New insights into the phenology, genetics and breeding system of critically endangered Nardostachys grandiflora DC

Abstract

The critically endangered Nardostachys grandiflora DC is a highly renowned medicinal and aromatic plant of the Western Himalayan region, traditionally used in treating heart palpitations, convulsions, epilepsy and insomnia, as well as being a source of aromatic oil used as incense, flavoring agent and in quality perfumes. This species is on the verge of extinction due to over-exploitation and habitat destruction and needs immediate preventive measures. Despite being essential for domestication, there have been few studies of the natural variation, breeding and genetic system of this species possibly due to its remote location and sparse populations. The present study has revealed that N. grandiflora, which is restricted to specific ecological niches in high alpine regions 3000–5000 m asl, is under stress due to its infrequent flowering nature, which limits generation of new variation, although some intra-population variability with regard to lamina margin (dentate and wavy) and type of stigma (capitate and trilobed) were observed. This limited variability prevents the species from colonizing new niches, as it mostly multiplies through rhizome ramets. Several interesting features have also been revealed, e.g. presence of bicolpate, tricolpate, tetracolpate and pentacolpate pollen grains as well as isobilateral, decussate and tetrahedral types of tetrad within the same or in different flowers. This study also establishes a new diploid chromosome number (2n = 78), the plant’s genomic allohexaploidy, up to 58% cross pollination; protoandry, its infrequent flowering nature, a high pollen ovule ratio of 6135 ± 327.86 per flower and poor seed set.

Keywords:

• Nardostachys grandiflora
• spikenard
• critically endangered
• medicinal and aromatic
• temperate Himalaya
• pollen grains

Introduction

Nardostachys grandiflora DC (syn N. jatamansi, N. chinensis) locally known as jatamansi and nard, a monotypic species considered as the most primitive genus in Valerianaceae (Weberling 1975; Government of Nepal 1993), is a perennial rhizomatous herb growing in sub alpine/alpine areas of the northwestern Himalayan region of Pakistan, India, Nepal, China and Tibet, mostly growing in steep, moist, rocky, undisturbed grassy slopes at 3000–5000 m asl (Chauhan and Nautiyal 2005). The species has been assessed as ‘Endangered’ in Arunachal Pradesh, Sikkim and Himachal Pradesh, India, ‘Critically Endangered’ in Uttrakhand and Kashmir Himalayas, India (Kaul 2001) and ‘Vulnerable’ in Nepal during the 2001 CAMP (Conservation Assessment and Management Plan) workshop (Bhattarai et al. 2002). It is listed in CITES (Convention on International Trade in Endangered Species of Wild Flora and Fauna) appendix II and the Red Data Book of Indian Plants (Nayar and Sastry 1988; Mulliken 2000).

Recorded use of N. grandiflora rhizomes in relieving pain dates back to the sixteenth-century “Compendium of Materia Medica” in China (Fu 1993). The fibrous rhizomes, used as incense, yield an important essential oil known as spikenard oil that is effective in treating leprous wounds (CSIR 1966; Thakur and Hussain 1989; Amataya and Sthapit 1994; Mulliken 2000; Mulliken and Crofton 2008) and incense sticks made from its rhizomes are sold in Middle East markets (Burbage 1981). In India, in traditional Unani and Ayurvedic systems of medicine, it is valued for its antispasmodic properties which are helpful in treating abnormal heart palpitations and constipation, as well as to regulate urination, menstruation and digestion (Jain 1994). Indian classical medical texts, such as Charak samhita and Shushruta samhita, have prescribed smoking pellets (small balls used for smoking) made from the rhizome for treating coughs, hiccups and asthma. The rhizomes are used as a nervine tonic, stimulant, external painkiller, antiepileptic, antiseptic and also in treating hysteria, convulsions, heart palpitations, high blood pressure, fever, anxiety, insomnia, asthma, acidity and,bronchial problems (Chauhan 1984; Government of Nepal 1993). Rhizomes are used by amchis (traditional medicine practitioners trained in Tibetan medicine) for treating complaints including epilepsy, wounds, coughs, colds and high blood pressure (Ghimire et al. 2005). In Pakistan, it is used to treat hysteria, epilepsy, neurosis, insomnia, constipation, scorpion stings, Bell’s palsy, Parkinson’s disease, indigestion and age related deafness (Kazmi and Siddiqui 1953; Zaman and Khan 1970; International Trade Centre 1982; Khan and Zaidi 1989). It has also been used as a single ingredient to treat hypertension, as an aphrodisiac and to aid memory (Mulliken 2000; Mulliken and Crofton 2008). The most recent review of this species has been provided by Kamini and Raina (2013).

