The conspicuously large bracts influence reproductive success in Thunia alba (Orchidaceae)

Abstract

In angiosperms, diverse floral traits are adaptations to various selective pressures and ecological functions. So far, studies on floral traits in orchids have focused primarily on the labellum but never on bracts. A bumblebee-pollinated and rewarding terrestrial or epiphytic herb, Thunia alba (Lindley) H. G. Reichenbach (Orchidaceae), has conspicuously large and curly bracts that enclose the spur and pedicel of flowers. We hypothesized that these large bracts could protect spurs against nectar robbers. To confirm this hypothesis, we experimentally removed the bracts to record the changes in visiting behavior of mutualistic pollinators and antagonistic nectar robbers and evaluated their effects on reproduction success. Our result revealed that Bombus breviceps, the only pollinator of T. alba, shifted to nectar robbery when the bracts were removed, and the proportion of robbed flowers also significantly increased. Thunia alba was found to be pollinator limited regardless of whether in intact treatment or removed bract treatment. Removal of bracts had no effect on the visiting frequency of B. breviceps, but it reduced male and female reproductive success. These findings indicate that, under complex environmental pressures with limited pollination, large bracts can protect against nectar robbers and enhance the fitness of T. alba.

摘要

大而显著的苞片影响了笋兰的繁殖成功

在被子植物中，多样的花部特征是为了适应各种选择压力和生态功能。到目前为止，对兰科植物花部特征的生态功能研究主要集中在唇瓣，而从未有对苞片的研究。本研究以一种熊蜂授粉并且提供花蜜报酬的地生或附生兰科植物笋兰(Thunia alba)为研究材料，其大而显著形似舟状的苞片包裹着花蜜距和花梗。我们猜想显著的大苞片能够保护蜜距，抵御盗蜜者。为了验证这一假设，我们通过实验去除苞片，记录传粉者和盗蜜者访花行为的变化，并评估了它们对笋兰繁殖成功的影响。研究结果表明，当苞片被去除后，唯一的传粉者短头熊蜂(Bombus breviceps)转变为盗蜜者，并且被盗蜜的花的比例也显著增加。无论是花苞片被去除的处理组还是未处理的对照组，笋兰都是受传粉者限制的。去除苞片对传粉者的访花频率并没有影响，但是显著降低了笋兰雄性和雌性的繁殖成功率。本研究表明，在复杂的授粉限制的环境压力下，大而显著的苞片可以防止花被盗蜜者盗蜜从而提高笋兰的整体适合度。

INTRODUCTION

The diverse floral traits of angiosperms are adaptations to various selection pressures however they may be opposed to mutualistic pollinators and antagonistic nectar robbers (Waser and Price 1983; Nilsson 1988; Irwin et al. 2001; Wu et al. 2021; Cha et al. 2023). Orchids are known for exhibiting an astonishing diversity of species and having evolved complex and elaborate floral traits through pollinator-mediated processes (Darwin 1877; Christenhusz and Byng 2016). Therefore, most orchid species have evolved specialized pollination systems that are pollinated by only one or a few pollinator species (Tremblay 1992). The reproductive success of orchids is influenced not only by pollinator limitations but also by many floral traits, such as flower size, color, rewards, spur characteristics, inflorescence height, and odor (Tremblay et al. 2005). Labellum structures of orchids are often highly modified, exhibit marked diversity, and are likely to serve different functions, such as a visual attractant and landing platform (van der Pilj and Dodson 1966; Rudall and Bateman 2002). Therefore, the labellum effect on reproductive success is widely studied in orchids (Dickson and Petit 2006; Benitez-Vieyra et al. 2009; Suetsugu et al. 2022). The bract is also an important floral structure in orchids, but its function and effect on reproductive success have never been reported.

The bract is a special trait of flowering plants that may fundamentally protect the reproductive tissues (Carlson and Harms 2007). Many studies have shown that bract functions are related to plant reproductive success and also provide protection against adverse abiotic factors, such as rainwash, low temperatures, and intense solar radiation (Armbruster 1997; Herrera 1997; Strauss and Whittall 2006; Sun et al. 2008; Yang and Sun 2009). In addition to the above resistance to abiotic factors, bracts may also have resistance to biotic factors, such as protection against seed predators and nectar robbers (Armbruster 1997).

