Bertholletia excelsa Bonpl.

  • Family

    Lecythidaceae (Magnoliophyta)

  • Scientific Name

    Bertholletia excelsa Bonpl.

  • Primary Citation

    Pl. Aequinoct. 1: 122-127, t 36. 1807

  • Common Names

    Brazil nut, Para nut, touka, castaña, castanha do Brasil, castanheira, tapa

  • Description

    Author: Scott A. Mori & G. T. Prance

    Type: Venezuela. Amazonas: Without locality, no date (st), Humboldt & Bonpland s.n.(holotype, P; isotypes, B-Herb. Willdenow, P-I0 107). According to the protologue, the type was collected from somewhere along the Orinoco River.

    Description: Trees, to 50 m tall, the trunk not buttressed. Bark fissured, the outer bark thick, the inner bark thick, reddish. Twigs 5-10 mm diam., those with fruits thickest, glabrous at maturity. Leaves scattered at ends of branches; petioles 20-35 mm long, very narrowly alate, caniculate adaxially, rounded abaxially; blades oblong, 17-36 x 6.5-15.4 cm, glabrous, discolorous, adaxially, papillate abaxially (best seen with SEM), the base rounded, the margins entire to slightly crenulate, undulate, the apex apiculate; venation brochidodromous, the midrib slightly impressed basally, prominulous apically, salient abaxially, the secondary veins in 25-45 pairs, plane to prominulous adaxially, prominulous abaxially, intersecondaries present, reaching one-half distance between midrib and margin, the tertiary veins reticulate. Inflorescences terminal or axillary, spicate or paniculate arrangements of spikes with 1-2 orders of branching, the rachis 12-19 cm long, ca. 5 mm thick at anthesis, 12-19 mm thick when bearing fruits; pedicel/hypanthium 2-3 mm long below articulation, absent above articulation, the bract caducous, shape, length, and width not recorded, bracteoles caducous, shape, length, and width not recorded. Flowers when leaves present, ca. 3 cm diam.; hypanthium truncate, puberulous, green, longitudinally oriented mucilage-bearing ducts absent; calyx-lobes 2(3) fused, apical slit present, valvate, obliquely oriented upward, ca. 10 x 10 mm, green, the margins entire; petals 6, shape and size not recorded, glabrous at anthesis, pale yellow to white; androecium zygomorphic, a staminal lip present, the staminal ring with numerous stamens, the filament length not recorded, clavate, white, the anther length not recorded, yellow, the appendage free ligule sometimes pinkish to purple on both surfaces (only visible in fresh flowers), the hood curved, outer surface smooth, white proximally, yellow distally (i.e., at entrance into flower), with numerous vestigial stamens, yellow, not stalked, proximal vestigial stamens swept inward, staminodes absent, posterior and anterior hood extensions present; ovary (3)4(6)-locular, the ovary summit truncate, the ovules ca. 4 per locule, inserted on basal septum, oblique, the style tapering to apex, oblique or geniculate, ca. 10 mm long, stylar collar absent. Fruits dehiscent, the opening smaller than seeds, falling to ground with seeds inside at maturity, usually globose, 10-12.5 (16) cm diam., the calyx-lobes not persistent, the calyx-lobe scars not visible, the infracalycine zone not separated from the supraclaycine zone, rounded to pedicel, the pericarp ca. 20 mm thick, smooth to slightly rough, brown, the operculum, 1.5–2 cm diam., falling inside fruit at dehiscence, the columella ca. 2 cm long, the opercular opening smaller in diameter than seeds. Seeds 3-angled, 10-25 per fruit, 4–5 x 2–2.5 cm, the testa 2–3 mm, thick, boney, dark brown, the primary veins running longitudinally along angles, the secondary veins horizontally oriented between primary veins; aril absent; endosperm not evident in seeds; embryo without cotyledons (macropodial).

    Common names: English: Brazil nut, Para nut. Spanish: castaña, castaña del Brasil, nuez del Brasil.. Brazil: castanha do Brasil, castanha-do-Pará, castanheira. Bolivia: tapa. French Guiana: châtaignes du Brasil, touka (Poiteaui, 1825). Surinam: Braziliaansche noot, ingi nooto (Taki-Taki), para noot. .

