| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
2Makino Herbarium, Graduate School of Science, Tokyo Metropolitan University, 1-1 Minami-Osawa, Hachioji, Tokyo 192-0397, Japan; 3Faculty of Integrated Arts and Sciences, The University of Tokushima, 1-1 Minami-josanjima, Tokushima 770-8502, Japan; 4Department of Biological Sciences, Botanical Gardens, Koishikawa, Graduate School of Science, University of Tokyo, 3-7-1 Hakusan, Bunkyo-ku, Tokyo 112-0001, Japan
Received for publication November 15, 2003. Accepted for publication August 26, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Bombus • Clematis stans • floral morphology • fruit set • pollen removal • Ranunculaceae • specialized pollination system • visitation frequency
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Schemske and Horvitz, 1984
; Herrera, 1989
; Waser et al., 1996
). When a plant adopts this strategy, its floral traits evolve to correspond to the morphology, physiology, and behavior of the specific pollinator that produces the highest reproductive success. However, nonspecific pollinators cannot use these plants because they lack the specialized morphology or physiology necessary to successfully pollinate the plants. If the presence of the specific pollinator is unstable, the specialized plants will suffer in terms of pollination and subsequent reproduction (Johnson and Steiner, 2000
). Thus, plants with specialized relationships achieve high pollination efficiency, but suffer a high risk of extinction.
Spur or floral tube lengths in various groups of flowering plants are well correlated to the lengths of the mouthparts (tongue) of certain insect pollinators (Grant and Grant, 1965
; Inouye, 1980
; Nilsson, 1988
; Suzuki, 1992
). In general, flowers conceal nectar deep within the spur or the tubular part of the corolla as a reward for pollinators. When tongue length exceeds flower depth, pollen transfer may be reduced, as physical contact between the body of the pollinator and the anthers or stigmas of the flower could be incomplete (Inoue, 1986
; Nilsson, 1988
). This situation can lead to coevolution between increasing flower depth and further elongation of the insect tongue (Nilsson, 1988
). In such a case, flowers are visited by a single pollinator displaying morphological correspondence between the length of its tongue and the depth of the flower, constituting a specialized pollination system. In some plant species, flower odor and color have also become adapted to particular pollinators in specialized relationships (Scogin, 1983
; Nilsson, 1992
; Raguso, 2001
). In general, plant-pollinator relationships have evolved toward specialized pollination systems (Stebbins, 1970
; Gilbert and Raven, 1975
).
Temporal changes in floral morphology are well-known strategies for enhancing the reproductive success of plants (see Silvertown and Gordon [1989]
for a review). Ipomopsis aggregata shifts its floral colors during the flowering season, corresponding to changes in the abundance of different pollinators (Paige and Whitham, 1985
). Color changes in flowers also increase pollinator visitation to new flowers (Gori, 1983
; Casper and La Pine, 1984
; Gori, 1989
) and reduce the number of flower visits per plant, thereby avoiding selfing (Jones and Cruzan, 1999
; Ida and Kudo, 2003
). The significance of floral odor changes is also known, i.e., plants pollinated by nocturnal visitors scent more strongly or exclusively at night, influencing pollinator visits (Silvertown and Gordon, 1989
). Floral structure changes in flower development are associated with protogyny and protandry, conditions that deter excessive self-pollination (Proctor et al., 1996
).
Through a preliminary investigation, we found that Clematis stans Sieb. et Zucc. (Ranunculaceae) appears to change its flower morphology temporally, which consequently decreases distance to nectar from floral opening (Dohzono and Suzuki, 2002
). Clematis stans can use two bumble bee pollinators, Bombus diversus Smith and B. honshuensis Tkalcu, visitation preferences of which are based on morphological correspondences between calyx tube length and proboscis length, suggesting specialized relationships. This led us to hypothesize that temporal changes in the calyx tube length of C. stans enables effective pollen transfer in a specialized relationship with bumble bee pollinators. Moreover, C. stans increases its chances of pollen transfer by exploiting multiple bumble bee pollinators, which reduces the risks of specialization and thus enhances its reproductive success.