The majority of N. grandiflora supplies come from Nepal (82 ± 5%), India (13 ± 5%) and Bhutan (5 ± 4%) and are exported to France, Spain, Germany, USA and others, mostly as unprocessed rhizomes, as well as essential oil and marc in smaller amounts (Mulliken 2000; Olsen 2005; Mulliken and Crofton 2008). India imports about 1000 t of rhizomes annually from Nepal (Olsen 2005; Mulliken and Crofton 2008) and nard oil fetches about 12,000–30,000 Indian rupees (≈US$240–600) per liter in international markets (Jhunjhunwalla 2010). N. grandiflora in trade is sometimes confused with Valeriana jatamansi (Mulliken 2000); also rhizomes of Valeriana jatamansi, Selinum candollei and Selinum vaginatum can be mixed in with it leading to adulteration (Singh et al. 2011).

Over-exploitation and habitat destruction have necessitated urgent steps to conserve and domesticate this species; for this suitable agro-practices are needed as well as better strains for optimal productivity. As is the case with most high-value temperate endangered medicinal species, genetic improvement has received almost no attention. Scientific studies have been limited to aspects of its taxonomy, distribution, ecophysiology, phytochemistry and ethnobotanical uses (CSIR 1966; Samant et al. 1998; Prakash 1999; Airi et al. 2000), but studies of its reproductive potential, genetic and breeding system, which are essential for its conservation and improvement, are lacking.

Materials and methods

Phenological studies were carried out during 2011–2012 on randomly selected plants sourced from Tirthan Wildlife Sanctuary, Kullu, Himachal Pradesh, India (altitude 2400–3000 m asl; 31° 41′55.75″ N, 77°31′46.90″ E).

The vegetative and floral studies were conducted as per standard literature (Lawrence 1951; Weberling 1989; Kuafman et al. 1989). Fresh leaves were preserved in a formalin:acetic acid:alcohol solution (5:5:90) at room temperature, dechlorophylled by boiling in chloral hydrate solution (50 g dissolved in 20 ml water) to determine the palisade ratio, vein-islet/vein termination number and stomatal index as per Fahn (1967).

Splitting of anther wall showing pollen grains was taken as sign of dehiscence, whereas a shining surface of stigma tip was considered as receptive stage (observed under hand magnifying lens). Pollen–ovule ratio was calculated as per Cruden (1977) and pollen viability was calculated on the basis of 1% acetocarmine staining test. Fruit and seeds were studied at maturity stage.

Young floral buds fixed in Carnoy’s fixative (absolute alcohol, glacial acetic acid and chloroform in the ratio of 1:1:1) for 24 h, washed and then stored in 70% absolute alcohol at low temperature before use for meiotic studies. Freshly prepared 1% acetocarmine stain was used for chromosomal staining by usual squash method.

Pollination studies were analyzed using a t-test (15 replications with minimum of 15 flowers per replication). For open pollination unopened healthy floral buds were tagged and left as such, whereas for assessing autogamy, healthy floral buds, about to open, were enclosed in butter paper bags. Increase in ovary size and its transformation into fruit was taken as the basis of fruit set. The number of flowers that finally transformed into fruit out of the total number of flowers tagged for a particular treatment/replication was used for calculating the fruit set percentage. As the fruit in this species is dry achene, the fruits that appeared firm on pressing between the fingers were considered as having developed seed and seed set was calculated on the basis of total number of flowers tagged. Germination test was performed at 23 ± 2°C in a growth chamber and radicle protuberances were taken as sign of germination.

