SEARCH

SEARCH BY CITATION

Keywords:

  • dichogamy;
  • geitonogamy;
  • inbreeding;
  • Orchidaceae;
  • pollen discounting;
  • self-pollination

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Temporal separation of male and female phases in hermaphrodite flowers (dichogamy) is proposed to reduce self-pollination, both within and among flowers. Darwin and later workers suggested that protandry (the most common form of dichogamy, in which the male phase precedes the female phase) should be most effective in reducing geitonogamous (between-flower) self-pollination when pollinators forage upwards from older female-phase flowers to younger male-phase flowers on vertical inflorescences.
  • 2
    We tested this hypothesis by manipulating the extent of protandry in artificial inflorescences of the orchid Satyrium longicauda Lindl. and using stained pollen to quantify self-pollination and pollen export.
  • 3
    Upper flowers of non-protandrous inflorescences received more self-pollen through geitonogamy than lower flowers, unlike protandrous inflorescences. Protandry reduced absolute levels of self-pollination, as the amount of removed pollen involved in self-pollination was three times greater in non-protandrous than in protandrous inflorescences. This high level of self-pollination reduced the pollen available for export, as the ratio of pollen export to self-pollination declined with increasing self-pollination, indicating the occurrence of pollen discounting.
  • 4
    This study represents the first direct measurement of the effects of protandry on the pollination process, and indicates that the evolution of protandry in plants could be driven strongly by the consequences of this trait for male mating success.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Approximately 80% of all flowering plants produce both female and male reproductive organs (Proctor, Yeo & Lack 1996). Potential benefits of hermaphroditism include economizing on resources, because floral advertising and rewards serve both sex functions simultaneously; and reproductive assurance through self-fertilization when mates and/or pollinators are rare or absent (Charnov, Maynard Smith & Bull 1976; Charnov 1982; Charlesworth & Charlesworth 1987; Lloyd 1987). The main disadvantage of hermaphroditism arises from the deleterious effects of pollinator-mediated self-pollination on both female and male function, including a reduction in the pollen available for export (pollen discounting: Holsinger & Thomson 1994; Harder & Wilson 1998), stigma or stylar clogging (Lloyd & Webb 1986), reduction in the availability of ovules for cross-fertilization (ovule discounting: Barrett, Lloyd & Arroyo 1996), reduced seed production following self-fertilization (seed discounting: Lloyd 1992; Herlihy & Eckert 2002), and inbreeding depression in the progeny of self-compatible taxa (Darwin 1876; Charlesworth & Charlesworth 1987).

Dichogamy, or temporal separation of male and female states, has long been considered a mechanism that reduces the likelihood of interference among genders in hermaphrodite flowers (Darwin 1876; Lloyd & Webb 1986; Bertin & Newman 1993; Harder, Barrett & Cole 2000). Although dichogamy is likely to reduce or eliminate within-flower self-pollination (cf. Lloyd & Webb 1986), its most important function may be to reduce self-pollination among flowers on the same inflorescence (geitonogamy). Protandry, in which the male state precedes the female state, should be most effective in terms of reducing geitonogamous self-pollination when flowers are arranged in racemose inflorescences and visited by insects that tend to move upwards (acropetal behaviour). In theory, older female-phase flowers would be visited first and receive mainly cross-pollen, whereas younger, male-phase flowers at the top would export pollen with little likelihood of it being deposited on self-stigmas (Best & Bierzychudek 1981; Bertin & Newman 1993; Jordan & Harder 2006). Despite the logic of this prediction, it has seldom been tested experimentally. Harder et al. (2000) manipulated the adichogamous flowers of Eichhornia paniculata to simulate the sexual segregation within inflorescences that would result from contrasting patterns of dichogamy. In this experiment, which involved upward-moving bumble bees, they found that outcrossing and male siring success, estimated from allozyme markers, were maximized with male (stigma-less) flowers above female (stamen-less) flowers on inflorescences, which is consistent with the predicted consequences of protandry for pollen fates. In another study using allozyme markers, Routley & Husband (2003) found that dichogamous inflorescences had greater outcross siring success than adichogamous inflorescences, but selfing rates did not differ between the two inflorescence types.

