A multi-year study of factors affecting fruit production in Aristolochia paucinervis (Aristolochiaceae)

Pollen limitation, resource limitation, fruit abortion, and predation have all been proposed as factors explaining low fruit set in hermaphroditic plants. We conducted a 5-year study combining field observations and pollination experiments to determine the causes of the low fruit set in Aristolochia paucinervis, a Mediterranean species with a specialized pollination system in two populations in SW Spain. Fruit initiation was markedly low, and between 28.6 and 75.0% of the flowering stems did not initiate any fruit. In most flowers, the number of germinated pollen grains was less than the number of ovules, and supplemental pollination significantly increased fruiting, indicating deficient pollination. In A. paucinervis, autonomous self-pollination seems to be a decisive factor in fruit production because the number of germinated pollen and the fruit set from flowers bagged before anthesis were similar to those in free-pollinated flowers. Only in 2005 did flowers that were successfully pollinated outnumber ripened fruits, suggesting that other factors limit fruiting. We found a significant positive correlation between tuber mass and fruit set. Deficient pollination and lack of resources could explain the low fruit set, but the relative consequences seem to vary spatially and temporarily.     

Testing hypotheses for excess flower production and low fruit-to-flower ratios in a pollinating seed-consuming mutualism

Pollinator attraction, pollen limitation, resource limitation, pollen donation and selective fruit abortion have all been proposed as processes explaining why hermaphroditic plants commonly produce many more flowers than mature fruit. We conducted a series of experiments in Arizona to investigate low fruit-to-flower ratios in senita cacti, which rely exclusively on pollinating seed-consumers. Selective abortion of fruit based on seed predators is of particular interest in this case because plants relying on pollinating seed-consumers are predicted to have such a mechanism to minimize seed loss. Pollinator attraction and pollen dispersal increased with flower number, but fruit set did not, refuting the hypothesis that excess flowers increase fruit set by attracting more pollinators. Fruit set of natural- and hand-pollinated flowers were not different, supporting the resource, rather than pollen, limitation hypothesis. Senita did abort fruit, but not selectively based on pollen quantity, pollen donors, or seed predators. Collectively, these results are consistent with sex allocation theory in that resource allocation to excess flower production can increase pollen dispersal and the male fitness function of flowers, but consequently results in reduced resources available for fruit set. Inconsistent with sex allocation theory, however, fruit production and the female fitness function of flowers may actually increase with flower production. This is because excess flower production lowers pollinator-to-flower ratios and results in fruit abortion, both of which limit the abundance and hence oviposition rates, of pre-dispersal seed predators.     

PATTERNS OF FRUIT PRODUCTION IN A NEOTROPICAL ORCHID: POLLINATOR VS. RESOURCE LIMITATION

Patterns of growth, and of flower and fruit production, were monitored over two years in a population of the orchid Aspasia principissa Reichb. f. in central Panama. Observations and experimental manipulations were used to determine the relative importance of pollinator and resource limitation on fruit production. Within a season, fruit production was limited by pollinator availability. Fruit set for hand-pollinated flowers was over six times greater than that for naturally pollinated flowers. However, in plants that produced more than one fruit, fruit size declined in subsequent fruits, indicating that resources could limit seed production within a season. Plants producing fruits in 1986, on average, produced smaller shoots and inflorescences in 1987 relative to plants that flowered but produced no fruit in 1986. Thus, plants are likely to be resource limited over their lifetimes. Most individuals of reproductive size (82.5%) did not produce fruit over a three-year period. The reproductive dominance of a few individuals in this population of Aspasia principissa may have important implications for understanding the population structure of the species and the high species diversity of orchids.     