We tested this hypothesis by comparing the components of reproductive success of C. stans among three flowering stages with different calyx tube lengths. We examined visitation frequency over two years and estimated pollen removal, pollen deposition, and fruit set in single-visit experiments.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
During flowering of male and female flowers, the four sepals gradually curl up on the upper half and are separated from each other, such that the calyx tube is shortened day by day, and the styles or stamens are gradually disclosed (Fig. 1). Stamen and pistil lengths do not change during the flowering period (Fig. 1). The calyx tube of the male flowers tends to be longer than that of the female flowers throughout the flowering period (Fig. 1). In a previous study, flowers were tentatively classified into three stages (L, M, or S), based on the length of the calyx tube. L-flowers represent the early stage of flowering, when the calyx tube is longer than the stamens or pistils; M-flowers occur in the intermediate stage, when the calyx tube is approximately as long as the stamens and pistils; and S-flowers occur in the late stage, when the stamen and pistil are exserted clearly from the calyx (Dohzono and Suzuki, 2002
). This stage classification was also used in this study.
|
). Bombus diversus (long proboscis) tends to visit flowers with longer calyx tubes more frequently, whereas B. honshuensis (short proboscis) prefers flowers with shorter calyx tubes (Dohzono and Suzuki, 2002
).
Study sites
Field observations were conducted in the summers of 1998 and 2000 at two sites near Kazuma (1000 m altitude) in Nishitama-gun, Tokyo Prefecture, and near Gotenba-guchi (1300 m altitude) in the foothills of Mt. Fuji, Shizuoka Prefecture, Japan. The distance between the two sites was approximately 45 km. The Kazuma site was located on the edge of a forest consisting of deciduous broad-leaved trees (Quercus serrata Thunb. ex Murray, Alnus firma Sieb. et Zucc.) and planted Cryptomeria japonica (L. fil.) D. Don and Chamaecyparis obtusa (Sieb. et Zucc.) Endl. Clematis stans and Impatiens textori Miq. co-occur at this site, and both are pollinated by B. diversus. In the Gotenba site, near the edge of a deciduous broad-leaved forest (Betula ermanii Cham., Alnus maximowiczii Call.), the dominant plant species are C. stans and Circium purpuratum (Maxim.) Matsum. The main pollinator of C. purpuratum is B. diversus. At these sites, the numbers of male and female plants were almost equivalent.
Visitation frequency
At the Kazuma site, during peak flowering in 1998 and 2000, we surveyed the visitation frequency of B. diversus and B. honshuensis during the three flower stages (18 hours on September 5, 6, and 8, 1998, and 21 hours on September 4, 9, 10, and 14, 2000). We established a quadrat that included 10 plants with approximately 400 flowers. Species of bumble bees foraging within the quadrat were identified and, for each flower visitor entering the quadrat, we recorded the number, stage, and sex of the flowers visited. The total number of open flowers and their stages in the quadrat were also recorded daily.
Pollen removal from male flowers
To estimate the pollen removal at each stage after a single visit by B. diversus and B. honshuensis, we counted the number of pollen grains remaining in the flowers. The following experiments were carried out in 1998 at Kazuma. Inflorescences of male plants at the bud stage were bagged with nylon nets to exclude pollinators. To prepare flowers of various stages, we removed the nets after some flowers had opened to present the flowers to bumble bees. After a single visit by either bumble bee species, we picked the flower, for which we recorded the species of bumble bee and flowering stages (L, M, or S), and measured the floral traits (calyx length, calyx tube length, and stamen length). The collected flowers were packed in paraffin paper. In the laboratory, samples were oven-dried at 60° for 48 h and preserved in a plastic container with silica gel. To count pollen grains, samples were treated by acetolysis (Kearns and Inouye, 1993
), and pollen grains (PN) were counted with a Coulter Z1 particle counter equipped with a 100-µm aperture tube.
