R.A. Bray
Introduction
Leucaena
Gliricidia
Calliandra
Sesbania
Desmodium and Codariocalyx
Conclusion
References
Although there is a wide range of genera that can be considered 'tropical tree and shrub legumes' (Williams 1983) few of these have been studied to any great extent. The best known is Leucaena leucocephala, which has become naturalised over large areas of the tropics. Other commonly grown tree and shrub legumes are Gliricidia sepium, Calliandra calothyrsus, Sesbania sesban and S. grandiflora, and many species of Desmodium (including Codariocalyx). In this section discussion will concentrate mainly on variation within Leucaena as that is where most information is available. References such as Le Houerou (1980) can be consulted for a broader perspective.
Leucaena leucocephala is probably the most widely grown tropical tree legume in the world. However, the genus Leucaena comprises a wide range of species and genotypes, some of which are very different from this well-known species. In the past, research and development has concentrated almost entirely on L. leucocephala, but it is likely that in the future more use will be made of the other species in the genus. The various species of Leucaena, their particular characteristics and what uses they might have, will be discussed here.
Leucaena is one of 60 genera in the tribe Mimoseae of the subfamily Mimosoideae. Although L. leucocephala has spread worldwide, the genus is indigenous only to Texas (USA), Mexico, Central America and the Caribbean, and the northern parts of South America. It is intolerant of highly acid or waterlogged soils, and is not often found at altitudes above 2,000 m. All species contain mimosine, a non-protein amino acid. It is a very variable genus, and taxonomic efforts in the past have created a bewildering array of 'species' (Britton and Rose 1928). Brewbaker (1987) simplified the taxonomy to include only 12 species but taxonomists currently recognise 16 species (Hughes 1991).
These are listed in Table 3.1.1, together with some of the characteristics by which they may be distinguished. Of these, pairs of pinnae, pinnule number, pinnule size and inflorescence size are likely to be the most useful in practical terms. A full key to the determination of species is given by Brewbaker (1987). However, recent taxonomic studies based on extensive collections in Mexico and neighbouring countries provide an alternative viewpoint by delineating new species and by again dividing some species into subspecies (see for example Hughes 1991). In the discussion below, the simpler terminology of Brewbaker will be used.
Characteristics of individual species
Leucaena leucocephala
This species probably originated in the Yucatan Peninsula of Mexico. One 'common' variety is now widely distributed throughout the tropical world, often appearing to be a weed. Extensive collections have now been made in Mexico, from a wide range of environments. The species is almost never found above an altitude of 1,500 m (and generally much lower), suggesting that there are unlikely to be frost tolerant types although regrowth after frost is vigorous enough to make it a commercial species in southern Queensland and south Texas, USA. Some accessions have come from very dry areas (rainfall c. 300 mm) but these have probably been growing in favoured situations (e.g. runoff from roadsides) and again there is little reason to hope to find genotypes well adapted to arid conditions.
Morphologically, the species is quite variable. It is generally accepted that there are two major varieties (in the botanical sense). These are the 'giant' or 'Salvador' type, and the 'common'. It is this latter type which has spread throughout the world, often as a weed. It is generally not very vigorous, but is a prolific seeder. Thus, as a general rule, it should be possible to find higher yielding types than those currently naturalised. in most tropical countries. The 'giant' types tend to be faster growing, with fewer branches, and moderate seed production.
Until the 1970s, only relatively small collections of germplasm existed. Early selections from the Hawaiian programme included the cultivars K8 and K28. These were selected mainly on the basis of good wood production, but when cut regularly produce high leaf yields also. In Australia, the cultivars Peru and Cunningham were developed. These were selected for high yield and branching, to provide a plant form suitable for grazing by cattle. In recent years, many hundreds of accessions have been collected in Mexico and screened in programmes all over the world. Superior lines have been isolated in most programmes. These include CPIs 58396, 61227, 85176 and 90814 in Australia, ML1 and ML2 in Malaysia, and a range of material such as K584 and K636 in Hawaii (see e.g. Bray et al. 1988). However, the recent problems with the leucaena, psyllid (see Section 6.1) have tended to obscure differences in potential productivity, although some lines such as K420 and K636 show more tolerance than others.
Table 3.1.1. Characteristics of Leucaena species (after Brewbaker 1987).
