Flowering Phenology and Mating System of Calanthe sieboldii

Abstract

The pollination characteristics and flowering phenology of Calanthe sieboldii were evaluated to elucidate its reproductive characteristics and breeding systems. Field observations and artificial pollination experiments were conducted to study the pollination biology in Xuancheng City, Anhui Province. Meanwhile, gas chromatography–mass spectrometry (GC-MS) was employed to analyze the species’ volatile organic compounds (VOCs). Key findings include the following: (1) the flowering period extends from mid-April to mid-May, with a population-level flowering duration of 29 days in 2017, individual plant flowering averaging 20.22 days, and single flower longevity ranging from 12 to 23 days (mean = 19.30 days); (2) the species exhibits deceptive nectar guides devoid of nectar, indicating food-deceptive pollination, with Bombus sp. identified as its primary pollinator; (3) the pollinial–ovule ratio and hybridization index suggest a high level of self-compatibility without autonomous self-pollination, with no significant difference in pollination success between self- and outcross populations; (4) GC-MS analysis identified methyl benzoate and acacia-related compounds as the primary VOCs of C. sieboldii. These findings provide valuable insights into the conservation and sustainable management of orchids, particularly C. sieboldii.

1. Introduction

As global climate change and habitat loss intensify, orchid species like Calanthe sieboldii, reliant on specific pollinators and reproductive strategies, face endangerment. Despite its uniqueness, the pollination biology, breeding strategies, and environmental responses are poorly understood. Clarifying these aspects is crucial for knowing its ecological adaptations, population dynamics, and guiding conservation efforts. In the case of most angiosperms, intricate interactions occurred between flowers and their pollinators, and pollination is a pivotal process during their life cycle. Over long evolutionary processes, these pressures have led to the development of specialized floral structures designed to attract pollinators [1]. These structures enable plants to meet the requirements of diverse pollinators and establish specialized pollination systems, thereby enhancing pollination success rates and minimizing pollen loss [2]. Furthermore, flowering plants have evolved several flower characteristics that are attractive to pollinators and utilize various pollinator resources, both of which are crucial for ensuring the functions of ecosystems that facilitate pollination. As one of the most diverse families in the plant kingdom [3], orchids have earned the interest of numerous researchers owing to their varied life forms, intricate structures, and complex interactions with insects [4]. Approximately one-third of orchid species utilized deceptive pollination strategies, such as mimicking food sources or sexual deception, without providing any reward to the pollinators [5,6,7,8]. In addition to nectar and pollen [9], some orchids produced substances such as oils and fragrances that attract pollinators [10]. Despite the availability of diverse plant resources, deceptive pollination is highly prevalent among orchids [5]. Moreover, certain orchids may employ a food-deceptive pollination mechanism by imitating rewarding plants, thereby tricking pollinators into visiting them without offering a reward [11,12].

C. sieboldii is a perennial herbaceous evergreen plant in the orchid family belonging to the genus Calanthe. The species commonly grows in mountain forests at elevations ranging from 400 to 1500 m and is distributed across East Asia, including China, the Republic of Korea, and Japan [13]. Studies had indicated that Calanthe and its related taxa exhibited a high degree of self-compatibility, as observed in species such as C. lyroglossa and C. mannii [14]. The morphology of the labellum is a critical distinguishing feature among species within the genus, and diversity in flower morphology increases the variety of pollination systems observed. Within the genus Calanthe, certain plants with spurs were pollinated by lepidopterans, where the pollen adheres to the proboscis of insects. Successful sexual reproduction depends entirely on specific pollinators. The study conducted on Tsushima Island, Kyushu, Japan, investigated the pollination biology of C. sieboldii, revealed that the plant employs a food-deception pollination mechanism, with Xylocopa appendiculata being specifically identified as its pollinator [15]. In China, only three species of Xylocopa, namely, X. chinensis, X. rufipes, and X. appendiculata, constitute effective plant pollinators. Consequently, the presence of these pollinator species limits the maintenance and expansion of C. sieboldii populations [16]. The current study endeavored to undertake a comprehensive evaluation of the pollination biology and flowering phenology of C. sieboldii, utilizing rigorous field observations and experimental analyses.

2. Materials and Methods

2.1. Study Site and Plant Material

Calanthe sieboldii was studied in a population located in Xuancheng (30°29′ N, 118°35′ E, 335–503 m), in southeastern Anhui Province, China. The climate of the study area is characterized by a subtropical humid monsoon climate, with the main associated plant species including Pteroceltis tatarinowii, Camellia sinensis, and C. graciliflora (Table S1). To detail the pollination biology of C. sieboldii, long-term observations were conducted at fixed points. These observations documented the annual growth characteristics, renewal, flowering, fruit morphology, and growth status of C. sieboldii from 2015 to 2017.