Thunia has about six species, and they are terrestrial or epiphytic orchids, mainly distributed in Southeast Asia. Out of which, only Thunia alba has been recorded in southwest China and has been found in Xishuangbanna, Yunnan Province (Gao et al. 2014). Thunia alba is a terrestrial or epiphytic species, occurring at altitudes ranging from 1200 to 2300 m and often found growing on tree trunks or rocks (Gao et al. 2014). It had a racemose inflorescence at the top of the plants, usually flowering in June or July, and mature fruits in November or December. In our field surveys on orchid species diversity in the Laobanzhang ancient tea garden in southwest Yunnan, we observed that T. alba has conspicuous large bracts that seem to enclose the spur of the flowers. In view of the above literatures, we propose the hypothesis that the large bract could protect the spur against nectar robbers, thus improving the reproductive success of T. alba. So as to assess the effect of bracts on reproductive success in T. alba, we have performed a manipulation experiment with bract removal from one population of T. alba. The following specific questions were addressed: (i) What are the pollinators of T. alba? (ii) What changes in pollinator visiting behavior occur after removing bracts? (iii) What is the effect of removing bracts on floral longevity, the proportion of robbed flowers, and reproduction success?

MATERIALS AND METHODS

Study sites

This study was conducted in an ancient tea garden in the Laobanzhang village in Xishuangbanna, Yunnan Province, southwestern China. The latitude and longitude of the study plot are 21° 43ʹ 41″ N and 100° 30ʹ 23″ E, with an elevation between 1700 and 1900 m a.s.l. The average annual rainfall of this region is between 1342 and 1540 mm, and the annual average temperature is about 18.7°C, which is a subtropical plateau monsoon climate. The study was conducted during the flowering and fruiting seasons of T. alba from 2022 to 2023.

Floral traits measurements

We have recorded the number of inflorescences, the number of flowers per inflorescence, and the flowering time. Furthermore, we have measured the size of bracts, pedicels, and spurs and observed the duration and color change of bracts during the flowering and fruiting periods. To study nectar volume and nectar sucrose concentration in flowers, we randomly bagged 10 inflorescences before anthesis in June 2022. We used 10-µl Sigma “micro-cap” calibrated capillary tubes (Sigma Chemical Co., St. Louis, USA) to measure the nectar volumes of 35 newly opened flowers during 12:00–15:00 in June 2023. Additionally, the nectar sucrose concentration of 25 samples, each of which has a mixture of two flowers, was measured with a hand-held, temperature-compensated refractometer (Eclipse, Bellingham & Stanley Ltd, UK) at the same time.

Bract manipulation experiments

To examine the role of the bracts in the reproductive phase of T. alba, we have randomly selected two populations distant over 300 m for two kinds of treatment. One with the intact (n = 40 inflorescences, 180 flowers) and another with removed bracts (n = 43 inflorescences, 187 flowers). All the inflorescences were selected from plants that had a similar reproductive phase.

To assess the effect of bract removal on flower physical longevity, we have recorded the single-flower longevity for both the treatments. To estimate the defense of bracts against nectar robbers, we have recorded the proportion of robbed flowers in both treatments.

Pollinator observations

The pollinators visit were observed during the flowering seasons in 2022–2023. Ten blooming inflorescences from both treatments were randomly selected and observed for pollinator visits from 09:00 to 19:00 per day. Each individual pollinator visiting the flowers of the two treatments was carefully observed and recorded to ascertain their visiting behavior, which included the frequency of visit, number of flowers visited, and time taken per inflorescence. All the visitor species were photographed during the observation periods. To identify the species, we have captured three to five individuals from each pollinator at the end of the observation period.

Breeding system and reproduction success

To detect self-compatibility, three different hand-pollination treatments were conducted at the study site in 2022. (i) Bagging: 47 flowers from 10 inflorescences were bagged and separated from the pollinators. (ii) Self-pollination: 49 flowers from 10 inflorescences were bagged before flower opening, and the flowers were hand-pollinated with pollen from the self or another flower on the same plant. (iii) Cross-pollination: 46 flowers from 10 inflorescences were bagged before flower opening, and the flowers were hand-pollinated with pollen from other individuals of the same species that were growing at least 50 m away. The fruit set in each treatment was recorded at the end of the flowering period.

To estimate the effect of bract removal on the reproductive success of T. alba, 40 inflorescences of intact treatment and 43 inflorescences of removed bract treatment were marked before anthesis in 2023, and pollinia removal, deposition, and fruit set were recorded at the ends of flowering.

Statistical analyses

Each experiment consisted of the specified number of individuals and was carried out once or twice during the year. The dates of length of bracts, pedicels, and spurs, flower longevity, frequency of pollinator visits, number of flowers visited, and time of visit were subjected to T tests. We used Yate’s continuity-corrected Chi-square test to analyze the difference in the proportion of nectar robbery visits between the two treatments. We used the generalized linear model (GLM) to compare the difference in proportion of robbed flowers, pollinia removal, deposition, and fruit set between the two treatments. Moreover, the comparison of fruit sets between the two treatments with hand cross-pollination was also subjected to the GLM. All statistical analyses were performed by SPSS ver. 25.0 for Windows.

