Use of Tropical Rainforests by Native Amazonians

Use of Tropical Rainforests by Native Amazonians
Author(s): Darna L. Dufour
Source: BioScience, Vol. 40, No. 9, Ecosystem Science for the Future (Oct., 1990), pp. 652-659
Published by: American Institute of Biological Sciences
Stable URL: http://www.jstor.org/stable/1311432
Accessed: 16/10/2009 10:53
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=aibs.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
American Institute of Biological Sciences is collaborating with JSTOR to digitize, preserve and extend access
to BioScience.
http://www.jstor.org
Use
of
Tropical Rain
Native
forests
by
Amazonians
These sophisticated and complex agriculturalsystems can serve
as models of sustainable agroecosystems
Darna L. Dufour
ndigenous peoples have lived in
the rainforestsof Amazoniafor a
long time, probablythousandsof
years. They were once more numerous and occupiedmore of Amazonia
than they do today. Those groups
known ethnographicallylive in interfluvial regions and share a broadly
similar subsistencepattern based on
horticulture, hunting, fishing, and
collecting.They provide the best example of sustainableuse of tropical
rainforestsunderlow populationdensities.
Of the ways in which Amerindians
use the rainforest, the best documentedare their diverse,multistoried
agriculturalplots, or swiddens.More
recentlystudied is their management
of swidden fallows, in which annual
crops are combined with perennial
tree crops and the naturalprocess of
reforestation.Some groupsalso modify what appearsto be primaryforest
by planting along trailsides and
campsites.The resultof their agricultural practicesis a mosaic of vegetational patches in different stages of
succession and under differing degrees of management. This vegetational mosaic is then used for food,
materials,and medicinals,as well as
for huntinggame animals.
The objectives of this article are
briefly to summarizethe history of
indigenouspeoples, describesome of
the contributions anthropology and
allieddisciplineshave madeto underDarnaL. Dufouris anassociateprofessor
of Colat theUniversity
of anthropology
orado, Boulder, CO 80309. ? 1990
American
Instituteof BiologicalSciences.
652
Much that has been
considerednatural
forest is probably the
result of centuriesof
human use
standinghow these peoples use tropical rainforests,and, finally, to compare this knowledge briefly with
patternsof use characteristicof other
populations. Many examples are
drawnfrom the TukanoanIndiansin
the Colombian Vaupes region in
Northwest Amazonia, the area with
which I am most familiar.
History of native Amazonians
When the Europeansarrived,the indigenous population of Amazonia
was much greater than it is today.
One estimateis that the Amerindians
numberedapproximately6.8 million
in an area of almost 10 million km2
(Denevan1976)1. Populationdensity
was highest in the floodplainsof the
majorrivers,or what is referredto as
Varzea,and along the Atlanticcoast.
Early explorers such as G. Carvajal
and F. Orellanareporteddensesettle'The use of Amazoniain this article follows
that of Denevan (1976). It refers to greater
Amazonia, which includes the tropical lowlandsand plateauseast of the Andesand north
of the Tropicof Capricorn,exceptfor the Gran
Chacoregion.It is an areaconsiderablylarger
than the drainageof the Amazonand its tributaries.
mentsalong the banksof the Amazon
in the 1540s and a level of social
organization referredto as a chiefdom (Roosevelt 1989). The remains
of large middens (refuse heaps), as
well as the extent and depth of terra
preta do indio2 (Indianblack earth)
provide additional evidence that the
riparianzones of major rivers were
heavily populated (Roosevelt 1989,
Smith 1980). The interiorforestsand
savannas, or terra firme, appear to
have been much more sparselypopulated (Denevan1976).
The indigenous population in the
floodplain declined rapidly, and, by
only 150 years after Orellana'sexpedition, the chiefdoms were extinct
(Roosevelt 1989). The severe depopulation that followed this contact is
assumedto have been the resultof a
combinationof disease, slavery, and
warfare. The rate of population decline in the more isolated interior
forest areaswas probablylower than
in the floodplain, because many of
these groups had only sporadiccontact with outsidersuntil well into this
century(Denevan 1976).
In the early 1970s, the indigenous
population of Amazonia was estimated at less than 500,000 (Denevan
1976). In Brazilalone, the population
dropped between 1900 and 1957
from approximately1 million to less
than 200,000 (Ribeiro 1967). Amerindians are now confined to interior
forests and savannas, and the once
2Terrapretado indio is a soil darkenedby the
ceramresidueof repeatedfires.It characterizes
ics and otherremainsof humanactivity(Smith
1980).
BioScienceVol. 40 No. 9
denselypopulatedfloodplainis home
to caboclos, the descendants of
detribalizedIndians and early European immigrants (Moran 1974,
Parkeret al. 1983). However,even in
terrafirme,it is clearthat no presentday indigenousvillagesare as largeas
some of the villages were in the past
1939, Smith1980). Un(Nimuendajti
of indigenoususe of Amderstanding
azonianecosystemsis limitedto terra
firmeforests, and these forests themselves may have been more heavily
used in the past.
