Previous Page Table of Contents Next Page

Germplasm survey, collecting and characterization of bamboo species - Sudhir Kochhar

Senior Scientist (Plant Breeding), National Bureau of Plant Genetic Resources, New Delhi - 110 012. INDIA.


Populations of economic plants comprise intra specific variability held in equilibrium that provides germplasm of interest for conservation and sustainable use. A greater human dependence on a few plant species and particularly some varieties, has resulted in active loss of diversity. As a result, concern for their safety has been shown, both locally and internationally. Exploration, collecting and management of diverse germplasm has, therefore, assumed great significance.

Bamboos represent a sign of household pride in rural areas. They grow and mature fast, have strength, elegance, aesthetics and other such values. This group of plants influences human life at all stages, in all moods and in all walks of the society. For example, cradle for the infant to coffin for the dead, baton for a penalty, helping stick for the handicapped, traditional arms like bow and arrow, agricultural tools for farming, handicrafts, ornaments, pulping for paper making and so on. This paper attempts to provide an understanding of the collecting and characterization of this diverse group of plants with a view to plan a systematic ex situ conservation programme and to classify various diverse types for sustainable use for different purposes.

Germplasm collecting

Exploration and collecting of diverse types constitutes a first step towards a systematic analysis of germplasm. It is important to follow a set of guiding principles while formulating strategy for collecting. This requires a working knowledge of the areas of occurrence/diversity of the plant species, the co-existing species, related wild species/ forms and other genepool. It is also useful to know if some specially adapted forms occur, particularly in marginal habitats. Here, it is important to differentiate between collecting the germplasm and the botanical specimens for floristic or taxonomic studies. Both of these have different objectives, approaches and techniques. The unit of collecting for a botanical specimen may be a taxon, i.e., species, sub-species or botanical variety, whereas the corresponding unit in case of germplasm collecting is a working sample derived from a population of variable phenotypes.

Kinds of exploration: By and large, germplasm collecting missions may be broadly categorized into two types, as follows: i) Specific - which may imply species specific or trait specific, and ii) Broad-based or non-specific - which may aim to collect maximum diversity of the species and relatives in a single trip. Such missions are, however, area/region and season specific.

The germplasm collecting logistics require that the available or theoretical practical knowledge on population sampling should be properly applied. Also, a practical know-how or an overall understanding towards plant as well as environmental diversity is essential, including the socioeconomics and ethnic rituals/culture of the native communities to bioprospect and collect useful types.

Planning for exploration and collecting: It is important to plan well ahead so that favourable results are obtained. A judiciously planned trip should cover the following details:

i) Pre-exploration review, to gather prior knowledge of the topography, eco-habitats, species occurring, flowering cycle of the species under consideration, whether semi-cultivated, semi-wild or wild, associated species in the area, specific diversity to look for, terrain, accessibility, lodging facilities, possible camping sites en route, cultural communities, ethnic groups and their socio-religious customs, the sociopolitical climate and willingness of local administration for survey support, previous survey and collecting trips made in the area and their results, possible local contacts/ resource persons, etc.

ii) Outlining the survey schedule: to fix up the itinerary, site/s for collecting, route of expedition, night halts, tools and equipment to be carried, strategy to be followed for sampling of germplasm and others.

iii) Collaborators and/or local contacts: This requires listing out institutes, agencies, organizations and/or individuals who may collaborate or assist in the area of survey. The local contact persons can effectively help the members of exploration team in getting effective guidance on the nature of diversity in the species, its spatial occurrence in contiguity, the local dialect and customs particularly related to collecting the samples.

iv) Duration of exploration: This should be logically chalked out in relation to the facilities and funds available, number of samples likely to be collected per day and number of days expected to be spent per site, depending upon topography, terrain, frequency of occurrence, expected climatic fluctuations, plant part/s to be sampled, handling of collected samples during the trip, etc. A 10-15 days trip would be ideally suited.

v) The exploration kit: It should invariably include the following:

(a) literature base, such as, regional flora, other reports, road map, etc.,

(b) site identification tools and equipment, such as, altimeter, compass, soil sampling kit, hand lens, binoculars, camera, flash, etc.;

(c) sampling materials, such as, a pair of heavy gloves, seed envelopes, cloth bags, polythene bags, tagging labels, drying sheets, old newspapers, plant press, rubber bands, stapler, cellotape, vasculum, scissors, knife, digging and cutting tools - a machete, cross-cut saw, secateur, pruning shears/saw, pickaxe, crow bar, spade, etc.;

(d) data recording materials, such as field note book, collecting sheets, metre rod, diary, pencils, data logger or pocket tape recorder;

(e) essential medicine kit and first aid box;

(f) other tools and equipment, such as, torch, plastic jars, hunter shoes, camping items, water bottle, etc.; and,

(g) a four wheel drive light vehicle.

Collecting the samples: Four main collecting sites have been described by Hawkes, (1980). Three out of these, namely, i) farmers' fields, ii) home gardens/back yards, and iii) wild habitats are directly applicable to bamboos whereas sampling from markets and farmers' stores has little relevance for this group of plants, apart from fresh bamboo shoots sold in the local market.

Chang gave a suggestive sampling strategy to the germplasm collectors, as follows:

Type of population

Sites to be covered per day

Plants/panicles to be sampled per site

Out breeding



Wild growing



Primitive (Traditional, unimproved farming system)



Slightly improved farming system



Intensive agriculture belt

More frequent

Lesser number

Collector's decision is important in sampling of accessions/specimens and it largely depends upon the pre-collection review, local assistance/help, spatial patterns of variation among individuals at a particular site of collection and within population phenotypic variation. Sampling methods must be so adopted to ensure that an adequate representation of variability is collected. Two basic approaches may be followed:

a) Extensive sampling approach: collecting samples at pre-determined spatial intervals along the survey route, say 5-10 km in a hilly terrain with varying topography/altitude, etc., and a longer distance, say 22-25 km, in relatively uniform eco-edaphically uniform area.

b) Intensive sampling approach: intensive collecting of samples at specific sites showing special situations, for example, if a particular species is flowering, or, for sampling of plants with particular traits.