Allozyme and other molecular markers provide only indirect estimates of the consequences of dichogamy for pollen fates, such as the rate of self-pollination, because post-pollination processes, such as self-incompatibility and inbreeding depression, can affect the presence of these markers in progeny. Pollen fates are best estimated by direct tracking of pollen itself, although this is very difficult in plants with granular pollen (Snow & Lewis 1993), with the rare exception of those with a pollen colour polymorphism (cf. Thomson & Thomson 1989). Orchids, on the other hand, offer unsurpassed opportunities to study pollen fates, because orchid pollen can be colour-stained (Peakall 1989; Johnson, Neal & Harder 2005) or microtagged (Nilsson, Rabakonandrianina & Pettersson 1992), allowing it to be tracked after dispersal. Pollen staining offers excellent insights into male function in orchids with massulate pollinia, as it does not appear to affect either pollinarium removal or subsequent pollen deposition on stigmas (Peakall 1989; Johnson et al. 2005).

Protandry in orchids was first described by Darwin (1862), who elucidated the unusual floral development in Spiranthes and Listera. In these genera, the stigmas of newly opened flowers are concealed either by the column being directed toward the lip, or by the horizontal position of the labellum. During this phase, which can last from 1 to 4 days, flowers act only as pollen donors. Thereafter, upward movement of the column or downward movement of the labellum exposes the stigma, and flowers can act as pollen receivers. This form of protandry is spread across the tribes Cranichideae, Orchideae, Neottieae and Cymbidieae (Darwin 1862; Ackerman 1975, 1977; Catling 1983; Sipes & Tepedino 1995; Singer & Sazima 2001a, 2001b; Singer 2002; Singer & Koehler 2003). We have observed that a different form of protandry, associated with gradual development of a sticky stigmatic mucilage, occurs in many orchids in the tribe Diseae. Flowers with this form of protandry can donate pollen, and are thus effectively male, from the first day of anthesis, but become effectively female several days after anthesis, when they develop the sticky stigmatic mucilage that allows pollen receipt (see Methods: Study species). Preliminary observations showed that this form of protandry occurs in the South African moth-pollinated orchid Satyrium longicauda Lindl.

The aim of this study was to determine the consequences of protandry for pollen fates, using S. longicauda as a study species. After first verifying that S. longicauda has protandrous flowers, we used pollen-staining techniques to compare pollen fates in artificial inflorescences of protandrous flowers vs those comprised of non-protandrous flowers. In particular we asked (1) whether protandry reduces self-pollination, and (2) whether self-pollination reduces the pollen available for export (pollen discounting sensu Harder & Wilson 1998).

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

study species

We performed our experiment in a large population of S. longicauda occupying 8 ha of moist montane grassland on the farm Wahroonga (30°08′ E, 29°37′ S, altitude approx. 1200 m), a South African Natural Heritage site, close to Pietermaritzburg. Flowering of S. longicauda at this site starts during late December, peaks during mid-January and finishes in early February. The fruits mature over 3 weeks. Flowering plants produce a single racemose inflorescence with pinkish white, scented, nectar-rewarding flowers (mean number of open flowers per plant = 18·6, N = 34).

The S. longicauda complex consists of several ecotypes that differ in their floral biology and pollination systems (S.D.J., unpublished data). The ecotype studied by Harder & Johnson (2005) is pollinated by both noctuid moths and hawkmoths, and has flowers that wilt rapidly after pollination. The form used in this study has large flowers pollinated almost exclusively by the hawkmoth Basiothia schenki Möschler (see Fig. 2a; S.D.J., unpublished data), and has flowers that remain open for up to 10 days after pollination. Voucher specimens of plants from the study population are deposited in the Bews Herbarium, Pietermaritzburg. The entrance to each of the two nectar spurs in S. longicauda flowers is situated behind the rostellum, such that the moth proboscis contacts one of the two viscidia, resulting in removal of the pollinium (Johnson 1997). Unlike many orchids, the pollinarium of S. longicauda does not undergo a bending movement to prevent self-pollination, so that pollen can be deposited on the stigma of any flower visited subsequently. A S. longicauda pollinium is segmented into many massulae (mean ± SE = 320 ± 10·8 massulae at our study site, N = 40 plants), so that pollen from a single pollinium can disperse to stigmas of many recipient flowers.

image

Figure 2. (a) Flowers of Satyrium longicauda being probed by a hawkmoth (Basiothia schenki) which is carrying pollinaria on the upper part of its proboscis. (b) Natural (left) and artificial (right) inflorescences of S. longicauda. Scale bars = 50 mm.

