Reproductive Biological Characteristics of Dendrobium Species

Dendrobium, as the second largest genera of Orchidaceae, is wildly distributed all over the world and has higher ornamental and medicinal values. Currently, research about the reproductive biological characteristics of the Dendrobium species are still limited. Therefore, only vegetative propagation (tissue culture) and sexual reproduction (flowering character, visitors, pollination and seed set) are reviewed in this article. Finally, the subjects of further research including flower morphology, pseudopollen, anther cap and lower fruiting rate are put forward.

Introduction

Dendrobium, which has about 1000–1400 original species with higher ornamental and medicinal values (Kuehnle 2006; Wang et al. 2007; He et al. 2008), is the second largest genera after Bulbophyllum of Orchidaceae. It is extensively distributed between Asia north of the equator and the Oceania regions (Ji et al. 1999; Wang et al. 2007). Seed and vegetative propagation are the major modes of reproduction of the Dendrobium species under natural conditions. However, Dendrobium species often have lower seed setting, and mycorrhizal symbiosis limits seed germination in seed propagation, resulting in the poor ability of natural reproduction (He et al. 2008; Li et al. 2009b). Dendrobium orchids are popular flowering potted and cut flowers around the world due to their flowering floriferousness, wide range in flower color, size, and shape, year-round availability, and lengthy postharvest life (Kuehnle 2006). The earliest record about Dendrobium species can be traced back to “Shen Nong’s Classic of Materia Medica” (also named as Shennong’s Herbal), a classic work on plants and their utilization, written in the Han Dynasty (206 B.C.–220 A.D.) in China (Wang et al. 2007). Previous studies have focused on the many chemical ingredients, such as alkaloid, polysaccharide, sesquiterpenoids, phenanthrenes and bibenzyls in the Dendrobium species (Zhang et al. 2003; Ma and Wang 2008; Li et al. 2010), which could be used to effectively treat eosinophilic gastroenteritis, cataract, arthritis, thromboangiitis obliterans and chronic pharyngitis (Wei 2005). Chemical ingredients contained in Dendrobium species are effective in promoting digestion, extending life span, regulating the immune system, dilating blood vessels, lowering or normalizing blood pressure, and treating against tumor (Cheng and Guo 2001; Zhang 2003; Ma and Wang 2008; Zou and Liu 2010; Ng et al. 2012). Although lots of knowledge has been displayed in tissue culture and rapid propagation of Dendrobium species, reproductive biological characteristics of the plant are still limited. In this present chapter, reproductive biological characteristics of Dendrobium species including vegetative propagation and sexual reproduction are reviewed. Besides, a perspective on unknown reproductive biological characteristics of Dendrobium species will be put forward at the end of this chapter.

Vegetative Propagation

Characteristics of Dendrobium Species

Dendrobium species is either epiphytic or lithophytic plant without or with few lateral branches in cylindrical or flat-triangular prism stems (Ji et al. 1999). The base of the pseudobulb has ability for tillering, which is determined by environmental conditions, ages of tuft, plant nutritional status and other factors. Generally, tiller growth increases with age of tuft when nutritional status and other conditions are not limited (He et al. 2009). In Dendrobium plants, pseudobulbs of different ages possess diverse functions varying among species (Table 1) (Tang et al. 2007; He et al. 2009). Dendrobium officinale has different characteristics of stems as shown in Table 2, compared with D. Nobile. There are two reserve buds on the base of pseudobulb in D. Nobile, from which the current buds will sprout. Usually, a pseudobulb and a bud together form tufted stems for growth and development. Buds sprout from the beginning of mid-February each year, and their detailed growth characteristics are shown in Table 3.

Propagation Modes

After reaching the age limitation, high buds and aerial roots often grow in nodes of the Dendrobium species, which produce a scaly bud on the base of stems. Therefore, the vegetative propagation of Dendrobium occurs primarily through high bud breeding, cuttings, layering, division propagations and tissue culture, in which high bud breeding and division propagation are more commonly used (Wang et al. 2007; He et al. 2008; Li and Miao 2009). 

Cutting Propagation

It has low cost and high yield, and can preserve the advantageous traits of the female parent. However, the buds have a different nutritional status, which easily lead to diverse seedling growth. Cuttings can be divided into different modes, such as one-year-old stem segment cuttings, rootless high bud cuttings and new bud cutting, according to the different shoot use for cuttings; Late mid-February to late March is the best time for cutting propagation (Wang et al. 2007).