Inadequate pollination services limit watermelon yields in northern Tanzania

Insect pollination plays a vital role for the yield of many crops, such as apples, strawberries and coffee, which are economically significant commodities on the global market. Yet, knowledge about the role of insect pollination is lacking for many cash crops that support the livelihoods of small-scale farmers in developing countries.To assess if yields of watermelon are pollen limited, we conducted a supplemental hand-pollinated experiment (using a pollen mixture of cross and self-pollen) in 13 small-scale farms in an agricultural landscape in the Kilimanjaro and Arusha regions in northern Tanzania. We assessed fruit set, fruit abortion and weight of the mature fruits stemming from hand pollinated and control flowers. To check if differences in yield responses between hand-pollinated and control treatments depended on resource availability, we also accounted for local soil conditions.We found that hand pollination (i) increased the probability of fruit initiation by 30%, (ii) reduced the probability of fruit abortion by 13%, (iii) increased the probability of flowers developing into mature fruits by 42% and (iv) increased average fruit weight by 1.3 kg (±0.15 SE). Our results indicate that our system is pollen limited, due to insufficient visitor frequency and/or inefficient pollinator species.Fruit initiation and fruit weight were positively related to soil carbon, irrespective of treatment. The influence of soil moisture was not consistent across the measured responses, and differed between hand-pollinated and the control treatments.We suggest that farmers in our focal area should focus on improving the quality of the landscape to sustain and enhance healthy pollinator communities ultimately improving yields. We also suggest that farmers should continue current practices with respect to fertilization. The role of soil moisture on fruit initiation and maturation should be investigated to ensure that the positive effects of pollen is not hindered by soil moisture conditions.     

Evidence for selective fruit abortion in Banksia spinulosa (Proteaceae)

Fruit set is consistently low in the genus Banksia (Proteaceae). One explanation for low fruit:flower ratios is that excess flowers allow plants to selectively abort inferior progeny thereby increasing the average quality of the seed crop. We examined whether self-pollinated fruits were aborted selectively in a partially self-compatible population of Banksia spinulosa. Inflorescences were divided in half and pollinated with cross pollen on both sides, self pollen on both sides or cross pollen on one side and self pollen on the other. Seed production was reduced significantly by 38% on fully self-pollinated compared to fully cross-pollinated inflorescences, indicating partial self-incompatibility or inbreeding depression. On inflorescences pollinated with both cross and self pollen, selfed seed set was reduced by 63% compared to crossed seed set. On the selfed half of these inflorescences, fewer fruits and seeds were produced, and more fruits had aborted seeds than on randomly selected halves of fully self-pollinated inflorescences. These results suggest that self-fertilized embryos were developed in the absence of outcrossed progeny but were selectively aborted when outcrossed progeny were present on inflorescences. To determine the amount of outcrossing occurring in the study population, outcrossing rates (t) of naturally pollinated plants were examined by electrophoretic assessment of progeny arrays. Values of I did not differ significantly from one, indicating complete outcrossing. High outcrossing, despite the high probability of self pollen receipt under natural conditions, is consistent with the conclusion that selective abortion occurs in B. spinulosa.     

Effects of pollen and nitrogen availability on reproduction in a woodland herb,Lysimachia quadrifolia

Summary Relationships between pollen loads, resource availability, and fruit and seed production were determined for Lysimachia quadrifolia ramets in two adjacent sites (the scrub site and the open site) in 1982, 1983, and 1984. Pollen loads limited % fruit set and seed production in the open site in 1982 and 1983. Reproduction in the scrub site was resource limited in 1982, as shown by an increase in % fruit set when one-half of the flowers on a ramet were removed prior to fruit initiation. In the scrub site in 1983, pollination of one-half of the flowers on a ramet decreased the % fruit set of the remaining, unpollinated flowers. Fruit production in the same site was limited by pollen in 1984. Addition of nitrogen fertilizer to the scrub site in 1984 had no effect on fruit and seed production. There was more variation in fruit set between sites than between years. There was no trend to greater fruit set or number of seeds/fruit on early flowers compared to late flowers on the same ramet. Correlations between measures of reproduction were positive or insignificant. These results demonstrate year-and site-specific variation in the factors that limit plant reproduction.     

Self‐pollination and parthenocarpic ability in developing ovaries of self‐incompatible Clementine mandarins (Citrus clementina)

