Genetic Diversity and Breeding Strategies of the Neem (Azadirachta Indica)

S. K. Kundu[1] and O. Luukkanen


The neem tree (Azadirachta indica) is an important multipurpose species with enormous potential. Rural people can easily get economic benefits through production of seeds and leaves within a short period of time. It is necessary to popularise the cultivation of neem to receive benefits and economic advantages. Conservation of the genetic resources of this species is essential for the improvement of its genetic quality and adaptability in ex situ and in situ conditions. The consequences of inbreeding and special adaptation to avoid inbreeding depression in this tree are discussed. A long-term breeding strategy is suggested.


The Neem (Azadirachta indica A Juss.) (Meliaceae) is a multipurpose tree native to Myanmar, the Indian subcontinent, and some parts of South-East Asia. For centuries, derivatives of the neem tree have been used in agriculture, health, medicine, cosmetics and livestock production. The recent surge of interest in this tree is due to the spectacular biological activity of ingredients found in the bark, leaves and seeds against a wide range of insects and fungal pathogens. Neem improvement requires understanding of the genetic nature of wild populations, their development and utility of genetic variability. The effective exploration, identification, documentation and use of genetic resources of neem are imperative for its efficient use.

Breeding neem for bioactive compounds or other multiple uses has not yet started systematically. A program on provenance trials for the International Neem Network has initiated the evaluation and improvement of genetic resources of the species in 1996, co-ordinated by the FAO (Anon 1998). Breeding objectives must be set on the basis of present and future needs. Thereafter, short, medium, and long term breeding programs need to be established. The ongoing international provenance trials would provide a great opportunity to select the best provenances for further improvement and domestication. The objectives of the present paper are to shed knowledge on genetic diversity and mating system in neem. Strategies for genetic conservation and long term breeding for the species are also suggested.

Genetic diversity

The SRIEG (Southern Regional Information Exchange Group) Meeting on ‘genetic diversity in commercial forest tree plantation’ concluded that genetic diversity is a fundamental tenet of the conservation ethic, and that genetic diversity is an important consideration when managing forest stands, ecosystem and landscapes (Libby et al. 1997). Genetic diversity is the most important component of biodiversity. It is the foundation of ecosystem stability and forest sustainability because genetic diversity provides raw material for evolution, survival and adaptation of the species, especially under changed environmental conditions. Genetic diversity needs to be assessed in long-term genetic resource collections, in breeding populations, in seed orchards or planting materials producing populations and in production populations (Muona 1990). For ecologically and socially sustainable forestry, monitoring of genetic diversity in forest trees is also important. Thus, genetic diversity includes all levels of variation harboured by plants, from morphology, physiology and biochemical to genes or to DNA sequences. Genetic variations of in neem have been investigated using morphological, physiological and biochemical traits (Kundu and Tigerstedt 1997; Kundu et al 1998; Kundu 1999a; Kundu and Tigerstedt 1999; Kundu 2000). Significance and brief descriptions of these studies are outlined below.

Provenance variation

Growth chamber and field investigations were conducted on the materials involved in the international provenance trials (Kundu and Tigerstedt 1997; Kundu et al 1998; Kundu 2000). Provenance variations in seed and plant traits were recognised in neem populations. Seed size, plant height, collar diameter, leaf area, leaflet ratio, and shoot:root survival rate were important and easily measured traits for an early evaluation of seed sources and indicated a potential for selection and breeding.

Clinal variation

Clinal variation along a humidity gradient was observed in the growth chamber experiment for shoot:root ratio, and number of leaves. Latitudinal clines were detected both in the growth chamber for leaflet ratio, as well as for collar diameters and survival rates in the field experiments. A regression analysis indicated a significant amount of variation for collar diameters due to the effect of latitude. Significant correlations between leaf number, shoot:root ratio with Mean Annual Rainfall (MAR) found indicated adaptation to water availability. These parameters were suggested for selection and breeding drought tolerance. The observed clinal variations provide a guideline for transfer of seeds and plant materials.

Physiological yield parameters

Yield is determined by the interaction of many physiological and biochemical processes, which could be manipulated through plant breeding and genetics. The higher dry matter productions in some provenances in neem were probably the results of high stomatal density. Nevertheless, the positive relationships between stomatal density or the total guard cell length per unit leaf area on the one hand and net photosynthesis, stomatal conductance, leaf area and whole-plant dry weight on the other hand, shed light on the possibility of using these physiological characters as early selection criteria (Kundu and Tigerstedt 1999).