50
"
40
30
20
10
Amerindian use of forests
for agriculture
Swiddenplots. The traditionalindigenous agriculturalsystem is based on
swidden cultivation (also known as
slash-and-burn or shifting cultivation). In essence,this system involves
felling and burninga patch of forest,
cultivatingand harvestingcrops for a
period of several years, and then allowing the forest to regrow for 15
yearsor morebeforethe site is cleared
again (Beckerman1987). The felling
and burning releases the nutrients
stored in the forest biomass and
makes them available to cultivated
crops. The long fallow restores soil
fertility,protectsthe physicalproperties of the soil, allows time for nutrient accumulationin the biomass,and
helps control agriculturalpest populations.
Amerindianswiddens are typically
small, 0.4-0.6-hectare (Beckerman
1987), polycultural plots (i.e., they
are planted in more than one crop
simultaneously).Multiplevarietiesof
the staple crops are planted in each
swidden. Most groups of Amerindians also have monocrop swidden
plots devoted to the dietary staple.
The predominatestaples are cassava
(ManihotesculentaCrantz)and plantains and bananas (Musa sp.).
Cassava, or manioc, is a perennial
shrub grown for its starchy roots. It is
native to the neotropics, and well
adapted to the low-fertility, acidic soils
common in terra firme (Cock 1985).
Cassava is cyanogenic, and varieties
are recognized as being either bitter or
sweet (i.e., containing high or low
amounts of cyanide). The bitter-sweet
distinction determines culinary use:
bitter varieties are elaborately processed to reduce cyanide levels before
October 1990
0
0
10
20
30
Subsample Units
numberof cassavavarietiesin subsample
unitsin fourTukanoan
Figure1. Cumulative
swiddens.Subsample
unitswererandomly
placedcircles,2 metersin radius.
being consumed, whereas the sweet
ones can simplybe peeled and boiled
(Dufour1988).
The agriculturalplots of Tukanoan
Indians in northwestern Amazonia
providea good exampleof traditional
swidden cultivation. Swidden plots
are felled by individualhouseholdsin
well-drained soils in both primary
and successionalforests.Most households fell swiddens in both types of
forest. Bittercassava,the dietarystaple, is denselyplantedover nearlythe
entire garden surface. Other crops,
such as taro (Colocasia spp.), sweet
potato (Ipomoea sp.), arrowroot
(Marantaruiziana),pineapple(Anana
sativa), chili peppers(Capsicumannuum), mafafa (Xanthosomamafafa),
lulo (Solanum sp.), bananas, and
plantains,are interplantedwhere microenvironmentalconditions(suchas
drainageand ash concentration)are
deemed most suitable. They are
plantedtoward the centerof the garden where the regrowthof the forest
will be most easily delayedby weeding. Coca (Erythroxylum coca),
which is a shrub,and tree crops such
as guama (Inga sp.), uvilla (Pourouma cercropiaefolia), and peachpalm (Bactrisgasipaes)are also interplanted with cassava toward the
centerof the garden.
Tukanoanswidden plots are polycultural. They are polyvarietal as
well, but the diversityof cassavavarietiesis much greaterthan the diversity of the other crop species. For
example, in identifying cultivated
plants in subsampleunits in four Tukanoan swiddens3 (Table 1), we
found that each containedbetween2
and 16 differentcrops,but as manyas
17 to 48 differentcassava varieties.
On average,the four most common
cassavavarietiesin each swidden accounted for only 38% of all cassava
plants identified.We also found that,
of all the cassava varieties reported,
only the one or two sweet ones in
each swidden were planted in a recognizablepatch. The remainingones
were all interplanted.We found 4 to
17 bitter varietiesper samplingunit
(12.57 m2), with a mean of 7.8 +
3.72 (Figure1).
Tukanoan swiddens are most intensivelyused and managedfrom approximately the 12th through the
24th month after burning. During
this time, the first cassava crop is
beingharvestedand some areasof the
swiddenare weeded for replanting.A
second, smaller cassava crop is
planted in swiddenswith good yields
as the firstis being harvested,so that
cassavaharvestingcontinuesthrough
approximatelythe 36th month. Other
3D. L. Dufourand R. Wilshusen,1986, unpublisheddata.
653
Table 1. Numberof crop species,cassavavarieties,and varietiesof other crops in sampleunits of four Tukanoanswiddenplots in Northwest
Amazonia.
Cultivars
Swidden
Size(ha)
Sampleunits
Cropspecies
Cassava
Other
0.35
20
17
48
25
Primaryforest"
0.15
10
17
1
2
Primaryforest
0.78
29
10
39
11
Rastrojot(60-year)
0.70
25
6
37
5
Rastrojo(25-year)
Unpublisheddatafrom 1986. Samplingunitswererandomlyselectedcirclesof 2-meterradius.All plantsin eachcircleweretaggedandthenidentified
by the gardenowner. Names of cassavacultivarswere systematicallycheckedfor overlapwith each gardenowner, but namingof cultivarsbetween
swiddenplots was not cross-checked.
