Andiroba

Related species
Description and phenology
Distribution, abundance and ecology
Uses and economic potential
The modern export market
Collection methods and yields
Propagation and cultivation methods
Research contacts

Paulo de T.B. Sampaio

Family: Meliaceae

Species: Carapa guianensis Aubl.

Synonyms: Carapa procera (separated from C. guianensis by Pennington et al. 1981)

Common names: Andiroba, A. Branca, A. do Igapó, A. Saruda, A. Vermelha, Andirova, Angirova, Comacari, Mandiroba, Yandiroba, Carape, Carepenha, Gendiroba, Jandiroba, Penaiba, Purga de Santo Inacio, Aboridan (Brazil), Caraba, Crabwood, Damerara, Brasilian Mahogany, Lowland Crabwood, Highland Crabwood, White Caraba, Karaba, Karapa, British Guiana Mahogany (Guyana); Bois Caille, Cachipou, Carape, C. Blanc, C. Rouge, C. Pune, Crapo, Andiroba Carapa (French Guiana); Kareppa (Suriname); Andiroba (Peru); Andiroba (Paraguay); Bateo, Cedro Bateo, Cedro Macho (Panama); other names found commercially: Bois Rouge Carapat, Crabbaum, Nandirova, Parama Mahogany, Mazabalo, Tangare, Guino (chico).

Related species

Carapa procera is more widely distributed than C. guianensis, being found both in South America and Africa. In the New World its distribution is restricted to northern South America, in ecological habitats similar to those favored by C. guianensis (see below). Its distribution in Africa is more ample, being found in both Central and Western Africa, with the same ecological preferences noted above (Pennington et al. 1981).

The botanical differentiation of C. procera from C. guianensis was done recently by Pennington et al. (1981). The morphological characters used to discriminate the two species are presented in the following key:

1. Flowers sessile, subsessile or, very rarely, short and thick pedicled, predominantly 4 meras with 8 anthers, 1 ovary, 4 [ocular and (2-) 3-4 (-6) ovules per locule; leaflets more elliptical, with a sharply pointed apex: C. guianensis;

2. Flowers always tenuously pedicled, predominantly 5 meras with 10 anthers (rarely with 6 petals and 12 anthers), 1 ovary, 5(-6) - locular and (2-) 3-6 (-8) ovules per locule; leaflets generally oblong with rounded apex: C. procera.

Description and phenology

The andiroba is a large tree, reaching 30 m in height, with a straight cylindrical stem, above frequently buttressed roots. It has a thick, bitter-flavored bark, which flakes off in large plates. The medium-sized crown is abundantly branched. The alternate leaves are compound, with long petioles; its length is 30-60 cm and it is up to 50 cm wide; the leaflets are opposite (up to 19 pairs), ovoid-oblong, with a short apical point, generally leathery in texture, dark green in color, smooth and flat, with an entire margin. The panicle inflorescence is axillary, generally near the branch tips, about 30 cm long. The flowers are subsessile, smooth petaled, cream colored (Loureiro et al. 1979, Rizzini & Mors 1976).

The fruit is a round to nearly round capsule, dehiscent along four sutures that separate upon impact with the ground, which liberates the seeds (sublet 1977, Loureiro et al. 1979). The 4-16 seeds have angular sides, due to mutual compression within the fruit. Each seed averages 21 g, although there is great variability due to seed number (Carruyo 1972, McHargue & Hartshorn 1983, Pinto 1963).

Andiroba flowers in the early rainy season (February to March) and fruits in the late rainy season (March to May) at the Curua-Una Experiment Station, south of Santarem, Pará (SUDAN 1979). At the A. Ducke Forest Reserve, Manaus, Amazonas, it also flowers in the early rainy season (December to March) and fruits in the mid to late rainy season (March to April) (Alencar et al. 1979).

Distribution, abundance and ecology

Andiroba is widely distributed in the Neotropics and Tropical Africa. In the Neotropics it is found from Central America south through Colombia, Venezuela, Suriname, Guyana, French Guiana, Brazil, Peru and Paraguay, as well as on several of the Caribbean Islands. In Brazil, it is found in the entire Amazon basin, preferentially in the varzeas (seasonally flooded white water terraces) and in the igapó (perennially flooded margins of black or clear water rivers and streams), frequently in association with ucuúba (Virola surinamensis) and Pará rubber (Hevea brasiliensis) (Cavalcante et al. 1986). It is occasionally cultivated on the terra firma (dry uplands), generally on well drained clay soils, where it grows well (Le Cointe 1947, Bena 1960, Carruyo 1972, Loureiro et al. 1979, Pennington et al. 1981).