2.2. Flowering Phenology and Floral Characteristics

One hundred and forty strains of C. sieboldii at the bud stage were randomly selected for flowering phenology monitoring in the spring of 2016 and 2017 [17]. Thirty blooming plants were randomly selected for observing the morphology of their inflorescences, and the diameter and the measurements of the petals, sepals, labella, and spurs were recorded using a Vernier caliper. The average values for these measurements were then calculated for the spring flowering period. A special channel is formed in Calanthe by the column wings adapting to the base of the labellum, which is crucial for effective insect pollination. Consequently, the width of the middle lobe was measured as the channel width, and the vertical distance from the therostellum to the labellum was measured as the channel height. Furthermore, 10 randomly selected, fully bloomed, and withered single flowers of C. sieboldii were subjected to color analysis using the Royal Horticultural Society Colour Chart (RHSCC), under the previously detailed controlled conditions, with an indirect light source. A portable colorimeter CS200 (Color Spectrum, Hangzhou, China) was used to measure the color parameters of different flower structures following black and white calibration, with laboratory values recorded for later creation of a color space map. At the peak blooming hours of 8:00, 10:00, and 12:00, ten newly opened flowers were randomly dissected and observed under an SZ810 continuous zoom stereo microscope (OPTEC, Chongqing, China) to determine the presence of viscous liquid within their spurs.

For the detection of volatile organic compounds (VOCs), the following procedure was adopted: Flowers at the early flowering stage were selected, and fragrance collection was conducted from 13:00 to 16:00 on a sunny day. Each sampling session lasted three hours, with three replicates and a control included. Prior to sampling, the air sampler was activated, and pumping was initiated until the majority of the air within the collection bags had been evacuated (Figure 1). After completing the cyclic sampling process, the adsorption tubes were collected and eluted into 2 mL Agilent sample vials using 0.5 mL of chromatography-grade dichloromethane. Then the fragrance components were analyzed via gas chromatography–mass spectrometry (GC-MS) (Agilent Technologies, Santa Clara CA, USA). The chromatographic conditions employed an HP-5 column (30 m × 250 μm × 0.25 μm) with high-purity helium as the carrier gas. The temperature program was set to start at 30 °C and held for three minutes then increased at a rate of 5 °C/min to 120 °C, followed by a ramp-up rate of 10 °C/min to 250 °C, which was maintained for 7 min. For mass spectral analysis, an EI ionization source was used with an ionization voltage of 70 eV and an ion source temperature of 270 °C. Data acquisition was performed in the full-scan mode (SCAN).

In this study, the internal standard method incorporating peak area and concentration standard curves was used to perform a quantitative analysis. After obtaining the total ion chromatogram (TIC), compounds were identified using the NIST standard reference database (NIST, Gaithersburg, MD, USA), with a match degree of over 80% considered significant. The composition of the measured gas was determined by comparing the retention times and characteristic ions of the samples with those of known standards in a mass spectrometry database. All ion peaks were identified after subtracting the interference background, and the experiment was repeated thrice to ensure consistency. To further determine the specific locations of fragrance glands on flowers, five fresh flowers were soaked in a 0.1% neutral red solution for staining periods of 12 and 24 h. Subsequently, the color changes of the flowers before and after dyeing were recorded.

2.3. Behavioral Observation and Species Identification of Pollinators

From 18 April 2017 to 1 May 2017, under rain-free conditions, plants in three randomly selected plots were continuously observed from 8:00 to 16:00 for a total of 112 h. The behaviors of various flower-visiting insects were recorded using an HDR-PJ90E Sony HD camcorder (Sony, Tokyo, Japan) and a Nikon D810 camera. The observation range was defined within a 1 m radius from the plant population outward. The recorded flower-visiting behaviors included passage (within 1 m of the flower), contact, staying, and entering. The observation content included Beijing time, insect flower-visiting behavior, landing method, visiting time, and the number of visits. Additionally, changes in pollen quantity and repeated flower-visiting phenomena were documented. Ten inflorescences were selected for bagging during the day and unbagging at night, repeated for five days. The pollen quantity of the plants was observed each morning, along with daily weather conditions, to investigate the flower-visiting insects and their visitation patterns. The species and numbers of insects were recorded, and specimens were captured as voucher samples. Three pollinators were randomly selected for morphological measurement (n = 3). The insect species were identified with the assistance of experts. Combining data from previous studies on flower morphological characteristics with the morphological traits of effective pollinators, the comprehensive pollination characteristics of C. sieboldii and the relationship with its pollinating insects were identified.

