Introduction

Bamboos, woody perennials of the grass family Poaceae, provide livelihood for millions of people around the globe. India is well endowed with 136 species of bamboo resources in 23 genera, extending over 13.96 million ha and is the second largest bamboo reserve in the world (FSI 2011). Bamboo is highly adaptable to a wide range of climatic and rainfall conditions and is one of the most suitable species for commercial forestry. Along with traditional uses of bamboo poles for rural building construction, fencing, agricultural purposes and household articles, in recent times, the ‘green gold’ has got immense demand as an industrial raw material (Khan and Hazra 2007). The multiple harvests possible around the year along with tremendous growth rate contribute towards wide-scale acceptability and commercial potential of bamboo species among the farming community. In India, the last few decades have witnessed a considerable interest in bamboo cultivation owing to the huge demand of source material from bamboo-based industries.

Even though India possesses 45% of global bamboo growth, the country contributes only 4% towards the global trade with an annual productivity of only 1 MT/ha (Tripathi et al. 2006). Overexploitation of available resources, poor performance of unsuitable species, forest fires, grazing, gregarious flowering, among others contributed to productivity decline albeit the existence of significant species diversity/growing habitats of bamboos in India (NBM 2013). National Bamboo Mission (NBM), Government of India, is trying to achieve a global share of 27% through improved productivity of bamboo plantations (Kumar et al. 2005). In this regard, NBM has recommended priority species based on their annual productivity and suitability in different agro-climatic zones throughout the country (https://www.nbm.nic.in). Guidelines issued by NBM aim at ensuring the use of superior planting stock for raising bamboo plantations through a network of certified high-tech bamboo nurseries in India. There are more than thousand nurseries in the country which provide planting stock to farmers for establishment of plantations. The crucial step in any productivity improvement program is proper identification, multiplication and supply of high-quality planting materials of suitable species with guaranteed productivity. It is a common observation, however, that plants being propagated in these nurseries are misidentified or that mixing of species occurs inadvertently. It is imperative to ensure correct species identity along with quality planting stock so as to make certain that it is suitable for a given site or purpose.

Unlike other plants, most bamboo species are semelparous, resulting in unpredictable flowering at long cycles after which all plants in the population die. Field identification of bamboo species thus traditionally rely on morphological characteristics of vegetative parts, mainly culm/culm sheath characteristics (Ohrnberger and Georrings 1986; Clark et al. 2007). Generally, in bamboo species, these characteristics manifest only when new culms emerge and juvenile plants do not possess these characteristics until after the 1st year or even later. Thus, identification of species is rendered difficult even for the best of nursery men. Further compounding the problem is the presence of variation between accessions within same species. Planting materials of bamboos handled in a typical bamboo nursery consist of plants in various age classes derived from seeds, macroproliferated seedlings (Kumar 1991), rooted culm or branch cuttings, rhizome transplants and micropropagated plantlets. Due to the absence of any consistent vegetative characteristics, unintentional mixing of species/clones is quite common at the nursery level. Therefore, to ensure that material used for planting conforms to the intended quality, NBM envisages a certification framework under which planting materials are characterized using various state-of-the-art tools (BTSG 2014). Methods to identify plant material at clonal level as well as at species level are equally important. Ideally these methods should be of the kind that is applicable to planting material at all stages of propagation, plantation and the harvested products. Morphology-based identification keys are very useful for quick identification at field, but it is well known that environmental plasticity in these characters is a serious limiting factor that prevents precise identification. DNA-based molecular tools are devoid of such environmental or developmental influences and can bring in more precision to protect the genuine interests of bamboo growers and industries.

