An e-publication by the World Agroforestry Centre

AGROFORESTRY A DECADE OF DEVELOPMENT  Printprint Preview

section 5
Research findings and proposals

Chapter 17
Leucaena: a multipurpose tree genus for tropical agroforestry

James L. Brewbaker
President, Nitrogen Fixing Tree Association
Professor, Department of Horticulture, University of Hawaii
3190 Maile Way, Honolulu, Hawaii 96822, USA

Introduction

The genus Leucaena has been popularized widely in the past decade. It has been the subject of an excellent book by Pound and Martinez Cairo (1983), of major international conferences co-sponsored by the Nitrogen Fixing Tree Association (NFTA) and the International Development Research Centre (IDRC (IDRC, 1982) and by NFTA (Leucaena Research Reports 7(2)), and of several national and international conferences (Kaul et al, 1981; PCARR, 1978; Indonesia, 1982).

Research on the genus is represented in over 3,000 publications, most of them included in the comprehensive three-volume bibliography of Oakes (1982-4), in the book by Pound and Martinez Cairo, and in reviews or bibliographies by Arellano (1979), Brewbaker and Hutton(1979),Halos(1980),Olveraef al. (1985) and Schroder (1986). Two excellent guides to production were co-authored by leucaena experts (NFTA, 1985a, 1985b); others include those of Banco de Mexico (1980), FAO (1983, 1985) and Proverbs (1985). The present review will focus on the past decade, primarily since the review paper of Brewbaker and Button (1979), and will seek to identify the people and institutions closely linked to leucaena's expanded use, and ask "Why do people plant leucaena?" Many references will be made to LRR, the Leucaena Research Reports, an annual publication of NFTA.

Historical perspective

Leucaena has often been regarded as a kind of "miracle tree", an appellation that makes sincere scientists wince. However, such names as subabul in India, lamtoro gung in Indonesia and giant ipil-ipil in the Philippines reflect the genuine affection that farmers, housewives, ranchers and scientists themselves have come to hold for the arboreal leucaenas.

Leucaena is not new to scientists, and certainly not to growers. It was a significant "alley crop" hedge in Asia, far from its native Americas, more than a century ago. It clearly was a common food and soil-restoration tree of early American Indian civilizations. In the first millenium AD, the Maya of the Yucatan peninsula may have relied heavily on leucaena as a green manure and possibly as a food crop (Brewbaker, 1979). Zarate (1984) considers the genus a significant source of leguminous pods long prior to Columbus. A major species, L. esculenta (Moc. and Sesse) Benth., got its name in recognition of its importance as food for highland Mexican Indians. The Mexican state of Oaxaca is a centre of great diversity of the genus, and owes its name to huaxin or Leucaena spp. (LRR 7(2):6).

Leucaena crossed the Pacific Ocean in Spanish galleons in the early 1600s, and by mid-1800 was pantropical and used in a variety of ways — food, fodder, shade, soil restoration. Superb early research on its soil-restorative properties occurred in Indonesia prior to 1910, as evidenced by Dijkman's early review (1950) encouraging use of leucaena for soil-erosion control. Takahashi and Ripperton (1949) published extensive studies of fodder management and improvement of koa haole (leucaena) in Hawaii. E. Mark Hutton and his colleagues developed an intensive leucaena improvement programme in northern Australia in the 1950s. Review papers that cover quite thoroughly the information on this early research and development include those of Brewbaker and Hutton (1979), Oakes (1968), Pound and Martinez Cairo (1983) and the U.S. National Research Council (1984).

New prominence

Leucaena has moved ahead of other leguminous trees in the past decade in both public image and farmer adoption. Three factors that have propelled it into new prominence are:

  1. Improved varieties of exceptional yielding ability;

  2. Improved communication about leucaena;

  3. Expanded demonstrations of leucaena's utility.

The new varieties were discovered through systematic surveys of leucaena germplasm collected from native habitats in the Americas and grown in Hawaii and Queensland, as described later in this paper. All leucaena are easily grown; they flower within two years, and are a delight to study and breed.

The improved communication began with a superb publication in 1977 on leucaena (NRC, 1984) which followed the first international conference on leucaena in the Philippines in 1976 (PCARR, 1978). This was followed in 1980 by the founding of the Leucaena Research Reports, an annual publication, at the University of Hawaii, with costs underwritten by the Council of Agriculture of Taiwan, Republic of China. Publication of LRR and related literature was assumed in 1982 by the Nitrogen Fixing Tree Association (NFTA), a non-profit organization presently composed of 1,200 associates. LRR has averaged 55 articles annually from more than 34 countries.

1_Leucaena a multipurpose tree for agroforestry


Expanded demonstrations of leucaena's growth and versatility in management and use have been the key to farmer acceptance (Figure 1). Alley-farming methods have been skilfully designed and farm-tested in Africa (Kang et al, 1984; Kang and Wilson, this volume); Indonesia (FAO, 1983) and India (Kaul et al., 1981).

New challenges

The role of leucaena as an agroforestry species is currently confronted with a new type of challenge — a psyllid insect pest, which has been reviewed thoroughly in LRR (7(2)). Apathy was leucaena's primary challenge prior to the 1970s, a decade that brought high oil prices and accelerated demands on tropical forests and the world's ecological health. The new challenge to leucaena is primarily sociopolitical. "Why plant leucaena — for what purpose?", will now be replaced by "Why plant leucaena — if it is not perfect?", since most laymen expect that somehow agroforestry trees must be "miracles" that have no need for research.

The test for leucaena, as the psyllid advances pantropically, will be primarily a test for policy makers asked to fund R & D on a non-food genus of major interest to the rural poor. The psyllid problem is soluble both by genetic and biological control (LRR 7(2)). It is not a significant problem in the Americas, where the psyllids are under adequate parasitization (Proverbs, 1985). Increasing commitment of funds and talents in long-range programmes on non-food crops like leucaena is essential if the constantly evolving biological challenges, such as the diseases and insects in tropical forest, are to be met.


Botany and genetics

Systematics and species collections

The genus Leucaena has over 50 names ascribed to it, most of them viewed as synonyms. Ten species were widely recognized as valid a decade ago (Brewbaker et al., 1972), to which two have since been added (LRR 6:78). The 12 species have been described recently in some detail from our collection of about 1,000 accessions and numerous artificial hybrids we have grown in Hawaii (LRR 7(2):7-20). It is unlikely that many species will be added to this list (Table 1), although several other populations are considered incipient species (LRR 7:110). The 12 species are easily differentiated by flower colour, leaflet size (Figure 2), growth habit, ecology, distribution and other traits (LRR 7(2):7). Chromosome counts have been made on the 12 species (Table 1). None of the species or populations tested have 2n = 26 or 2n = 28, counts that characterize most mimosoids. The importance of polyploidy is thus implicit in the high base number of the genus.

Several major expeditions to collect leucaena germplasm have been made in the past decade by our teams (1978,1985) with support from the University of Hawaii, USD A and IBPGR, by Bob Reid for CSIRO in 1980-1982, and recently by Colin Hughes for the Oxford Forestry Institute. Hawaii's collection of 905 accessions (80 percent indigenous) can be ascribed to the 12 species of Table 1 with rare exceptions. The Australian collection includes at least 700 accessions (Bray et al, 1984). Unique accessions include those ascribed to the taxon L. cuspidata, a highland shrub of N.E. Mexico similar to L. pallida. Another unique accession identified by C. Hughes (LRR 7: 110) was a large tree that otherwise resembled the shrubby L. shannoni.

The important commercial species, L. leucocephala (2n= 104) has long been recognized as a polyploid, and our repeated efforts to identify segregating monogenes (e.g., isozymes) in this species have met with failure. Earlier references to simple genetic segregations in this species are now viewed with suspicion. The species is presumed to be an amphidiploid, which Sorensson (1987) suggests most logically to be built on the 2n=52 diploids L. leucocephala and L. diversifolia. Pan (LRR 7(2):6) has shown that the 2n= 104 L. pallida (taxon conserved over synonyms dugesiana, oaxacana and paniculata due to priority in descriptive literature) is an amphiploid derived from the species L. diversifolia and L. esculenta. An extensive set of traits confirms this origin, including distributional, ecological, and morphological conditions. Pan (LRR 5:88) further identified 2n=52 and 2n=104 races within L. diversifolia, distinguished by their distribution and morphology (Figure 2).


Table 1 Species of the genus Leucaena

2_Leucaena a multipurpose tree for agroforestry


3_Leucaena a multipurpose tree for agroforestry

Breeding systems

The common leucaena is self-seeding, a fact probably underlying its early use as a food. This is both a blessing and a curse, permitting us rapid seed multiplication but also giving the species an undesirable element of weediness. Most accessions are less seedy than the common type, of which K29 and K636 are outstanding in form, but they have not been used widely due to difficulty in seed increase. The psyllid tolerance of K636 may encourage its wider use (LRR 7(2):29,84).

In contrast, nearly all other Leucaena species have proven self-incompatible and thus completely outcrossing, with the exception of low self-fertility in L. retusa (LRR 7(2):6). The tetraploid form of L. diversifolia is self-compatible, in contrast to the uniformly self-incompatible diploids in this species (LRR 5:88). The operation of "S" alleles of the gametophytic type was confirmed, and a theory presented for derivation of self-compatibility in the two tetraploid species (Bewbaker, 1983; LRR 7:114). Self-incompatible forms of both species might be expected following hybridization of self-compatible parents. S1 plants could be exploited in several ways, notably in production of hybrid seed and in development of broad genetically based synthetics. They might be further desirable in reducing seediness, quite probably to the advantage of wood yields.

Arboreal and shrubby varieties

Two decades ago, only Leucaena leucocephala was recognized to have merit in agroforestry-type land use. It was the only species distributed pantropically, represented by one variety — a seedy shrub with virtually no genetic variation. Improved strains have now replaced the common type, most of them of the "giant" or arboreal type referred to as "Salvador-type" (LRR 1:43). The first of these, K8, was released by the author in 1967 in Thailand and the Philippines, and it was described as the "Hawaiian Giant" (Brewbaker, 1975) largely because the common weedy type had been referred to as "Hawaiian"(which it is not, having come from Mexico by way of the Philippines). A later Australian release of M. Mutton, "Cunningham", resulted from crosses of the Salvador type and the more branched Peru type, and proved to be an excellent fodder variety. Batson et al. (1984) have collected Caribbean accessions, classifying over 90 percent as the common type.

The arboreal ("giant") and shrubby ("common") forms are now viewed as taxonomically distinct: var. glabrata and var. leucocephala, respectively (LRR 7(2):6). The "giant" or glabrata variety is arboreal, with glabrous branches and generally large vegetative and floral parts, and is widely distributed by man in N.E. and N.W. Mexico. The "common" or leucocephala variety is indigenous to a restricted area of S.E. Mexico. Hybrids are arboreal but early to flower and seedy like the commons. Similar morphological variants (Figure 2) occur in L. lanceolata, var. lanceolata (shrubby) and var. sousae (arboreal). Hughes (LRR 7:110) collected a variant of L. shannoni that makes impressive arboreal growth in Hawaii. A Panamanian tree described as L. multicapitula also appears to be an arboreal variant of the South American shrub, L. trichodes.

Genetic improvement

Genetic improvement has focused on L. leucocephala, a self-pollinated species whose polyploidy must be stressed as a factor impeding rates of genetic gain (Brewbaker, 1983). Improvement of biometrical traits requires large progenies and high selection indices. Most released varieties have been derived directly from accessions in Latin America. These have been marked by high uniformity of single-tree progenies. Two major breeding programmes are those of Hutton (1983) and Brewbaker (1983), the former focusing on acid tolerance and the latter on species hybridization and tolerance to cold (see later sections).

