Comments on Prehistoric Agriculture in Amazonia
William M. Denevan
William M. Denevan is professor emeritus of geography and
environmental studies at the University of Wisconsin-
We know very little about pre-European agricultural
techniques in Amazonia. This is unfortunate, given that state-
ments regarding prehistoric demography and settlement pat-
terns, including size, location, and duration, are based in part
on assumptions about food productivity. There is only scat-
tered physical evidence, and information from ethnohistory
and ethnography is of limited value.
Anthropologists have generally portrayed surviving indi-
genous hunters and gatherers (foragers), shifting cultivators,
and other traditional economies in Amazonia as representa-
tive of prehistoric food-production systems. Even where such
groups have clearly undergone considerable acculturation, it
has been suggested that their food-getting ecologies (and set-
tlement behavior) have changed little since prehistoric times,
even with changes in crops and tools (Meggers 1995: 35). This
perspective, however, is coming under increasing attack. Few
groups anywhere have been isolated from the world econ-
omy and technology, directly or indirectly, not only today but
Archaeologist Anna Roosevelt has stated that in Amazonia
"theories about pre-Conquest subsistence cannot be tested
with ethnographic data" and that "present-day Indians' re-
source management modes may not be representative of
prehistoric ones" (Roosevelt 1989:31). Beckerman (1987:88)
points out that in Amazonia "the systems we see operating
today are for the most part tiny remnants of what was once a
much larger system of farmers and fields." There are com-
parable statements by Colchester (1984:311) and Roe (1994:
198-200), among others. Furthermore, most surviving Indians
are located in the terra firme high forests of the interfluves,
where resource conditions (soils, game, and fish) are relatively
poor, whereas numerous prehistoric Indians, now mostly
extinct, were located in or adjacent to the resource-rich flood-
That low-productivity shifting cultivation was the dom-
inant system is indeed only an assumption. There is no direct
evidence for it. We know that forests were cleared, but this
does not necessarily indicate shifting cultivation. I have ar-
gued in an article in 1992 that stone axes were so inefficient
for clearing forest that long-fallow shifting cultivation, so
common today, would not have been feasible. Too much time
and energy were required, especially for large hardwood
trees. Initial clearings would have been made where trees
were small, as along streams and at tree falls; at forest patches
dominated by palms, bamboo, or lianas; and at sites of former
human activity (villages, camps, trails, fields). Once initiated,
a small clearing could be gradually enlarged to a considerable
size. Once established, a field would likely have been used for
many years, given the labor required for clearing a new field
with stone axes. Thus most terrafirme agriculture, I argue, was
permanent or semi-permanent until metal axes were intro-
duced making frequent shifting of fields practical. Soil fertility
was maintained by ash from in-field burning, organic addi-
tives, integration with tree crops, and the creation of fertile
anthropogenic soils. Even without soil improvement, manioc
production can continue for many years on poor soil, weed
and pest invasion being more of a problem than soil fertility.
Forms of Prehistoric Cultivation
We can designate three general types of prehistoric fields
on the basis of habitat. The first is floodplain cultivation.
Clearly, annual crops were obtained from playas, islands, and
natural levees during periods of low water. The early explor-
ers of the Amazon River gave some indication of this
(Meggers 1996:125-126). However, as Meggers (1996:12,28-29)
and others have pointed out, periodic major floods every five
to ten years cover the entire floodplain destroying most or all
crops. Consequently a safety valve was necessary, and this
probably was the adjacent bluffe, the well-drained edges of
terra firme, and especially bluffs that impinged against navi-
gable channels so that easy access was provided to the main
river channel (Lathrap 1970:44; Meggers 1991:199; Denevan
19%). Thus there apparently was a complementary system in
which villagers farmed both the good soil/high risk flood-
plains (vdrzea) and the poor soil/low risk bluff fringe (terra
The second field type is the raised field. Remnants of
these ridges, platforms, and mounds are located in seasonally
flooded savannas in northern Bolivia (Denevan 1966; Erickson
1995), the Orinoco Llanos in Venezuela (Zucchi and Denevan
1979), and the coasts of Guyana, Suriname, French Guiana,
and northern Amapa, Brazil (Rostain 1991). They may well
have existed on Marajo Island and elsewhere, but if so they
have since been destroyed or buried under sediment. The
raised fields we discovered in the Llanos de Mojos in Bolivia
in 1961 were the first indication of intensive prehistoric
Culture & Agriculture 54 Vol. 20, Nos. 2/3 Summer/Fall 1998
agriculture in Amazonia outside the major floodplains. These
fields are as much as 25 m wide, 350 m long, and number in
the tens of thousands. They were probably cultivated con-
tinuously, or nearly so, given practices with similar fields in
Mexico and in the Old World tropics today. Fertility could
have been maintained by additives of rich organic muck that
accumulated in the ditches between the fields.
