Floral Longevity and Nectar Secretion of Platanthera chlorantha (Custer) Rchb. (Orchidaceae)

Abstract

Flowering and nectar secretion were studied in Platanthera chlorantha in two years. Nectar was secreted and accumulated in this orchid’s spur, originating from part of the labellum. The nectary spur was, on average, 32 mm long. It produced 6·86 µl nectar in 1999 and 7·84 µl in 2000. The number of flowers per inflorescence and the volume of nectar secreted per flower were not correlated. Nectar secretion and flower longevity differed depending on pollination and flower position in the inflorescence. Among pairs of pollinated and unpollinated flowers there was no difference in the volume of nectar produced; however, the life span of pollinated flowers was shorter than that of unpollinated ones. Within an inflorescence, the lowest‐positioned flowers had the largest nectar production and the longest life compared with flowers positioned higher up.

INTRODUCTION

Orchid flowers are highly dependent on their pollinators for reproductive success. Many flowers utilize characteristic shapes and colours or secrete species‐specific scents to attract pollinators (Kaiser, 1993). However, nectar is the most frequently sought reward for pollen vectors of orchids. Neiland and Wilcock (1994, 1995, 1998) indicated that nectariferous orchids are more successful in setting fruit than nectarless species. Pollination systems such as mimicry, nectar deceit or offering alternative floral rewards, e.g. pollen or floral oils, appeared less successful than nectar in improving fruit set.

White moth flowers of Platanthera chlorantha attract pollinators using a characteristic scent and offer them copious amounts of nectar (Nilsson, 1978; Stpiczyńska and Pielecki, 2002). The ‘investment’ in nectar production and secretion has been shown to require a considerable expenditure of energy (Pyke, 1991). According to Southwick (1984), nectar production in Asclepias syriaca may use up to 37 % of a plant’s photosynthetic energy. Koopowitz and Marchant (1998) estimated that the average energy expenditure for nectar production per plant of Aerangis verdickii was 684 J during one season. Thus, it may be expected that after pollination the process of nectar secretion should be reduced or inhibited completely.

In numerous orchid species, reproductive success is pollinator limited (Zimmerman and Aide, 1989; Johnson and Bond, 1992; Calvo, 1993; Mattila and Kuitunen, 2000). Low pollinator visitation may be one reason why orchids produce long‐lived flowers. However, as in nectar production, the costs of flower longevity may also be high (Ashman and Schoen, 1997).

There is little information available concerning the pattern of nectar secretion by orchids during the lifespan of individual flowers or on fluctuations in nectar production pre‐ and post‐pollination. The aim of this study was to monitor nectar secretion during the life span of P. chlorantha flowers, with special emphasis on any changes occurring after pollination. The research concentrated on: (1) the dynamics of nectar secretion and accumulation in relation to lifespan/age of flowers and pollination status of flowers; (2) determining the effect of flower position in the inflorescence on nectar secretion and flower longevity.

MATERIALS AND METHODS

Study site and plant material

The study was conducted during two seasons on Platanthera chlorantha (Custer) Rchb. (Orchidaceae) originating from a forest ‘Las Stary Gaj’ (Tilio‐Carpinetum) near Lublin, Poland, and covering a ground area of approx. 50 m².

During two seasons, whole plants were transplanted into pots and transported to the laboratory before flower opening and the beginning of nectar secretion. Plants were grown at 20–25 °C and at a relative humidity of 70–75 %. All flowers on the inflorescences were numbered from the bottom to the top; labels with numbers were attached to the bracts. For each individual flower, the start of nectar secretion, the day of anthesis (petals fully expanded) and wilting of the perianth were noted. The spur length in each flower was measured using a ruler.