The tremendous advancement of molecular marker technologies holds promise to address this issue and yet only limited progress has been achieved with regard to traceability systems in forest reproductive material employing molecular markers (Botta et al. 2001, 2004; Konnert and Behm 2006; Degen et al. 2010). The last decade had witnessed unraveling of potential of molecular techniques as a tool to supplement species identification methods in plants as well as in animals. DNA barcoding, process of species identification based on short standard conserved region of the genome (Hebert et al. 2004; CBOL 2009) is foremost when it comes to identification at species level and to resolve taxonomic problems. Plastid gene sequences such as rbcL, matK, rpS4, rpL16, among others have contributed immensely to the current understanding of bamboo systematics and phylogeny (Kelchner and Clark 1997). However, the feasibility of recommended conserved plastid barcode regions for species discrimination in bamboos has been reported only in a few instances. In temperate woody bamboos, four barcoding loci, namely matK, rbcL, psbA-trnH and ITS2 were analyzed and the combination of rbcL + ITS2 suggested as a potential barcode region for species discrimination (Cai et al. 2012). Failure of matK to discriminate Bambusa species due to interspecific hybridization and polyploidy was recently reported (Das et al. 2013). Low discriminatory power of core barcode region (rbcL + matK) as well as greater discriminatory power of trnG-trnT spacer in bamboos was also suggested (Zhang et al. 2013). Sosa et al. (2013) recommended matK + psbI-psbK as discriminant barcode loci in some temperate bamboos.

The present study therefore envisages developing DNA barcodes for commercially important priority species of bamboos recommended by NBM, Government of India for large-scale multiplication in accredited bamboo nurseries.

Materials and methods

Plant sampling

Leaf samples were collected from mature clumps of bamboo species from reserve forests and protected areas throughout the distribution zones in India (Table 1). For the preparation of voucher specimens, twigs with a few leaves and culm sheaths were collected. Thirteen commercially important bamboo species, viz., Bambusa balcooa Roxb., B. bambos Voss, B. nutans Wall ex Munro, B. pallida Munro, B. tulda Roxb, B. vulgaris var. vulgaris Schrad ex Wendle, Dendrocalamus asper Baker ex Heyne, D. giganteus Munro, D. hamiltonii Nees and Arn ex Munro, D. strictus Nees, Melocanna baccifera Kurz., Ochlandra travancorica Benth and Oxytenanthera parvifolia Brandis ex Gamble were selected for the development of DNA barcodes. Leaf samples from multiple accessions were collected and dried in silica gel. In addition to NBM listed priority species, several species distributed in Northeastern India such as those of genera Bambusa (B. jaintiana, B. mohanramii, B. teres, B. multiplex), Dendrocalamus (D. brandisii, D. hookeri, D. longispathus), Oxytenanthera (O. monadelpha) as well as Melocanna clarkei, which are generally in cultivation were also considered for the development of DNA barcodes. Samples were authenticated at Kerala Forest Research Institute (KFRI) and voucher specimens were deposited at KFRI herbarium.

Table 1 Details on bamboo species and their distribution included in the present study
Full size table

DNA extraction, PCR amplification and sequencing

Total genomic DNA was extracted from either fresh or silica dried leaves using modified cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle 1990) as well as using DNeasy Plant Mini Kit (Qiagen, Germany) according to the manufacturer’s protocol. Seven candidate barcode loci of the plastid genome (four coding regions such as matK, rbcL, rpoB, rpoC1 and three intergenic spacers namely psbA-trnH, psbK-psbI and atpF-atpH were evaluated to identify discriminant DNA barcodes for bamboo species. Primer details and reaction conditions standardized for DNA amplification through polymerase chain reaction (PCR) are listed in Table S1.

Amplification of genomic DNA was performed in a PTC-100 thermocycler (BIO-RAD, India) in a final volume of 20 μL reaction mixture containing 50–100 ng DNA, 10X Taq buffer with 1.5 mM MgCl2, 200 μM dNTPs, 10 pm of each primer, and 2U Taq DNA polymerase (Invitrogen, Bangalore). The amplified products were resolved in 2 % agarose gel and documented using a gel documentation system (Syngene, UK). PCR products were further purified using a Nucleospin Gel and PCR Clean-up kit (Macherey Nagel, USA) and quantified using Nanodrop (Thermo Scientific, USA). Sequencing was performed using Sanger dideoxy chemistry in both forward and reverse directions (Chromous, Bangalore).

Sequence analysis

Chromatograms were edited and trimmed using BioEdit software (Hall 1999). Edited sequences were aligned using ClustalX of Clustal W packages (Thompson et al. 1994) and submitted to GenBank (https://www.ncbi.nlm.nih.gov/genbank/) as well as BOLD https://www.barcodinglife.org.