'Tall forests,not cut in the memoryof currentinhabitants,is assumedto be primaryforest.
tRastrojois local Spanishfor successionalforest.Ages are estimatedfrom informants'historiesof use.
crops are harvestedas they mature
and are needed.Tree cropssuch as
uvillaand peachpalmmaturein approximatelythreeyearsand areharvested until forest regrowthdominatesthe plot.
Householdsestablishone or two
new swiddenplots each year, and
thereforetheyhaveaccessto a numberof swiddensin differentstagesat
any time: a newly plantedswidden
with immaturecrops, a cassavaproducingswidden,andone or more
older swiddenor fallowswith fruit
trees, fish poisons, and medicinal
plants.Someof theplotsarecontiguous and forma mosaicof patchesof
successionalvegetation.Othersare
of
widelydispersedto takeadvantage
areasof particularly
goodagricultural
soil and/orserveas basesforhunting
andfishing.
Swiddenfallows.The transitionbetweena swiddenand a swiddenfallow is not sharp,and,certainlyfrom
the Amerindianpoint of view, the
plot is not abandonedafterthe principalcrophas beenharvested.DenevanandPadoch(1988)examinedthe
swiddensof BoraIndianslivingin the
humidtropicalrainforestsof Peruat
differentstagesof regrowthfromthe
timeof cutting.Theyfounda continuum from a swiddendominatedby
cultivatedgrasses,forbs,andshrubs;
to an orchardfallowphase,in which
of fruittrees,
therewasa combination
smallercultigens,andnaturalvegetation;to a forestfallowthatstillconusefulplants,but
tainedeconomically
moreandmore
becameprogressively
forest(Table2).
likethe surrounding
Someof the usefulplantswerenondomesticated
plantsthatappearedin
Tukanoan
womansievinggratedcassavarootsto extractthestarch.Thisextractionis the natural reforestation sequence
of
the
processused to makecassavabread,the dietarystaplein Northwest and were protected.
part
The Bora, therefore, practice a
Amazonia.
654
BioScienceVol. 40 No. 9
form of agroforestry in which perennial tree crops are combined with
natural forest regrowth (Denevan et
al. 1984). Similar practices have been
reported for other Amerindians and
appear to be widespread (Balke and
Gdly 1989, Eden and Andrade 1987,
Harris 1971, Posey 1984).
Forest-fields and trailside plantings.
Posey (1984) has documented among
Kayapo Indians in Brazil a much
broader system of forest management, which he refers to as nomadic
agriculture. Traditionally the Kayapo
were seminomadic and spent much of
the year trekking along an extensive
system of trails between the Tocantins and Araguaya rivers and the
north-south limits of the Planalto and
Amazon rivers.
Anthropologists had assumed that
the Kayapo lived by hunting and collecting wild plant foods during these
treks. However, Posey (1984) found
that the trailsides and campsites were
actually planted with "numerous varieties of yams, sweet potatoes, Marantacea, Cissus, Zingiberaceae,
Araceae, Cannaceae, and other unidentified, edible, tuberous plants," as
well as fruit trees and medicinal
plants. This planting was a conscious
attempt to replicate naturally occuring concentrations of resources in primary forest. The plants used included
Table 2.
both domesticates and semidomesticates (i.e., plants transplanted from
primary and successional forest).
These semidomesticates were also
transplanted to old swiddens and naturally occuring forest gaps.
Ecological effects of traditional agricultural practices. The ecological effects of traditional Amerindian agricultural practices are still poorly
understood. Swidden cultivation, as
practiced by Amerindians with short
cropping and long fallows, is considered a relatively benign disturbance
(Herrera et al. 1981), and it does not
seriously impair ecosystem function
(Uhl 1987).
At the level of the individual swidden plot, a number of traditional
farming practices are considered beneficial. For example, the use of shade
trees, mixed cropping of species that
differ in phenology, dense spacing of
crops, and fallowing all help preserve
soil organic matter, which is a critical
factor in the maintenance of soil fertility in the deeply weathered and
leached soils common in Amazonia
(Ewel 1986). The species richness of
the plots is assumed to confer pest
protection and decrease the risk of
complete crop failure (Ewel 1986; c.f.
Brown and Ewel 1987). Planting
many varieties of a staple crop may
fulfill the same functions (Balke and
Gdly 1989, Beckerman 1983, Boster
1983, Parker et al. 1983).