L. diversifolia
There are two main subdivisions of this species. These are the 'diploid' and 'tetraploid' groups, which have 52 and 104 chromosomes respectively. The diploid group occurs naturally in southern Mexico and Guatemala and, due to self-incompatibility, members of this group are obligately outcrossing. The tetraploid group occurs only in Veracruz, and is self-fertile. Tetraploids can generally be distinguished morphologically from diploids by having larger, more open flower heads, fewer florets per inflorescence, more pairs of pinnae and pairs of pinnules, and longer pinnules (Pan 1984). There has not been extensive testing of L. diversifolia germplasm. Most trials have been concerned with wood yield, but some accessions (e.g. CPI 46568) can yield as well under regular cutting as L. leucocephala (Bray et al. 1988). However, there are some indications that this species may not survive well under dry conditions. Because it often occurs naturally in the mountains, it has been evaluated as a source of cold tolerance, and some lines (CPI 46568) again show promise in this regard (R. Wheeler, personal communication). However, it is unlikely that any really frost tolerant genotypes exist. Although the genus is not found on acid soils, it has been used successfully in the breeding of lines that grow well in such conditions (Hutton 1990). It is also potentially useful as a source of psyllid resistance (Section 6.1) and some lines isolated from old plantations in Indonesia show promise in this regard.
Little is known about the value of this species as animal feed (although its mimosine content is less than that of L. leucocephala), or about its agronomic limitations. Its seeds are considerably smaller than those of L. leucocephala and consequently it may be difficult to establish in the field. Seed production of the diploid types is sometimes sparse, and any seed orchard needs to contain a reasonable number of trees (30 at a minimum) to ensure adequate pollination and to minimise inbreeding. The diploid group often have a characteristic almost two-dimensional branching habit, and may have potential as a support for growing vines or other scrambling plants.
L. pallida
This species has only recently been recognised (Pan 1985) and probably is an ancient hybrid between L. diversifolia and L. esculenta. It has some cold tolerance (being found mainly in mountain areas) and psyllid resistance. It yields reasonably well under cutting (Wheeler and Brewbaker 1989), but its real agronomic potential is not known. It is another outcrossing species, and thus seed orchards should contain a reasonable number of trees.
L. pulverulenta
Interest in this species has largely been due to its low mimosine content and arboreal habit. However, it can produce yields of edible material approaching that of L. leucocephala Hybrids between L. leucocephala and L. pulverulenta can be very vigorous, with moderate levels of mimosine (Bray 1984). Such hybrids have long been used as shade trees for coffee.
Little is known about the other species although about one thousand accessions, including all 16 species, have been grown in University of Hawaii trials. However, they are all browsed in the wild, and therefore have some potential as forage plants. The pods and seeds of some species (especially L. esculenta) are eaten by humans. Leucaena retusa and L. greggii come from relatively cool climates, and may have potential for these areas. Other species, such as L. lanceolata, have attractive foliage and flowers, and may have potential as ornamentals. Others (L. macrophylla) may have potential as timber producers. Some features of all 16 species are given in Table 3.1.1.
The potential of hybrids
Most Leucaena species are outcrossing, the exceptions being L. leucocephala and tetraploid L. diversifolia. This means that seed harvested from any of the cross-pollinated species will not breed true. In fact, all Leucaena species, even the predominantly self-pollinated ones, will cross readily with other species. This has both advantages and disadvantages. If seed is harvested from a nursery containing a range of species, hardly any of it will be 'true to type'. It is common to find a wide range of different plant types in samples of open pollinated seed, and, if these are interspecific hybrids, they may not look like any of the known species, being intermediate between the two parents. These interspecific hybrids are often sterile, but fertility levels are sometimes sufficient to enable enough seed to be produced for selections to be made. The ability to make interspecific crosses easily is of course of great interest to the plant breeder, but it is likely to be a long and difficult job to combine the desired characters of any two parents, together with satisfactory seed production. By careful selection of parents, it may be possible to produce large quantities of seed. This has been suggested for hybrids between L. leucocephala and L. pulverulenta (Bray 1984), which have considerable potential as high yielding forage plants. Brewbaker et al. (1990) have also discussed this possibility in a wider context. The Hawaiian program is well advanced in the development of cool tolerant and psyllid resistant hybrids based on the interspecific crosses of L. leucocephala with L. pallida (KX2) and with L. diversifolia (KX3).