2.4. Pollen Viability and Stigma Receptivity

(1)

Influence of Different Culture Media on In Vitro Pollen Culture

Pollen viability was determined using an in vitro pollen culture method. The culture medium was prepared according to the experimental design, autoclaved, and placed in a 60 °C incubator for future use. A flower was selected, and the anther cap was removed using a dissection needle to extract the pollen block. The pollen block was repeatedly dispersed by alternating between mashing with a dissection needle, grinding with a glass slide, 40% ultrasonic treatment, and centrifugation at 80 rpm. After the pollen was dispersed, the volume was adjusted to 2 mL to prepare a pollen suspension. A pipette was used to transfer 1 mL of culture medium into the wells of a 3 × 4 12-well cell culture plate. After cooling and solidifying, 0.1 mL of the pollen suspension was added. The plate was covered and sealed in a self-sealing bag then placed in a constant temperature incubator at 25 °C for a period of time. Pollen germination was observed and counted under a B203 LED optical microscope (OPTEC, Chongqing, China).

The methodology for pollen germination rate involves calculating the germination percentage by determining the ratio of the number of pollen grains that have successfully germinated within a defined field of view to the total number of pollen grains present in that same field. Pollen viability is the ratio of the quantity of germinated pollen to the total number of pollen grains. The length of the pollen tubes was measured using ImageJ 1.49. A minimum of 10 pollen tubes from germinated pollen grains were selected, and the average length was computed.

For the high-concentration medium, single-factor designs were utilized, labeled G-1 to G-8 (Table S2). Both orthogonal and single-factor designs were employed to develop low-concentration media. Three factors—sucrose, boric acid, and calcium nitrate—were tested at three different levels to identify the optimal combination of culture medium components that would yield the highest pollen germination. The sucrose concentrations in the high-concentration and low-concentration culture media were set at 0–250 g/L and 10–30 g/L, respectively. The concentrations of boric acid and calcium nitrate were kept relatively low. In the low-concentration series of culture media, experimental media D-10, D-11, and D-12 were included as part of the single-factor design (Tables S3 and S4).

(2)

Stigma Receptivity Test

A solution of benzidine hydrochloride and hydrogen peroxide (1% benzidine: 3% hydrogen peroxide: water = 4:11:22, by volume) was used to assess stigma receptivity. The 1% benzidine solution was prepared by dissolving benzidine in pure ethanol, followed by the gradual addition of water; a 1% benzidine sulfate solution was also prepared for future use. Stigmas were collected from flowers that had been open for 1 day, 5 days, and at the end of flowering. The stigmas were placed in the wells of a 3 × 4 12-well cell culture plate, and the reaction solution was added until the stigmas were fully submerged. After 3 min, the presence of bubbles around the stigmas was observed using an SZ760 stereo microscope (OPTEC, Chongqing, China). Strong receptivity was indicated by the production of a large number of bubbles around the stigma and blue coloration of the stigma.

2.5. Hybridization Index (OCI) and Pollen–Ovule Ratio (P/O)

The outcrossing index (OCI) was calculated, encompassing an evaluation of three key parameters: (1) plant flower diameter less than 1 mm, denoted as 0; 1–2 mm, denoted as 1; 2–6 mm, denoted as 2; and if it is greater than 6 mm, denoted as 3; (2) according to the time interval between stamen maturity and stigma availability, both male and female or pistil are first memorized as 0, with stamen first memorized as 1; (3) according to the spatial position of the stigma and the anther, the same height of the two is denoted as 0, and spatial separation is denoted as 1. Finally, when OCI is 0, it corresponds to cleistogamy; when OCI is 1, obligate autogamy; when OCI is 2, facultative autogamy; when OCI is 3, self-compatible and pollinators are needed; and when OCI is 4, partial self-compatibility, outbreeding, and pollinators are required [17].