This study aimed to determine if self‐pollination is needed to trigger facultative parthenocarpy in self‐incompatible Clementine mandarins (Citrus clementina Hort. ex Tan.). ‘Marisol’ and ‘Clemenules’ mandarins were selected, and self‐pollinated and un‐pollinated flowers from both cultivars were used for comparison. These mandarins are always seedless after self‐pollination and show high and low ability to develop substantial parthenocarpic fruits, respectively. The time‐course for pollen grain germination, tube growth and ovule abortion was analyzed as well as that for carbohydrates, active gibberellins (GA₁ and GA₄), auxin (IAA) and abscisic acid (ABA) content in the ovary. ‘Clemenules’ showed higher pollen grain germination, but pollen tube development was arrested in the upper style 9 days after pollination in both cultivars. Self‐pollination did not stimulate parthenocarpy, whereas both un‐pollinated and self‐pollinated ovaries set fruit regardless of the cultivar. On the other hand, ‘Marisol’ un‐pollinated flowers showed greater parthenocarpic ovary growth than ‘Clemenules’ un‐pollinated flowers, i.e. higher ovule abortion rate (+21%), higher fruit set (+44%) and higher fruit weight (+50%). Further, the greater parthenocarpic ability of ‘Marisol’ paralleled higher levels of GA₁ in the ovary (+34% at anthesis). ‘Marisol’ ovary also showed higher hexoses and starch mobilization, but lower ABA levels (?64% at anthesis). Self‐pollination did not modify carbohydrates or GA content in the ovary compared to un‐pollination. Results indicate that parthenocarpy in the Clementine mandarin is pollination‐independent with its ability to set depending on the ovary hormone levels. These findings suggest that parthenocarpy in fertile self‐incompatible mandarins is constitutively regulated.     

Reproduction in an Orchid Can Be Resource‐Limited over its Lifetime1

Most orchids studied thus far show long‐term resource adjustments to increases in fruit production within a flowering season, but none of these offers rewards to their potential pollinators. If nectar production is energetically expensive, then resources utilized to produce fruits and seeds may be even more limited in pollinator‐rewarding orchids than in non‐rewarding ones. Thus, resource adjustments may be more dramatic or entirely different in nectar producing plants. In this study, we performed artificial hand‐pollinations for two consecutive flowering seasons in die nectar producing orchid Comparettia falcata, and tested whether or not fruit set, seed set, and seed viability were limited by the quantity of pollinations or by resources. In addition, we compared mechanisms of short‐term (fruit abortion within seasons) and long‐term consequences (percent change in leaf length and change in flower number per plant between seasons, probability of shoot and inflorescence production, and mortality) between hand‐ pollinated and unmanipulated plants. The relationships among plant traits related to vegetative size and reproduction also were examined. Hand‐pollinations showed some negative effects. Fruit set was higher in hand‐pollinated plants in the first season but was similar to the controls in the second. Seed set was significantly lower and abortions were higher than in unmanipulated plants. On the other hand, some of our measurements were unaffected by die hand‐pollination treatment. Unexpectedly, there were no significant differences between groups in percent change in leaf length, change in flower number per plant between seasons, or die probability of shoot and inflorescence production. Although there was a strong correlation between leaf size and die number of flowers produced within a season, associations between leaf size and traits related to current or future reproduction were not consistent. Like other epiphytic orchids, pollination limitation occurred within a single season in C. falcata., but increases in fruit production also resulted in reduced lifetime fitness as estimated by a compounded fitness index. Contrary to all other epiphytic orchids studied, long‐term adjustments to increased fruit production in C. falcata through reduction in future growth or flower and inflorescence production were either minor or lacking. Our results suggest that the nature of plant strategies associated with resource constraints during sexual reproduction may be dependent on whether or not plants have evolved traits that are costly.     

Factors constraining fruit set in Mandevilla pentlandiana (Apocynaceae)

The reproductive success of Mandevilla pentlandiana was studied to disclose its reproductive strategy, and to determine the links between nectar production, breeding system, fruit set and inflorescence size. The plant produces many inflorescences with a large number of flowers but initiates few fruits (9%). This vine is self-compatible but not autogamous. Given that no significant differences could be detected considering many traits (ripe and abortive fruit sets, fruit quality, and seedling survival) between the pollination treatments (self-, cross-and natural-), the low natural fruit set was not related to pollen limitation. Fruits were not distributed at random within inflorescences (earlier fruits had the highest probability of maturation) but there were no significant differences in fruit quality according to different fruit positions. Conversely, the time of fruit initiation influenced most of the fruit-traits. Many developing fruits were aborted (20%). An increase in the probability of abortion was detected when the whole inflorescence was hand pollinated. In addition, a positive correlation was detected between the abortions and the number of ripe fruits which developed before them. Looking at our data from an evolutionary perspective, we argue that a theoretical inflorescence size, corresponding to the intersection point between the mean values of fruit number and fruit set per inflorescence, can be assumed to indicate the optimum inflorescence size that maximizes equally both female and male functions. Comparison between the theoretical and the observed mean inflorescence size suggests, that for M. pentlandiana , pollen donation may be the primary evolutionary factor behind excess flowers.     