Pollination methods have been improved primarily through the work of the author's colleagues Pan and Sorensson (Brewbaker, 1983). They stress the value of early-bird activities, emasculating (as needed) at dawn and pollinating with fresh pollen taken at anthesis. Prior-day emasculation (LRR 5:29) and various detergent dips have worked less well. Dry pollen or stigmas can be moistened with glycerine to enhance results.

Many qualitative traits have been scored for their appearance in parents and hybrids of leucaena, but relatively few carried through to genetic analysis. Dominance in Fl characterizes arboreal over shrubby, early over late flowering, pink over white flowers, yellow over white flowers, pubescent branches and leaves over glabrous, flower odour over none and pendulous inflorescence over erect. A small list of gene loci has been developed for leucaena due to the work on 2n=52 L. diversifolia of Pan (1984), as follows:

Locus                                Alleles                          Description
S                                       Multiple                         Gametophytic, personate self-incompatibility
Pub                                   2                                   Pubescent leaflets, dominant over glabrous
P x 1                                 2                                   Peroxidase, roots, co-dominant alleles
P x 2                                 2                                   Peroxidase, roots, presence vs. absence
P x 3                                 2                                   Peroxidase, roots, presence vx absence
P x 4                                 2                                   Peroxidase, roots, co-dominant alleles


An early report of monogenic segregations for the arboreal vs. shrubby trait of L. leucocephala deserves verification. It is to be hoped that identification of genes and mapping will be accelerated as the diploids and interspecific hydrids come under increased study.

Species hybridization

One of the most exciting advances in our knowledge of leucaena has come through species hybridization in the past decade. Leucaena leucocephala has been crossed successfully with all other species except L. greggii (not tested). At least 51 species hybrids are now growing and under study (LRR 7:13), and their psyllid tolerance is shown in Table 2, following Sorensson and Brewbaker (LRR 7:13). With rare exceptions, the hybrids have been partially or highly fertile, allowing further hybridization and backcrossing for gene transfer. The entire genus is proposed to be an "effective gene pool" available for breeding improvement. Species hybrids have been made largely in our arboreta (LRR 7(2):29) and in those of Mutton and Bray (Bray el al, 1984; Hutton, 1983), and many are impressive.


Table 2 Hybrids produced among species of the genus Leucaena as scored for resistance to Heteropsylla cubana

4_Leucaena a multipurpose tree for agroforestry

The occurrence of unreduced gametes leading to polyploid progenies in leucaena has been documented by Sorensson and is a probable source of both amphiploid and autoploid populations. Large pollen grains are believed to represent such unreduced gametes, and occur with measurable frequency in many species.

Plant chemistry and physiology

Major discussions of leucaena's fodder and wood chemistry occur in other sections of this paper, notably of fodder quality, mimosine and gums. Pound and Martinez Cairo (1983) provide an excellent review of the chemistry of leucaena. Allelopathy has been attributed to leucaena leaves, based on inhibitions of seed germination with mimosine itself (LRR 3:65) and with leaf extracts (LRR 3:57). The extracts contained several common phenolics of toxic potential, e.g., ferulic and coumaric acids (LRR 3:57). Extrapolations to the field are fraught with hazards, and there has been no clear documentation of leucaena-induced allelopathy under field conditions. Leucaena suppresses nearby herbs largely through competition for light and water.

Physiological research on leucaena has been relatively limited. Photosynthetic rates increased linearly with light flux, and saturated at 40,000 lux at a rate of 25.8 mg CO2 dm-2 hr-1 in Taiwan (LRR 2:45), and at least 40 mg in a study in India (LRR 6:42), generally similar to other C3 plants. Leucaena has been described as a facultative shade plant that made major adjustments in stomatal density and conductance, light-saturating photo-synthetic rate, and leaf dimensions when grown under shade. Leaflet folding and stomatal responses to avoid direct sunlight and drought stress were strongly related to effective mycorrhizal associations (Huang et al, 1985). Transpiration rates were calculated in the range of 0.25-0.57 ml cm-2 (LRR 2:45). Leaf yields as kg dry matter per hectare were linearly related to solar radiation in Hawaii by the regression Y = -13 + 3.83X over the range of 8-18 MJ m-2 day -1 (LRR 6:88), roughly tripling from summer to winter in Hawaii. It is less clear how closely wood increments regress on intercepted irradiance. Early winter short-day flowering occurs in L. esculenta and L. macrophylla in Mexico and Hawaii, while other species appear day-length-neutral in flowering. L. esculenta has flowered irregularly when planted near the University of Hawaii's Food Science building, perhaps in response to ethylene stimulants.


Ecology and soils

Agroforestry in the tropics is often viewed as a system of band-aids or palliatives for injuries to the ecosystem. Deforested and poorly fanned tropical soils are spoken of as "degraded" and agroforestry is expected to restore them miraculously. The genus Leucaena is one of impressive adaptability in the tropics, but is thus being asked to grow luxuriantly on many types of generally impoverished soils. Like most plants, it grows best on the best soils, but has major problems with soils of low pH, low P, low Ca, high salinity, high aluminium saturation and waterlogging.

Acid soils

The "cerrado" soils (Oxisols and Ultisols) of South America pose a challenge to leucaena, and generally to most fodder legumes. These soils range from pH 5 down, with Ca deficiency and Al toxicity (over 50 percent saturation), and deficiencies of N, P, K, S, Mg and Zn (Hutton, LRR 3:9). Hutton (1983,1984) has concluded that leucaena has adequate tolerance of aluminium in the soil solution, but is more sensitive to deficiencies of calcium. While other major elements are translocated both to roots and shoots, Ca is mono-directional and translocated only from roots and shoots. The Ca content is very low in the Oxisols of Brazil, and the inefficient root absorption of Ca by leucaena is considered the major deterrent to growth on these soils (LRR 3:9). Even with 11 ha-1 of dolomite or 0.81 ha-1 of gypsum, leucaena died within four years, evidently due to Al inhibition of Ca uptake (LRR 7:28). Gypsum was the preferred source of Ca and S to supplement rock phosphate on these acid soils (Hutton, 1984). Liming of surface soils was adequate, however, to permit good leucaena growth in low Al acid subsoils in Hawaii and in pot trials (Olvera and Blue, 1985).

Acid-tolerant leucaena include L. diversifolia and L. shannoni, marked by superior root absorption of Ca. This tolerance of acidity was transferred to their hybrids with L. leucocephala (Hutton, 1984). Tolerance was tested to acid, Al-toxic soils using root growth in culture by Oakes and Foy (1984), and a wide range of varietal responses noted, which did not correlate directly to Al liberation. Great reduction in growth occurred with concentrations over 4 ppm, and were little influenced by Ca concentrations (Chee and Devendra, 1983).

Acid Oxisols restrict leucaena growth severely even when Al is not a major problem (LRR 2:69; Chee and Devendra, 1983). Relative yield of leucaena increased linearly with pH on two acid soils ameliorated with lime in a continuous-function design, from about 20 percent yield at pH 5 to 100 percent at pH 7 (Munns and Fox, 1977). Manganese solubility, which is known to induce Fe deficiency, was high in the tested Oxisol, but could not be invoked as a cause of lime response. Yields of 16-month-old leucaena K 29 correlated with pH in trials near Taipei, Taiwan (Hu and Chen, LRR 2:48), ranging linearly from 3.21 ha-1 (dry weight) at pH 4.7 to 15.51 ha-1 at pH 8. In contrast, no effect on one-year-old trees' growth was shown for liming treatments up to 91   ha-1, that changed pH values from 4.6 to 8.4 in Indonesia, on a soil with adequate levels of Ca and P. It has been tempting to accuse acidity of causing poor leucaena growth in soils later shown to be acutely P deficient (LRR 7:117).

Saline soils

Leucaena can often be found growing on coral outcroppings very near the ocean, but never growing well on the sand dunes or on inland sodic soils, an apparent reflection of high Ca demands and low Na tolerance. Trees stagnated on saline sodic soils of pH 9.5 in India that showed salt concretions (LRR 7:66). Growth in pots was severely depressed by NaCl at 6 g l-1 (LRR 5:77). Irrigation with saline water (2 mmhos) in Rajasthan, India, produced only fair growth (2.5 m in one year, LRR 6:54), and 21 varieties showed small differences in growth — none convincingly related to Na tolerance.

Cold tolerance

The genus Leucaena includes species that vary widely in cold tolerance, as evidenced by their regions of origin (Table 1). L. leucocephala does not vary widely in cold-tolerance, although selected varieties have survived well, regrowing from the crown after frosts at Gainesville, Florida, USA (LRR 5:84). Outstanding tolerance of frost has been found among hybrids of L. leucocephala and 4n L. diversifolia in Louisiana, USA (Brewbaker, unpublished). L. retusa survives frost routinely in south Texas, when L. leucocephala and L. pulverulenta are killed to the crown (LRR 5:76). Under severe frost, L. leucocephala was killed while L. pulverulenta regrew from coppice and L. retusa survived well (LRR 7:119). Tolerance of low temperatures appears to differ genetically in leucaena from frost tolerance, with species like L. leucocephala showing poor growth under low mean temperatures. Under Hawaii's relatively low variation in temperature, L. diversifolia completely outgrows L. leucocephala at temperatures below 22° C mean annual, or where soils are acidic (Brewbaker, 1986a), and the hybrids perform as well as the L. diversifolia parent (Figure 3). Near Brisbane, Australia (mean temperature 19°C), R. Bray also observed superior cold tolerance of the diversifolias, although many accessions were low in fodder yield, while leucocephalas were nipped by frost but grew back vigorously (LRR 3:1; 5:3).

5_Leucaena a multipurpose tree for agroforestry


Drought and other factors

Tolerance of trees to drought and their yield response to water are subjects of much speculation but few data. Leucaena probably maximizes its yields when under no moisture stress.

Physiological studies reveal a high ability to tolerate drought, largely through avoidance responses of leaflet folding and leaf drop. Irrigation in the dry season at the Bharatiya Agro Industries Foundation (BAIF) in Maharashtra State, India, led to continuous function responses to amount of water and tree density. At this institute, Relwani and associates (LRR 4:38) calculated water use to be 3201 kg-1 of wood for trees at a 1 x 1 m spacing. They consider it a superior legume tree for reforesting the Deccan Plateau.

Leucaena tolerates fast fires and can regrow from the crown after burning to the crown by slower fires. Survival was superior to that of N-fixing trees such as Prosopis and Casuarina spp. during fires following the El Nino related droughts of 1983-1985. As agriculture increasingly overwhelms tropical forests, fire tolerance becomes a sine qua non for survival.


Symbionts, diseases and insects

Rhizobia

Rhizobia of leucaena have the reputation of being rather uncommon, rather specific, and predominantly fast-growing and acid-producing, the latter traits being linked inconclusively to leucaena's intolerance of acid soils. None of these features could now be considered inviolable (Bushby, 1982). Few studies seem to have contended fully with the fact that Leucaena leucocephala is an amphiploid (2n= 104), doubtless based on amphiploid parents, with great likelihood of highly reiterated gene sequences. Intragenotypic as well as intraspecific genetic variability is to be expected for traits such as rhizobial strain specificity.

Leucaena seedlings nodulate in most tropical soils, although temperate soils and relatively sterile tropical soils often lack appropriate bacteria. Several laboratories provide inocula, and preference has been shown for strains such as NGR 8 (New Guinea), CB 81 (Australia) and TAL 1145 (Hawaii). Recent reviews of rhizobial research on leucaena include those of Halliday and Somesagaran (1983) and Schroder (1986).

The rhizobia of leucaena are certainly no longer to be considered exceptionally species specific (Trinick, 1980). Those derived from Leucaena leucocephala evidently nodulate all other Leucaena species, with some exceptional reactions involving Leucaena retusa (Thoma, 1983). They also nodulate a number of legumes, including Acacia albida, A. nilotica, A. Senegal, A. raddiana, Calliandra calothyrsus, Desmanthus virgatus, Gliricidia sepium, Macroptilium and Vigna sp. (Halliday and Somesagaran, 1983; Pound and Martinez Cairo, 1983; Dommergues 1982; LRR 2:43). In turn, rhizobia from the following species have been able to nodulate leucaena: Acacia farnesiana, A. mearnsii, A. Senegal, Dalea sp., Mimosa invisa, M. pudica, M. scabrella, Neptunia sp., Piptadenia sp., Prosopis juliflora, Sabinea sp. and Sesbania grandiflora.