RESULTS

Floral traits

Each individual plant of T. alba produced several inflorescences with an average of 4.42 ± 1.10 (n = 83) large, white flowers (Fig. 1a). The yellow stripes on the labellum acted as a nectar guide to attract the insects. The bracts were large (length: 4.01 ± 0.17 cm, width: 2.90 ± 0.21cm, n = 32) and curly, wrapping the pedicels (2.97 ± 0.18 cm, n = 30) and spurs (1.40 ± 0.09 cm, n = 31; mean ± SD, Fig. 1b). Our results showed that the bracts are significantly longer than the spurs and pedicels (t_spur = 76.44, t_pedicel = 23.14, all P < 0.0001). The bracts were persistent and lasted up to 2 months. They continued from the flower budding period until the fruit set, and after pollination, when the immature fruits grew up to 5–6 cm long, the bracts began to wither (Fig. 1c), thus they did not affect seed dispersal. The color of the bracts was also observed to change from light green to yellow-white from the budding period to the end of the flower period. The spurs were observed to be filled with nectar. The nectar volume and sugar content were recorded at 4.29 ± 1.11 µl (n = 35) and 38.93 ± 3.06% (mean ± SD; n = 25), respectively. The single-flower longevity was not significantly different between the intact (13.27 ± 1.03 days, n = 30) and removed bract (13.07 ± 0.93, n = 30) treatments (mean ± SD; t = 0.78, P = 0.44).

Changes in pollinator behavior

On observing T. alba for more than 70 h from 2022 to 2023, it was found that the Bombus breviceps is the only and effective pollinator at the study site, as they were found to carry the pollinaria of T. alba (Fig. 1d). Under natural conditions, most of the pollinators have landed on the labellum, then climbed into the flowers along the passage to forage nectar. Only B. breviceps carries pollinia on the back when the pollinator contacts the stigma (Fig. 1e). But when the pollinator (B. breviceps) visits the flowers of the removed bracts, it makes a hole in the middle of the spur to steal the nectar directly, and fails to carry the pollinia, and also fails to complete pollination in its successive flower visits (Fig. 1f). We have recorded that B. breviceps had 16 normal visits and 3 nectar robbery visits in intact treatment, while in removed bract treatment, there were 5 normal visits and 21 nectar robbery visits. The number of visits by nectar robbery was significantly higher in the removed bract treatment than the intact treatment (χ² = 18.62, df = 1, P < 0.0001, Fig. 2). Moreover, our results showed that the proportion of robbed flowers was significantly higher in the removed bracts treatment (90.25% ± 1.88%, n = 43) than in the intact treatment (15.43% ± 3.13%, mean ± SE; n = 40, P < 0.0001, Fig. 3). The visiting frequency of B. breviceps showed no significant difference between the two treatments (Intact: 2.48 ± 0.31, n = 21; Removed bract: 2.39 ± 0.25, n = 23, t = 0.21, P = 0.83), but the visiting number of flowers per inflorescence was significantly higher in the removed bract treatment (3.54 ± 0.21, n = 39) than the intact treatment (2.54 ± 0.16, n = 37, t = 3.72, P < 0.001), and the visiting time per flower was significantly lower in the removed bract treatment (Intact: 5.53 ± 0.49 s, n = 15; Removed bract: 3.06 ± 0.21 s, n = 16; mean ± SE; t = 4.60, P < 0.001, Fig. 4).

Effect of removed bracts on reproduction success

In our hand-pollination treatments, all T. alba flowers that were bagged before opening failed to develop fruits (0%, n = 10). Of the flowers that were self- and cross-pollinated by hand, 90.98% ± 3.61% (n = 10) and 92.67% ± 2.86% (mean ± SE; n = 10) flowers set fruit, respectively. Our results showed that pollinia removal, deposition, and fruit set were significantly lower in the removed bract treatment (15.47% ± 3.85%, 9.73% ± 3.37%, 8.18% ± 3.39%, n = 43) than the intact treatment (56.50% ± 2.65%, 30.79% ± 2.13%, 28.71% ± 2.08%, n = 40; mean ± SE; χ²_removal = 78.74, χ²_deposition = 28.72, χ²_{fruit set} = 27.38; all df = 1, all P < 0.001, Fig. 5). Fruit set was significantly lower for both treatments and when compared with the hand of cross-pollinated (χ²_intact = 85.01, χ²_removed = 346.84; all df = 1, all P < 0.001).