    Distribution: Guyana and Amazonian Colombia, Venezuela, Peru, Brazil, and Bolivia. Bertholletia excelsa may also occur in French Guiana but all collections from this Department of France are most likely from cultivated trees. In addition, there is no evidence that native trees exist in the Atlantic coastal forests of Brazil. Humans have planted Brazil nut trees in so many places that it is difficult to tell where they are and where they are not native. Shepard and Ramirez (2011), based on ecological, phytogeographic, genetic, linguistic, and archeological data, make a very strong case for a northern/central Amazonian origin of the Brazil nut and its subequent spread by humans to the southwestern Amazonia. Brazil nut trees are cultivated in botanical gardens far outside their natural range (e.g., Hawai, Florida, Taiwan, Trinidad), and minor plantations have been established in Malaysia (Müller, 1981) and Ghana (Mori, 1992).

    Ecology: Bertholletia excelsa favors non-flooded (terra firme) forests and grows only in areas with a three- to five-month long dry season (Müller, 1981). In eastern Amazonian Brazil, Brazil nut trees are often associated with tall tree species such as Dinizia excelsa and Cedrelinga cateniformis and canopy trees such as species of Manilkara, Protium, and other Sapotaceae (Salomão, 2009). Most fruits are produced on trees that have diameters greater than 50 cm (Staudhammer et al., 2013). Trees of B. excelsa occur in stands known as manchales in Peru (Sánchez, 1973) and castanhais (Dias, 1959) in Brazil. The concept of stands of Brazil nut trees was, however, challenged by Wadt et al. (2005) in their study of Brazil nut spatial patterns in southwestern Amazonia. Densities of Brazil nut trees vary considerably throughout Amazonia; in eastern Amazonian Brazil, Miller (1990) found from nine to 26 reproductive trees per hectare, while Mori et al. (2001) documented only one reproductive tree over 10 cm DBH in a 100 hectare plot in central Amazonian Brazil. In a very intensive study of the ecology of B. excelsa in eastern Brazil, Salomão (2009) studied two plots and found at a density of 1.5 and 0.005 trees, respectively. This supports other research in which great variation in Brazil nut tree density has been reported, even though the plots studied may be relatively close together. In a study of three hectares in the lower Rio Purus region, Haugaasen & Haugaasen (2010) showed that the Brazil nut trees accounted for between 12.0 and 21.8% of the total basal areas. Most research indicates that trees of this species are gap dependent as suggested most recently by Salomão (2009), and many of these gaps may owe their origin to pre-Colombian Indians (Miller, 1990; Müller et al., 1980; Salomão, 2009; Scoles 2011a, 2011b, 2012). According to Cotta et al. (2008), Amazonian inhabitants (caboclos) believe that Brazil nut trees establish themselves and reach high densities in swidden fallows and persist long after the forests appear to be old growth forests. Salomão (2009) points out that in older growth forests the trees are nearly all of the same size in contrast to secondary forests in which they are represented by a greater variation of size classes. Cotta et al. (2008) tested this hypothesis and found that B. excelsa seedlings reached higher densities in fallow than in mature forest (12.7 and 5.3 individuals per hectare, respectively). They concluded that swidden fallows could be managed for Brazil nut production, a system that they argue could increase the income of local people, contribute to the sustainablity of Brazil nut extraction, and promote forest conservation (Cotta et al., 2008), a viewpoint that has been supported by Scoles et al. 2011a, 2011b, 2012). According to Salomão, a typical Brazil nut tree in a mature forest on the Almeida Plateau in NW Pará has a DBH of 234 cm, a height of 32.3 meters, possesses a biomass of 18.4 tons dry weight, and is 253 years old (Salomão, 2009). Small trees grow rapidly when they are exposed to light but do not reach large diameters until they have reached or exceeded the canopy layer. The largest diameters recorded have reached 525 cms DBH and the tallest trees may reach up to 50 m (Salamão, 2009) . In a plot with 1,140 trees the trees were aggregated and in another plot with only seven trees they were not aggregated (Salomão, 2009). The paper by Salomão (2009) should be consulted because of the original data presented and because of its citation of past studies of the ecology of the Brazil nut.