The rhizobial reactions of L. retusa are most unusual (Halliday and Somesagaran 1983; Thoma, 1983). This yellow-flowered, frost-tolerant acacia-like shrub of Texas nodulates well in Texas. It did not nodulate in Hawaii until recently, and nodulated ineffectively with a strain derived from L. pulverulenta (Thoma, 1983), suggesting a more specific rhizobium than for other leucaenas.

Leucaena rhizobia are now distinguished from the "cowpea miscellany" group as Rhizobium loti, and are recognized to be quite diverse in type. Most of them have fast growth (3-5 hr generation time) and acid reactions in culture similar to temperate legume rhizobia. They are usually slightly acid or neutral in culture, but range widely in both acid formation and rate of growth. The thesis that alkali-producing, slow-growing rhizobia are necessary for acid-soil tolerance has been challenged by Halliday and Somesagaran (1983). They found only fast-growing, acid-forming rhizobia in the acid soils, and none of the isolates from alkaline soils survived in acid media (LRR 2:71). Despite inoculation with acid-tolerant strains (e.g., TAL 1145), leucaenas did not thrive in acid soils (Chao et al, 1985). Nodule numbers dropped log-linearly with increasing free aluminium in acid soils (LRR 4:54). Genetic variation of leucaena's rhizobia has been reported by many authors, with wide ranges in growth rates and characteristics. Many strains have shown N-fixation as good or better than common inoculants NCR 8 or CB 81 (LRR 2:71; 6:14).

Nodule number and seedling weight are highly correlated in leucaena (LRR2:19;4:18, 57). Nodules appear on seedlings within five weeks if inorganic N is withheld (LRR 2:25). Transplant shock in leucaena results in loss of nodules and re-inoculation (LRR 3:91; 6:95). Strains used in inocula are often recovered with difficulty from soil after a year's growth. Rhizobial persistence is considered fairly high in soil or in compost and charcoal mixtures at room temperature (LRR 5:68; 6:73).

Mycorrhiza

Mycorrhiza inoculation may be more important than rhizobial inoculation when leucaena is transplanted into sterile soils (LRR 2:84; 6:97). Leucaena roots have poorly developed root hairs, and appear to rely heavily on mycorrhiza for nutrient uptake, notably of phosphate. Several species of Glomus and Gigaspora can infect leucaenas (LRR 7:61), and strain differences are significant but variable (LRR 4:83; 7:61, 94). Impressive responses occur for growth of shoot and root and of leaf area on sterilized soils inoculated with mycorrhiza (LRR 4:86, 6:97). Huang et al. (1985; LRR 4:83, 86; 5:79) have conducted extensive studies of leucaena mycorrhiza, finding Glomus fasciculatus best among their isolates. They used single leaflets for phosphate analysis (LRR 5:79). Mycorrhiza plants not only had improved uptake of minerals — notably P, K and Ca — but better stomatal conductance and responses of leaflet folding and orientation (Huang et al., 1985).

Diseases

Although leucaenas have a reputation for high disease resistance, they are not without problems (Pound and Martinez Cairo, 1983). Three diseases have occurred with sufficient severity to attract major research interest — leafspot, gummosis, and seedling rots. Leafspot due to Camptomeris leucaenae (Stev. and Dalbey) Syd. has been known since 1919 and is periodically serious in Latin America. It was studied intensively by Lenne (LRR 1:8; 2:18) in Colombia, who found six Leucaena species to be completely resistant but not L. leucocephala (Pound and Martinez Cairo, 1983). Colletotrichum spp. also occur as secondary leaf pathogens (LRR 3:58; 7:48).

Gummosis from stems has been reported from the Indian subcontinent, and attributed to Fusarium spp., notably F. semitectum (Singh, 1981; Singh et al., 1983) and F. acuminatum Ell. & Ev. The incidence was higher in common and Peru types, but rare in the Salvador types (LRR 3:25,33; 7:48). The fungus may be a secondary pathogen, but is probably systemic and associated with disease symptoms only under stress. Fungicide application controlled the disease (LRR 6:38). Gum exudation of the acacia-type has also been observed without evident pathogenesis, e.g., on the seedless hybrids of L. esculenta x L. leucocephala in Hawaii. These gums may have great potential value, as discussed later in this paper.

Seedling and plant damping-off occurs on leucaenas in nurseries and under poorly drained conditions, involving several common Pythium, Rhizoctonia, and Fusarium spp. (LRR 1:28; 3:58; Pound and Martinez Cairo, 1983). Fusariums were also associated with leucaena stem and root rots in Taiwan (LRR 5:64) and Colombia (LRR 3:14), while Ganoderma lucidum caused root rots on moist sites in India (LRR 4:35; 7:65). Phytophthora drechsleri infections stimulated by storm damage led to cankers and some tree death in Hawaii (LRR 1:56). A collar rot of young trees has been attributed to Sclerotium rolfsii in Florida.

Many different fungi and bacteria (notably Pseudomonas fluorescens) can occur on leucaena pods and seeds, notably following insect attack (LRR 1:8; 3:14; 4:70). None are known to be seed transmissible.

Psyllids and other insects

Insect damage to leucaena is uncommon except to pods and seeds (Pound and Martinez Cairo, 1983). Two insects have caused severe damage to leucaena trees over the past decade — ants and psyllids (jumping plant lice).

Termites or harvester ants make it virtually impossible to establish leucaenas without pesticide use in some areas such as East Africa. Ground-baiting with insecticide-treated grass is one novel solution. In contrast, leucaenas are highly resistant to root-knot nematodes (LRR 4:92; Vicente et al., 1986), but may house species as yet unidentified (LRR 2:17).

Since 1980, the leucaena psyllid, Heteropsylla cubana Crawford, found its way out of its native Latin America, via Florida to Hawaii and thence into the Pacific and S.E. Asia (Figure 4). It has become the most severe pest of leucaena on record, despite the fact that it has been virtually ignored in Latin America due to heavy predation and parasitization (Proverbs, 1985). The psyllid was first described in 1914 and then observed attacking leucaena in the Dominican Republic in 1980 (Martinez Cairo, personal communication), in Florida in 1983 (LRR 5:86), in Hawaii in April 1984 (LRR 5:91), in the South Pacific in 1985 and in S.E. Asia and Australia in 1986 (LRR 7:6).

A section of Volume 7 of LRR was dedicated to invitational papers reviewing the status of the psyllid. W.C. Mitchell and D.F. Waterhouse (LRR 7:6) traced the spread of the psyllid. An international symposium was conducted by NFTA on "Biological and genetic control strategies for the leucaena psyllid" in Hawaii in late 1986, and a special issue of LRR was dedicated to the proceedings of this workshop (LRR 7(2): 109). Major review papers included those by J. Beardsley on the psyllids (LRR 7:2,7(2): 1), by L. Nakahara et al. on the predators (LRR 7(2):39; 7:9), and by Sorennson and Brewbaker (LRR 7:13; 7(2):29), Bray (LRR 7(2):32) and Pan (LRR 7(2):35) on resistance. High psyllid resistance characterized several leucaena species and their hybrids (Table 2).

The leucaena psyllid has caused a severe setback to fodder production in S.E. Asia, with less obvious but significant effects on wood production. Native predators and parasites appeared to be rare or slow to find the insect. Only partial control was effected in Hawaii by two predatory beetles, both introduced early this century to control other pests. A Caribbean parasitoid, Psyllaephagus, however, shows specificity and excellent appetite for H. cubana, and may prove to be the major biological control agent of that region (LRR 7:9). It was released in Hawaii in June 1987. Genetic control is expected to provide the only viable option for some areas of the world, if predators cannot be introduced. Rapid seed increase and deployment of psyllid-resistant germplasm has been advised (LRR 7(2):88). Seed insects are common pests of leucaena, including the seed borer Araecerus fasciculatus (LRR 4:70), Cathartus grain beetles and bruchids (Pound and Martinez Cairo, 1983). Other pests appear to be largely under biological control where earlier reported, such as the black twig borer, mealy bugs of pods, and the pantropical moth Ithome lassula whose larvae feed on the florets (LRR 2:11). Three scale insects caused damage on Taiwanese leucaenas, of which Hemiberlesia implicata was the most serious, involving as many as 20 percent of trees sampled (LRR 3:55). Leaf feeding by the beetle, Apogonia rouca, was reported in India (LRR 7:67). In Hawaii the rose beetle, Adoretus sinicus, may feed at night on the large-leaflet Leucaena spp. (Figure 2). Several types of inchworms have also been observed on leucaenas around the world, but none appear to be serious. The psyllid problem, however, should put scientists on alert for insect pests and diseases that may move pantropically and present new challenges in the future for leucaena production and improvement.

6_Leucaena a multipurpose tree for agroforestry


Establishment and management

Establishment and fertilization

Establishment methods for leucaena have been refined and are often shown to be site-specific. Direct seeding has been recommended where soil moisture conditions permit and economic weed control can be maintained (LRR 4:78). The mature seeds have a water-impermeable coat, carry 5-7 percent moisture, and store well at room temperature (LRR 2:59; 3:87; 4:67). Seed germination continues to be the subject of articles on methodology despite the thorough coverage in early literature (NFTA, 1985a,b). Co-planting with a crop like maize can produce excellent results (LRR 1: 19), and seeds have been successfully mixed in the planter box of machine drills (LRR 7:26). We use alachlor as a preplant herbicide for such mixed plantings. Hormonal treatments of seed or seedlings can accelerate early growth slightly (LRR 1:53; 3:83). On the farm, leucaena is often much easier to plant than to protect due to its attractiveness to grazing animals; thorn hedges and wire enclosures can be used.

The use of bare-root ("stump", "bare-stem") cuttings from seedling beds has worked variously well in Taiwan (LRR 3:45), Honduras (LRR 6:20), Thailand (LRR 4:78) and Java, Indonesia (LRR 3:45). An important trick used in Taiwan is to roll the cuttings (10 cm below ground, 25 cm above) in mud, drying them before shipping to planting site. Cuttings can be stored two weeks under wet burlap without loss (LRR 3:45).

Fertilization practices for leucaena are site-specific, but routinely involve recommendations of phosphate and lime applications to accelerate early growth. Significant responses to both Ca and P have been reported on a wide array of tropical soils (Brewbaker and Button 1979, Pound and Martinez Cairo, 1983). Responses to Ca and pH have been detailed in an earlier section. Responses to P are less predictable, and the effectiveness of mycorrhizal associations must be involved. No responses to K have been reported. Sulphur response can be great, and gypsum and rock phosphate are recommended for S-deficient acid soils also low in Ca or P. In low-fertility soils generally, Hutton has recommended application of 100-200 kg ha-1 superphosphate and 100-200 kg ha-1 dolomitic lime plus supplemental Mo, Zn, Cu and Bo. Mineral deficiency symptoms occurred at about 50 percent of normal leaf concentration for K, Ca, Mg and S in leucaena, and at about 80 percent of normal for N and P (LRR 1:6).

Production and tissue culture

Vegetative propagation of leucaena has been successful in relatively few locations, apparently reflecting critical environmental requirements or possibly systemic fungi. Hu and colleagues have mastered most field management problems with leucaenas in Taiwan (Hu, 1978; Hu and Kiang, 1983). Hu has directed replicated trials on every aspect of leucaena management for pulpwood, providing careful cost analysis of each step (Hu and Kiang 1983). Hu obtained excellent rooting of cuttings under mist spray in Taipei greenhouses, using 15 cm leafy twigs having terminal shoots intact from one-year-old trees (LRR 2:50). Leafless twigs from older trees failed to root; this has been the experience in Hawaii also.