Further, some of the common procedures followed for sampling invariably involve the following:
i) Random sampling: Collecting one or a few rhizome(s)/propagules from any one or a few of the clumps of the same species growing at the site. Samples may be collected from several spots within the site avoiding deliberate collecting of sturdy/more vigorous types.

ii) Coarse grid sampling: The collector may walk past across the sampling site in a zigzag fashion and pick up samples at random, avoiding sampling from the periphery.

iii) Clustered sampling/fine grid sampling: It involves collecting several samples within a small area showing distinct phenotype/s. The probability of sampling variability related to micro-geographical differences in environment gets enhanced in this way. However, in case of bamboos, when only rhizomes are to be collected, there is a severe limitation to the sample size per species per site. The frequency of sampling and the size of each sample, nevertheless, would depend upon the collectors' spot decisions.

Collecting, packaging and transport of bamboo germplasm samples: Bamboos require unique attention of the collectors as compared with other plant species. First priority is to look for a (sporadic) flowering clump, and collect seeds. Such samples shall be collected in a routine way. Second priority is to collect rhizomes. Their collection has to be done by digging. Precaution must be taken to ascertain that the rhizome is not damaged while extracting. It is preferred to cut 1-2 peripheral culm(s) in the clump selected for sampling, at a height of about 0.5 m from ground level, in case of tall and sturdy species and little shorter for species having smaller and thinner culms. Then, dig around with the help of crowbar and spade so that the whole rhizome mass is exposed. Then cut the rhizome from rest of the underground mass, extract it softly without causing physical injury and level/fill the soil around the clump.

Packaging for short distance transportation may be done in hessian cloth or corrugated sheets. The rhizome mass should be first covered with soil or thick moss and then wrapped with hessian cloth or corrugated sheets depending upon availability. It should be tightly tied at the neck with hessian rope or thin wire. An aluminium foil label inscribing accession number, collector's name and date of collection should be tied with the rope/wire. Same number should also be written with ball pen on exposed portion of the culm to be doubly sure. Packaging for medium to long distances should be done using hard baskets or wooden baton frames, to ensure safe transportation. For short distances, collectors have their own light vehicular transport and for medium to long distances train or air lifting is required. Attempts should be made to carry it as personal luggage in basket packages in train whereas cargo for transportation should be booked under very safe packaging in wooden containers.

The third option, in case appropriate rhizome mass is not available, is to collect 2-3 nodal cuttings with active buds and then make a subsequent attempt to raise nodal plantlets by filling the cavity with 200 ppm boric acid solution (Kochhar and Prasad 1986). Collection of leaves, culm sheaths and single node branches with pruned branchlets should also be sampled for collecting the material for the herbarium and to record further details (Kochar et al. 1990).

Recording the passport data: It is important to record appropriate data at the sampling site. A suitable passport data sheet should be adopted/developed so that most salient passport descriptors are recorded in the field, in a universal/compatible format. NBPGR adopted its suitable passport data sheet from FAO/IBPGR(IPGRI)/Hawkes (1980) which is appended (Appendix I). The details are adjusted to the needs of bamboo. The same format was mainly used during the survey and collecting missions for bamboos in northeast India 1984-1990. It is understood that the collector's data should invariably include, accession identifiers, site identification, including geographical location, species occurrence, habitat, including edaphic data, sampling features, plant traits, custodian community or contact person's particulars, etc. An exhaustive coverage should be given on the original description of the material in the passport data so that it is appropriately represented in the database for future reference and effective, sustainable use.

Population-Genetic principles underlying sampling: Four basic, conceptual classes of populations may be delineated, depending upon frequency and prevalence of particular alleles (Brown and Marshall 1995) as follows:

Class I: Widespread and carrying common allele
Class II: Widespread but carrying rare allele
Class III: Localized but having common allele
Class IV: Localized and also carrying rare allele
The priority for inclusion of samples representing each of these classes may vary. It will also influence the sample size and frequency of sampling in each case. It is understood that in case of Class I populations, random sampling alone should be appropriate. The representation of alleles from Class II populations in the collection can be assured if the sample size at each site is kept big enough. Similarly, to cover Class III variables in the collections made, due coverage to various geographical and ecological sites should be promptly determined. Nevertheless, accessions with localized and rare allele (Class IV) have to be included in the collection through non-random sampling.

In the case of bamboos, however, the sample size is limited to 2-3 rhizomes, depending upon available land area for planting. Therefore, collector's judgment on variables to be sampled or quality of sample being collected plays an important role. The post-collection data should also be recorded for the parent rhizome and culm characteristics so that a comparison can be made at a later date. Quick operations should be ensured in case seed samples or seedlings are collected, also ensuring a minimum time-loss for storage/planting.

Characteristics of bamboos

Gramineae or Poaceae is a large family with many subfamilies (Fig. 1). Bamboos typically depict an understorey in the forest vegetation, representing peculiar ecotypic climax. A bamboo plant is a giant grass. Like grass, it also has culms, branchlets, culm sheaths, thin, elongated leaves as prominent vegetative parts. However, the underground part in bamboos is woody and massive forming the rhizome, which is of two basic types, the caespitose, which gives rise to pachymorph shoots that grow close together with a clump habit, or leptomorph, which runs underground in a reticulate fashion and gives rise to shoots at frequent intervals. This branching habit constitutes one of the primary criteria of classifying bamboos (Fig. 2). They are commonly called sympodial (clump) and monopodial bamboos. Typically, bamboos have woody, tall, cylindrical and jointed single culms clustering together into a clump. Most of the bamboo plant species show this type of growth habit (Fig. 3). However, in certain species the plant habit is different, that is the culm arises vertically from a reticulate underground rhizome, which gives an overall appearance of a clump. Similarly, the culm sheath, the leaf sheath and the ligule are also important marker traits because these are specific to individual species and may vary greatly among different species. It becomes important to understand bamboos in greater detail for their biological and economic features (i.e. end-uses) so that their exploration and collecting is more effectively carried out for conservation and sustainable utilization.