Forest regeneration during the fallow is considered a key to the sustainability of swidden systems (Ewel
1986). Amerindian management of
the successional process, by the selective weeding out of certain trees and
the protecting and planting of others,
appears to have a greater impact on
species diversity than it does on forest
regrowth in stature and biomass (Uhl
1987). Uhl (1983) has suggested that
heavily managed successions in areas
such as house gardens may grow to
forest stature as fast or faster than
natural ones (Uhl 1983).
Recovery to primary forest in terms
of biomass and species diversity,
however, is slow. Current estimates
are that in areas like San Carlos de
Rio Negro it will take 100 years or
more for traditionally farmed sites to
return to primary forest, and sites
suffering greater disturbance will take
even longer (Saldarriaga 1985, Uhl
and Murphy 1981).
Given this long period of recovery,
and the practice of cutting swidden
plots yearly, it is clear that human
settlements will be surrounded by a
complex mosaic of agricultural and
agroforestry plots, as well as forest in
various stages of regrowth. In addition, in some areas, the species composition of the forests themselves will
Harvestable plants found in Bora fields and fallows at different stages of regrowth after cutting. (Adapted from Denevan et al. 1984.)
Stage
Newly planted field (0-3 mo)
New field (3-9 mo)
Mature field (9 mo-2 yr)
Transitional field (1-5 yr)
Transitional fruit field (4-6 yr)
Orchard fallow (6-12 yr)
Forest fallow (12-30 yr)
Cultigens
None
Maize, rice, cowpeas
Manioc, some tubers,' bananas, cocona,
other quick-maturing crops
Replanted manioc, peanuts, pineapple,
guava, caimito, uvilla, avocado, cashew,
coca, barbasco, chili peppers,
miscellaneous tubers
Peach palm, bananas, uvilla, caimito, guava,
annatto, coca, some tubers, propagules of
pineapple, other crops
Peach palm, some uvilla, macambo,
propages
Macambo, umari, breadfruit, copal
Other useful plants
Dry firewood
Various useful early successional species
Some useful vines and herbs in abandoned edge
Useful medicinals and other plants within field;
on edges, seedlings of useful trees appear;
saplings of Cercropia and Ochroma lagopus in
abandoned edges
Many useful soft construction woods and
firewoods, palms (including Astrocaryum),
useful vines, and understory aroids
Useful plants as above and self-seeding Inga
Self-seeding macambo and umari; high forest
successional species appearing; some
hardwoods becoming harvestable (e.g., Iriartea
sp.); many large palms (Astrocarym hunicungo,
Euterpe sp., Jessenia bataua)
Old fallow (30+ yr)
Numerous construction, medicinal, handicraft,
Macambo, umari
and food plants
Plant identifications are as follows: Annatto (Bixa orellana), avocado (Persea americana), banana (Musa sp.), barbasco (Lonchocarpus sp.),
breadfruit (Artocarpus incisa), caimito (Pouteria caimito), cashew (Anacardium occidentale), chili pepper (Capsicum sp.), coca (Erythroxylon coca),
cocona (Solanum sp.), copal (Hymenaea courbaril), cowpeas (Vigna unguiculata), guava (Inga sp.), macambo (Theobroma bicolor), maize (Zea
mays), manioc (Manihot esculenta), peach palm (Bactris gasipaes), peanut (Arachis hypogaea), pineapple (Ananas comosus), rice (Oryza sativa),
umari (Poraqueilba sericea), uvilla (Pourouma cercropiaefolia).
*These tubers include cocoyams (Xanthosoma sp.), sweet potatos (Ipomoea batatas), and yams (Dioscorea trifida macrocarpa).
October 1990
655
swiddenin Northwest
cassavarootsin a representative
Tukanoanwomanharvesting
Amazonia.Thepalmate-leafed
plantat lowerleft is cassava.
be the result of human endeavorsto
increasethe density of useful plants.
Such vegetationalmosaics may offer
advantagesin termsof pest protection
and be a way of risk spreading(Eden
and Andrade 1987, Ewel 1986). For
indigenous populations, these landscapesoffer clear advantagesin hunting and collecting.
Amerindianuse of forests for
hunting, fishing, and collecting
Hunting, fishing, and collecting are
integral components of the subsistence pattern of Amerindians that
656
successional vegetation of swidden
are attractiveto such gameanimalsas
rodents,peccary,and deer.
Crop losses due to predation are
routinely compensated for by overplanting (Balkeand Gdly 1989, Carneiro 1983). Tukanoan women, for
example, plant extra sweet-cassava
for a small rodent, boo (Dasyprocta
punctada, which averages approximately 2 kg in weight), and only
complainwhen "he" seems to be eating more than "his" share. This rodent accountedfor more than 20% of
all animalskilledduringthree-monthlong observationperiodsin 1977, and
women huntingin gardenswith dogs
were responsiblefor almosthalf of all
kills (Dufour 1981).
The fruit trees in swidden fallows
are attractive to a number of large
game animals, especiallytapir (Tapirus terrestris)and peccary (Tayassu
spp.; Balie and G6ly 1989, Chagnon
and Hames 1979, Denevan et al.