Layering Propagation

Pseudobulbs are cut away from the female parent and laid onto the water grass in a 2–3 cm depth rectangular plastic box. The stems can be cut off after germination and rooting. Each stem can propagate 4–5 seedlings (Wang et al. 2007).

High Bud Breeding

It also can keep the advantageous traits of the female parent and its operation is very simple. However, very few numbers of high buds restrain propagation (Wang et al. 2007). Under field conditions, D. loddigesii produce new seedlings by means of high bud breeding (He et al. 2009). D. officinale propagate mainly by lateral buds, thus leading to a lower reproductive rate (Jin 2009).

Division Propagation

It also can keep the advantageous traits of the female parent and seedlings can be in blooming in current year. This method is very simple and survival rate is very high. But it is difficulty in large-scale cultivation due to the lower propagation and the easily damaged roots (Wang et al. 2007).

Tissue Culture

Tissue culture can also keep the advantageous traits of the female parent and grow regularly and fast. However, it needs higher technical skills, cost and longer time. The common culture mediums include MS, B5, N6, 1/2MS, KC (Knudsoc), W (White) and VW (Vacin and Went). NAA and BA are commonly used as the hormones of culture medium. Moreover, addendums such as potato, banana juice and casein are often used in culture medium (Wang et al. 2007). At present, explants of tissue culture contain seeds, roots, pseudobulb segments, leaves, protocorm and artificial seeds (Song et al. 2004; Wang et al. 2006; Li et al. 2010).

As mentioned before, seed germinations of Dendrobium species are difficult under natural conditions due to lack of mycosymbiosis, whereas in a synthetic medium, comprising of inorganic salts, sugars and agar without mycorrhizae it germinates successfully (Wang et al. 2007). Callus induced by seed and protocorm cultured by embryo have stronger differentiation potential and higher quality, respectively (Li et al. 2010). Seed culture is generally carried out before capsule dehiscing because of the easily separated powdered-seed, sowing and sterilization (only for capsule epidermis) at this period. The capsule after removal of persistent perianth and carpopodium is scrubbed with 70% alcohol for a minute or two, soaked in supernatant of 10% calcium hypochlorite or 2% chlorinated acetic acid for 20 min, and then washed 3 times with aquae sterilisata. Afterwards, we can use four methods for inoculation in a super clean bench: (1) cutting the peel and sowing it in culture medium surface; (2) sowing with inoculating needle after the capsule placed in 0.1% mercuric chloride for disinfection for 3 min; (3) seeds are transferred into 3% to 6% hydrogen peroxide, shaken for 20–30 min, sowed with a sterile pipette, and then evenly distributing them in the medium; (4) cutting at 1 cm of both the front and the back ends of the capsule, taking out the seed block, immersing in sterile water, shaking and then sowing seeds as droplets (Wang et al. 2007).

In order to improve the viability and production, stem sections with strong reproductive capacity should be selected for Dendrobium cultivation (Li and Xiao 1995). The stem sections as appropriate explants are washed with tap water and detergent on the super clean bench, scrubbed and sterilized for 15 s with 70% alcohol, put them into 0.1% mercuric chloride (including 2 drops of Tween-60), and then sterilized for 20 min. Afterwards, they are rinsed with sterile water for 5 s, gently shaken and washed 6 times, then placed in the corresponding medium (Wang et al. 2007). Cultivation conditions, such as hormones, pH value of cultivation medium and addendum, play different roles in Dendrobium cultivation. Coconut milk can promote the differentiation of buds as well as proliferation of shoots and protocorm. Moreover, banana juice can improve rooting (Jiang et al. 2005; Liao et al. 2006). Decreased basic element contents of cultivation medium are beneficial for the growth of protocorm, but do not benefit protocorm differentiation. In addition, NAA, IAA, IBA and GA3 can induce the growth and differentiation of protocorm, whereas acidic cultivation medium act against morphological variation of protocorm (Hou and Guo 2005). Cultivation mediums of protocorm, shoots and rooting added 1 mg/L AgNO3 can improve the proliferation of protocorm, the differentiation of buds, and the growth of seedlings, and thus increase the survival rate of transplant (Li et al. 2007).

