Economic catalysts to ecological change


Material Information

Economic catalysts to ecological change working papers : 39th Annual Conference, Center for Latin American Studies, University of Florida, February 1990
Physical Description:
178 p. : ; 28 cm.
University of Florida -- Tropical Conservation and Development Program
Latin American Conference, 1990
TCD, Tropical Conservation and Development Program, Center for Latin American Studies, University of Florida
Place of Publication:
Gainesville, Fla
Publication Date:


Subjects / Keywords:
Economic development -- Environmental aspects -- Congresses -- Latin America   ( lcsh )
Ecology -- Economic aspects -- Congresses -- Latin America   ( lcsh )
bibliography   ( marcgt )
conference publication   ( marcgt )
non-fiction   ( marcgt )


Includes bibliographical references.
General Note:
Cover title.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 50025081
lcc - HC130.E5 L38 1990
System ID:

Full Text



FOREW ORD ............ .......................................................................................................... 3

CONFERENCE PARTICIPANTS.......................................................................................5

Christopher Uhl.......................................................................................................... 7

A nna Roosevelt.................................................................................................................29

Javier A. Simonetti ard Luis E. Cornejo..................................................................... 65

Douglas Southgate and C. Ford Runge........................................................................79

DEPLETION, 1967-1985
Ana Doris N Capistrano..................................................................................................93

Steven E. Sanderson.......................................................................................................125

Ronald Foresta................................................................................................................ 149

A lcida Rita Ram os...........................................................................................................161




Economic Catalysts to Ecological Change in Latin America

Economy and ecology, inextricably linked, are responsible for the two
great crises of our generation: the loss of biological diversity and the need for
appropriate economic development for improved human welfare. The loss of
biological diversity, especially in the tropics, has reached epidemic
proportions. Development pressure on tropical rainforests, savannah,
wetlands, and marine environments threaten the unique biological
characteristics of such ecosystems. This loss is not offset by gains in human

Indeed, conservation and development are both losers if development is not
ecologically sensitive. Disruption of traditional patterns of land occupancy and
resource use has resulted in fuelwood shortages and forced migration into
marginal and ecologically sensitive lands. Upland watersheds have been
degraded by overgrazing and soil sedimentation of reservoirs, disruption of
irrigated agriculture downstream, and loss of crops, land and human life.
Renewable natural resources are overexploited and lost to future generations.

Nowhere are these crises more evident than in Latin America where the
conflict between ecological and economic concerns has become increasingly
publicized. On February 9-10, 1990, Kent H. Redford and Steven E. Sanderson
convened the Center for Latin American Studies 39th Annual Conference
entitled "Economic Catalysts to Ecological Change in Latin America." In
addition to the CLAS, the conference received support from the Program for
Studies in Tropical Conservation, the Tropical Conservation and Development
Program, the Pew Charitable Trusts Initiative for Integrated Approaches to
Tropirnl Conservation and Sustainable Development, and the departments of
Political Science and Wildlife and Range Sciences.

The conference brought together a wide variety of scholars and
practitioners representing ecology, archeology, anthropology, economics,
political science, and representatives of international conservation and
development organizations. Rarely has such a broad range of expertise been
convened to discuss questions that require, by their nature, input from many
different disciplines. The diversity of interests was typified by the fact that
one of the convenors was a biologist and the other a political scientist.

The conference built on ten formal presentations, each accompanied by a
substantive response by another scholar and public participation. A keynote
address entitled "The political economy of ecological change in Latin America:
Global issues and responsibility" was delivered by Osvaldo Sunkel, UN Economic
Commission for Latin America, Santiago, Chile. The last session was reserved
for a free-ranging discussion on the research and policy agenda raised during
the two previous days.

This volume "Economic Catalysts to Ecological Change" presents eight of the
papers presented at the conference. This format allows wide dissemination of
the papers without constraining the authors from placing their papers in
other publications.


William Bal6e, Museu Goeldi, Bel6m, Brazil.

Enrique Bucher, Centro de Zoologia Aplicada, Universidad de C6rdoba, C6rdoba,

Ana Doris Capistrano, Dept. of Food and Resource Economics, University of
Florida, Gainesville, FL.

Luis Cornejo, Universidad de Chile, Santiago, Chile.

Ron Foresta, Dept. of Geography, University of Tennessee, Knoxville, TN.

Henry Gholz, Dept. of Forestry, University of Florida, Gainesville, FL.

Gary Hartshorn, World Wildlife Fund, Washington, D.C.

Gary Lynne, Dept. of Food and Resource Economics, University of Florida,
Gainesville, FL.

Ricardo Ojeda, C.R.C.Y.T., Mendoza, Argentina.

Alcida Rita Ramos, Departamento de Antropologia, Universidade de Brasilia,
Brasilia, Brazil.

Kent H. Redford, Center for Latin American Studies, University of Florida,
Gainesville, FL.

John Robinson, Wildlife Conservation International, New York, New York.

Anna Roosevelt, American Museum of Natural History, New York, New York.

Carlisle F. Runge, Dept. of Agricultural and Applied Economics, University of
Minnesota, St. Paul, MN.

Steven E. Sanderson, Dept. of Political Science, University of Florida,
Gainesville, FL.

Marianne Schmink, Center for Latin American Studies, University of Florida,
Gainesville, FL.

Javier Simonetti, Universidad de Chile, Santiago, Chile.

Douglas Southgate, Dept. of Agricultural Economics, Ohio State University,
Columbus, OH.

Osvaldo Sunkel, UN Economic Commission for Latin America, Santiago, Chile.

Chris Uhl, Dept. of Biology, Penn State University, University Park, PN.

William Vickers, Department of Sociology and Anthropology, Florida
International University, Miami, FL.






The Amazon is blanketed in green and beneath this mantle lies billions
of cubic meters of wood whose overall value before sawing might be some six
hundred billion dollars.1 After sawing, the value of this wood could easily
approach two trillion dollars.2

The eastern Amazon began to be aggressively developed in the early
1960s with the opening of the Beldm-Brasilia highway, but it was not a wood
crisis that spurred Brazil to open its frontiers. Indeed, the first government-
sponsored settlers were not loggers, but farmers and ranchers. They
regarded the forest as an obstacle and used the age-old technique of cutting
and burning to clear the land. Immense quantities of wood were wasted in
this clearing process. Browder (1988) estimated that between 1966 and 1983,
SUDAM (Superintendency for Amazon Development) subsidized ranchers,
alone, destroyed an estimated 193 million m3 of marketable roundwood (48
million trees). He concludes that this was four times greater than the total
volume of industrial roundwood extracted from the Brazilian Amazon between
1975 and 1980.

While this seems like strange economic behavior on the surface,
Amazon settlers have responded in a rational way to a peculiar set of
government incentives (Mahar 1989; Hecht 1985). The only way to acquire
title to land in Amazonia has been through clearing. Hence, newcomers to
Amazonia rush to lay claim to as much land as possible. INCRA, the
government agency that oversees land titling, initially established that one
could claim 6 ha for every ha cleared thereby providing a built-in incentive
to clear ever-larger tracts. Moreover, land speculation has been an
impoiLa.t part of the deforestation dynamic. Once land is titled the owner is
free to sell it--frequently at substantial profits while paying minimal land-
sale taxes.

It has only been in the 1980s that the value of the forest for its wood is
coming to be appreciated. Two developments have catalyzed this awakening.
First, it has taken some 20 years to establish a reliable transport and
communication system in the eastern Amazon making the establishment of
sawmills and marketing of sawn products easier and less risk prone. Second,
wood supplies in the south of Brazil have been steadily declining in recent
years. For example, in the twelve-year period, 1976-1988, total roundwood
production in Brazil's Southern States (Parand, Santa Catarina, and Rio Grande
do Sul) decreased from 15 million m3 (47% of Brazil's total roundwood
production) to 7.9 m3 (17% of total) (Anuirio Estatistico). During this same
period, roundwood production in the North region (Amazonia) has increased
from 6.7 m3 (21%' of Brazil's total) to 24.6m3 (54% of this total) (Fig. 1A).

The eastern Amazon, in particular the northern part of Para State, has
been a natural place for this burst of logging activity because it is adjacent to
the Northeast with its burgeoning population and paucity of wood resources
and also close to the major port city of Bel6m. Moreover, logging requires a
rich array of support services and, in the North of Pard, a whole cast of
middle men has been available to cut the wood, haul it to the mill, provide
transport of sawn wood to cities, use wood "waste" to make charcoal, etc. This

has meant that sawmills do not need large amounts of capital to get started. As
a result of these factors, Pard has been, by far, the largest producer of
roundwood in Amazonia and in the entire country (Fig. 18).

Hence, Pard (the eastern Amazon) is ideally suited for a study
examining economic catalysts (in this case, demand for wood in Brazil and
abroad) to ecological change. In this paper, we present preliminary results
from research that we have been conducting since November, 1988,
examining the ecological, social, and economic impacts of logging in Parl.


Logging techniques and economies are complex and varied in the
North of Pard and appear to be representative of the range of logging
initiatives occurring in Amazonia as a whole. We have identified four major
forms of logging in the North of Pard (Table 1).

The first two types of logging occur in floodplain forest. One is low
intensity (i.e., highly selective) logging that characteristically occurs in the
more remote (pristine) regions of the vdrzea. In this case, only one species is
extracted, Virola surinamensis. This low intensity logging has been going on
for generations. Local people are involved in extraction but are obliged to
give the harvested boles to large land barons in exchange for subsistence
needs provided at a "company store." The land owners sell the wood to large
sawmills that are involved in export (e.g. Georgia Pacific).

Recently, a second, much more intensive type of logging has become
common in the varzea. In this model, small family-run mills process all
available wood (some 75 species) down to 15 cm in diameter. Over time, virzea
stands normally pass through these two logging stages.

Prior to the construction of highways in Amazonia, terra firme, or
upland, forests were inaccessible to loggers. Over the last three decades,
however, terra firme forests have become an important source of wood in
Amazonia. In areas where road servicing is recent, such as along Pard 150
(connecting Belem and Marabi), logging is highly selective and therefore
non-intensive. Wood harvesting is done by small operators with little capital
and crude machinery. These small-scale extractors and mill operators only
turn a profit when the wood resource is nearby and of relatively high value.
With time, these mills generally relocate closer to the wood resource. Older
frontier areas, such as along the Bel6m-Brasilia Highway, are characterized
by much more intensive logging. This intensive terra firme logging occurs
with the arrival of well-equipped, more sophisticated mills that oversee the
harvesting and processing of all species of value down to 45 cm in diameter.
These mills are frequently integrated vertically (i.e., participate in all aspects
of logging from forest extraction to sawing and selling). Some 150 species are
harvested. Any given stand of terra firme forest will likely pass successively
through both of these logging phases.

In the text which follows, we will provide four case studies: two
focusing on the varzea, treating both intensive and non-intensive logging
there, and two focusing on terra firme, likewise treating both intensive and

non-intensive logging. We wish to stress at the outset that the material that
we are presenting is the result of a very preliminary analysis of only a
portion of our data. Hence, our final results and interpretations may differ
somewhat from those presented here.


Logging has traditionally been associated with riverine environments
in Amazonia (Rankin 1985) because humans have traditionally settled in
varzea areas and water provides an easy medium for log transport.

Case Study #1: Non-intensive Varzea Logging

We studied traditional, non-selective virzea logging in the
municipality of Anajas in the central part of the Island of Maraj6 (Fig. 2).

Social milieu. The residents of the Anajis region derive their income
from a combination of activities, including rubber tapping, harvest of forest
fruits, and logging. Fruit harvest and rubber tapping predominate in the dry
season while logging is the main economic activity in the wet season. The
land at Anajas is generally in large holdings with the extractors living as
tenants on these holdings. These extractors are obligated to sell their
extracted products to the land owner (or one of his agents) in exchange for
the right to live on the land. The land holder, in turn, provides goods and
services to extractors (i.e., "avamiento" system). The extractors receive low
prices for the raw materials that they produce while paying high prices for
the goods that the land holders provide. Hence, most are perpetually in debt
with little opportunity for upward mobility.3 As Brazilian workers become
better informed and organized, exploitive labor practices should diminish.