    Phenology: Brazil nut trees flower during the dry season into the wet season. In the eastern part of the Amazon Basin, flowering begins at the end of the rainy season in September and extends to February with peak flowering in October, November, and December (Maués, 2006; Moritz, 1984; Santos & Absy, 2010). Toward the end of the rainy season, usually in July, the leaves of Brazil nut trees begin to fall. The new growth flushes from directly below the inflorescences of the previous year, and the new inflorescences are produced at the apex of the current growth flush. Some trees become totally defoliated in the dry season (Maués, 2006). Large numbers of flowers are produced daily over a relatively long period. The flowers open between 4:30 and 6:00 a.m., but the anthers start to dehisce within the bud several hours before the flowers open. The petals and androecia fall in the afternoon of the day that the flowers open (Mori, 1992). Individual trees produce a large number of flowers each day over a period of three to eight weeks (Maués, 2006). Moritz (1984) and Maués (2006) state that it takes 14 to 15 months for the fruits to develop after they have been set. Thus Brazil nut fruits fall mostly in January and February in eastern Brazil, during the beginning of the rainy season (Maués, 2006). Without special treatment, the seeds take from 12 to 18 months to germinate because of their exceeding boney seed coat (Müller, 1981). Bayma et. al. (2014) found in Roraima that Brazil nut trees flower for 5-6 months in the dry season and that their peak fruiting time ranges from four to six months in the wet season.

    Pollination: The flower of the Brazil nut is closed because 1) the androecial hood arches over and presses against the summit of the ovary and 2) the petals form a cup that presses against the sides of the androecium thereby blocking entrance into the flower by insects not strong enough to push open the hood and petals; thus, Bertholletia excelsa is only efficiently pollinated by robust bees of the genera Bombus, Centris, Epicharis, Eulaema, and Xylocopa (Maués, 2006; Moritz, 1984; Müller, et al., 1980; Nelson et al., 1985; Santos & Absy, 2010). Bees begin to visit the flowers at day break and peak bee visitation usually takes place between 5:30 and 11:00 a.m. (Maués, 2006; Santos & Absy, 2010). It is thought that the bees visit the flowers to collect the nectar that is produced in the hood of the androecium, but this has not yet been clearly demonstrated. A recent paper (Santos & Absy, 2010) lists the pollinators that the authors have recorded visiting the flowers and discusses the role that stingless bees have on the pollination of the Brazil nut. Their study emphasizes the interaction beween stingless bees (i.e., species of Trigona and its relatives) with pollination by Xylocopa frontalis and Eulaema mocsaryi. They report that the former's visits to the flowers of the Brazil nut are not interferred with as much as are those of the latter species, and suggest that stingless bees may play a role in promoting outcrossing by driving pollinators away from one tree and into other trees because of conflicts the pollinators have with stingless bees.

    Dispersal: At maturity the large, round, woody fruits of Bertholletia excelsa fall to the ground with from 10 to 25 seeds trapped inside. The seeds are removed from the capsules and are dispersed by agoutis (Dasyprocta spp. (Huber, 1910; Miller, 1990), acouchis (Myoprocta spp.), and less frequently by squirrels (Haugaasen & Haugaasen, 2010; Miller, 1990). Because the seeds are protected by the woody fruits and because the boney seed coats (Tsou & Mori, 2002) are difficult to open, only animals with sharp teeth and/or a strong biting force are able to consume them. The agoutis and squirrels eat some of the seeds and cache others for future consumption. Some of the cached seeds are forgotten by the animals, and it is these seeds that may germinate and grow into the next generation of trees. A study by Haugaasen et al. (2010) shows that the black agouti (Dasyrprocta fuiginosa) is the principal seed disperser in the area of their study in Amazonian Brazil. The agoutis hide the seeds in diffuse, single-seeded caches at usually less than 10 meters from where they collected them, but sometimes as far away as 50 meters. The removal of entire fruits from under trees has been shown to account for greater dispersal distance than by the transport of single seeds (Haugaasen et al., 2012). The seeds were sometimes secondarily or tertiarily dispersed by either the same agoutis or by other agoutis (Haugaasen & Haugaasen, 2010). In the dry season, a period of fruit scarcity, more seeds were eaten and fewer were cached.

