancestry. These findings provide insights into the complexity of the genomic architecture of phenotypic differences and are a first step towards identifying loci that are associated with both trait variation and reproductive isolation in nature.     

IRIS NELSONII (IRIDACEAE): ORIGIN AND GENETIC COMPOSITION OF A HOMOPLOID HYBRID SPECIES

Nuclear and chloroplast DNA variation was assayed for two populations of Louisiana irises (Bayou Teche and Young's Coulee) that demonstrated extreme morphological variation and for a sample of the putative hybrid species, Iris nelsonii. The genetic markers examined in this analysis were diagnostic for either Iris fulva. Iris hexagona, or Iris brevicaulis. These data demonstrated that the two morphologically variable populations were hybrid associations involving all three of these species and that all three of these species were involved in the origin of I. nelsonii. The distribution of genetic variation in I. nelsonii was significantly different from that present in either of the two hybrid populations. I. nelsonii demonstrated significantly fewer foreign markers than the two hybrid populations. This finding is in accord with the prediction that I. nelsonii is a hybrid species that has undergone stabilization with regard to genetic recombination and segregation. Although the genotypic makeup of I. nelsonii was significantly different from other parental and hybrid populations, individual plants from this species cannot be unequivocally differentiated from either I. fulva or certain hybrid genotypes. This reflects the paradoxical nature of genotypic variation in hybrid species. Thus, a hybrid species may include genotypes that overlap with both parental and contemporary hybrid populations. In the case of I. nelsonii it is necessary to utilize additional information (morphological, chromosomal, ecological) to identify plants belonging to this taxon. One hybrid population (Young's Coulee) is suggested as a paradigm for the progenitor population that gave rise to I. nelsonii.     

Pollen competition between two sympatric Orchis species (Orchidaceae): the overtaking of conspecific of heterospecific pollen as a reproductive barrier

The frequency of hybrid formation in angiosperms depends on how and when heterospecific pollen is transferred to the stigma, and on the success of that heterospecific pollen at fertilising ovules. We applied pollen mixtures to stigmas to determine how pollen interactions affect siring success and the frequency of hybrid formation between two species of Mediterranean deceptive orchid. Plants of Orchis italica and O. anthropophora were pollinated with conspecific and heterospecific pollen (first conspecific pollen then heterospecific pollen and vice versa) and molecular analysis was used to check the paternity of the seeds produced. In this pair of Mediterranean orchids, competition between conspecific and heterospecific pollen functions as a post‐pollination pre‐zygotic barrier limiting the frequency of the formation of hybrids in nature. Flowers pollinated with heterospecific pollen can remain receptive for the arrival of conspecific pollen for a long time. There is always an advantage of conspecific pollen for fruit formation, whether it comes before or after heterospecific pollen, because it overtakes the heterospecific pollen. The conspecific pollen advantage exhibited in O. italica and O. anthropophora is likely to result from the reduced germination of heterospecific pollen or retarded growth of heterospecific pollen tubes in the stigma and ovary. Overall, the results indicate that our hybrid zone represents a phenomenon of little evolutionary consequence, and the conspecific pollen advantage maintains the genetic integrity of the parental species.     

Mating system and the hybridization between self-compatible Phlox cuspidata and self-incompatible Phlox drummondii

Mating system can impact the frequency of hybridization and therefore the maintenance of species diversity. I evaluate the effects of weak self-incompatibility (SI) in Phlox cuspidata and SI in Phlox drummondii on mating success within species and on hybridization dynamics between species under controlled conditions. The effects of SI on hybridization frequency were assessed by manipulating the relatedness of conspecific pollen and the relative timing of pollen deposition in mixed-donor interspecific pollinations. Selfing as opposed to outcrossing increased hybridization by 16% in P. cuspidata maternal plants and by 48% in P. drummondii maternal plants because self pollen did not compete as well against heterospecific pollen. The relative timing of conspecific versus heterospecific pollen deposition also impacted hybridization. In self-compatible P. cuspidata, the deposition of self pollen 5 h earlier than heterospecific pollen decreased hybridization by 28%. In self-incompatible P. drummondii, a 5 h delay in the deposition of compatible conspecific pollen increased hybridization by 32%. In this hybrid system, early self-pollination can decrease hybridization (but increase inbreeding) by P. cuspidata maternal plants, and SI may increase hybridization by P. drummondii maternal plants.     