Extraction process. We studied traditional logging on three properties
near the village of Luciano, several hours by traditional river boat up the
Mok6es river from Anajis. We observed that felling is done by hand and
felled trees are divided into 4-5 m lengths. Logging roads are made by
placing sections of tree stem (2 m long and 10-20 cm in diameter) crosswise
along the logging trails (with a 1.5 m spacing between cross pieces). The
loggers, working in groups of 5-10, then push the logs down the trail as if the
log were a train sliding over the ties (poles).

The work is time consuming. Through interviews and first-hand
observations, we determined that an average of 7.1 hours are required to cut
and harvest each Virola log. One half of this time (3.6 hours) is spent in
manually pushing the log out of the forest to the edge of the water. An
additional 0.5 hours per log, on average, is spent in guiding the logs out to the
main trunk streams where they can be sold. The remaining time (3.0 hours)
is divided between building and poling of logging roads (31% of total logging
time) and felling and preparing the logs for transport (11%). In terms of
labor efficiency, 7.5 hours are required to harvest and transport each m3 of
Virola [given that the average volume of a log in the AnajAs region is 0.94
m 3(n=88)]. The selling price per m3 of wood is approximately US$ 4.00
(March, 1989). Hence, workers engaged in this activity were earning

approximately US$ 0.50/hour ($4.00 per m3/7.5 hours work = $0.53/hour of

Because of the large amount of time devoted to pushing the logs out of
the forest, there are limits on how far loggers can go into the forest in search
of Virola. For example, we determined that the total time to cut, harvest, and
transport one Virola log located 50 m within the forest is 4.1 hours (or 4.4
hours/m3). Earnings in this case are approximately $1.00/hour. But if the
log is 1000 m from the edge of a forest stream, the total time increases to 14.5
hours (or 15.5 hours/m3) with an hourly earning of only US$ 0.25. For this
reason, most extractors place the maximum distance for viable logging (given
current prices) at about 500 m (9 hours/m3 harvested; average nourly return,
US$ 0.42).

Both Virola processors and harvesters concur that the availability of
high-quality Virola has declined over the last two decades. Mousasticoshvily
(internal document), in interviews of 64 loggers on Maraj6 Island in 1989,
reports that 95% expressed difficulty in finding good quality Virola. And the
principal processing industry, EIDAI, has revealed that the mean size of logs
that it buys has declined in recent years.

One technique for providing access to high-quality Virola is to
construct artificial canals that extend deeper into the forest (i.e., to literally
bring the transport medium to the forest resource). Several companies and
individuals with ample capital are now financing canals. These canals are
generally dug by hand with only 7 m being opened per man-day of work.
Hence, the average time to build a canal 500 m into the forest might be some
555 hours or more than 3 months work. Individuals cannot afford to invest
this amount of time (with no return), but with big-mill financing, canals are
an option.5

Wood processing. There are no Virola processing mills in the Anajas
region--all logs are floated to regional mills in Breves and Belem (Fig. 2)
where they are transformed into boards, plywood or veneer. The relatively
inexpensive cost of riverine log transport means that mills can be at great
distances from the wood resource. For example, Trevo, a large sawmill
company at Macapi, buys wood from Tabatinga, more than 2000 km up the
Amazon. By contrast, on dry land, transport distances beyond 100 km are
usually regarded as non-economical.

Ecological impacts. The highly selective nature of varzea logging is
illustrated in Fig. 3. Only Virola surinamensis was harvested in these sites.
The average number of trees harvested per site was 13 (average wood volume
harvested = 27 m3). On an areal basis, we estimated that about 2 trees or 4.9 m3
per hectare were harvested.

The ecological impacts of this highly selective logging were small.
Canopy openings amounted to only 2% of total logged forest area, and logging
roads occupied only 4% of total site area (S = 0.43; n = 3). For every tree
harvested, 14 woody plants 10 cm dbh (diameter at breast height) were
severely damaged. Even so, this comes to a total damage of 32 trees per
hectare or, on average, 10.5% of all trees > 10 cm dbh in these stands. Most of

this damage (approximately 70%) occurs in the cutting of trees to open the
logging roads and to make the cross poles to slide the logs over. Of the trees
damaged, 66% were judged to have economic value.

Specific impacts on the species, Virola surinamensis, must also be
considered. This species is known to grow well in open disturbed areas and is
present in post-logging forest regeneration. However, it seems likely that
the very best Virola individuals are being selectively removed in logging
operations leaving behind less desirable germplasm. Unfortunately, there is
no systematic effort by Amazonian researchers or institutions to collect and
cultivate superior germplasm of Virola.

We determined that there was still a substantial amount of valuable
wood in these forests after logging. Based on the sampling of 10 plots, 20 X 50
m each, in each of the three study sites (total sample area = 1 ha/site), the
total volume of trees 40 cm dbh, considering only species that are usable in
sawmills, was 90 m3 (s.d. = 34; n = 3).

Wood asa catalyst of ecological change. In remote areas of the virzea
where selective Virola logging is being conducted within the social context of
the "aviamento system," it is hard to argue that wood is acting as a new type of
economic catalyst producing unfamiliar ecological changes. Rather, we
believe that wood simply represents one of a string of forest products that
have been commercially extracted over the past one hundred years (others
include: fruits, oils, fibers, resins, latexes, and recently, heart of palm).
Because Virola extraction is highly selective in this region, causing little
damage to other forest products, this logging is benign and within the
confines of the extractivist tradition.

Case Study #2: Intensive Varzea Logging

In areas with good access to regional markets, selective Virola
extraction, as described above, represents just the first stage of logging. Once
the valuable Virola trees have been removed, other species are gradually

Social milieu. Intensive virzea logging has a different social and
economic structure. To understand, consider first that land, in of itself, is
generally not accorded much value in the Amazon estuary. Indeed, the large
land owners of the estuary are frequently absent for long periods, only
returning to exert their economic control when some new source of wealth
appears. Hence, as the valuable Virola is logged out and rubber prices
plummet, there is not much room for exploitation by the landed class and
varzea inhabitants are given a freer hand in the management of their
economic affairs.

We studied intensive virzea logging on the Canaticu River which
drains into the Para River at Curralinho (Fig. 4). A striking feature of this
intensive logging is that it is being conducted by the region's residents, not
"outsiders." We compared the yield per hour of work in four important
income generating activities: farming (i.e. cassava meal production), rubber
tapping, palm heart extraction, and logging) and found that returns on

logging were several fold greater than for either farming or rubber tapping
but only half of those rendered from palm heart cutting.6

Extraction process. Forest extractors on the Canaticu River go after a
specific size class of tree--those between 15 and 45 cm in diameter. Those
trees beyond 60 cm in diameter are too large to be processed by the local mills.
Hence, some larger trees are left at the end of these logging operations. In
our interviews of 20 mills on the Canaticu River, we found that some 60
specimens were being sawn in August, 1989. Extraction activities are.
concentrated in the rainy season when water transport is easier.

Because the trees of interest are small, fewer people are needed to push
them out of the forest, and polled logging roads are less elaborate (indeed,
sometimes loggers just carry the logs out on their shoulders. Once the logs
are out to the main river channels, the extractor ties several logs together
and paddles the string to a nearby mill.

Wood processing. Small, family-run sawmills are common throughout
the Amazon estuary. For example, there are some 350 such mills in the
municipality of Breves, just to the east of Curalinho (Mousasticoshvily, 1989,
internal document); and in a survey of just the Canaticu River, we
encountered 44 such mills (Fig. 4).

While small sawmills have been present for decades in this region,
their numbers have increased dramatically in recent years (Fig 5). For
example, in interviews conducted in 51 small mills in the municipalities of
AfuA, Breves, and Curralinho, we found that 83% of all small mills (n = 42) had
been established in the 1980s. The remainder had been established in the
1970s (i.e., no mill dated prior to 1970). We found that all owners were from
the region with most having been involved in extractive activities and/or
farming in the past. By comparison, there was a total of 20 large sawmills
(i.e., those with a band saw and a monthly production capacity > 250 m3) in
these three municipalities. Of the nine that we interviewed, most (78%) had
been established in the 1970s (Fig. 5).

The capacity of the small mills was low, producing, on average, only 17
m3 of sawn boards per month and the quality of the wood produced was poor.
However, because the boards produced are of small dimensions (standard
board: 17 cm wide X 3.8 m long and 19 cm thick) and of variable quality,
wastage is low: only 48% (s.d. = 9.3; n = 34) of each sawn log is lost as scraps
and sawdust.

The Canaticu mills are simple and inexpensive to mount. The total cost
to establish a mill in 1989 was approximately US$ 2,100.00 including the
materials to build an open-sided shelter for the mill ($170.00); the motor
($850.00 for a refurbished motor); the pulleys, axels, belts, blades and
hardware associated with the saw ($990.00); and the labor to actually build the
shelter and mount the mill ($110.00).7 Occasionally, the mill owners reported
that they had mounted their mills slowly with savings garnered over several
years, but the more common pattern (65% of all mills interviews) was for
older mill owners to provide financing to those wanting to establish a mill in
exchange for a set amount of production. Generally, 1.5-2 years was
sufficient time to pay off sawmill inplantation costs.

That these mills continue to be built is an indication that small-scale
mills can be profitable. Yet, the profit margins are not great. For example,
profits were approximately US$105.00 per month in August, 1989, when we
conducted our survey. Many mills function for only part of the year when
logs are cheapest and most available. Moreover, it is not uncommon for mills
to fail (24% of all mill sites on the Canaticu River were abandoned; most of
these defunct mills had functioned for less than a year before being

Ecological impacts. Because these mills specialize in sawing small
diameter logs, trees are frequently cut when they are in their most
productive growth phase (i.e., in general, trees accumulate biomass slowly
during their early life; but when they reach the canopy (at' about 25 cm dbh),
annual increments increase until later in life when, once again, annual
growth increments decline). Eventually, as more and more wood is extracted
from the forest, its structure is inverted from the normal situation of closed
canopy punctuated by occasional gaps to large openings with a few very
large trees sticking up at widely spaced intervals. We estimated mean tree
volume for trees > 40 cm dbh at only 25 m3 in three heavily logged stands in
the Curralinho region as compared to 90 m3 (as reported above) after Virola
logging at Anajis, and many stands that we saw literally had no useful wood

The many canopy gaps created by this more intensive logging create
conditions that favor the growth of light-loving species. Palms and vines, in
particular, seem to be favored by heavy logging in the varzea. Hence, in
intensively logged stands, it is likely that the relative abundance of species
will shift: widely dispersed pioneer-type species will be at a relative
advantage compared to slow-growing, poorly dispersed species.

Wood as a catalyst to ecological change. In the case of this more
intensive virzea logging, the demand for inexpensive sawwood products is
clearly acting as a catalyst to ecological change. Sawmills occupy the
margins of most streams in the Amazon estuary and most of these mills appear
to have been established in the last decade. A large fraction of the sawn wood
goes to Belem where it is bought by middlemen who then shunt it to the
construction trade, either in Para or in the Northeast.

Because virtually all useful wood between 15 and 60 cm dbh is
extracted, forests are left in a low, open state--a situation which may .lead to
new stands with grossly altered physiognomies and species compositions.