Statistical analysis was conducted as per Gomez and Gomez (1984). Ocular and stage micrometers (ERMA, Tokyo, Japan) were used for micro-measurements and microscopic examination was made using an Olympus trinocular research microscope (model CH20iBIMF, New Delhi, India).

Results

Phenology

Plants (Figure 1a) of N. grandiflora are erect perennial rhizomatous herbs with an aerial stem (attaining a height of 20.55 ± 2.28 cm), consisting of a rosette of radical leaves at the base and a seasonal pubescent, quadrangular, hollow flowering shoot bearing cauline leaves topped by floral heads. Flowering and fruiting occurs during June to September, with underground perennation occurring from October onwards; sprouting occurs the following year after the snow melts. The perennating rootstock consists of stout, cylindrical rhizomes (Figure 1b) covered with reddish brown remains of dry petioles of previous radical leaves and adventitious roots.

Figure 1. N. grandiflora (a) plant; (b) rhizome; (c) inflorescence.

Radical and cauline leaves are mostly estipulate, petiolated (radical) or sessile (cauline), simple, lanceolate (radical) or ovate (cauline), glabrous, longitudinally veined and lamina entire with mostly wavy smooth margin having acute tip and attenuate base. However, a single plant having lamina with dentate margin was also observed in the natural population. Histological features including the palisade ratio, vein islet and termination number/mm² and stomatal type/index (Figure 2d) are determined here for the first time in this species and are given in Table 1.

Figure 2. N. grandiflora (a) transverse section of ovary; (b) capitate stigma; (c) trilobed stigma; (d) stomata.

Table 1. Some leaf (radical) constants of N. grandiflora.

Feature	Observation
Palisade ratio	6.47 ± 0.35
Vein-islet number	1.77 ± 0.16 per mm^2
Vein termination number	1.35 ± 0.16 per mm^2
Type of stomata	Anomocytic (Figure 2d)
Size of stomata (mm)	0.02 ± 0.001 × 0.035 ± 0.001
Stomatal index	22.24 ± 1.06

There is a single flowering shoot per plant (Figure 1c), which bears two to five condensed dichasial cyme heads, with the main largest terminal head having eight to 25 flowers and lateral heads having three to nine flowers, where each cyme head is subtended with a pair of ovate, greenish involucre bracts. Bud opening is asynchronous and not more than two to eight flowers are open at any time in the main terminal cyme with only one to two having dehiscing anthers.

Flowers are bracteate, sessile, complete, actinomorphic (or irregular due to the absence of the fifth stamen and with only single ovule in one of the three locules of ovary), bisexual and epigynous. Floral bracts are pubescent, oblong and pale green with size decreasing from basal to terminal flowers. A gamosepalous calyx, at the rim of the thalamus cup, is constituted of five, distinct, valvate, dentate and pubescent lobes which are purple towards the margins and green at the middle. Calyx size increases after fertilization and remains persistent till fruit dispersal. The corolla is tubular at base with five free valvate lobes above, campanulate, purple with pinkish tinge and whitish hairs inside, becoming pale white later.

Four well-developed free, epipetalous, haplostamenous and oppositipetalous stamens represent the androecium. One extra rudimentary anther is also present in some flowers. Anthers are bithecous, exerted, dorsifixed, kidney shaped and violet colored, dehiscing by longitudinal slits. Filaments are purple above and with a white hairy base.

The epigynous gynoecium is represented by three fused carpels. Ovary is trilocular (Figure 2a) bearing single anatropous ovule showing apical placentation. Linear style bears capitate (Figure 2b) as well as trilobed (Figure 2c) papillate stigmas in different flowers, and is purple colored above and whitish towards the base. Stigmatic tip is papillate and violet colored. Fruit is dry pubescent indehiscent achene with persistent calyx bearing a single seed.

Qualitative, quantitative and phenological parameters of this species are summarized in Tables 2 and 3.

Table 2. Qualitative and quantitative features of N. grandiflora.