Optimal resource allocation model for excessive flower production in a pollinating seed-predator mutualism

Many plants produce excessive flowers and several hypotheses have been proposed for adaptive significances of this behavior. Here, I develop a simple resource allocation model for plants in a mutualism with pollinating seed-predators to examine a novel hypothesis that excessive flower production can be favored to “dilute” seed predation by the pollinators. Pollinators visit flowers to deposit pollen and oviposit on them, and their offspring feed on a portion of the seeds, leaving the remainder intact. Further pollinator visits increase seed mortality by over-oviposition. Excessive flower production is favored if it decreases pollinator-visit frequency per flower, while it incurs decrease in seed production because of the resource trade-off. I examine three plant strategies: (1) no abortion, the plant allocates resource to all pollinated flowers to mature; (2) selective abortion, the plant aborts flowers depending on how many times they were visited by pollinators; and (3) random abortion, the plant indiscriminately aborts a fraction of pollinated flowers irrespective of how many times they were visited. I show that the random abortion strategy can perform much more effectively than the no-abortion strategy when the amount of resource is small, the production cost per flower is low, and the pollinator density is high, although the selective abortion strategy is always the best. This “predator dilution” effect has not been considered with regard to previous excessive flower production hypotheses.     

Effects of Cross-pollination and Flower Removal on Fruit Set in Macadamia

Macadamia racemes were cross-pollinated and had flowers removedto determine whether manipulation of initial fruit set wouldaffect final fruit set or yield. Examination of flowers collected7 d after pollination indicated that the cross-pollination treatmentgenerally induced a small (13-19%) but significant increasein the number of pollen grains per stigma. All flowers possessedsome pollen grains, but in all cases cross-pollination approximatelydoubled the percentage of flowers with a pollen tube at thebase of the style (42-68% of flowers on cross-pollinated racemescompared with 21-35% of flowers on control racemes). Fruit setafter 14 d was always improved by cross-pollination, and thistranslated into increased final fruit numbers on two out offour occasions. In these two cases, kernel weights were increasedby 24 and 31% in the cross-pollinated treatments. This indicatesa possible xenic (paternal parent) effect in macadamia. Flowerremoval resulted in retention of a greater percentage of flowersas fruits, so that only severe reductions in flower number decreasedfinal fruit numbers. Improved efficiency of cross-pollinationin macadamia may increase both nut numbers and nut size.Copyright1994, 1999 Academic Press Macadamia integrifolia Maiden and Betche, Macadamia tetraphylla L. A. S. Johnson, Proteaceae, macadamia, pollination, pollen tube growth, flower number, fruit drop, fruit set, xenia     

Limited carbohydrate availability as a potential cause of fruit abortion in Rubus chamaemorus

Fruit abortion can be caused by a range of abiotic and biotic factors. To gain a better understanding of the causes of the high fruit abortion frequency in cloudberry ( Rubus chamaemorus L.), we manipulated different sources of carbon, that is, leaves and rhizome. We also manipulated flower number to see if competition between floral ramets explained fruit abortion in cloudberry. Reducing the number of flowers had no impact on fruit abortion frequency. In fact, the species forms an extensive rhizome network with only a few ramets per clone and competition between floral ramets is unlikely. Ramet defoliation had limited impact on fruit abortion, but successful fruit development was affected by rhizome length. The longer the rhizome, the higher the chances to mature a fruit. These results suggest that current photoassimilate production by the reproductive ramet alone is insufficient to insure fruit development. Carbon can come from other ramets but distances are usually high between ramets. Fruit production might thus depend on the use of stored carbohydrates in the rhizome to balance insufficient photosynthetic contribution during fruit production.     