Distribution: Bamboos are worldwide in distribution and occur from tropical lowlands to the high mountains, unlike cereals and other grasses which are restricted to narrow habitats and growing conditions. However, it should be clearly noted that different species or groups of closely related species of bamboos have different adaptability ranges (Kochhar and Rana 1993). Over 60 genera and nearly 1500 species of bamboos have been described. Bambusa Schreber. represents the type genus of bamboos, having nearly 70 described species. It is prevalent in the tropical and sub-tropical Asia, Africa and America. Arundinaria Michaux. has 150 species prevailing in the warm tropics. However, this genus has been critically re-examined and several new genera have been separated. Following this, Phyllostachys Sieb. et Zucc. has a large number of species (40), which cover temperate areas of India, China and Japan. Other more common genera, having 20 species or more, are Dendrocalamus Nees., Dinochloa syn. Melocalamus Benth. Gigantochloa Kurz., Oxytenanthera Munro and Semiarundinaria Makino et Nak. All these genera, except Oxytenanthera, are confined to the Asia Pacific region. A few species of Oxytenanthera are present in Tropical Africa (Kochhar and Rana 1993).

Biology of bamboos: Few systematic studies have been carried out on this aspect and also, few details are published in literature (Gamble 1986; McClure 1966; Kondas 1982; Soderstrom and Young 1983, and others). McClure (1966) categorized bamboos in two distinct groups, namely pachymorph and leptomorph, each requiring different propagation procedures. These groups also have different morphological characteristics with respect to rhizome, form of constituent axes, form of lateral buds, growth habit of individual axis, typical clump habit, culm origin, culm branches (mid-culm) range and transverse venation of leaf blades (Fig. 3). Further, the genera represented in the two groups have shown varying adaptation to temperature. This is, however, to indicate that biology of bamboos may not be typical of the group of plants as such but may vary from species to species.

Bamboos have a clump habit or a runner habit like other grasses and the aerial stems are called culms. The height of a bamboo clump varies from about 1 m to 30 m. The culm habit varies from species to species. e.g. B. tulda has got erect clumps, D. hookerii clumps are semi-spreading and culms arching over whereas the culms of Cephalostachyum fuchsianum are tall and climbing on the tree trunks. The colour varies from dark green to pale green on yellow (B. vulgaris): streaked black on green or yellow on green colour, (Oxytenanthera spp.). There are a few exceptions to the shape of culms, e.g. the culms of Phyllostachys manii and Arundinaria hirsuta are D-shaped particularly at young ages and those of B. quadrangularis are almost square shaped.

The nodes and internodes are distinct in culms (Fig. 3). The internodes are sequentially jointed in a straight line but in a few cases these are somewhat zigzag due to uneven growth of alternate nodes. Further, the internodes may be slender, thicker near the lower nodes or have streaked constrictions. The nodes may be inconspicuous to very prominent or bulged. In some species there are 1-2 nodal rings of white powdery mass or a ring of root scar around nodes. In a few Arundinaria species there is a prominent ring of spines on the nodes. The internodes are hollow but in some cases solid culms are found up to a few metres of culm length e.g. C. fuchsianum. Branching occurs at upper nodes in many species. In Cephalostachyum pergracile the culms are slender for more than 2/3rd of clump height whereas in B. tulda and B. balcooa profuse branching begins from the lower nodes. In Dendrocalamus sp. the branching is so prominent that it almost appears like full culms arising from nodes.

Branching may occur within the culm sheaths or it may break through the base of the sheath, known as intravaginal and extravaginal branching, respectively. Further, branches may arise singly, in pairs or in clusters of 3 or more. Invariably the branching is above the nodal line (Fig. 4). These may arise as primary dominant branch followed by secondary or tertiary orders or on a semi-circle plane called aspidate branching. In a few species like Dendrocalamus membranaceus the branches are spiny whereas in others like B. tulda a single branch in the middle is modified into sickle shaped thorn like branchlet at every node. The branches are in general angled upwards but in a few species these may be horizontal or angled downwards to a shorter length or appressed. The internodes are covered with culm sheaths with varying distances and patterns.

In the early stages of culm-emergence, the entire shoot is covered with closely overlapping, ascending or horizontal blades, which dilate at the stem elongation stage. The culm sheaths are deciduous to persistent and vary in shapes and sizes. A culm sheath has three distinct parts; sheath proper, imperfect blade and a ligule in the inner margin of the joint between these two. A pair of lateral appendages called auricles are also present which constitute marker traits. These further possess hairy appendages or cilia. The sheath blade is sometimes covered with dark brown coloured appressed ciliate mass on its entire dorsal surface or in patches. The culm sheaths vary in form from thin papery membranaceous to thick leathery and glazed on inner surface.

Leaves are compound pinnate, imparipinnate, 7-13 leaflets, dark green to green on dorsal surface and light green to white powdery on the ventral surface. Leaf sheath and ligule vary from species to species and can be used as marker traits. Venation is prominent up to tertiary level but in some cases only secondary and intermediate veins are prominent on either side of the midrib and small dotted glands between intermediate veins give false appearance of tertiary veinlets. The size of leaves varies considerably between the species as well as at different growth stages.

Inflorescences show great variation among species, in general these are compound inflorescence: racemose e.g. B. pallida or cymose (pseudospikelets in head) e.g. D. hamiltonii or Pseudostachyum polymorphum (Fig. 4). Inflorescence emerges on leafy branch terminals or on completely dry culms. The panicles may be closed, spicate or open. The spikelets vary in number but usually they all contain two or more empty glumes; a flowering glume; a palea, keeled or convolute; lodicules, stamens 3-6 or more; and the ovary surmounted by style and stigma. Fruits of bamboo are caryopses and resembling wheat grain very closely, in which pericarp is thin and adnate e.g. B. pallida, B. tulda. In other cases like D. hamiltonii or Pseudostachyum polymorphum the pericarp is thick, leathery and the seed is almost the size of a monkeynut seed or even bigger. In Melocalamus sp. and Melocanna baccifera the seed is a berry and sometimes is equal to the size of an apple.