For pairwise genetic distance (PWG) method, interspecific as well as intraspecific genetic distances were determined by MEGA v.6.0 using Kimura two-parameter distance model (K2P) adopting complete deletion option (Tamura et al. 2013). The interspecific divergence between species was calculated using three parameters: (1) average interspecific distance, (2) average theta prime (θ') and (3) minimum interspecific distances. Intraspecific parameters such as (4) average intraspecific distance, (5) theta (θ) and (6) coalescent depth were also calculated to characterize intraspecific divergences (Meyer and Paulay 2005). Barcoding gap was calculated by plotting intraspecific distances against interspecific divergences for each species (Meier et al. 2006). A blind sampling test was performed with twelve samples of unknown identity to check the efficiency of the selected barcode region in discriminating bamboo species.

In the tree-based analysis, neighbor-joining (NJ) trees were constructed for the studied bamboo taxa using most discriminant psbA-trnH spacer barcode region, adopting K2P parameter in MEGA v.6.0. Bootstrap support was estimated with 1000 heuristic replicates (100 random addition cycles per replicate, with tree bisection reconnection and branch swapping) to test the reliability of inferred phylograms. All positions containing gaps and missing data were eliminated from dataset (complete deletion option).

Results

DNA barcode amplification, sequencing and alignment

An ideal DNA barcode is expected to have adequate conserved regions, high PCR amplification efficiency, and enough variability for species identification (CBOL Plant Working Group 2009). All the evaluated DNA barcode regions (matK, rbcL, rpoC1, rpoB, psbK-psbI, atpF-atpH and psbA-trnH) were successfully amplified with 100 % PCR efficiency using CBOL (2009) recommended primers (Fig. S1). The edited sequences after homology searches were deposited in the GenBank and accession numbers were provided (Table 2).

Table 2 GenBank accession numbers and barcode regions of Dendrocalamus, Bambusa, Melocanna, Oxytenanthera genera
Full size table

Six of the DNA barcode regions, viz., rbcL, matK, rpoB, rpoC1, psbK-psbI and atpF-atpH displayed exactly identical sequences in all analyzed species of Bambusa, Dendrocalamus, Oxytenanthera as well as Melocanna and Ochlandra. Hence, these DNA barcode regions cannot be useful for species certification of planting materials in bamboos. Multiple sequence alignment (MSA) of psbA-trnH intergenic spacer barcode region showed species-specific nucleotide differences in most studied bamboo taxa, viz., Bambusa, Dendrocalamus, Melocanna and Ochlandra. The blind sampling test carried out proved the effectiveness of this barcode region. Two specimens from each of the six bamboo species (Bambusa tulda, B. vulgaris, B. bambos, Dendrocalamus strictus, D. stocksii, Melocanna baccifera) randomly collected by a third party from a bamboo nursery maintained at KFRI, Peechi Campus were correctly identified to the species level. Additionally, the intergenic spacer psbK-psbI also showed species discrimination in the genus Melocanna.

Sequence length and basic sequence statistics like conserved sites, variable sites, singletons and informative sites of psbA-trnH spacer region based on CLUSTALX alignment as well as with alignment explorer in MEGA v.6.0 are provided in Table S2. The maximum number of variable sites among species was in genus Bambusa (23) followed by Dendrocalamus (12) and the least was in genus Melocanna (8).

Sequence analysis of Bambusa

Multiple sequence alignment (MSA) of psbA-trnH barcode showed nucleotide differences unique to species in most cases (Fig. 1). The major nucleotide changes were transitions/transversions as well as insertions/deletions of nucleotides in the analyzed barcode region. In most species, deletions of mononucleotide thymine repeats in various numbers were obvious. Among ten species of the genus Bambusa, B. multiplex had maximum level of unique nucleotide changes. B. balcooa/B. vulgaris/B. pallida as well as B. nutans/B. teres shared the type of deletion of thymine mononucleotides along with other specific nucleotide changes. B. balcooa, B. tulda and B. vulgaris had a similar transition event (G>A). Similarly, B. nutans, B. tulda, B. vulgaris, B. multiplex and B. jaintiana had a similar transition event (C>T) (Table 3).