The third general form of prehistoric agriculture is terra
firme cultivation. We can speculate that, using stone axes,
long-fallow shifting cultivation in the upland forests was rare;
instead semi-permanent short-fallow systems were more
common, and these were integrated with and rotated with
permanent gardens and managed agroforests, both domi-
nated by useful perennial trees and especially by fruit trees.
There are numerous mentions of the importance of fruit trees
in the sixteenth-century accounts. The most intensive terra
firme agriculture, however, was probably on anthropogenic
black earths, terra preta do indio, and I want to give particular
attention to them.
These soils have long been known, but they remain
poorly studied and seldom have been mapped (Woods 1995:
159-160). However, their locations are well known to local
people who farm them because of their high fertility. The
black color of the soil is apparently due to ash from repeated
in-field burning and from household fires. Fertility is en-
hanced by relatively high levels of organic material, calcium,
and phosphorus, with higher pH and moisture levels than for
surrounding soils. Terra preta soil has been dated to 450 B.C.
near Manaus and 200 B.C. on the Rio Ucayali in Peru (Eden,
On bluffe, terra preta sites have been described which are
80, 90, 200, and 350 ha in extent, though averaging 20 ha
(Smith 1980:560,1999:26). Anna Roosevelt (1987:157) believes
there are 500 ha (5 km2
) of terra preta underlying the city of
Santarem. Some sites are several kilometers long, thus sup-
porting sixteenth-century reports of large riverine villages
extending for several leagues (Denevan 1996:659). Meggers
(1995:27-28), on the other hand, argues that only patches of
these large sites were occupied at any one time. Furthermore,
only portions of terra preta sites are actually middens from
past villages (Woods and McCann 1999).
On interior terrafirme, small terra pretas are described by
Smith (1980:563) along the Trans-Amazon Highway as being
only 1 to 2 ha or less, suggesting single or several houses that
remained in place long enough to create terra preta. Katzer
(1944:35-38) reported 50,000 ha of mostly small terra preta sites
between the Rio Tapajos and the Rio Curua-Una. This pattern
seems supportive of the Tropical Forest Model of small terra
firme forest communities and low population densities in con-
trast to large riverine villages (Raymond 1994:177). In recent
years, however, several studies have demonstrated that terra
preta sites on terrafirme can be enormous.
In 1996, geographer/ethnobotanist Joseph McCann and
geographer/soils archaeologist William Woods examined terra
pretas in the vicinity of Santarem and between the Rio Tapajos
and the Rio Arapiuns, where some sites are very large (over
120 ha at Oitavo Bee south of Santarem; Woods and McCann
1999; McCann and Woods, n.d.). Smith (1980) and others have
suggested that terra pretas are primarily prehistoric middens
containing ceramics, bones, ash, and other settlement refuse.
McCann and Woods, however, found sites which are not uni-
form middens. Sectors with ceramics are separated by large
sectors of black or brown soils without ceramics.1
Woods believe that the non-midden terra pretas were formed
by long-term agricultural activity including in-field burning,
mulching, and composting. Phosphorus and calcium levels
are lower than for the midden terra pretas, but organic matter
content is high. The non-midden terra pretas consequently are
still more fertile than the surrounding natural soils. Once
established, both types of terra preta would have been sought
out for cultivation, as they still are today.
McCann and Woods see no natural explanation for terra
They occur on a wide variety of slopes and sub-
soils; they are both clayey and sandy; they are surrounded by
the typical terra firme red and yellow tropical soils.3
abandoned for long periods they still retain their fertility. In
fact, there is some evidence that they are not only self perpet-
uating but actually expand because of intense microbiological
activity. "We suggest that some combination of long-term
mulching and frequent burning produces the heightened or-
ganic content and dark color of the [non- midden terra preta]
expanses. Though perhaps temporarily reducing near-surface
soil biota,firemore importantly contributes charcoal and ash,
which increase soil pH and thereby suppress Al activity toxic
to soil biota" (Woods and McCann 1999). "This increased
micro-biological activity over time is the key to terra preta
formation and persistence" (Woods, pers. comm.).