Measurements of nectar secretion

The dynamics of nectar production were determined over the whole secretory period by monitoring changes in nectar level inside the almost transparent spur. The length of the nectar column in the spur was measured in each individual flower three times per day: 0800, 1400 and 2000 h. Measurements were carried out on 102 flowers from five inflorescences in 1999 and on 79 flowers from 15 inflorescences in 2000. To estimate the volume of nectar that had accumulated inside the spur, the length of the nectar column was measured in 44 additional flowers from 15 inflorescences in 1999. Subsequently, nectar from each flower was collected using 20 µl glass microcapillary tubes, and expelled separately onto Whatman No. 1 chromatography paper. The volume of nectar in the spur of individual flowers was calculated on the basis of the nectar spot diameter (Dafni, 1992). Nectar levels in the spur were related to estimated nectar volumes by the linear regression: y = 0·3066x + 0·7633 (r² = 72·52; t = 10·53; d.f. = 42, P < 0·001), where y is the nectar volume and x is the nectar level in the spur. The same equation was used in 2000. Using this equation we were able to estimate nectar volume in the same individual flower over the length of the study without removing the nectar from the spur.

Effect of pollination

Flower longevity and the dynamics of nectar secretion during the life span of the flower were compared between pairs of neighbouring pollinated and unpollinated flowers. To make that comparison, flowers were hand‐pollinated with one pollinarium, typically originating from a different plant (geitonogamy was performed for five flowers), on the second day of anthesis. In 1999, comparisons involved 18–26 pairs of flowers from five inflorescences, four to six pairs per inflorescence, depending on number of flowers (the difference in number of pairs resulted from the fact that, in the case of 16 flowers, detailed data on the start of nectar secretion and anthesis were not noted). In 2000, nectar secretion throughout the lifetime of the flower was observed on 22 pairs of pollinated and unpollinated flowers from 15 inflorescences, with one or two pairs per inflorescence.

Capsule formation was observed in the pollinated flowers from the above‐mentioned pairs and in the additionally pollinated 38 flowers from which nectar was sampled for chemical analysis.

Effect of flower position in the inflorescence

To determine the effect of flower position in the inflorescence on nectar secretion and flower longevity, data were collected in 1999 from five inflorescences that had been divided into three equal parts (low, middle and high), each part consisting of two to five flowers depending on inflorescence size. In 2000, comparisons were carried out on two or three flowers from the three different regions of 15 inflorescences. Differences in sample size resulted from the fact that, in some flowers, spurs had started to abort during the course of nectar secretion or secreted only immeasurably small amounts of nectar.

Statistical analysis

Results were compared using Stat Graphics Plus 1.4 Win software, using one‐way ANOVA, analysis of regression and 95 % Tukey confidence intervals. Data concerning pairs of pollinated and non‐pollinated flowers were compared using the Kołmogorov–Smirnov non‐parametric test.

RESULTS

In 1999, blooming of P. chlorantha started on 1 June and ended on 23 June; however, owing to the early, warm spring in 2000, blooming commenced earlier and lasted longer, from 19 May to 25 June.

Nectar secretion

The nectary spur was, on average, 31·5 mm long, and this figure did not differ significantly between years. The coefficient of variation in spur length was low, being 3·65 and 5·47 in 1999 and 2000, respectively. Externally, nectar was observed to appear inside the spur about 1·5 d before anthesis. It started to accumulate in the apex of the spur as small droplets. Examined flowers accumulated, on average, of 6·9 ± 4·1 µl nectar in 1999 and 7·8 ± 2·7 µl in 2000 (Table 1). The total number of flowers produced by each inflorescence varied widely, and ranged from five to 25. The positive correlation between the number of flowers in the inflorescence and the maximum volume of nectar secreted per flower in 1999 and 2000 was statistically insignificant (r² = 10·74, d.f. = 100 and r² = 23·05, d.f. = 77, respectively, P > 0·05). In both years, about one‐third of spurs in the flowers monitored were fully filled with nectar. When spurs were fully filled with nectar, secretion did not end, and a large droplet of nectar usually appeared above the spur entrance or even on the labellum surface.