Gliricidia is classified in the tribe Robinieae, subfamily Papilionoideae. This genus has not been collected to the same extent as Leucaena. There are considered to be only four species (Polhill and Sousa 1981), of which G. sepium (common name 'gliricidia') is the only species of real agronomic potential; it is only recently that an effort has been made to collect a representative range of germplasm of this species. This has been undertaken by the Oxford Forestry Institute, which has organised a series of trials using 28 provenances collected from a range of environments in Central America and Mexico. Collection sites ranged from 7 to 19° N. altitude from sealevel to 1,650 m, and rainfall from 650 to 3,500 mm. From preliminary results, it is clear that there is considerable variation in growth form and yielding ability. However, the performance of the best lines is fairly consistent across environments. One provenance from Guatemala, Retalhuleu, showed superior production for both leaf and wood while another from Guatemala, Monterrico, showed poor growth in terms of wood production but was outstanding for leaf production (Simons and Dunsdon 1992). There is also considerable variation in digestibility within this collection (Bray et al. 1993) suggesting that it may be possible to isolate superior lines with both high yield and high digestibility. Gliricidia is a cross-pollinated species; seed supply is often a problem as seed set is sporadic and uncertain in environments without a marked dry season. However, since the usual method of propagation is by cuttings, this is a difficulty largely confined to experimental situations. One danger arising from vegetative propagation is the existence of large monogenotypic stands, with no inherent variation to combat new pests and diseases. Such a situation is undesirable, and efforts should be made to ensure diversity by planting from a wide range of clones. Even widespread distribution of a single elite variety is perhaps best avoided.
Calliandra (Ingeae, Mimosoideae) contains more than 200 species (Williams 1983). Some of these have been grown for horticultural purposes, but only two have had any widespread agricultural use, C. calothyrsus and C. tetragona. The latter, a white flowered species, has proved to be slower growing than the red flowered C. calothyrsus, and will not be considered further.
Until recently, C. calothyrsus has not been comprehensively collected in the wild and testing has only been on a very narrow genetic base. The extensive plantings already made in Indonesia can be traced to only one or two accessions introduced in the 1930s. However, the geographical range of the species has been found to be more extensive than was formerly believed (Chang and Martinez 1984, MacQueen 1991). Collecting activities in Central America by the on have now secured over 20 provenances, mainly C. calothyrsus, but also including such agronomically unknown species as C. acapulcensis, C. grandiflora, C. houstoniana, C. juzepczukii, and C. physocalyx (D.J. MacQueen, personal communication). Testing of these provenances could reveal variation in yield, adaptation and feeding value.
The genus Sesbania (Sesbanieae, Papilionoideae) is indigenous to much of the tropical world; it is estimated that there are 70 species (Williams 1983) of which 10 are indigenous to Australia and 13 are woody perennials. The two most promising perennial species are S. grandiflora and S. sesban, both being used extensively in traditional agroforestry systems. Sesbania grandiflora is a tetraploid species native to Asian countries including India, Malaysia, Indonesia and the Philippines where it is commonly grown on bunds between rice paddies, along roadsides and in backyard gardens. Sesbania formosa is a closely related tetraploid species found in northern Australia. Sesbania sesban is a diploid species that occurs throughout Africa and most of western and southern Asia; it is a highly variable species, with two subspecies and four varieties recognised. Sesbania species are adapted to a wide range of environments and are noted for tolerance to alkaline and saline soils and to waterlogging. Germplasm collection and screening has been initiated by ILCA, CSIRO (Wood and Larkens, 1987) and the University of Hawaii, but systematic regional collections have been limited to Tanzania. Considerable variation exists, both in dry matter production and quality (see Evans and Rotar (1987) and Macklin and Evans (1990) for detailed discussions).
Desmodium (Desmodiae, Papilionoideae) is one of the largest of the tropical legume genera; it has been estimated that there are about 150 species with forage potential (Williams 1983). Some authorities include Codariocalyx, Dendrolobium and Phyllodium in the genus Desmodium, while others prefer to maintain them as distinct genera. The distribution of Desmodium is pantropical, with southeast Asia as an important centre of species diversification. Of the shrubby species, D. rensonii, D. discolor and Codariocalyx gyroides (syn. D. gyroides) are the most promising. However, of these only C. gyroides has been studied to any extent (Jones 1984). In this species the outstanding accession is CPI 76104, which has been circulated around the world, gaining many different identifying numbers in the process. More extensive collections are needed to document adequately the extent of variation.
Because of the diversity within all tree and shrub legume species, it is important, when planning experiments or evaluating results, to be certain of the identity of the particular cultivars or accessions being used. It is not sufficient to quote yields of a particular species (e.g. L. leucocephala) but is necessary to include all known details (e.g. L. leucocephala CPI 61227). Only by doing this is it possible to get accurate comparisons between various experiments.