The extremely impoverished life of riverine inhabitants is provoking
this move to non-sustainable forest exploitation by the region's residents.
Because present-day prices of traditional extractive products (e.g., rubber,
virola and andiroba oil, forest fruits) are extremely low, the region's
inhabitants face a difficult choice--cut their forest or migrate to urban
centers. Many are opting for migration: Breves, Macapi and Bel6m are filled
with former vArzea inhabitants. Those that stay don't see logging as an
opportunity to get ahead, so much as a way to stay where they are for a bit


In upland forest, logging also occurs in both intensive and non-
intensive forms. Intensive logging is generally conducted by well capitalized
firms and occurs in areas that have been settled for several decades (old
frontiers) and that have a good infrastructure. By contrast, non-intensive
terra fire logging is concentrated in remote frontier areas and is usually
conducted by a host of actors each with only small amounts of capital.8

Case Study #3: Terra Firme Non-intensive Logging

We studied non-intensive terra firme logging along Para Highway 150
in the environs of Tailandia (North Central Pari) (Fig. 2).

Social milieu. PA-150, connecting Belem and Marabd, was first opened
in the late 1970s. In the vicinity of Tailindia, a 6 km wide strip of land on
both sides of this road was set aside for colonist settlement. Terra firme
settlement projects, such as this one, have been generally unsuccessful in
Amazonia. The reasons for this are in part cultural. Terra firme colonists are
recent immigrants from other regions (at Tailindia, 62% of the colonists are
from the Northeast) and generally lack an understanding of Amazon ecology
and potential forest uses. Hence, these colonists do not have the diverse
economic options that are available to varzea residents, who are old-timers
with a sophisticated knowledge of their ecosystem. If these terra fire
newcomers cannot earn a living selling the traditional cassava flour and rice
crops from their slash and burn farms, they must move on in hopes of other
opportunities. And, indeed, in inventories of 349 properties in November,
1988, we found that 68% of the original colonists were no longer present. Of
the present residents that we interviewed, we found that 57% (n = 34 families)
were not able to make enough on farming to satisfy their subsistence needs.
For example, families clearing 3 hectares per year earned an average of only
US$1200.00 from the sale of farm surpluses, a sum that provided only 75% of
their subsistence needs.

The first sawmills were established in Tailandia in the late 1970s, and
with the asphalting of Para Highway 150 in 1985, the number of mills has
risen sharply (e.g., 70% of the 48 sawmills that we tallied in the environs of
Tailandia were established between 1986 and 1989). The wood industry
appears, in the short run, to provide new opportunities for these colonist
families by providing an additional income source, either through revenues
from selling trees or through employment in the mills.

In interviews with 59 families along side roads, we found that 86% were
involved, either actively or passively, in logging activities. Those passively
involved (61%) simply sold trees in their forest tracts from time to time (price
per tree in January, 1989 was US$5.00). Because the logging is highly
selective, there are generally only a few desirable trees per hectare. This
combined with the fact that colonist holdings are only 50 hectares in size
means that "passive colonists" do not stand to make much money selling trees.

At present, tree sales are frequently made in times of crisis to cover the cost
of medicines or food. By contrast, the so-called "active colonists" (25%)
actually participate in the extraction process in collaboration with other
middle men. These colonists divide their time between logging (dry season)
and agriculture (wet season). For these families, logging represents a chance
for upward mobility. Some families have acquired chainsaws and/or old
trucks.9 Most were using additional cash to buy forested lots from
unsuccessful colonists to assure a supply of wood in the future.

Overall, the new market for Amazon wood is affecting colonist families
at Tailindia to varying degrees by providing them with: 1) cash in times of
emergency; 2) a chance to get involved in a new economic activity on their
land; and 3) job opportunities in the mills for family members not needed on
the farm.

Extraction process. Wood transport is a critical limitation to terra firme
wood extraction. While a hard-surface all-weather road is essential to
guarantee the smooth flow of sawn wood to distant markets, loggers still must
gain access to the trees themselves. Not surprisingly, loggers have become
the main catalysts for road (and bridge) building in the Taillndia region. Of
272 km of side-roads that we surveyed, two-thirds had been built by loggers,
frequently in exchange for partial logging rights on the lands of ranchers
and colonists.

Extraction is semi-mechanized in these frontier settings. Logging
teams usually consist of 3-4 people. Chainsaws are used to cut trees and then
narrow roads are cut by hand (following the path of least resistance) to the
base of the cut trees. Later a truck comes and the cut logs are hand-winched
onto the truck bed.

We accompanied four logging teams, each for a period of eight days,
and determined that the total time necessary to make all the preparations for
a tree, with an average "Francon" volume of 8.5 m3, to be hauled out of the
forest averaged 13 man hours (n = 14), or 1.5 man hours/m3. The time
required to actually fell the tree and divide the bole into manageable logs is
small (6% of total time); 38% of total time was devoted to the building of roads
within the logged area to provide logging trucks access to the felled trees.
Because equipment is old, spare parts are absent, and work crews are poorly
coordinated in these frontier areas, a surprising 56% of the total on site time
is spent in non-production related activities.

Given these time expenditures, a logging team of three working 8
hours/day produces, on average, some 15.8 m3 (Francon volume) of wood
which has a value of US$9.00 m3 when sold in the forest to wood transporters.
Approximately 45% of the total production value of the wood is spent in
extraction costs (e.g., purchase of logging rights, fuel, chainsaw
maintenance, etc.). Therefore, hourly renumeration is $3.24 [((15.7 m3 X
$9.00)/(24 hours)) X .55 = $3.24/hr]. These terra firme production values
compare well to varzea selective logging: work efficiency is four times
greater (1.5 hours/m3 vs. 7.5 hours/n3) and earnings per hour of work are
some six times higher ($3.00 vs. $0.50).

Wo processing. The market for Amazonian wood, in addition to being
a boon for some enterprising colonists, is also providing opportunities for
ambitious, hard-working lumbermen. For example, many mill owners that we
talked with were new to the sawing trade although they had been involved in
the forest phase of timber extraction in the past. Because a rich network of
middle men exists in these frontier zones (Fig. 6), these mills can depend on
others to harvest the wood and transport it. Hence, many specialized actors,
each with only a little capital, are able to apply their resources in
complementary ways allowing the complex task of lumbering to be realized.

The unsophisticated nature of frontier mills is reflected in the quality
of wood produced and in the low efficiency of bole-wood use. We found that
the final sawn product in most mills was generally only of modest quality,
most appropriate for the domestic market. Moreover, only about 30-40% of
the total wood volume processed actually results in sawn wood (i.e., for every
three logs processed, a volume of sawn wood equal to one log is generally
produced).10 This occurs because boards of only a few sizes are sawn and
because of lack of training and experience.

An index of the low capital levels of frontier mills is their general
inability to purchase sufficient log stocks to operate during the rainy season.
Logs are available for purchase only during the six-month dry season.
Sawmills must stockpile wood during these months if they intend to saw
throughout the rainy season. Because the purchase of logs represents 50% or
more of total operating costs of these mills, the capital outlay to secure ample
stocks is not trivial. At Tailindia nearly half of the 48 mills that we
interviewed are compelled to close during part of the wet season for lack of
stocks (capital).

Ecological impacts. We studied the ecological impacts of logging on
three sites in the environs of Tailandia and found that the primary impacts of
this activity are minor although significant secondary impacts do exist.

An average of 2.0 trees per hectare were harvested in the three study
areas. The average volume harvested per hectare was 16 m3. Logging roads
occupied, on average, some 6% of the study areas with, on average, 57 m of
logging road constructed for each tree harvested. Canopy openings
attributed directly to logging were only 8% of total logged forest area.

The actual number of trees 10 cm dbh damaged was 582 per hectare or
29 per each tree harvested. Expressed in terms of volume, 1.2 m3 of bole were
damaged for each m3 of wood harvested. Most damage (46%) resulted from the
cutting of trees to open logging access roads within the forest. While the
number of trees damaged is small when considering what is still left in the
forest (i.e., only 11% of the total trees 10 cm dbh were damaged during
logging), most of this damage (55%) is concentrated in the open gaps created
in the extraction process. If damage were minimized, these gaps would
provide ideal conditions for tree growth.

Based on inventories of all trees 40 cm diameter in two hectares in each
of the three study sites, we estimate that an average of 127 m3 of wood per
hectare (s.d. = 37; n = 3) is present in post-logged stands in this frontier
region. Dividing this wood into quality groups and applying current prices,

the value of this wood in log form (prior to sawing) is approximately
US$2,600.00 per hectare.

Wood a a catalyst ecological change. While primary impacts of low-
intensity logging at Tailindia appear to be small, the presence of a logging
economy in this frontier region contributes to deforestation. Indeed, without
the security that logging provides, it is likely that Tailindia would have failed
as a colonization center (just as the Transamazon settlements failed in the
early 70s) (Moran 1989). However, with a logging economy in place, colonists
are able to persist for longer periods while continuing to cut forest each year
to produce cassava meal and rice for both home consumption and sale. The
low price of these basic food stuffs (US$0.12 per kg cassava flour; US$0.04 per
kg rice) combined with the low productivity of cleared forest plots means that
large plots (frequently > 5 ha) must be cleared each year to help meet basic
needs. In the process, the forest, with some 125 m3/ha of usable wood, is lost.

In terra firme frontier areas, then, wood can act as a powerful "pull
factor" by attracting: 1) lumbermen who have small amounts of capital and
who develop a rich network complementary activities (Fig. 6); and 2)
desperately poor peasants who come from other regions in hopes of a better
life. The extraction process, while careless, does not threaten the overall
integrity of the forest. It is the secondary impacts of spontaneous
colonization and total deforestation associated with slash and burn
agriculture that could compromise the ecology of the region for many years
to come.

Case Study #4: Intensive Terra Firme Logging

Two hundred kilometers to the east of ParA Highway 150 and twenty
years older, runs the Bel6m-Brasilia Highway--the oldest frontier in the
Brazilian Amazon (Fig. 2). Selective logging, such as that now occurring at
Tailindia, began in the early seventies along this Highway and during the
1980s logging became increasingly intensive. Now well more than 100 tree
species are being harvested. We have focused our study of intensive logging
in the municipality of Paragominas (27,000 km2).

Social milieu. The municipality of Paragominas was almost entirely
forested in 1960 but by the end of that decade thousands of square kilometers
of forest had been cut to establish cattle ranches. Satellite images,
interpreted at the Amazon Center for Remote Sensing in Bel6m, reveal that, in
the 25-year period from 1960 to 1985, 24 percent of this municipality was
converted to pasture.

Unfortunately, the planted pasture grasses began to lose vigor within
three to four years. Ranchers were then left with the option of abandoning
their old pastures or reforming their land (i.e., fertilizing and planting better
adapted forage varieties). Ranchers with little capital sold off their lands;
those with better financial means set about the task of reforming their
pastures. Although these "reformed pastures" had a longer life than the first
generation pastures (i.e., 6-8 years vs. 3-4 years) they also degraded
gradually, and by the early 1980s, ranching in this region faced severe
economic and ecological constraints as government-sponsored financing
schemes for cattle raising became more difficult to arrange and maintain.

When it looked as if the viability of cattle raising in Amazonia would be
put to a true test, forest timber appeared as a new form of subsidy. In the
1980s, ranchers have been able to sell the logging rights to their forest
reserves to loggers. The revenues so derived are sometimes used to reform
degraded pastures or for other ventures. The price for logging rights vary
between US$20-60 per hectare, depending on distance from the mill and the
quality of the forest. The cost to reform 1 ha of degraded pasture is
approximately US$150.00-300.00. Hence, the long-term economic prospects of
using the forest to finance further pasture reformation are not very
encouraging, particularly if pastures need to be reformed every 5-10 years.1 1

Initially, ranchers and loggers paid little attention to each other, or, at
best, blamed each other for deforestation. Recently, however, the distinction
between "loggers" and "ranchers" at Paragominas has become blurred. This
is because ranchers are becoming increasingly involved in the extraction
process (e.g., purchasing logging trucks and bulldozers for extraction).
Meanwhile, mill owners are beginning to invest in land in the region, in part
to secure future wood stocks, but also because land and cattle appear to be
secure investments. In a political context, as well, the distinction between
these groups is diminishing. This is because they now realize that the biggest
threat to their continued domination of this landscape is not each other, but,
rather, outside forces concerned with rational land management. While these
important shifts in land control and alliances are occurring, one fact remains
unchanged at Paragominas: land holdings are still immense in this
municipality, with small producers relegated to marginalized conditions, a
characteristic that appears to be typical of old frontiers in Amazonia.