We used stamen number to estimate the number of pollen grains present prior to a visit. The following experiments were carried out in 1998 at Gotenba. The mean (±1 SE) number of stamens in male flowers was 17.13 ± 0.19 (N = 151). Linear regression of original pollen number (PO) on stamen number (S) showed a significant positive correlation (PO = 19.8 x S – 208.4, r2 = 0.26, P < 0.0005, N = 45). We also estimated the number of pollen lost naturally during the flowering stages. We used the bagged flowers, from which pollinators had been excluded, and measured flower morphology (calyx tube length and calyx length) after collecting flowers of various stages. We calculated the number of pollen lost (L) as the original pollen number – remaining pollen number. Linear regression of L on calyx tube length/calyx length (C) showed a nonsignificant correlation (L = 51.1 x C + 69.22, r2 = 0.042, P = 0.24, N = 35), indicating that pollen lost during flowering was negligible. Thus, the number of pollen grains removed from a flower was calculated as PO – PN.
Pollen deposition and fruit set
We counted pollen grain deposition on stigmas and examined fruit set, following a single visit by B. diversus and B. honshuensis, and compared them among the three flower stages. The following experiments were carried out in 1998 at Kazuma.
Female flowers were bagged with nylon nets at the bud stage to prepare nonpollinated flowers, as described for male flowers. In the morning, we removed the nets to present the flowers to bumble bees. After each visit, we tagged the flower, noting its flowering stage (L, M, or S), measured its floral traits (calyx length, calyx tube length, and stamen length), and then bagged it again. After 5 d, we cut off the style, including the stigma, mounted it on a glass slide, stained it with fuchsin glycerin-jelly, and counted pollen grains on the stigma under a microscope (Kearns and Inouye, 1993
). We then counted the total number of styles and the number of pollinated stigmas, calculating the proportion of pollinated stigmas. This proportion may reflect the fruit set, because the flowers of C. stans have compound pistils and the number of styles is equal to the number of ovules (17 ovules per flower on average). Ovaries were left intact until maturation. After approximately 1 mo, the fruits (achenes) matured. The fruit set was defined as the proportion of mature ovaries to the total number of styles in each flower.
To estimate the maximum fruit set per flower, possible pollen limitation, and pollen receptivity of the stigma at each flowering stage, we conducted hand-pollination experiments in 1998 at Gotenba. Flower buds were arbitrarily selected and bagged to exclude pollinators. After flowering, the pollen grains from male flowers were applied to the stigmas of female flowers at each flowering stage. We then treated these flowers, as described in the single-visit experiment.
Data analyses
To determine whether pollinators preferred a particular flower stage, we used a chi-square test and a Fisher's exact test to compare the observed and expected number of visits. The expected number of visits was calculated based on the number of open male and female flowers among the three stages (i.e., random visits).
In male flowers, the numbers of pollen grains removed were compared among the three stages and between the two bumble bee species with a Kruskal-Wallis test, as a two-way ANOVA was difficult to apply due to the shortage of specific data (see Results). If the difference was significant (P < 0.05), we performed pairwise comparisons between two stages using a Mann-Whitney U test with a sequential Bonferroni correction (Sokal and Rohlf, 1995
).