To obtain the P/O value by dividing the average pollen quantity per single flower by the number of ovules per single flower. For the total amount of pollen of a single flower, 10 open 1-day pollen blocks were randomly selected at the full flowering stage. The pollen was transferred to a 2 mL centrifuge tube, crushed with a dissecting needle, and dispersed by an ultrasonic cleaning instrument. After that, the pollen was diluted to 1 mL in distilled water, mixed well, and then 10 μL of the pollen liquid was placed on a slide under a microscope. The total pollen count from 10 visual fields was calculated using the software ImageJ and Photoshop CS 4.0, and the mean value was calculated from 5 measurements. The total pollen of a single flower was 100 × n. Statistics of seed number and embryo rate of a single fruit: 10 ripe capsules were randomly selected at the fruit maturity stage. The capsules were cut open with surgical blades, all the seeds were scraped out, and the seeds were fully shaken and mixed into a seed suspension. According to the method of pollen statistics, the number of seed embryos from 10 seeds was observed and calculated under the microscope. The count was repeated 5 times to calculate the average value and the total number of seed embryos [18]. The relationship between the P/O ratio and breeding systems was delineated as follows: cleistogamy was noted at a P/O range of 2.7 to 5.4, obligatory selfing occurred within the range of 18.1 to 39.0, and facultative selfing was identified when the P/O ranged from 31.9 to 396.0. Additionally, facultative outcrossing was observed at a P/O interval of 244.7 to 2588.0, while obligate outcrossing manifested within the range of 2108.0 to 195,525.0 [19].

2.6. Breeding System Inspection

Selected plants were bagged (using transparent pearl gauze bags measuring 20 × 30 cm) before flowering. Three to four flowers per plant were treated as follows: (1) no treatment to check for spontaneous self-pollination; (2) emasculation to test for agamospermy; (3) artificial self-pollination, including within the same flower and the same clone (pollination within the same flower on the same plant and between different flowers on the same plant); and (4) artificial cross-pollination using pollen from plants more than 10 m away. The fruit set rate was determined by the following formula: (Number of fruits set/Total number of pollinations) × 100%. This metric provides a quantitative assessment of the proportion of successful pollinations that result in fruit formation. This rate was examined and statistically analyzed for each treatment after the flowering period in autumn. Additionally, control flowers within the population were left untreated, and the natural fruit set rate was recorded after the flowering period.

3. Results

3.1. Flowering Phenology and Single Flowering Stage Dynamics

The flowering period of the C. sieboldii population occurred between April and May. In 2017, the first flowers began blooming on 15 April. The peak flowering stage was reached on April 18, with approximately 60% of the plants in bloom. Subsequently, on 10 May, the population entered the final stage of flowering, with approximately 7.87% of the plants continuously blooming, and all flowers had withered by 13 May. The duration of the flowering period of the C. sieboldii population was 29 d (Figure S1). At the individual level, the flowering period ranged from 12 to 27 d, with an average duration of 20.22 ± 0.24 d (n = 140). For individual flowers, the flowering period ranged from 12 to 23 d, averaging 19.30 ± 2.61 d (n = 30).

The inflorescent bud and three leaves were enveloped by two foliaceous bracts and initially underwent an elongation phase. When the inflorescence buds reached approximately 15 cm, the leaves spread, and the inflorescence emerged from the middle of the upper leaves of the pseudobulb and elongated. The leaves were completely unfolded to a length of approximately 20 cm and turned dark green. The peduncle grew to a length ranging from 35 to 50 cm. The flowers opened from the lower part of the inflorescence, with a rapid opening rate initially and slowed down when 1–2 flowers remained, and the size of the last 1–2 flowers was reduced (Figures S2 and S3).

3.2. Floral Characteristics

The scape measured 40 to 60 cm in length and bore 1–20 slightly fleshy, hairless flowers that were dispersed and relatively evenly arranged. The flowers located in the lower part of the plant were more pendulous than those in the upper part. The florets were symmetrical, with individual flowers measuring approximately 6 cm in diameter. The middle sepal was oblong, whereas the lateral sepals were slightly narrower and oblong. The petals were subelliptical with an acuminate base and featured a distinct mark akin to that of the midvein. The apices of the sepals and petals were often rolled backward. The labellum was trilobed and displayed a callus at the base, with large and flat lateral lobes and smooth surfaces. The central lobes were pubescent with three keel-like structures, and an inverted “Y”-shaped and pubescent callose appendage was present at the base (Figure 2). The spur was approximately 9 mm in length, and the short stigma was connected to the labellum to form an entrance to the spur. The characteristics of C. sieboldii were clearly observed using line mapping (Table S5). Post-pollination occasionally caused serious morphological changes. The peduncle curved downward, and the ovary gradually expanded after pollination. The perianth began to fade, shrink, and droop; concurrently, brick-red spots emerged. These changes effectively prevented further visits by the pollinators. Inside the successfully pollinated flowers, pollen grains swelled and became translucent once in contact with the stigma. Subsequently, the pollen tube initiated germination, leading to the completion of fertilization. By contrast, flowers that were not pollinated fell after flowering.

