Evaluation of pollinator effectiveness based on pollen deposition and seed production in a gynodieocious alpine plant,Cyananthus delavayi

Examining variations in pollinator effectiveness can enhance our understanding of how pollinators and plants interact. Pollen deposition and seed production after a single visit by a pollinator are often used to estimate pollinator effectiveness. However, seed production is not always directly related to pollen deposition because not all pollen grains that are deposited on a stigma are compatible or conspecific. In the field, we tested pollinator effectiveness based on pollen deposition and the resulting seed production after single visits by different pollinator groups in a gynodieocious alpine plant Cyananthus delavayi (Campanulaceae). Our results showed that mean pollen deposition was generally inconsistent with mean seed production when comparisons were performed among different pollinator groups and sexes. In general, the correlations were not significant between pollen deposition and seed production in both perfect and female flowers after single visits by halictid bees, bumble bees, and hoverflies. We suggest seed set of virgin flowers after single visits is a more reliable indicator of pollinator effectiveness than pollen deposition and would be a better indicator of pollinator effectiveness for future studies.     

The distyly in Primula elatior (L.) Hill (Primulaceae), with a study of flowering phenology and pollen flow

Flowers of Primula elatior exhibit all the dimorphic traits typical of distyly. The difference in pollen production between pin and thrum flowers is not thought to be caused by the pollen size dimorphism. The percentage of pin pollen which is probably viable exceeds that of thrum pollen. In the population studied the morph ratio is 1:1, and plants are randomly dispersed with regard to morph type. Bumblebees are the main pollinators, and analysis of the stigmatic pollen load in naturally |xjllinated flowers shows that pin stigmas receive many more pollen grains than thrum stigmas. Intramorph pollination is considerable in pins, while thrums are subject to some differential pollination. Emasculation experiments have shown that intraflower pollination in the thrum morph is negligible. Aging of pin flowers has a reducing effect on their total pollen loads, while in thrums there is no change in total pollen load after the first days of anthesis.     

The function and adaptive significance of tristyly in Pontederia cordata L. (Pontederiaceae)

Darwin proposed that the adaptive significance of tristyly is to promote insect-mediated pollination among floral morphs with anthers and stigmas at equivalent levels (legitimate pollination). Strong pollen trimorphism in Pontederia cordata enables an evaluation of this hypothesis. In an investigation of pollen flow patterns in different parts of the North American range of the species, legitimate pollination of some morphs was observed in all but one population investigated. However at only one site (Ft. McCoy, Florida) was Darwin's hypothesis confirmed for the three floral morphs. The long-styled morph most frequently exhibited legitimate pollination, the mid-styled morph was intermediate, and the short-styled morph rarely experienced legitimate pollination.
Temporal variation in pollen flow was observed during the flowering season of P. cordata . Total stigmatic pollen loads decreased as the density of inflorescences and activity of pollinators declined. In spite of variable total loads the legitimate component was relatively constant in the long- and mid-styled forms, while in the short-styled form this measure increased as the season progressed. Little variation was detected between stigmatic samples collected at different times during a single day.
Legitimate pollination may be inhibited by local foraging of pollinators and the spatial segregation of floral morphs. At Paugh Lake (Ontario) over 75% of all bumble bee flights are among the five nearest neighbours of a particular inflorescence. In the same population, there is a probability of over 70% that the nearest three neighbours of an inflorescence are of the same floral morph. Despite these influences, legitimate pollination occurs in P. cordata , suggesting substantial pollen carryover.     

Heterostylous morph differences in pollen transfer and deposition patterns in Primula sieboldii on a visitation by a queen bumblebee, measured with a semi-natural experimental system

Abstract Under a semi-natural setting the between-morph pollen exchange patterns were studied in distylous Primula sieboldii flowers by measuring pollen removal from the anthers on a single visit by a Bombus diversus tersatus queen, and stigmatic pollen deposition along the sequence of the visitation of the opposite-morph flowers by the bee. Despite the twofold larger number of pollen grains produced in a single flower of the long-styled morph compared to that of the short-styled morph, no significant difference in pollen removal from a flower was found between the morphs. The stigmas of the long-styled morph received significantly more opposite-morph pollen grains than those of the short-styled morph on a single visit by the bee. Sufficient legitimate pollen grains, surpassing the ovule number, were loaded on the stigmas of 27% and 17% of visited flowers of the long- and short-styled morphs, respectively. The short-styled morph could more efficiently donate pollen to the opposite morph stigmas than the long-styled morph.     