Fig. 1
figure 1

Multiple sequence alignment of psbA-trnH barcode region in the genus Bambusa

Full size image
Table 3 Nucleotide differences in psbA-trnH spacer region in the genus Bambusa
Full size table

In distance-based analysis using psbA-trnH, basic statistical parameters (average interspecific distance, theta prime and minimum interspecific distance) were employed to characterize interspecific divergence. The intraspecific variations were calculated by employing average intraspecific distance, mean theta and coalescent depth (Table 4). Average interspecific distance was 0.0509 and DNA barcoding gap was 0.0485 for the genus Bambusa. Even though species-specific nucleotide differences could be identified for each Bambusa species, differences were located in the non-coding intergenic spacer regions or in mononucleotide repeats.

Table 4 Genetic divergence parameters in the genus Bambusa, Dendrocalamus and Oxytenanthera using MEGA v 6.0
Full size table

Sequence analysis of Dendrocalamus

Out of the seven analyzed barcode regions, only psbA-trnH showed species-specific nucleotide differences in the genus Dendrocalamus (Fig. 2). Basic statistical parameters used to characterize interspecific and intraspecific distances are provided in Table 4. The average interspecific distance was 0.0267 and DNA barcoding gap was 0.0264 for the genus Dendrocalamus.

Fig. 2
figure 2

Multiple sequence alignment of psbA-trnH barcode region in the genus Dendrocalamus

Full size image

Unlike in Bambusa, Dendrocalamus species had a lower number of nucleotide changes and also shared some of the nucleotide changes. Both transitional and transversional nucleotide changes were present only in D. hookeri in addition to two major deletions. Transversion of G > C was observed both in D. hookeri and D. brandisii. An inversion of GTA nucleotides was specifically observed in psbA-trnH sequence of D. stocksii. Both D. asper and D. longispathus showed only a deletion of T mononucleotide (Table 5).

Table 5 Nucleotide differences in psbA-trnH spacer region in the genus Dendrocalamus using MEGA v 6.0
Full size table

Sequence analysis of Melocanna and Ochlandra

Among genus Melocanna and Ochlandra of subtribe Melocanninae, only psbA-trnH and atpF-atpH spacer regions showed nucleotide differences which were species specific out of the seven analyzed barcode regions. Both M. baccifera and M. clarkei differed in terms of C>T and T>C transitions and G>C transversion in their psbA-trnH sequences. Additionally, M. baccifera had two specific deletions of GTATTG and TTATTTT sequences (Fig. 3). psbA-trnH sequence of O. travancorica was more similar to M. clarkei than M. baccifera. Both O. travancorica and M. clarkei shared G nucleotide at two sites, whereas it had undergone a transversional change (G>C) in M. baccifera. Similarly, a cytosine and an adenine nucleotide present in O. travancorica and M. clarkei, respectively, had undergone transitional changes in M. baccifera (C>T and A>G). On the contrary, both O. travancorica and M. baccifera shared three major deletions such as GTATTTG, ATT and GTGGGTATTTTTTTTTT (Fig. 3). Even though, two genera shared many nucleotide changes among them, each of them had unique species-specific nucleotide changes as well. Statistical parameters were employed to characterize interspecific divergence and average intraspecific distances (Table 6).

Fig. 3
figure 3

Multiple sequence alignment of psbA-trnH barcode region in the genera Ochlandra and Melocanna

Full size image
Table 6 Genetic divergence parameters in the genus Melocanna using MEGA v 6.0
Full size table

Sequence analysis of Oxytenanthera

Among seven analyzed barcode regions, only psbA-trnH barcode showed species-specific nucleotide differences in the genus Oxytenanthera. Species-specific differences were mostly in mononucleotide repeats of psbA-trnH, which consist of three base pair and single base pair deletions, respectively, in O. monadelpha and O. parvifolia (Fig. 4). Statistical parameters to characterize interspecific/intraspecific distances were employed (Table 4).

Fig. 4
figure 4

Multiple sequence alignment of psbA-trnH barcode region in the genus Oxytenanthera

Full size image

Tree-based analysis

In the tree-based analysis, ten species of the genus Bambusa showed species-specific clustering with two major groups except for B. balcooa and B. teres (Fig. 5). Similarly, out of the seven species in genus Dendrocalamus, three species such as D. asper, D. giganteus and D. longispathus were clustered together as a complex, while the remaining species formed monophyletic clusters with greater than 60 % bootstrap support (Fig. S2). Three taxa of Melocanninae subtribe such as M. clarkei, M. baccifera and O. travancorica formed monophyletic clusters with 64 %, 96 % and 99 % bootstrap support values, respectively, thus psbA-trnH spacer region can be effectively utilized as a discriminant barcode for these species (Fig. S3). Similarly, O. parviflora and O. monadelpha grouped into two separate monophyletic clusters with 100 % bootstrap support (Fig. S4).