The coloration of various floral components is described as follows: The petals exhibit a vivid yellow hue, corresponding to shades 9–13 on the standardized color chart. The sepals are characterized by a yellow coloration, falling within the range designated as 7-A. The labellum, the distinct petal-like appendage, displays a yellow-tinged white tone, classified as 7-B, while its pleated regions exhibit an orange-red coloration, identified as 164A on the color chart. Fading flowers featured brick-red spots. Based on the laboratory measurements of the flowers, the predominant color of the flowers was yellow (Figure 3). The floret was dissected randomly, and it was found that there was a brick-red honey guide at the stigma; its spur was 9 mm, and it had dry, light white hair within the spur with no nectar secretion, belonging to the false honey guide. Fragrances such as those of orange flowers were obviously observable during 10:00 and 14:00 on sunny days. A total of 32 organic compounds were successfully isolated and identified (Table S6), including two alcohols, four terpenoids, one aldehyde, one acid, fifteen alkanes, four alkenes, and two esters. Notably, benzyl methyl ether, acenaphthylene, 6,9-heptadiene, β-elemene, caryophyllene, and α-pinene were identified as the predominant active components of the neroli flower’s volatile fraction. After further neutral red staining, it was confirmed that the main stained part on the flower was the stigma (Figure 4B), followed by the part near the spur on the pleat.

3.3. Species, Frequency, and Behavior of Flower-Visiting Insects

C. sieboldii employed food-deceptive pollination mechanisms. During the observation period, the insect needed to meet the following three conditions to be identified as a pollinator of the plant. The first is the ability to carry away pollen. The second is the ability to carry the pollen to other plants. Finally, the morphological characteristics of the plant and its pollinators were compared, which allowed us to identify Bombus sp. as the pollinators of the plant (Table S8).

The flower-visiting behavior of bumblebees includes passing, staying, and entering. On a sunny day, when the temperature is right, Bombus sp. spread out (Figure S4A). Bombus sp. fly close to the flower (Figure S4B), whirr their wings from a distance, and then fly to the flower of the inflorescence. Its front and middle feet grasp the flap of the labellum and it pushed its beak into the distance of the orchid so that the head of Bombus sp. rests on the top of the stamen. A sticky disk pulls pollen (occasionally with a cap) tightly to the head of Bombus sp. (Figure S4C). Bombus sp. carries pollen chunks on top of its head and flies to the florets next to it. This species then transfers one to six pollen chunks or empty caps to the sticky stigma. Other small bees visit the flower, but they cannot carry the pollen because of the size mismatch (Figure S4D).

During the observation period (between 08:00 and 16:00), Bombus sp. entered the labellum of the plant 29 times and made contact 20 times. A maximum of nine pollination visits by Bombus sp. were recorded (Figure 5). The average widths of the entrance (the spur was approximately 9 mm in length, and the short stigma was connected to the labellum to form the entrance to the spur) and the head of Bombus sp. were 5.95 mm and 8.26 mm, respectively. The flower was marginally broader than the head, and the entrance height was comparable to the head height. The width of the entrance was slightly wider than that of the head of Bombus sp., the height of the entrance was similar than the head height of Bombus sp., and the difference between the flower’s spur-like structure and the distance of Bombus sp.’s peak was not significant (p = 0.000 < 0.001, p = 0.123 > 0.05, and p = 0.872 > 0.05, respectively) (Table S7). In this study, C. Sieboldii displayed no nectar, magnificent inflorescence, labellum, flower spur, or fragrance. Combined with all the floral characteristics and the capture of pollinating insects, we inferred that the pollinating insects of C. sieboldii were deceptive pollinators, and the pollinating insects were Bombus sp.

3.4. Pollen Viability n and Stigma Receptivity

Comparison of liquid and solid media: Liquid media are not ideal for the culture of C. sieboldii pollen as they lead to the aggregation and entanglement of pollen grains, making statistical analysis difficult. In contrast, solid media allow for uniform germination and growth of pollen at their respective positions, facilitating the analysis of germination data. In the high-concentration media series, it was observed that with sucrose concentrations of 0, 100, 200, and 250 g·L⁻¹, a concentration of 250 g·L⁻¹ severely inhibited pollen germination (Figure S5). High concentrations of sucrose suppress pollen germination.