Download figure to PowerPoint

Although pollinia can be removed from S. longicauda flowers on the first day of anthesis, the stigmas gradually develop a sticky mucilage over several days. In preliminary tests to establish the influence of stigmatic mucilage on pollen receipt, we brushed a single pollinium over the stigma of each open flower on 34 inflorescences that had been bagged from the bud stage, and recorded the deposition of pollen on stigmas. The mean (±SE) percentage of open flowers in a male phase (when stigmas do not accept pollen) was 32 ± 14·5%. The most massulae adhered to stigmas in lowermost flowers on inflorescences (Fig. 1). To estimate the beginning of the female phase, we marked two flowers at anthesis on each of 10 plants, then on each of the following 4 days brushed pollinia over the stigmas and recorded the number of adhered massulae. The stigma was regarded as receptive if at least one massula adhered to the stigma. Flowers reached the female phase after 3·2 ± 0·7 days (N = 20).

image

Figure 1. Hand-pollination experiment demonstrating changes in mean (± SE) stigmatic pollen receipt with flower position (n = 12 plants). The solid curve indicates non-linear regression of original data, y = 206·1/{1 + exp[–(x − 8·4)/1·8]}.

Download figure to PowerPoint

experimental design

We assessed the consequences of protandry for pollen dispersal by constructing artificial 10-flowered inflorescences of two kinds: ‘protandrous’ inflorescences resembled the natural inflorescences with the four uppermost flowers having non-sticky stigmas and the six lowermost flowers having sticky stigmas. The ‘non-protandrous’ inflorescences consisted entirely of flowers with sticky stigmas. Inflorescences of both kinds were made by mounting virgin S. longicauda flowers spirally on a wooden stick with thin white wire (Fig. 2b). Virgin flowers were harvested from c. 30 natural inflorescences that had been bagged from the bud stage. To minimize the chance of flowers wilting, artificial inflorescences were constructed within a few hours prior to evening foraging by hawkmoths.

Before exposing inflorescences to pollinators, we marked pollinia by injecting 0·5 µl of histochemical stain into each anther sac of all flowers with a calibrated microsyringe (Peakall 1989). Stains used were gentian violet (pre-mixed medicinal prescription-Alpha) initially added to protandrous inflorescences, and rhodamine B (0·2% concentration) initially added to non-protandrous inflorescences. These stains were switched between the treatments mid-way through the experiment, even though Peakall (1989) and Johnson et al. (2005) showed that staining does not affect the probability of pollinarium removal and deposition.

The protandrous and non-protandrous artificial inflorescences were placed in pairs within the S. longicauda population. Pairs of artificial inflorescences were separated by at least 20 m to limit the possibility of stained cross-pollen being mistaken for self-pollen on a plant that had also been stained with the same colour. After artificial inflorescences had been exposed to pollinators for 24 h, we transported them to the laboratory and counted removed pollinaria and stained and unstained massulae deposited per stigma. We also examined the stigmas of all plants (c. 100) in the population and recorded the presence of any stained massulae. We measured the distance between any plant that had received stained massulae and the nearest artificial inflorescence the pollinia of which had been stained with the same colour. Flowers that had received stained massulae were removed from the population after each exposure. This procedure was repeated 12 times during January 2005 and 2006 using the same patches, resulting in 35 and 34 replicates of successfully visited protandrous and non-protandrous inflorescences, respectively.

hawkmoth behaviour

Nectar accumulates in unvisited S. longicauda flowers, such that older flowers with sticky stigmas contain more nectar than newly opened flowers with non-sticky stigmas (7·9 vs 1·9 µl, respectively, N = 20 flowers of each kind). Thus our protandrous inflorescences would, on average, have contained slightly less nectar per flower, at least when they are visited for the first time. To assess whether this difference affected hawkmoth behaviour, we observed hawkmoths foraging on the two kinds of artificial inflorescence. For each inflorescence on which we observed a hawkmoth visit, we counted the probed flowers and measured the duration of probes of individual flowers. We also recorded the typical number of flowers probed on natural inflorescences. We used single-factor anova to compare the mean number of flowers probed on protandrous, non-protandrous and natural inflorescences. Probing times per flower for protandrous and non-protandrous inflorescences (these data were not collected for natural inflorescences) were compared using a t-test for independent samples.

analysis of pollen fates

Injection of stains into anthers did not always stain all massulae in a pollinium. However, the type of stain had no influence on the proportion of stained massulae per pollinium (t = 0·99, df = 38, P > 0·32) or the mean number of pollinaria removed per plant (interaction between treatment and stain type: F1,54 = 0·36, P > 0·54). To avoid underestimating pollen fates because of understaining, we used the mean number of stained massulae per pollinium as a correction factor (mean ± SD = 291 ± 66·8, N = 40) to calculate the actual number of stained massulae removed from flowers.