In female flowers, the flower stages, pollinator species, and their interaction effects on pollen deposition, the proportion of pollinated stigmas, and fruit set were tested by a two-way, fixed-effects model ANOVA. Multiple comparisons were performed using Scheffé's test. The hand-pollination experiment showed no significant differences in fruit set among the three stages (one-way ANOVA, F = 0.11, P = 0.89, N = 66); therefore, we pooled the data to compare them to fruit set following a single visit. In addition, this indicated that the stigma of any flowering stage is receptive to pollen. Pollen deposition was log transformed, and the proportions of pollinated stigmas and fruit set were arcsine transformed prior to analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
The sex of flowers significantly influenced the visitation pattern of the two pollinators: B. diversus and B. honshuensis consistently preferred male to female flowers (Table 2). Both species visited male flowers more frequently than female flowers on six out of seven and three out of five days, respectively (Table 2). These trends were also evident on the remaining days, although they were not significant (Table 2).
|
|
Pollen deposition and fruit set
A two-way ANOVA revealed that flower stage and pollinator species had significant effects on pollen deposition (Fig. 4A, Table 3). The amount of pollen deposited differed between bumble bee species, such that B. honshuensis deposited more pollen grains on stigmas than did B. diversus (mean ± 1 SE: B. honshuensis, 130.90 ± 19.85, N = 57, B. diversus, 112.63 ± 34.38, N = 48, Scheffé's test: P < 0.01; Fig. 4A). In contrast, the pattern of pollen deposition onto flowers in all three stages did not differ between the two bumble bee species, as indicated by a nonsignificant interaction term of flower stage x pollinator species (Fig. 4B, Table 3). Both bumble bee species deposited more pollen grains onto S-flowers than onto M-flowers (Scheffé's test: P < 0.05; Fig. 4A), but no difference was detected between the other flower stages.
|
|
Fruit set after a single visit by bumble bees also differed between the two pollinators; flowers visited by B. honshuensis showed higher fruit set than did those visited by B. diversus (Scheffé's test, P < 0.05; Fig. 4B). In contrast, the pattern of fruit set among the three stages did not differ between the two bumble bee species, as indicated by a nonsignificant interaction term of flower stage x pollinator species (Fig. 4B, Table 3). The difference in fruit set among flower stages paralleled the proportion of pollinated stigmas in the flowers (Fig. 4B). Both bumble bee species made a greater contribution to fruit set of S-flowers than M- flowers (Scheffé's test, P < 0.05; Fig. 4B). S- and L-flowers had the highest fruit-set values when they were visited by B. diversus and B. honshuensis, respectively (Fig. 4B).
The average fruit set in hand-pollinated flowers was 0.96 ± 0.05 (N = 66), which was higher than in any flower stage following a single visit (Mann-Whitney U test, P < 0.001 in all cases; Fig. 4B).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Inouye, 1980
; Nilsson, 1988
). Bumble bees choose flowers based on the energy value of rewards contained within the flowers; they can more efficiently forage in flowers whose sizes match their proboscis lengths (Heinrich, 1979
; Inouye, 1980
; Waser, 1983
). In C. stans, nectar is secreted daily from the base of the calyx tube in both male and female flowers as a reward for pollinators (Dohzono and Suzuki, 2002
); in addition, the length of the calyx tube decreases temporally (Fig. 1), and bumble bees forage in these flowers by inserting the proboscis into the calyx tube. Therefore, visitation preferences of the two bumble bee species can be inferred by morphological matching between the lengths of the calyx tube and the proboscis, indicating a specialized relationship (Fig. 2). In addition, it was found that bumble bees favored male over female flowers (Table 2), even though female flowers contain more nectar (higher sugar concentration) than do male flowers (Dohzono and Suzuki, 2002
). This may be explained by the fact that male flowers contain pollen grains as an additional reward for the pollinators and for their larvae (Heinrich, 1979
). Many studies have reported that bumble bees favor male flowers (male phase in dichogamous plants) that provide both pollen grains and nectar (Kay et al., 1984
; Delph and Lively, 1992
; Cresswell and Robertson, 1994
; Dohzono and Suzuki, 2002
).