Factors affecting pollen germination rate in C. sieboldii, ranked from most to least influential: Calcium nitrate, sucrose, and boric acid. The optimal pollen culture medium composition was 30 g·L⁻¹ sucrose + 60 mg·L⁻¹ H₃BO₃ + 60 mg·L⁻¹ Ca(NO₃)₂·4H₂O.

Factors affecting pollen tube growth length, ranked from most to least influential: Boric acid, sucrose, and calcium nitrate. Through correlation analysis, the Pearson correlation value was 0.600, indicating a significant positive correlation at the 0.01 level (two-tailed) between pollen germination rate and pollen tube length. The optimal pollen culture medium for promoting pollen tube germination was 30 g·L⁻¹ sucrose + 60 mg·L⁻¹ H₃BO₃ + 15 mg·L⁻¹ Ca(NO₃)₂·4H₂O. In the orthogonal combinations, Group 2 and Group 7 showed the best germination results, both of which closely matched the theoretically predicted optimal medium.

In both culture media formulations, the combination of 45 mg·L⁻¹ H₃BO₃, 15 mg·L⁻¹ Ca(NO₃)₂·4H₂O, and 25 mg·L⁻¹ sucrose performed best. Lower sucrose concentrations can promote the in vitro germination of Hibiscus tiliaceus pollen (Table S9).

In the low-concentration pollen culture medium series, the overall germination rate of pollen from flowers that opened on the first day was low after 24 h of cultivation. Data from the orthogonal design table were selected for analysis. After conducting an intergroup effect test, sucrose (p = 0.000 < 0.01) and calcium nitrate (p = 0.000 < 0.01) had a highly significant effect on the pollen germination rate of Hibiscus tiliaceus, while boric acid (p = 0.169) had no significant effect. Sucrose (p = 0.004 < 0.01), calcium nitrate (p = 0.004 < 0.01), and boric acid (p = 0.0007 < 0.01) all had a highly significant effect on pollen tube growth length.

Reaction solutions prepared with benzidine and benzidine sulfate both elicit a significant amount of bubbling from active stigmas. Fresh stigmas exhibit visibly glossy and crystalline mucus (Figure 6A). Following a three-minute reaction period, stigmas collected on both the first and fifth days display abundant bubbling (50 and 60 bubbles, respectively), indicating strong receptivity (Figure 6B,C). In contrast, stigmas harvested during the final stage of flowering produce fewer bubbles (only 10), suggesting weaker receptivity.

3.5. Pollen–Ovule Ratio and Hybrid Index

The OCI of C. sieboldii was calculated as four, indicating that the plant exhibits a combination of partial self-incompatibility and pollinator-mediated outcrossing. The pollen count per flower averaged 1,178,915 ± 94,462 (n = 10), while the average ovule count per flower was 50,408 ± 3832 (n = 10). The pollen/ovule (P/O) ratio ranged from 16.05 to 32.23, with a mean of 23.39 ± 1.87. The breeding system of this plant displays both obligate and facultative self-fertilization tendencies.

3.6. Breeding System

Considering potential factors that can lead to plant death and fruit rot, such as late-stage sterility, rain, high temperatures, and animal predation, the fruit set was recorded once a week after artificial pollination treatments and again in autumn. The results revealed that the rates of ovary enlargement in the spontaneous autogamy test (flowers bagged without emasculation) and agamospermy test (flowers emasculated and bagged) for C. sieboldii were 0% and 0.83%, respectively (n = 120), suggesting the absence of spontaneous autogamy and the possible occurrence of agamospermy. In contrast, the rates of ovary enlargement for artificial self-pollination (including within-flower and within-plant pollination) and artificial cross-pollination were both 100%, while that for natural cross-pollination was 1.64%. Notably, there was no significant difference in the ovary enlargement rates between the selfing and outcrossing treatments (Table S10). Furthermore, the fruit set rates for artificial pollination (39.02%, 82.05%, and 73.00%) were significantly higher than those for natural fruit set (5.45% and 4.60%), indicating a pollinator limitation in C. sieboldii.