Isozymes variation

Isozyme techniques have been widely used in plants as genetic markers for studying the population structure and genetic diversity, and for identifying species or genotypes. They are particularly useful for detecting genetic variation within and between populations, geographical or ecological races, clines, phylogenetic relationships, pollen neighborhoods, estimating mating system, and other genetic characteristics at the population level (Nevo 1978; Hamrick et al. 1979; Loveless 1992; Muona 1990). Recently, isozymes variations have been reported in neem (Kundu 1999a). Allelic differences were clearly marked between the populations. The Maletdehydrogenase (MDH-4) and 6-phosphogluconate dehydrogenase (6PGD-1) were suggested as useful marker loci for identifying genotypes or provenances. The total genetic diversity was high (HT = 0.57) in comparison with other forest trees.

Mating system and polycarpy

Many tropical species possess mechanisms that ensure or encourage outcrossing. Allozyme data demonstrated that tropical tree mating systems range from mixed outcrossing and selfing (Murawski et al. 1990), to predominantly or completely outcrossing (O’Malley and Bawa 1987; Murawski et al. 1990). A high level of outcrossing rate (tm = 0.90) in neem has been detected by Kundu 1999b. The study indicates that A. indica possesses a predominantly allogamous mating system. Insect pollination, presence of ‘protandry’, ‘andromonoecy’ and ‘polycarpy’, and absence of possible biparental inbreeding, were the probable causes of such a high degree of outcrossing. In the seed study of Kundu 1999b, on average, an endocarp (trilocular) produced only 0-2 seeds so that 4-6 ovules were eliminated during seed development. The presence of the empty endocarps, as well as fewer seeds per endocarp than expected, indicate that inbreds are selected against. An analysis of seeds per endocarp showed significant heterogeneity indicating genetic differences in tolerance to inbreeding. The ‘polycarpic’ endocarp structure was suggested as a mechanism to make natural barriers against widespread selfing. Nevertheless, the recent studies on outcrossing rate in neem could be useful in formulating breeding strategies and designing seed orchards.

Genetic conservation

The purpose of conserving plant genetic resources for food, agriculture and forestry is to make these resources available for utilization in plant or tree improvement programs. The improvement potential in tree productivity, quality characteristics, resistance to diseases and pests, and adaptation to climatic and edaphic stress are all based on the existing genetic variability. The existing genetic variability could be used to produce new gene combinations that add to increased genetic diversity.

More genetic variability is expected to occur in neem in its natural distribution-range. This diversity needs to be assessed, including its southward extension into drier parts of the peninsular and eastwards to the upper Irrawady valley in Myanmar (Arora 1993). Emphasis should be paid to the species diversity and intraspecific variability prevalent in the natural range of A. indica, which is widely distributed, and the species A. siamensis, which is more confined to Indo-China and Thailand. Ecogeographic survey and study of such regions may lead to identification of more variable types and natural hybrids. A variant individual tree showing narrow serrated leaflets has been identified and grown at the Arid Forest Research Institute (AFRI), Jodhpur (India). Several Melia species possess a dwarf habit, short stem, well-branched canopy, good flush etc. Such variability may need to be studied for its potential use.

Field gene banks (in situ) of representative diversity needed to be established, particularly at the area of diversity. Some of these conservation areas should be kept free of human interference to allow the plant populations to undergo evolution in line with environmental changes. Other conservation areas may be managed to some extent to prevent their shrinking due to competition from other species.

For ex situ conservation, sampling from the populations should maintain the maximum level of genetic diversity. Studies have indicated enormous genetic variation in neem at the species, population and individual levels. Germplasm conservation should concern all levels of genetic diversity. Clinal variations in some traits suggest that germplasm sampling should get more attention for populations along the latitudinal and humidity gradients.

It would appear best to integrate ex situ germplasm conservation into breeding programs through the management of long-term breeding populations. This would help to reduce the cost and promote the efficient use of available resources. Future research is needed on seed storage (both cryogenic and normal) to enable it to retain its viability for a longer period because neem seeds are recalcitrant.