The underground part of a bamboo plant is mostly the rhizome. The rhizomes are knotty, thick, solid growths, covered with golden papery overlapping sheaths throughout their surface, which form an entangled network running underground, occasionally pushed above ground or specialized as props for culms. These are woody, arched slightly and upturned sharply like the handle of a walking stick. The rhizomes may be narrower at the proximal end called neck and broader at the distal end from where the culm emerges out.

As noted earlier there are two forms of plant habit; pachymorph or the clump habit e.g. B. tulda, B. balcooa, B. pallida or leptomorph i.e. scattered giving rise to one or few culms at one point and then extending as underground runner to a short distance where it again sprouts up into a few more culms at one point and so on. The axial growth of rhizome can vary from determinate to indeterminate habit. The rhizomes of a clump forming bamboo are sympodial with short or long necks whereas the other type is monopodial. In some cases the amphipodial rhizome system is common in which sympodial rhizome clusters may arise on long axial monopodial rhizomes. The roots are present at nodal rings or throughout the surface of the rhizome. These may be extended to a few lower culm nodes. The roots are profuse, coaxial and adventitious. Lateral buds may be dome shaped, boat shaped or even absent. The size of rhizomes vary from a few cm in diameter and length to 30 cm and thickness is more or less proportional to the size of culms.

Growth habit: In the older clump the most active region is the youngest rhizome. The mother and grandmother rhizomes influence the growth and development of the young one directly in a successively diminishing manner. Positive geotropism is reported in the growth behaviour of successive rhizomes whereby the axes get deeper and deeper into the soil (Kondas 1982). In the case of very old clumps the entire rhizome mass rises 1-2 m above the ground level with persistent soil erosion. In such cases the rhizome surface turns green with photosynthetic activity. There is no proper understanding about the effect of rhizome depth on the transformation of root-to-shoot portions. Kondas (1982) reported that the rhizome shows a normal curvature at a particular soil depth, it tends to bend downwards as a sickle before reemerging if the depth is shallow and emerges straight without a curvature if placed deeper due to soil heaping. Culm-to-culm distance also shows corresponding variation due to the varying culm emergence. No conclusive evidence can be recorded unless more species are systematically studied. Rhizomes in older caespitose clumps are seen at times above the ground whereas some bamboo clumps along the road cuts show deeply sunken rhizomes.

Aerial parts of bamboos remain dormant in extreme winter seasons. Growth is active in the spring and rainy seasons. Culm growth follows a sigmoid curve pattern. At the initial stage growth is slow when a conical protrusion is seen above the ground, covered with imbricating sheaths furnished with blades. This stage continues for nearly 15 days subsequent to which maximum culm elongation takes place in next three months. After that a marginal increase in culm height is attained. The shoots thus developed are tender and branchless and mature in 3-4 years. Some deformities like development of clumplets with aerial prop-rooting on the upper nodes or branches are also seen in bamboos. A clear idea about the growth phases of bamboo clump is hardly known except for a few observational reports. However, Seth and Mathauda (1956) reported high clump mortality and low production on complete felling of the clump. They further observed that growth and development depends on the type of felling cycle and total clump age. B. bambos clump reaches full size in 13 years. Annual yield of 250-450 culms or 3.5-5 tonnes per ha is attainable in tall growing bamboos.

Variation: Within species variation has been reported in several species of bamboos (Kondas 1982; Soderstrom and Young 1983; Kochhar et al. 1989). Kondas (1982) suggested more careful survey and collecting trips were necessary to identify useful germplasm with known characteristics. Collection and conservation of variants is essential to conduct further studies. Intra-specific variability has been recorded in B. pallida and D. hamiltonii, conserved at Basar, Assam. Further, variation in character manifestation of species was reported to be influenced by eco-habitats (Kochhar et al. 1989). The authors observed large variation among different characters and combination of characters in the natural populations growing in the mid-hills and the plain valley areas, for three species viz; B. tulda, B. pallida and D. hamiltonii. Such knowledge and practical experience, would help in developing future guidelines. A guide to collecting bamboos (Soderstrom and Young 1983) gives useful hints on the biology of bamboos along with their collection and characterization approaches.

Seedling variations occur due to genetic or developmental causes. Kondas (1982) reported that erect and very erect seedlings are capable of attaining normal culm size (mature early for cutting stage) as compared with grassy and grassy-erect types. Bir Bahadur et al. (1978) reported that the seedlings which show left-handed folding of the first leaf above coleoptile were more vigorous than their right-handed counterparts. They further reported that the ratio of left: right handedness was hereditary even though the individual trait was not so. Kochhar et al. (1989) compared plants raised from seeds with plants obtained from nodal cuttings and observed that the latter were tall and showed longer root length and shorter internodes.

Population behaviour: Kochhar et al. (1989) reported population behaviour of 3 species viz., B. tulda, D. hamiltonii and B. pallida at mid hill and valley land conditions. These species performed better in the plains as compared to mid hill conditions. There was a wide range but low coefficient of variation particularly for height and culm diameter. The culm diameter (girth) and thickness were positively related to internode length and the diameter of hollow was negatively related to the diameter (girth) of culm. At the population level, culms with long internodes were correspondingly thicker, stout and less hollow. Internodal volume is a direct criterion that can be used to determine the growth vigour of culms. Further studies are required to establish genetic and environmental relationships with population biology.


The diverse germplasm collected needs to be evaluated for various qualitative and quantitative traits so that the accessions may be subjectively used. It requires a universal or a generally agreed upon format for data input, units of measurement or ranges for grading information, their rationalization and coding to develop a good database. For this, use of a minimum set of descriptions need to be developed and the descriptors have been advocated as an integral component of a PGR handling system.