ASSORTATIVE MATING AND NATURAL SELECTION IN AN IRIS HYBRID ZONE

The phenology of different genotypes and the distribution of genetic variation among flowering plants and their progeny were examined to assess the levels of assortative mating and selection in a hybrid population of Iris. This study and a previous survey of RAPD nuclear markers and chloroplast markers indicate that the population consists of parental genotypes and recombinant hybrid genotypes that are similar to the parental species (I. fulva and I. brevicaulis), although lacking intermediate genotypes. Early in the season only I. fulva genotypes produced flowers, but as flowering in these plants decreased, the hybrid genotypes and I. brevicaulis genotypes began flowering, resulting in a 24-d period of coincidental flowering. The genotypic distribution of seeds produced during the period of flowering overlap contained a high frequency of intermediate genotypes that were not present in the adult generation. The degree of effective assortative mating was examined by comparing the observed progeny genotypic distributions with expected distributions from a mixed-mating model. The model included selfing and random outcrossing to the nearest plants that had pollen-bearing flowers on the day the recipient flower was receptive. The observed genotypic distribution of progeny from plants with I. brevicaulis chloroplast DNA (cpDNA) was not significantly different from the expected distribution. For I. fulva genotypes, however, there were higher than expected frequencies in the extreme genotypic classes, although intermediate genotypes were absent, indicating that these plants were preferentially mating with similar genotypes. Compared with the extreme genotypes, a larger proportion of the intermediate seed progeny produced were aborted, indicating that intermediate genotypes have lower viability. On the basis of the observed progeny genotypes and genetic disequilibria estimates for the adults and the progeny, there appears to be a pattern of effective asymmetrical mating in this population. This asymmetry is most likely due to pollen-style interactions that reduce the fertilization ability of genetically dissimilar pollen, or preferential abortion of genetically intermediate zygotes by I. fulva-like genotypes. The lack of any apparent discrimination by I. brevicaulis-like genotypes creates a directional exchange of nuclear genetic elements that will have implications for introgression and the evolution of hybrid genotypes.     

Reproductive isolation and hybrid pollen disadvantage in Ipomopsis

One cause of reproductive isolation is gamete competition, in which conspecific pollen has an advantage over heterospecific pollen in siring seeds, thereby decreasing the formation of F1 hybrids. Analogous pollen interactions between hybrid pollen and conspecific pollen can contribute to post-zygotic isolation. The herbaceous plants Ipomopsis aggregata and I. tenuituba frequently hybridize in nature. Hand-pollination of I. aggregata with pollen from F1 or F2 hybrids produced as many seeds as hand-pollination with conspecific pollen, suggesting equal pollen viability. However, when mixed pollen loads with 50% conspecific pollen and 50% hybrid pollen were applied to I. aggregata stigmas, fewer than half of the seeds had hybrid sires. Such pollen mixtures are frequently received if plants of the two species and F1 and F2 hybrids are intermixed, suggesting that this advantage of conspecific over hybrid pollen reduces backcrossing and contributes to reproductive isolation.     

Asymmetrical conspecific seed-siring advantage between Silene latifolia and S. dioica

Background and Aims

Silene dioica and S. latifolia experience only limited introgression despite overlapping flowering phenologies, geographical distributions, and some pollinator sharing. Conspecific pollen precedence and other reproductive barriers operating between pollination and seed germination may limit hybridization. This study investigates whether barriers at this stage contribute to reproductive isolation between these species and, if so, which mechanisms are responsible.

Methods

Pollen-tube lengths for pollen of both species in styles of both species were compared. Additionally, both species were pollinated with majority S. latifolia and majority S. dioica pollen mixes; then seed set, seed germination rates and hybridity of the resulting seedlings were determined using species-specific molecular markers.

Key Results

The longest pollen tubes were significantly longer for conspecific than heterospecific pollen in both species, indicating conspecific pollen precedence. Seed set but not seed germination was lower for flowers pollinated with pure heterospecific versus pure conspecific pollen. Mixed-species pollinations resulted in disproportionately high representation of nonhybrid offspring for pollinations of S. latifolia but not S. dioica flowers.

Conclusions

The finding of conspecific pollen precedence for pollen-tube growth but not seed siring in S. dioica flowers may be explained by variation in pollen-tube growth rates, either at different locations in the style or between leading and trailing pollen tubes. Additionally, this study finds a barrier to hybridization operating between pollination and seed germination against S. dioica but not S. latifolia pollen. The results are consistent with the underlying cause of this barrier being attrition of S. dioica pollen tubes or reduced success of heterospecifically fertilized ovules, rather than time-variant mechanisms. Post-pollination, pre-germination barriers to hybridization thus play a partial role in limiting introgression between these species.     