Tissue cultures of leucaena have gone through the traditional phases of development, but cannot be considered field applicable. Callus growth was achieved from germinating seeds (LRR 1:54), and hormonal treatments were perfected leading to shoot differentiation and somatic embryoids (LRR 4:88). Seed-derived explants were carried without callus formation through BA (benzyladenine) and NAA (naphthalene acetic acid) media to vigorous shoots (LRR 3:81), and with added ascorbic acid to healthy plantlets (LRR 4:37). Explants from mature tissues have not been carried through tissue culture. Nodal branch cuttings will produce shoots on BA media, and rooting was obtained by subsequent transfer to media with NAA (LRR 5:22) or with auxin treatments of shoot explants prior to rooting in charcoal-supplemented media (LRR 5:37). Further refinements, including embryoid and protoplast culture and adaptation to large-scale propagation, are in progress LRR 6:32).

Grafting is a relatively simple and historic technique for leucaenas (Dijkman, 1950), but has been refined in the lab of Versace (LRR 3:3). The best results involved cleft and whip and tongue grafts, rather than bud or T-shield grafts, using four-month seedlings as stock and taping with grafting tape. Several types of interspecific grafts were successful.

Nursery methods

The most effective nursery methods employ long narrow plastic tubes with 100-500 g of medium. A potentially serious tapered open base on dibble tubes leads to aerial root-pruning and excellent taproot growth when transplanted (LRR 1:57). Dibble tubes and "root-trainers" are now extruded in several countries with ribbing to prevent coiling. A fine layer of crushed stone or sand on top of the soil or soilless mix (1 part vermiculite to 1 part peatmoss is common) prevents damping-off and weed growth. Growth is superior in larger tubes, but nursery and transplanting costs increase accordingly. Yields increased with increasing pot size (up to 40 kg per pot) in both sand and a black soil, with heights of four-month-old seedlings ranging from 20 cm in 500 g pots to 2 m in 40 kg pots (LRR 6:12). Foam rubber 3 cm cubes also worked well for leucaena seedlings (LRR 7:88).

Nursery fertilization regimes have been refined to optimize growth. Urea is minimized to levels that maintain good growth but do not fully suppress nodulation (0.5 g per seedling, LRR 7:38), and phosphate levels are critical (1 g superphosphate per seedling). Growth of 6-8 weeks under cover followed by 6-8 weeks without cover is commonplace. Full sunlight creates seedlings that are stunted but tough.

Weed control

Weed control is the major expense in tropical tree establishment, even with fast-growing leucaena at high density on good soil. Simultaneous planting for fodder with grasses such as Brachiaria, guinea grass (Panicum maximum) or Paspalwn spp. can give effective weed control (Pound and Martinez Cairo, 1983). Establishment of solid-stand forests requires herbicides or weeding by hand or machine. Herbicide recommendations are imprecise for leucaena, in part due to differing experiences with levels of toxicity and effectiveness of weed suppression. Land preparation prior to planting is critical in reducing weed and weed-seed populations. High-input methods prior to transplanting in Hawaii involve disking, application of glyphosate (Roundup) to regrowth and use of pre-emergence herbicides.

Pre-emergence herbicides that are recommended in the literature include alachlor, bentazone, dalapon, metabenzthiazuron, monuron, nitrofen, phenoxalin, simazine and trifluralin, to which leucaena is fairly or highly resistant. Especially effective treatments have included oryzalin (Surflan) at 2.8 kg ha-1 (LRR 5:105), simazine at 5 kg ha-1, phenoxalin at 3.5 kg ha-1 (Pound and Martinez Cairo, 1983), nitrofen (Tok) at 4.5 kg ha-1 (LRR 1:50), and trifluralin (Treflan) at 1.5 kg ha-1 (LRR 1:50). Post-emergence grass control with fluazifop (Fusilade) is effective at 2 kg ha-1 over the leucaena (LRR 6:1) and bentazone (Basagran) at 2 kg ha-1. Fluazifop is currently the favoured post-plant herbicide for many types of trees. Simazine, dalapon, diuron and oxyfluorofen have also been used post-emergence for grass control (LRR 5:105).

Fodder and wood management

Fodder yields of leucaena from different varieties and different management systems are discussed in a later section. Low hedge management with 2-4 month cutting intervals is now about standard for leucaena, moisture permitting (Figure 5). Nutrient replacement regimes have not been well defined for solid fodder plantings ("protein banks") of leucaena, but must be devised to restore nutrients to the soil that are lost in harvest (Brewbaker and Hutton, 1979). Notable are the losses in fodder harvest of soil P, K, Ca, Mg and selected micronutrients. Alley farming with animal presence in the field ensures partial return of such nutrients, as do any alley-cropping systems involving complete return of leucaena foliage as green manure (although this rarely occurs).

Wood is harvested from leucaena in almost all ways known, on cycles ranging from 6 months to 10 years, and including use of roots and small branches, so that generalizations are difficult. Wood yields maximize at high population densities of 10,000 ha-1 on 2-4 year cycles that also minimize weed competition. Leucaenas coppice readily with rare loss of a tree, and are cut for fodder or debranched in many ways. Coppicing studies are poorly recorded in the literature. Pecson (personal communication) observed no significant differences among methods of coppicing on different varieties and densities. Leucaenas had a strong tendency to return to a fixed stand of co-dominant stems at any one site irrespective of management and varietal variables. Under Hawaii conditions of 1,500 mm rain per annum, the arboreal varieties stabilized around 15,000 co-dominant stems ha-1, and numbers of coppice shoots were directly correlated with stem diameter. One-year trees coppiced at two sites in Thailand averaged about 5,000 stems ha-1 at one site and 19,000 ha-1 at a second (LRR 5:70). Branch wood is considered valuable as fuelwood in some systems, and branch numbers varied from 30-80 per tree at 11 densities in India (LRR 6:23).

7_Leucaena a multipurpose tree for agroforestry


Use as animal feed

Almost half the published literature on leucaena deals with its use as animal fodder. The increasing role of meat and animal products in tropical diets is clearly a cause, as is the increasing market for leucaena leaf meal (LLM) in temperate countries. Tropical poultry feeds have long used leucaena as a yolk-colouring device, and attempts are being made to improve the digestibility and energy value of LLM as feed for all non-ruminant animals. A primary interest in ruminant use has been in the toxic by-product of mimosine digestion, dihydroxypyridine (DHP).

Fodder yields

Leucaena fodder yields vary greatly for different ecosystems but less so for different management systems, and good summaries exist of the extensive early literature (Oakes 1968; Brewbaker and Hutton, 1979; Pound and Martinez Cairo, 1983). Fresh herbage yields exceed those of other shrubby tropical legumes (LRR 4:77; Brewbaker, 1985a, 1986b). Yields are comparable to the best herbaceous legumes, ranging from 40 to 801 ha-1 when moisture is not limiting (LRR 2:19; 3:39; 6:40; 7:19; Brewbaker et al, 1972; Brewbaker 1976; Hedge, personal communication; Hogberg and Kvarnstrom, 1982) and from 20 to 50 t ha-1 in seasonally dry tropics (LRR 3:31; 4:25, 31, 69, 77, 79; 5:3) and in frost-affected subtropics (Othman et al., 1985). Much lower herbage yields usually reflect serious constraints of soil fertility (Chee and Davendra, 1983).

Variety, harvest intervals, cutting heights and planting densities are major variables for the hedge management considered here. Favoured varieties are the Salvador or Peru types as they outyield common varieties by 20-100 percent (Brewbaker et al., 1972; Brewbaker, 1976; LRR 2:19; 3:39; 4:3). When cut to very low stubble heights, yields are reduced for all varieties but the common types perform relatively well (Guevarra et al., 1978). L. diversifolia produced competitive fodder yields in cooler climates with less than 20° C mean temperature (LRR 3:1), and its hybrids with L. leucocephala and those of L. pulverulenta x L. leucocephala (LRR 4:1) were excellent (Bray, personal communication).

Harvest intervals produce significant variations in fodder yield and quality. Some varieties branch and flower rapidly and must be harvested earlier than the favoured Salvador and Peru types. Percentage leaf in herbage dropped as cutting intervals increased (from 67 percent at 30 days to 42 percent at 150 days) in Mauritius studies by Osman (LRR 2:33,35; 3:49; 7:91). Edibility by cattle, as measured by weighing uneaten fractions, was reduced in proportion to leafiness from 100 percent at 30 days to 85 percent at 120 days (LRR 7:91). Yields maximize at 70-90 day harvest intervals (LRR 3:31; 4:25; 7:91) depending on temperature. Yields on a per-day basis were highly correlated with temperature in Australia (LRR 5:3; Bray et al., 1984) and Hawaii (Brewbaker et al., 1972; Hedge, personal communication).

Cutting or stubble heights have been evaluated at BAIF in India, and lead to a consistent recommendation for maintaining hedges above 60 cm (LRR 4:25,41). Cutting heights above 60 cm allow retention of some green foliage and vigorous lateral meristems. The common shrubby leucaenas can be cut to 10 cm without appreciable yield loss (Guevarra et al., 1978), but the more arboreal varieties should not be cut below 25 cm (LRR 4:3,41; 5:3; 7:91). Widely spaced high hedges can provide high-quality fodder (LRR 4:69) at the expense of yield.

High plant densities are recommended for solid fodder plantings, 50-150 x 103 ha-1, if moisture is not limiting. Yields maximize when the trees are coppiced to produce a highly branched shrub that can grow quickly to intercept all light on the field after harvest. One metre between rows and 0.1 m between plants is a standard for comparison (LRR 3:40).

Fodder quality

Leucaenas are among the highest quality fodder trees in the tropics (Brewbaker, 1985a). Pound and Martinez Cairo (1983), Jones (1979) and Brewbaker and Button (1979) summarized many publications on herbage, leaf meal and seed meal. Herbage taken at peak quality has the following percentage values of dry matter: digestibility 55-70, crude protein 20-25, ash 60-10, N-free extract 30-50 (fibre 25-35, NDF 20, ADF15, cellulose 10, Ugnin 5), fat 6, mimosine 4.5, tannins 1.5-2.5, Ca 0.8-1.8, P 0.23-0.27, and silica < 1.

All leucaenas are known to be highly attractive to animals as browse and feed. Introduction of cattle to the Americas probably decimated leucaena populations, although deer damage can be great (buffaloes were north of the leucaenas). Removal of feral cattle pressure can lead to explosive populations of the common seedy types. Extensive analyses have been made of fodder components of L. diversifolia, L. lanceolata, L. pulverulenta, L. shannoni and L. trichodes at the Indian Grasslands and Fodder Research Institute (LRR 7:43) and the Tamil Nadu Agricultural University, India (LRR 3:21), showing only minor variations in fodder quality and mineral contents (LRR 2:53; 7:43), often related directly to tissue maturity (LRR 4:24). Varietal differences in Leucaena leucocephala were also small (LRR 5:14), and appeared to be related more to stage of growth than to genotype. High digestibility and comparability of leucaena's amino-acid composition to that of alfalfa are stressed in such studies (LRR 2:53). Proximate analysis of seeds and pods revealed values comparable to foliage except for elevated fibre levels (LRR 3:21). Feed mills for the processing of leucaena into pellets or meal are becoming common, and the pellets have an international market (Manidool, 1983; PCARR 1978; FAO, 1985).

Dihydroxypyridine (DHP) and mimosine

Toxins are a cause célèbre for this generation, and leucaena's mild toxin mimosine still frightens many animal growers into ignoring this "alfalfa of the tropics" (Holmes, 1981). Mimosine is an amino acid found in high concentrations in leucaena tissues (2-6 percent of dry weight) whose toxicity has been covered thoroughly in the early literature. A series of definitive experiments in ruminant nutrition by Raymond J. Jones and his colleagues has clarified most doubts about mimosine and its DHP breakdown products. These studies permit accurate prediction of ruminant nutritional problems and provide a ready cure.