Foraging behavior affects pollen removal and deposition in Impatiens capensis (Balsaminaceae)

Flowers of most plant species are visited by a variety of animals. Some of these visitors are effective pollinators while others remove resources without transferring pollen. Studies comparing the effectiveness of different visitors as pollinators often compare taxa without considering variation in behavior within a taxon. Wilson and Thomson (Ecology 72: 1503-1507, 1991) documented the effects of honey bees and bumble bees on the pollination dynamics of Impatiens capensis. They found that pollen-collecting honey bees removed large numbers of pollen grains from anthers but deposited little of it on stigmas; bumble bees, which sought nectar, removed less pollen but deposited more of it on stigmas. It is unclear whether the low pollen transfer efficiencies of honey bees are explained by their morphology or by their pollen-collecting behavior. We repeated the work of Wilson and Thomson at a site where honey bees were foraging for nectar, not pollen. We measured the quantity of pollen remaining in anthers, the number of pollen grains deposited on stigmas, and seed production after single visits by honey bees and bumble bees. The differences between the taxa disappeared when they were foraging in a similar manner. Our results clearly demonstrate the importance of foraging behavior on the pollination effectiveness of floral visitors.     

Pollen flow in the distylous<Emphasis Type="Italic"> Palicourea fendleri</Emphasis> (Rubiaceae): an experimental test of the Disassortative Pollen Flow Hypothesis

The Disassortative Pollen Flow Hypothesis proposed by Darwin postulates that the relative position of anthers and stigmas in distylous flowers enhances pollen flow between flowers of different morphs (legitimate pollination), in comparison to flow between flowers of the same morph (illegitimate pollination). In order to test this hypothesis, we measured pollen transport, mediated by a trained Copper-rumped Hummingbird (Amazilia tobaci), between flowers of the distylous Palicourea fendleri under laboratory conditions. In individual tests, we offered to the hummingbird a pollen donor flower and two emasculated recipient flowers in a controlled sequence. After each foraging bout, we counted the number of pollen grains transported from the donor flower to the stigmas of both recipient flowers. In agreement with Darwin's hypothesis, we found that hummingbirds transport pollen of "pin" flowers in significantly higher numbers to legitimate "thrum" stigmas, even if previously visiting a "pin" flower. However, "thrum" pollen was deposited in greater numbers on illegitimate "thrum" stigmas. We interpret this asymmetry largely as the consequence of floral morphology; pollen flow was greater between anthers and stigmas that exhibit greater spatial matching. In P. fendleri, the position of floral organs along the corolla tube does not always precisely correspond. In our experimental system, the probability that the pollinator extracts a pollen grain from the anther and the probability of self-pollination were both dependent on the type of floral morph. We discuss the relevance of the latter findings in relation to other studies of pollen flow in heterostylous species.     

Temporal changes in calyx tube length of Clematis stans (Ranunculaceae): a strategy for pollination by two bumble bee species with different proboscis lengths

We examined the adaptive significance of a temporal decrease in the calyx tube length of Clematis stans, a dioecious species pollinated by Bombus diversus (long proboscis) and B. honshuensis (short proboscis). We compared visitation frequency, pollen removal, pollen deposition, and fruit set after a single visit among three flower stages, differentiated by calyx tube length. Bombus diversus frequently visited and removed significantly more pollen from long flowers. Bombus honshuensis visited and tended to remove more pollen from short flowers. Both pollinators deposited more pollen in short flowers, resulting in higher fruit set. These results indicate that size correspondence between the proboscis and the calyx tube enhances visitation frequency and pollen removal, but not pollen deposition. Because a single visit does not fertilize all ovules of a flower, multiple visits by two bumble bee species may increase seed production and genetic diversity of offspring. By temporally changing calyx tube length, C. stans can use two bumble bee pollinators and maintain specialized relationships with each. This strategy may be adaptive when the pollinator fauna fluctuates, and is economical because it eliminates costs required to produce different types of flowers. This constitutes a novel pattern of temporal specialization in flower-pollinator relationships.     