We observed that the abundance of bumble bees varied daily at this study site (Fig. 2, Tables 1, 2); furthermore, the population size of B. honshuensis varied from year to year in many localities (Hiei and Suzuki, 2001
; Suzuki et al., unpublished data). A number of studies have demonstrated that the visitation frequency of pollinators varied among and within years, and even within populations (e.g., Herrera, 1989
; Utelli and Roy, 2000
; Ivey et al., 2003
). By depending on multiple pollinators, a plant is able to maintain frequent pollination, and can produce fruits during times of fluctuating pollinator abundance and/or species composition (Waser et al., 1996
). Therefore, the ability to use multiple pollinator species may be advantageous to C. stans in that it ensures pollinator visits and subsequent reproduction. Moreover, temporal changes in calyx tube length induce multiple visits to a flower by two bumble bee species, and influence pollen removal and pollen deposition, as discussed below.
Pollen removal
Pollen removal was influenced by the morphological correspondence between calyx tube length and proboscis length. Bombus diversus removed more pollen grains from flowers with long calyx tubes (Fig. 3), whereas B. honshuensis tended to remove more pollen grains from flowers with short calyx tubes (Fig. 3). Bumble bees attach pollen grains to their faces (prementum, clypeus, and mandible), and size correspondence between calyx tube length and proboscis length enhances the fit between the anther and the face of the pollinator. Thus, size correspondence is likely to increase pollen removal. In several studies, pollen removal was reported as being positively correlated with male reproductive success, as measured by paternity analysis (Broyles and Wyatt, 1990
; Ashman, 1998
), and the amount of pollen removed is considered a good marker of male reproductive success (Ashman, 1998
). On the other hand, it has also been reported that multiple visits increase the reproductive success of male plants (Harder and Thomson, 1989
; Harder, 1990
). For example, Galen (1992)
reported that floral characteristics that enhance pollinator visitation frequency have positive effects on paternity and gene flow in Polemonium. In either case, our data suggest that male reproductive success in C. stans peaks in flowers with long calyx tubes that are visited by B. diversus, and is possibly increased again in flowers with short calyx tubes that are visited by B. honshuensis (Fig. 3). However, with regard to male reproductive success in C. stans, the relative importance of the effects of pollen removal and visitation frequency remains to be determined.
Pollen deposition and fruit set
Pollen deposition, the proportion of pollinated stigmas, and fruit set are not always influenced by a morphological match between calyx tube length and proboscis length. In particular, B. diversus, which has a long proboscis, was less effective in L-flowers; this differed from the results in male flowers (Figs. 3, 4). A loose fit between the stigma and the face of B. diversus may result in this incongruence, because the calyx tubes of L-flowers were consistently shorter in female flowers than in male flowers (Fig. 1; Dohzono and Suzuki, 2002
). In addition to this, the nectar-sucking behavior of bumble bees influences pollen deposition and fruit set (Fig. 4). Because the glossa of B. diversus is a little shorter than the calyx tubes of female L-flowers (glossa length: 9.92 ± 0.37 mm ; Dohzono and Suzuki, 2002
), bumble bees show two types of foraging behavior: extending the prementum together with the glossa, or pushing the head into the calyx tube by extending only the glossa. In the former case, the flower may receive a small amount of pollen because of the loose contact between the face of the bee and the stigma. In the latter case, the flower may receive a large amount of pollen because the face of the bumble bee comes in close contact with the stigma. Therefore, the small or exceptionally large amounts of deposited pollen that we observed may correspond to the nectar-sucking behavior of B. diversus, i.e., extending the glossa with the prementum or extending only the glossa, respectively.