4. Discussion

Flowering phenology, which encompasses parameters such as flower number, flowering duration, and onset timing, significantly influences the success of plant pollination [21,22]. For C. sieboldii, this phenology can be segmented into five distinct stages: budding, pre-anthesis, anthesis, full bloom and post-anthesis. The duration of the flowering period of C. sieboldii population was 29 days. The flowering intensity of this species, which is intimately tied to both plant characteristics and environmental factors [23,24,25], ranges from 50% to 70%, exceeding that of some tropical and subtropical plants, such as Plumbago auriculata Lam. and Eremosparton songoricum [26,27], which have been reported to exhibit flowering intensities of 30–40%. The concentrated flowering pattern of C. sieboldii facilitates the attraction of a diverse and abundant insect pollinator community. However, this characteristic also poses potential risks. Specifically, the heightened pollen transfer among individuals, both within and between populations, may inadvertently promote self-inbreeding, ultimately leading to a decline in population fitness [28].

Pollination syndromes have emerged as a powerful tool for elucidating the intricacies of floral morphology and its implications for pollination biology. These syndromes encompass traits such as symmetry, size, nectar production, spur presence, coloration, and fragrance, each of which plays a pivotal role in shaping pollinator–plant interactions [29]. Notably, symmetrical flowers, exemplified by the zygomorphic blooms of all orchid species, including C. sieboldii, tend to elicit higher visitation rates and foster pollinator specialization [30]. The symmetrical structure of the flowers of Calanthe triplicata, which belongs to the group of rewardless plants, attracts pollinators through deceptive pollination. Like most orchids, Calanthe triplicata has symmetrical floral structures, which are advantageous for attracting pollinators. Generally, nectar production is significantly related to flower size, and pollinators tend to visit larger flowers [31]. Larger flowers in Silene virginica and Brassavola nodosa show higher reproductive success, while in Tolumnia variegata, reproductive success is not related to flower size [32,33]. However, the relationship between spur length and pollination rates can be complex. For instance, in Platanthera bifolia, shorter spurs initially facilitated higher pollination rates, but these were later surpassed by flowers with longer spurs [34]. Specialized pollination systems in long-spurred orchids, which are often pollinated by long-tongued moths, restrict the range of potential pollinators and the natural mating rates [35]. Similar to most deceptive pollination mechanisms in orchids with nectar spurs, Calanthe triplicata has nectar spurs but no nectar reward. Bombus sp. upon visiting, extend their proboscis into the nectar spur to extract nectar, possibly completing the pollination process in the act.

Flower color also plays an important role in attracting and guiding pollinators. Specifically, Bombus sp. are known to favor yellow, blue, and white hues [36,37]. Moreover, flower color provides visual signals to increase pollination efficiency [38]. In this study, significant differences were observed in the flower colors of C. sieboldii between the early and late flowering stages, with a transition from yellow to brick-red. The attractiveness to Bombus sp. showed a strong attraction-weakening process, consistent with the aforementioned pattern. In addition, after pollination, the flower stalk bends, the petals wither, and the perianth closes and droops, causing pollinators not to revisit the flowers.

Flower fragrance is another important factor in attracting pollinators [39]. Flowers of the genus Ophrys release a scent that mimics the pheromones of female Campsoscolia sp., deceiving wild bees into pollinating them through sexual deception [40]. In this study, chemical analysis of the volatile components of C. sieboldii revealed the presence of terpenes, esters, aliphatic compounds, and aromatic compounds, including methyl benzoate, acacia, and 6,9-heptadiene, which were similar to the results found in another study [41]. By using 1:10,000 neutral red staining and a scanning electron microscope to observe the microstructure of C. yaoshanensis, it was determined that the scent of C. yaoshanensis was located in the sepals, stigma, and labellum folds of the petal tip [42]. Unlike C. yaoshanensis, the volatile emission sites of C. sieboldii in this study (as indicated by neutral red staining) may be primarily located in the stigma, followed by the part near the spur on the pleat. This hypothesis requires further optical microscope observation and analysis.