In the process of tissue culture, the selections of the types of medium, addendum, hormones and culture conditions mainly are based on Dendrobium species and the stage of culture. Cultivation condition for spring Dendrobium: (1) seed culture: N6 culture medium (Wang et al. 2007; Zhou and Liu 2010); (2) stem induction: medium contains MS + 2.0 mg/L 6BA + 0.1 mg/L NAA, 3.0% sucrose and 0.21% crystal agar, temperature (24 ± 2°C), light 14 h/d, 500–2000 lux; (3) root induction: medium is the same as above except 0.1 mg/L NAA is substituted 0.05 mg/L for 0.1mg/L NAA (Zou and Liu 2010). Cultivation condition for autumn Dendrobium: (1) shoot tip or vegetative buds culture: medium contains MS + 5 mg/L BA + 1 g/L activated carbon (pH 5.2–5.5), light 8–12 h/d, 1500–2000 lux, at 26°C; (2) vegetative buds or protocorm subculture: medium contains MS + 1 mg/L BA + 0.5 g/L activated carbon, light 8–12 h/d, 1500–2000 lux, at 26 ± 2°C; (3) monoclonal cultures: medium contains 1/2 MS + 0.05 mg/L BA + 0.5% activated carbon + 0.5 mg/L NAA + 5% banana juice (pH 5.2–5.5), light 12 h/d, 4000–5000 lux, at 26 ± 2°C (Zou and Liu 2010). Cultivation condition for D. Candidum: (1) protocorm proliferation: medium contains 1/2 MS + 0.2 mg/L KT + 1.0 mg/L NAA + 0.5% activated carbon; (2) protocorm differentiation: medium contains 1/2 MS + 0.5 mg/L NAA + 0.5% activated carbon; (3) rooting: medium contains 1/2 MS+ 3.0 mg/L NAA + 0.5% activated carbon (Jiang et al. 2003); (4) test-tube seedling: medium contains B5 or 1/2 MS, 10% banana extract, 2 mg/L NAA and 2% sucrose (Liu and Zhang 1998). Cultivation condition for D. Nobile: the basic conditions of seed germination and seedling development is light intensity of 1000–1500 lux and photoperiod of 10–12 h at 18–27°C (Song et al. 2004). Furthermore, the type and concentration of disinfectant and disinfection time also have effects on non-symbiotic germination of Dendrobium seed (Song et al. 2004).

Sexual Reproduction

Flowering Character and Phenology

Dendrobium species has one or more flowers with normal structure (3 sepals, 3 petals and 1 gynostemium) at the top of inflorescence axes. Compared to the sepal and petal, the labellum structure is more complex and obviously differs in different species. For example, D. fimbriatum: sunken labellum, edged with fringed beard, covering with pubescence; D. loddigesii: labellum with pubescence, outward circle is white, edge destined hairy red tassels; D. hercoglossum: labellum epidermis with tall and slender paper structure; D. jiajiangense: labellum with pseudopollen (He 2008; Wang et al. 2009; Pang et al. 2012). Dendrobium species have various colors including red, yellow, white, green, pink and purple. Thus, the flowers can be subdivided according to labellum colors (Wang et al. 2007; Li et al. 2009a). Many Dendrobium species have fairly strong, pleasant scents. Of the 140 species evaluated, 40% produced scents ranging from floral to fruity to herbaceous (Kaiser 1993). However, most of them have no scents, such as D. devonianum (Lian and Li 2003). Furthermore, partial Dendrobium species have a small amount of nectar, such as D. fimbriatum and D. setifolium (Inoue et al. 1995; Wang et al. 2009). Flowers of Dendrobium species have significant interspecific differences. Flowers characteristics of some Dendrobium species are presented as Table 4.

During cultivation, Dendrobium can be divided into spring Dendrobium and autumn Dendrobium, according to the flowering period. Spring Dendrobium, mainly as a potted flower plant, generally is a deciduous species and flower from internodes in spring. Autumn Dendrobium, mainly as cut-flower, generally, is an evergreen species and about 20 flowers bloom from the stem tip (Zou and Liu 2010). The flowering period of Dendrobium is affected by environmental factors such as temperature and humidity (Table 5). Furthermore, pollination also impact flowering (Lian and Li 2003; Jin 2009; Wang et al. 2009). In some species, the flower begins to wilt once the pollinia is removed or pollinate to stigma. Under natural conditions, an unfertilized flower of D. Jiajiangense could generally open for about 10 d but fertilized flowers would wither in about 4 d (Pang et al. 2012).