The creation of a brown soil from intensive, long-lasting
agricultural activity, in contrast to an origin from middens,
has been suggested in several other studies. These studies,
however, do not emphasize that agricultural terra preta may
be much more extensive than midden terra preta. The Dutch
soil scientist, W.G. Sombroek (1966:175), mentions a brown or
less dark soil, which he calls terra mulata, in the Belterra area
east of the Rio Tapajos, without artifacts, occurring in bands
around black terra preta. He believes that "It seems likely that
this soil has obtained its specific properties from long-lasting
cultivation." His map shows terra preta along a bluff backed
by a much larger area of terra mulata. At Araracuara on the Rio
Caqueta in the Colombian Amazon, brown anthropic soil
formed where "cultivation was done in the mature or second-
ary forest in a semi-intensive way, probably located away
Culture & Agriculture 55 Vol. 20, Nos. 2/3 Summer/Fall 1998
from the living area, and always done in the same place to
improve the soil" (Andrade 1986:54, in Mora et al. 199177).4
The Araracuara project is a long-term study by Colombian
archaeologists and ecologists (Mora et al. 1991). The project
has investigated terra preta sites on a high bluff. Settlement at
one of these sites, Abeja (6 ha.), was nearly continuous for 800
years (1565-775 B.P.) Pollen indicates large quantities of fruit
trees, along with maize, manioc, and other crops. The soil
scientists involved believe that permanent cultivation was
made possible by the transport to fields of alluvial silt from the
river and incorporation of organic matter consisting of do-
mestic waste, dead leaves, wood, weeds, and algae—a major
Another study of upland terra preta sites is the 1996 disser-
tation by archaeologist Michael Heckenberger in the Upper
Xingu Basin. As in other parts of southwestern Amazonia, in-
cluding Mojos in Bolivia, many prehistoric and historical sites
are encircled by defensive moats, several moats in some in-
stances. He found 19 sites in forests near streams, with terra
preta soils. One, Nokugu, is 40 ha (15 ha of terra preta), and
another, Kuhikugu, is 50 ha (mostly dark brown terra preta).
(Recent Kuikuru villages near both ancient villages are only
a few hectares in size.) Besides the ditches there are mounds
and causeways. At Nokugu, there is cultural and occupational
continuity from A.D. 950 to after 1500, with a large plaza fixed
in place the entire time, and a possible late population of 2,500
given the apparent occupation of the full area (except for the
plaza) within two moats. Why the second, outer moat?
Heckenberger believes that the only reason for digging a
second moat was because population growth filled the area
within the inner moat. This is part of his argument for the full
site being occupied atone time, rather than being periodically
reoccupied by small villages. These large villages were pro-
bably supported by relatively intensive agriculture, including
use of terra preta soils (Heckenberger 1996:40, 47, 54, 98-100;
Without question, then, there were large prehistoric occu-
pation sites in Amazonia, as evidenced by terra preta middens.
Some may represent periodic reoccupations of different sec-
tors, others not. Others apparently were dispersed clusters of
houses surrounded by agriculture, as evidenced by non-mid-
den brown soils. But even if a terra preta site of 100 ha was
only 20 percent occupied at one time by houses and associ-
ated activity areas, 20 ha is still a very large village compared
to the much smaller villages of today.
In any event, it is becoming apparent that large areas of
terra preta are not middens, but rather are zones of cultivation
that: (1) were often intensive; (2) could have sustained rela-
tively large villages; and (3) were maintained over long per-
iods and thus were associated with either permanent settle-
ments of whatever size, and/or villages that shifted within the
agricultural zone. The fertile terra pretas created, both black
and brown, themselves became the foci of cultivation and
hence settlement, a self-perpetuating process. Such a site
could have begun as a single household at a natural tree-fall
farm plot, gradually being enlarged with stone axes, the ulti-
mate village size possibly having nothing to do with natural
soils but rather with particular local or regional social events
Today in the uplands new terra preta is seldom created by
either Indians or settlers.5
(One example would be in long-
lasting house gardens.) Considerable time is required for terra
preta formation, decades or more, whether from middens or
from semi-intensive cultivation practices. Now Indian villages
are usually moved too frequently and cultivation is of too
brief duration for terra preta to form.