Pollinators and Reproductive Success of the Wild Cucurbit Cucurbita maxima ssp. andreana (Cucurbitaceae)

Abstract: We studied the reproductive success and pollinators of Cucurbita maxima ssp. andreana in different disturbed habitats where it grows naturally. Data were obtained from three populations. One grew within a soybean crop, the other within a corn crop, and the third in an abandoned crop field. Cucurbita maxima ssp. andreana is an annual vine with a flowering period from December to April. Male flowers appear first, thereafter female and male flowers appear together. Flower lifetime (9 h) was similar in male and female flowers. The pollinator guild was comparable for the three populations but some differences in the frequency of the insect species were observed. Native bees were the main pollinators in the population in the abandoned field, while beetles pollinated the populations in crop fields. These differences were not linked with the pre-emergent reproductive success, fruit and seed set, or fruit quality. This is a self-compatible plant. Fruit and seed set and fruit traits (total mass, width and length of fruits, number of seeds per fruit, and seed mass) did not show significant differences between hand-cross and hand-self pollinated flowers. This wild cucurbit is a generalist with respect to pollinator guild, and flower visitors seem to be highly efficient in pollen transference. Cucurbita maxima ssp. andreana is well adapted to disturbed habitats because plants ripened fruits successfully, regardless of the group of insects visiting flowers.     

荒漠植物准噶尔无叶豆结实、结籽格局及其生态适应意义

连续2a采用野外记录的方式对荒漠植物准噶尔无叶豆的结实、结籽格局进行了比较研究。结果表明:植株内花序生成格局表现为不同大小的花序在总花序中的比重与其结实率存在显著正相关,其中含有中等花数目的花序所占的比重及其结实率均最高,此种分布格局能够最大程度的保证繁殖成功;果序内果实生成格局表现为单花着生位置(从近柄端算起)与其结实量占总结实量的比例成线性负相关,花序基部的结实比例最高,顶部结实比例近为0;荚果内种子生成格局表现为:中间位置的胚珠败育率最低。另外,年际间荚果内仅有(1.08±0.03)粒和(1.07±0.03)粒种子能够完全成熟,此种结籽格局是准噶尔无叶豆保证后代质量的最佳策略。     

EFFECT OF DEFOLIATION ON GENDER EXPRESSION AND FRUIT SET IN PASSIFLORA INCARNATA

We examined the effect of defoliation on gender expression and fruit set in a north-central Florida population of the andromonoecious vine, Passiflora incarnata, during the 1984 flowering season. At three times during the flowering season (May, June, July), the leaves adjacent to flowers at four stages of development were removed, and subsequent sex and fate of each flower were determined. Defoliation affected gender of the flower by significantly decreasing the probability that the styles of the flower would deflex and the flower would thus function only as a pollen donor. Flowers were sensitive to defoliation at any stage before anthesis, though the sensitivity appeared to decrease in the most mature category of flower buds. Fruit set of hermaphroditic flowers remained unaffected by defoliation, but the probability of fruit set and gender expression were significantly influenced by the time of the flowering season. We conclude that the local photosynthate environment determines flower gender in Passiflora, but branch or entire plant photosynthate resources can compensate for local resource fluctuations and play an important role in fruit set and flower bud abortion. The sexual lability of Passiflora incarnata appears to be an adaptation to uncertain resource levels at a fine scale, caused by nearby developing fruits and the possibility of herbivore defoliation.     

Genetic engineering of parthenocarpic fruit development in tomato

Parthenocarpy was engineered in two genotypes of Lycopersicon esculentum Mill. by using the DefH9-iaaM chimeric gene. The parthenocarpic trait consists of fruit set and growth in the absence of fertilization. Seedless parthenocarpic fruits were obtained from emasculated flowers, and fruits with seeds from pollinated flowers. All parthenocarpic tomato plants analysed expressed the DefH9-iaaM gene during flower development. The fruit set percentage of emasculated transgenic flowers was similar to that of control plants. In 7 out of 8 independent transgenic plants, the fresh weight of fruits derived from pollinated or emasculated flowers did not significantly differ from that of fruits obtained by pollination of the control plants. The pH of the parthenocarpic fruit was generally unaffected and the soluble solid concentration was either unchanged or increased. Thus, the DefH9-iaaM gene is a genetic tool that might be used to improve tomato productivity.     