Our results also showed that B. honshuensis deposited more pollen grains on the stigma than did B. diversus (Fig. 4, Table 3). There are two possible reasons for this. The first lies in the nectar-sucking behavior of B. honshuensis. The proboscis length of B. honshuensis, which corresponds to the calyx tube length of female M-flowers, is shorter than that of B. diversus. When B. honshuensis forage in female L- and M-flowers, they extend either the prementum with the glossa or only the glossa (glossa length: 7.52 ± 0.16 mm ; Dohzono and Suzuki, 2002
), respectively. In either case, they must put their face into the stigma cluster, resulting in higher pollen deposition and fruit set (Fig. 4). The second reason is that the grooming behavior of bumble bees influences pollen deposition and fruit set. Bumble bees groom the pollen from various parts of the head, including the mouthparts and the antennae, using the forelegs; subsequently, the pollen is transferred to the pollen baskets on the hind legs for the larvae (Heinrich, 1979
). In our study, we observed the grooming behavior more often in B. diversus (I. Dohzono, personal observation). Therefore, the proportion of pollinated stigmas and fruit set among flowers visited by B. diversus may be lower than that of B. honshuensis (Fig. 4B). Because the fruit of C. stans is an achene, and each has only one seed, an important factor related to seed production is the number of pollinated stigmas in each flower. Thus, larger pollen deposits, a higher proportion of pollinated stigmas, and subsequently higher fruit set in the three stages may be the result of characteristic behavior of the two bumble bee species.
Comparing fruit set resulting from hand-pollination to fruit set following a single visit indicated that a single visit by bumble bees could not fertilize all ovules in a flower (Fig. 4B), suggesting that multiple visits will increase seed production. By changing calyx tube length, female flowers of C. stans have the opportunity to be visited by two bumble bee species, thereby compensating for the less effective pollen deposition in L-flowers by B. diversus, as compared to the highly effective pollination in M- and S-flowers by B. honshuensis. This strategy is advantageous to female plants in that it increases the genetic diversity of their offspring. Female plants are likely to increase their reproductive success by enhancing the quality or genetic diversity of their offspring; this is because male fitness is enhanced by increasing pollen removal, while female fitness is limited by ovule number (Marshall and Ellstrand, 1989
; Arnold, 1994
).
Adaptive significance of temporal changes in calyx tube length
Specialized pollination systems have evolved to enhance the pollination efficiency of the specific pollinator, which has led to morphological correspondence between flowers and pollinators (Grant and Grant, 1965
; Inouye, 1980
; Nilsson, 1988
). However, specialized plants are likely to suffer the risk of losing their chance of reproduction when the abundance of the specific pollinator decreases (Johnson and Steiner, 2000
). Clematis stans used two bumble bee species by changing its calyx tube length temporally, which increased the visitation frequency of the bumble bees (Fig. 2, Table 1) and enhanced the reproductive success of the plant (Figs. 3, 4). These results suggest that C. stans has established specialized relationships with each pollinator and has maintained reproductive stability by avoiding the risks of specialization. Therefore, our original hypothesis was supported.
Previous studies have suggested that morphological changes in flowers increase pollinator visits to new flowers (Gori, 1983
; Casper and La Pine, 1984
), which enhances reproductive success (Gori, 1989
; Jones and Cruzan, 1999
; Ida and Kudo, 2003
), or that they reduce selfing or geitonogamy (Jones and Cruzan, 1999
; Ida and Kudo, 2003
) by controlling pollinator visits within an individual plant. However, this is not the case with C. stans.
Ipomopsis aggregata can use two pollinators (hummingbirds and hawk moths) by changing flower color in accordance with changing pollinator abundance. However, corolla colors of individual plants, not individual flowers, "shift" from darker to lighter during the flowering season. Darker colors attract hummingbirds earlier in the season. After emigration of the hummingbirds, lighter colors attract the other pollinator, a hawk moth (Paige and Whitham, 1985
). In C. stans, the abundance of two bumble bee pollinators did not change drastically, unlike the situation with I. aggregata. Instead, the pollinator abundance varied daily (Fig. 2, Table 2). Under these conditions, temporal changes in calyx tube length may be more adaptive than flowers that do not change their shape; it is also economical because it eliminates costs required to produce different types of flowers.