Pollen and Stigma Biology

Gynostemium of Dendrobium is formed by androecium and pistil. Androecium, on top of the gynostemium, is made up by anther cap and pollinia. Pistil locate at the lower part of the gynostemium. Waxy pollen, 4 pollinia (2:1) is formed by 4 single pollinia, the anther cell is covered by anther cap. The anther cap fallen off by external force (Lian and Li 2003; Pan et al. 2010). Obvious differences in pollen vary within Dendrobium species. Previous studies have found that the order of pollen size is D. officinale > D. huoshanense > D. moniliforme. Pollinia surface has less and more sculpture in D. officinale and D. huoshanense respectively, while the surface of D. moniliforme is smooth (Wang and Wang 1989). There are differences in pollen vitality among species or during different flowering periods. Throughout the whole flowering period, pollen always has vitality in some species, such as D. speciosum, whose pollen vitality are greater than 95% (Slater and Calder 1988). The others have significant differences during different flowering periods. For example, pollen vitalities of D. crepidatum, D. chrysotoxum, D. moniliforme and D. Nobile are as below: bud stage (11.5%–32.0%), flowering (55.5%–84.0%), full flowering period (31.0%–71.5%) and flower withering period (0–9.5%) (Pan et al. 2009). For D. Candidum, the proportion of pollen with vitality in the bud stage is 29.4%, reaches the maximum 70.6% in the first day of flowers and then decreases. After the flowers have been open for a week, the proportion of pollen with vitality is 31.9% and decrease to 21.8% 12 d later (Zhu et al. 2011). In D. hercoglossum, pollen reaches the maturing stage within 4 d of flowering and drop rapidly on 12–14 d. Pollen within 2 d of flowering or 2 d impending withering of the flower is not suitable for pollination. The optimal period for pollen collection is 5–12 d after flowering (Li et al. 2009b). In D. Nobile, pollen has the strongest germination and seed rate after the first day of flowering. Pollinia are still available after 9 d of flowering (Wang et al. 2006). Stigma receptivity also differs among some species or flowering periods. The stigma of D. speciosum is receptive at anthesis and remains receptive until flower senescence (Slater and Calder 1988). Stigma of D. secumdum becomes receptive after the flowers have been open for 10–12 d (Kerr 1909). The best pollination periods of D. hercoglossum and D. Nobile are 6–10 d and 1–6 d after flowering respectively (Wang et al. 2006; Li et al. 2009b). The success of pollination is determined by stigma receptivity and pollen vitality, and pollen vitality is easily affected by environmental factors including temperature, humidity, and storage time. Pollen can be stored for a relatively longer period in optimum temperature condition, but its effective storage duration varies among Dendrobium species. Previous studies have indicated that pollen vitality can keep longer under dry conditions and decrease with storage duration. For instance, under dry conditions, the optimum effect for pollen storage of D. Candidum is at 4°C. The proportion of pollen with vitality is 48.7% after 8 d and 21.2% after 20 d respectively. In addition, the rate of decrease of pollen vitality is larger in early storage than later, for example, D. officinale (Zhu et al. 2011). To improve the rate of fruit set of Dendrobium by artificial pollination, pollen with higher vitality should be collected and stored in optimum conditions. Pollen of Dendrobium, as a two karyotype, theoretically is easy to save due to thick outer wall, desiccation tolerance and long life (Shi 1994).

Pollinia collected during disaccord flowering season of the parents, may be stored for artificial pollination (Wang et al. 2006; Zhu et al. 2011).