Reproductive biology and effect of nectar robbing on fruit production in Macleania bullata (Ericaceae)

Hummingbird-pollinated flowers are frequently subjected to nectar robbing. In this paper, I examine the impact of nectar robbing on plant reproductive success on a hummingbird-pollinated species. After studying the basic aspects of the floral morphology and reproduction of Macleania bullata (Ericaceae) in a tropical montane wet forest in southwest Colombia, I examined the percent of flowers robbed and the effect of nectar robbery on fruit set. The flowers of this species are typical for plants pollinated by long-bill hummingbirds. They are protandrous and open for four days. Fruit production requires a pollinator visit; fruit set following pollinator exclusion was zero. Fruit set following xenogamous pollen transfer (36.8%) differed significantly from that of population controls (11.9%) and of autogamous pollen transfer (6.3%). Nectar volume, sugar concentration and sugar production were measured at daily intervals from bud opening until the fading of flowers. Daily nectar production (both volume and amount of sugar) varied considerably with flower age. Sugar production peaked on the second day, coinciding with the male phase. The frequency of nectar robbing in the studied population was very high (75% of examined flowers) and was positively correlated with reduced fruit set. I discuss the probability of a relation between reduced fruit set on robbed flowers and an energetic investment. Robbing by non-pollinating visitors can suppose the plant to re-synthesize more nectar. The high incidence of nectar robbing impugns the advantage of specialization.     

A Top-Down Hierarchy in Fruit Set on Inflorescences in Iris fulva (Iridaceae)

Abstract: In Iris fulva, the apical flower on an inflorescence opens first, and the flowering sequence then proceeds from the most basal flower upward to the apex (acropetally). Flowering order and flower position are thus partially decoupled, and this provides an opportunity to separately investigate the effects of these two factors upon fruit formation. We recorded natural patterns of fruit set and seed production, and found that fruit set patterns were determined by what appears to be a form of apical dominance. The first, apical flower had the highest fruit set and seed production. Fruit set decreased towards the base of the inflorescence, with later-opening flowers having a higher fruit set. This is contrary to the pattern usually observed in other acropetally flowering plants, which is a higher fruit set for basal, early flowers. By performing additional hand pollinations and counting pollen grains on naturally pollinated flowers, we found that pollen deposition was not a major factor limiting fruit set, and that it could not explain the large difference in fruit set between the apical and basal flowers. Removing the first flower after it had wilted increased fruit set in the remaining flowers, but mainly in the more apical flowers. Only by removing the two topmost flowers could we obtain an increase in fruit set for the basal flower. The fact that the basal flower rarely sets fruit, despite being closer to resources, suggests that the apical meristem is either a strong sink for resources or produces hormones that form a gradient along the inflorescence, which is comparable with apical dominance.     

Effects of floral display and plant abundance on fruit production of Ryncholaelia glauca (Orchidaceae)

Flowering plant density can increase number of visits and fruit set in multi-flowering plants, however this aspect has not been studied on few flower species. We studied the effects of individual floral display and plant density on the fruit production of the epiphytic, moth-pollinated orchid, Ryncholaelia glauca, in an oak forest of Chavarrillo, Veracruz, Mexico. Species is non-autogamous, and produced one flower per flowering shoot each flowering season. We hypothesized that orchids with more flowering shoots and those on trees with clumps of conspecific should develop more fruits than isolated ones. R. glauca population flowers synchronouly, and individual flowers last up to 18 days, with flowers closing rapidly after pollination. Individuals produced few flowers per year, although some plants developed flowers in both seasons and fewer of them developed fruits both years. There was no relationship between flower number per orchid, or per host tree, with the number of fruits developed per plant. Host trees with flowering and fruiting orchids were randomly dispersed and the pattern of distribution of flowering and fruiting plants was not related. Apparently, pollinators visit the flowers randomly, with no evidence of density dependence. The fruit set of R. glauca was as low as fruit set of multi-flowered orchids moth pollinated, suggesting that fruit set on moth pollinated orchids could be independent of the number of flowers displayed.     

Are reward polymorphisms subject to frequency‐ and density‐dependent selection? Evidence from a monoecious species pollinated by deceit

Abstract Most deceit‐pollinated species involve floral dimorphisms characterized by the presence of rewarding male flowers and nonrewarding female flowers. It has been proposed that this polymorphism establishes the conditions for the action of frequency dependent selection (FDS). The tendency of foraging animals to aggregate in areas of high resource density suggests that pollination efficiency and fruit production may be positively influenced by flower density (density dependent selection, DDS). In this paper we offer a graphical model describing the effects of FDS and DDS on a monoecious species pollinated by deceit. We test the FDS and DDS assumptions and the predictions of the model using field observations and experimental populations of Begonia gracilis in which population sex ratio and flower density were controlled. We found a marked effect of both FDS and DDS on pollinator visitation, fruit‐set, and on the probability of female flowers to setting fruits. We conclude that these two types of selection have had a strong influence on the evolution of deceit‐pollinated species.