1984, Dufour 1981, Posey 1984).
The Kayapo purposefully disperse
their gardens so as to attract game
animals over a large area (Parkeret
al. 1983, Posey 1982, 1983, 1984).
For this reason,Posey (1984) suggests
that old swiddens be called "gamefarm-orchards."
The effect of indigenous hunting
practiceson animal biomass and diversity is a complex question. Posey
(1982) has argued that certain game
species would not occur in forest unmodified by humans, and several of
the importantmammalssuch as deer,
tapir, and collaredpeccarymay reach
higher densities in modified areas.
Further,not all animalsthat frequent
the swiddensare taken for food (Ross
1978), and researchershave described
a numberof ways in which the Amerindians may be regulating their
huntingof game animals(Balde1985,
Beckerman1980, Reichel-Dolmatoff
cannot be considered apart from
swidden agriculture(Baldeand Gdly
1989). In nutritionalterms, these activities providethe sourcesof dietary 1976).
The only long-term study of the
protein, fat, and other nutrientsthat
are critical supplementsto the high- effects of indigenous hunting is Vickers' (1988) documentation of the
carbohydratestaples.
Faunalresources.Amerindianshunt a
wide variety of animals, the numerically most important of which are
primates and rodents (Redford and
Robinson1987). One of the explicitly
recognizedfunctionsof swiddenplots
is to attract game animals for hunting. The roots, tubers, and low-
hunting returns in a Siona-Secoya village in northeastern Ecuador over a
10-year period. His data suggest that
some species were being depleted locally. These included the woolly monkey (Lagothrix lagotricha), a large
forest understory bird called the
curassow (Mitu salvini), and a large
ground-dwelling bird, the trumpter
BioScienceVol. 40 No. 9
seasonally important in the diets of
some groups (Dufour and Zarucchi
1979).
Plant foods are collected from the
entire range of successional vegetation types, from nondomesticated
herbs such as Phytolacca rivinoides,
which grow in newly burntswiddens,
to the seedsof the rubbertrees,Hevea
sp., which grow in primary forests
(Dufour 1981). In collecting plant
parts(fruits,nuts, and seeds)fromthe
litteron the forestflooror fromliving
trees, humans are competing with
other herbivores,but the impact of
their activityon the forest is minimal.
When trees are felled to harvestfruit
or other parts, the impact is greater.
Localizeddepletionof productssuch
as cedar (Cedrelaodorata)for canoes
and palms for roof thatch has been
agesin NorthwestAmazonia.
Obliqueaerialviewof swiddenplotsof different
documented(Vickers1988).
Some of the more important collected plant foods appearto be from
(Psophia crepitans). Other species, dammed.Theeffectsof thesepoisons anthropogenicforests, that is, forests
such as peccaries, tapir, deer, other appear to be temporary (several that are the result of human disturbance. Balke (1989) has argued that
primates, other birds, rodents, and hours)andhighlylocalized.
babassu
Small
COLLECTING.
evishow
did
vertebrates,
palm (Orbignya phalerata)
not, however,
reptiles,
dence of depletion. These data sup- suchas frogs,as wellas invertebrates,forests and Brazil nut (Bertholletia
portthe suggestionthat the patchesof are importantfaunalresourcesthat excelsa)forests,amongothers,should
successional vegtation created in are collected.The use of insectsfor be consideredanthropogenic,because
swidden cultivationmay allow some foodis widespread
(Dufour1987,Po- the predominanceof these treesin the
game species to survive near human sey 1978).The morecommonlycol- forest is associatedwith evidence of
settlements (Redford and Robinson lectedand consumedinsectsappear
Attaspp.),terto be ants (especially
1987).
FISHING.The majorityof the Am- mites,andlarvaeof bothColeoptera
erindiansin Amazonia rely on fish, (especiallyBuprestidae,Curculionand Lepirather than game, as their principal idae, and Scarabaeidae)
harvestantsand
sourceof animalprotein.In blackwa- doptera.Tukanoans
ter areaswhere riversare small, such termitesat low but constantrates.
as northwesternAmazonia, fish can Nests areneverdestroyed,and some
be consideredpart of the forest eco- colonies in favorablelocationsare
system, because nonpredatory fish activelyprotected.Palmgrubs(Curfeed primarily on forest products culionidae)are a managedresouce:
(Knoppel 1970). Tukanoan Indians palms are cut with the expectation
recognizethe importanceof foreststo that theywill be invadedby weevils
fish (Chernela1985, Dufour 1981). and the larvaecan be harvestedat a
The flood forest, or igapo, is an im- laterdate.
portant feeding ground for fish. The
collect
Indianshave protectedit from defor- Floralresources.Amerindians
a wide range of plants and plant
estation (Chernela1985).