DHP was known to be a product of mimosine catabolism. A key discovery by Jones and colleagues was that DHP was goitrogenic (Hegarty et al, 1979), its action marked by elevated levels of serum thyroxine in urine (Jones and Bray, 1983). Secondly, it was shown that further catabolism of DHP was not completed normally in ruminants of Australia, Papua New Guinea and certain other regions in the world, its accumulation being marked by loss of appetite, goitre and related toxic symptoms. It was shown that mimosine was converted to 3,4 DHP by enzymes in the leaf (LRR 2:31) and during animal chewing (Lowry, 1983), or in the rumen. A second isomer, 2,3 DHP, could also be formed in the rumen from 3,4 DHP (LRR 5:2). Megarrity (1978) then isolated a strain of bacteria from Hawaiian goats that degraded the DHP. They transferred the bacterium successfully to unadapted ruminant animals and reversed leucaena's toxicity. International service was provided by Jones to identify those regions of the world where DHP toxicity occurred, i.e., lacked the appropriate bacteria.

Fodder use is a significant feature of many, perhaps most, agroforestry species. Knowledge of toxins and their detoxification, e.g., as in leucaena, may become of great value to their deployment. The Jones' model should serve as a classic in the field of ruminant toxicity of tropical trees, many of which contain alkaloids, tannins and flavonoids at high levels as part of their defence mechanism against herbivores and tropical insects. Ruminant bacterial degradation is expected to become an option for some of these toxins, following identification and introduction of appropriate bacteria essential to control toxicity. It. is probable that genetic removal of an alkaloid like mimosine would lead to unacceptable levels of insect susceptibility, but this has not been verified. Its concentration could be halved by plant breeding (Gonzalez et al., 1967). Mimosine's chemical properties and activities are subjects of extensive continuing research. Mimosine is not carcinogenic or mutagenic, but inhibits many enzymes (LRR 3:67; 5:72) involving at least three mechanisms — chelation of metals, interaction with pyridoxal-phosphate enzymes, and inhibition of DNA synthesis (LRR 6:63). It does not act as a trypsin inhibitor. It can act as a fungicide on pathogens such as Sclerotium rolfsii (LRR 5:73), but is better known for its depilation of horses in the tropics.

Methods have been refined to quantify mimosine and DHP. The classical ferrous sulphate colorimetric method for mimosine of Matsumoto and Sherman (as modified in Hawaii, LRR 2:66) was automated by Megarrity (1978) and modified by Jones (LRR 1:3) to collect samples directly into HC1 to minimize loss of activity. Mimosine elutes with isoleucine in amino-acid analysis, and performic acid treatments destroy the mimosine, thus permitting its estimation by this method (LRR 3:78). DHP could also be estimated through ascending paper chromatography (LRR 1:16), and ion-exchange chromatography in Na citrate buffer gave an excellent estimation of mimosine and DHP (LRR 1:39). High performance liquid chromatography for both compounds was performed at very low concentration in the laboratories of Acamovic and D'Mello in Edinburgh (LRR 2:62; 6:75) and Tangendjaja and Wills (1980) in Indonesia. The extraction of crystalline mimosine and of DHP was also improved by resin column methods (LRR 5:50). Like other free amino acids, mimosine values maximize in juvenile leaves on vigorously grown plants before the dry weights stabilize (LRR 7:34).

The reduction of mimosine levels by pretreatments of feedstuffs has long been of interest, but must now be offset by the interest created in DHP. Mimosine is converted to DHP by an enzyme that is inactivated by heat or by drying in leaves (Lowry, 1983, LLR 7:77). Animal chewing thus initiates conversion that is stopped in the stomach of monogastric animals. Simple leaching with standing or preferably running water lowers mimosine dramatically. Sun drying under hot wet conditions leads to significant loss, and can be accelerated by heat under pressure (LRR 4:62). Silage processing reduced mimosine to 50 percent (LRR 1:17). The addition of sugars and organic acids accelerated this degradation down to 10 percent (LRR 7:85). Dietary supplementation to reduce mimosine toxicity is discussed in one of the later sections. Lowry (1983) stresses that conflicting leucaena toxicity data in ruminant and non-ruminant animals clearly involve mimosine and DHP.

Ruminant animals

A most thorough review of ruminant animal use of leucaena has been provided by Pound and Martinez Cairo (1983), who stress the role of leucaena as a supplement to diet rather than as its basis. A value of 30 percent of diet is adequate to fulfil its primary role as supplemental protein, roughage and mineral source. In any location it must first be determined that DHP is catabolized by ruminant animals (i.e., no elevated thyroxine level in urine, no loss of appetite on leucaena) or appropriate action taken. Even when fed 100 percent leucaena plus salt supplement, mature animals able to detoxify DHP remain thrifty and gain weight (LRR 7:26; Pound and Martinez Cairo, 1983).

A major agroforestry system with leucaena involves discrete blocks of leucaena ("protein banks") and of grass, or leucaena plants or hedges in a grass pasture. Liveweight gains often maximize on leucaena-supplemented pastures, with several studies reporting gains of 0.6-0.8 kg per day, the range influenced by dietary energy levels (Jones and Jones, 1982). Milk production in the tropics is often based on pen-fed animals receiving diets of low palatability, thus supplemental leucaena leads to increased milk yields (LRR 2:39,40; 3:21). Gains of suckling calves given supplemental leucaena and molasses/urea were comparable to controls in Mexico, Mauritius and Dominican Republic trials, but at less expense (Pound and Martinez Cairo, 1983). Leucaena hay is widely used as a supplemental feed to goats, fattening or milk cattle, and buffaloes with good response (LRR 3:99; 5:12; 7:99). Many short reports of the use of leucaena in tropical pastures or hay rations appear in the journal Tropical Animal Production.

Non-ruminant animals

Leucaena leaf meal (LLM) and leucaena seed meal (LSM) are prepared by grinding leaves and seeds as poultry and small-animal feeds. Both products have led to problems when fed at high levels, problems now ascribed as much to tannins and low metabolizable energy (ME) as to mimosine. Extensive early literature (D'Mello and Taplin 1978) records the widespread use and problems of leucaena leaflets or leaf meal as a poultry-feed supplement.

Broiler gains and egg laying are reduced by levels in excess of 40 g kg-1 of feed (LRR 2:41, 2:47). At this low intake, xanthophyll contents are so high (750 mg kg-1) that egg colouration is much better than with yellow maize (LRR 6:76), even though less than 50 percent of the xanthophyll is biologically available (Taplin et al. 1981; LRR 6:76,99). Use of higher levels of LLM reduced weight gains and slowed maturity rates, with few other toxic symptoms. Causes cited for these effects include mimosine, tannins, low metabolizable energy and saponins.

Daily feeding of mimosine at a level equal to that of 200 g kg-1 of LLM did not delay chick development, however (LRR 7:83). While other studies have indicated that ferric sulphate reduces toxic effects of LLM, including mortality and weight loss (LRR 6:70; 2:60), this may not result from mimosine precipitation per se. D'Mellow and colleagues in Edinburgh have conducted a series of LLM treatments to reduce nutritional toxicity. Aluminium salts given to chicks on LLM led to full excretion of mimosine, less completely so with Fe2 and Fe3 salts (LRR 1:38). Ferric sulphate reduced the depression in chick growth by half (150 g kg-1 LLM), and added polyethleneglycol (PEG) reduced this further by half (LRR 2:60).

Tannin levels of leucaena leaves are not high (1-1.5 percent), but are presumed to be neutralized by PEG. Tannins in seeds (1.5 percent) were reduced 75 percent by 30 minutes cooking in water, and over 85 percent by boiling in NaOH (LRR 7:77). The addition of cholesterol to LLM chick diets led to normal chick growth (LRR 3:72), suggesting that saponic toxicity was involved. However, LLM did not reduce plasma levels of cholesterol in the studies of Tangedjaja and Lowry (LRR 3:57) as would be expected if saponins were a problem. Trypsin inhibitors were invoked as a possible cause of the LLM toxicity, and were eliminated by heat treatments of 30 minutes at 120° C (LRR 7:97). Heat treatments of LLM under pressure are known to improve protein digestibility and weight gains of rats, also eliminating mimosine (LRR 4:62). Heated and ferric-treated LLM had 50 percent mimosine reduction (LRR 7:97). Rats showed reduced weight gains above 15 percent LLM and above 20 percent if LLM was sun dried, but ferrous sulphate treatments had no effect (LLR 4:59). Szyszka (LRR 5:5) calculated non-toxic average daily intake values (g kg-1) for mimosine, based on LLM or LSM, for the non-ruminant animals — broilers 0.16, layers 0.35 and rabbits 0.30.

Metabolizable energy (ME) of LLM is very low, about 15 percent of the expected 20 MJ kg-1 dry matter (LRR 2:63). The causes again are elusive, and include gums, fibre, tannin, mimosine and DHP. Fat or tallow is added normally to increase ME of LLM-supplemented diets. Use of a hemicellulase to digest the gums did not affect ME (LRR 4:82).

Leucaena seed meals have been successfully fed to ducks (LRR 7:24) and chicks at modest intake levels. Lowered mortality and improved growth on the heat-treated meal (boiled one hour) occurred with chicks, but ferrous sulphate treatments were ineffective (LRR 3:66).

Swine cannot be fed high levels of LLR unless mimosine levels are very low or iron is fed, notably during gestation in gilts (Pound and Martinez Cairo, 1983). Leucaena can be fed up to 10 percent in fattening rations and provides a valuable source of vitamin A. Swine intestines have been used in permeation studies in which mimosine permeability was inhibited by ferric ions and also by low levels of lactic acid (LRR 7:86). When LLM was heat-dehydrated and ground finely, mimosine levels averaged 1.5 percent and no deleterious effects were observed in swine diets up to 16 percent LLM (LRR 2:46). Digestibility coefficients were reduced about 20 percent for dry matter, protein and gross energy on diets supplemented with 35 percent LLM (LRR 3:76). A maximum of 20 percent LLM was suggested for weaner pigs, with slight improvements in gain noted upon addition of 1.5 percent ferric sulphate or 3 percent polyethylene glycol or both (LRR 2:74).

Rabbits also tolerate low levels of LLM, which they accept readily, but feeding results are variable. Liveweight gains dropped linearly as LLM levels were raised to 100 percent of diet (Pound and Martinez Cairo, 1983). Molasses supplementation showed that low metabolizable energy was a major cause. Feeding studies with mimosine supplementation showed unusually high tolerance (>0.24 mg kg-1 body weight per day) (LRR 5:7). Leucaena compared favourably to alfalfa in digestibility of crude protein, fibre and cell-wall constituents, but voluntary consumption was low and urine turned red-black (LRR 2:73). It must be stressed that DHP has not been distinguished from mimosine in non-ruminant nutritional studies, and may be the basis for conflicting results from LLM and mimosine-feeding trials.


Soil improvement

Leucaena groene bemester (Dutch for "green manure") was the subject of many articles in the early 1900s (Dijkman, 1950), and represents perhaps the most important of leucaena's many uses in agroforestry systems. These articles referred to the use of leucaena and other legumes as companion trees ("shade trees", "nurse trees") in coffee and other plantation-tree crops. In recent times, use of leucaena in "alley farming" is also assuming great importance as the tropics are progressively denuded (Kang et al, 1984; Kang and Wilson, this volume).