Strategies for long-term breeding

Breeding strategies are normally evaluated in terms of genetic gain expected for traits of importance, usually over a period of time. Expected gain could be utilised by a proposed selection and propagation system. Several long-term breeding strategies are now available, which are designed to retain sufficient genetic variability to counteract the risks of inbreeding in future generations. In recent years, different strategies and methods have been proposed and used to widen the genetic base of breeding populations for long-term breeding. The Multiple Population Breeding Systems (MPBS) and Hierarchical Open Ended systems (HOPE) have been suggested for breeding allogamous tree species (Namkoong 1989). In the former, intensive recurrent selection is practiced within multiple independent populations, to create broad differences among them in relation to their source germplasm, their traits and adaptibilities. The latter case, HOPE, is maintained as a very large base population, which is open-ended as far as receiving new materials is concerned. MPBS may be the better choice for neem because it is better suited for combining the highest possible genetic gains and highest possible genetic diversity. It also gives the breeders more options for shifting breeding goals with changing environments and demand (Namkoong 1989; Eriksson et al 1993).

Tree breeders evaluate the genetic resources available for improvement, and select genes of great utility and economic importance and package them in genotypes that can be used to establish commercial plantations (Libby 1973). Tree breeders also identify superior genotypes in existing provenances and propagate them clonally to tap both additive and non-additive genetic effects governing commercial traits. In the study of Kundu 1999b, the neem tree has been shown to be an allogamous species. For long-term progress in neem breeding, a general flow chart may be suggested (Fig 1). The following major approaches may be included in a long-term improvement program.

Fig. 1. A long-term breeding strategy in A. indica.

1. Introduction of provenances (40-50 trees/provenace)

2. Establishment of seed production areas (SPA)

3. Genetic testing and establishment of seed orchards (SO) include

a) selection of plus trees in the provenances
b) progeny testing of plus trees and
c) establishment of seed orchards for long-terms genetic gains

4. Clonal strategies include

i) clonal propagation for selected plus trees for large genetic gains
ii) establishment of clone trials and clone banks

5. Conservation of genetic diversity

From the above long-term breeding population, short-term breeding populations (5-100 individuals) could be drawn, which are adapted to the local climate having the desired traits. Some promising genotypes may be possible to select for clonal tests directly (Fig 1). The long-term and short term breeding populations will be maintained by continuous population improvement, which would ensure a wide genetic base and long-term progress in neem breeding. The essential features of this long-term breeding population are that the improvement will result from mild selection, which increases the frequency of desirable genes without the loss of neutral genes.

Breeding objectives and methods

Breeding strategies differ considerably with the aims and objectives of the tree improvement program. The neem can be bred for many purposes such as higher fruit yield and other desired agronomic traits. The first category includes the uses as medicine, pesticides, fungicides, nematicides, cosmetics, fodder and organic manure. The latter category includes timber and fuelwood, agroforestry species, shelterbelts, avenue trees, drought and disease resistance.

Multiple stems that produce high biomass, high wood density, and large quantities of fruit with high limonoid and oil content, should be the criteria for selecting provenances for fuel wood production, charcoal, pharmaceutical and pesticide industries. For early establishment of a plantation, especially in dry areas, provenances of neem that show high survival rate and fast growth may be the best choice. In the long-term for utilization as timber, provenances that have a straight stem and more promising intraspecific hybrids could be among the selection criteria. For agroforestry species, a narrow crown with deep-rooted habits; for shelterbelts, resistance to high wind-run and persistent leaf habits; for avenue trees, a larger crown with evergreen features could be selection criteria.

Net photosynthesis is also an important trait affecting yield. Leaf and leaf stomatal characters, stomatal density, and total guard cell length per unit leaf area, as reliable indicators of stomatal conductance, were suggested as important components for yield. Leaflet ratio, shoot:root ratio, and leaf area could be easily measured traits in breeding for drought tolerant neem.

The strategy to improve several characters simultaneously depends on whether they are controlled by a single major gene or groups of genes. If the characters are positively correlated, improving of one character would improve the other and selection becomes efficient. Independent characters could be selected successively or simultaneously. Strongly negatively correlated traits should be carefully estimated, since improving one character always amounts to lowering the level of other(s). To improve several characters simultaneously, index selection is probably the best choice.

In neem breeding, most of the characters mentioned are probably quantitative in nature. The recent studies indicate that seed diameter, leaflet ratio, and leaflet area were all independent. These traits could be improved without having negative effects on other. In the preliminary studies, no significant negatively correlated traits were observed. The economic traits of interest in this species could be growth, straight bole, crown form, and seed oil, including limonoid contents.

