Yields, harvesting and processing methods


Babassu is a low yielder in most ecosystems, averaging only 1.5 MT/ha/yr in Maranhão. There are indications, however, that adequate management of natural stands can double this. Nonetheless, even 3 MT/ha/yr is a low yield. As mentioned above, however, Orbignya oleifera yields an average of 5 MT, so there is potential for increasing the yield of the genus if it is to be commercially exploited.

The babassu fruit is generally harvested after it has fallen from the bunch. People gather the fallen fruit and either split it immediately to remove the kernel or take it home for later kernel separation. The fruit can be stored for subsequent kernel extraction, but a beetle can destroy one or more kernels.

Kernel extraction has traditionally been carried out by women and children, who take a whole fruit and split it lengthwise by pounding it with a mallet against an up-turned axe blade. Two or three splits are necessary to extract all the kernels. This labor intensive process generally yields per day 58 kg of whole kernels per person, with many damaged ones resulting.

The numerous industrial efforts for babassu in Maranhão and elsewhere have attempted to develop mechanical methods for extracting the kernels. Unfortunately, babassu is not a cooperative fruit with which to work, as fruit vary widely in size and in number of kernels contained in each. Consequently, machine extraction yields mostly damaged kernels that must be processed immediately to avoid rancidity.

As the babassu industry moves from an oil-extraction process to integrated use of the fruit (see Anderson et al. 1991), machines to separate the husk from the mesocarp and the kernel from the endocarp are being developed. Village or farm level machinery is also being developed. Current work is extremely promising at the village level, as this allows the rural inhabitants to obtain a greater percentage of final price for the material that they sell. Although Anderson et al. (1991) do not provide much detail, because of the incipient nature of this project, this is the most likely route that could be followed to reinvigorate the babassu industry and to permit babassu to return to prominence as a valued renewable resource of Maranhão and elsewhere.

In 1989-1990, the first village-level babassu processing plant was set up near Bacabal, Maranhão, in the area of an agricultural cooperative. The Agricultural Research Enterprise of Maranhão (EMAPA) and the Financiadora de Estudos e Projetos (FINEP) collaborated on the development of a village scale splitting and processing technology. Simple, but durable machines were developed to separate the husk and mesocarp from the endocarp, and to crack the endocarp to extract the seeds. A village level oil press, designed to handle small amounts of kernel at each pressing, could then be used to immediately extract oil and avoid rancidity. Small kilns were designed to make charcoal and extract; tars.

The model proposed by EMAPA for that area included the introduction of mechanization and integral processing of fruits at the village level, and the organization of fruit collection and transportation to the processing area. In this system, fruits are gathered by cutting off the mature infrutescences (just after the fruits start falling), and by using the traditional system of collecting the fruits which have already fallen to the ground. By using this system, a person can collect, on average, 25 infrutescences/day, equivalent to » 625 kg of fruits. Animals carry the fruits to the roads, and the fruits are transported to the processing area by trucks. Using small-scale equipment with capacity of processing » 15 MT/day, the working area for fruit collection must have approximately 3 km of radius which corresponds to » 2,850 ha of continuous area. The area must also have an effective palm coverage of approximately 50-60% and a mean productivity of 1,800 kg of fruit per hectare in order to satisfy the processing capacity of the equipment. The machinery consists of a dehusking machine which separates epicarp and mesocarp (capacity of 2 MT/hr), a breaking machine (1.5 T/hr), and as the final step a machine to separate the small pieces of broken endocarp from the kernels.

If the processing plant works as planned, and the cooperative can market the kernel oil, kernel press cake, starch, charcoal and tar, the members of this cooperative stand to receive five times what they would have received for fresh kernels from the same amount of fruit. This economic incentive could help reestablish babassu's prominence in the local community.

Additionally, members of the cooperative were organized to manage their babassu zones as agroforestry systems. At densities of up to 80 palms/ha, subsistence crops can be planted and yield well under the palms. Rice, maize and manioc are the traditional crops of this area. Anderson et al. (1991) point out that a shifting cultivation cycle of only 4-5 years can be established under babassu because the palm produces such a large foliar biomass. Every 5 years the leaves are cut and burned to provide nutrient-rich ash for planting crops. After the harvest, the palms recuperate and continue to yield fruits that help sustain the farm family until the next cropping cycle.