Extraction process. Extraction at Paragominas is frequently carried out
in teams consisting of four people: a bulldozer operator who makes roads and
drags cut boles out of the forest; an assistant who links the downed boles to
the bulldozer using a steel cable; and a chainsaw operator with an assistant,
who helps locate desirable trees and clears the rubble of branches and lianas
from the base of the trunk to allow cutting.

Machinery is frequently new, allowing the work to be done efficiently.
On average, 75-100 m3 of wood are produced per day or 3-4 m3 per man hour
of work. Extraction rates per man hour are four times greater than at

In an evaluation of the logging process on a 52 hectare stand (1630 m3
wood harvested), over a 21 day period, we estimated that profits to the logger
were approximately US$9,000.00 (Uhl and Vieira 1989). Hence, profits can be
substantial in this sector.

While Paragominas logging is more rapid and more profitable than the
more labor-intensive extraction at Tailindia, it is less efficient in energetic
terms. Some 15,000 kcal are expended for every m3 that is cut and prepared
for truck transport in mechanized logging operations at Paragominas,
whereas only 3000 kcal/m3 are expended in non-intensive operations in
Tailindia (based on time/labor studies). The involvement of bulldozers, that
consume 130-150 1 of diesel per day, in intensive logging is the main reason
for this difference.

Processing. In the town of Paragominas, alone, there were 83 sawmills
in January, 1990. Most of these mills were well capitalized--having
equipment and employees involved in forest extraction. Most also function
throughout the year and many are involved in secondary wood processing
(e.g., making flooring, paneling, molding, etc.). Sawn wood is generally of
high quality and many mills export a portion of their production.12

Ecological impacts. We evaluated ecological impacts in three
intensively logged areas but have results from only one area analyzed to date.
In this site, we found that 140 trees, distributed among approximately 45
species, were harvested in a 15.9 hectare area (Fig. 7). The number and
volume per hectare was 9.3 trees and 67 m3, respectively (compared to 1.7
trees and 15 m3/ha at Tailindia on a similar sized site).

Impacts of this intensive logging are much greater than under
selective conditions at Tailindia. For example, we found that 23% of the total
study area was covered with logging roads as compared to 5% under selective
logging at Tailindia (Table 5). Moreover, canopy openings increased
dramatically in the intensively logged site from 16 to 56% of the area;
whereas canopy alteration attributable directly to logging at Tailandia was
less than 10%. Not surprisingly, tree damage was much greater in this
intensive-extraction study site: 226 trees/ha 10m dbh were severely damaged
in the extraction process compared to 41 trees/ha in the non-intensive

As extraction intensity increases, damage appears to increase in linear
fashion. As evidence for this, note that the number of trees and volume of
wood harvested under intensive extraction was 4-5 times greater than in the
non-intensive study area (Table 2) and, correspondingly, the indices of
damage (e.g., number of trees damaged/ha, logging roads/ha, etc.) were 4-5
times greater.

This more complete harvest and associated damage meant that the
volume of residual wood (trees 40 cm dbh) with economic value was relatively
small (about 13 m3/ha), only 10% of the residual log volume estimated for
Tailindia. Because post-logged forests are very open (canopy cover, 50%), the
regeneration of pioneer species and vines is encouraged. Indeed, there is
concern that dense tangles of vines could dominate these stands far into the

Secondary impacts are also associated with intensive logging. Of
particular concern is fire. Forest cutting for any purpose leaves slash on the
ground, increasing potential fuel loads; and the opening of the forest canopy,
by increasing the amount of radiation reaching the forest floor and
increasing vapor pressure deficits, dries the slash. During the six-month dry
season, a period of five or six rainless days is sufficient to dry fuels below a
threshold of combustion in these logged forest stands whereas non-logged,
virgin forest will not burn even after very prolonged droughts (Uhl and
Kauffman, in press). As fire is commonly used for pasture establishment and
weed control, an ignition source is widespread during the latter part of the
dry season. In summary, these logged ecosystems represent an entirely new
and unique fire environment in Amazonia: fuel loads are greatly increased

from forest slash, the microclimate is decidedly drier, and anthropogenic
ignition sources are common.

Wood Is a catalyst Lo ecological change. At Paragominas, wood
completely dominates the regional economy. The municipality's 400 sawmills
create jobs for some 12,000 workers. And at least this many people are
employed in jobs directly associated with logging (i.e., forest extraction, wood
transport, secondary wood processing (e.g., furniture making), etc.).
Browder (1989), in a study of the economic impacts of the lumber industry in
Rolim de Moura, Rond6nia, estimated that 60% of all the lumber industry jobs
were outside the sawmills.

Wood has created a boom economy at Paragominas. Much of the wood
money flows to the mills themselves where profit margins appear to run at
about 25%. These profits are being invested in both the purchase of
equipment and virgin forest tracts (i.e., they are being applied to assure that
wood resources will be available in the future and to increase the rate at
which wood can be extracted in the present). Nobody at Paragominas
seriously invests in forest management. Hence, it is inevitable that the
region's forests will be completely harvested over the next decade or two and
this boom period will pass.

But Pard's wood resources, properly managed, could benefit the state
indefinitely. In Table 3, we contrast what could happen in Pard with
foresight and planning and what is happening. We argue that logging could
have favorable ecological, economic, social and political impacts, far
different from those presently characterizing this activity.

The main impediments to sustainable logging in the eastern Amazon
are not economic or ecological. They are social and political. With a properly
designed set of incentives and disincentives, together with a rigorous
protocol of vigilance, the wood industry could be effectively disciplined and
Para's inhabitants, both human and wild, would be immensely benefitted. The
Wood Project, by providing relevant information on the economic, social, and
ecological consequences of logging, hopes to be of assistance to those
involved in the formulation of land-use policy in the State of Para.


Figure 1. (A) Roundwood production in Amazonia compared to the rest of
Brazil. The "Amazon" includes: the states of Rondonia, Acre, Amazonas, and
Pard; the territories of Roraima and Amapi; and parts of Maranhdo, Goids, and
Mato Grosso. (B) A comparison of roundwood production among the Amazon
states of Pard, Rond6nia, Acre, and Amazonas.

Figure 2. The Wood Project's main study regions in the state of Pard provide
examples of both high intensity (Furos) and low intensity (Anajis) varzea
logging, and high intensity (Paragominas) and low intensity (Tailindia and
Tucumd) terra firme logging.

Figure 3. The skeleton of logging trails established in the selective extraction
of Virola on three properties in the remote central region of Maraj6 Island,

Figure 4. Distribution of family-run sawmills along the Canaticu River,
Maraj6 Island, Pard.

Figure 5. Pattern of establishment of small and large sawmills over the past
twenty years on Maraj6 Island, Pard.

Figure 6. The web of middlemen that act between the forest resource and the
sawmills in frontier logging zones of Pard.

Figure 7. The network of logging roads and the distribution of harvested
trees associated with both intensive (Paragominas) and non-intensive
(Tailfndia) logging in eastern Pari.

TABLE 2. A comparison of ecological impacts of non-intensive and intensive
logging in terra firme forest in eastern Pari, Brazil.*


Total area (hectares) 15.5 15.9

Total no. species harvested 7 45

No. trees harvested/ha 1.7 9.3

Volume (m3) harvested/ha 14.9 67.4

No. trees (>10 cm dbh) 41 226

No. trees (>10 cm dbh) 24 24
damaged/tree harvested

% surface area covered by 5.3 22.8
logging roads

length of road (m)/tree harvested 59 43

% canopy opening attributed to 5-10 40

*Data are from one site in each of the two logging intensity types.

TABLE 3. Undesirable (actual) and desirable (possible) outcomes of logging
on environment and society.


Actual situation: damage a significant portion of the "future forest;"
compact soil; create conditions favorable to vine growth and fire.

Desirable situation: remove old-growth trees providing light and
nutrients for suppressed trees; creating more habitat diversity and therefore
creating more diverse ecosystems.


Actual situation: poverty (low wages); insecurity of work; boom-town
syndrome--in the end, people and land used up.

Desirable situation: increased employment; opportunities for growth
of local added-value operations (e.g., furniture making).


Actual situation: increase class discrimination; high-risk work; high
pollution environment; surplus of cabarets and shortage of doctors and

Desirable situation: people establish permanently in region; future
importance of wood for region is recognized; logging industry invests in
community well-being.


Actual situation: loggers aligning with ranchers to form strong lobby
to gain complete domination over land and its richers with no responsibility
for sustainable uses.

Desirable situation: class of hard-working people using a resource that
could be sustainably harvested operating in a centralized and predictable


1. To estimate the volume of wood and its potential worth is to "play with
numbers" because the information base is so scanty, but order of magnitude
calculations are useful in defining the playing field. Our estimate assumes
that the, total forested area of Amazonia is 4 million square km with 3.5 million
square km of forest still available to be logged. We further assumed that these
3..5 million square km have, on average, 125 m3 of harvestable wood per
hectare with an average value of $15.00 per m3 before sawing (3,500,000 X 100
X 125 X 15 = $656,250,000,000).

2. Here we assume that 3 m3 of roundwood are required to produce 2 m3 of
sawn wood (typical of more efficient sawing operations) and that the value
of the sawn product is four times greater than the roundwood
($656,250,000,000 X 0.67 X 4 = $1,759,000,000,000). It should be noted in these
calculations that we are computing the value of a resource that is renewable.
Hence, the net present value of the forest would be many times greater than
these calculations indicate.

3. As one index of the difficult conditions of these workers, we found that
very few extractors could afford to buy rubber boots for their work (cost,
approximately US$12.00), even though they stressed, to a man, the need for
boots to avoid puncture wounds from spiny palms and reduce the danger of
snake bite.

4. This is an overestimate of worker earnings because it does not consider
travel time to the work area (frequently in excess of two hours) nor does it
consider cost of equipment (i.e., axe and machete).

5. We were told that water buffalo are used to haul logs out of the varzea
forest in the region between Anajas and Afua, but we have no idea how
successful or widespread this is.

6. Because of the lucrativeness of palm heart extraction, many areas of the
estuary have been intensively exploited and no longer have harvestable palm
heart available. As palm hearts become scarce (an inevitable occurrence),
the relative attractiveness of lumbering will further increase.

7. Frequently, mill operators use the motor from their river boats to run
their sawmills, thereby reducing the cost of mill establishment.

8. Not all selective logging operations in Para are run on a low capital base.
Indeed, when the wood resource is sparsely distributed but very valuable,
such as with Mahogany in the south of Pari, small logging outfits have
insufficient capital and political/social savvy to conduct the logging. Hence,
Mahogany logging is conducted by very large companies, each constructing
hundreds of kilometers of logging road each dry season to reach ever more
distant Mahogany trees. The present transport distance approaches 400 km,
but the value of sawn Mahogany, at US$600.00-800.00 / m3 is compensatory.
The "Logging Project" will study Mahogany logging in 1990.

9. Frequently, these items are acquired in exchange for wood. As a typical
example, one colonist that we interviewed acquired an old logging truck in
1988 in exchange for a promise to deliver 200 m3 (approximate value
US$3,600.00) of hardwood to the mill of the seller.

10. This scrap wood is frequently used to make charcoal which, at present, is
being sold to iron ore smelters at Maraba, 300 km to the south of Tailindia.

11. Some well-capitalized ranchers are, nonetheless, reforming their
pastures for a second time (i.e., implanting third generation pastures). Great
hope is being placed in the newly introduced forage, Brachiaria brizantha.
With continued field research and field experimentation, prospects for
sustainable cattle ranching in this region are improving.

12. At present, we are studying the wood processing sector at Paragominas
and can do no more than provide a qualitative assessment here.