Fig. 5
figure 5

Neighbor-joining tree of selected Bambusa species

Full size image

Discussion

Bamboo is very fast growing with 20 times more yield than any other timber tree species and is the most preferred industrial raw material (https://life.gaiam.com/article/how-eco-friendly-bamboo). Over the last few decades, commercial bamboo plantations in India have significantly increased and the planters are really concerned about reliability and identity of species/clonal material multiplied in nurseries all over India. Among the hurdles that nursery managers face in producing quality planting material, the difficulty in precise identification of species is particularly vexing. A further confounding problem is the relative lack of discriminatory morphological features in juvenile plants maintained in nurseries. The misidentification of species suitable for different agro-climatic zones can lead to a significant decline in the productivity (Sharma 2008). It can also lead to significant loss to farmers if wrong species identity is discovered much later especially when the species is unsuitable for a specific end use for which it was grown. This has necessitated the need for a certification agency at national level and with Bamboo Technical Support Group (BTSG), Kerala Forest Research Institute (KFRI) proposed to National Bamboo Mission (NBM), Government of India, a certification framework and guidelines and recommends an integrated approach which includes DNA barcoding for precise species identification (BTSG 2014). A species-specific DNA barcode thus can serve as a valid certification tool to ensure species identity and productivity in the commercial bamboo plantations.

Out of the seven recommended DNA barcode regions evaluated, six DNA barcodes were substantially the same across all species of genera Bambusa, Dendrocalamus, Ochlandra and Melocanna. Because of low ability for species discrimination, most working groups had suggested the use of rbcL in conjunction with other gene regions (Chase et al. 2007; Hollingsworth et al. 2009). Similarly, matK had proved its utility as a potential barcode in closely related groups, such as Compsoneura (Newmaster et al. 2007), orchids (Lahaye et al. 2008), sedges (Starr et al. 2009) and Acacia (Newmaster and Ragupathy 2009), but universality of this barcode region remains uncertain in various taxa. In this study, rbcL and matK could not differentiate bamboo species. Even though rpoB and rpoC1 has been recommended as suitable barcodes, due to low interspecific divergence, these barcode regions were reported as inappropriate supplementary barcode loci (Lahaye et al. 2008). The present barcode analysis also revealed low discriminatory power of rpoB and rpoC1. In addition to the candidate barcode regions described above, other plastid barcoding regions such as atpF-atpH, psbK-psbI and trnT-trnL were also recommended for species identification (Taberlet et al. 2007). psbA-trnH has the potentiality as a suitable marker for species discrimination between closely related taxa due to high rate of sequence variation present generally in this intergenic spacer region (Kress and Erickson 2007; Newmaster et al. 2007). It has also been recommended as one of the best performing loci for various taxa in terms of PCR amplification success, sequencing and species resolution (Lahaye et al. 2008). In the present analysis, species-specific nucleotide differences were observed in psbA-trnH barcode region of genera, Bambusa and Dendrocalamus, Ochlandra and Melocanna. Thus, psbA-trnH can serve as a DNA barcode region for species identification of various bamboo taxa taken up for this study.

Bambusa balcooa and B. vulgaris shared the same type of deletion in thymine mononucleotides. These species are widely cultivated and morphological features are greatly influenced by selection process. Both the species are different in the absence of transverse veinlets in lemma with ovate oblong lodicules in the former and presence of transverse veinlets in lemma with narrowly oblong lodicules in the latter. B. vulgaris formed a very distinct cluster in tree-based analysis. Likewise, B. balcooa and B. tulda had a similar transversion event (G > A). In B. balcooa and B. tulda, inflorescence is clustered at nodes and glumes persistent and shorter than spikelet. In NJ tree, B. balcooa, B. tulda and B. teres grouped together into a complex. Both B. teres and B. balcooa are arborescent densely tufted clump-forming species with glabrous culm sheaths. Even though B. teres is distinct in having glabrous culm sheath proper, similar auricles erect at top of sheath proper, long ciliate ligule having white hair underneath blade, and dense hair at incurved leaf apex, Majumdar (1989) treated B. teres under synonym of B. tulda. B. tulda is a widely distributed species and morphologically highly variable in vegetative and reproductive characters.