Descriptors: The term “Descriptor” relates to a 'unit of information' in a PGR database set to describe features of a germplasm accession, such as its origin, description, and other aspects of conservation management, utilization, etc. It is used in corollary to the term 'Character' or 'Trait' as used by plant breeders to record various qualitative and quantitative data in their programmes. In computer terminology, a descriptor refers to a “Field” which constitutes basic unit of a “Record” that describes a single accession for a particular trait of expression or unit of description in the database. The common or uniform format used produces a universally understood language for the PGR data and helps to establish efficient information storage, retrieval and communication for encoding, transformation and interpretation to assist in effective germplasm utilization.

A world lead for providing descriptor lists on standard format has been undertaken by IPGRI. In finalizing such lists, advice from respective crop experts, national organizations, collection managers, crop advisory committees, throughout the world, was taken and the existing lists examined. Over seventy descriptors on various crops/ group of crops have been published by IPGRI, which aim to encourage collecting of minimal data that should be ideally available for each germplasm entry on Passport, Characterization, Preliminary evaluation and Further evaluation sectors. The characters of concern to plant breeders can be broadly divided into two functionally and genetically distinct groups, viz., observable or qualitative characters and non-observable or quantitative characters. The traits covered under characterization are, in general, qualitative, easy to score and theoretically, need to be scored only once. Further evaluation would cover the data on the quantitative traits. Such details need to be recorded over years, including locations and activities and their average values should be preferably included in the database.

Descriptor states: A “Descriptor state” represents relative state of details of a descriptor. It refers to a set of alternatives for a qualitative (or graded), logical or descriptive record or suggests measurements to record quantitative data specifically to describe a germplasm accession. The lists of minimal descriptors are accompanied by standards for qualitative states, measurement techniques, time and units of measurement for quantitative descriptors and encoding methods. A “Code Dictionary” invariably provides details for interpretation of coded data.

Descriptor can more closely describe a germplasm accession. For example, High (1), Medium (2) or Low (3) and High (1), Moderately High (2), Medium (3), Moderately Low (4) or Low (5) may be used as two alternate sets of descriptor states for a particular descriptor. In case of bamboos, the range of expression for each descriptor state is clearly indicated so as to minimize sampling error and maintain uniformity in data recording. To cite an example for plant height in wheat, it is more commonly known that the Tall and the 1D, 2D and 3D dwarf phenotypes grossly represent >1.25 m, 1.00-1.25 m, 0.75-1.00 m and <0.75 m plant height, respectively. Similarly, in chickpea, the descriptor states for leaflet size may be recorded to express small, medium and big types for the ranges of expression of upto 3 mm, 3-5 mm and >5 mm, respectively. For other temporal or measurable quantitative descriptor states, unit and system for measurement, intensity of colour, etc., need to be clearly specified for coding of data. There is a need to standardize the various quantitative descriptors based upon collective judgement of bamboo workers.

The internationally accepted norms for coding of descriptor states are mostly annexed with the IPGRI germplasm descriptors. These should be followed to arrive at uniform data sets. Since a PGR database is expected to be huge, having a more general application as compared with plant breeding data, the collection and documentation data on germplasm accessions needs rationalization and also standardization of data in terms of terminology and measurement. A more common choice of descriptors in relation to types of measurement data is given below:



Basis of observation


Height, Days to flower, 100- seed weight, Number of pods per plant

Direct measurement of attribute


Harvest index, Per cent oil, protein or sugar

Combination of two direct measures or inference from a single measurement


Assigning a related value from a standard scale

Susceptibility to diseases or pests, Leaf or seed shape


Flower colour, seed pattern, etc.

Assigning qualitative character states into arbitrary classes

As per the existing scheme/s, recording of passport data, and characterization and preliminary evaluation is managed directly by the institutes related to PGR activities whereas the 'Further characterization and evaluation' is invariably recorded by users. Such information should be ideally passed on as a feedback to the database management system to help in a better long term management and use of accessions. Such a stage is, however, far from attainable, in totality, due to changed priorities by the germplasm users.

Use of descriptors in field characterization: Descriptor details should be such that these cover the entire range of annual growth season, without crowding with other details. It is important for effective planning to record data judiciously with efficient utilization of available manpower. Recording of characters like initial vigour, mortality, etc., should be done within a month after planting, emergence of new shoots and rhizomes during main growth initiation periods, which vary from species to species. In observations on bamboos in N.E. Indian collections at Basar, three distinct periods were noted, i.e., Spring (February to March), Rainy season (July-August) and Fall season (October). It is very important to observe rate of growth of individual culms/internodes and average over clumps. This may be done by recording data in the main growing season, at least every week, or even at shorter intervals depending on facilities available. Morphological characters like leaf colour, size, shape, stem colour, surface, etc., should be recorded after the active growth period is over, i.e., after 4-6 months. Data recorded on canopy and phenological traits like fresh sprouting, culm elongation, culm maturity, flowering, clump death, etc., could be recorded over a longer period. Multiple records should be kept on most of the morpho-agronomic quantitative traits and their average values used for interpreting data.


Availability of an elaborate database is essential for a sustainable and effective germplasm utilization. IPGRI descriptor lists and also documentation standards suggested by different PGR systems and institutes facilitate uniform database records. Scientists can document their accessions according to standard terms thereby resulting in compatible databases followed by effective germplasm use of the selected accessions. Data validation for the descriptor states is important and this may be done as per accepted definitions, standards, thesaurus and data dictionary. This ensures the eventual validity of data in terms of permissible limits for item whether it be of a character, numerical, data or logical field. The memo records descriptive, untabulated, text information accompanying the data, which can not be generalized.

Two basic types of database management systems could be identified, viz., hierarchical and relational. A hierarchical system tends to be extremely complex, posing several management difficulties arising due to multiple superior-subordinate relationships between data elements in the hierarchical (tree) structure. In comparison, a relational database is much simpler because of its two-dimensional outlay in which inter-relationship across the tables/data files can be smoothly worked out.