Visitors and their Behavior

Dendrobium species attract pollinators by a variety of ways, such as color, smell (D. speciosum), rest and mating place (D. loddigesii), providing reward: nectar (D. setifolium, D. devonianum, D. finisterrae), pseudopollen (D. jiajiangense, D. unicum), and shelter (D. jiajiangense) (Kjellsson and Rasmussen 1987; Inoue et al. 1995; Davies and Turner 2004; He 2008; Li et al. 2009a; Kamińska and Stpiczyńska 2011; Pang et al. 2012). However, some species such as D. infundibulum, have neither nectar nor any other reward for pollinators, and no odor could be perceived. They can simulate surrounding plants at the same flowing period in order to attract pollinators (Kjellsson et al. 1985). The flower of D. sinense mimics the alarm pheromone of honey bees in order to attract prey-hunting hornets for pollination (Brodmann et al. 2009). Other species such as D. speciosum, can take advantage of bonding factors: visual and olfactory, to attract pollinators. Potential pollinators of D. speciosum are attracted to the plant by large, bright, finely segmented, highly aromatic flowers, and the flowers on all plants in an area open almost synchronously (Slater and Calder 1988). A larger number of studies in different research areas have shown that there may be a variety of visitors in some areas and also rare or no visitors in other areas. The types and characteristics of visitors and the position of pollen carried by pollinators vary among species (Table 6) (Kjellsson and Rasmussen 1987; Inoue et al. 1995; He 2008; Kamińska and Stpiczyńska 2011; Pang et al. 2012). Visiting time of pollinators is affected by many factors. The specific climatic conditions during which pollinators visit the flowers are important for visitors, some visitors’ visit activities become intensive in warm weather (Pang et al. 2012). For example, in D. specious, osmophores scattered over the perianth produce a strong, sweet scent in sunny weather, this has an indirect effect on the pollinators’ behavior (Slater and Calder 1988). Some pollinators only visit flowers on sunny days, and the others have visiting action in rainy days until heavy rain (Slater and Calder 1988; Pang et al. 2012). There are differences in the peak period of visiting time of flowers among species, such as D. jiajiangense: 14:00–15:00, D. loddigesii: 11:00–14:00. The frequency of visitors is also affected by flowering phase, such as in D. loddigesii, there are more visitors at the full-blossom time than the end of the flowering period (He 2008; Pang et al. 2012). The behavior of visitors is different. Some visitors coming from a distance fly into flowers directly and leave quickly. They have no observation and testing behavior and don’t return to the same flower or circumjacent Data from (Kjellsson et al. 1985; Slater and Calder 1988; Inoue et al. 1995; He 2008; Brodmann et al. 2009; Pang et al. 2012)

The others have more testing behavior before visiting the flower and repeatedly visit the same flower (He 2008). Generally, visitors seldom fly into the flower passages and only pollinators will fly into the flower passages (Kjellsson et al. 1985; He 2008; Pang et al. 2012). The differences in visiting behavior will lead to variation of visiting time. For example, when Andrena parvula visit D. jiajiangense, visiting-time of a flower for seeking rewards is 37 ± 5 s, for rest or exercise on labellum is 33 ± 9 s, and for feeding the labellum fluff is 6 ± 1.5 s. Pang et al. (2012) found that Andrena parvula spend the night in the flower passage of D. jiajiangense to avoid rain or the heat from the sun, with its tails usually placed inwards and its head facing outwards. The flowers of D. jiajiangense provide shelter, and their male and female reproductive success rate is higher, 19% and 27% respectively, than those that don’t provide shelter (Pang et al. 2012). Different visitors of Dendrobium species have different visiting frequencies and efficiency as well as flower numbers (Kjellsson et al. 1985; He 2008; Pang et al. 2012). Of all the visitors, only those who removed pollinia and spread it to the stigma effectively are potential pollinators (Kjellsson et al. 1985; He 2008; Pang et al. 2012). It has been reported that the majority of Dendrobium species have specific pollinators (Richards 1997; Brodmann et al. 2009). The body size and structure of the pollinator form an ideal mechanical fi t with the height of flower passage and flower morphology (Table 7), and thus resulting in skillful removal of pollinia and accurate pollination (Figs. 1 and 2) (Slater and Calder 1988; He 2008; Pang et al. 2012). Body structure of the pollinator can also adapt to pollination, except for the adaptation between body size and flower passage. For example, pollinator of D. infundibulum, the mesonotum is usually hairy in bumble bees, but in B. eximius the hairs are absent from a narrow median area. The pollinia exactly fi t the bald patch on the back of the bee. The attachment of the pollinia to the bee appears to be very effective: bees were frequently observed flying around with D. infundibulum pollinia (Kjellsson et al. 1985).

Theoretically, pollination can be completed effectively via ingenious cooperation between flowers and the behavior of pollinator (Figs. 3 and 4). However, few researches have focused on the whole process of pollination, which is only reported in D. infundibulum by Kjellsson (1985). The gynostemium of D. infundibulum is short and erect. The functional stigma is a cavity formed mainly by the median stigma lobe, referred to as the rostellum. The dorsal side of the apex of the rostellum, the rostellum projection, carries a vesicle containing a sticky substance, which is released when the pollinator pushes the apex of the anther upwards. When a visiting bee with pollinia on its mesonotum starts backing out of the flower, the pollinia are caught by the slightly emarginated apex of the rostellar projection and are scraped off into the stigmatic cavity (Kjellsson et al. 1985). Flowers vary in size within the six recognized varieties of D. speciosum (vars speciosum, curvicaule, grandiflorum, hillii, capricornicum and pendunculatum) and are pollinated when visited by bees of appropriate size (Slater and Calder 1988). Although obligate pollination can improve the pollination efficiency in plants, the natural fruit set of Dendrobium is lower on account of scarce pollinators in the field conditions. The reproductive fitness of Dendrobium species, such as D. infundibulum will be increased through extended flowering season to make up lacking pollinators (Kjellsson et al. 1985).