In other species of the genus Clematis, temporal modification of floral morphology has never been described. In Japan, C. stans is closely related to C. speciosa (Makino) Makino. The flowers of C. speciosa are pendulous and have four sepals that do not curl up (calyx tube length: 20.39 ± 0.40 mm, N = 15; I. Dohzono, unpublished data). Clematis speciosa is distributed in the lowlands (below approximately 800 m altitude) of western Honshu, where only B. diversus occurs and visits the flowers. Therefore, calyx tube modification of C. stans may have evolved in response to the occurrence of two bumble bee species, B. diversus and B. honshuensis. Additionally, C. stans is dioecious, whereas C. speciosa has hermaphroditic flowers that are self-compatible (I. Dohzono, unpublished data). Such differences in sexuality may also affect the evolution of pollination systems that rely on changing calyx tube length.
|
|
FOOTNOTES |
|---|
5 dohzono-ikumi{at}c.metro-u.ac.jp
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Ashman T.-L. 1998 Is relative pollen production or removal a good predictor of relative male fitness? An experimental exploration with a wild strawberry (Fragaria virginiana, Rosaceae). American Journal of Botany 85: 1166-1171[Abstract]
Broyles S. B. R. Wyatt 1990 Paternity analysis in a natural population of Asclepias exaltata: multiple paternity, functional gender, and the "pollen-donation hypothesis". Evolution 44: 1454-1468[CrossRef][ISI]
Casper B. B. T. R. La Pine 1984 Changes in corolla color and other floral characteristics in Cryptantha humilis (Boraginaceae): cues to discourage pollinators?. Evolution 38: 128-141[CrossRef][ISI]
Cresswell J. E. A. W. Robertson 1994 Discrimination by pollen-collecting bumblebees among differentially rewarding flowers of an alpine wildflower, Campanula rotundifolia (Campanulaceae). Oikos 69: 304-308[CrossRef][ISI]
Delph L. F. C. M. Lively 1992 Pollinator visitation, floral display, and nectar production of the sexual morphs of a gynodioecious shrub. Oikos 63: 161-170[CrossRef][ISI]
Dohzono I. K. Suzuki 2002 Bumblebee-pollination and temporal change of the calyx tube length in Clematis stans (Ranunculaceae). Journal of Plant Research 115: 355-359[CrossRef][ISI][Medline]
Galen C. 1992 Pollen dispersal dynamics in an alpine wildflower, Polemonium viscosum. Evolution 46: 1043-1051[CrossRef][ISI]
Gilbert L. E. P. H. Raven 1975 Coevolution of animals and plants. University of Texas Press, Austin, Texas, USA
Gori D. F. 1983 Post-pollination phenomena and adaptive floral changes. In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 31–45. Van Nostrand Reinhold, New York, New York, USA
Gori D. F. 1989 Floral color change in Lupinus argenteus (Fabaceae): why should plants advertise the location of unrewarding flowers to pollinators?. Evolution 43: 870-881[CrossRef][ISI]
Grant V. K. A. Grant 1965 Flower pollination in the phlox family. Columbia University Press, New York, New York, USA
Harder L. D. 1990 Behavioral responses by bumble bees to variation in pollen availability. Oecologia 85: 41-47[CrossRef][ISI]
Harder L. D. J. D. Thomson 1989 Evolutionary options for maximizing pollen dispersal of animal-pollinated plants. American Naturalist 133: 323-344[CrossRef][ISI]
Heinrich B. 1979 Bumblebee economics. Harvard University Press, Cambridge, UK
Herrera C. M. 1989 Pollinator "abundance" morphology, and flower visitation rate: analysis of the "quantity" component in a plant-pollinator system. Oecologia 80: 241-248[ISI]
Hiei K. K. Suzuki 2001 Visitation frequency of Melampyrum roseum var. japonicum (Scrophulariaceae) by three bumblebee species and its relation to pollination efficiency. Canadian Journal of Botany 79: 1167-1174