Table 1.

Maximum accumulation of nectar as nectar column length and corresponding nectar volume in P. chlorantha flowers during the two seasons

Nectar accumulation	Mean ± s.d. nectar column length (mm)	Mean ± s.d. nectar volume (µl)	CV (%)
1999 N = 5; n = 102	19·88 ± 10·1	6·86 ± 4·1	41·99
2000 N = 15; n = 79	23·08 ± 9·37	7·84 ± 2·7	35·88

N, Number of inflorescences; n, number of flowers; CV, coefficient of variation.

The dynamics of nectar secretion in an individual flower could be differentiated into four phases. In phase I, secretion was very slow and the nectar level in the spur increased constantly at a rate of approx. 1–3 mm a day. This phase was the longest, and lasted 7–16 d. Phase II characterized rapid secretion, when the nectar level in the spur increased by over 20 mm during 6–12 h and nectar heights reached a maximum. Cessation of nectar secretion occurred during phase III and lasted for approx. 0·5–2 d; nectar levels in the spurs remained unchanged. During the final stage, phase IV, which lasted approx. 2 d, the nectar level began to fall constantly since nectar was resorbed, and the spur emptied. A decrease (of approx. 1–2 mm) in nectar level within the spur during phases I–III was noted on rare occasions. This experiment, in which the nectar level was measured three times a day, did not show any daily peaks in nectar secretion, i.e. there was no regular pattern of enhanced nectar secretion resulting in increasing nectar levels in the spur.

Effect of pollination

Differences in the dynamics of nectar secretion and flower longevity occurred after pollination (Table 2). The duration of the secretory period was shorter in pollinated flowers, and these flowers also reached the peak of nectar secretion faster. The maximum accumulation of nectar in the pairs of pollinated and unpollinated flowers did not differ significantly. Differences in persistence of the phase of cessation and the phase of resorption in pollinated and unpollinated flowers were also insignificant.

Table 2.

Comparison of nectar secretion and floral longevity between pairs of hand‐ pollinated and unpollinated flowers

	1999		2000
Year	N	Mean ± s.d.	N	Mean ± s.d.
Presence of nectar in the spur (d)
Pollinated	18	9·6 ± 2·4	22	11·5 ± 1·9
Unpollinated	18	11·9 ± 2·4	22	16·4 ± 2·5
		n.s.		K–S = 2·72**
Maximum nectar volume (µl)
Pollinated	26	6·71 ± 2·6	22	7·98 ± 2·7
Unpollinated	26	7·42 ± 3·0	22	8·06 ± 2·6
		n.s.		n.s.
Peak of secretion (d)
Pollinated	18	7·0 ± 1·3	22	9·0 ± 1·2
Unpollinated	18	8·22 ± 2·1	22	14·0 ± 2·6
		K–S = 2·29**		K–S = 2·67**
Nectar cessation (h)
Pollinated	26	41·3 ± 20·6	22	53·27 ± 19·8
Unpollinated	26	41·5 ± 17·4	22	48·64 ± 14·4
		n.s.		n.s.
Nectar resorption (h)
Pollinated	26	47·9 ± 16·6	22	39·72 ± 12·7
Unpollinated	26	41·8 ± 17·9	22	38·09 ± 11·8
		n.s.		n.s.
Flower longevity (d)
Pollinated	18	8·28 ± 1·5	22	9·8 ± 1·6
Unpollinated	18	11·33 ± 2·85	22	15·95 ± 2·3
		K–S = 2·0**		K–S = 4·4**

K–S, Kołmogorov–Smirnov test; **P < 0·01.

When the dynamics of secretion in the pairs of neighbouring flowers (pollinated and unpollinated) were compared, it was found that the level of nectar began to increase quickly 3–5 d after pollination and only as it came close to its end did secretion in the neighbouring, unpollinated flower accelerate.