 Post-pollination Response

Generally, when pollinia is removed or spread into stigma, the perianth closes around the column and the flower color may change. This response, by hiding the central target area, prevents prospective visitors being specifically attracted to previously pollinated flowers. The senescent perianth intensifies or dulls in color, increasing the visual contrast between the inflorescence and the surrounding vegetation (Kjellsson et al. 1985; Slater and Calder 1988; Slater 1991; Jin 2009).

Breeding System and Fruit Set

Previous studies have indicated that Dendrobium showed high incompatibility in interspecific pollinations (Table 8). The majority (72%) of the 61 species that were self-pollinated showed self-sterility. Self- and interspecific incompatibility is expressed by flower abscission and not by inhibition of pollen germination or pollen tube growth (Johansen 1990). Till now, self-compatible and partly self-compatible Dendrobium species include 19 species, i.e., D. albosangnineum, D. bilobuatum, D. brymerianum, D. crystallinum, D. erostelle, D. exile, D. formosum v. giganteum, D. gibsonii, D. heterocarpum, D. infundubulum, D. nathanielis, D. pendulum, D. phalaenopsis, D. salaccense, D. senile, D. tetrondon, D. tortile, D. fi mbriatum, and D. jiajiangense (Johansen 1990; Pang et al. 2012). Spontaneous self-pollinatione exists in D. brymerianum, D. erostelle and D. telrdon (Johansen 1990), but most Dendrobium species have no self-pollination and apomixes. Therefore, these species are pollinator-dependent for fruit set in wild state. Dendrobium species have more flowers and less fruit. Except higher fruiting rate in D. jiajiangense (46.1%), the others are relatively lower, such as D. sinense (13%), D. fi mbriatum (11.11%), D. Candidum (0.31% and 0.86%) and D. loddigesii (0.48%) (Lian and Li 2003; He 2008; Zhu et al. 2011; Pang et al. 2012). The main reasons of lower fruiting rate of Dendrobium species are as follows: (1) special flower structure: four pollinia are covered by two anther cap that drop through external forces; (2) lacking or even no pollinators in the field conditions; (3) there are spatial segregation between stigma and androecium and some species have well developed rostellar projections that autogamy seems mechanically impossible; (4) under natural conditions, few Dendrobium species have self-pollination and apomixes; (5) It is also considered to be a combined result of self-infertility and the absence of rewards offered by the flower (Kjellsson and Rasmussen 1987; Johansen 1990; Liang and Li 2009; Pang et al. 2012). Many factors impact the fruit set of Dendrobium species, such as nutritional status of the female parent, parental attributes and pollination time (Lian and Li 2003; Pan et al. 2010; Zhu et al. 2011). Dendrobium species is a package plant. It can grow vigorously and accumulate more nutrients than individuals (Lian and Li 2003). The nutritional status of Dendrobium may affect fruiting rate. In the study of artificial pollination on D. devonianum and 4 medicinal Dendrobium (D. crepidatum, D. chrysotoxum, D. monilifome and D. Nobile), fruiting rate was significantly associated with the number of tree/individual and the number of inflorescences and flowers. Within a certain range, more plants and less inflorescences and flowers result in higher fruiting rate and larger pods (Liang and Li 2009; Pan et al. 2010). Zhu et al. (2011) observed that fruiting rate of direct crossing and reciprocal crossing (I, mother × G, father) were 85.0% and 89.5%, on the contrary, were 88.5% and 92.0%. The fruiting rate of D. Candidum is 100% when pollination is timely carried out (Zhu et al. 2011). In four medicinal Dendrobiums (D. crepidatum, D. chrysotoxum, D. monilifome, D. Nobile), the fruiting rate of artificial selfing and hybridization in flowering period is higher than that in bud period, e.g., flowering period: selfing (25%–75%), hybridization (16.6%–50%); bud period: selfing (14.2%–50%), hybridization (16.6%–33.3%) (Pan et al. 2010). Most of Dendrobium plants are cross–pollinated. In D. chrysanthum, if the pollinator fails in removing the pollen-masses when visiting the flower, the plant can skillfully finish fertilization through self-pollination (Darwin 1862). Artificial pollination experiments found that different pollination modes have different effects on fruit set and seed quality. Selfing can bring various degree of inbreeding depression (Wang et al. 2009). The different pollination modes can lead to different fruiting rate. In general, there are significant differences between selfing and hybridization (Wang et al. 2009). In D. devonianum, fruiting rates of different plexus cross-pollination, the cross-pollination, inflorescence pollination, as well as selfing are 50%, 11.11%, 4.12% and 3.33% respectively (Lian and Li 2003).  