Browder, J. 0. 1988. "The social costs of rain forest destruction: a
critique of the 'hamburger debate." Interciencia 13: 115-120.

Browder, J. 0. 1989. "Lumber production and economic development
in the Brazilian Amazon: regional trends and a case study."
Journal of World Forest Resource Management 4: 1-19.

Hecht, S. B. 1985. "Environment, development and politics: capital
accumulation and the livestock sector in eastern Amazonia."
World Development 13:668-684.

Mahar, D. J. 1989. Government policies and deforestation in Brazil's
Amazon region. World Bank publication.

Moran, E. F. 1989. "Government-directed settlement in the 1970s: an
assessment of Transamazon highway colonization." In D. A.
Schumann and W. L. Partridge, Eds., The Human Ecology of
Tropical Land Settlement in Latin America. Westview Press,
pp. 172-198.

Rankin, J. M. 1985. "Forestry in the Brazilian Amazon." In G. T. Prance
and T. E. Lovejoy, Eds., Amazonia. key environment series.
Oxford: Pergamon Press, pp. 369-392.

Uhl, C. and J. B. Kauffman. 1990. "Deforestation, fire susceptibility, and
potential tree responses to fire in the eastern Amazon." Ecology

Wood Project, in press. A atividade madeireira em expansao em uma
regiio de fronteira: o caso de Tailindia. PA 150. IDESP.





It is often said that the lowland tropics have only a limited potential for
sustained resource exploitation by humans, and this idea is important in
scientists' recommendations for planning development and conservation
there. The basis for this statement, however, is theoretical and does not derive
from adequate empirical data about the actual character of particular
resources and their response to long-term use. In fact, a variety of data show
that certain areas of the neotropical lowlands are suitable for sustained-yield
production of food and materials and were used for this purpose for hundreds
to thousands of years in the past. Since some lowland areas are not suitable for
intensive use and have undergone damage from recent overuse, the problem
for the human ecologist is to define the different types of areas and the
conditions and methods conducive to sustained-yield use. Looking critically at
non-empirical resource evaluations and testing them with evidence of the
actual history of land use in different areas of the neotropical lowlands is an
important first step in approaching the problem.


The general idea of the unsuitability of tropical lowland resources in
Latin America for sustained use is based primarily on interpretations of the
land use of contemporary native peoples and rural smallholders and on the
present vegetation in their territories. In the last 100-200 years, when
scientists have worked in the area, there have been large areas of tall forest,
used primarily for shifting cultivation and foraging by native peoples or
Mestizo peasants living in small, dispersed settlements. Because of the
compelling influence of this "ethnographic present," the tall forest is
considered to be the original state of the habitat, and the shifting cultivation
and foraging are considered the traditional and appropriate uses for it.

Safe, nondestructive human exploitation of the forest is said to be
limited to foraging and shifting cultivation (Meggers 1971; Sioli 1973) because
of the scarcity of nutrients in tropical soils and the importance of the forest
vegetation in circulating the nutrients (Herrera 1985). The organisms in the
forest ecosystem are considered to have developed efficient ways to use and
protect a scarce supply of nutrients over millions of years of evolution. The
lack of nutrients in the forest ecosystem is attributed to the senility of the
characteristic Oxisol and Ultisol soils (Soil Survey Staff 1975), thought to have
developed during millions of years of weathering of acid, resistant
precambrian rocks under humid tropical climates. The forest subsoils are
found to be acid, kaolinitic, aluminum-rich, and extremely poor in nutrients
and organic matter, and most of the available nutrients for plant growth are
held in a tight cycle between the top surface of the soil and the body of the
vegetation. Consequently, the presence of the forest is considered necessary
for sustained yield land use systems in the tropics, because it is supposed to be
the main supply of nutrients and their only protection from leaching by the
abundant rain.

Historical information, however, shows that the rain forest has not
always been there and that a number of much more intensive land uses have
been carried out in the lowland tropics of Latin America through history and
prehistory. Both in prehistoric times and during the colonial period, large
areas that are now "virgin" forest were not tall forest, in some cases because of
catastrophic natural disturbances of the vegetation and in others because of

extensive human interference. In earlier times large areas of the lowlands
were exploited by methods more land intensive than swidden cultivation and
small-scale foraging and often supported larger populations than are in
evidence today. For various reasons--demographic, cultural, social, political,
and economic--many of these other land use methods are not in evidence
today, although they flourished for hundreds and in some cases thousands of
years in the past. Some other types of intensive land use are still in existence
today but are seldom studied by human ecologists because they are considered
non-native and inappropriate for the environment. Plantations of cash crops,
intensive subsistence agriculture based on annual crops, large-scale cattle
ranching, and highly intensive fishing and plant collecting are some of the
most important of these. Contrary to the common assumptions about the
unproductivity of the tropical lowlands, intensive prehistoric subsistence
systems supported large groups of people for long periods of time, and in
historic times the commercial production systems have made enormous
fortunes for the owning classes. There is no evidence that lack of soil
nutrients constitutes a limitation on the use of intensive systems.

The key reason why such systems have been viable in the lowland
tropics is that the land on which they were used is not the stereotypical
leached, nutrient-poor tropical soils of terra firme rain forest, but rather it is
land with a wide variety of vegetation and soils developed in high-nutrient
sediments. Heretofore, the areal and ecological importance of high-nutrient
ecosystems in the American tropics has been underemphasized by scientists,
who often imply that the great majority of the area c. 98% was nutrient-
poor terra firme rain forest (Prance and Lovejny 1985; Meggers 1971, 1984,
1985; Meggers, Ayensu, and Duckworth 1973; Sioli 1984). Actually, current
knowledge indicates that more than 20% of the Neotropical region has much
more favorable land use capability, a proportion similar to that in many more
heavily occupied parts of the world. This proportion of productive land
represents an enormous area and would not be considered a situation of low
resource potential anywhere else in the world. To project the characteristics
of the poor-soil terra firme tropical forests onto the tropical region as a whole
is to misrepresent its overall resource potential.

To point out that intensive systems of land use were stable for long
periods of time in the American tropics is not to say that some of the systems
were not damaging to the environment. All land use causes some damage, and
some methods, such as large-scale cattle ranching, have definitely
impoverished local wildlife and vegetation in the 300 years they have been in
place in the American tropics. Others, such as the intensive cultivation of food
crops, carried on for more than 2000 years in some areas, transformed the
habitat biologically, eliminating tall forest over large areas and replacing it
with agricultural fields and low, secondary growth. And yet, after a few
hundred years of abandonment, the tall forest returned to such a convincing
state of luxuriance that ecologists and anthropologists have assumed that it was
a virgin forest. Future research may show that the long periods of
environmental disturbance changed the species mix in plant and animal
populations and affected soil conditions, and some anthropologists have
pointed out that tropical vegetation types previously assumed to be natural, are
actually the product of past human interference (Anderson 1983; Bal6e 1989).
However, it appears that subsequent abandonment of the land permitted tall
forest resurgence in relatively short periods of time, between 50 and 100
years. How the process of deforestation and regrowth affected animal

populations specifically is unknown. Although the tall, regrown forest
appears to be inhabited by abundant wildlife, we do not have any information
about what was there before. This question is one that zooarchaeological
studies can contribute to answering in the future. It seems likely that the
return of the forest and animals would not have been possible in the absence
of substantial forest refuges.

The new information about the history of tropical environments and
land use in the Americas are important for general tropical ecology because
they show that the "primeval" tropical forest is not very ancient, in many
cases only 5,000-10,000 years old, and that it evolved under conditions of
instability, rather than stability. Another important implication of the new
historic information is that the modern post-Pleistocene environments
actually formed under the influence of human exploitation. Previously, it has
been assumed that the selective forces creating tropical ecosystems were
primarily natural, and ecologists attempting to understand the processes of
ecosystem formation and maintenance have tried to study habitats that had
escaped human influence. But as Peter Feinsinger has pointed out, this
approach is no longer tenable. The new information about the history of the
environment and land use suggests that basic explanations for the formation
of particular tropical habitats need to be revised fundamentally.

The historical data about tropical environments and land use are also
important for development planning because they reveal that a larger
proportion of the land can support human populations than we thought and
that a wider range of alternative land use modes are available for use in
tropical lands in the future. If we do not use this historical information to
increase production in the land suited for intensive use, there will be no hope
of protecting the more fragile land and its human populations and organisms.

The new historical evidence illustrates some of the causal relationships
that have been responsible for change in patterns of land-use over time and
therefore provide potential strategies for changing land use in the future.
One interpretive consideration turns out to be of great causal importance in
these human ecological problems: the changing socio-political and
demographic context of land use. The evidence does not support the view
(Meggers 1954; Steward 1949; Sanders and Price 1968) that the tropical
environment has limited political development. Nor, however, does it support
the view that large, centralized projects are more efficient than small,
independent, local ones, and it suggests that the large ones are more harmful
environmentally and more problematical sociologically. The historical
analysis thus exposes the realities that lie behind traditional theoretical and
ideological questions about the relative value of plants and animals, people,
classes, and nations for the long term.

The findings about the history of tropical resource use in Latin America
may be surprising or uncomfortable to some. Especially for those fighting
against great odds to ensure the preservation of the tropical forest and the
rights of native peoples, it may be difficult to accept that not all the luxuriant
forest is ancient or virgin and that for long periods in the past it was used in
ways that they might not consider ecologically sound. For some national and
international agencies seeking to develop the tropical resources, it may be
difficult to accept that the centralized, large-scaled projects that they promote
may not be more ecologically, economically, or sociologically viable than the

aggregate actions of anonymous traditional rural small holders and
indigenous peoples.

The discordance between our hopes and expectations and the actual
history of the environment and its use only underscores the need to
continually test general theories about tropical human ecology with both
experimental and historical data. The workings of the world almost always
turn out to be more complex than our theoretical models, which are often
based on inadequate knowledge and intuitions limited by researchers' cultural
and national contexts and political ideologies. But it would be dangerous to
ignore the lessons of tropical habitats' past history. Planning for the future
survival of tropical habitats and their inhabitants will rely on decisions based
on a firm grasp of causality.


The Yucatan Peninsula

An important tropical American region that has undergone great
changes in vegetation and land use over the last few thousand years is the
Yucatan Peninsula, especially the Pet6n, the forest heartland of classic Mayan

The peninsula is presently covered with forest, savanna, and swamp
vegetation; mainly dry, deciduous forest and scrub savanna in the north and
semi-deciduous and swamp forest in the south. Here, the fact that 19th and
20th century Mayans were subsisting on slash-and-burn cultivation in the
forests led scholars to conclude wrongly that Mayans had always done so and
that the land was not apt for intensive, clear cultivation (Meggers 1954;
Sanders and Price 1968). The supposed poverty and fragility of the Mayan
environment were considered by anthropologists to have influenced the
evolution of Maya society by preventing the intensive agriculture to support
the development of urban population centers and centralized state
organization. In the absence of archaeological investigation on ancient
subsistence, the prehistoric subsistence system was modeled upon tropical
forest ecosystems. Archaeologists theorized that ancient Mayans relied on the
produce from polycultural swiddens and tree cropping (Bronson 1966; Marcus
1982; Puhleston 1968; Wiseman 1978), and Mayanists generally accepted the
theories (Willey 1982) although they actually were not proven by
archaeological evidence.

The palynological and archaeological records and the patterning of
prehistoric agricultural facilities and settlements, however, show that for at
least 1500 years, intensive cultivation of maize was the primary method of
agricultural food production in the Mayan lowlands, not agroforestry or slash
and burn cultivation of starchy crops. The data show that by about A.D. 200,
the Pet6n was heavily occupied by the Maya, with nearly continuous rural
habitation and numerous large ceremonial and population centers.
Subsistence intensification increased with time during this occupation,
reaching a peak reliance on maize cultivation during the latest period, A.D.
900-1550, according to the human bone chemistries (Table 1; White and
Schwarcz 1989). As evidence of the great intensity of food production,
landscape studies have uncovered tens of thousands of square kilometers of

agricultural earthworks and constructions, including raised, ditched, and
stone-walled fields, stone-lined terraces, irrigation canals, and causeways
(Harrison and Turner 1978).