DISCUSSION

Bombus breviceps is widespread throughout Asia (Williams et al. 2009). According to the previous studies, B. breviceps is the main pollinator of Amomum subulatum (Zingiberaceae; Deka et al. 2011, 2014), and also plays the nectar-robber role in many species (Irwin 2010). The lack of fruit production in the bagging treatments indicated that T. alba needed a pollinator to transmit its pollinia. In over 2 years of observation, we found that nectar-foraging B. breviceps is the only pollinators of T. alba at the study site. But when the bracts of the flowers are removed, the visiting behavior of B. breviceps changes from entering through the labellum passage to forage for nectar to making holes in the spurs to rob the nectar. Therefore, the role of most B. breviceps changed from pollinator to nectar robber in the pollination process. From the structure of the flowers, the length of the bracts is significantly longer than the pedicels and spurs. Thus, the spurs are protected by large and curly bracts to defend against nectar robbers. The results also support this conclusion that the proportion of robbed flowers was significantly higher in the removed bracts treatment than the intact treatment. A low proportion of robbed flowers was also found in intact treatment, possibly because bumblebees have learning behavior (Leadbeater and Chittka 2008), after visiting the flowers of removing bracts and then visiting the intact flowers of smaller bracts. Due to resource limitation or intraspecific variation, T. alba may produce smaller bracts, which provides the possibility of nectar robbery.

Bracts have been proposed as highly effective displays of long-distance visual signals and as being associated with honest signaling of reward production (Borges et al. 2003; Pélabon et al. 2012), thus being considered able to attract pollinators. Our results showed that the visiting frequency of pollinators and single-flower longevity between intact and removed bract treatments showed no significant difference. Thus, the bracts removal had no obvious negative effects on floral structure or pollinator attraction. The bract was not able to attract pollinators in T. alba. However, removed bracts had a significant effect on the number of flowers the pollinators visited per inflorescence and the time required to visit each flower. Because robbery can improve the efficiency of nectar robbers foraging for nectar (Irwin 2010), and B. breviceps visits flowers with removed bracts almost entirely in the form of nectar robbery, they can visit more flowers in a shorter period of time.

Orchid pollen has evolved specialized pollinarium structures, which make it possible to estimate both female (pollen deposition or fruit) and male (pollen removal) pollination success (Nilsson et al. 1992; Ackerman et al. 1997). In the removed bracts treatment, pollinia removal, pollinia deposition, and fruit set of T. alba were significantly lower, suggesting that the effect of removed bracts on reproductive success is negative, whether for female or male components. The fruit set of intact and removed bract treatments was significantly lower than that of hand cross-pollination. Thus, the large bracts may enhance the female and male fitness of T. alba under environmental pressures of limited pollination.

Previous studies suggest that the photosynthesis of bracts has a very important contribution to seed maturation in Carpinus laxiflora (Betulaceae; Hori and Tsuge 1993), and cup-like bracts that fill with rainwater can help protect the flowers from seed herbivores in Pedicularis rex (Orobanchaceae; Sun and Huang 2015). In our study species, the bracts showed a light green in early flowering, but they are curly, so the area of the bracts for photosynthesis may be small. Moreover, the duration of the bract ranges from the bud stage to the fruit stage, which is nearly 2 months. How these traits affect the fruits and seeds may also require further investigation. In orchids, spur length and pollinator traits may coevolve (Darwin 1862; Wallace 1981), and spur elongation is advantageous to improve reproduction success (Nilsson 1988). However, the elongation of the spur undoubtedly increases the opportunity for nectar robbery by nectar robbers. Our study shows that the large bracts can protect against nectar robbers and thus reduce the possibility of nectar robbing that does not provide pollination services. So, the evolutionary relationship between bracts and spurs in orchids will be a very interesting research direction.

To our knowledge, conspicuously large bracts can be observed in all six species belonging to the Thunia genus. Analyzing the evolution of large bracts and their significance in pollinator interactions through phylogenetic comparative studies across the genus may offer valuable insights. Moreover, the presence of morphological similarities in other orchid species raises the intriguing prospect that they might also serve a similar protective function. Future research could explore whether other orchid species exhibiting similar bract morphology employ them as a defense against nectar robbers.

Acknowledgements

We thank Sheng Zhang for providing valuable advice on this study, and Rengasamy Anbazhakan for reviewing the language of the manuscript.

Authors' Contributions

J.-Y.G. designed the experiment and wrote the manuscript; S.-M.W. conducted field surveys, collected and analyzed the data, and wrote the manuscript. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the Joint Special Project on Construction of “First-Class Universities and Disciplines” of Yunnan University (202201BF070001-017). Shi-Mao Wu is mainly engaged in the reproduction and conservation of orchids.

Conflict of interest statement.

The authors declare that they have no conflict of interest.

REFERENCES