In D. fi mbriatum, fruiting rates of the cross-pollinations in interspecific populations, intraspecific populations, the same tree and selfing are 76.4%, 58.62%, 21.74% and 11.11% respectively (Wang et al. 2009). In D. Candidum, fruiting rates of artificial hybridization and selfing are 82.6% and 7.3% respectively (Zhu et al. 2011). In D. hercoglossum, fruiting rate of selfing is only 4.3% (Li et al. 2009b). Fruit abortion will happen after pollination, whose differences depend on pollination modes. In D. Candidum, the numbers of fruit abortion after selfing and hybridization are 153 and 37 respectively. The peaks of abortion after selfing and hybridization are within 5 d and 20 d. The numbers of fruit abortion after 5 d of selfing and hybridization account for 96.1% and 45.9% of the total numbers in D. Candidum, respectively (Zhu et al. 2011). The self-pollinations are monitored and after 10 d four varieties of D. speciosum (D. speciosum var. speciosum, D. speciosum var. curvicaule, D. speciosum var. grandiflorum and D. speciosum var. pendunculatum) aborted all developing capsules (Slater and Calder 1988). Seed quality is affected by the pollination modes. Normally, seed quality of hybridization is superior to that of selfing (Wang et al. 2009; Pang et al. 2012). In fruit obtained from selfing of D. fi mbriatum, seeds with no or imperfect embryos account for about 2/3 of the total, only 1/3 of them have vitality. In fruit obtained from hybridization, the percentage of seeds with activated embryos is significantly higher than those obtained from selfing. Seeds with activated embryos obtained from interspecific populations are significantly higher than intraspecific populations; their percentages are 96.70% and 70.05% respectively (Wang et al. 2009). Seed viability of D. jiajiangense under the treatment of hand-cross-pollination was the highest, and that of seeds resulting from hand-self-pollination was lowest (Pang et al. 2012). In D. fi mbriatum, seed germination rates of self-pollinations, cross-pollinations, interspecific populations and intraspecific populations cross-pollinations are respectively 77.42%, 72.73%, 98.50% and 92.17% (Wang et al. 2009). Artificial-assisted-pollination strongly improve fruiting rate of Dendrobium, and artificial selective pollination also enhance qualities of fruit and seed (Pan et al. 2010; Zhu et al. 2011).