During this period of occupation, large areas became deforested. From
about A.D. 300 until the conquest, when most of the agricultural constructions
were built, those areas that have been sampled have pollen profiles heavily
dominated by herbs, with only rare tree pollen (Brenner, Leyden, and Binford
1990; Wiseman 1978). The post-conquest segments of the profiles, which
coincide with widespread population decimation among the Indians, are
dominated by tree pollen, in contrast to those of the prehispanic period. The
prehistoric geography and paleontology show, then that forests that we had
assumed were virgin were in fact reduced greatly under intensive cultivation
by the populous native societies and grew back fully only after the disruptions
of conquest led to the abandonment of intensive cultivation over much of the
land, due to depopulation.

Scholars had thought that the Mayan lowlands were not fit for sustained
intensive use but the manifest evidence of long-term intensive use falsifies
that hypothesis. Books on the Mayans by anthropologists had stated that the
forest soils of the peninsula are poor, acid, and lacking in nutrients, due to the
rainy tropical climate (Coe 1966; Hammond 1982). But this evaluation of
Yucatan soils did not come from natural scientists, who early on had described
the soils as limey and rich in organic matter (e.g., Lundell 1937). Empirical
soils studies had never demonstrated that the soils were unsuitable for
intensive cultivation. Preliminary studies commissioned by anthropologists
showed that topsoils under cultivation were somewhat more acid and more
subject to weeds than those under forest (Cowgill 1961), but this is true of most
soils under cultivation in rainy regions worldwide and is not an indicator of
general soil acidity or infertility. As the geological and soil maps show, the
forest soils of the Yucatan peninsula are for the most part calcareous, not acid,
for the predominant soils substrate is a limestone shelf of marine origin
(Lopez Ramos 1983; West 1964). In the calcareous soil environment, organic
matter is preserved, and expanding lattice clays of high exchange capacity

In addition, far from being extremely rainy and subject to severe soil
leaching, the Mayan lowlands have relatively low average rainfall in
comparison with other parts of the tropics, ranging from less than 500 mm in
the north of Yucatan to c. 1300 mm in the central Pet6n. Figures of over 3000
mm, often quoted as characteristic of the central Mayan lowlands, are not from
the central lowlands at all but from the mountain slopes to the south. In the
lowland areas, seasonal and periodic drought is one of the most important
problems that farmers face. The forest soils of the Mayan heartland,
therefore, are not characterized by a high degree of chemical leaching or
acidity. The soils of the swamps and wet savanna bajos of the Yucatan tend to
be influenced by lime, like the non-flooded land, but their hydrology and
chemistry produces hydromorphic soils with pH that fluctuates from alkaline
to moderately acid depending on changing soil moisture. Despite the
secondary literature, no part of the Yucatan Peninsula is characterized by the
archetypal highly leached, highly acid, aluminum rich, nutrient-poor soils of
the humid tropics. The quality of the soil varies in the Yucatan Peninsula, but
the predominant types in the area belong to groups that are successfully used

for permanent subsistence agriculture and commercial agriculture elsewhere
in the world.

Interestingly, the conclusion that the land in the Yucatan Peninsula
was not apt for intensive use was made in the absence of definitive empirical
soil data to support the conclusion. The origin of the idea actually was cultural
evidence: the lifeways of the Mayan Indians. Anthropologists made the
pejorative ecological judgement about the land to a large degree because the
Indians appeared to be using it only for swidden cultivation, which was
thought to be an adaptation to poor soil. What the anthropologists and
geographers did not take into consideration was that the practices of the
Indians that they were observing were not necessarily representative of those
of pre-conquest times, due to substantial demographic, economic, and political
changes that had taken place. The anthropologist's treatment of Mayan Indian
communities as if they were isolated, indigenous societies obscured their
connection with their neighbors and their placement in history. During the
European conquest, the best of their land had been taken by Europeans for
cattle ranching and for commercial plantations of sisal and other industrial
crops. Indians had been relegated to poorer land or fled to refuges distant
from transport, and their labor and production were appropriated for use by
on the European-owned land (Farriss 1984). In addition, increased indigenous
mortality after the conquest led to rural depopulation, another incentive to
long-fallow cultivation, which is less labor intensive than permanent
cultivation and therefore a preferred method where population pressure is not
a factor. The relatively low population density of present-day Mayans is thus
partly due to the effect of introduced diseases and partly due to the takeover of
land by Europeans. It is not an adaptation to generally poor land.

Thus, the state of the environment and the intensity of use of the land
in the Yucatan Peninsula has changed greatly through time in response to the
density and organization of the human population and its social, economic, and
political context.

The Isthmian Region

Another tropical American region that has seen great changes in
environment and land use is the Isthmian region. The Pacific side of the
isthmus in Central America has a climax vegetation of dry and deciduous
forests. The pollen and archaeobotanical records (Bartlett and Barghoorn
1973; Bennett 1968; Colinvaux 1990; Ranere and Hansell 1978) show that the dry
forests were extensively burned by humans as early as c. 5000 years ago, and
that deforestation was widespread from about 2000 years ago, due to intensive
maize cultivation (Cooke 1972; 1975). Since the European conquest, the dry
forests have been under cultivation or grazing almost continuously. At
present they are used for a range of different types of cultivation and animal
husbandry. Their main ecological problem is lack of water, and the
consequent limitation on maintaining forage and ground-cover to prevent
erosion and on providing water to cattle and crops.

The moister rain forests in the central highlands and Atlantic side of
lower Central America were also deforested during prehistoric times. Due to
gaps in the palynological and archaeological records in these areas, we do not
know many details about the history of the forest. During the Pleistocene, the
area was deforested for long periods due to climatic conditions, and pollen

profiles of c. 11,000 contain evidence of clearing by fire (Colinvaux 1990),
possibly as a foraging strategy. By late prehistoric times, large areas were
under clear cultivation. Numerous independent written records from the first
fifty years of the European conquest period make it clear that when the
conquerors first reached the area, Indians were clear-cultivating large areas
of regions that are now covered with luxuriant "virgin" forests; the major
crop, as elsewhere, was maize (citations summarized in Roosevelt 1979). These
wetter areas were reforested when the Indian populations contracted, due to
disease and military defeat during the European conquest. European colonists
moved mainly into the areas of dry and montane forests, more apt for their
preferred production systems, and the moister lowland areas were primarily
left as an Indian refuge, except where there were resources for mining. With
their lower post-conquest population densities and the need to be mobile to
avoid interference by intruders, the Indians primarily used the land for
foraging and shifting cultivation (e.g., Wafer 1903 cited in Roosevelt 1979).
With the improvements in Indian political representation, rural
transportation, and access to outside markets that took place during this
century, Indians have expanded their subsistence economies to include cash
crops such as coconuts and tourist crafts (Herrera 1989; Salvador 1976; Stout

Scholars assume that the land under moist forest in the Isthmian region
is not suitable for more intensive production, because the shifting cultivation
carried out by the native peoples is supposed to be an adaptation to extremely
poor tropical soils. However, because of the heterogeneous geology of the
area, which includes mafic igneous deposits, as well as ancient marine and
riverine sediments, there are few areas with such nutrient deficient soil that
they could not be put into intensive agricultural production for subsistence or
sale, as is the practice on similar soils in other tropical regions, such as
Colombia, El Salvador, Southeast Asia, Hawaii, and parts of Africa. However, the
present Indian manner of land use, which combines subsistence production in
swiddens with foraging, crafts for the lucrative tourist market, and some cash
cropping of starchy crops and tree products, is probably a better way to
conserve the biota of the region, not to speak of the Indians themselves, who
have unassailable ancestral rights to occupy the land and decide the nature of
land use in the area.


Examples of long-term patterns of change in environmental conditions
and land use in the tropics are found in the history of the interfluvial forests
of the Amazon.

Current Indigenous Occupation of the Interfluvial Forests

Much of the historical data for Amazonia comes from the forests of the
Upper Amazon in Peru and Ecuador. In the last hundred years or so, these
interfluves have primarily been inhabited by Indians, such as the Campa,
Amahuaca, Piro, Yagua, Yaminagua, Machiguenga, and the Jivaroan speakers,
who live dispersed in small family groups, subsisting mainly on slash-and-
bur cultivation and hunting and fishing. Some of these Indians, particularly
the Jivaroans, are engaged in constant feuding and raiding and are considered
fierce fighters who have been successful at protecting their territory from

intruders. Political integration is slight for all these Indians today, and the
dispersed extended family settlement is considered the main corporate social
unit (Farabee 1922; Fejos 1943; Carneiro 1970; Descola 1989; Harner 1973;
Johnson and Earle 1987).

Indigenous lifeways and cultural patterns are quite similar in other
interfluvial forest areas of the Amazon and Orinoco basins, such as the Guiana
and Brazilian shields, although the size and degree of dispersal of households
and length of residence in a place varies, For example, Yanomamo shifting
settlements in the Guiana Shield in Venezuela and Brazil sometimes include
several extended families (Chagnon 1968; Smole 1976), and the Kuikuru of the
Upper Xingu in Brazil are known to have lived in a large village for over 100
years (Carneiro 1957; 1961). Ge-speaking groups such as the Northern Kayapo
of Brazil are also reported to have lived for long periods in large villages with
several thousand people in multifamily houses (Posey 1987).

The practice of swidden cultivation of manioc or introduced starchy
crops is widespread in the Amazon interfluves, though in pockets some
peoples, such as the Amahuaca (Carneiro 1970) and the Xavante (Flowers 1983),
primarily cultivate maize as a staple in swidden fields. Scholars have usually
assumed that the land in the Amazon interfluves cannot support more
intensive land use than this (Gross 1975), and the peoples' culture is
considered to represent a type of long-term homeostatic adaptation to the
forest (Descola 1989). Some feel that the intensive warfare is a mechanism to
disperse population to protect subsistence resources (Ross 1978), and, some
have suggested that cultural ideological systems, such as food prohibitions,
also operate to disperse population for ecological reasons (Ross and Ross 1980).

Because of the assumption that the Amazon Indians' lifeway is a very
ancient adaptation to the tropical forest, the Indians have become important
foci of theories about the ecology of early human evolution and the lifeways
and social organization of early human development stages (Chagnon 1988;
Hill and Hurtado 1989). Scholars have held these Indians up as an example of
how cultures do not evolve to complexity in the poor environment of the
tropical forest (Johnson and Earle 1987), and, taking their cue from the
evolutionists, natural scientists assume that the land is not suitable for
intensive use, because it has been weathered under the rainy tropical climate
for many millions of years, and thus must be very low in nutrients for plant
growth (Prance and Lovejoy 1985; Sioli 1973, 1984).

However, the history of culture, human population, and environment in
the Amazon interfluves over the last 10,000 years does not support these

Soil Resources in the Interfluves

First of all, geological studies show that the surface sediments of large
areas of the Upper Amazon interfluves in Bolivia, Peru, and Brazil,
particularly what used to be called the "old alluvium" (Lathrap 1970), are not
Tertiary in date but actually consist of recent sediments descended from the
Andes less than 5000 years ago in the Middle Holocene (Campbell and Frailey
1984; Campbell et al 1985). What happened is that geomorphological and
hydrological changes that took place during the end of the Pleistocene caused
cataclysmic erosion on the Andean slopes and foothills and the widespread

deposition of deep layers of fresh sediments in the Upper Amazon. These large
areas of recent sediments from the Andes are not particularly leached or poor,
and the forests upon them are obviously not ancient virgin forests. In fact,
forests were destroyed over large areas of the Upper Amazon during these
geomorphological events and their remains, including whole trees, logs,
branches, and leaves, are to be found sealed in under the deep layers of recent
sediment. Radiocarbon dates from the stratigraphy in these areas of the Upper
Amazon show that the extreme cycles of erosion and deposition began between
30,000 and 20,000 and ended sometime in the middle Holocene, c. 5000 B.P. Such
great fluctuations in habitat would be expected to have changed the
distribution and composition of the animal populations of the area, and future
studies of the vertebrate faunas sealed in the various strata will give
information about the nature of faunal changes.

