Piquiá

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

Charles R. Clement

Family: Caryocaraceae

Species: Caryocar villosum (Aubl.) Pers.

Synonym: Caryocar butyrosum (Aubl.) Willd.

Common names: Piquiá, pequiá, piquiá verdadeiro, amêndoa de espinho (Brazil); arbre à beurre (French Guiana); suari, soari (Suriname); bats sauari (Guyana) (Cavalcante 1988); piqui-a, pekea (Malaysia) (Lane 1957).

Related species

Caryocar nuciferum L. - The souari or butter nut of the Guianas (often confused with the piquiá in southern Suriname) has occasionally entered the world market (mostly Europe) as an edible nut, almost exclusively from wild collected trees. Although a larger fruit, it has a lower mesocarp to endocarp ratio than the piquiá, which makes the seed (nut) more important. The souari nut has a flavor like sweet almonds (Rosengarten 1984). It occurs in Suriname, Guyana and Venezuela, where it is also reputed to be cultivated (Leon 1987).

Caryocar brasiliense Gam. - The pequí or piquí has a smaller fruit, with a similar mesocarp to endocarp ratio as the piquiá, It is widely distributed in the Brazilian cerrado (sparsely forested savannas) and adjacent transition zones, where it is extremely popular and occasionally cultivated. It is used principally as a flavoring for other dishes, especially rice, although it may also be consumed cooked or made into a liquor. The seed (nut), although edible, is seldom used (FAO 1986).

Caryocar glabrum (Aubl.) Pers. - the piquia-rana (Brazil) or almendro (Peru) has very little mesocarp but is widely used for its seed (nut) in Peru, where it is frequently commercialized. There is no record of cultivation (FAO 1986).

Description and phenology

The piquiá, is one of the largest forest trees in most of its distribution, frequently attaining 40-50 m when occurring as an emergent above the canopy, with an unbranched, straight trunk supporting an open leafy crown. When grown in the open, trees are smaller, with abundant branching and a wide crown. The trunk of emergents may attain 2.5 m in diameter at breast height and has a rough, fissured gray-brown bark. The taproot is thick and long, although most roots are superficial. The leaves are trifoliate, with the central, broadly elliptical, apically pointed leaflet 822 cm long by 6-12 cm wide, and the 2 laterals slightly smaller, on a relatively long (4-15 cm) petiole.

The large fruit vary in size, from smaller than a baseball to as large as a softball (6-8 cm long by 6-9 cm diameter), and weight (150-750 g; average 300 g); they may contain 4 seeds, although 1 or 2 are most frequent. The thin skin is gray-brown and moderately smooth. The pericarp (an outer, loose mesocarp) is 1 trick (1-2 cm) and fleshy, representing about 65% of fruit weight. The oily mesocarp is moderately thick (0.5-2 cm) and dense, creamy-beige to creamy-yellow, representing about 10% of fruit weight, in which are embedded the endocarp spines. The tough spiny endocarp is composed of short (0.5-1.5) spines arising from a very tough inner layer that surrounds the kernel, representing about 23% of fruit weight. The white, oily kernel is large (2-3 cm long by 1-2 cm diameter) and somewhat soft, although also somewhat crispy, with a pleasant flavor, and represents about 3% of fruit weight. The mesocarp/endocarp forms a hard, oily, kidney-shaped stone, which represents about 35% of fruit weight.

The piquiá, flowers during the dry season [July-November in Manaus and Belém (Cavalcante 1988, FAO 1986, Prance & Silva 1973), December-February in Roraima (Prance & Silva 1973)], with fruiting 6-8 months later [March-May in Manaus and Belém (Cavalcante 1988, Prance & Silva 1973)]. Leaves are shed annually at-the start of the dry season (FAO 1986).

Distribution, abundance and ecology

The piquiá's modern distribution is limited in the north by the Guiana shield, extending from Cayenne (French Guiana) in the east, through Pará and Roraima, to the middle Negro River in Amazonas (Brazil) in the west, and in the south by the Brazilian shield, extending from São Luiz (Maranhão, Brazil) in the east, through Pará and Amazonas, to northern Rondônia and eastern Acre (Brazil) in the west (Prance & Silva 1973).