Possible Cropping Systems on Terra Firme
I am suggesting that the short-cropping, long-fallow shift-
ing cultivation so widespread today was uncommon in pre-
historic Amazonia because of the inefficiency of the stone axe,
especially in the mature hardwood forests of the terra firme.
Indian shifting cultivation now has a short cropping period,
reflecting poor soil, pest invasion, game depletion, and social
friction, but it is made possible by the steel axe which makes
clearing new plots a relatively easy process—a matter of a few
weeks to create a field large enough (0.5-2.0 ha) to feed a fam-
ily. Even on the fertile terra preta soils, shifting cultivation is
now the norm because it is easier than coping with the ag-
gressive weeds associated with permanent cultivation. (The
same is true on limestone and volcanic soils in Yucatan and
Indian shifting cultivation as we know it today is the pro-
duct of the steel axe, and a]so the machete. What then was the
nature of prehistoric upland agriculture? We do not know
and may never know. However, there are several possibilities:
(1) House gardens: permanent plots of mixed annuals and
perennials around houses, with careful weed control and soil
management using household refuse for fertilizer. Lathrap
(1977), in his classic article "Our Father the Cayman, Our
Mother the Gourd," maintained that the earliest agriculture in
Amazonia was in such house gardens. He believed that the
first gardens were along or near rivers; however, they were
undoubtedly also an important form of prehistoric crop pro-
duction hi the interfluve forests since they do not require fre-
quent clearing. Today Indian house gardens are poorly devel-
oped in most forest villages given the frequency of village
(2) Intensive swiddens: located on sites where tree clearing
was relatively easy, such as naturally disturbed places with
young secondary growth of softwoods. A present-day exam-
ple of such swiddens would be the highly diverse (polycul-
tural) conucos described by Harris (1971) for the Waika
Culture & Agriculture 56 Vol. 20, Nos. 2J3 Summer/Fall 1998
(Yanomamo) of the Upper Orinoco, which are cultivated for
up to six years. Such fields contrast with the monocultural
swidden dominated by a single species, usually manioc,
which is the common form of tropical-forest Indian field
today. Most current monocultural fields are long fallow and
are only used for only one to three years. Beckerman (1983:4-
6) gives several reasons for the monocultural field, but he does
not consider the role of the steel axe in making short-lived
Intensive swiddens (short fallow or semi-permanent)
were possible, given known indigenous practices today.
Weeds could have been controlled by shading and hand
weeding; pests could have been reduced by crop diversifica-
tion; soils could have been artificially maintained to varying
degrees by short fallowing, mulching/composting, in-field
burning, and other internal and external organic inputs
(Hecht and Posey 1989:180-186); and fertile soils could have
been unintentionally created by settlement and cultivation ac-
tivity in theformof terra preta. Labor inputs often would have
been high, but probably not as high as that required by long-
fallow shifting cultivation, with frequent field shifts, necessita-
ting frequent and difficult tree clearing with stone axes. This
assumes that time and energy are critical factors, and they
usually are (Hill and Kaplan 1989:331).
(3) Patch cultivation: the planting of small natural clear-
ings, such as tree falls, or easily cleared vegetation such as
bamboo. This could have been done by bands of nomadic for-
agers who returned periodically to small plantings. House-
holds could also have been permanently settled at small clear-
ings, as suggested by terra preta sites of 1 ha or less. In addi-
tion, people in large villages with surrounding fields could
have obtained additional production from small outlying
patches. The Kayapo villagers today plant both domesticates
and semi-domesticates in natural clearings created by tree
falls, as well as along trails and at camp sites (Posey 1984:117,
122). Another report of tree-fall planting is for the semi-
nomadic Nambiquara in the Guapore Valley in 1968: "we
came to a place where a huge tree had fallen, taking several
smaller trees with it. Among the tangle of fallen limbs, tobacco
plants were growing" (Price 1989:127).