Determining population structure and hybridization for two iris species

Identifying processes that promote or limit gene flow can help define the ecological and evolutionary history of a species. Furthermore, defining those factors that make up “species boundaries” can provide a definition of the independent evolutionary trajectories of related taxa. For many species, the historic processes that account for their distribution of genetic variation remain unresolved. In this study, we examine the geographic distribution of genetic diversity for two species of Louisiana Irises, Iris brevicaulis and Iris fulva. Specifically, we asked how populations are structured and if population structure coincides with potential barriers to gene flow. We also asked whether there is evidence of hybridization between these two species outside Louisiana hybrid zones. We used a genotyping‐by‐sequencing approach and sampled a large number of single nucleotide polymorphisms across these species' genomes. Two different population assignment methods were used to resolve population structure in I. brevicaulis; however, there was considerably less population structure in I. fulva. We used a species tree approach to infer phylogenies both within and between populations and species. For I. brevicaulis, the geography of the collection locality was reflected in the phylogeny. The I. fulva phylogeny reflected much less structure than detected for I. brevicaulis. Lastly, combining both species into a phylogenetic analysis resolved two of six populations of I. brevicaulis that shared alleles with I. fulva. Taken together, our results suggest major differences in the level and pattern of connectivity among populations of these two Louisiana Iris species.     

Pollinator behaviour and the evolution of Louisiana iris hybrid zones

Pollinator preference may influence the origin and dynamics of plant hybrid zones. Natural hybrid populations between the red‐flowered Iris fulva and the blue‐flowered Iris brevicaulis are found in southern Louisiana. The genetic structure of these populations reflects a lack of intermediate genotypes. We observed pollinator behaviour in an experimental array with five plants each of I. fulva, I. brevicaulis, their F₁, and the first backcross generation in each direction, to obtain data on flower type preferences and transitions between flower types. The most abundant visitors were Ruby‐throated Hummingbirds (Archilochus colubris) and workers of the bumblebee Bombus pennsylvanicus. Hummingbirds visited I. fulva twice as often as I. brevicaulis and visited hybrids at intermediate frequencies. Bumblebee workers preferred the purple‐flowered F₁s and visited plants of I. fulva and the backcross to I. fulva more often than I. brevicaulis and its backcross. Overall, F₁ flowers were visited most frequently. Both hummingbirds and bumblebees visited nearest neighbours in almost 80% of the interplant movements. This meant that a majority of movements were between different flower types, rather than between plants of the same type. Findings from the present study suggest that pollinator preference is not a major causal factor for the lack of intermediate genotypes in natural iris hybrid populations. Instead, pollinator behaviour in our array promoted mixed mating between flower types belonging to different pollination syndromes. However, owing to predominant nearest‐neighbour visitation, the spatial distribution of parental and hybrid genotypes (in concert with pollinator behaviour) will have a strong influence on mating patterns and thus the genotypic structure and evolution of Louisiana iris hybrid zones.     

DIFFERENTIAL POLLEN-TUBE GROWTH RATES AND NONRANDOM FERTILIZATION IN HIBISCUS MOSCHEUTOS (MALVACEAE)

The prevalence of nonrandom fertilization due to postpollination events has rarely been studied in natural populations, despite important implications for outcrossing rates, mate choice, and plant fitness. Nonrandom paternity within fruits can be caused by both unequal fertilization and unequal embryo abortion. Using self-compatible Hibiscus moscheutos, we studied the potential for nonrandom fertilization by comparing growth rates of pollen-tubes from different donors. The branched style of Hibiscus allowed within-flower comparisons between pollen donors. Relative pollen-tube growth rates were determined by applying pollen from pairs of donors to different stigmas on adjacent stylar branches. We then measured the number of callose plugs per tube in cross-sectional transects across the style after 3 hr. We demonstrate that rates of callose plug formation can be used as a sensitive indicator of relative pollen-tube growth rate. Differences between pollen donors were common and repeatable. Self-pollen-tubes grew slower than outcross pollen-tubes in some crosses and faster in others. Allozyme variation in glucose phosphate isomerase was used to show that individuals with fast-growing pollen-tubes sired a disproportionate number of seeds following mixed pollinations (up to 72%). Since seed abortion was negligible, we conclude that variation in pollen-tube growth rates leads to nonrandom paternity within fruits.     

Reproductive significance of indirect pollen-tube growth in Dalechampia (Euphorbiaceae)

Most species of Dalechampia have expanded stigmatic surfaces that extend from the stylar tip part way down the sides of the elongated styles. Pollen grains landing at the stylar tip grow tubes directly down the style to the ovary. Pollen grains landing on the lateral stigmatic surfaces, however, exhibit indirect pollen-tube growth: the tubes grow first to the stylar tip, bend 180°, and then grow to the ovary. Indirect pollen-tube growth appears to be reproductively important: more than half the arriving pollen lands on the lateral stigmatic surfaces, and virtually all of these grains produce tubes that grow indirectly to the ovary; pollen on the lateral surface fertilizes seeds about as effectively as pollen at the tip (with direct tube growth). Indirect pollen-tube growth may be the result of a correlated response to selection for expanded stigmatic surfaces in Dalechampia, or it may be an adaptation to increase the intensity of pollen competition.