Henry Wickham claimed that the piquiá, was very abundant near Santarém (Pare, Brazil), at the mouth of the Tapajós River (Lane 1957). Otherwise the species is represented by less than one tree/ha in areas where it is common. Higuchi et al. (1985), for example, found only 0.1 tree/ha north of Manaus, although Rodrigues (1963) found 0.4 trees/ha nearby and FAO (cited by Loureiro et al. 1979) found 0.4-0.6 trees/ha along the Belém Brasília highway in southeastern Pará. Its distribution suggests that the Amerindians may have helped disperse the species, although they appear not to have created the concentrated stands that they did with the Brazil nut (Bertholletia excelsa), except in the case noted by Wickham. More detailed study of the species' distribution may turn up other cases of exceptional abundance.

The piquiá, is adapted to the heavy, nutrient-poor but well-structured, well-drained clay soils (oxisols and ultisols) of the non-flooded Amazonian plateaus (terra firma). The species is most common in Köppen's "Am" climatic zone, with 1700-2500 mm of rainfall and a 1-2 month dry season; it also occurs in "Aw" climate areas, with 1200-1700 mm and a 3-4 month dry season and in "Af" climate areas, with 2500+ mm and no dry season.

The species is a prominent component of the high forest, frequently found as an emergent (similar to Brazil nut). Pereira & Pedroso (1972a) suggest that it does not regenerate well in the shade of the high forest, a common characteristic of canopy and emergent species. In full sun, growth is rapid (Clement 1982, Loureiro et al. 1979).

Uses and economic potential

Principal uses
Secondary uses

Principal uses

The principal commercial use of the piquiá, is its wood. In 1972, piquiá, represented 1.1% (26,540 m³) of the timber commercialized in Manaus [Volatron (1976), cited by Corrêa & Corrêa 1979]. Loureiro et al. (1979) describe the wood: heavy (0.8-0.85 g/cm³); light whitish- to yellowish-beige color; rough grain because of compact interwoven fibers; medium texture, moderately easy to work and finish; rot resistant; and good quality for ship-building, civil construction, and general carpentry [although susceptible to warping if not dried correctly (Clement, personal observation)].

The principal popular use of the piquiá, is its fruit. Henry Wickham commented on the avidity with which the Amerindians near the mouth of the Tapajós River collected these (Lane 1957). Throughout its distribution, the caboclos know the location of most trees and visit them frequently during harvest season.

The fruit are boiled for about one hour in salted (if available) water. The pericarp becomes loosened and is discarded. The hard, oily mesocarp is gnawed or cut-off with a knife and the endocarp and kernel are generally discarded because of the difficulty of removing the seed (nut) with primitive tools (Cavalcante 1988, FAO 1986). Aguiar et al. (1980) report 72% oil, 3% protein, 14% fiber and 11% other carbohydrates [dry weight basis (50% water)] in the mesocarp, which make it a good source of energy but poor in protein.

Secondary uses

The piquiá's major potential is as an oil crop, although this is currently only a very minor use. Lane (1957) explained that Henry Wickham [famous for having taken Pará rubber (Hevea brasiliensis) to British Asia] spent a significant part of his life trying to promote the development of piquiá, as an oil crop in Asia.

Numerous authors have studied the oil of the mesocarp and the seed, but the piquiá, never attracted the attention of the agricultural research community as a potential oil crop. Table 1 gives several physical-chemical analyses of these oils and Table 2 gives available fatty acid data.

Table 1. Physical-chemical characteristics of piquiá mesocarp and seed oils. LeCointe (1927) and Peace (1941) analyzed samples from eastern Amazônia, while Georgi (1929) and Eckey (1954) analyzed samples from Malaysia, whose germplasm originated in eastern Amazônia.

In general, the piquiá, mesocarp and seed oils are very similar to oil palm (Eleais guineensis) mesocarp oil (Eckey 1954). Pesce (1941), however, recommended the piquiá seed oil for use in the cosmetic industry. If this oil turns out to be especially useful for this market it will command a higher price than the mesocarp oil, which can serve as a substitute for palm mesocarp oil, although this market is currently saturated.

Table 2. Fatty acid composition of piquiá mesocarp and seed oils. Both samples obtained from Malayan plantations, whose germplasm originated in eastern Amazônia.

Just after World War I, the piquiá was finally taken to Malaya, where the Birkhall Estate plantation (Kuala Ketil, Kedah) was created (Lane 1957). Growth of the piquiá seedlings was rapid and they started to yield within 7 years. Yields, however, were lower than expected. The germplasm available was unselected and, obviously, from generally unproductive plants. This venture failed because of this and low world prices for vegetable oils. It was the only attempt ever made to plant piquiá commercially.