In addition, to the areas that received the blanket of recent sediments,
there are substantial areas in the Andean foothills of the Upper Amazon and
the Brazilian and Guiana shields with subsoil composed of limestones and basic
igneous rocks (Derby 1877; Petri and Fulfaro 1988; Putzer 1984; Irion 1984a and
b; Schubert et al. 1986). In addition to having this nutrient-rich geological
substrate, many of these areas have comparatively low rainfall for the tropics,
ranging from 1500-2500mm a year, and the vegetation varies from tropical
rain forest to savanna, with the majority of the area under semideciduous or
dry tropical forest. Thus, the soils not only have better nutrient supply than
had been thought, but they are also not subject to the advanced degree of
chemical weathering that takes place in older soils in rainier areas.

The residual soils formed on these sediments and rocks under these
conditions cannot accurately be termed poor, acid leached tropical forest soils.
Rather than falling mainly in the nutrient-poor, kaolinitic aluminum rich,
acid tropical forest soils classes, such as the Oxisols and Ultisols, which are
characteristic of much of the shields and Tertiary alluvium in central
Amazonia, these geologically heterogeneous interfluvial areas have soils of
the classes Alfisol, Mollisol, Vertisol, Inceptisol, and Entisol groups., which
lack aluminum toxicity and have montmorillonite or illite as predominant soil
clays, rather than kaolinite. In contrast to the classic Oxicols and Ultisols,
these soils have abundant sources of nutrients in the subsoil. Several scholars
have mentioned some of the cultural implications of the existence of high-
nutrient soils in archaeological studies in the Upper Amazon (Allen 1968;
Porras 1987), but the implications of the soils' empirical characteristics have
not been brought into our general interpretations of land use capability in the

Deficiencies of the soil, nevertheless, could not be the reason the the
extensive land use practices and shifting settlement of the Indians in the
Upper Amazon and is not a barrier to future development in the manner
usually assumed. What is known about the history of resource exploitation in
the area gives some indication of the potential of the area for other patterns of
settlement and land use.

The History of Resource Exploitation in the Interfluves

Although the Upper Amazon interfluves are presently covered with tall
forest cultivated mainly with shifting cultivation of starchy crops, the
paleoecological and paleodietary record offers conclusive evidence that this

was not always so, showing a pattern of deforestation under intensive maize
cultivation over the last two thousand years of prehistory. In the Ecuadorian
Amazon, herb and secondary vegetation pollen becomes common in profiles
from the last two thousand years before the present, along with finely divided
charcoal and extremely abundant maize pollen (Bush and Colinvaux 1988; Bush
et al. 1989). Independent evidence that maize, not manioc, was a staple food in
the Upper Amazon during the thousand years before conquest is found in
isotopic studies of prehistoric human skeletons from both riverine and upland
sites in the Ucayali basin of Peru. The prehistoric bone chemistries of people
of the period c. 500-1600 have the pattern of carbon and nitrogen isotope
ratios associated with staple maize consumption, not manioc (Roosevelt 1989:
55, Table II). There are as yet no reliable dates for the pollen profiles of
earlier Holocene times in the Upper Amazon, c. 8000-2000 B.P., but the profiles
show evidence of slash-and-burn cultivation (Bush and Colinvaux 1988; Bush
et al. 1989), and the bone chemistries of the time are less influenced by maize
consumption (Roosevelt 1989: Table II). The data of the earlier prehistoric
periods seem more comparable with present subsistence patterns than do the
intervening two thousand years of later prehistory.

This historical evidence, like the soil data, suggests that the modern
Indian's practice of shifting cultivation could not be merely an adaptation to
poor soils that do not permit more intensive land use, since there was local
deforestation and intensive cultivation production for periods of thousands of
years in prehistory. In addition, today's focus on manioc as a staple could not
be because the land does not permit the cultivation of other crops, because
maize, not manioc, clearly was the staple for the two thousand years before the
European conquest in the area.

In accord with this evidence, critical ethnographic studies show that
there is no empirical evidence that ecology is" the factor restricting present
Indian population density and settlement aggregation. There is, for example,
no correlation between variation in land fertility and the patterning of land
use or settlement in the Jivaro areas (Descola 1989). Long ago, Carneiro (1961)
pointed out that cultivation practices in the interfluves of the Brazilian shield
are carried out well below agricultural carrying capacity. Many writers have
noted a diminution of and change in species representation in the vicinity of
older, larger settlements, which suggests that if game were the limiting factor
in subsistence, the shifting settlement pattern would make sense. However,
this begs the question of why people do not rely on plant cultivation for
protein today, as they did in prehistoric times in the same area. Although
intensive maize cultivation may not be viable in areas of impoverished terra
firme Oxisols and Ultisols, there is no known ecological barrier to more
permanent agriculture in the extensive nutrient-rich soils of the Upper
Amazon, and, as mentioned earlier, such areas were used for staple maize
cultivation for over 2000 years before the conquest.

In line with this evidence for lack of correlation between soil fertility
and ethnographic settlement patterns, the archaeological evidence of sites in
the Amazon interfluves shows that Indians have not always lived in small or
dispersed settlements. Archaeologists have found the remains of substantial
villages in the Brazilian shield, where today Indian settlements are much less
substantial; and the archaeology shows a marked cultural and ecological
disjunction between the current and prehistoric Indians there (Wust 1987,

1989). The patterns of prehistoric settlement in the Peruvian, Bolivian, and
Colombian upland interfluves are poorly known, because archaeologists have
not yet systematically studied site layouts and regional settlement systems. But
the archaeological evidence for the Ecuadorian Amazon shows that, far from
being demographically dispersed and unevolved socio-politically, prehistoric
populations in the montana were regionally politically integrated and erected
truly monumental ceremonial earthwork complexes many square kilometers
area (Porras 1987) during the two thousand years before conquest (Athens
198p). If, in this area, the Jivaroan speakers do not carry out similar corporate
projects, it cannot be because the land does not permit it, since prehistoric
peopled did so on a large scale. Some other causes must be involved.

The European Conquest of the Amazon Interfluves

Historical processual factors have not traditionally been very important
in our explanations of the ecological adaptations and socio-political
organization of Amazonian Indians. However, a number of important recent
studies reveal that in many parts of the interfluves Indian settlement patterns,
population density, subsistence, and social organization have changed greatly
during the last several hundred years.

Indigenous oral histories in many regions record the fact that the
present inhabitants came in from the outside, either displacing the
autochthonous people or coming into a vacuum caused by depopulation or out-
migration of groups previously in the area. Ethnohistorians are uncovering
numerous accounts of the wholesale movements of peoples over large
distances and considerable reorganization and cultural change in response to
contact (Rodrigues 1988; Whitehead 1988, 1989; Porro 1981, 1983-84. 1989).
Ethnoarchaeologists are finding substantial evidence that some present-day
ethnographic groups have not been in their traditioriai 'territories for very
long and that their predecessors had very different cultures and lifeways (e.g.,
Wust 1987, 1989).

Rather than being ecologically determined by poor resources, the
present patterns of the Indian occupation seem an adaptation to their recent
demographic history and sociopolitical and economic context, which is very
different from that before the conquest. Europeans penetrated deeply into the
interfluves at different times in the first three hundred years of the conquest.
Modern ethnographic accounts for the interfluves often date first contact of
groups as occurring in the 20th century, but there are numerous historic
records of earlier contacts in those areas (e.g., Posey 1987; Rodrigues 1988;
Roth 1924; Schomburgk 1922-23; A. Leeds, personal communication), now
forgotten. Introduced diseases and European military conquest of the early
contact period caused steep declines in population in many areas, which may
have influenced some of the changes in subsistence patterns. There has been
no intrinsic lack of native population growth, since populations with access to
treatment for the foreign diseases have grown rapidly in some areas (Hern
1979). The introduction of new production systems to the Indians, such as
cattle-raising (Descola 1981, 1982) has altered both settlement patterns and
social organization. In addition, outsiders' interference with indigenous
political organization today prevents the development of native politics and
ritual centers. The police forces of the modern nations severely limit native
political integration by fiercely putting down the rise of native political,
military and ritual leaders and rebellions (Brown 1989; Whitten 1976), and the

dispersal of Indian settlement patterns may reflect the avoidance of
missionaries' strictures and the aggressive labor demands of landowners, who
are invading their territory in large numbers and appropriating land
supposedly not under use (ie., land used for swidden cultivation and foraging).
Thus, the distinctive indigenous lifeways of the interfluves today cannot be
explained purely in terms of ecology without consideration of indigenous
peoples' patterns of interaction with outsiders through time.

These changes in environment and land use through time mean that
land use evaluation and development planning based purely on present
indigenous patterns in the interfluves will not be an accurate and realistic
reflection of ecological potentials. Later on, I will discuss the range of
possible ecological impacts of various human occupations, but first, a look at
the environment and land use in the Amazon floodplains.


Scholars have been more equivocal in attributing severe ecological
limitations to human use of resources in the Amazon's extensive floodplains,
but nonetheless, the general point of view is that these, too, inhibited cultural
evolution due to ecological limitations (Meggers 1971, 1984, 1985; Sioli 1973,

Resources of the Amazon Floodplains

Early in Amazonian studies, both geographers and anthropologists
projected the stereotypical nutrient-poor, acid, aluminum rich soils of the
ancient terra firme rain forest upon the floodplains, and argues that they, too,
had limited agricultural potential (Denevan 1966; Meggers 1971). However, the
empirical evidence from the floodplains does not support this view.

The floodplains of Amazonia are very extensive really and include both
the present active channels of the rivers and their earlier beds. The recent
hydromorphic soil areas of the Amazon are the largest area of such soils in the
world. The sediments of these areas derive primarily from erosion of alkaline
rocks in the Andes mountains in the Upper Amazon watershed and from the
extensive areas of Carboniferous and Cretaceous limestones and diabase dikes
in the Lower Amazon (Irion 1984a; Derby 1877; Petri and Fulfaro 1988; Putzer
1984). The soils formed on these annually renewed sediments are rich in
nutrients for plant growth and the soil clays, primarily montmorillonite and
illite (Irion 1984b), are conducive to nutrient exchange with plant roots. The
distinct regime of all these rivers provides dry land for extensive plant growth
during the dry season and renewal of the fertility of the soils with new
sediment deposits during the wet season. Both the flora and fauna supported
on these soils have high levels of harvestable biomass, and the soils are not
characterized by severe nutrient deficiencies, although they are subject to
drought and flooding. The soils do not fall in the typical soil class groups of
poor terra firme land but rather in the many soil classes that support the
majority of intensive subsistence agriculture and cash cropping elsewhere in
the world (references summarized in Roosevelt 1980).

Prehistoric Occupation of the Amazon Floodplains

The resources of the Amazon floodplains did not generally limit long-
term occupation of the area by humans during prehistoric times, according to
the archaeological evidence. Although anthropologists had originally
assumed that indigenous people had been in Amazonia less then 1000 years
(Steward 1949; Meggers and Evans 1957; Roosevelt 1990c), they appear to have
been living in the area for more than 10,000 years, since the end of the
Pleistocene and possibly before (references summarized in Roosevelt 1989).
The subsistence of the earliest inhabitants varied through time and space.
Groups in higher, drier, cooler areas emphasized hunting of larger game and
those in lower, wetter areas focused on fishing and water mammal hunting,
plant collecting and small game hunting. Subsistence changed through time
toward more intensive use of resources. In some interfluves, early foraging
trended toward broad-spectrum exploitation of small game, fish, shellfish, and
plants. Intensive exploitation of aquatic resources along the major rivers
supported large and relatively stable settlements and pottery-making long
before agriculture, starting more than 7000 years ago (Roosevelt 1989;
Roosevelt et al. 1990). The middens of the specialized aquatic foragers of the
floodplains are so extensive that they have served as a commercial source of
lime for fertilizer for more than 200 years (Monteiro de Noronha 1862;
Ferreira Penna 1876; Smith 1879; Roosevelt et al. 1990). In time, crops such as
manioc appear to have become staple foods in the floodplains, and stable
villages had proliferated along the waterways by about 400 years ago. In some
areas, native grasses were put under cultivation as populations grew and in
the last 2000 years of prehistory, maize became the staple crop in many areas
of the floodplains.