    Predation: The moth, Lusura altrix (Stoll 1782), causes considerable defoliation of Brazil nut trees in central Amazonian Brazil. A pest outbreak of this moth reported by Haugaasen (2009) appeared to follow abnormal weather conditions, and defoliation disrupted normal phenological patterns. See the collection of Mori et al. 25637 for images of the adult and larvae of what appears to be the same moth species photographed on Brazil-nut trees in the Peruvian Amazon (Haugaasen, 2009). Müller et al. (1980) observed that a yellow-colored beetle of the genus Cyclocephala entered a closed flower and destroyed its "sexual organs." The dispersal system of the Brazil nut is based on predation by agoutis and squirrels because some of the seeds that have been buried for future consumption are not relocated by the animals that cache them, and, thus, germinate to grow into the next generation of trees. In addition, capuchin monkeys sometimes open weaker-walled, older fruits and eat the seeds, spiny rats infrequently eat seeds, and peccaries may occasionally eat seeds (Haugassen et al., 2010). Finally, large macaws (Ara chloroptera, A. macao, and A. ararauna) prey upon the seeds of this species while they are still green, causing the loss of 10% of the fruit crop.

    Field characters: Betholletia excelsa is distinguished from all other species of Lecythidaceae by the cylindric trunk, fissured bark, oblong leaves with a whitish cast on the abaxial surface caused by microscopic cuticular papillae; two-lobed calyx at anthesis; woody fruits that fall to the ground with the seeds inside; and seeds with a boney testa and no apparent cotyledons.

    Taxonomic notes: The authorship of this species has been attributed to Humboldt and Bonpland, but the epithet should be credited to Bonpland alone. Although there is considerable variation in fruit shape and size there is no justification for recognizing more than one species of Bertholletia. The early developing seedlings and saplings of Bertholletia excelsa obtain their nutrients from the hypocotyl. In young saplings the hypocotyl appears as a swollen part of the stem. As nutrition is absorbed from the hypocotyl, and as the stem growths in diameter, the swollen part disappears.

    Uses: Brazil nuts are the seeds of the Brazil nut tree (Bertholletia excelsa) and are the only tropical nut from lowland rain forests sold throughout the world (Salomão, 2009). In 2008, 78,000 tons of Brazil nut seeds were harvested and 40% of that harvest was gathered in Brazil (Freitas-Silva & Venâncio, 2011).They are harvested almost entirely from wild trees during a five to six month period in the rainy season. Baider (2000 cited in Malu & Peres, 2005) found 103 to 270 fruits per tree in a forest in southeastern Amazonian Brazil and a mean seed number per fruit of 17.5. In this area, there are 1.65 fertile trees per hectare. Based on studies of three sites in eastern Amazonian Brazil, Miller (1990) determined that the number of capsules produced per tree ranges from 63-216. Removing lianas that climb into Brazil nut trees or are rooted in the vicinity of them triples fruit production 10 years after the lianas were cut (Kainer et al., 2014). These authors, however, point out that lianas should only be removed in the proximity of Brazil nut trees because lianas play an important role in the ecology of tropical forests. Because Brazil nuts fall to the ground with the seeds trapped inside of the fruit, the seeds are easy to harvest and, thus, the Brazil nut has become an important non-timber forest product in Brazil, Peru, and Bolivia. In 1998, the world trade in Brazil nuts was estimated to be worth 57 million dollars (Zuidema, 2003). The fruits are gathered immediately after they fall in order to avoid insect and fungal attack of the seeds and to control the number of seeds carried away by agoutis and other animals. Brazil nuts are attacked by aflotoxins that are toxic to consumers of infected seeds; thus imported seeds are strictly monitored by countries that import them (Freitas-Silva & Venâncio, 2011). Although the harvest of Brazil nuts has often been cited as a non-timber forest product that is sustainable, a paper by Peres et al. (2003) provides evidence that contradicts this belief. They found that groves which have been heavily harvested for long periods are not regenerating themselves as indicated by the near absence of juvenile trees. In contrast, groves that have either not been harvested or lightly harvested have much higher percentages of juvenile plants. The Peres et al. hypothesis (2003) has recently been convincingly contested by a series of papers by Scoles et al. (2011a, 2011b, 2012) in which they concluded that there is ample recruitment of Brazil nut trees in areas that have been harvested over long periods of time. In times past, the bark of the Brazil nut used to be beaten and used for caulking ships (Spruce, 1850). Native Amazonians steep their fishing nets in water with Brazil nut bark which dyes the nets to make them less visibile to fish (S. A. Mori, pers. obs.). Pieces of the pericarp (husks) and seed coats are burned to generate energy; for example, Michael Nee (pers. comm. 28 Nov 2011) observed them being used for generating electricity in Riberalta, Bolivia. Stoian (2004) claims that 40% of the electricity in Riberalta is generated by burning waste products from Brazil nut factories. Brazil nut trees provide an excellent timber, but at least in Brazil and Peru havesting them is against the law. Moreover, Brazil nut trees are considered as endangered species and protected because of their endangered status. Locatelli et al. (2004) and Salomão et al. (2006) suggest that the Brazil nut would make an excellent timber tree to plant on degraded Amazonian landscapes. After 18 years, they found that Brazil nut trees grew to 27.7 meters in height and 45.2 to 60 cm in diameter in expermental plantations in Rondônia. Scoles et al. (2011) and Staudhammer et al. (2013) have reported that Brazil nut trees grow best in areas with the greatest amount of light, such as is found in disturbed habitats. Most of the Brazil nut production is provided by trees that have reached 50 cm in diameter and only 20% of the trees under 50 cm diameter produced fruits while 96% of them at or over 50 cm diameter produced fruits (Staudhammer et al., 2013). Their work supports the suggestions that Brazil nut trees need disturbed habitats to establish initial growth and that this species can be used to regenerate disturbed forest habitat in the Amazon. More detail on the economic botany and agricultural practices of Bertholletia excelsa can be found in Cornejo-Valverde (2003), Kainer et al. (1998, 1999, 2006), Mori and Prance (1990) and Mori (1992), Müller and associates (Müller, 1981; Müller et al. 1980, 1982), Salomão et al. (2006), Stoian (2004), and Zuidema (2003). Vaz Pereira et al. (1981) provide a list of references dealing with various aspects of Brazil nut biology and agriculture. A case study of the importance of the Brazil nut to the economy of northern Amazonian Bolivia by Stoian (2004) demonstrates how much rural Brazil nut collectors and urban workers in factories processing Brazil nut depend on this non timber forest product for their income.