OtherTukanoanpractices,such as productsas foodandfor usein houscraftproducthe useof fishpoisons, may negatively ing, tool manufacture,
affectthe local fishpopulations,but it tion, and medicine.The role of colis not clear to what extent. The fish lectedplantfoods in the diet ranges
poisons are the crushedroots, stems, from trail snacks and emergency
and/orleavesof a varietyof wild and foods to importantsourcesof nutricultivatedplants.Theiruse was tradi- ents. Palmfruitsand Brazilnuts are
tionallycontrolledby the village sha- well-knownexamplesof wild plant
man (Reichel-Dolmatoff1976) and foods.Lesswell-knownareoil seeds
typically restricted to small forest suchas Erismajapuraandthelegume Tukanoanboyfishingwithhookandline
streams that could be temporarily Monopteryxangustifolia,which are froma smalldugoutcanoe.
October 1990
657
some of the resourcesof the forest.
Changingviews of native
Amazonianresourceuse
Seedsof Monopteryx
a seasonally
wildplantfoodin Northwest
angustifolia,
important
Amazonia.
human settlement. Babassu palms
tend to be the dominant,or at least an
important species in burned forest
clearings because of the manner in
which they germinate:the apicalmeristem grows downward, rather than
upward, and remains protected undergroundfor a year or more (Anderson and Anderson1985, Balke1989).
Brazil nut forests, at least in some
areas, appear on or near terra preta
sites (Balke1989). Furthermore,Kayapo actually plant Brazil nuts as a
sourceof food for themselvesand the
game they hunt (Posey 1985). Balke
(1989) estimatesthat at least 11.8%
of terrafirmein the BrazilianAmazon
is coveredby anthropogenicforests.
Contrastswith
nonindigenousfarmers
Indigneouspeoplesare not the only
oneswho havea detailedknowledge
of thelocalAmazonianecosystemsin
whichthey live. The long-termresidentsof Amazonia,thecaboclos(also
riberenos, mestizos, or campesinos),
do as well.Thecaboclosarethe rural
peasantryof Amazonia, and they are
now the principalinhabitantsof the
vcrzea, the narrow but productive
floodplainof Amazonianrivers(Parker et al. 1983). Their use of resources
resemblesthat of Amerindians,
from
river and cultivated subsistence and
market crops on the restingas (river
levees). On the poorer soils of the
higher areas, they had intensively
managed for subsistence crops the
swiddenplots, whichgraduallyturned
into orchardsand then forest fallows.
Some orchardswere used for as long
as 30 years, and the forest fallows
were weededselectivelyto maintaina
high proportionof usefulspecies.
Like Amerindians,caboclos use a
wide variety of forest products as
food, medicinals,fiber, and building
materials.They also collect nontimberforestproductssuchas palmfruits,
Brazilnuts,andrubberfor commercial
sale (Padoch1988, Parkeret al. 1983).
At least for Brazil nuts and rubber,
their collection practicesare productive and environmentally
conservative,
and they provideexamplesof the sustainable use of Amazonian forests
(Fearnside1988).
The newest immigrantsto Amazonian rainforestsare the colonos, or
colonists. Colonos are the subsistence
farmerswho went and still go to the
Amazon as part of resettlement
schemes promoted by government
agencies(Moran 1988). The majority
see the rainforest for the first time
when they arrivein the Amazon,and,
understandably, they have little
knowledgeof how to makea livingin
whom many are descended,but they it.
are more oriented to market econoIn his study of Brazil'sTransamamiesandtypicallyquickto respondto zon Resettlement Scheme in the
market opportunities (Padoch 1988, 1970s, Moran (1988) found that the
Parkeret al. 1983).
colonos treatedthe forestas an enemy
Forexample,the caboclos,studied rather than a resource.They cleared
by Padoch (1988) along the Ucayali more land per year, but cultivated
River in Peru, recognizeda complex less, and they were less successful
set of ecological zones in the riparian agriculturallythan caboclos living in
environment
andusedthemto advan- the same area. They gradually
tage. They cash-croppedrice on the learned, however, to clear less land,
seasonallyfloodedmud flats of the work it more intensively, and use
658
Recent studies have considerablyrefined understandingof the ways in
which Amerindiansuse the tropical
rainforest.Originallyanthropologists
and ecologists envisioned swiddens,
fallows, and forests as more or less
separateentities. Now, however, we
understandmore clearly the process
of swiddens becoming forests, the
length of time involved, and the degree to which human managementis
part of the transition.
The distinctionsbetween domesticated and wild plants, or naturaland
managedforest, are also not as sharp
as we thought they were. Much of
what has been considerednaturalforest in Amazoniais probablythe result
of hundredsof yearsof humanuse and
management (Posey 1984, Smith
1980). We arenot certainhow specific
human activities may have changed
Amazonianecosystemsover the long
term,but they werecertainlyan essential component. Future researchwill
have to take into account the long
historyof occupationand use of these
forestsby Amerindians.