The combination of village-level processing and an adequate agroforestry management of the babassu stands with annual crops appears to be sustainable both ecologically and economically in this part of Maranhão. This could contribute to raising the extremely low standard of living in rural Maranhão to levels that could create a permanent rural economy based on babassu. These same technologies are planned for use in the first extractive reserve for babassu being organized in Maranhão

Propagation and cultivation methods


The embryo in the babassu seed is straight and basal, measuring, an average 1.0 cm long by 1.5 mm wide (Pinheiro & A. Neto, 1985). The babassu seed has a type of germination known as " remote tubular, " according to classification of Gatin (1906), and it germinates hypogeally, since the cotyledon does not emerge from the seed. A cotyledonary axis (comprising petiole and sheath) elongates 20-60 cm downward from the kernel into the soil, carrying the undeveloped plantlet (plumule and radicle) below the surface. The haustorium (Tomlinson, 1960) enlarges within the seed and occupies the space formerly filled by the endosperm. The radicle (that precedes the plumule) persists for a limited period, reaching 20-40 cm in length, and is replaced by adventitious roots. The first leaf grows out through an opening in the sheath.

Fire and shade stimulate germination, both conditions being common in the forest-cerrado transition zone where babassu is most common. The palm forms an extensive root system that can extend to several meters in depth and a dozen meters or more from the stem. The apical meristem remains buried in the soil as it expands in diameter, usually requiring several to many years to produce enough leaves to provide sufficient diameter to start stem growth. This period encompasses the seedling and stemless juvenile phases of babassu growth and the plant is difficult, if not impossible, to kill at this stage because the meristem is buried in the soil. As one might imagine, it is also extremely difficult to manage this type of growth in a nursery.

One of the great difficulties when working with palms in nursery production is obtaining reliable germination levels. Considerable work on germination of the babassu seed has been carried out at EMAPA, Maranhão Brazil (Frazao & Pinheiro 1985; Pinheiro 1986; Pinheiro & Araujo Neto 1987ab).

Most work with babassu has focused on managing natural stands, rather than creating new plantations (although this is also being attempted). As mentioned in the above discussion, babassu is an excellent candidate for agroforestry systems; it is also an excellent silvo-pastoral component.

In either case, the large number of seedlings and stemless juveniles need to be controlled. Because these plants are immune to fire and slashing, they must either be harvested for palm heart or sprayed with systemic herbicides. A mix of juveniles and young fruiting plants can then be identified for further management; the older fruiting plants and purely staminate plants can be eliminated as they contribute little or nothing to stand productivity. A stand density of 80 plants will permit adequate light to reach the soil for either annual crops or pasture to grow under them. Meat yields from cattle are reputed to be higher in pasture under babassu stands, probably because of a combination of nutrient cycling from the palms, shade for the cattle and the generally better management practiced in the area to control the palms.

No information has been published on babassu response to fertilizer application, although trials are underway in Bacabal to determine the result of fertilization. Since the palm does better where soils are naturally richer, it is expected that it will respond well to any fertilizer applied. In both agroforestry and silvo-pastoral systems, some fertilizer will need to be applied to restore nutrients lost by export from the area. This might only be necessary at relatively advanced ages, however, because of babassu's deep and extensive root system that is reputed to be an efficient deep soil to surface nutrient pump. Additional research is necessary on this topic.

Research contacts


Michael J. Balick, Institute of Economic Botany, The New York Botanical Garden, Bronx, New York 10458-5126, USA.

Anthony B. Anderson, Ford Foundation, Praia do Flamengo, 100-12° andar, 22210 Rio de Janeiro, RJ, Brazil.

José Mário F. Frazão, Empresa Maranhense de Pesquisa Agropecuária - EMAPA, Rua Henriques Leal, 149 - CENTRO - CEP 65010, São Luis, MA, Brazil.

Claudio Urbano B. Pinheiro, Institute of Economic Botany, The New York Botanical Garden, Bronx, New York 10458, USA.

Peter H. May, Av. Epitacio Pessoa, 3400/706, 22471 Rio de Janeiro, RJ Brazil.

Antonio Mariano de Campos Mendes, Secretaria de Agricultura do Estado do Maranhão CEP 65.400, Codó - Maranhão Brazil.

Lidio Coradin, Centro Nacional de Recursos Genéticos e Biotecnologia, CENARGEN/EMBRAPA, SAIN, Parque Rural, Caixa Postal 10.2372, CEP 70.700, Brasilia, DF, Brazil.