Although the study was conducted during 2 years, we used pooled data as no pollen fate parameter varied significantly between years. We used the non-parametric Mann–Whitney U-test to compare responses between treatments for several variables with poorly defined distributions, including the number of pollinia removed from an inflorescence, the number of cross-massulae deposited on flowers of artificial inflorescences, proportions of removed massulae involved in self- and cross-pollination, proportions of deposited massulae resulting from self-pollination, and proportions of dispersed massulae involved in cross-pollination.

The relations between pollen removal and self-pollination, pollen removal and pollen export, and self-pollination and pollen export were explored using simple linear regression of ln(X + 1) transformed values. The significance of differences in regression slopes for the two inflorescence types were assessed using ancova. We assessed the independent effect of self-pollination on pollen export with analyses of general linear models that considered treatment as a categorical factor and pollen removal and self-pollination as continuous covariates. As we applied logarithmic transformation to both dependent and independent variables, the partial regression coefficient (b) for covariate variable indicates whether the value of the dependent variable proportionately increases (b > 1) or decreases (b < 1) with increases in the value of the predictor. Tests concerning partial regression coefficients involved single-sample t-tests.

To test whether removal of at least one pollinarium or receipt of at least one self-massula differed with flower position within the two types of artificial inflorescence, we used analysis of generalized linear models (genstat, Lawes Agricultural Trust, IACR, Rothamsted, UK; Payne et al. 1993) with the logit link because of the binary nature of the response variable. We calculated ratios of mean deviance changes (quasi-F values), which follow the F distribution asymptotically (Payne et al. 1993).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

pollinator behaviour

The hawkmoth B. schenki visited S. longicauda inflorescences only during the first 30 min after dusk. Hawkmoths typically started foraging from the lowermost flowers on inflorescences and always moved upwards on inflorescences, but on several occasions feeding commenced on the middle or upper flowers.

Overall, hawkmoths behaved similarly on natural and experimental inflorescences. Moths probed equivalent numbers of flowers on protandrous (mean ± SE = 4·1 ± 0·8), non-protandrous (5·7 ± 0·9) and natural inflorescences (5·9 ± 0·8; single-factor anova, F2,56 = 1·66, P > 0·05). In addition, their probing time (s) per flower did not differ significantly between protandrous (2·2 ± 0·2) and non-protandrous inflorescences (2·5 ± 0·2; t-test, t = −0·98, df = 25, P > 0·05).

pollen fates

Moths removed more pollinaria from protandrous inflorescences than from non-protandrous ones, but this difference was not significant (Fig. 3a). Flower position within the inflorescence affected pollinarium removal similarly for both protandrous and non-protandrous inflorescences; the proportion of flowers with at least one pollinium removed did not vary significantly with flower position (interaction between inflorescence type and position: quasi-F1,67 = 0·003, P > 0·05; Fig. 4a).

image

Figure 3. Pollen fates of protandrous and non-protandrous artificial inflorescences. (a) Number of pollinia removed; (b) number of self-massulae deposited to flowers of artificial inflorescences; (c) proportion of removed massulae involved in self-pollination; (d) proportion of removed pollen involved in cross-pollination; (e) proportion of self-pollen in total pollen load deposited on stigmas of artificial inflorescences; (f) proportion of dispersed massulae involved in cross-pollination. Box-plots indicate medians, 10th, 25th, 75th and 90th percentiles. Sample sizes are given above each box-plot. Z values are for Mann–Whitney U-tests.

Download figure to PowerPoint

image

Figure 4. Relations of the mean proportions of flowers with (a) at least one pollinarium removed; (b) receipt of at least one self-massula to flower position within artificial inflorescences of Satyrium longicauda. The lines represent logistic regression solutions.

