C.H. Muller and his colleagues at CPATU, in Belém, have executed the most complete research project undertaken with Brazil nut to date. They have developed seed germination techniques, grafting practices, field establishment and density recommendations; collected elite germplasm for breeding; and started a selection program. The following discussion is based upon their work, as outlined by Mori & Prance (1990b) and discussions with C.H. Muller.
The seed of the Brazil nut may take 12-36 months to germinate and even then only attain 60% germination. Figueiredo et al. (1980) and Frazão et al. (1984) used physical and chemical method. to scarify the seed and enhance germination. Some of these worked, but not well. Finally they simply removed the shell (after soaking for 24 hours in water), treated the seed with fungicide for 60-90 minutes and sowed the seed in sand or sawdust germination beds. Germination starts within one month and after 6 months attains 80%.
Because the Brazil nut seedling has a vigorous taproot, the seedling must be transplanted early and carefully to a deep plastic nursery bag. Unfortunately, root bound plants are still common because the taproot will bottom-out and start to curve around the bag. An untested method is to transplant the seedling into a screen bottomed or funnel-shaped tube (dibble tubes) raised above ground level. When the taproot reaches the bottom of the tube, it would be "air-pruned" and stop growing. If this air-pruned root is not damaged in any other way, it will start growing again when planted out. This idea is currently expanding rapidly in North American horticultural circles and can be executed with extremely rudimentary materials. With locally available materials it could be used anywhere in Amazônia
The young seedling should be hardened-off and planted out when about 40-60 cm tall. A planting pit about 40 cm on each side, containing a mix of surface soil, well-rotted manure and 100 g of treble super phosphate, should be prepared at the end of the dry season and the seedling planted out at the beginning of the rainy season.
Because the seedling should be 2 cm thick at 20 cm above ground level to graft, most grafting is done in the field after the seedlings are established. If the air-pruning technique is developed, it may be possible to grow seedlings large enough in the nursery to allow grafting. In either case, patch buds from orthotropic branches should be selected from highly productive trees with desirable characteristics. The Forkert grafting method has given the best results. If grafted plants are to be used, it is also worthwhile to obtain seed for root stock from highly productive trees that appear never to have been fertilized and, if possible, are found growing in less than favorable environments. These kinds of plants are probably very efficient at scavenging nutrients from Amazônia's nutrient-poor ecosystem. A grafted plant may start fruiting after only 8-10 years in the field, versus 15-20 for seedling plants. This differential precocity suggests that future plantings be directed towards grafted plants, although this will require more skills on the part of the farmer, a large number of high quality clones to choose from (6 at a minimum) and more information on the biology and reproduction of clonal plantings.
A seedling monoculture density of 65-100 trees/ha (10 × 15 m or 10 × 10 m) appears to be suitable, although there are no experimental results yet; grafted trees may do well at higher densities (8 × 10 m or even 8 × 8 m). Muller (1981) recommends 10 × 15 m for mixed cropping schemes and 15 × 20 and 20 × 20 m for establishment in pastures. Posey (1985) reports on Kayapó agroforestry plantings but does not give densities. Observed densities in "natural" castanhais, however, suggest that only 1520 trees/ha are used. More research on planting in mixed cropping, agroforestry and forest-enrichment environments is required. G. Hartshorn's (WWF-US, pers. com.) comments on the Brazil nut with regards to surviving and growing well in abandoned secondary forest, however, suggest that this species may be easy to manage in alternative agricultural systems.
Padoch et al. (1987) report that the Brazil nut is being planted in a market-oriented agroforestry system near Iquitos, Peru. Although Brazil nut is not the major species there, this does show that Amazonian agroforest farmers, without government-sponsored extension support, are following the Amerindian traditions. This is obviously because there is money to be made.
The future of Brazil nut as a crop can be three-fold: as an extractivist product; as an agroforestry/forest management component; and as a modern monoculture plantation crop. As already mentioned, the first option will help conserve tropical forests and the cultures of numerous tropical-forest peoples, both Amerindian and caboclo. The second option is the preferred route to enrich already deforested areas with Brazil nut and provide for the long-term capitalization of the Amazonian farmer. The final option is for the already capitalized investor and may become limited by pests and diseases (Andrade & Cardoso 1984, Freire & Ponte 1976) in the future.
MSc. Carlos H. Muller, Centro de Pesquisas Agropecuárias do Trópico Úmido - CPATU/EMBRAPA, Cx. Postal 48, 66040 Belém, Pará, Brazil.
Dr. Scott A. Mori, Institute of Economic Botany, The New York Botanical Garden, Bronx, NY 10458-5126, USA.