(4) Agroforestry: forest manipulation via intentional and
unintentional planting and management of perennial crops
along trails, campsites, fallow swiddens, and other activity
areas (Posey 1985; Denevan and Padoch 1988). These "food
forests" contained domesticates, semi-domesticates, and spon-
taneous tree growth, much of which was managed for useful
species. Suggestions of this can be seen in the Kayapo "forest
fields" and the Bora "forest orchards," as well as in the anthro-
pogenicforestsof the Huastec in Mexico (Alcorn 1984) and the
"tree gardens" in Bocas del Toro, Panama (Gordon 1982:52-98).
Gordon points out that the introduction of the machete was
detrimental to such forest management. Clearing and weed-
ing by hand better allows for decisions as to what plants are
to survive and which are to be destroyed, whereas slashing by
machete tends to be less selective. He believes that polycul-
tural milpas were an integrated component of anthropogenic
forests. However, while agroforestry systems do not require
frequent clearing, they are "shade" systems which suppress
weeds; thus they are not conducive to production of staple
annual crops. Hence population densities would have re-
mained low, unless associated with primary fields.
These four models of terra firme agriculture with a stone-
axe technology in reality were likely manifested by numerous
transitional forms and combinations, varying with habitat,
mobility, time, and demography. These activities, plus forag-
ing, contributed to the creation of anthropogenic forests, or
semi-managed forests, with a larger than natural number of
useful plants present. The Amazon forest was not pristine in
1492, nor is it today. Probably all of these forms of agriculture
and agroforestry were present in the terra firme in a mosaic of
variable population densities that may have included sectors
of sparse semi-nomadic foragers; small but permanently set-
tled households and extended families; and in some selected
places large and permanent fields and associated villages,
such as on Amazon bluffs, in the Upper Xingu Basin, and in
the Rio Arapiuns Basin, where there are large terra preta sites.
Thesefieldsand villages could have originated in small clear-
ings which were enlarged over a long period of time by grad-
ually eliminating the trees at the periphery. By people coming
back repeatedly and frequently to the same site and/or by the
shifting of houses and fields within the same site, the inter-
fluve terra preta black and brown soils could have been
The evidence is still preliminary, but it appears that on
terra firme relatively intensive prehistoric agriculture pro-
duced a long-lasting, self-perpetuating, fertile, anthropogenic
soil (terra preta), which made possible continuing cultivation
that in some places supported relatively large and fairly per-
manent villages. How widespread large and small terra preta
sites are we do not yet know. Regardless, the implications for
both past and present tropical agriculture are startling6
Upland Amazonian soils are considered by many to be
mostly too infertile to support permanent cultivation
(Meggers 1996:18-23; Gross 1983:445-446; Lamb 1987:434-
440). However, Sombroek, Andrade, and Woods/McCann be-
lieve that permanent or semi-permanent agriculture itself
created fertile soils. What a paradox!
Culture & Agriculture 57 Vol. 20, Nos. 2J3 Summer/Vail 1998
*In one sampling of 18 soil tests in four terra preta sites, Woods (pers.
comm.) found only three midden types of terra preta, which is about 17
Falesi (1974:210-214) discusses possible natural origins of terra preta.
Soil color and other characteristics of terra preta vary considerably.
Undoubtedly different kinds of terra preta originated under different
human circumstances and on different original soils (Sombroek 1966:
However, Eden et al. (1984:137), who also worked on the soils during
the early stages of the Araracuara Project, stated that: "there is no reason
to assume that the soils themselves were originally the direct result of
agricultural activity." Thus there is a difference of opinion with Andrade.
^abst (1993:142) in his study of villages of five contemporary terra firme
Indian groups in the eastern Amazon found little evidence of active terra
preta formation in sectors of refuse accumulation and no intentional
effort to create terra preta.
Woods and McCann (1999) say that: "Once infused with the self-
perpetuating life force of an active soil biota and an adequate nutrient re-
tention capacity, under proper management additional inputs may not
be necessary to maintain a reasonable fertility." Thus, "agricultural sys-
tems more in-tensive than shifting cultivation seem possible" (Woods,
Sombroek (1966:261) did not get excited over the implications for de-
velopment suggested by terra preta produced by cultivation. He says that:
"Theoretically, it would be possible to attain gradually a level of soil or-
ganic matter which is comparable to that of the Terra Preta soil;" how-
ever, he goes on to say that "whether it will be economically justifiable,
is questionable." The possibility remains to be demonstrated, obviously.