Forest inventories: in various parts of Amazônia, done by FAO in 1956-1961 (SUDAM 1975), found 0.24-4.91 trees/ha. RADAMBRASIL (1974/78) also found the species in diverse areas.

Glerum & Smith (1965) found an average of 1.2 trees/ha near Curua-Una, Pará, with diameter at breast height (DBH) above 45 cm. On the Ilha de Marajó and in Pará's Zona da Mata), IDESP (1975) found 1.87 trees/ha in terra firma primary forest, 4 trees/ha in transitional forest between terra firma and varzea and 1.6 trees/ha in primary varzea forest, in all cases for trees with DBH ³ 25 cm.

In SUFRAMA's Agricultural District, Manaus, Amazonas, an average of 3 trees/ha were found in igapó forest and 7 trees/ha in terra firma forest, all with DBH ³ 25 cm (PROFLAMA 1972). At INPA's Tropical Silviculture Experiment Station, in the same Agricultural District, Jardim (1985) found only 0.125 trees/ha with DBH ³ 20 cm. Three hundred kilometers further north, in Presidente Figueiredo, Amazonas, Poyry (1984) found only 0.4 trees/ha (DBH of 30-60 cm) in terra firma forest and 0.6 trees/ha in igapó forest in a 8000 ha sample. In the same area, at the Balbina Hydroelectric Project, Rio Uatumã, Amazonas, INPA's Tropical Silviculture Dept. found only 1 tree/ha with DBH ³ 10 cm (INPA 1983).

In the Projeto Esperança area, Novo Aripuanã, southern Amazonas, Jardim & Fernandes (1983) found 0.3 trees/ha with DBH ³ 25 cm. Further west, in the Polo Juruá-Solimões, Amazonas, Hosokawa (1981) found 1.2 trees/ha with DBH £ 50 cm and 1.1 with DBH 2 ³ 50 cm.

Uses and economic potential

Principal use
Secondary uses

The wood of andiroba is its principal use, being considered one of Amazônia's noble timbers, and the seed oil is used in popular medicine as a home remedy. The oil has other potential uses, however, which will be explained.

Principal use

The wood is moderately heavy (0.70-0.75 g/cm³); the heartwood is dark red to reddish-brown immediately after cutting; the outer wood is pale brown; the grain is quite regular, the texture is a little thick, the smell and flavor are indistinct. The wood works very well" accepting a finish exceptionally well, and is very popular on the local market for furniture and cabinetry; when prepared as charcoal it has high calorific potential (SUDAM 1979).

A combination of air and oven drying is recommended to avoid warping. It is not resistent to soil rot (Loureiro et al. 1979). Preservation tests have classified andiroba as difficult to impregnate, due to a very low absorption rate and only superficial penetration (SUDAM 1979).

Secondary uses

Andiroba seeds contain approximately 74% kernel and 26% shell (Pinto 1963). The kernel contains 56% liquid oil, which is a transparent yellow color, that will solidify to a consistency of petroleum jelly at temperatures below 25°C. The fresh oil contains about 9% glycerine, principally olein and palmitin, as well as other glycerines in lower proportions (Loureiro et al. 1979).

Kernel analysis gave the following results: humidity - 40%; proteins (N × 6.25) 6.2%; fats - 33.9%; glycerines (by difference) - 6.1%; fiber - 12.1%; minerals - 1.8%. Its physical-chemical constants and fatty acid composition are given in Table 1.

Table 1. Physical-chemical constants and fatty acid composition of andiroba oil.

Andiroba oil is extremely bitter, both when pressed cold or hot. When obtained by hot press, it becomes cloudy when left standing, condensing out a solid white fat composed principally of palmitin. This solid fraction is used for making sodium soaps in northern Brazil.

Small quantities of oil are used to sooth muscular distentions, skin tumors and other superficial skin ailments. The Amerindians use it as an insect repellent (Pinto 1963).

The modern export market

Table 2 contains some export information extracted from the IBGE Annual Statistical Summaries (IBGE 1976, 1979, 1982, 1985). The external market for these kernels is clearly quite small. The use to which they are put outside of Brazil is not explained in the literature cited here.

Table 2. Exports of andiroba kernels (MT) from 1974-1985 (IBGE 1976/79/82/85).