Alley farming

Alley farming is an agroforestry system of growing row crops between closely spaced woody hedges (Kang et al., 1984). An historic practice in Indonesia and the Philippines (Figure 6) involved maize planted between contour strips of leucaena (LRR 1:13, 20). Dairies in Hawaii had an alley-farming system of pangolagrass (Digitaria decumbens) between leucaena hedges before 1940, allowing animals free access to both grass and legume (Takahashi and Ripperton, 1949). Agronomic research has led to management systems that optimize total yields from these systems, with demonstration of their effectiveness in a wide array of cropping systems (Kang et al, 1984). Maize and leucaena are very compatible in such a system. Leucaena hedges are cut and the green manure ploughed under prior to planting, then cut again in 6 weeks to eliminate competition for light and provide added nutrient-rich green leaf manure to the maize. After maize harvest, the leucaena can be cut for fodder and fuelwood, the hedge acting to reduce erosion and build terraces. The system has been called lamtoronisasi in E. Indonesia, where Viator Parera has guided demonstration and improvement of the practice (LRR 1:13). Stem-girdling of the leucaena to induce litterfall and allow light penetration during the critical grain-fill period is recommended by Parera (LRR 4:45; 5:51).

Trials at the International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria, have included several shrubs suitable for alley cropping; leucaena has generally been superior. Alley widths have been varied and 4 m preferred, with intra-row spacing to create a dense hedge. Pruning every 5-6 weeks is advised, with periodic low cutting to stimulate branching. Tractor-pulled rotary mowers, tilted slightly, have been very effective where available. The hedge is managed to maximize shade-control of weeds and minimize later shading of the intercrop. Continuous alley cropping for six years showed that maize yields could be kept at 2 t ha-1 with leucaena prunings alone (Kang et al, 1984). Summarizing nitrogen yields of hedgerow fodder, Torres (LRR 4:50) calculated an average of 45 g N per one metre of row, and suggested this was adequate for yield improvement only on marginal farms. With an average tropical maize yield of only 1.2 t ha-1, however, any procedure increasing grain yields to 2 t ha-1 under continuous rather than periodic cropping could have a substantial impact on tropical maize production (Brewbaker, 1985b).

8_Leucaena a multipurpose tree for agroforestry


The Indian Grassland and Fodder Research Institute at Jhansi has also conducted a series of leucaena alley-farming trials with crops such as maize, sorghum, buffel grass (Cenchrus ciliaris), millet and napier grass (Pennisetum purpureum) with generally salutary results (LRR 3:29, 30; 4:20; 5:24; 6:36). Competition for moisture can reduce yields of the row crop, so the total value of crop and hedge becomes a determinant in farmer acceptance (LRR 2:30; IITA, 1986). Similar results were obtained with leucaena and sweet potato, where light interception by the hedge reduced crop yields but crop plus fodder made a better total cash return (LRR 3:52). Leucaena can be used with kenaf (Hibiscus cannabinus) as alley crop (LRR 6:1) and with cassava, but was not effective combined with sugarcane (LRR 6:35).

Hedge trimmings from leucaena can be carried as excellent green manure to row crops like maize and rice (LRR 2:41,42). Leucaena leaves mature over a period of 2-4 weeks and the leaflets, pinnae and midribs dehisce in 3-5 months. The litter is fragile and quickly decomposed, with N half-life of 7 days if buried (Guevarra et al, 1978; Kang et al, 1984). Soil-incorporated leaves completed N-release in three weeks in India, with somewhat higher values under aerobic rather than anaerobic conditions (LRR 3:54). Maize grain yields to leucaena green manure regressed linearly on application rates over a wide range (C. Evensen, cited by Brewbaker, 1985b). One kg of N in leucaena mulch produced an increase of 14.8 kg of grain, while one kg of N as incorporated green manure produced an increase of 24.1 kg of grain. In other studies, excellent yield responses of maize to green manure have been reported (LRR 1:13, 22; 4:33; 5:59; Kang et al., 1984).

Companion tree

Companion or "nurse" trees are planted for two major purposes with coffee, tea, cacao, teak and other tree crops — to confer shade and wind protection in early growth, and to provide nitrogen, mainly through N-fixation. Leucaena was a major coffee and cacao shade tree prior to modern use of inorganic nitrogen. Successful companion planting requires care in spacing and management. The Salvador leucaenas are generally too vigorous as nurse trees unless boldly coppiced, while L. diversifolia (4N) and its hybrids deserve wider appraisal. Examples are reported, not all successful, of leucaena as nurse tree to teak (LRR 1:21), mahogany (LRR 1:5,21; 7:82,102) and coconut (LRR 5:62,7:72). Its use for social forestry plantings along with eucalyptus would seem well advised.

Van Den Beldt assessed both litterfall and biomass development of L. leucocephala in Hawaii under stand densities ranging from 10 x 103 to 40 x 103 ha-1 (LRR 3:95). Annual litterfall averaged 8.54 t ha-1 (dry matter) with no significant effects of stand density. Nutrient analysis of the litter showed that 100 kg N, 7 kg P, 16 kg K, 200 kg Ca and 12 kg S were recycled annually per hectare through such litterfall.


Tree and wood use

Leucaena production for wood use expanded most prominently in the past decade in two ways — for pulpwood in Taiwan (Hu, 1987; Tai et al, 1984) and for dendrothermal energy in the Philippines (Denton, 1983). Although much less heralded, the use of leucaena in village agroforestry systems for fuelwood, for poles (e.g., propping bananas) and for housing and furniture wood has probably expanded more rapidly as a result of the spread of the arboreal varieties (NRC, 1984; NFTA 1985b).

Wood yields

Wood yields of Leucaena leucocephala are greater than those of species with which it has been compared in most short-duration (3-5 yr) trials. Height growth and mean annual wood increments fall in the range of 3-4 m yr-1 and 20-60 m yr-1 for many trials. Only the Salvador types are now grown, with wood yields many times that of the common types (LRR 6:49). Failures are almost entirely related to two factors — poor soil (acid, low Ca, low P) and low temperatures. Fresh-weight data from 2-4-year-old trees illustrate the limitations of L. leucocephala in comparison with an acid-tolerant acacia (Acacia auriculiformis) (Table 3). In cooler climates the use is recommended of species such as L. diversifolia whose hybrids with L. leucocephala greatly outyield both parents in the cooler tropics (Brewbaker, 1986a). L. collinsii is a fast-growing species that deserves evaluation in cooler tropics, while L. lanceolata var. sousae has promise for lowland tropical wood use.


Table 3 Wood yields (fresh) of Leucaena species and Acacia auriculiformis in diverse environments

9_Leucaena a multipurpose tree for agroforestry

The forte of leucaena in agroforestry systems is its ability to maintain high yields over a wide range of harvest cycles, stand densities, and systems of management and mismanagement. Van Den Beldt and associates have conducted extensive collaborative trials of arboreal varieties under diverse spacings (Hu and Kiang, 1983; Van Den Beldt, 1983; LRR 1:29,53,55; 3:62,96; 4:93). Annual yields changed relatively little over a wide range of densities and age at harvest, optimizing between 10 and 20 thousand per hectare. Optimal spacing changed in relation to harvest age, with a drop in mean annual increments occurring earlier at high densities. Increased densities did not affect height appreciably; instead, the plant diameter, specific gravity and wood strengths were lower, and relative moisture and amount of bark were higher at the high plant densities (Van Den Beldt, 1983; Hu and Kiang, 1983; LRR 3:68).

Volumetric formulae for leucaena have been produced from many trials. Van Den Beldt (LRR 4:93) compared these and concluded that few deviated significantly from the formula of Kanazawa et al. (1983): tree volume=0.5 D2 H. In this formula, tree volume is in m, D is diameter at breast height in m, and H is height in m. Leucaena trees at 1 x 2 m spacing grow with D in cm roughly equal to H in m, and thus a 10 m tree is about 0.05m3 (about 50 kg wet weight) and a 5 m tree about 0.006 m3 (6 kg).

Hu (LRR 1:29) and Van Den Beldt (LRR 4:93) calculated form factors relating stem volume to a cylinder with the same dimensions under varying conditions of tree age, density and site. These factors were very similar except for very young trees (1.5 yr), and generally fit well the Kanazawa equation.

Log transformation and corrections for intercept improved the fits only slightly. Our data as t ha-1 yr-1 in Table 3 use equations that recognize Y axis intercepts when relating yield in kg per tree (Y) with the formula: Y=0.5+46.5 D2H for LEU, and Y=0.72+46.5 D 2H for DIV where LEU and DIV refer to L. leucocephala and L diversifolia, respectively (Brewbaker, 1986a). Hu (LRR 3:62) found the Kanazawa equation suitable for Taiwan trials in which five different Salvador varieties were statistically similar in yield (mean 32 m3 ha-1 yr-1). Log transformations have also been used to compute total biomass dry matter a bit more precisely: in one example: log wt. = -0.462 + 0.663 log (D2H) (LRR 7:53).

Wood yield data are now widely available for Salvador leucaenas in different sites all over the world. They range up to 100 m3 ha-1 yr-1 (Table 3), and would average about 15 m3 ha-1 yr-1. Some of the more extensive data involving varietal or spacing differences are from Thailand (LRR 4:81), BAIF in India (LRR 4:38; 40,49, 50), Jammu in India (LRR 2:22; 3:27), the Philippines (LRR 1:27; 6:82), Taiwan (LRR 2:53; 3:59, 62), Costa Rica (LRR 3:15), Jamaica (LRR 6:60), Hawaii (LRR 1:55; 4:93) and Florida (LRR 5:84), the latter being impressive despite winter frosts that killed plants to the crown.

Fuelwood

Leucaena has become a popular fuel wood in Asia and less widely so in the rest of the tropics (Figure 7). Fuelwood properties are widely respected and well documented in the literature (NRC, 1984; NFTA 1985b; Pound and Martinez Cairo, 1983). Heating values on oven-dry basis average about 19.4 MJ kg-1 (4,640 kcal kg-1) and the wood burns steadily with little smoke, few sparks, and less than 1 percent ash. Moisture at harvest depends on tree age, ranging from about 55 percent at 1 year to 45 percent at 4 years and 35 percent at 7 years (LRR 6:49). Specific gravities as dry matter/displacement volume average 0.5-0.6 for 4-year-old trees.

Economic conversion of wood to electrical energy remains an elusive target in the tropics, but leucaenas and eucalyptus generally remain the favoured trees due to their rapid growth and wide adaptability (Brewbaker, 1980, 1984). Only in the Philippines has an extensive dendrothermal scheme based on leucaena been developed. Denton (1983) has written in detail of this scheme, developed by the Philippine National Electrification Administration. The ambitious plan involved many plantations of about 1,000 ha in size, each including a woodfuel-burning power plant of 3-5 MW peak. At least two of these became operational, but problems of plant construction and tree production have plagued the scheme. The sites were on marginal, usually deforested lands. It is certain that none were chosen because of their high site index for leucaena! Many had no leucaenas growing at all in the area, and several had been abandoned by shifting agriculturalists as worthless. Preliminary trials of species and provenances were generally bypassed, as were careful soil analysis and assessment of nutritional demands. The demands of transplanting and weeding were generally underestimated, coming as they did at the peak of annual farm activities.

10_Leucaena a multipurpose tree for agroforestry

11_Leucaena a multipurpose tree for agroforestry


The success at one site, at Balinao, provides important lessons for others (Denton, 1983). It was located on calcareous hills in an area where shrubby leucaena thrives. Local farmers were familiar with the tree, and management was excellent. Yield attained about 30 m3 ha-1 yr-1, as was predicted. Transportation to power plant has been a major cost factor, with an ineffective overhead hauling system for the small trees (Figure 8). Harvesting with machettes has led to accelerated cutting of small trees. The lack of a suitable tool for felling small trees is a limitation on such schemes. Chainsaws are too expensive and costly to maintain, heavy axes are like scythes (not designed for the steamy tropics!), and small bow saws are unfamiliar.

Charcoal production from leucaena has made little advance in recent years except in the Philippines (Guevarra in PCARR, 1978). The tree makes excellent charcoal, with heating values of 29 MJ kg-1 and good recovery values (25-30 percent). The Mabuhay Vinyl Corporation of the Philippines is a major producer of leucaena charcoal, with several types of improved kilns in operation. The addition of ground leucaena charcoal to fuel oil for diesel engines was found to involve no harmful agents in the ash (LRR 2:76).