A catalogue is essentially a hard copy of a relational database in which the relationship between two different tables (or correspondingly between two different files in the database) can be established on the basis of common field characters or descriptors, such as, the accession identifiers, viz., Collector number, Accession number, other Identification number, etc. Several DBMS software packages, such as dBASE III Plus, dBASE IV, FOXBASE, FOCUS, ORACLE, UNIFY, INGRESS, SYBASE, QBE, QUEL, RELATE/3000, etc. have been marketed and currently newer versions, such as, Microsoft Access are more popular. However, relational database management vis a vis cataloguing is machine (software) independent, easily maintained and processed and is also globally accepted due to good performance record and security measures for protecting the data during hardware failure, abuse or unauthorized use. dBASE III PLUS or dBASE IV is an appropriate choice for a small to medium-size database vis a vis catalogue and is widely used whereas in order to handle large database, the ORACLE database management system provides a better and powerful package in relational technology.

A model data sheet for indexing and retrieval of bamboo germplasm

A data sheet for indexing and retrieval of bamboo germplasm is presented below. This was prepared based on a brief practical working experience on bamboos. It is desirable that to begin with, a few working descriptors and their descriptor states (grades) be adopted out of various description areas, namely, source data, planting data, phenological parameters, histological characteristics, quality traits and management aspects, which are given below. A sustained and systematic collection of data on the suggested lines and as per the given format could be helpful in ultimately detecting and interpreting diversity and desirable variations in bamboos for further improvement:

A. Source data

1. Institute accession number
2. National accession number
3. Species name
4. Common name
5. Synonym
6. Source/origin
7. Source location, longitude, latitude
8. Source altitude
9. Habitat
10. Topography
11. Soil type
12. Rainfall
13. Frequency
14. Habit
15. Important characteristics
B. Planting Data
1. Date of planting
2. Location
3. Locality number
4. Survival (Yes/No)
5. Date of first culm emergence
6. Rhizome stock (number)
7. Rhizome buds (number)
8. Rhizome size (length x width)
9. Number of culms
10. Culm diameter
11. Culm thickness
12. Internode length
C. Phenological Data
1. Date of observation

2. Growth stage

3. Remarks

Key to growth stages: 1) Dormant, 2) Culm emergence, 3) Juvenile culm 4) Leaf sheath persistence, 5) Young culm attaining maximum average height of clump, 6) Appearance of leaves, 7) Appearance of branches, 8) Flowering initiation, 9) 75% flowering, 10) Seed formation, 11) Seed ripening, 12) Seed maturity and 13) Clump death.
D. Morphological Characteristics
1. Year of observation

2. Month & date of observation

3. Clump circumference (m)

4. Clump diameter (m)

5. Area of clump (sq. m.)

6. Density of clump (grade).

Key to density grades: 1) Very dense, 2) Dense, 3) Medium, 4) Poor, 5) Congested, 6) Scattered.
7. Habit (grade)
Key to clump habit grades: 1) Very erect, 2) Erect, 3) Arching over, 4) Decumbent, 5) Scandant, 6) Grassy.
8. Clump height
- maximum (m)
- average (m)
9. Number of culms/clump
- Juvenile
- 1 year old
- 2 year old
- 3 year old
- Old
- Total.
10. Number of culms cut/clump
- Old (above 3 years)
- Young (1-2 years)
- Juvenile
- Total.
E. Rhizome characteristics
1. Date of observation

2. Habit (grade)

Key to habit grades: 1) Running overground, 2) Running underground 3) Specialized as props for culms.
3. Form (grade)
1) Pachymorph (dense clump habit), 2) Leptomorph (scattered clump habit).
4. Lateral buds (grade)
1) Dome shaped, 2) Boat shaped, 3) Absent.
5. Axis growth (grade)
1) Determinate, 2) Indeterminate.
6. Position of roots (grade)
1) At nodal rings, 2) Throughout the surface.
7. Diameter (cm)
- Near neck
- Near bend
8. Circumference (cm)
- Near neck
- Near bend.
9. Length (cm)
F. Culm characteristics
1. Date of observation

2. Origin (grade)

1) Distal to rhizome, 2) Proximal to rhizome, 3) Lateral to rhizome
3. Colour (grade)
1) Dark green, 2) Green, 3) Light green, 4) Pale green, 5) Striped and 6) Other.
4. Surface (grade)
1) Glabrous, 2) Glaucous, 3) Bristled, 4) Spiny at nodes.
5. Height (m)
- Maximum
- Minimum
- Average
6. Diameter (cm)
- 5th internode
- Maximum
- Minimum
- Average
7. Thickness (cm) 5th internode

8. Culm to culm distance (cm)

9. Internode length (cm)

- 5th Internode
- Maximum
- Minimum
- Average
10. Shape of internode (grade)
1) Slender, 2) Thicker near lower node, 3) Streaked constrictions 4) Ridged surface 5) Other (specify).
11. Shape of node (grade)
1) Level nodal line, 2) Dipping nodal line, 3) Biannular nodal line.
12. Number of nodal rings

13. Number of nodal buds

14. Bud position

1) Above the node, 2) On the node, 3) Below the node.
15. Bud size and shape.

16. Arrangement

1) All in one series, 2) Alternate positions, 3) Others.
G. Branching characteristics
1. Date of observation

2. Origin (grade).

1) At nodal line, 2) Above nodal line, 3) From a specialized overgrowth.
3. Habit (grade)
1) Angled upwards, 2) Level, 3) Angled downwards 4) Appressed 5) Other.
4. Development (grade)
1) Intravaginal, 2) Extravaginal, 3) Both.
5. Shape (grade)
1) Simple, 2) Curved, 3) Constriction at nodal points, 4) Spiny and Thorny.
6. Number of branches

At lower half of culm

- Primary
- Secondary
- Tertiary
At upper half of culm
- Primary
- Secondary
- Tertiary
7. Length of branch (cm)

At lower half of culm

- Primary
- Secondary
- Tertiary
At upper half of culm
- Primary
- Secondary
- Tertiary
Key to Pattern of branching grades: 1) Single, 2) Bunch of same size 3) Bunch of unequal branches, 4) Tufted (Aspidate) and 5) Others.
H. Culm sheath characteristics
1. Date of observation