Plant part	Qualitative	Quantitative
Stem	Erect, represented by flowering shoot.	Up to 20.55 ± 2.28 cm (height)
Rootstock	Perennial rhizomatous with adventitious roots.	L: 6.9 ± 0.41cm
		B: 6.35± 0.5 mm
Leaves
Radical	Rosette arrangement, petiolated, simple, estipulate, lanceolate, lamina entire with acute tip and attenuated base.	L: 11.9 ± 0.56 cm
		B: 1.85 ± 0.13 cm
Cauline	Paired, opposite, sessile, simple, estipulate, ovate, lamina entire, acute tip and attenuated base showing decurrent attachment.	L: 4.28 ± 0.22 cm
		B: 1.07 ± 0.11 cm
Inflorescence	Dichasial cyme with condensed cyme heads.	Length of flowering shoot: 20.55 ± 2.28cm
		No. of flowering shoots/plant: 1
Flower	Bracteate, actinomorphic (or irregular), bisexual, complete, epigynous.	L: 10.8± 0.20 mm
		B: 3.3± 0.18 mm
Calyx	Gamosepalous with five dentate lobes which are green in centre and purplish towards margin. Odd sepal posterior.	Lobe size: 1.5–2 mm
Corolla	Petals fused below forming tube with five lobes above, campanulate, odd petal anterior and purplish with pink tinge.	L:
		Total: 7.9 ± 0.12 mm
		Fused: 4.9± 0.06 mm
		B: 3.3 ± 0.18 mm
Androecium	Four stamens, haplostamenous, epipetalous, oppositipetalous. Anthers dorsifixed, kidney shaped with two violet colored anther lobes. Staminode present in some flowers.	Filament length: 8.79 ± 0.19 mm
		Anther size: 2.1± 0.09 × 1.3± 0.12 mm
Gynoecium	Tricarpillary syncarpous, trilocular with single fertile anatropous ovule in one locule showing apical placentation and the rest empty. Style linear and papillate stigma trilobed or capitate.	Ovary
		L: 2.11 ± 0.05
		B: 1.94 ± 0.05 mm
		Style
		L: 9.3 ± 0.28 mm
		Stigma: 0.4 × 0.31 mm
		Ovule size: 0.47 × 0.36 mm
Fruit/seed	Achene	L: 4.3 ± 0.18
	Pale brownish colored.
		B: 2.1 ± 0.05 mm
		100 fruit/seed weight: 0.389 gm

Abbreviations: L, length; B, breadth.

Table 3. Pheno-phases of N. grandiflora.

Pheno-phases	Time of year
Re-sprouting	May onwards
Development of inflorescence	June onwards
Flowering	June–August
Peak flowering	Mid July to first week of August
Fruit formation	August onwards
Seed ripening and dispersal	September–October
Perennation	Mid October to May

Breeding system studies

Anther, stigma maturity and pollen ovule ratio

Anther dehiscence is initiated (Figure 3a) 20–25 days after floral bud appearance and dehiscing anthers are exerted out of corolla tube. The stigma tip is initially positioned below the level of anthers, and later gets pushed out of corolla tube due to increased stylar growth after complete anther dehiscence. At this stage the stigma tip becomes visibly glossy, indicating receptivity (Figure 3b).

Figure 3. N. grandiflora (a) anther dehiscence; (b) stigma receptive stage.

The number of pollen grains ranges from 5320 to 7840 (average 6135 ± 327.86) per flower. With only a single ovule per flower, pollen ovule ratio works out to be 6135 ± 327.86 per flower.

Male meiosis

A uniform number of 39 bivalents at diplotene and metaphase-I (Figure 4a, 4c) followed by 39–39 segregation of chromosomes at anaphase-I (Figure 4b) stage was observed in all the pollen mother cells studied. The bivalents were characterized by both interstitial as well as terminal chiasmata and no disjunctional abnormalities were observed. However, three types of tetrads, i.e. isobilateral (Figure 5a), decussate (Figure 5b) and tetrahedral (Figure 5c) were observed in the same as well as different anthers. The normal meiosis is reflected in high pollen stainability of 92.12%.

Figure 4. N. grandiflora (a) metaphase-I; (b) anaphase-I; (c) diagramatic representation of metaphase-I.