The seed (nut) is also a secondary product, mostly because it is so difficult to remove from the endocarp. It has a good flavor, slightly sweet, and could surely find a market. The souari nut market is currently small, but piquiá could enter this as an alternative source of this slightly better-known nut. It would not be difficult to develop a machine to break-off the endocarp spines and then crack the remaining endocarp to liberate the seed in good condition. Toasting or dehydration would be necessary to inactivate enzymes that would otherwise rancify the seed oils and causes free fatty acids to develop in storage (Lane 1957).

The pericarp is rich in tannins [34% dry wt. basis (Georgi 1929)] of the pyrogallol type. This type of tannin was imported into the US in the 1960's in large quantities for use in leather tanning (Thorenstensen 1969). Tannins of diverse types are also being used for making adhesives for use in the wood products industry; based on current available information, piquiá tannins have not yet been researched. Because of the large quantity of pericarp, this could become an important by-product of any piquiá processing project.

Loureiro et al. (1979) mention a popular medicinal use of a tea made from the bark for relieving fevers and as a diuretic, however, there is no supporting evidence to prove medicinal values.

Collection methods and yields

The fruit falls from the tree when ripe and needs only to be collected from the ground. Because it is rich in oils, it is necessary to collect and process the fruit as soon as possible to avoid rancidity. The Birkhall Estate (Malaya) developed a machine to remove the mesocarp from the endocarp (Lane 1957). Similar machinery is easily available today and could be modified to work at the community level.

Le Cointe (1947) reported a single tree producing 6,000 fruit near Belém, but yields are generally between 100 and 300 fruit on the nutrient-poor soils near Manaus. A yield projection may be made from the Manaus data, using the following assumptions: 100 trees per hectare; unselected germplasm; no agronomic treatments; 300 g per fruit; 50% humidity in fruit. This projection yields 3-9 tons of fresh fruit per hectare, from which can be obtained 90-270 kg of seeds (nuts) and 1-3 tons of fresh mesocarp/endocarp stone for direct human consumption, or 330-990 kg of tannins, 105-315 kg of mesocarp oil and 30-90 kg seed oil. Even if yields were twice as good in Malaya, it is easy to see why their effort failed: oil palm produced 2 tons per hectare of mesocarp oil in those days and produces 5 tons today (Hartley 1977). To be a valid option in either monoculture or agroforestry systems, these yields must be increased dramatically. Nonetheless, if Le Cointe's (1947) report is correct, it would not be difficult to find good germplasm and, combined with grafting and good agronomy, raise yields.

Propagation and cultivation methods

Pereira & Pedroso (1972b) obtained 60% germination in 25 weeks when the cleaned stone was sown at a depth of 10 cm in sandy loam, although CTM (1979) reports a similar percentage in only 3-4 months. At 30 weeks from sowing, the seedlings averaged 75 cm in height (Pereira & Pedroso 1972b).

Barbosa, Vastão & Clement (unpublished) found that piquiá grafts easily, especially with the modified forkert method as used in rubber. Grafting would make selection and propagation of elite material extremely easy and could significantly raise yields in plantations, either monoculture or agroforestry.

In silvicultural trials in full sun, growth is rapid: 1.0 m per year at 5 × 5 m in Manaus (Clement 1982) and 1.14 m in Pará (CTM 1979). CTM (1979) also report an average volumetric increment of 10.7 m³ per hectare per year in Pará, although spacing is not given. For a hardwood species, these growth rates are extremely high. Clement (1982) reports slightly lower growth rates at 10 × 10 m after 5 years and attributes this to crown expansion, which seems to be the Malaysian experience also (Lane 1957).,

Pereira & Pedroso (1972a) and CTM (1979) suggest that natural regeneration of piquiá is difficult, which means that seedlings grow slowly in the shade of the forest. If piquiá were to be used in a forest enrichment scheme, it would require relatively large gaps or east-west oriented roads in the forest to grow quickly enough to provide a return to the planter.

Given these rapid growth rates and the piquiá's popularity in most of Amazônia, it would appear feasible to include this species as a multipurpose agroforestry component or forest management component. All the research remains to be done, however.

Research contacts

MSc. Charles R. Clement, Depto. Ciências Agronômicas, Instituto Nacional de Pesquisas da Amazônia - INPA, Cx. Postal 478, 69011 Manaus, Amazonas, Brazil.

Dr. Noeli P. Fernandes, Dept. Silvicultura Tropical, Instituto Nacional de Pesquisas da Amazônia - INPA, Cx. Postal 478, 69011 M Manaus, Amazonas, Brazil

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

Dept. Silvicultura, Superintendência pare o Desenvolvimento da Amazônia - SUDAM, Belém, Pará, Brazil.