Download figure to PowerPoint

We found strongly contrasting patterns in the incidence of receipt of self- and cross-pollen in the two inflorescence types. Non-protandrous inflorescences received twice as many self-massulae as protandrous inflorescences (Fig. 3b). The median proportion of removed pollen deposited on self-stigmas in the case of non-protandrous inflorescences was 2·8 times greater than for protandrous inflorescences (Fig. 3c). However, because non-protandrous inflorescences also received more cross-pollen than protandrous inflorescences, the proportion of self-pollen in the total pollen load deposited on stigmas did not differ significantly between inflorescence types (Fig. 3e). Protandry decreased the incidence of geitonogamy. The minimum estimate of geitonogamy, based on the number of flowers that received self-pollen, but from which no pollinarium was removed, was significantly lower in protandrous inflorescences (median = 1, lower quartile = 1, upper quartile = 2, N = 29) than in non-protandrous ones (median = 2, lower quartile = 1, upper quartile = 3, N = 34; Mann–Whitney U-test: Z = 2·24, P < 0·05).

Self-pollination varied positively with pollen removal for both inflorescence types (protandrous: R2 = 0·22, P < 0·05; non-protandrous: R2 = 0·34, P < 0·005; Fig. 5a). The slopes of these relationships did not differ significantly (F1,46 = 0·34, P > 0·05). The proportion of flowers that received at least one self-massula differed significantly according to position in the two inflorescence types (interaction between inflorescence type and position: quasi F1,67 = 33·1, P < 0·0001; Fig. 4b), with self-pollination decreasing towards the top in protandrous inflorescences, but increasing towards the top in non-protandrous inflorescences.

image

Figure 5. Relations between pollen removal, self-pollination and pollen export to other plants for protandrous (•) and non-protandrous (○) artificial inflorescences: (a) self-pollination vs pollen removal; (b) pollen export to other plants vs pollen removal; and (c) pollen export to other plants vs self-pollination. Lines represent linear regression solutions for protandrous (solid line) and non-protandrous (dashed line) treatments. In (c) the dotted line depicts an equal ratio of self-pollination and pollen exported to other plants.

Download figure to PowerPoint

The proportion of removed pollen that was exported to stigmas of other plants did not differ significantly between inflorescence types (Fig. 3d). However, the proportion of pollen exported to stigmas of other plants as a fraction of total pollen export to all stigmas was significantly higher for protandrous plants (Fig. 3f). Pollen export to other plants increased with pollen removal for both inflorescence types (protandrous: R2 = 0·49, P < 0·001; non-protandrous: R2 = 0·28, P < 0·01; Fig. 5b), but pollen export increased more strongly with pollen removal in protandrous plants (regression slopes: 2·04 vs 0·87; F1,46 = 5·23, P < 0·05).

In general, pollen export increased with self-pollination for both inflorescence types (protandrous: R2 = 0·25, P < 0·01; non-protandrous: R2 = 0·33, P < 0·005; Fig. 5c). The slopes of these relationships did not differ significantly (F1,46 = 0·02, P > 0·05) and were <1, indicating that pollen export declined proportionally with increasing self-pollination. The relationship between pollen export and self-pollination remained significant when pollen removal and treatment were included as predictor variables using multiple regression. The partial regression coefficient for pollen removal (b ± SE = 1·01 ± 0·29) did not differ significantly from 1 (t46 = 0·03, P > 0·05), indicating that pollen export used the same proportion of removed massulae, regardless of the number of massulae removed. The partial regression coefficient associated with self-pollination (b ± SE = 0·28 ± 0·12) was significantly <1 (t46 = −5·9, P < 0·0001), so the ratio of exported massulae to self-deposited massulae declined with increasing self-pollination. This indicates that self-pollination significantly reduced a plant's opportunities to export pollen.