Figure 5. N. grandiflora (a) isobilateral tetrad; (b) decusate tetrad; (c) tetrahedral tetrad.

Spherical to oval shaped pollen grains were mostly tricolpate (Figure 6a). However, bicolpate (Figure 6b), tetracolpate (Figure 6c) and pentacolpate (Figure 6d) pollen grains were also observed within the same anther (Figure 6e), with some anthers having only tricolpate pollen grains. Pollen stainability was found to marginally increase with increase in pollen aperture number, with the lowest (86.38 ± 0.83%) observed in bicolpate and the highest (99.1 ± 0.8%) in pentacolpate pollen grains (Table 4).

Figure 6. N. grandiflora (a) tricolpate pollen grain; (b) bicolpate pollen grain; (c) tetracolpate pollen grain; (d) pentacolpate pollen grain; (e) all pollen morphs in one anther.

Table 4. Dynamics of different type of pollen grains in the same flower.

Type of pollen grain	Bicolpate	Tricolpate	Tetracolpate	Pentacolpate
Shape	Spherical-oval	Spherical-oval	Spherical-oval	Spherical-oval
No./flower	0–955	4528–6135	0–604	0–48
% stainability	86.38± 0.83	92.12 ± 1.72	97.34 ± 0.85	99.1 ± 0.8
% of pollen grain in flower	0–15.56	73.8–100	0–9.84	0–0.74
Length (mm)	0.075 ± 0.002(0.06–0.09)	0.062 ± 0.03(0.06–0.07)	0.074 ± 0.004(0.05–0.11)	0.090 ± 0.001(0.08–0.098)
Breadth (mm)	0.054 ± 0.001(0.05–0.06)	0.062 ± 0.001(0.05–0.07)	0.074 ± 0.004(0.05–0.11)	0.090 ± 0.001(0.08–0.094)

Values in parenthesis are ranges.

Pollination studies

The potential of open and autogamous pollination was evaluated to determine their effect on fruit set and seed germination. Insect pollinators, including different species of bees, flies, butterflies, and ants were observed visiting the flowers. Maximum fruit and seed set of (18.94% and 8.08%) were obtained in open pollination which was statistically higher than self pollination (4.52% and 1.85%). However, seeds from both these sources exhibited about 71% germination (Table 5).

Table 5. Effect of different pollination methods on fruit set, seed set and seed germination in N. grandiflora.

Treatment	Fruit set (%)*	Seed set (%)*	Seed germination (%) **
Open pollination	18.94 ± 0.18	8.08 ± 0.16	71.64 ± 0.42
Autogamy	4.52 ± 0.24	1.85 ± 0.17	71.09 ± 0.30
T calculated value	53.63	29.28	1.21

^* Out of total number of flowers studied

^** Out of total seed set

Discussion

The floral characters of plant species are indicators of the type of breeding system, having a bearing on conservation and domestication; such studies are lacking in N. grandiflora. Most of the morphological and floral characters studied agreed with earlier records (Clarke 1885; Amatya and Sthapit 1994; Dutta and Jain 2000). However, some plants with a dentate margined radical leaf lamina, in contrast to the common wavy margin, were present in wild populations, which is a first record for this species. Leaf variants (narrow and broad) have also been reported in another endangered temperate Himalayan perennial herb Picrorhiza kurroa (Nautiyal and Nautiyal 2004). Similar variations have also been reported in Gaultheria fragrantissima (Vijayakumar and Paulsamy 2010).

Some leaf features, including palisade ratio, vein islet number, vein termination number and stomatal type/index/size have been studied for the first time during the present study (Table 1). Anomocytic stomata were observed, as in taxa of Ranunculaceae, Malvaceae, Scrophulariaceae, Cucurbitaceae and Geraniaceae (Fahn 1967).

The presence of capitate as well as trilobed stigmas in different flowers, indicating the existence of different floral types in this species, is also reported here for the first time. Similar features have been reported in Valeriana jatamansi (a gynodioecious species of family Valerianaceae) wherein female flowers posses trifid stigmas and bisexual flowers posses unifid stigmas (Raina 2006; Nawchoo et al. 2012).