Another point worthy of emphasis is the need for diversity in tree legume plantings. The damage wrought by the leucaena, psyllid is a good example of the perils of relying too much on one species or one cultivar. Future plantings should contain not only a range of species, but probably a range of varieties within species. This should safeguard against major disasters due to disease or insects.
Bray, R.A. (1984) Evaluating hybrids between Leucaena leucocephala and L. pulverulenta as potential forage plants. Australian Journal of Experimental Agriculture and Animal Husbandry 24, 379-385.
Bray, R.A., Cooksley, D.G., Hall, T.J. and Ratcliff D. (1988) Performance of 14 Leucaena lines at five sites in Queensland. Australian Journal of Experimental Agriculture 28, 69-76.
Bray, R.A., Ibraham, T., Palmer, B. and Schlink, A.C. (1993) Yield and quality of Gliricidia sepium accessions at two sites in the tropics. Tropical Grasslands 27, 30-36.
Brewbaker, J.L. (1987) Species in the Genus Leucaena. Proceedings of a workshop on biological and genetic control strategies for the leucaena psyllid. Leucaena Research Reports 7, 6-20.
Brewbaker, J.L., Sorensson, C.T. and Wheeler, R.W. (1990) New tree crops from interspecific Leucaena hybrids. In: Napompeth, B. and MacDicken, K.G. (eds), Leucaena Psyllid: Problems and Management. Proceedings of an international workshop held in Bogor, Indonesia, pp. 105-110.
Britton, N.L. and Rose, J.N. (1928) Mimosaceae. North American Flora 23, 122-128.
Chang, B. and Martinez, H. (1984) Germplasm resources of Calliandra calothyrsus Meissn. in Central America and Panama. Forest Genetic Resources Information, FAO, Rome, 13, 54-58.
Evans, D.O. and Rotar, P.P. (1987) Sesbania in Agriculture. Westview Tropical Agriculture Series No. 8. Westview Press, Boulder and London, 192 pp.
Hughes, C.E. (1991) Leucaena improvement through use of lesser known species. Leucaena Research Reports 12, 98-100.
Hutton, E.M. (1990) Field selection of acid-soil tolerant leucaena from L. leucocephala x L. diversifolia in a tropical oxisol. Tropical Agriculture (Trinidad) 67, 2-8.
Jones, R.M. (1984) Yield and persistence of the shrub legumes Codariocalyx gyroides and Leucaena leucocephala on the coastal lowlands of south-eastern Queensland. CSIRO Division of Tropical Crops and Pastures, Tropical Agronomy Technical Memorandum No. 38, 9 pp.
Le Houerou, H.N. (ed.), (1980) Browse in Africa. ILCA, Addis Ababa, Ethiopia, 491 pp.
Macklin, B. and Evans, D.O. (eds), (1990) Perennial Sesbania Species in Agroforestry Systems. Proceedings of a workshop held at ICRAF in Nairobi Kenya. NFTA Special Publication 90-01, 241 pp.
MacQueen, DJ. (1991) Exploration and collection of Calliandra calothyrsus as a foundation for future genetic improvement. Nitrogen Fixing Tree Research Reports 9, 96-98.
Pan, F.-J. (1984) Tetraploidy in Leucaena diversifolia. Leucaena Research Reports 5, 88-90.
Pan, F.-J. (1985) Systematics and genetics of the Leucaena diversifolia (Schlecht.) Benth. complex. PhD thesis, University of Hawaii, Honolulu, Hawaii.
Polhill, R.M. and Sousa, M.S. (1981) Robinieae. In: Polhill, R.M. and Raven, P.H. (eds), Advances in Legume Systematics. Royal Botanic Gardens, Kew, p. 283.
Simons, A.J. and Dunsdon, A.J. (1992) Evaluation of the potential for genetic improvement of Gliricidia sepium. Report to ODA on Forestry Research Project R-4525. Oxford Forestry Institute, 176 pp.
Wheeler, R.A., and Brewbaker, J.L. (1989) Results from the international psyllid trial network. Leucaena Research Reports 10, 11-15.
Williams, R.J. (1983) Tropical legumes. In: McIvor, J.G. and Bray, R.A. (eds), Genetic Resources of Forage Plants. CSIRO, Melbourne, p. 17.
Wood, I.M. and Larkens, A.G. (1987) Agronomic and phenological data for a collection of Sesbania species grown in south-east Queensland, Australia. Genetic Resources Communication No. 11, CSIRO, Brisbane.