1984 Huastec Mayan Ethnobotany. Austin: University of Texas
1986 Investigacion arqueologica de los antrosoles de Araracuara.
Fundacion de Investigaciones Arqueologicas Nacionales.
Bogota: Banco de la Republica.
1983 Does the Swidden Ape the Jungle? Human Ecology 11: 1-12.
1987 Swidden in Amazonia and the Amazon Rim. In Comparative
Farming Systems. B.L. Turner and Steven B. Brush, eds. Pp.
55-94. New York: Guilford Press.
1984 Rethinking Stone Age Economics: Some Speculations
Concerning the Pre-Columbian Yanoama Economy. Human
Denevan, William M.
1966 The Aboriginal Cultural Geography of the Llanos de Mojos of
Bolivia. Ibero Americana 48. Berkeley: University of California
1992 Stone vs Metal Axes: The Ambiguity of Shifting Cultivation in
Prehistoric Amazonia. Journal of the Steward Anthropological
1996 A Bluff Model of Riverine Settlement in Prehistoric Amazonia.
Annals of the Association of American Geographers 86:654-
Denevan, William M. and Christine Padoch, eds.
1988 Swidden-Fallow Agroforestry in the Peruvian Amazon. Ad-
vances in Economic Botany, Vol. 5. New York New York
Eden, Michael J., Warwick Bray, Leonor Herrera, and Colin McEwan
1984 Terra Preta Soils and their Archaeological Context in the
Caqueta Basin of Southeast Colombia. American Antiquity
Erickson, Clark L.
1995 Archaeological Methods for the Study of Ancient Landscapes
of the Llanos de Mojos in the Bolivian Amazon. In Archae-
ology in the Lowland American Tropics: Current Analytical
Methods and Applications. Peter W. Stahl, ed. Pp. 66-95.
Cambridge: Cambridge University Press.
Falesi, Italo Claudio
1974 Soils of the Brazilian Amazon. In Man in the Amazon, Charles
Wagley, ed. Pp. 201-229. Gainesville: University Presses of
Gordon, Burton L.
1982 A Panama Forest and Shore: Natural History and Amerindian
Culture in Bocas del Toro. Pacific Grove: Boxwood Press.
Gross, Daniel R.
1983 Village Movement in Relation to Resources in Amazonia. /?i
Adaptive Responses of Native Amazonians. Raymond B.
Hames and William T. Vickers, eds. Pp. 429^149. New York
Harris, David R.
1971 The Ecology of Swidden Cultivation in the Upper Orinoco
Rain Forest, Venezuela. Geographical Review 61:475-495.
Hecht, Susanna B. and Darrell A. Posey
1989 Preliminary Results on Soil Management Techniques of the
Kayapo Indians. Advances in Economic Botany 7:174—188.
New York New York Botanical Garden.
Heckenberger, Michael J.
1996 War and Peace in the Shadow of Empire: Sociopolitical
Change in the Upper Xingu of Southeastern Amazonia, A.D.
1400-2000. Ph.D. dissertation. Pittsburgh: University of
1998 Manioc Agriculture and Sedentism in Amazonia: The Upper
Xingu Example. Antiquity 72:633-648.
Hill, Kim and Hillard Kaplan
1989. Population and Dry-Season Subsistence Strategies of the Re-
cently Contacted Yora of Peru. National Geographic Research
Culture & Agriculture 58 Vol. ID, Nos. 2/3 Summer/Fall 1998
1944 A term preta. Boletim da Seocao de Fomento Agrfcola no Estado
do Para 3(2):35-38.
Lamb, F. Bruce
1987 The Role of Anthropology in Tropical Forest Ecosystem Re-
source Management and Development, journal of Developing
Lathrap, Donald W.
1970 The Upper Amazon. New York: Praeger.
1977 Our Father the Cayman, Our Mother the Gourd: Spinden
Revisited, or a Unitary Model for the Emergence of Agricul-
ture in the New World. In Origins of Agriculture. Charles A.
Reed, ed. Pp. 713-751. The Hague: Mouton.