Plants from which hawkmoths removed pollinaria exported pollen to as many as eight recipient plants, with a median of three recipients. The numbers of recipient plants did not differ significantly between the two inflorescence types (Mann–Whitney U-test: Z = 0·13, P > 0·05). Protandrous inflorescences dispersed pollen to more flowers (median = 10, lower quartile = 9, upper quartile = 15) than did non-protandrous inflorescences (median = 8, lower quartile = 5, upper quartile = 17), but this difference was not significant (Mann–Whitney U-test: Z = 2·20, P > 0·05). Most pollen was exported to nearby plants, with 61 and 64% of all dispersals occurring within a 30-cm radius around protandrous and non-protandrous inflorescences, respectively (Fig. 6). This localized pollen dispersal probably reflects the combined effects of hawkmoths moving short distances between plants and limited carry-over of pollen between successively visited flowers.

image

Figure 6. Pollen dispersal for protandrous (shaded bars) and non-protandrous (open bars) inflorescences of Satyrium longicauda.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Our results reveal that protandry has profound consequences for pollen fates in S. longicauda. Most evident was that more self-pollen is deposited on upper flowers when inflorescences lack gender segregation (Fig. 4b), indicating that protandry reduces geitonogamous self-pollination between lower and upper flowers. For the inflorescence as a whole, total self-pollination increased by a factor of two and the proportion of removed pollen that becomes involved in self-pollination was increased by a factor of three when inflorescences lack protandry (Fig. 3c). The absence of protandry thus could have implications for both female function (by leading to increased inbreeding) and male function (by reducing pollen export). Indeed, the species is susceptible to inbreeding depression, as self-pollinated fruits contain fewer seeds with embryo than those arising from cross-pollination (unpublished data).

Protandry may affect female function weakly in S. longicauda because, although it decreases self-pollination markedly, it also decreases pollen import. Thus the typical proportion of self-pollen on stigmas, an important determinant of the selfing rate in self-compatible species, is apparently not affected by protandry (Fig. 3e). Although the absence of other pollen-fate studies precludes testing the generality of this finding, it is interesting that Routley & Husband (2003) found that protandry had no effect on the selfing rate in Chamerion angustifolium. These experimental results are also consistent with broad surveys of Bertin & Newman (1993) and Routley et al. (2004), which show that protandry is equally common in self-incompatible and self-compatible plants, and therefore unlikely to be linked evolutionarily with avoidance of inbreeding. Similarly, the model of Sargent et al. (2006) indicates that interference among genders in hermaphrodite flowers can favour dichogamy even in the absence of inbreeding depression.

Although not quantified in this study, ovule discounting (Barrett, Lloyd & Arroyo 1996) is another mechanism whereby self-pollination can have implications for female function in flowers. In particular, the lower absolute levels of self-massulae deposited on protandrous inflorescences (Fig. 3b) may free up ovules for cross-fertilization following additional pollination events that occur after most of the plant's own pollinaria have been removed. This could enhance female function, as cross-fertilized ovules have significantly lower levels of embryo abortion than do self-fertilized ovules (unpublished data). However, with the possible exception of ovule discounting, the observed reduction in the proportion of removed pollen involved in self-pollination probably has greater implications for male function.

Self-pollination, especially geitonogamy, can limit male mating success by reducing the opportunities for pollen export. The strong positive relationship between self-pollination and pollen export in S. longicauda seems at first to contradict the principle of pollen discounting. However, as pointed out by Harder & Wilson (1998), this positive relationship simply reflects that pollinators remove variable amounts of pollen that has the potential to be exported (cf. Johnson et al. 2005). Thus removal of a large number of pollinaria creates more opportunities for both more self-pollination and pollen export. Instead, the evidence for pollen discounting lies in analysis of the slopes of these relationships, which were significantly <1 (Fig. 5c), indicating that pollen export to other plants declines proportionally with self-pollination. Consequently a removed massula has less chance of reaching a stigma on another plant if self-pollination claims many massulae, compromising pollen export per removed pollinarium in S. longicauda. This conclusion is consistent with the contention of Lloyd (1988, 1992) that geitonogamous self-pollination reduces outcross siring opportunities for plants. Thus by reducing geitonogamous self-pollination, protandry increases male mating success (Fig. 3f), as also demonstrated by Harder et al. (2000).