Fruit and Seed

After successful pollination of Dendrobium, the sepal usually closes within 2 d; the ovary turns green and expands after a certain period, e.g., D. devonianum 7 d, D. officinale 4–5 d and D. loddigesii 4–5 d. However, flowers become withered, ovary turns yellow and fall off after unsuccessful pollination (He 2008; Liang and Li 2009; Wang et al. 2009). The growth and development of ovary change with time (Table 9). For example, after pollination of D. officinale, the prototype of fruit will be observed after 6 d, grow rapidly within 20 d, keep stable after 60 d and the fruit mature after 70 d (Zhu et al. 2011). After pollination of D. devonianum, fruit grow rapidly within 7–15 d (width increase faster than length), and then almost stop growing and enter maturity period (Liang and Li 2009). The maturity time of the capsule of the same Dendrobium species is different at different years, e.g., the average time of D. secundum was 67 d in 1986 and 87 d in 1987 (Johansen 1990).  Fruit of Dendrobium: Capsule with edges, oval or strip shape, gynostemium-tipped fruits. Containing 0.1–1 million seeds, yellow, fi ne as dust, 0.3–0.4 μg each seed (Wang et al. 2007). The capsule automatically cracks after maturity and spreads the seeds out. The seeds, no endosperm, having after-ripening phenomena, germinate with difficulty under natural conditions, and their germination rates are usually less than 5% (Zhang et al. 2000; Wang et al. 2006). Seed germination of Dendrobium need mycorrhizal symbiosis (Guo et al. 2000; Song et al. 2004; Swamy et al. 2007; Li et al. 2010), which vary at different growth stages in natural state. Epulorhiza sp., Mycena dendrobii, Rhizoctonia sp., Mycena orchidicola, and Gliocladium sp. can improve seed germination of D. Candidum. In addition, its seedlings are enhanced by Cephalosporium sp., Epulorhiza sp. and Gliocladium sp. (Guo et al. 2000). Guo et al. (2000) also found that Gliocladium sp., Epulorhiza sp. and Mycena dendrobii can form mycorrhiza structure with roots of D. Nobile, thus improving the growth of the seedlings. Seed of Dendrobium: Under light microscope they appear as irregular spindle, mid-intamescentia, both ends attenuate: one end is broad obtuse and the other is apiculate. There are also differences in the length of both ends among species. Seed structure is formed by testa and embryo. Yellow embryo is ellipsoidal and occupies the middle part of seed. When the fruit is ripe, there are no differentiations of radical, hypocotyl, germ and cotyledon in the original embryo period. Testa is paper like and transparent, located in the middle of seed. It wraps up the embryo which stores gas at both ends (Wang and Xiao 2010). There are great differences in characteristics of seeds among different species, such as the size of seed and embryo as well as the proportion of the gas chamber (Table 10) (Swamy et al. 2007). The qualities of fruit and seed of Dendrobium are affected by pollination modes, plant hormones and nutritional status (Johansen 1990; Zhang et al. 2008; Wang et al. 2009). Generally, fruit and seed from hybridization are superior to those from selfing (Wang et al. 2009). Moreover, plant hormones have effects on fruits and seeds. Earlier studies found that, fruit treated by 2,4—D after one month of artificial pollination, needed 6 month until maturity, significantly enlarged, e.g., length and width of fruit treated were 1.5–3 cm and 0.5–1 cm larger than those that are untreated. Meanwhile, fruit treated by 2, 4—D can produce more seeds than those that are untreated (Fig. 5) (Zhang et al. 2008). After application of NAA or IAA on D. phalaenopsis, parthenocarpic fruit formation occurred. No seeds were formed in these parthenocarpic developed capsules (Johansen 1990). The seed longevity will be shortened by high humidity and temperature conditions. In order to gain a longer store life, the specific procedures are as follows: (1) the capsule is put into sterile tubes with 0.1% mercuric chloride for 10 min and disinfected; (2) the capsule is carefully taken out using sterile tweezers and washed 3 times with sterile water; (3) immediately, it is dried for 1–3 d in a sterile petri dish, transferred into sterile tubes and tamponed; (4) the tubes are placed into desiccators and stored in the refrigerator at 10°C or below 0°C (Wang et al. 2006).

Perspective

Dendrobium, as the second largest genera of Orchidaceae, consists of 1000–1400 original species with higher ornamental and medicinal values (Kuehnle 2006; Wang et al. 2007; He et al. 2008). However, only about 70 species reproductive biological characteristics have been reported to date. Therefore, there are lots of species whose reproductive biological characteristics are unknown and worthy of researching in the future. It is most important to probe the following interesting questions. Flower morphology: diversity and function. Flower morphology of Dendrobium varies remarkably among species. For example, there exist great differences in the flower morphology between D. aphyllum and D. brymerianum. It is unclear why they have various flower morphologies and what is the specific value in propagation adaptation? Pseudopollen: reward or deception. Many studies have paid attention to the morphological characteristics and chemical composition of pseudopollen covering in labellum. It is speculated that pseudopollen may be a reward for pollinators because it contains starch and protein. However, till date there has been no direct experimental evidence. Therefore, further studies are also needed in order to fully understand how the pseudopollen plays a role in pollination. 

Anther Cap: Structure and Function

The structure of anther cap obviously varies among the Dendrobium species though why this is so, is still unknown. Darwin reported that pollinators can pollinate the flowers of D. dixanthum, but D. chrysanthum will complete self-pollination skillfully, when pollinators do not take away the pollinia. Moreover, Liang and Li (2009) also found that abscission of anther cap needed external force. However, the mechanism of anther cap-releasing pollinia is rarely mentioned in the current study. Moreover, Pang et al. (2012) observed that flower passage would be blocked up by the fallen anther cap; whose adaptive significance in reproductive biology remains mysterious.

Lower Fruiting Rate: Causes and Countermeasures

Dendrobium species have very low fruiting rates, for example, D. sinense (13%), D. fi mbriatum (11.11%), D. Candidum (0.31% or 0.86%) and D. loddigesii (0.48%) (He 2008; Liang and Li 2009; Zhu et al. 2011; Pang et al. 2012). Although some scientists analyze the reasons resulting in lower fruit set, corresponding achievements are insufficient to some extent. Most of Dendrobium species are endangered plants in nature. Why are those rare and how will they be protected l are still worthy of investigation.