McCann, Joseph M. and William I. Woods
n.d. Suelos oscuros antropogenicos en la Amazonia: Implicaciones
para la sostenibilidad. In Desarrollo sostenible en la Amazonia:
Mito o realidad? Mario Hiraoka and Francis Kahn, eds. Lima:
Instituto Frances de Estudios Andinos, in press.
Meggers, Betty J.
1991 Cultural Evolution in Amazonia. In Profiles in Cultural Evolu-
tion. A. Terry Rambo and Kathleen Gillogly, eds. Pp. 191-216.
Anthropological Papers No. 85. Ann Arbor: Museum of
Anthropology, University of Michigan.
1995 Judging the Future by the Past: The Impact of Environmental
Instability on Prehistoric Amazonian Populations. In Indi-
genous Peoples and the Future of Amazonia: An Ecological
Anthropology of an Endangered World. Leslie E. Sponsel, ed.
Pp. 15-43. Tucson: University of Arizona Press.
1996 Amazonia: Man and Culture in a Counterfeit Paradise. Re-
vised edition. Washington, D.C.: Smithsonian Institution
Mora, Santiago, Luisa Fernanda Herrera, Ines Cavelier, and Camilo
1991 Cultivars, Anthropic Soils and Stability: A Preliminary Report
of Archaeological Research in Araracuara, Colombian
Amazonia. University of Pittsburgh Latin American Archaeol-
ogy Reports No. 2.
Pabst, Eije E.
1993 Terra Preta: Ein Beitrag zur Genese-Diskussion auf der Basis
von Gelandearbeiten bei Tupi-Volkern Amazoniens. Doctoral
dissertation. Kassel, Germany: Kassel University.
1984 A Preliminary Report on Diversified Management of Tropical
Forest by the Kayapo Indians of the Brazilian Amazon.
Advances in Economic Botany 1:112-126. New York: New
York Botanical Garden.
1985 Indigenous Management of Tropical Forest Ecosystems: The
Case of the Kayapo Indians of the Brazilian Amazon. Agro-
forestry Systems 3:139-158.
Price, David A.
1989 Before the Bulldozer: The Nambiquara 1 ndians and the World
Bank. Cabin John, Maryland: Seven Locks Press.
Raymond, J. Scott
1994 The Intellectual Legacy of Donald W. Lathrap. In History of
Latin American Archaeology. Augusto Qyuela-Caycedo, ed.
Pp. 173-182. Aldershot, England: Avebury.
Roe, Peter G.
1994 Ethnology and Archaeology: Symbolic and Systemic Disjunc-
tion or Continuity? In History of Latin American Archaeology,
Augusto Qyuela-Caycedo, ed. Pp. 183-208. Aldershot,
Roosevelt, Anna C.
1987 Chiefdoms in the Amazon and Orinoco. In Chiefdoms in the
Americas. Robert D. Drennan and Carlos A. Uribe, eds. Pp.
153-185. Lanham, Maryland: University Press of America.
1989 Resource Management in Amazonia before the Conquest:
Beyond Ethnographic Projection. Advances in Economic
Botany 7:30-62. New York: New York Botanical Garden.
1991 Les champs sureleves amerindiens de la Guyane. Paris: Centre
ORSTOM de Cayenne.
Smith, Nigel J.H.
1980 Anthrosols and Human Carrying Capacity in Amazonia.
Annals of the Association of American Geographers 70: 553-
1999 The Amazon River Forest: A Natural History of Plants, Ani-
mals, and People. New York Oxford University Press.
1966 Amazon Soils: A Reconnaissance of the Soils of the Brazilian
Amazon Region. Wageningen: Centre for Agricultural Publica-
tions and Documentation.
Woods, William 1.
1995 Comments on the Black Earths of Amazonia. Papers and Pro-
ceedings of the Applied Geography Conferences. F. Andrew
Schoolmaster, ed. Vol. 18, pp. 159-165. Arlington.
Woods, William I. and Joseph M. McCann
1999 The Anthropogenic Origin and Persistence of Amazonian
Dark Earths. Yearbook, Conference of Latin Americanist
Geographers 25, in press.
Zucchi, Alberta and William M. Denevan
1979 Campos elevados e historia cultural prehispanica en los Llanos
Occidentals de Venezuela. Montalban (Caracas) 9:565-736.
Culture & Agriculture 59 Vol. 20, Nos. 2/3 Summer/Fall 1998