Collection methods and yields

A mature andiroba tree may produce 180-200 kg of kernels/ year (Rizzini & Mors 1976). In a monoculture, at moderate density (6 × 8 m) tan area of 3 × 4 m is used in some silvicultural research (see below)], a yield of 25-50 kg/tree might be possible. This extrapolates to a yield of 5-10 MT/ha/yr, which contains 50% oil. The seeds are collected from the ground around the tree, preferably within a couple of days after falling to avoid predation by mammals and insects (SUDAM 1975).

In Amazônia, andiroba oil is extracted in small processing plants by the following process: the seeds are ground into small sections; these are dried in an oven at 60-70°C until humidity reaches » 8%; the mass is then pressed at 90°C, in hydraulic presses of the Cage Press" or "Expeller type. Double pressing rarely yields more than 30% oil from seeds at 8% humidity, a rather disappointing yield (Pinto 1963).

Propagation and cultivation methods

Andiroba seed germinates well (85-95%) if sown soon after falling from the tree (Viana 1982), starting within 6 to 10 days after being sown in sand covered by sawdust. Germination is complete within 2-3 months (Viana 1982) and the seedlings can be transferred to plastic nursery bags; alternatively, they can be sown directly into the nursery bags (Loureiro et al. 1979, Alencar & Magalhães 1979). Direct sowing is easy and more efficient because of the size of the seed (-60 seed/kg) and its rapid germination. This practice is recommended when the plantation area is near the seed production area, since the seed only remain viable for 2-3 months. Direct sowing, however, may be difficult because of rodent attack; the rodents are attracted by the large nutrient reserves of the seed (SUDAM 1975). Nursery or field grown seedlings are susceptible to shoot borer (Hipsiphylla grandella) attack (SUDAM 1979).

Vianna (1982) concluded that the best conditions for storing andiroba seed is in a humid storage area (14°C and 80% R.H.) or in dry storage (12°C and 30% R.H.), with the seed placed in plastic bags. In either condition, they remain viable for up to 7 months.

Fernandes (1985) analyzed the A. Ducke Forest Reserve (Manaus, Amazonas) experiments and suggested that andiroba could be harvested on a 18-23 year cycle, based upon height, diameter and volume data in plots on yellow latosools, planted at 3 × 4 m. He found yields of 152.5-189.5 m³/ha. At the Curua-Una Experiment Station, Pará, in full sun, planted at 2.5 × 2.5 m, the andiroba showed 80% survival, and average annual increments of 1.8 m in height, 1.1 cm in diameter and 11.4 m³ in volume; the trees started to fruit at 10 years (SUDAM 1979).

Andiroba has been recommended for second growth enrichment near Belém, Pará, because, after 48 months in the field, it presented an annual growth increment of 1.65 m in height and 1.9 cm in diameter (Yared & Carpanezzi 1981). While these data are of short duration, others (see below) show similar results and suggest that andiroba may have great potential as a forest enrichment and/or agroforestry species. Its seed oil could supply annual income to the agroforester or forest manager after the tenth year until the trees are large enough to cut for timber, at which time the pros and cons of timber versus oil could be weighed. The species also has great potential to help recuperate the large degraded areas in the more humid parts of Amazônia.

Volpato et al. (1972) mention its great potential for planting in full sun, both to obtain oil and wood, and its potential for enrichment planting for wood. Alencar & Araujo (1980) concluded that growth in full sun gave superior diameter increments but inferior height increments when compared to trees planted in primary forest enrichment. They explained that the lower height increments are due to intense attack by H. grandella or excessive insolation. The former reason appears to be more likely, because the H. grandella attack destroys apical dominance and causes excessive branching.

Very little work has been done with andiroba in agroforestry systems or in multipurpose forest management. Its potential as a multipurpose (oil-wood) crop seems great.

Research contacts

MSc. Paulo de T. B. Sampaio, Departamento de Silvicultura Tropical, Instituto Nacional de Pesquisas da Amazônia - INPA, Cx. Postal 478, 69011 Manaus, Amazonas, Brazil.

Departamento de Silvicultura, Centro de Pesquisas Agropecuárias do Trópico Úmido - CPATU/EMBRAPA, Cx. Postal 28, 66040 Belém, Pará, Brazil.

Departamento de Silvicultura Superentendência pare o Desenvolvimento da Amazônia - SUDAM, Belém, Pará, Brazil.