These hitherto unknown leaf and floral features have enriched our knowledge about this species and have the potential of providing leads in any further genetic improvement programs.

Phenology

Phenological episodes, e.g. sprouting (May onwards after snow melting), flowering (July to August), seed setting (August to October) followed by perennation were observed as per earlier reports (Ghimire et al. 2005; Chauhan et al. 2008). However, the asynchronous flowering pattern in this species is significant as it is generally encountered in species that require pollinator services, e.g. insects, for effecting allogamous pollination; asynchronous flowering overcomes the risk of pollinator limitations. Another important phenological feature observed in this species was its infrequent flowering behavior. Out of an approximate population size of about 3500 plants growing at the study site, only 50–60 plants bore flowers in a season and this trend was repeatedly observed for two seasons (2011 and 2012). This limits the possibility of good seed production and genetic variation.

Anther dehiscence and stigma receptivity

The temporal and spatial separation between male and female phase to a large extent determines the breeding system of a species. Anther dehiscence prior to the stigma becoming receptive indicates the protoandrous nature of this species. Stigmas become receptive only after complete anther dehiscence, at which stage they protrude out of the corolla tube presenting themselves to pollinators. In Gentiana kurroo, another temperate critically endangered species, stigmatic lobes remain adpressed to each other till complete anther dehiscence, unfolding later and becoming receptive (Raina et al. 2003). Protoandry has been reported in many other temperate Himalayan medicinal plants, e.g. Aconitum heterophyllum, A. balfourii, Inula racemosa, Podophyllum hexandrum, Digitalis purpurea, Digitalis lanata, Digitalis grandiflora, and Rheum emodi (Behera 2002; Wafai et al. 2005; Wani et al. 2006, 2009; Nazir et al. 2008; Nautiyal et al. 2009a, 2009b). However, Chauhan et al. (2008) reported N. grandiflora as protogynous.

Temporal separation of the maturity of two sexes within a flower can minimize fitness losses from self-pollination within flowers, reduce interference between male female functions and increase female fitness (Lloyd and Webb 1986). In N. grandiflora dichogamy appears to play a key role in ensuring cross fertilization, which is crucial for generating variation in an otherwise infrequent flowering and mostly vegetatively multiplying species.

Pollination studies and breeding behavior

A maximum fruit set of 18.94% occurred in open pollinated flowers, but only 4.52% fruit set was observed in passive autogamy. The same trend was noticed in seed set, with 8.08% and 1.85% seed set in open pollination and passive autogamy respectively, indicating preference of its flowers for cross pollination. Based on fruit/seed set percentage, it is estimated that flowers of N. grandiflora exhibit lower chances of self pollination (23.86–42%), compared to higher chances of cross pollination (58–76%). According to a general hypothesis in alpine botany, selfing rates should increase with increasing altitudes (Schroter 1926; Bliss 1962; Garcia and Totland 2009; Korner and Paulsen 2009), as pollinator abundance and activity have been shown to become limiting factors for successful pollination due to hostile climatic conditions at higher altitudes (Arroyo et al. 1982, 2006; Bingham and Orthner 1998; Medan et al. 2002). However, Eritricium nanum has higher cross pollination rates at high altitudes, and higher selfing rates at lower altitudes (Wirth et al. 2010); also, in Chaetanthera renifolia (Asteraceae) of the Andean region, long-lived stigmas allow extended cross pollination (Diaz et al. 2011), leading to higher cross pollination rates at high altitudes. The infrequent flowering habit observed in this species limits the chances of generating new genetic combinations, as the species mostly multiplies vegetatively. Hence the meager number of flowers produced in a growing season, as well as offsetting pollinator limitation, is more adapted to cross pollination than self pollination.

Our findings support the conclusion of Chauhan et al. (2008) that this species exhibits both cross as well as self pollination, but disagree with Chauhan et al. (2008), who found that mostly self pollination occurs. Wirth et al. (2010) could not demonstrate a higher selfing rate in the alpine species Eritrichium nanum (Boraginaceae) and called for a critical re-examination of the hypothesis of higher selfing in alpine species.