Bumble bee behavior and selection on flower size in the sky pilot,Polemonium viscosum

Summary In alpine Polemonium viscosum, plants having sweet-scented flowers are primarily pollinated by queens of the bumble bee species, Bombus kirbyellus. In this paper we ask whether two aspects of the pollination effectiveness of bumble bees, visitation rate and pollination efficiency, vary significantly with flower size in sweet-flowered P. viscosum.(i) Bumble bees visited plants with large flowers on 80–90% of encounters, but visited those with smaller flowers on only 49% of encounters. (ii) However, the gain in pollination that large-flowered plants obtained via increased visitation was countered in part because bumble bees deposited fewer outcross pollen grains per visit on stigmas of large flowers than on those of small ones. When both visitation rate and pollination efficiency are taken into account, the predicted value of a single bumble bee encounter declines from 1.06 seeds for flowers larger than 18 mm in diameter to 0.55 seeds for flowers smaller than 12 mm in diameter. Our results suggest that bumble bee pollinators of P. viscosum prefer flower morphologies that are poorly suited for precise pollination. Such behavioral complexities are likely to place constraints on the evolution of optimal floral design.     

Effectiveness of buzz pollination in Pedicularis chamissonis: significance of multiple visits by bumblebees

Pollination success of plants is highly susceptible to the frequency of visits and foraging behavior of pollinators. Pollination of the nectarless flowers of Pedicularis species depends on bumblebee workers collecting pollen by vibrating the anthers (buzz pollination). However, little is known about the efficiency of the pollination system. Foraging behavior, pollen removal from anthers and pollen deposition on stigmas of P. chamissonis were studied to assess the effectiveness of buzz pollination in an alpine snowbed population of northern Japan. Although bumblebees tended to visit most of the flowers open at a given time within inflorescences during a single visit, pollen removal rate at the first visit was about 20%, and buzzing period decreased with increasing number of previous visits, resulting in a decreasing proportion of pollen removed per visit as the number of visits increased. These trends enable plants to provide pollen for more pollinators. The number of pollen grains deposited on stigmas was not saturated during the first visit and increased with additional visits. Irrespective of weak self-compatibility, evidence of interference between self and outcross pollen was lacking for seed production. Therefore, buzz pollination in P. chamissonis acts as a mechanism that improves the chance of cross-pollination upon multiple visits if pollinator visitation is frequent.     

The timing and effectiveness of sequential pollinations in Hibiscus moscheutos

In many species with animal-pollinated flowers, pollen arrives on the stigma in pulses and late-arriving pollen may be precluded from fertilizing ovules. When seed set per fruit is not pollen limited, the fate of pollen from later cohorts is affected by the time between pollinator visits, variation in pollen tube growth rates, and the amounts of pollen deposited relative to the availability of stigmatic area and unfertilized ovules. In a natural population of Hibiscus moscheutos, we found that consecutive pollinator visits to individual flowers occurred within 15 min of each other in more than half of our observations. We then conducted hand-pollination experiments using equal doses of early and late pollen bearing unique allozyme markers for paternity analysis (each dose was more than sufficient to result in full seed set). When pollen was applied 15 or 30 min after an earlier pollen load, the proportions of seeds sired by late-arriving pollen were reduced by 13–30% and 21–57%, respecitively. A few pollen-tubes from the late pollen load sired seeds even when a large dose of pollen was applied 1–2 h earlier, suggesting that the performance of pollen grains within each cohort was highly variable. The advantage of arriving early was greater when the first pollen load was applied at 0930 hours as compared with 1130 hours. We found no effect of previous pollination on the growth rates of late-arriving pollen-tubes. Taken together, these results demonstrate that pollen from later visits can compete with pollen from earlier visits, but the success of late-arriving pollen typically declines to very low levels after an interval of about 2 h. Given that surplus pollen often arrives on stigmas by mid-morning, we conclude that male reproductive success is more likely to be affected by the timing of pollen dispersal and pollen-tube competitive ability than by the total amount of pollen that is exported from flowers.     