Orchids with their aggregated pollen may be particularly vulnerable to the deleterious effects of pollinator-mediated self-pollination. Lack of floral rewards (Johnson et al. 2004), pollinarium bending (Peter & Johnson 2006) and pollination-induced reduction in display sizes (Harder & Johnson 2005) have all been suggested as mechanisms that promote cross-pollination in orchids. Our study shows that protandry in orchids can have a similar function. This idea is supported by the occurrence of various forms of protandry in other orchids, such as Spiranthes (Catling 1983; Sipes & Tepedino 1995), Piperia (Ackerman 1977), Goodyera (Ackerman 1975), Prescottia stachyodes (Singer & Sazima 2001a), Sauroglossum elatum, Mesadenella cuspidate (Singer 2002) and Erythrodes arietina (Singer & Sazima 2001b), which are also nectar-rewarding and have sectile pollinia lacking other mechanisms that could prevent self-pollination. An exception is Notylia nemorosa, which produces fragrance as a reward for Euglossine bees and has solid pollinia (Singer & Koehler 2003). Delayed stigma stickiness, the mechanism of protandry that we have demonstrated to occur in S. longicauda, may be fairly widespread among orchids, but has probably been under-reported because it is a cryptic trait that is not evident from casual inspection of flowers. In conclusion, our study represents the first direct measurement of the effects of protandry on the pollination process and offers insight into the pollen fates that shape the evolution of dichogamy in hermaphroditic plants. In particular, our results confirmed that protandry reduces absolute levels of self-pollination and enhances the efficiency of pollen export by reducing the fraction of removed pollen that is involved in self-pollination.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank L.D. Harder for very helpful comments on the manuscript. This work was supported by postdoctoral funding from the South African National Research Foundation to J.J., and grants AV0Z60870520 to the Institute of Systems Biology and Ecology AS CR and MSM 6007665801 to the Faculty of Biological Sciences of University of South Bohemia.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Ackerman, J.D. (1975) Reproductive biology of Goodyera oblongifolia (Orchidaceae). Madroño 23, 191198.
  • Ackerman, J.D. (1977) Biosystematics of the genus Piperia Rydb. Botanical Journal of the Linnean Society. 75, 245270.
  • Barrett, S.C.H., Lloyd, D.G. & Arroyo, J. (1996) Stylar polymorphisms and the evolution of heterostyly. Narcissus (Amaryllidaceae). Floral Biology: Studies on Floral Evolution in Animal-pollinated Plants (eds D.G.Lloyd and S.C.H.Barrett), pp. 339376. Chapman & Hall, New York.
  • Bertin, R.I. & Newman, C.M. (1993) Dichogamy in angiosperms. Botanical Reviews 59, 112152.
  • Best, L.S. & Bierzychudek, P. (1981) Pollinator foraging on foxglove (Digitalis purpurea): a test of a new model. Evolution 36, 7079.
  • Catling, P.M. (1983) Pollination of Northeastern North American Spiranthes (Orchidaceae). Canadian Journal of Botany 61, 10801093.
  • Charlesworth, D. & Charlesworth, B. (1987) Inbreeding depression and its evolutionary consequences. Annual Review of Ecology and Systematics 18, 237268.
  • Charnov, E.L. (1982) The Theory of Sex Allocation. Princeton University Press, Princeton, NJ, USA.
  • Charnov, E.L., Maynard Smith, J. & Bull, J.J. (1976) Why be an hermaphrodite? Nature 263, 125126.
  • Darwin, C. (1862) On the Various Contrivances by which Orchids are Fertilised by Insects, 2nd edn revised. John Murray, London.
  • Darwin, C.R. (1876) The Effects of Cross and Self Fertilisation in the Vegetable Kingdom. John Murray, London.
  • Harder, L.D. & Johnson, S.D. (2005) Adaptive plasticity of floral display size in animal-pollinated plants. Proceedings of the Royal Society of London B 272, 26512657.
  • Harder, L.D. & Wilson, W.G. (1998) A clarification of pollen discounting and its joint effects with inbreeding depression on mating system evolution. American Naturalist 152, 684695.
  • Harder, L.D., Barrett, S.C.H. & Cole, W.W. (2000) The mating consequences of sexual segregation within inflorescences of flowering plants. Proceedings of the Royal Society of London B 267, 315320.