A natural hybrid (A. siamensis × A. indica) found in Thailand indicates that hybridization among related species is possible and promising for further neem improvement. Natural or artificial hybrids could be introduced both in long and short term breeding programs. When breeding neem for biochemicals it may be advantageous to advance to the F2 generation and select in the recombinant generation.

The usefulness of polyploid and mutation breeding in neem are still questionable. Gene transfer techniques could be useful for producing transgenic neem (Naina et al. 1989). Genes that control drought tolerance and genes for high 'azadirachtin-A' contents could be usefully exploited.

The establishment of seed production areas (SPO) from provenance trials is featured as the most effective and time-efficient strategy for short-term neem improvement. For long-term sustainable tree production as well as fruit yield, seed orchards (SO) are suggested.


Anon, 1998. A report on the workshop of the international neem network. Yangon, Myanmar, 28 July - August 1997. For Genet Res 25: 60-62

Arora, R.K., 1993. Genetic diversity and ethnobotany. Neem Res Dev. 3: 33-37

Eriksson G., G. Namkoong and J.H. Roberds, 1993. Dynamic gene conservation for uncertain future. For Ecol Manage 62: 15-37

Hamrick, J.L., J.B. Linhart and Y.B. Mitton, 1979. Relationships between life history characteristics and electrophoretically detectable genetic variation in plants. Ann Rev Ecol Syst 10: 173-200

Kundu, S.K. and P.M.A. Tigerstedt, 1997. Geographical variation in seed and seedling traits of neem (Azadirachta indica A. Juss.) among ten populations studied in growth chamber. Silvae Genetica 46:129-137

Kundu, S.K., Q.N Islam, C.J.S.K. Emmanuel and P.M.A. Tigerstedt, 1998. Observations on genotype x environment interactions and stability in the international neem (Azadirachta indica) provenance trials in Bangladesh and India. For Genet 5: 85-96

Kundu, S.K. and P.M.A. Tigerstedt, 1999. Variation in net photosynthesis, stomatal characteristics, leaf area and whole-plant phytomass production among ten populations of neem (Azadirachta indica). Tree Physiol 19: 47-52

Kundu, S.K., 1999a. Comparative analysis of seed morphometric and allozyme data among four populations of neem (Azadirachta indica). Genet Res and Crop Evol. 46(6): 569-577

Kundu, S.K. 1999b. The mating system and genetic significance of polycarpy in neem tree (Azadirachta indica). Theor and Appl Genet. 99(7/8):1216-1220

Kundu, S.K., 2000. Evaluation of provenance variation on early growth and survival of neem (Azadirachta indica) in Bangladesh and India. Journal of Trop For Sci. 12(3): 509-523

Libby, W.J., 1973. Domestication strategies for forest trees. Can J For Res 3: 265-276

Libby, W.J., F. Bridgwater, C. Lantz and T. White, 1997. Genetic diversity in commercial forest tree plantations: introductory comments to the 1994 SRIEG meeting papers. Can J For Res 27: 397-400

Loveless, M.D., 1992. Isozyme variation in tropical trees: patterns of genetic organisation. New For 6: 67-94

Muona, O., 1990. Population genetics in forest tree improvement. In: Brown, A.H.D., M.T. Clegg, A.L. Kahler and B.S. Weir (eds), Plant population genetics, breeding and genetic resources. Sinauer Press, Sunderland, MA. USA. pp 282-298

Murawski, D.A., J.L. Hamrick, S.P. Hubbell and R.B. Foster, 1990. Mating systems of two bombacaceous trees of a neotropical moist forests. Oceologia 82: 501-506

Namkoong, G., 1989. Population genetics and the dynamics of conservation. In: Knutson L, and A.K. Stoner (eds), Biotic diversity and germplams preservation, global imperatives. Kluwer Academic Publisher, Dordrecht. pp 161-181

Naina, N.S., P.K. Gupta and A.F. Mascarenhas, 1989. Genetic transformation and regeneration of transgenic neem (Azadiarchta indica) plants using Agrobacterium tumefaciens. Curr Sci 58: 184-187

Nevo, E., 1978. Genetic variation in natural populations: patterns and theory. Theor Pop Biol 13: 121-177

O’Malley, D.M. and K.S. Bawa, 1987. Mating system of a tropical rain forest tree species. Am J Bot 74: 1143-1149

[1] Divisional Forest Officer, Coastal Forest Division, Bhola, Bangladesh. Tel: +88 0491 55416; Fax: +88 049155192; Email: Kundu98@bdonline.com