The longevity of pollinated flowers was significantly shorter than that of unpollinated ones (Table 2). Thickening of the ovary was noticeable 4–7 d after pollination in 86 of the observed flowers. The perianth dehisced completely, usually 6–7 d after pollination (range 5–10 d, s.d. = 1·33). At this time, capsules became characteristically erect and pressed against the inflorescence axis. Hand‐pollinated flowers always set fruit whilst unpollinated flowers did not form any capsules.

Effect of flower position

Nectar secretion and flower longevity varied depending on flower position in the inflorescence (Table 3). The period of secretion in flowers from the higher section of the inflorescence was shorter than that in the lowest flowers, and the maximum volume of secreted nectar was also lower (however, the difference was non‐significant in 2000). The higher positioned flowers had the shortest longevity.

Table 3.

Differences in mean longevity of nectar presence in the spur, mean maximum volume of nectar per flower and mean flower longevity as a function of flower position in the inflorescence

	Flower position
	Low	Middle	High	LSD
Presence of nectar in the spur (d)
1999	11·2a	10·8a	8·2b	2·23
N = 5	n = 22	n = 22	n = 22
	F = 6·03, d.f. = 65, P = 0·004
2000	16·2a	15·7a	13·3b	2·39
N = 15	n = 21	n = 25	n = 18
	F = 4·69, d.f. = 63, P = 0·0128
Maximum level of nectar (µl)
1999	7·8a	6·2ab	5·6b	2·15
N = 5	n = 22	n = 26	n = 24
	F = 3·15, d.f. = 75, P = 0·049
2000	8·7a	8·1a	6·6a	–
N = 15	n = 19	n = 23	n = 14
	F = 2·1, d.f. = 55, P = 0·133
Flower longevity (d)
1999	10·2a	10·26a	8·3b	1·9
N = 5	n = 22	n = 23	n = 21
	F = 4·53, d.f. = 65, P = 0·0145
2000	15·1a	14·6ab	12·2b	2·7
N = 15	n = 21	n = 25	n = 26
	F = 4·01, d.f. = 71, P = 0·0225

Data analysed by one‐way ANOVA. Distribution into homogeneous group with 95 % Tukey confidence interval.

Means within a row followed by the same superscript are not significantly different.

N, number of inflorescences; n, number of flowers.

DISCUSSION

Nectary spur length and nectar secretion

In orchids, length of the nectary spur is closely related to the length of the pollinator tongue (Nilsson, 1981, 1992; Robertson and Wyatt, 1990; Johnson, 1997; Johnson and Steiner, 1997). The spurs of species (e.g. Platanthera chlorantha) visited by Lepidoptera are relatively long. In Lublin, flowers of P. chlorantha developed spurs with an average length of approx. 32 mm. The spur length in this species noted by Nilsson (1978, 1981, 1985) ranged from 17 to 45 mm. A spur length greater than 20 mm occurred in Platanthera species pollinated by Lepidoptera (Smith and Snow, 1976; Inoue, 1986a; Robertson and Wyatt, 1990), whereas P. stricta, a species pollinated by different insects with relatively short mouthparts, possessed a short spur, only 3·97 mm long (Patt et al., 1989).

Some differences in spur length in P. chlorantha growing in wooded regions and open ground were noted by Nilsson (1978); however, the differences were small, only approx. 3·5 mm. In this study spur length was characterized by a low coefficient of variation. Similarly, observations by Inoue (1986a) revealed that the coefficient of variation in spur length of P. metabifolia was small, the spur being under strong genetic control, rarely subjected to phenotypic plasticity. The presence of a spur increases effective visitation. In P. mandarinorum spp. hachijoensis, Inoue (1986b) showed that experimental shortening of the spur restricted pollinator activity, and plants with shortened spurs had decreased reproductive success, proportional to the spur‐length reduction. Moreover, the experiment of Neiland and Wilcock (1994) indicated that the presence of the spur alone, even without nectar, increases the probability of pollination.