  • Herlihy, C.R. & Eckert, C.G. (2002) Genetic cost of reproductive assurance in a self-fertilizing plant. Nature 416, 320323.
  • Holsinger, K.E. & Thomson, J.D. (1994) Pollen discounting in Erythronium grandiflorum: mass-action estimates from pollen transfer dynamics. American Naturalist 144, 799812.
  • Johnson, S.D. (1997) Insect pollination and floral mechanisms in South African species of Satyrium (Orchidaceae). Plant Systystematics and Evolution 204, 195206.
  • Johnson, S.D., Peter, C.I. & Ågren, J. (2004) The effects of nectar addition on pollen removal and geitonogamy in the non-rewarding orchid Anacamptis morio. Proceedings of the Royal Society of London B 271, 803809.
  • Johnson, S.D., Neal, P.R. & Harder, L.D. (2005) Pollen fates and the limits on male reproductive success in an orchid population. Biological Journal of the Linnean Society 86, 175190.
  • Jordan, C.Y. & Harder, L.D. (2006) Manipulation of bee behavior by inflorescence architecture and its consequences for plant mating. American Naturalist 167, 496509.
  • Lloyd, D.G. (1987) Allocations to pollen, seeds, and pollination mechanisms in self-fertilizing plants. Functional Ecology 1, 8389.
  • Lloyd, D.G. (1988) Benefits and costs of biparental and uniparental reproduction in plants. The Evolution of Sex (eds R.E.Michod and B.R.Levin), pp. 233252. Sinauer, Sunderland, MA, USA.
  • Lloyd, D.G. (1992) Self- and cross-fertilization in plants. II. The selection of self-fertilization. International Journal of Plant Sciences 153, 370380.
  • Lloyd, D.G. & Webb, C.J. (1986) The avoidance of interference between the presentation of pollen and stigmas in angiosperms. I. Dichogamy. New Zealand Journal of Botany 24, 135162.
  • Nilsson, L.A., Rabakonandrianina, E. & Pettersson, B. (1992) Exact tracking of pollen transfer and mating in plants. Nature 360, 666667.
  • Payne, R.W., Lane, P.W., Digby, P.G.N. et al. (1993) GENSTAT 5 Reference Manual. Clarendon Press, Oxford, UK.
  • Peakall, R. (1989) A new technique for monitoring pollen flow in orchids. Oecologia 79, 361365.
  • Peter, C.I. & Johnson, S.D. (2006) Anther cap retention prevents self-pollination by elaterid beetles in the South African orchid Eulophia foliosa. Annals of Botany 97, 345355.
  • Proctor, M.C.F., Yeo, P. & Lack, A. (1996) The Natural History of Pollination. Timber Press, Portland, OR, USA.
  • Routley, M.B. & Husband, B.C. (2003) The effect of protandry on siring success in Chamerion angustifolium (Onagraceae) with different inflorescence sizes. Evolution 57, 240248.
  • Routley, M.B., Bertin, R.I. & Husband, B.C. (2004) Correlated evolution of dichogamy and self-incompatibility: a phylogenetic perspective. International Journal of Plant Sciences 165, 983993.
  • Sargent, R.D., Mandegar, M.A. & Otto, S.P. (2006) A model of the evolution of dichogamy incorporating sex-ratio selection, anther-stigma interference, and inbreeding depression. Evolution 60, 934944.
  • Singer, R.B. (2002) The pollination biology of Sauroglossum elatum Lindl. (Orchidaceae: Spiranthinae): moth-pollination and protandry in neotropical Spiranthinae. Botanical Journal of the Linnean Society 138, 916.
  • Singer, R.B. & Koehler, S. (2003) Notes on the pollination biology of Notylia nemorosa (Orchidaceae): do pollinators necessarily promote cross pollination? Journal of Plant Research 116, 1925.
  • Singer, R.B. & Sazima, M. (2001a) The pollination mechanism in three sympatric Prescottia (Orchidaceae: Prescottinae) species in Southeastern Brazil. Annals of Botany 88, 9991005.
  • Singer, R.B. & Sazima, M. (2001b) Flower morphology and pollination mechanism in three sympatric Goodyerinae orchids from southeastern. Brazilian Annals of Botany 88, 989997.
  • Sipes, S.D. & Tepedino, V.J. (1995) Reproductive biology of the rare orchid, Spiranthes diluvialis– breeding system, pollination, and implications for conservation. Conservation Biology 9, 929938.
  • Snow, A.A. & Lewis, P.O. (1993) Reproductive traits and male fertility in plants: empirical approaches. Annual Review of Ecology and Systematics 24, 331351.
  • Thomson, J.D. & Thomson, B.A. (1989) Dispersal of Erythronium grandiflorum pollen by bumblebees: implications for gene flow and reproductive success. Evolution 43, 657661.