    Etymology: The species epithet refers to the lofty growth form of this species.

    Conservation: IUCN Red List: vulnerable A1acd+2cd, ver 2.3 (assesed in 1998) (IUCN, 2009). Plantas Raras do Brasil: not on list (Giulietti et al., 2009). Salomão et al. (2006) have demonstrated that the Brazil nut is an excellent species for reforestation of highly degraded areas. Nineteen years after a bauxite mine had been abandoned, these authors found that the largest trees had reached 60 cm DBH and heights of 28 meters. Diameter growth was 0.11 cm per year for the smallest trees and 3.24 cm per year for the largest trees and height growth ranged from 0.16 to 1.47 m per year. Besides reforesting the degraded areas, the Brazil nut trees produced fruits that enhanced the income of local people. These observations indicate that Brazil nut trees can become established and grow to large sizes in disturbed habitats which suggests that they are resilient to disturbance and are less likely to be driven to extinction by the activities of humans.

    Source: Based on Prance in Mori & Prance, 1990. This species pages was last updated on 02 March 2014 by Scott A. Mori.

    Acknowledgements: We are grateful to B. Angell, C. Gracie, T. Paine, M. Rothman, and W. Thomas for allowing us to use their images to illustrate the characters of this species.

  • Floras and Monographs

    Bertholletia excelsa Bonpl.: [Article] Mori, S. A. & Prance, Ghillean T. 1990. Lecythidaceae - Part II: The zygomorphic-flowered New World genera (Couroupita, Corythophora, Bertholletia, Couratari, Eschweilera, & Lecythis). With a study of secondary xylem of Neotropical Lecythidaceae by Carl de Zeeuw. Fl. Neotrop. Monogr. 21: 1-376.

    Bertholletia excelsa Bonpl.: [Article] Mori, S. A. & Lepsch da Cunha, Nadia M. 1995. The Lecythidaceae of a central Amazonian moist forest. Mem. New York Bot. Gard. 75: 1-55.

  • Narratives

    Leaf morphology and anatomy of Bertholletia excelsa.

    Wood anatomy of Bertholletia excelsa.

    Inflorescence and flower morphology and anatomy of Bertholletia excelsa.