Pulpwood, poles and miscellaneous

Leucaena has been evaluated highly as pulpwood, and initial commercial production in Taiwan was carefully planned and economically successful (Hu, 1987; Tai et al, 1983). Pulping properties are suitable to both paper and rayon production. Compared with other tropical hardwoods, leucaena was above average in cellulose and low in lignin, low in bark (5 percent), had a medium-high specific gravity (>0.5) and a higher pulp yield, up to 75 percent under neutral sulphite semichemical processing (LRR 2:51). The wood pulp strength was greater than most hardwoods, with almost 50 percent greater ring-crush resistance. Fibre values are similar to other tropical hardwoods and produce papers with good printability but low tearing and folding strengths. Pulping data are summarized by Hu (1987) and include reports from Taiwan (LRR 2:57), the Philippines (Bawagan and Semana, 1978), India (Relwani, 1983), and the USA (unpublished data).

Poles, mine props, furniture, parquet flooring and chips for particle board are among increasingly popular uses of leucaena. Small furniture and craftwood items are strong and light in weight, and the close-grained wood finishes well and darkens to a golden brown with age (LRR 5:48.) Furniture has also been made from press-wood of glued-up 5 mm strips, and had a specific gravity of 0.58 with good strength (LRR 3:63).

"Grow your own house" could be a motto for leucaena pole and mortar houses in India, pictured in NFTA (1985b). The wood is termite-susceptible, but accepts preservatives. Poles serve a wide array of agroforestry needs, from props for bananas (PCARR, 1978) to crop-support posts for yams, pepper, and other vines.

Ornamental and shade uses of arboreal leucaenas are extensive despite seediness of early varietal releases. All taller species (Table 1) make suitable ornamentals, and seed production is lower on self-incompatible trees and on selected varieties and triploid species hybrids. Isolated self-incompatible trees are seedless, as are certain species hybrids, e.g., L. esculenta x L. leucocephala. The ability to achieve mature height in 3-5 years makes the leucaenas attractive for shade, and loppings provide useful fodder and branch wood, although juvenile growth is wind fragile (LRR 6:23). L diversifolia 4N is an attractive shade tree for subtropical climates with no seed problem (LRR 5:88). The leucaenas crown out too widely to make good windbreaks, although they do not have offensive lateral root development.

Erosion control is a concomitant of alley farming, much of which is on fragile uplands. Contour strips of leucaena must be planted very densely, with small branches laid laterally along the hedge to serve effectively in erosion control (LRR 1:13; 5:51; NRC, 1984). Leucaena is not easily used on rice paddy bunds, due to problems with shading of rice and tolerance by leucaena of waterlogged soils.


Food and other uses

Food

A broadened view of leucaena that embraces all species should bring renewed attention to uses other than fodder, wood and soil amendment. The favoured species for food have traditionally been L. esculenta (guaje rojo) and L. macrophylla (guaje verde). Both are midland or highland Mexican diploids that flower heavily in short days. Pods are marketed in winter at physiological maturity of the seeds. L. esculenta has very large beans with low mimosine. Those of L. macrophylla are similar in size and mimosine content to L. leucocephala, of which the "giant" varieties are also widely used for green pods, even in S.E. Asia. Zarate (1984) considers food use of the leucaenas most important in interpreting distributions of the species. The leucaena seeds are tasty and are often consumed fresh. Nothing is known about human conversion of mimosine through DHP, and whether DHP might be a human toxin of significance (Lowry 1983). Mimosine is lost on preparation through diffusion (LRR 5:53) or following cooking or fermentation, and can be precipitated and entirely removed in preparations involving ferrous ions, as in a rusty pot. Other methods for reducing mimosine values were discussed earlier.

Food use of leucaena leaf tips is common in parts of S.E. Asia (Thailand, Malaysia, Indonesia), normally served after brief stir-frying. They make a significant protein contribution, and are often used with noodle dishes and generous portions of chilli sauce. Leucaena seedlings ("sprouts", i.e., roots) of about three days are prepared as a food in Java, Indonesia. They showed high protein and ascorbic- and amino-acid values with lowered mimosine in studies of Dewi Slamet and associates (LRR 5:53). Tempeh is a fermented food that is also prepared with considerable patience from leucaena seeds in Indonesia (LRR 3:100). Phytic acid, a trypsin inhibitor, was found at a level of 7 mg g-1 in leucaena seeds prepared for tempeh, and two-day fermentation sharply reduced these levels by 70 percent (LRR 7:75). Tempeh samples prepared in our lab by nutritionist M. Whiting were free of mimosine, but otherwise of rather controversial food acceptability.

Protein concentrates and gums

Proteins have been concentrated with difficulty from leucaena leaves, due primarily to galactomannan gums in the leaves that require high dilution and other steps (LRR 2:81). Precipitation of the gums caused by procyanadins or condensed tannins led to reduced protein yields. However, protein processing removed mimosine entirely (LRR 3:93; 5:96). The leaf protein concentrates were low in nutritional value to rats, and tannins were held responsible (LRR 1:45).

Galactomannan gums of leucaena seeds are similar to those of gum arabic (Acacia Senegal), guar (Cyamopsis tetragonoloba) and carob (Ceratonia siliqua). Gum contents of the whole seeds are in the range of 25-30 percent (LRR 5:16). Pure seed galactomannan was isolated readily by Leshniak and Liu (LRR 2:75), and found to be of high uniformity with 1.3 mannosyl residues to 1 galactosyl. It had high haemagglutinating activity on human and rat erythrocytes that was further increased by reducing the galactosyl residues (LRR 2:77). The gum was believed to be a gelling agent comparable to algal carageenan. Mimosine contaminating the gum was easily removed by dialysis, or could be avoided through trichloracetic acid purification with dialysis of the galactomannan (LRR 2:79).

Gums arise from the leucaena stem ("gummosis") under ill-defined conditions of injury and disease, but notably in India with Fusarium infections. Anderson et al. (1983) have carefully analysed these and other mimosoid gums, and found leucaena to be of potential coihmercial value and very similar to gum arabic. Amino-acid fractions of the protein in leucaena and the acacia gums were very similar (LRR 7: 108).

Miscellaneous uses of leucaena include use of seeds for ornaments, and for practical items like hot pads. The Leucaena species vary widely in attractiveness to bees, and provide pollen but not nectar for honey production. Most of the self-sterile species are actively bee-pollinated, and those with heavy scent — L. lanceolata, L. shannoni — seem particularly attractive and might make distinctive honeys.


Why do people plant leucaena?

A good example of the spread and role of leucaena in rural development is given by the work of Manibhai Desai, Director of the non-profit Bharatiya Agro Industries Foundation (BAIF) of Pune, India. He has inspired a generation of Indian scientists to dedicate their efforts to improving the livelihood of the rural poor. Leucaena is a major tool in this programme of revegetation, water management and animal improvement, a programme that earned Desai the prestigious Magsaysay Award in 1985.

Often featured in their excellent publication, The BAIF Journal, leucaena is the subject of extensive development and research activities at BAIF. The giant leucaenas were carefully appraised in many types of management systems prior to large-scale seed increase (to 40 tons by 1986, equivalent to 800 million seeds) and distribution among India's rural community. Desai and his scientific staff have unquestionably inspired tree planting in the past decade on a scale previously believed impossible for the small farmer.

Although fruit trees are planted by rural people of the tropics, the planting of trees for fuel or fodder is an unfamiliar practice — such trees are to be hunted and gathered. Backlash can develop, as with eucalyptus in India, when the tree planting appears to serve the rich or to bring social benefits only to a select society. The fast-grown small legume trees —leucaenas, gliricidias, calliandras, acacias — have a special attraction when they provide both fodder and wood, and can be managed so as to optimize returns from either or both (Brewbaker, 1984).

Why plant leucaena? Hedge (1985) of the BAIF staff lists the features that have made it so effective in India:

  • Fast growth, rapid generation time;

  • Multiple uses, easily interplanted with field crops;

  • Quick returns and high profit;

  • N-fixation and soil improvement;

  • Strong root system, drought and salt tolerance;

  • Cheap to establish;

  • Few diseases and pests.

Some of the successful schemes with leucaena pictured in the US National Academy of Science publication on leucaena (NRC, 1984) illustrate Hedge's points. In the hilly islands of Cebu, Philippines, and Flores, Indonesia, leucaena is widely alley cropped to stabilize shifting agriculture and stop erosion. Markets help motivate the plantings — for pelleted leaf meal on Cebu and fodder to support banteng cattle production on Flores. Markets for woodfuel in pottery making or bakeries motivate leucaena production in central Thailand and the Philippines. Many plantings of leucaena, however, are small and on small farms with the immediate intention of feeding farm animals and serving as home fuel, postwood and other uses.

Leucaena is self-advertising in at least three important ways that greatly simplify its extension and demonstration:

  • Animals relish the fodder;

  • The rapid tree growth is astounding even to tropical farmers;

  • Green manure effects are often plainly obvious.

Alley-cropping trials can show the improved colour and yield of crop plants near the leucaena hedge sufficiently well to convince the most skeptical farmer of the power of green manure. Overcoming the initial farmer reluctance to plant a non-food crop is a difficult hurdle, and almost impossible if the tree is slow-growing and the market is uncertain.

In many sites, perhaps surprisingly, the giant leucaenas have had only to be planted to tell their own story and gain wide grower acceptance. Nevertheless, the key is to inspire that first tree planting. People like Manibhai Desai in India, Harold Watson in the Philippines (also an awardee of the Magsaysay Foundation), Viator Parera in Indonesia (Parera, 1983), and B.T. Kang, G.F. Wilson and colleagues in Nigeria (Kang et al, 1984) are providing that insoiration.


REFERENCES

Anderson, D.M.W., M.M.E. Bridgeman, E.I.G. Brown and J.A.M. Anderson. 1983. The gum exudate from a cultivar of Leucaena leucocephala Lam. de Wit. Intl. Tree Crops J. 2: 291-295.

Arellano R. J. A. 1979. Bibliografia sobre Leucaena leucocephala. Centre Invest. Agric. Peninsula de Yucatan, INIA, Merida, Yucatan, Mexico.

Banco de Mexico. 1980. Leucaena (huaje), leguminosa tropical Mexicana. Usos y potential. FIRA, Banco de Mexico.

Batson, H.F., T.U. Ferguson and K. A.E. Archibald. 1984. Variability in leucaena and its potential in the Caribbean. Faculty of Agriculture, University of the West Indies, St. Augustine, Trinidad.

Bawagan, P. V. and J. A. Semana. 1978. Utilization of ipil-ipil for wood. International Consultation on Ipil-ipil Research, PCARR, Los Banos, Philippines.

Bray, R.A., R.J. Jones and M.E. Probert. 1984. Shrub legumes for forage in tropical Australia. In E.T. Craswell and B. Tangendjaja (eds.), Shrub legume research in Indonesia and Australia. ACIAR, Canberra, Australia.

Brewbaker, J.L. 1975. "Hawaiian Giant" koa haole. Hawaii Agriculture Experimental Station, Misc. Publication 125.

——.1976. Establishment and management of leucaena for livestock production. In Production de Forrajes, Memoria de Seminario International de Ganaderia Tropical. Banco de Mexico, Mexico City.

——.1979. Diseases of maize in the wet lowland tropics and the collapse of the classic Maya civilization. Economic Botany 33: 101-118..

——.(ed.). 1980. Giant leucaena energy tree farm: An economic feasibility analysis. Hawaii Natural Energy Institute Publication 81-04, Honolulu, Hawaii.

——.1983. Systematics, self-incompatibility, breeding systems and genetic improvement of Leucaena species. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

——.1984. Short-rotation forestry in tropical areas. In H. Egneus and A. Ellegard (eds.), Bioenergy 84, Vol. I. Bioenergy state of the art. Elsevier Applied Science Publication 58-78.