In P. chlorantha, nectar appeared in the spur before anthesis and was available until the end of the flower’s life. Flowers secreted a maximum quantity of nectar exceeding 6 µl. Moth‐pollinated orchid species examined to date have produced various amounts of nectar in theirs spurs, ranging from 0·5 to 20 µl (Robertson and Wyatt, 1990; Galetto et al., 1997; Johnson, 1997; Koopowitz and Marchant, 1998; Luyt and Johnson, 2001). Regardless of the volume of nectar produced, nectar availability significantly increases the effective visitation rate and reproductive success in orchids (Rodrígues‐Robles et al., 1992; Ackerman et al., 1994; Neiland and Wilcock, 1998).

The number of flowers in the inflorescence of P. chlorantha was not connected with the volume of nectar produced per flower. Some authors have shown that abundant flowering and fruiting in one season may result in lower future growth and flowering of plants, connected with resource limitation (Montalvo and Ackerman, 1987; Snow and Whigham, 1989; Ackerman and Montalvo, 1990; Mattila and Kuitunen, 2000).

Monitoring nectar secretion three times a day throughout the secretory period did not reveal any daily peaks in nectar production by P. chlorantha. Accordingly, in this species, nectar production was not related to the period of pollinator activity. However, the lack of variability in nectar production during defined periods of the day might also result from the conditions in which the experiment was carried out (in the laboratory with relatively constant temperature and humidity). Coincidence of daily pollinator activity with daily maximum nectar production has been observed in Mystacidium venosum (Luyt and Johnson, 2001). However, it should be pointed out that in Platanthera, the pollinators’ activity is stimulated not only by nectar, but also by scent, emitted periodically during the dusk and night (Nilsson, 1978).

Effect of pollination

The experiment involving pairs of pollinated and unpollinated flowers showed that pollination affected the duration of nectar secretion, by shortening the life span of the flower, but did not affect the volume of nectar produced. However, comparing nectar production in all pollinated vs. unpollinated flowers showed that pollinated flowers secreted more nectar that unpollinated ones (Stpiczyńska, 2001). Measurements of nectar column length in pollinated and virgin flowers of Aerangis verdickii did not reveal any significant differences (Koopowitz and Marchant, 1998). Similarly, hand‐pollination had no influence on nectar production in Alstroemeria aurea, which might reflect the low cost of nectar production (Aizen and Basilio, 1998). Surprisingly, nectar secretion in P. chlorantha was enhanced 3–5 d after pollination and, as it came close to its end, secretion in the neighbouring, unpollinated flower accelerated. Such consumption of energy for nectar secretion in pollinated flowers seems unprecedented. It might be because the flowers remained attractive in order to attract additional pollinators. Pollination in P. chlorantcha did not restrict stigma receptivity (Stpiczyńska, 2003). The presence of nectar after pollination undoubtedly enhances the probability of additional pollen load. Neiland and Wilcock (1995) showed that multiple pollinations were advantageous to fruiting and seed set in orchid flowers.