2. Duration (grade)

1) Persistent, 2) Caducous 3) Deciduous and 4) Others.
3. Texture (grade)
1) Soft, 2) Hard, 3) Leathery, 4) Others.
4. Surface (grade) (upper)
1) Glaucous, 2) Glabrous, 3) Hairy, density, colour, 4) Spinules present.
5. Sheath blade posture (grade) (lower) 1) Glaucous, 2) Glabrous, 3) Hairy present/ absent.
1) erect, 2) Horizontal, 3) Reflexed.
6. Size of sheath blade (cm)
- Length Near node
- Breadth Near sheath blade
- Shape
7. Sheath blade, sheath joint ring (grade)
1) Bulged 2) Prominent, 3) Inconspicuous.
8. Ligule (descriptive)

9. Auricles (descriptive)

10. Appendages (descriptive)

11. Cilia (descriptive)

I. Leaf characteristics
1. Date of observation

2. Colour (grade)

- Dorsal
- Ventral

Dorsal: 1) Very dark green, 2) Dark green, 3) Green, 4) Light green, 5) Pale green, 6) Outgrowths, 7) Others: Ventral: 1) Light green 2) Whitish green 3) White due to powdery surface 4) Outgrowths 5) Others

3. Texture (grade)
1) Leathery, 2) Soft, 3) Brittle.
4. Surface (grade)
1) Glaucous 2) Glabrous, 3) Hairy.
5. Habit (grade)
1) Stiff, 2) Flexuous, 3) Erect 4) Pendant.
6. Size (length x breadth) (cm)
- Lower leaves
- Upper leaves
7. Venation (grade)
1) Prominent, 2) Conspicuous, 3) Inconspicuous
8. Glands (Grade)
1) Prominent, 2) Minute, 3) Inconspicuous, 4) Other.
9. Stalk (Grade)
1) Long (above 1 cm), 2) Medium (0.5 cm) 3) Small (below 0.5 cm), 4) Absent (sessile).
10. Sheath (grade)
1) Laminar, 2) Spinose, 3) Hairy, 4) Ligulose and 5) Absent.
J. Floral characteristics
1. Date of observation

2. Inflorescence (grade)

1) Head Pseudospikelet) 2) Raceme, 3) Spikelet, 4) Closed panicle 5) Spicate panicle 6) Open panicle 7) Incipient
3. Habit (grade)
1) Erect, 2) Arching, 3) Drooping and 4) Scandent.
4. Occurrence (grade)
1) Throughout the clump 2) On the upper leafy branches, 3) Others (specify location)
5. Colour (grade)
1) Green, 2) Straw, 3) Purple, 4) Others.
6. Size of panicle/head (length x breadth) (cm)

7. Size of flower (length x breadth) (cm)

8. Ligule (description)

9. Lemma (description)

10. Palea (description)

11. Anthers (description)

12. Ovary, style and stigma (description)

13. Seed (description)

K. Histological parameters:
1. Stem-wall cuticle (description)
2. Mesophyll, number of layers, thickness
3. Endodermis, number of layers, thickness
4. Intercellular materials (description)
5. Cell wall thickness
6. Intercellular characteristics (description)
7. Cell diameter
8. Length of cell
9. Length of fibre.
L. Wood quality parameters
1. Strength
2. Elasticity
3. Thickness
4. Smoothness
- Outer surface
- Inner surface
M. Pulp quality parameters
1. Fibre length
2. Fibre width
3. Alkali test
4. Acid test
5. Pulp recovery percent
N. Cutting Management
1. Date

2. Number of culms present in clump

- Number of culms cut
- Old
- Below 3 years
- Below 1 year
3. Length of cut culms (average)
- Old
- Below 3 years
- Below 1 year.
4. Weight of cut culms (average)
- Old
- Below 3 years
- Below 1 years.
5. Quality (suitability) (grade).
1) For regeneration through nodal cutting 2) For timber, 3) For handicraft, 4) Splitted culms 5) Others (specify).

In order to enhance studies on germplasm management of bamboos that would accumulate pertinent information and analysis to strengthen conservation and sustainable utilisation of this group of plant species, it is suggested that appropriate periodical data be recorded on a minimal set of descriptors. Analysis and interpretation of such data accumulated over a sufficiently long period of 5-10 years would provide useful and practical information on bamboos. The diversity of characters in bamboos is very vast. The above outline may provide only basic details and detailed notes on other tests or characters should also be built up.


Anon. 1989. Annual Progress Reports 1981-1988 on bamboos for AICRP on UU & UEP, ICAR Complex for NEH Region, AP Centre, Basar (Distt. West Siang).

Arora, R.K. 1991. Plant exploration and germplasm collection. Pp. 55-93 in Plant Genetic Resources - Conservation and Management. (R.S. Paroda and R.K. Arora, eds.). IBPGR Regional Office for South & Southeast Asia, New Delhi.

Bir Bahadur, K. Lokendra Rao and M. Madhusudan Rao. 1978. Left and right handedness in seedlings of Bambusa arundinacea Wild. Curr. Sci. 47(16):584-86.

Brown, A.H.D. 1989. The case for core collections. Pp. 136-156 in The Use of Plant Genetic Resources. (A.H.D. Brown, O.H. Frankel, D.R. Marshall and J.T. Williams, eds.). Cambridge Univ. Press., Cambridge, U.K.

Brown, A.H.D. and D.R. Marshall. 1995. A sampling strategy: theory and practice. Pp. 75- 92 in Collecting Plant Genetic Diversity (L. Guarino, V. Ramanatha Rao and R. Reid, eds.). CAB International in collaboration with IPGRI, in association with FAO/IUCN/UNEP. Wallingford, U.K.

Clark, M., S. Zhang and P. Wendel. 1995. Relationships of the subfamilies of the Poaceae. Syst. Bot. 20(4):436-460

Gamble, J.S. 1986. The Bambusae of British India. Annuals of Royal Botanical Garden, Calcutta, Vol. VII. PP 133 + 119 plates.