The agriculturalsystems of Amerindians and caboclos have proven to
be more sophisticatedand complex
than we imagined.Thereis a growing
recognition that these agricultural
systems have a great deal to offer in
the design of sustainableagroecosystems (Denevan et al. 1984, Ewel
1986, Hart 1980). Furtherstudy of
these systems that explicitly recognizes and incorporates the detailed
knowledge and long experience of
Amerindiansand caboclos is needed.
Suchstudy will requirethe collaboration of anthropologists and ecologists.
Acknowledgments
Funding for research in the Northwest Amazon was provided by the
National Science Foundation (BSN8519490). I am very grateful to R.
Wilshusen for his invaluable assistance in mapping and sampling Tukanoan swiddens and his helpful
comments on the manuscript. W.
Bale kindly provided prepublication
copies of two of his papers.
BioScienceVol. 40 No. 9
Referencescited
Anderson, A., and S. Anderson. 1985. A "tree
of life" grows in Brazil. Nat. Hist. 94(12):
40-47.
Bale, W. 1985. Ka'apor ritual hunting. Hum.
energy flow in a population of tropical forest
horticulturalists. Ph.D dissertation, State
University of New York, Binghamton.
. 1987. Insects as food. Am. Anthropol.
89: 383-397.
Posey, D. A. 1978. Ethnoentomological survey
of Amerind groups in lowland Latin America. Fla. Entomol. 61: 225-229.
1982. Keepers of the forest. Garden 6:
18-24.
. 1983. Indigenous ecological knowlCyanide content of Cassava
Ecol. 13: 485-510.
(Manihot esculenta, Euphorbiaceae) cultiedge and development of the Amazon. Pages
.(1988. Tukanoan Indians in Northwest
. 1989. The culture of Amazon forests.
vars used
225-256 in E. Moran, ed. The Dilemma of
by
In D. A. Posey and W. Bale, eds. Special
Amazonia. Econ. Bot. 42: 255-266.
AmazonianDevelopment.Westview Press,
issue: ResourceManagementin Amazonia: Dufour, D. L., and J. L. Zarucchi. 1979.
Boulder, CO.
Indigenousand Folk Strategies.Adv. Econ.
. 1984. A preliminary report on diverMonopteryx Angustifolia and Erisma
Bot. 7: 1-21.
sified management of tropical forest by the
Japura: their use by indigenous peoples in
and
A.
1989.
forest
Bale, W.,
Managed
Gdly.
the northwest Amazon. Bot. Mus. Leaf.
Kayap6 Indians of the Brazilian Amazon. In
succession in Amazonia: the Ka'apor case. In
Harv. Univ. 27: 69-91.
G. T. Prance and J. A. Kallunki, eds. EthnoD. A. Posey and W. Balie, eds. Resource
Eden, M. J., and A. Andrade. 1987. Ecological
botanyin the Neotropics.Adv.Econ.Bot. 1:
Managementin Amazonia:Indigenousand
112-126.
aspects of swidden cultivation among the
FolkStrategies.Adv. Econ.Bot. 7: 129-158.
Andoke and Witoto Indians of the Colom. 1985. Indigenous management of
Beckerman, S. 1980. Fishing and hunting by
bian Amazon. Hum. Ecol. 15: 339-359.
tropical forest ecosystems: the case of the
the Bari of Colombia. Pages 67-109 in R. B.
Ewel, J. J. 1986. Designing agricultural ecosysKayap6 Indians of the Brazilian Amazon.
Hames,ed. Studiesin Huntingand Fishingin
tems for the humid tropics. Annu. Rev. Ecol.
AgroforestrySystems3: 139-158.
the Neotropics. Bennington College, VT.
Syst. 17: 245-271.
Posey, D. A., J. Frechione, J. Eddins, L.
. 1983. Bari swidden gardens: crop
P. M. 1989. Extractive reserves in
Francelino da Silva, D. Myers, D. Case, and
Fearnside,
segregation patterns. Hum. Ecol. 11: 85Brazilian Amazonia. BioScience 39: 387P. Macbeath. 1984. Ethnoecology as applied
101.
393.
anthropology in Amazon development.
I. 1987. Swidden in Amazonia and the
Hum. Organ.43: 95-107.
Amazon rim. Pages 55-94 in B. L. Turner Hames, R. B. 1979. A comparison of the efficiencies of the shotgun and the bow in neoRedford, K. H., and J. G. Robinson. 1987. The
and S. B. Brush, eds. Comparative Farming
tropical forest hunting. Hum. Ecol. 7: 219game of choice: patterns of Indian and coloSystems. Guilford Press, New York.
252.
nist hunting in the neotropics. Am. AmthroA
1983.
of
the
Boster, J.
comparison
diversity
pol. 89: 650-667.
of Jivaroan gardens with that of the tropical Harris, D. R. 1971. The ecology of swidden
cultivation in the Upper Orinoco rain forest,
Reichel-Dolmatoff, G. 1976. Cosmology as
forest. Hum. Ecol. 11: 47-68.