Previous research predicted and confirmed the unique pollination strategy employed by Paphiopedilum dianthum, which mimics a breeding site to deceive hoverflies, its pollinators [1]. Similarly, our investigation into floral traits led us to anticipate that plants of Calanthe would be pollinated by Lepidoptera and Hymenoptera. Upon examining C. sieboldii specifically, we discovered several floral characteristics that collectively suggested a mechanism of food deception: a labellum, a short spur devoid of nectar, a fragrance emitted during the day, and no signs of pollination at night. Based on these findings, we hypothesize that large hymenopterans are the probable pollinators of this species. Our findings suggest that C. sieboldii potentially mimics the local species P. illicioides to deceive pollinating insects. This hypothesis is supported by multiple lines of evidence. First, both plants exhibit bright yellow flowers, albeit with structural differences in their spurs (a true spur in the orchid versus a spur-like tubular structure in P. illicioides). Second, the primary components of their floral fragrances are strikingly similar, dominated by β-ocimene and caryophyllene [43,44]. Third, despite both being visited by bumblebees, only P. illicioides offers nectar rewards, suggesting a potential deception strategy by C. sieboldii. Lastly, the slightly earlier blooming period of C. sieboldii compared to P. illicioides may confer an advantage in terms of reproductive success, as early spring flowering and extended bloom durations have been shown to enhance plant fitness in species such as Calypso bulbosa and C. delavayi [45,46]. The combination of these factors—earlier and prolonged flowering, a compatible pollination syndrome featuring similar fragrances, and the absence of competing flowering species in their habitat—likely contributed to the success of the food-deceptive pollination strategy in C. sieboldii. However, it is noteworthy that self-fertilization was also observed in this study, indicating potential redundancy or adaptability in the reproductive strategies employed by this orchid species. This study underscores the complexity of pollination systems and highlights the need for further investigation into the ecological and evolutionary drivers of such flexibility.

The mating system of a species offers valuable insights into the patterns and frequencies of reproductive interactions within a population [47]. While self-fertilization can be advantageous for populations facing limited or scarce pollinator availability, it can also lead to inbreeding depression, which reduces offspring fitness and hinders outcrossing rates, ultimately resulting in the inefficient utilization of reproductive resources. Conversely, outcrossing promotes genetic diversity and enhances the adaptive potential of offspring, conferring long-term fitness benefits [48,49]. Members of Calanthe often exhibit a mixed mating system, characterized by a relatively high fruit set (a minimum of 86.7%) achieved through artificial self- and cross-pollination experiments [50,51,52]. However, under natural conditions, fruit set rates are significantly lower (a maximum of 6.5%), suggesting a limitation in pollination efficiency. Similarly, C. sieboldii also displays a mixed mating system, with artificial pollination experiments revealing no significant differences in fruit set rates but consistently higher rates compared to natural pollination under identical conditions. This finding is in line with observations in other orchids that employ deceptive pollination mechanisms, where pollinator constraints contribute to relatively low pollination efficiency in C. sieboldii.

Climate consistently exerts profound effects on plant reproduction by modulating flowering phenology and pollinating activity [28,53,54,55,56]. Specifically, lower temperatures and reduced sunlight can initially delay flowering onset and prolong the flowering period, whereas high humidity and poor airflow can exacerbate flower and fruit abscission rates. Furthermore, rainy weather and low temperatures diminish insect activity, thereby delaying pollination and potentially impacting reproductive success. In the present study, C. sieboldii growing in shady forests with acceptable airflow displayed a relatively long flowering period but exhibited scattered blooms. In contrast, plants located on roadsides, forest edges, or sunny slopes, which were exposed to longer periods of direct sunlight, flowered earlier and more synchronously but had accelerated life cycles. These findings suggest that varying light conditions can significantly influence flowering patterns and potentially reproductive output in C. sieboldii. Climate conditions also significantly influence pollinating insects, primarily by determining the temporal patterns of insect visits to flowers (Table S11). In turn, this alters the timing of pollen dispersal, with far-reaching consequences for plant reproductive success. For instance, in the case of Cephalanthera rubra, its primary pollinators are solitary bees that engage in deceptive pollination [57]. Warm treatments accelerated the developmental rate of these solitary bees but yielded smaller adults with reduced flight capabilities [58]. This size reduction was accompanied by a shortened flight distance, which constrained pollen flow distances and potentially intensified the disadvantages of self-fertilization. These findings underscore the intricate interplay between climate, pollination ecology, and plant reproduction, emphasizing the need for further research to better understand the mechanisms underlying these complex interactions.

5. Conclusions

In conclusion, our study provides a comprehensive insight into the reproductive biology of the studied Calanthe, elucidating key aspects of its flowering phenology, floral characteristics, floral chemistry, and pollination mechanisms. The unique flowering patterns and extended flower longevity documented in this research suggest sophisticated temporal strategies employed by the plant to optimize pollinator attraction and facilitate subsequent reproductive success. The chemical analysis of floral volatiles has revealed a suite of compounds that likely play a pivotal role in mediating the intricate interactions between pollinators and the orchid, thereby emphasizing the significance of chemical signaling in reproductive ecology. Furthermore, our pollination experiments and observations have challenged conventional understandings of plant reproductive strategies by revealing an unexpected combination of high self-compatibility and outcrossing potential.