Rapid resorption of nectar once peak secretion was achieved was noted in the spurs of P. chlorantha, irrespective of whether or not pollination had been effected. Generally, during the first three phases of the secretory period, a decrease in nectar level, which might be indicative of evaporation, was not observed. Moreover, a sharp decrease in nectar sugar content in the oldest flowers was noted (Stpiczyńska, 2001; Stpiczyńska and Pielecki, 2002). Nectar resorption is not an unusual floral feature and has been noted in other species: resorption of uncollected nectar was described in Brassica napus (Burqúez and Corbet, 1991), Cucurbita pepo (Nepi et al., 1996a, b, 2001a), Eucalyptus sp. (Davis, 1997) and Carum carvi (Langenberger and Davis, 2002). Nectar was also resorbed near the end of a flower’s lifetime in Combretum fruticosum (Bernardello et al., 1994), Mandevilla pentlandiana (Torres and Galetto, 1998) and Linaria vulgaris (Nepi et al., 2001b, 2003). Plants can reabsorb nectar from old, wilting flowers to utilize the carbon in seed development. Constituents from reabsorbed nectar have been used in production of a stigmatic exudate (Shuel, 1961). Resorption of sugars reduces the viscosity of nectar, meaning pollinators can take nectar quickly (Cruden at al., 1983; Nicolson, 1995). Moreover, if a pollinator does not remove the nectar, reabsorption may be a form of protection against nectar thieves or may eventually reduce any negative effects of post‐pollination visits (Burqúez and Corbet, 1991). In Mystacidium venosum, nectar reabsorption during the night followed a decrease in pollinator activity (Luyt and Johnson, 2001). Aerangis verdickii is an example of an orchid species in which reabsorption of ungathered nectar, 48 h after pollination, has been demonstrated (Koopowitz and Marchant, 1998). However, nectar resorption in P. chlorantha was not a direct consequence of pollination, and it was observed 6–7 days afterwards.

Floral persistence is a feature demonstrated in research on European orchids, whose reproduction depends upon specific pollinators, usually uncommon in the natural environment, and the flowers can still remain attractive to visitors following pollination (Ashman and Shoen, 1994; Neiland and Wilcock, 1995). P. chlorantha has long‐lasting flowers, with the perianth remaining fresh for about 2 weeks unless it is pollinated. Pollinated flowers are shorter‐lived, although the perianth did not exhibit changes for 4–5 d after pollination. These observations are in agreement with observations on P. bifolia (Tollsten and Bergström, 1989; Tollsten, 1993) and Mystacidium venosum (Luyt and Johnson, 2001). In the case of other orchid species, such as Stenorrhynchos orchioides, Pelexia bonariensis and Beladea dutraei, unpollinated flowers lasted 2–3 weeks, but complete floral senescence was observed within 2–4 d after pollination (Galetto et al., 1997). As suggested by these authors, floral senescence was probably connected with reallocation of resources to developing seeds. Flower fading may also depend on pollen dosage. When flowers of Cleistes divaricata receive a low quantity of pollen, wilting of the perianth was as slow, as in unpollinated flowers, thus increasing the chances of additional pollination (Gregg, 1991). Similarly, removal of pollinaria may induce flower senescence, as in Chloraea alpina (Clayton and Aizen, 1996) or Mystacidum venosum (Luyt and Johnson, 2001), and nectar removal in Stenorrhychos orchioides (Galetto et al., 1997) significantly affected flower longevity.

Effect of flower position

The total amount of nectar produced was related to flower position in the inflorescence. Flowers from lower regions of an inflorescence produced larger amounts of nectar than those positioned higher up. The distribution of nectar reward may be a crucial factor in determining the foraging behaviour of pollinators (Cruden et al., 1983). Differences in nectar distribution may also limit geitonogamy, making some flowers in the inflorescence unattractive to pollinators. This may be significant for self‐compatible species, such as P. chlorantha. However, the regular pattern of nectar distribution, with distal flowers producing less nectar, was probably connected with resource limitation.

The longevity of lower‐ or middle‐positioned flowers was greater than that of upper‐positioned ones. This difference could result from resource limitation in the upper flowers. However, according to Ishii and Sakai (2001), prolonged longevity of lower‐ and middle‐positioned flowers could be advantageous because it results in a large inflorescence display size. A large display attracts more pollinators and increases reproductive success (Rodríguez‐Robles et al., 1992; Murren and Ellison, 1996). Prolonged floral longevity of the upper flowers would not benefit the plant and would only require resources to maintain the flowers.

ACKNOWLEDGEMENTS

I thank Dr Arthur Davis (Department of Biology, University of Saskatchewan) for valuable and helpful comments on the manuscript.

Received: 22 January 2003; ; Returned for revision: 18 March 2003. Accepted: 16 April 2003    Published electronically: 12 June 2003