Morphologically distinct species like B. multiplex, B. mohanramii, B. jaintiana as well as B. pallida displayed unique barcodes and also showed distinct species-specific clusters in the derived NJ tree. In mature state, B. pallida is a quite distinct species with triangular culm sheath with sheath proper truncately cut at top, long imperfect blade as broad as top of sheath proper and lanceolate spikelets with 3–8 fertile florets. B. jaintiana is a shrubby erect bamboo found in loose clumps. Culms are green becoming orange with age and while young ones are white powdery. B. mohanramii differs remarkably in other vegetative and floral aspects like culm sheath short than internodes, auricles with short rounded shape with somewhat matching culm sheath to B. balcooa (Kumari and Singh 2009). Bambusa multiplex is a morphologically variable widely cultivated perennial species with slender and erect woody culms, nodal roots, bractiferous inflorescence and caryopsis fruit. Among ten species of the genus, B. multiplex had maximum level of unique nucleotide changes.

Dendrocalamus brandisii and D. giganteus had same type of GTTTTT nucleotide deletions and both of which have morphologically prominent auricles in culm sheath and its sheath is glabrous. Some unique nucleotide changes present in D. stocksii such as inversion of GTA to ATG and insertion of AA nucleotides are absent in other Dendrocalamus species. D. stocksii was initially known as Oxytenanthera stocksii which was shifted later on to genus Dendrocalamus based on similar morphological features such as basal nodes bearing aerial roots, erect culms and short internodes, large panicle of spicate heads, keeled palea, among others (Kumar et al. 2004). D. brandisii with its unique nucleotide changes is a morphologically distinct species with its mature culm smooth ashy-gray to greenish-gray colored, loosely spaced and thornless which formed a well-defined clade in the tree-based analysis. D. asper, D. giganteus and D. longispathus have morphologically distinct characteristics but they grouped together in the phylogenetic tree generated.

So far, only universal molecular markers were used for species/cultivar certification of tree species. For example, RAPD markers were employed for certification of Populus species (Sanchez et al. 1998), ISSR markers for Picea species (Nkongolo et al. 2005), Eucalyptus species (Balasaravanan et al. 2006), certification of lupine cultivars (Nam et al. 2014) as well as for genetic fidelity testing in Saccharum officinarum (Thorat et al. 2018). A reliable and affordable certification tool based on SSRs was reported for the certification of chestnut varieties to prevent its commercial misuse (Botta et al. 2001), commercial cultivars of Populus (de-Lucas et al. 2007), characterization of olive cultivars (Muzzalupo et al. 2009), registration and certification of planting materials in Eucalyptus (Torres-Dini et al. 2011), to differentiate Chilean and foreign commercial rice varieties (Becerra et al. 2015), discrimination of Panax species/cultivars (Jo et al. 2016) and for the certification of Albania olive (Muzzalupo et al. 2018).

Forest certification schemes, state agencies such as customs offices, forest enterprises producing timber have relied on molecular methods to improve the traceability of timber and offering opportunities to identify false declarations of timber origin (Finkeldey et al. 2010). Even though, the necessity of a viable molecular method for the certification of planting material through vegetative propagation has been discussed and suggested by various research groups (Alvarez et al. 2001; Rajora and Rahman 2003; Fossati et al. 2005), no reports are available on actual use and implementation of DNA barcode tool for certification of vegetative propagation material.

Conclusion

A molecular tool that is not influenced by age, phenological and physiological status, is useful for species certification of planting stock produced in bamboo nurseries. This study could demonstrate the efficiency of DNA barcoding as a reliable supplementary tool in an integrated approach for the proposed certification system in bamboos. psbA-trnH DNA barcode region can thus be utilized to authenticate species identity against a database of barcodes of bamboo species generated for this purpose. The tool can be integrated into the framework to comply with requirement for confirming species identity under the certification program and give farmers an assurance of the quality intended in label issued by accredited nurseries. To achieve assured productivity in bamboo species, any national certification agency set up for this purpose can utilize psbA-trnH DNA barcode region to tag species identity and to prove the authenticity of multiplied planting materials in all NBM recommended priority bamboo species.