——.1985a. Leguminous trees and shrubs for Southeast Asia and the South Pacific. In G.J. Blair, D.A. Ivory and T.R. Evans (eds.), Forages in Southeast Asia and South Pacific Agriculture. Canberra, Australia: ACIAR.

——.1985b. The tropical environment for maize cultivation. In A. Brandolini and F. Salamaini (eds.), Breeding strategies for maize production improvement in the tropics. Rome, Italy: FAO/UN.

——.1986a. Performance of Australian acacias in Hawaiian nitrogen-fixing tree trials. In J.W.Turnbull (ed.), Australian Acacias in developing countries. ACIAR Proceedings No. 6.

——.1986b. Nitrogen fixing trees for fodder and browse in Africa. In B.T. Kang (ed.), Alley farming for humid and subhumid regions of tropical Africa. Ibadan, Nigeria: IITA.

Brewbaker, J.L., D.L. Plucknett and V. Gonzalez. 1972. Varietal variation and yield trials of Leucaena leucocephala (koa haole) in Hawaii. Hawaii Agriculture Experimental Station Research Bulletin 166.

Brewbaker, J.L. and E.M. Hutton. 1979. Leucaena — Versatile tree legume. In G.A. Ritchie (ed.), New agricultural crops, AAAS Selected Symposium. Boulder, Colorado: Westview Press.

Bushby, R.V. A. 1982. Rhizosphere populations of Rhizobium strains and nodulation of Leucaena leucocephala. Austr. J. Exp. Agric. An. Husb. 22: 293-298.

Chao, C.C., C.C. Young and W.E. Cheng. 1985. Effects of rhizobia inoculant and liming on the growth and nodulation of Leucaena leucocephala (Lam.) de Wit in an acid slopeland. J. Agric. Assn. China 131: 35-41.

Chee, Wong C. and C. Devendra. 1983. Research on leucaena forage production in Malaysia. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

D'Mellow, J.P.F. and D.E. Taplin. 1978. Leucaena leucocephala in poultry diets for the tropics. World Rev. Animal Prod. 14: 417.

Denton, F.H. 1983. Wood for energy and rural development— The Philippine experiences. Denton, 3181 Readsborough Ct, Fairfax, Virginia.

Dijkman, M.J. 1950. Leucaena — A promising soil-erosion-control plant. Econ. Hot. 4: 337-349.

Dommergues, Y. 1982. Ensuring effective symbiosis in nitrogen-fixing trees. In P.H. Graham and S.C. Harris (eds.), Biological nitrogen fixation technology for tropical agriculture. Cali, Colombia: CIAT.

FAO. 1983. Leucaena leucocephala: The Indonesian experience. FAO Regional Office for Asia and the Pacific.

——.1985, Production of leafmeal from ipil-ipil (leucaena). FAO Regional Office, Bangkok, Field Document 4.

Gonzalez, V., J.L. Brewbaker and D.E. Hamill. 1967. Leucaena cytogenetics in relation to the breeding of low-mimosine lines. Crop Sci. 7: 140-143.

Guevarra, A.B., A.S. Whitney and J.R. Thompson. 1978. Influence of intra-row spacing and cutting regimes on the growth and yield of leucaena. Agron. J. 70: 1033-1037.

Halliday, J. and P. Somesagaran. 1983. Nodulation, nitrogen fixation and Rhizobium strain affinities in the genus Leucaena. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Halos, S. C. 1980. Abstract on leucaena. Forest Research Institute. (Philippines), Ref. Ser. 8.

Hegarty, M.P., C.P. Lee, G.S. Christie, R.D. Court and K.P. Haydock. 1979. The goitrogen 3 hydroxy-4(lH)-pyridone, a ruminal metabolite from Leucaena leucocephala: effects in mice and rats. Aust. J. Biol. Sci. 32: 27-40.

Hedge, N. 1985. Leucaena for energy plantation. BAIF Journal 5: 37-42.

Hogberg, P. and M. Kvarnstrom. 1982. Nitrogen fixation by the woody legume Leucaena leucocephala in Tanzania. Plant Soil 66: 21-28.

Holmes, J.H.F. 1981. Toxicity of Leucaena leucocephala for steers in the wet tropics. Trap. Anim. Health Prod 13: 94-100.

Hu, Ta Wei. 1987. Use of nitrogen fixing trees for pulpwood. Waimanalo, Hawaii: NFTA.

Hu, Ta Wei and T. Kiang. 1983. Leucaena research in Taiwan. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Huang, Ruey-Shyang, W.K. Smith and R.S. Yost. 1985. Influence of vesicular-arbuscular mycorrhiza on growth, water relations, and leaf orientation in Leucaena leucocephala (Lam.) de Wit. New Phytol. 99: 229-243.

Hutton, E.M. 1983. Selection and breeding of leucaena for very acid soils. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

——.1984. Breeding and selecting leucaena for acid tropical soils. Pesq. Agropec. Bras. 19: 263-274.

IITA. 1986. Alley farming for humid and subhumid regions of tropical Africa. Proceedings of a Workshop. Ibadan, Nigeria: IITA/ILCA.

IDRC. 1982. Leucaena research in the Asia-Pacific Region. Proceedings of Workshop, IDRC and NFTA, Singapore, November 1982.

Indonesia, Government of. 1982. National leucaena seminar. Jakarta, Indonesia.

Jones, R. J. 1979. The value of Leucaena leucocephala as a feed for ruminants in the tropics. World Animal Rev. 31:13-23. Jones, R. J. and R.M. Jones. 1982. Observations on the persistence and potential for beef production of pastures based on Trifolium semipilosum and Leucaena leucocephala in sub-tropical coastal Queensland. Tropical Grasslands 16: 24-29.

Jones, R.J. and R.A. Bray. 1983. Agronomic research in the development of leucaena as a pasture legume in Australia. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Kanazawa, Y., A. Sato and R.S. Orsolino. 1983. Above-ground biomass and the growth of giant pil-ipil (Leucaena kucocephala Lam. de Wit) plantations in northern Mindanao Island, Philippines. JARQ 15(3): 209-217.

Kang, B.T., G.F. Wilson and T.L. Lawson. 1984. Alley cropping — A stable alternative to shifting cultivation. Ibadan, Nigeria: IITA.

Kaul, R.N., M.G. Gogate and N.K. Mathur (eds.). 1981. Leucaena kucocephala in India. Proceedings of National Seminar at BAIF, Urulikanchan, Maharashtra, India.

Lowry, J.B. 1983. Detoxification of leucaena by enzymic or microbial processes. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Manidool, C. 1983. Leucaena leaf meal and forage in Thailand. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Megarrity, R.G. 1978. An automated colorimetric method for mimosine in leucaena leaves. J. Sci. FoodAgric. 29: 182-186.

Munns, D.N. and R.L. Fox. 1977. Comparative lime requirements of tropical and temperate legumes. Plant Soil 46: 533-548.

NRC. 1984. Leucaena: Promising forage and tree crop for the tropics. 2nd edn. Washington, D.C.:USNAS.

NFTA (Nitrogen Fixing Tree Association). 1985a. Leucaena forage production and use. Waimanalo, Hawaii: NFTA.

——.1985b. Leucaena wood production and use. Waimanalo, Hawaii: NFTA.

Oakes, A.J. 1968. Leucaena kucocephala: description, culture, utilization. Advancing Frontiers of Plant Science (India) 10: 1-114.

——.1982-4. Leucaena bibliography. Three volumes, Germplasm Resources Lab., USDA, Beltsville, MD.

Oakes, A.J. and C.D. Foy. 1984. Acid soil tolerance of leucaena species in greenhouse trials. J. Plant Nutr. 7: 1759-1774.

Olvera E.M., M. Benge and S.H. West. 1985. World literature on leucaena Leucaena kucocephala Lam. de Wit. Monograph 11, Agricultural Experiment Station, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, Florida.

Olvera, E. and G. Blue. 1985. Establishment of Leucaena kucocephala Lam. de Wit in acid soils. Trop. Agric. 62: 73-76.

Othman, A.B., M.A. Soto, G.M. Prine and W.R. Ocumpaugh. 1985. Forage productivity of leucaena in the humid subtropics. Proc. Florida Soil Crop Sci. Soc. 44: 118-122.

Pan, F.J. 1984. Systematics and genetics of the Leucaena diversifolia complex. Ph.D. Thesis, University of Hawaii, Honolulu, Hawaii.

Perera, Viator. 1983. Leucaena for erosion control and green manure in Sikka. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

PCARR. 1978. International consultation on ipil-ipil research. Los Banos, Philippines: Philippines Council for Agriculture and Resources Research.

Pound, B. and L. Martinez Cairo. 1983. Leucaena: Its cultivation and use. London: Overseas Development Administration.

Proverbs, G. 1985. Leucaena, a versatik plant. Caribbean Agric. Res. Dev. Inst., Barbados, West Indies.

Relwani, L.L. 1983. Kababul (Leucaena kucocephala) — a renewable source of fuel, timber and pulp. In R.N. Kaul, M.G. Gogate and N.K. Mathur (eds.), Leucaena leucocephala in India. Dehra Dun, India: Vanguard Press.

Sanzonowicz, C. and W. Couto. 1981. Effect of calcium, sulphur and other nutrients on dry matter ield and nodulation of Leucaena leucocephala in &"ceTrado"soft. Pesq. Agropec. Bros. 16: 789-794.

Schroder, B.C. 1986. Recent advances in leucaena research. Misc. Paper, Department of Agronomy, University of Puerto Rico, Mayaguez.

Singh, S. 1981. Gummosis and canker in Leucaena leucocephala. In R.N. Kaul, M.G. Gogate and N.K. Mathur (eds.), Leucaena leucocephala in India. Dehra Dun, India: Vanguard Press.

Singh, S., S.N. Khan and B.M. Misra. 1983. Gummosis, brown spot and seedling mortality in Subabul: Disease incidence and pathology of the host. Indian Forester 109: 185-192.

Sorensson, C.T. 1987. Interspecific hybridization and evaluation in the genus Leucaena. M.S. Thesis, University of Hawaii.

Tai, K.Y., Tao Kiang and K.H.K. Chen. 1983. Promotion of giant leucaena in eastern southern Taiwan. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Takahashi, M. and J.C. Ripperton. 1949. Koa haole, its establishment, culture, and utilization as a forage crop. Hawaii Agric. Exp. Sta. Bull 100:56.

Tangendjaja, B. and R.B.H. Wills. 1980. Analysis of mimosine and 3-hydroxy-4(lH)-pyridine by high-performance liquid chromatography. J. Chromatography 202: 317-318.

Taplin, D.E., J.P.F. DTvlellow and P. Phillips. 1981. Evaluation of leucaena leaf meal from Malawi as a source of xanthophylls for the laying hen. Trop. Science 23: 217-226.

Thoma, P.E. 1983. Leucaena-Rhizobium compatibility and nitrogen fixation. Desert Plants 5: 105-111.

Trinick, M. J. 1980. Relationships amongst the fast-growing rhizobia of Lablabpurpureus, Leucaena leucocephala, Mimosa spp., Acaciafamesiana and Sesbaniagrandiflora and their affinities with other rhizobial groups. J. Appl. Bacterial. 49: 39-53.

Van Den Beldt, R.J. 1983. Effect of spacing on growth of leucaena. In Leucaena research in the Asia-Pacific region. Ottawa: IDRC.

Vicente, N.E., N. Acosta and B.C. Shroder. 1986. Reaction of Leucaena leucocephala to populations of Meloidogyne incognita and M.jovanica from Puerto Rico. /. Agric. U.P.R. 70:157-158.

Yost, R.S. and R.L. Fox. 1979. Contribution of mycorrhizae to P nutrition of crops growing on an oxisol. Agron. J. 71: 903-908.

Zarate, P.S. 1984. Taxonomic revision of the genus Leucaena Benth. from Mexico. Bull. International Group Study Mimosoideae 12: 24-34.