Ida T. Y. G. Kudo 2003 Floral color change in Weigela middendorffiana (Caprifoliaceae): reduction of geitonogamous pollination by bumble bees. American Journal of Botany 90: 1751-1757
Inouye D. W. 1980 The effect of proboscis and corolla tube lengths on pattern and rates of flower visitation by bumblebees. Oecologia 45: 197-201[CrossRef][ISI]
Inoue K. 1986 Experimental studies on male and female reproductive success: effects of variation in spur length and pollinator activity on Platanthera mandarinorum ssp. hachijoensis (Orchidaceae). Plant Species Biology 1: 207-215
Ivey C. T. P. Martinez R. Wyatt 2003 Variation in pollinator effectiveness in swamp milkweed, Asclepias incarnata (Apocynaceae). American Journal of Botany 90: 214-225
Johnson S. D. K. E. Steiner 2000 Generalization versus specialization in plant pollination systems. Trends in Ecology and Evolution 15: 140-143
Jones C. E. M. B. Cruzan 1999 Floral morphological changes and reproductive success in deer weed (Lotus scoparius, Fabaceae). American Journal of Botany 86: 273-277
Kay Q. O. N. A. J. Lack F. C. Bamber C. R. Davies 1984 Differences between sexes in floral morphology, nectar production and insect visits in a dioecious species, Silene dioica. New Phytologist 98: 515-529[CrossRef][ISI]
Kearns A. C. D. W. Inouye 1993 Techniques for pollination biologists. University Press of Colorado, Niwot, Colorado, USA
Marshall D. L. N. C. Ellstrand 1989 Regulation of mate number in fruits of wild radish. American Naturalist 133: 751-765[CrossRef][ISI]
Nilsson L. A. 1988 The evolution of flowers with deep corolla tubes. Nature 334: 147-149[CrossRef]
Nilsson L. A. 1992 Orchid pollination biology. Trends in Ecology and Evolution 8: 255-259
Paige K. N. T. G. Whitham 1985 Individual and population shifts in flower color by scarlet gilia: a mechanism for pollinator tracking. Science 227: 315-317
Proctor M. P. Yeo A. J. Lack 1996 The natural history of pollination. Timber Press, Portland, Oregon, USA
Raguso R. A. 2001 Floral scent, olfaction, and scent-driven foraging behavior. In L. Chittka and J. D. Thomson [eds.], Cognitive ecology of pollination, 83–105. Cambridge University Press, UK
Schemske D. W. C. C. Horvitz 1984 Variation among floral visitors in pollination ability: a precondition for mutualism specialization. Science 225: 519-521
Scogin R. 1983 Visible floral pigments and pollinators. In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 277– 293. Van Nostrand Reinhold, New York, New York, USA
Silvertown J. D. Gordon 1989 A framework for plant behavior. Annual Review of Ecology and Systematics 20: 349-366[CrossRef][ISI]
Sokal R. R. F. J. Rohlf 1995 Biometry. Freeman, New York, New York, USA
Stebbins G. 1970 Adaptive radiation of reproductive characteristics in angiosperms. I. Pollination mechanisms. Annual Review of Ecology and Systematics 1: 307-326
Suzuki K. 1992 Bumble bee pollinators and pollination ecotypes of Isodon umbrosus and I. shikokianus (Lamiaceae). Plant Species Biology 7: 37-48
Utelli A. B. B. A. Roy 2000 Pollinator abundance and behavior on Aconitum lycoctonum (Ranunculaceae): an analysis of the quantity and quality components of pollination. Oikos 89: 461-470[CrossRef][ISI]
Waser N. M. 1983 The adaptive nature of floral traits: ideas and evidence. In L. A. Real [ed.], Pollination biology, 241–285. Academic Press, New York, New York, USA
Waser N. M. L. Chittka M. V. Price N. M. Williams J. Ollerton 1996 Generalization in pollination systems, and why it matters. Ecology 77: 1043-1060[CrossRef]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
2) and Fisher's exact test. Expected values are shown in parentheses. The null hypothesis of bee visitation is independent of flowering stages