Hawkes, J. G. 1976. Manual for Field Collectors (seed crops). IBPGR, FAO, Rome, Italy.

Hawkes, J.G. 1980. Crop Genetic Resources - A Field Collection Manual. IBPGR/EUCARPIA. Univ. of Birmingham, U.K.

Kochhar, S. 1986. Some aspects of standardization of technology for bamboo preservation, cultivation and utilization in NEH region. Approach paper for AICRP on UU & UEP. Presented at III Annual workshop, NBPGR, N. Delhi, June, 6-7.

Kochhar, S. 1986. Various steps in bamboo germplasm studies and their interrelation. Presented at Seminar on Bamboos and Canes, NEC, Shillong. Nov. 26-27.

Kochhar, S. and R.N. Prasad. 1986. Standardization of technology for bamboo preservation and cultivation in NEH region. I Experimental approach. Presented at Seminar on Bamboos and Canes, NEC, Shillong. Nov. 26-27.

Kochhar, S., B. Mal and R.G. Chaudhary. 1990. Population aspect of the phenological behaviour of bamboo germplasm. Pp. 51-58 in Proc. III Intl Bamboo workshop. (I.V. Ramanuja Rao, R. Gnanaharan and C.B. Sastry, eds.). Bamboos: Current Research KFRI, India and IDRC, Canada.

Kochhar, S., R.N. Prasad, B. Mal, R.G. Chaudhary, M. Rai and B.D. Sharma. 1990. Bamboo germplasm - Diversity and conservation in northeast India. Indian J. Pl Genet. Resources 3(2):21-26.

Kochhar, S. and R.S. Rana. 1993. Bamboo germplasm resources of India - their documentation and conservation. Ind. J. Pl. Genet. Resources. 6(2):93-107.

Kochhar, S., R.S. Paroda, B. Mal, R.G. Chaudhary and M. Rai. 1994. Chemically induced variations in bamboo (Bambusa pallida). Ind. J. Agric. Sci. 64(11):775-778.

Kochhar, S., B.K. Sarma, B. Mal and R.K. Mahajan. 1994. Morpho-physiological studies on bamboo germplasm. I. Variability and selection. Indian J. Hill Fmg. 7(2): 176-182.

Kochhar, S. 1997. Concept of descriptors and descriptor states and its application particularly in field characterization. Lecture delivered at NBPGR-FAO-IPGRI Regional Training on germplasm conservation and utilization. IPGRI South Asia Office, NBPGR, N. Delhi.

Kondas, S. 1982. Biology of two Indian bamboos, the culm potential and problems of cultivation Ind. Forester. 108(3):179-83.

Marshall, D.R. and A.H.D. Brown. 1975. Optimum sampling strategies in genetic conservation. Pp. 53-80 in Genetic Resources for Today and Tomorrow. (O.H. Frankel and J.G. Hawkes, eds.). Cambridge Uni. Press, Cambridge.

McClure, F.A. 1966. The Bamboos: A Fresh Perspective. Harvard Univ. Press. Cambridge.

Seth, S.K. and G.S. Mathauda. 1956. The All India Co-operative investigation (1934- 48) on the management of Bamboo (Dendrocalamus strictus Nees. Proc. IX. All India Silviculture conference. Dehradun c.f. Annotated Bibliography on bamboos (Suppl.) 159.

Soderstrom, T.R. and S.M. Young. 1983. A guide to collecting bamboos. Ann. Missouri Bot. Gard. 70:128-36.

Fig. 1. Different subfamilies of the family Poaceae.

Fig. 2. Rhizomes, branching.

Fig. 3. Culm, branching and branch leaves.

Fig. 3. Buds and different types of branching.

Fig. 4. Inflorescences.

Appendix I

A standard format for passport data sheet (adopted for Bamboos)
[as adopted by NBPGR from FAO, IPGRI and Hawkes (1980)]























SOURCE: 1. Natural wild 2. Disturbed wild 3. Plantation 4. Village groves 5. Private garden

STATUS: 1. Wild 2. Weedy 3. Cultivated

FREQUENCY: 1. Abundant 2. Frequent 3. Occasional 4. Rare

MATERIAL: 1. Rhizomes 2. Culms 3. Inflorescence 4. Roots 5. Live Plants 6. Herbarium 7. Others

SAMPLE TYPE: 1. Population 2. Few plants 3. Individual plant

SAMPLE METHOD: 1. Bulk 2. Random 3. Selective (non random) 4................

HABITAT: 1. Forest 2. Disturbed forest 3. Partly disturbed 4................

PLANT CHARACTERISTICS: ...........................................................................

ADDITIONAL NOTES: .......................................................................................

USES: ..................................................................................................................


1. Irrigated .......... 2. Rainfed .......... 3. Arid ..........

4. Wet............ 5.............

SEASON: 1. Spring-Summer 2. Autumn 3. Winter


SOIL COLOUR: 1. Black 2. Yellow 3. Red 4. Brown 5................

TEXTURE: 1. Sandy 2. Sandy loam 3. Loam 4. Silt loam 5. Clay 6. Clay loam 7. Silt 8................

STONINESS: 1. Stony 2. Pulverized 3................

DRAINAGE: 1. Well drained 2. Poorly drained 3................

pH.: 1. Normal ( ) 2. Alkaline ( ) 3. Acidic ( )


ASPECT: 1. Level 2. Crest 3. Escarpment 4. Rounded summit 5. Upper summit 6................

SLOPE: 1. Mild slope 2. Lower slope 3. Open depression 4. Closed depression 5. Terrace

TOPOGRAPHY: 1. Swamp 2. Flood plain 3. Level 4. Undulating 5. Hilly dissected 6. Steeply dissected 7. Mountainous 8. Valley

PLANT COMMUNITY ASSOCIATED: ................................................................

MAJOR SPECIES ASSOCIATED: ......................................................................

NAME OF COLLECTOR: ....................................................................................

COLLECTION NO. AND OTHER DETAILS: .....................................................

Previous Page Top of Page Next Page