BUMBLE BEE POLLINATION AND FLORAL MORPHOLOGY: FACTORS INFLUENCING POLLEN DISPERSAL IN THE ALPINE SKY PILOT,POLEMONIUM VISCOSUM (POLEMONIACEAE)

Pollen dispersal success in entomophilous plants is influenced by the amount of pollen produced per flower, the fraction of pollen that is exported to other flowers during a pollinator visit, visitation frequency, and the complementarity between pollen donor and recipients. For bumble bee-pollinated Polemonium viscosum the first three determinants of male function are correlated with morphometric floral traits. Pollen production is positively related to corolla and style length, whereas pollen removal per visit by bumble bee pollinators is a positive function of corolla flare. Larger-flowered plants receive more bumble bee visits than small-flowered individuals. We found no evidence of tradeoffs between pollen export efficiency and per visit accumulation of outcross pollen; each was influenced by unique aspects of flower morphology. Individual queen bumble bees of the principal pollinator species, Bombus kirbyellus, were similar in male, female, and absolute measures of pollination effectiveness. An estimated 2.9% of the pollen that bumble bees removed from flowers during a foraging bout was, on average, deposited on stigmas of compatible recipients. Significant plant-to-plant differences in pollen production, pollen export per visit, and outcross pollen receipt were found for co-occurring individuals of P. viscosum indicating that variation in these fitness related traits can be seen by pollinator-mediated selection.     

Effect of bumble bee (Hymenoptera: Apidae) pollination intensity on the quality of greenhouse tomatoes

Laboratory studies were conducted to assess tomato, Lycopersicon esculentum Mill. (Solanaceae), quality in relation to the level of buzz-pollination by bumble bees. Studies were conducted in commercial tomato greenhouses in the Leamington, Ontario, area to categorize bruising of tomato anther cones by bumble bees into five levels of bruising. The number of pollen grains per stigma was determined for each bruising level, and the bruising level was found to be a good predictor of stigmatic pollen load. Experimental flowers were pollinated by bumble bees and assigned to bruising levels based on the degree of anther cone discoloration. Fruit set, tomato weight, minimum diameter, the number of days until ripe, roundness, weight, percentage sugars, and number of seeds were assessed and compared among bruising level. Fruit set in flowers receiving no pollination visits was 30.2%, whereas, 83.3, 84.4, 81.2, and 100% of the flowers set fruit in bruising levels 1, 2, 3, and 4, respectively. Minimum diameter, number of seeds, and tomato weight all increased from no bruising to different levels of bruising. There was no increase in weight or diameter above a bruising level of 1, and no increase in the number of seeds per fruit after a bruising level of 2. We found that pollination of tomato flowers greater than a bruising level of 2 (corresponding to approximately one to two bee visits) did not result in a significant increase in quality.     

Sex differential nectar secretion in protandrous Alstroemeria aurea (Alstroemeriaceae): is production altered by pollen removal and receipt?

We examined diurnal and nocturnal nectar secretion across sexual stages in protandrous Alstroemeria aurea, a bumble bee-pollinated herb with long-lived flowers native to the southern Andes. We found the following patterns: (1) most nectar was produced diurnally and (2) three times more sugar was secreted during the male than female phase, not only because the male phase lasted longer but also because the rate of nectar production was higher. This 3:1 ratio in nectar production matched the ratio of the minimum number of bumble bee visits required on average to saturate male (pollen removal) vs. female (seed set) functions. Standing crop of nectar, on the other hand, did not differ greatly between male- and female- stage flowers left open to visitors, because the high-production male-phase flowers were visited more frequently than female- phase flowers. In an experiment concurrent with the repeated nectar sampling of individual flowers over their life-span, we removed pollen from anthers or deposited pollen on stigmas by hand. Neither treatment, designed to mimic effects of visits by Alstroemeria's native bumble bee pollinator, affected nectar production. The absence of plasticity in nectar secretion in relation to pollination events may reflect a low cost of nectar production, or may result from developmental constraints related to the evolution of the synchronous protandry that characterizes A. aurea.     

Behavioral responses by bumble bees to variation in pollen availability

Summary Pollen-collecting bumble bees (Bombus spp.) detect differences between individual flowers in pollen availability and alter their behavior to capitalize on rewarding flowers. Specific responses by bees to increased pollen availability included: longer visits to flowers; visits to more flowers within an inflorescence, including an increased frequency of revisits; an increased likelihood of grooming while the bee flow between flowers within the inflorescence; and more protracted inter-flower flights, probably because of longer grooming bouts. The particular suite of responses that a bee adopted depended on the pollen-dispensing mechanism of the plant species involved. Bees buzzed previously-unvisited Dode-catheon flowers longer than empty flowers. In contrast, pollen availability did not significantly affect the duration of visits to Lupinus flowers, which control the amount of pollen that can be removed during a single visit. Simulation results indicate that the observed movement patterns of bumble bees on Lupinus inflorescences would return the most pollen per unit of expended energy. The increased foraging efficiency resulting from facultative responses by bees to variation in pollen availability, especially changes in the frequency and intensity of grooming, could correspondingly decrease pollen dispersal between plants.     