Other direct evidence of the mostly cross pollinating behavior of N. grandiflora established during the present study includes attractive floral attributes, flowers aggregated into a head, dichogamy (protandrous) and receptive stigma protruding out of corolla tube. Insect pollinators such as species of bees, flies, butterflies and ants, were actively observed visiting the flowers.

Meiotic behavior

The present observation of n = 39 and 2n= 78 chromosomes indicates a new chromosome number for this species; two earlier reports give x = 13 in Nardostachys (Engel 1976) and 2n = 26 in N. grandiflora (Goldblatt 1984). As per Stebbins (1971), base numbers (x) above 12 are secondary in origin (denoted as X₂) and x=13 in N. grandiflora (Engel 1976) should be denoted as X₂ =13. Family Valerianaceae exhibits a dysploid series of five basic chromosome numbers: x = 15 in American Valerianella; x = 13 in Nardostachys; x = 11 in Patrinia and Valeriana; x = 8 in Cantranthus, Fedia, Valeriana and Valerianella; and x = 7 in Valeriana and Valerianella (Engel 1976). With X₂ = 13, the cytotype studied here appears to be a hexaploid and due to regular meiosis is also a genomic allohexaploid. The present studies as well as earlier reports in this species reveals the existence of at least two cytotypes (i) 2n = 26 (Goldblatt 1984); and (ii) 2n = 78 (present study). This wide variation in chromosome numbers points to the possibility of the existence of other intermediate cytotypes that have evaded detection so far. If these are located, a new source of genetic variability would be opened up in this species.

Meiotic events observed in this species followed a normal pattern which was reflected in high pollen viability of 92.12% and compares well with earlier reports of 81.89% (Chauhan et al. 2008). However, the course of meiosis from telophase-II to pollen formation is eventful. Three type of tetrads – isobilateral, decussate and tetrahedral – were observed within the same flower. Different types of tetrad emerge as a result of successive or simultaneous cytokinesis during microsporogenesis and generally dicots are considered to follow simultaneous cytokinesis post telophase-II stage (Furness and Rudall 2004), but in the present cytotype it appears that some of the pollen mother cells undergo successive cytokinesis as well. Furness and Rudall (2004) also found that different tetrad shapes influence pollen aperture numbers, which may be the reason for the occurrence of bi, tri, tetra and pentacolpate pollen grains observed in this species. Of the different type of pollen grains observed, the tricolpates were far in excess (>73.8%) of other types. This preeminence of tricolpate pollen grains has been explained by Beiatrice et al. (2010) on the basis of tetrad type and cytokinesis pattern in angiosperms. In Clypeola aspera (Brassicaceae) 3- and 4-colpate pollen have been observed in the same individual (Keshavarzi et al. 2012). Beiatrice et al. (2010) demonstrated the role of cytokinesis in pollen aperture pattern ontogeny in Epilobium roseum (Onagraceae) and Paranomous reflexus (Proteaceae).

Increased pollen aperture number offers a potential advantage because it increases the number of prospective germination sites, thus facilitating contact between at least one aperture and the stigmatic surface (Furness and Rudall 2004). It appears that N. grandiflora also has this advantage. Also, pollen grains with more apertures (4-and 5-colpate) were more stainable in comparison to 2- and 3-colpate ones, indicating better fitness potential of such pollen grains. The pollen ovule ratio is also a good indicator of cross ability pattern of any species (Cruden 1977). The pollen ovule ratio (6135 ± 327.86) per flower compares well with the earlier report of 5222.4± 922.63 (Chauhan et al. 2008) and indicates obligate outcrossing (Cruden 1977). This also supports our findings that N. grandiflora favors cross pollination.

Acknowledgements

The study was conducted under All India Co-ordinated Research Project (AICRP) on Medicinal, Aromatic Plants and Betelvine (MAPB) of Indian Council for Agriculture Research (ICAR), New Delhi, and financial support from ICAR is acknowledged.

Disclosure statement

No potential conflict of interest was reported by the authors.