Venezuela. Geogr. Rev. 61: 475-495.
ecological analysis: a view from the forest.
Brown,B. J., andJ. J. Ewel. 1987. Herbivoryin
Man 11: 307-318.
R. D. 1980. A natural ecosystem analog
Hart,
and
successional
complex
simple tropical
to the design of a successional crop
D. 1967. Indigenous cultures and lanRibeiro,
approach
68:
108-116.
ecosystems. Ecology
system for tropical forest environments. Bioguages of Brazil. Pages 69-76 in J. H. HopCarneiro, R. 1983. The cultivation of manioc
tropica12: 73-83.
per, ed. Indiansof Brazil in the Twentieth
among the Kuikuru of the upper Xingu.
C.
F.
E.
and
Institute for Cross-Cultural ReH.
Herrera,
Century.
R.,
Jordan,
Medina,
in
65-111
R.
B.
Hames
and
W.
T.
Pages
search, Washington, DC.
Klinge. 1981. How human activities disturb
Vickers,eds. AdaptiveResponsesof Native
the nutrient cycles of a tropical rainforest in Roosevelt, A. 1989. Lost civilizations of the
Amazonians. Academic Press, New York.
Amazonia. Ambio 10: 109-114.
lower Amazon.Nat. Hist. 98(2): 75-83.
Chagnon, N., and R. B. Hames. 1979. Protein
Ross, E. 1978. Food taboos, diet, and hunting
Kn6ppel, H. 1970. Food of central Amazonian
deficiency and tribal warfare in Amazonia:
fishes: contribution to the nutrient ecology of
new data. Science 20: 910-913.
strategy: the adaptation to animals in AmaAmazonian rain forest streams. Kiel Amazozon cultural ecology. Curr. Anthropol. 19:
Chernela, J. 1985. Indigenous fishing in the
niana2: 257-352.
1-19.
neotropics: the Tukanoan Uanano of the
blackwater Uaupis River basin in Brazil and Moran, E. F. 1974. The adaptive system of the Saldarriaga, J. G. 1985. Forest succession in
Amazonian Caboclo. Pages 136-159 in C.
the upper Ri6 Negro of Colombia and VenColombia.Interciencia10: 78-86.
ezuela. Ph.D. dissertation, University of TenWagley, ed. Man in the Amazon. University
Cock,J. H. 1985. Cassava:New Potentialfor a
of Florida Press, Gainsville.
nessee, Knoxville.
Neglected Crop. Westview Press, Boulder,
. 1988. Following the Amazon highCO.
Smith, N. J. H. 1980. Anthrosols and human
ways. Pages 155-162 in J. S. Denslow and C.
carrying capacity in Amazonia. Annals of the
Denevan, W. M. 1976. The aboriginal populaAssociation of American Geographers70:
tion of Amazonia. Pages 205-234 in W. M.
Padoch,eds. Peopleof the TropicalRainforest. University of California Press, Berkeley.
553-566.
Denevan,ed. The Native Populationof the
Americas. University of Wisconsin Press, Nimuendajui, C. 1939. The Apinaye. Anthro- Uhl, C. 1983. You can keep a good forest
Madison.
down. Nat. Hist. 92(4): 71-79.
pological series no. 8. Catholic University of
. 1987. Factors controlling succession
America, Washington, DC.
Denevan, W. M., and C. Padoch. 1988. The
Bora agroforestry project. In W. M. Denevan
Padoch, C. 1988. People of the floodplain and
following slash and burn agriculture in Amand C. Padoch, eds. Special issue: Swiddenforest. Pages 127-140 in J. S. Denslow and
azonia.Ecology 75: 377-407.
Fallow Agroforestryin the PeruvianAmaC. Padoch,eds. Peopleof the TropicalRain- Uhl, C., and P. Murphy. 1981. A comparison
zon. Adv. Econ. Bot. 5: 1-7.
of productivities and energy values between
forest. University of California Press, BerkeDenevan, W. M., J. M. Treacy, J. B. Alcorn, C.
slash and burn agriculture and secondary
ley.
Padoch, J. Denslow, and S. F. Paitan. 1984.
Parker, E., D. A. Posey, J. Frechione, and L. F.
succession in the upper Ri6 Negro region of
Indigenous agroforestry in the Peruvian AmDa Silva. 1983. Resource exploitation in
the Amazon.Agro-Ecosystems7: 63-83.
azon: Bora Indian management of swidden
Amazonia: ethnoecological examples from Vickers, W. T. 1988. Game depletion hypothfallows. Interciencia 9: 346-357.
four populations. Ann. Carnegie Mus. 52:
esis of Amazonian adaptation: data from a
163-203.
Dufour, D. L. 1981. Household variation in
native community. Science 239: 1521-1522.
October1990
659