Subdioecy in Consolea spinosissima (Cactaceae): breeding system and embryological studies

The breeding system and the embryology of Consolea spinosissima, a tree-like opuntioid endemic to Jamaica, were investigated. Morphological and embryological studies revealed that the species is subdioecious, with three sexual morphs present in the 150 × 120 m plot studied at Hellshire Hills, Jamaica. The female morph has pistillate flowers with open stigma lobes, no pollen grains, and sets fruit. The male morph has cryptic staminate flowers with closed stigma lobes, viable pollen grains, and a nonfunctional gynoecium that does not set seed. The weak hermaphrodite morph has low fruit set and "perfect" flowers that superficially resemble the functionally staminate flowers of the male morph. These perfect flowers reach anthesis with viable pollen grains, with no or only a few functional ovules, and with the style supporting pollen tube growth. Embryological studies showed that the critical stage for sex determination occurs earlier in pistillate than in staminate and perfect flowers. Anthers of pistillate flowers abort prior to microspore tetrad formation, whereas ovules of the staminate and perfect flowers degenerate after the complete maturation of the embryo sac. Based on flower structure and embryological data, we hypothesize that the ancestor of C. spinosissima is/was hermaphroditic.     

POLLINATOR EFFECTIVENESS AND SEED SET IN POPULATIONS OF AGALINIS STRICTIFOLIA (SCROPHULARIACEAE)

Examining variation in pollinator effectiveness and seed production resulting from single pollinator visits can provide a deeper understanding of how pollinators may influence reproduction in plant populations. When comparing populations, differences in the number of seeds produced from single pollinator visits to flowers may not always be attributable to differences in pollen deposition, but rather to differences in plant fecundity or resource availability. Pollinator effectiveness and seed production were studied for two populations over a 4-year period and were measured using single bee visit manipulations of flowers. No significant difference in pollinator effectiveness (pollen deposited on stigmas) was observed between the two populations. However, a significant difference between the two populations was observed in the number of seeds produced per flower. The Wellhouse population produced approximately three times as many seeds/flower from a single pollinator visit as did the Firefly Meadow population. Within each population, pollinators (Bombus pennsylvanicus and Apis mellifera) did not differ in the number of pollen grains deposited on stigmas or seeds produced per flower from single visit experiments. Differences in plant density, pollen viability, and ovules per flower also could not account for a significant amount of the variation. A resource augmentation experiment (water and fertilizer application) resulted in a decrease in seeds per flower per bee visit for the water treatment at the Wellhouse population only. For both populations, pollen deposition, pollen viability, and ovules per flower were unaffected by the resource augmentation. Alternative possibilities for the observed differences in seeds per flower per bee visit are discussed.     

Pollination biology and the impact of floral display, pollen donors, and distyly on seed production in Arcytophyllum lavarum (Rubiaceae)

In animal-pollinated plants, two factors affecting pollen flow and seed production are changes in floral display and the availability of compatible mates. Changes in floral display may affect the number of pollinator visits and the availability of compatible mates will affect the probability of legitimate pollination and seed production. Distyly is a floral polymorphism where long-styled (pin) and short-styled (thrum) floral morphs occur among different individuals. Distylous plants frequently exhibit self and intra-morph incompatibility. Therefore changes in morph abundance directly affect the arrival of compatible pollen to the stigmas. Floral morph by itself may also affect female reproductive success because floral morphs may display differences in seed production. We explored the effects of floral display, availability of neighboring compatible mates, and floral morph on seed production in the distylous herb ARCYTOPHYLLUM LAVARUM. We found that floral display does not affect the mean number of seeds produced per flower. There is also no effect of the proportion of neighboring legitimate pollen donors on seed production in pin or thrum flowers. However, floral morphs differed in their female reproductive success and the thrum morph produced more seeds. Hand pollination experiments suggest that differences in seed production between morphs are the result of pollen limitation. Future research will elucidate if the higher seed production in thrum flowers is a consequence of higher availability of pollen donors in the population, or higher efficiency of the pin morph as pollen donor.     

Pollen competition in a natural population of Cucurbita foetidissima (Cucurbitaceae)

The pollen competition hypothesis predicts that when the number of pollen grains deposited onto stigmas exceeds the number of ovules, selection can operate in the time frame between deposition and fertilization. Moreover, because of the overlap in gene expression between the two phases of the life cycle, selection on microgametophytes may alter the resulting sporophytic generation. The extent to which pollen competition occurs in nature has been unclear, because tests of the predictions of the pollen competition hypothesis have used cultivars and/or artificial growth conditions and hand-pollination techniques. In this study we used a wild species, Cucurbita foetidissima, in its natural habitat (southern New Mexico) to determine the amount and timing of the arrival of pollen onto stigmas, the relationship between pollen deposition and seed number, and the effects of the intensity of pollen competition on progeny vigor. We found that ～900 pollen grains are necessary for full seed set and that a single visit by a pollinator results in the deposition of 653.0 ± 101.8 pollen grains. About 29% of the flowers receiving a single pollinator visit had 900 or more pollen grains on its stigma. Moreover, within 2 h of anthesis, >4000 pollen grains were deposited onto a typical stigma, indicating that multiple pollinator visits must have occurred. Fruits produced by multiple visits had greater seed numbers (206 vs. 147) than fruits produced by a single visit. Finally, the progeny produced by multiple pollinator visits were more vigorous than those produced by single visits with respect to five measures of vegetative growth (MANCOVA, Wilks' lambda = 0.96, F(6,370) = 2.54, P < 0.02. These data demonstrate that conditions for pollen competition exist in nature and support the prediction that pollen competition enhances offspring vigor.     

Pollination Rates and Pollen Tube Growth in a Vulnerable Plant,Liriodendron chinense (Hemsl.) Sarg. (Magnoliaceae)

The limited seed production of insect-depended plant, Liriodendron chinense was once considered to be pollen-limited and insufficient cross pollination. In this study, we counted pollen grains deposited on stigmas in three populations in Guizhou, Hunan and Zhejiang provinces of China respectively. Over 61.9% stigmas were pollinated. From 1994 to 1996, the mean number of pollen grains on each stigma ranged from 4.4 to 42.6, much more than ovules(2) in each pistil. Based on observations of three years, both the pollination rate and pollen quantity on stigmas significantly affected seed set. When flowers opened without stamens dehiscencing at the early stage of anthesis, stigmas received considerable quantity of pollen grains. Pollen grains from different sources were able to germinate and pollen tube growth rates were not greatly variable. It is very likely that pollen grains arrived first would fertilize eggs. Since only several pollen tubes went through the stylar canal, the potential pollen competition may exist. In this case, there would be strong selection on floral syndrome which benefit early insect visits. Pollen grains from the early visits would have a greater chance to fertilize ovules than those from later visits, which implies that cross pollination is the predominant breeding system of this plant. The conclusion was also confirmed by following four artificial experiments. Three treatments, including flowers bagged, netted or with the perianth removed, all reduced seed set clearly, but flowers with the stamens removed (emasculation)had a higher seed production by open pollination. As the rates of deposited stigmas in three populations were 6～8 times more than full seed set, we considerthat lower seed production in this plant may not mainly be due to pollen limitation.     

Ecological and selective effects of stigma-anther separation in the self-incompatible tropical tree Ipomoea wolcottiana (Convolvulaceae)

We evaluated the effect of stigma-anther separation (SAS) in flowers of the self-incompatible tropical tree Ipomoea wolcottiana (Convolvulaceae). We measured the deposition of self pollen on the stigma and the growth of pollen tubes. We also evaluated phenotypic selection on female fertility using univariate regressions of SAS with fruit production. SAS varied among individuals in two years, with a range of 0–6 mm. There was a negative correlation between SAS and the number of self pollen grains on the stigma; trees with SAS of 0–2 mm received averages of ca 80–130 self grains. The effect of loading the stigma with self-pollen was tested by allowing isolated flowers to self-pollinate, then cross-pollinating by hand: there was a negative correlation between the pollen load and the proportion of grains forming pollen tubes. The highest proportion of grains forming tubes was found with loads of <10 grains, while the minimum frequency of tubes occurred with loads of >50 grains. The results suggest that more self pollen deposition leads to fewer tubes from outcross pollen. There was a significant effect of SAS on fruit production in both years. The analyses suggested stabilizing selection; very small SAS is disadvantageous due to the high load of self pollen, while very large SAS probably led to damage to the style during pollinator visits. We suggest that herkogamy may be a key character for determining the success of fruit production in self-incompatible hermaphroditic species.