The stratigraphy and archaeological site distributions in a number of
areas show that the region's resources allowed large aggregated Indian
settlements where now there are only sparse populations (Erickson 1980;
Boomert 1976, 1980; Nimuendaju 1949; Roosevelt 1987, 1980, 1990a, 1990b). The
Amazon main channel banks are covered with ancient garbage fill sites, and
in swampy areas towns were raised up on large earth platforms of several
hectares, and wide roads were built on causeways many kilometers long. The
large, deep many-hectare occupation sites, occupied for hundreds and in some
cases thousands of years, are now for the most part unoccupied except by
isolated peasant families. As many as 20-40 archaeological earth habitation
mounds cluster within a few miles, where only a few families live today. The
prehistoric sites are so numerous and large that their black soil garbage
dumps constitute a commercially important agricultural soil taxon and support
a widespread type of anthropogenic forest rich in fruit trees. The remains of
hearths and houses in the sites indicate permanent occupations of many
multifamily houses, indicating large resident populations. Such large
facilities and structures are not to be found among the sparse and more mobile
Indian peoples of today.

In addition to the changes in population and settlement pattern, the
prehistoric botanical and human osteological evidence from a number of areas
show that areas where the floodplain Indians and peasants are now living by
shifting cultivation of starchy crops and foraging, in antiquity people lived
by intensive cropping of seed crops and fishing. Thus, the Shipibo and
Cocama of the Upper Amazon, the Carib of the Middle Orinoco, and the

Mundurucu of the Tapajos basin eat manioc as their staple crop, but their
immediate prehistoric predecessors ate mainly maize (Heriarte 1964; Roosevelt
1980, 1989, 1990b; van der Merwe et al. 1981). In aid of cultivation, tens of
thousands of square kilometers of large earthworks were erected, for raised
and drained fields, canals, and causeways. On the evidence of zooarchaeology,
the prehistoric populations appear to have been using the fish resource more
intensively than it is being used today in the Lower Amazon, because the
average size of fish taken was much smaller in prehistoric times than today.

The complex cultures that achieved this intensification of subsistence,
population growth, and engineering feats appear by at least 2000 years ago in
the Upper Amazon, Lower Amazon, Middle Orinoco, and Guiana plains, and they
were long-lived cultures, with series of radiocarbon dates lasting more than
1000 years for several of them (references summarized in Brochado 1980;
Lathrap 1970; Roosevelt 1980, 1989, 1990a, b). These were stable societies
compared to the great Inca and Aztec empires of the temperate regions of
Latin America, which lasted a scant hundred years or so. The dense lowland
occupations no doubt greatly affected their habitats, but there was no
environmental deterioration sufficient to cause their premature demise, and
most continued to grow, develop, and expand until they were cut off by
European conquest.

The Floodplains since the European Conquest

Few Indians still live in the major floodplains of Amazonia, which were
depopulated by diseased and military defeats early on in the conquest, and the
land was rapidly taken over by a ruling elite of Europeans supported by an
underclass of mixed geographic and racial origin (Moreira Neto 1988;
Rodrigues 1988; Jones 1984; Vickers 1984; Stocks 1984). During the early
colonization, Indians were collected into mission centers and lost population
rapidly through mortality from epidemic disease and malnutrition. They also
lost their indigenous culture through the process of deculturation and
miscegenation imposed by the missionaries and colonial administrators.

During the period from about 1750 to the late nineteenth century, after
the expulsion of the religious orders, the Europeans deployed the local
Amazonian population in forced labor for extensive construction and
transportation works, extractive industries, cattle ranching, plantation
agriculture, and domestic service. Much of the production of the floodplain
went to support the owning class and for export abroad. The abolition of
slavery ended the plantation era by eliminating the main legal form of forced
slave labor, but other industries continued, often conducted by debt peonage
or a kind of feudal tenantry. The collection of rubber was an exceedingly
important export industry until its collapse after the establishment of
competitive industries in southeast Asia. Cattle-raising has long been a
primary industry in the floodplain savannas of the Bolivian Llanos de Mojos,
the Orinoco Llanos of Colombia and Venezuela, and the Marajo plains in Brazil,
producing meat and hides mainly for local urban markets. On Marajo, coffee
and food production for the Belem market is also carried out (Anonymous
1906). Along the Guiana plains, intensive wet rice cultivation has been a
major productive industry. Intensive, seasonal fishing is a widespread
regional industry that mainly feeds local populations and urban centers. Only
the extractive industries and timbering still seem to feed into the world market
today. In these areas of the Amazon, like the interfluves, the main staple food

today is manioc. Supplemented by fish among the poor and by beef among the

In the floodplains, Indians remain today mainly in those areas that
were not permanently settled by colonists during the early historic era. In
Peru, Ecuador, and Colombia, settlers entered during boom times, such as
during the early exploration for gold and the rubber era, and then for the
most part left during the busts. In floodplain areas such as the northwest
Amazon, the poverty of soils between the rivers seems to have limited the
permanent intrusion of Europeans for ranching and plantation agriculture
(Koch-Gruneberg 1967). In contrast to the Brazilian Amazon mainstream,
Indians such as the riverine Panoans in the Ucayali in the Peruvian Amazon
and the Tucanoans and Arawakans in the northwest Amazon still live in
indigenous communities subsisting on native crops and fishing. Manioc
and/or plantains tend to supply most of the calories in their diet, though in
the Shipibo area, where the soils are better than in the northwest Amazon,
native people cash crop the maize that their prehistoric predecessors subsisted
on, as well as introduced seed crops such as rice. The floodplain Indian groups
do not live much differently than the interfluvial Indians, despite the
significant differences in their habitat.

Essentially, the historical information about native lifeways shows that
recent native land uses cannot be assumed to have been the only ones possible.
In fact, the majority of indigenous uses are no longer in evidence today. One
cannot, therefore, infer that their present modes of land use are the only ones
viable ecologically.


One of the most important influences on the character of land use in
tropical Latin America has been political and social organization.

Effects of Centralized and Stratified Organization on Land Use

Land use patterns on societies with centralized and stratified political
organization tend to be more intensive than those under unstratified,
segmentary systems. Some theorists have claimed that large administrations
organized by centralization, specialization, and hierarchy are necessary for
the running of complex, intensive production systems to support large
regional populations (Johnson 1978; Johnson and Earle 1987; Sanders and Price
1968). Such functionalist arguments have been attractive to scholars partly
because of the importance of concepts of adaptation in studies of human
evolution and partly because of the scholars' context in modern stratified
societies. The reasoning behind such theories is that what exists must be
adaptive, necessary, and functional. However, many studies of the evolution of
intensive land use systems have shown that they often are developed and
maintained for long periods by uncentralized, unstratified organizations
(Higham 1989; Lees 1973; Enge and Whiteford 1989). Such systems are
vulnerable to takeover by centralized and hierarchical political organization,
which often promote such systems to support their projects. But such
economic systems do not need those kinds of organizations to run them, and
those organizations do not necessarily, in the long run, operate them for the
benefit of the populace (World Bank 1973).

Stratified and centralized socio-political organization has several effects
on land use. It tends to create population pressure because it skims off much
of the productivity of the land for the benefit of a few affluent and powerful
people. Thus the available resources cannot support as many people as they
could if land-use were not stratified. Socioeconomic stratification also often
narrows the economy by focusing on systems appropriate for cash cropping
and commercial production and eliminating subsistence production. In
stratified socioeconomic systems large areas of land are owned by outsiders
unaware of or uninterested in the welfare of the local community, and local
people who were supporting themselves on the land are dispossessed of land
and become indigent. Although production on large landholdings can be
costly and inefficient, the large owners usually have more political pull and
can get government subsidies and tax benefits that small holders or
subsistence farmers cannot. Centralized and stratified political systems and
the urban centers that they foster are historically an important cause of
chronic disease, nutritional stress, and high mortality rates among human
populations (Pelto and Pelto 1983; Roosevelt 1984). The amelioration of health
and nutrition in industrial nations during the last century is merely a
reversal of the deleterious effects of the rise of preindustrial and colonial
civilization and the improvement has occurred at the expense of health status
in third world nations.

Large-scale stratified economic systems also tend to be more harmful
ecologically than smaller scale, segmentary ones. Their larger scale of
operations and greater intensity of land use give them greater ability to
impact the environment. National and international projects can achieve a
greater scale of operations because their connection to the resources,
populations, and technologies of the world system gives them ways to
circumvent the impact of local scarcities and ecological problems on their
businesses. With the outside economic links, local degradation does not
necessarily hurt operations, which can be supported by production in other
regions. If a project eventually becomes unprofitable because of deteriorating
local conditions, it can be moved to another locale. In small, isolated
communities, people's land use can cause ecological damage, but its extent is
limited by the fact that there is no outside means of support or resupply to
maintain a high level of destructive resource use for long periods.

The large, specialized production systems also tend to be much less
stable than smaller, less specialized systems, partly because of their adverse
ecological impacts, partly because they are so narrowly based, and in part
because they are subject to uncontrollable changes in world markets or
political systems. Because the larger systems are run by outsiders for purposes
related to their activities in the larger world system, they are less sensitive to
adverse local changes caused by their operations and therefore are less able to
adjust to deal with them effectively.

Thus, in our consideration of the relation of political organization and
land tenure to natural resource preservation, we should not assume that large,
centralized and hierarchical organizations are the most appropriate or
effective ones in the long run. In fact, it appears that such systems may be
the ones most likely to implement destructive land use and to deprive local
populations of the ability to support themselves and produce goods useful to
the society.

Impact of the Colonial Expansion in Tropical Latin America

One of the major influences on current demographic patterns and
resource use in the Neotropics has been the European conquest. From a global
point of view, no other event since the Ice Age had such a rapid and enormous
impact on the use and character of the environment and the distribution of
peoples and cultures in what is now Latin America.

In the first two hundred years of conquest, military defeat of Indians
and the diseases introduced from Europe and Africa depopulated large areas,
which then returned to forest. The conquest also led Indians to become more
mobile so as to be less accessible to the missionaries and landowners in search
of converts and laborers. The early post-conquest depopulations and
dislocations had other repercussions, leading many groups to abandon the
intensive maize cultivation late prehistoric times and go back to long-fallow
cultivation of starchy root crops, an earlier prehistoric pattern. Some refugee
groups even gave up agriculture entirely for a time and lived by foraging in
the forests. Between about 1750-1900, Europeans settled the ecologically and
economically desirable areas in large numbers and put the remaining Indian
populations to work or brought in African laborers. Independent native
populations persisted only in distant hinterland areas.

The long term effect of the colonial expansion can be viewed globally as
the formation of international stratified societies in which the industrial
nations are the elite and the third world nations, the underclass. The
industrial nations expanded abroad in search of markets for their products and
sources of cheap labor, food, and raw materials. The imposition of colonial
power greatly expanded the geographic reach of states and increased
economic stratifir'tion further. Either directly or indirectly, the conquerors
reorganized economies and land-tenure systems to create abundant cheap
labor to produce for their support and for export. In that process, rural
populations commonly lost control over the land they inhabited and were
forced into the laboring class.

These changes under colonial expansion caused a decline in the
standard of living for many people in the conquered areas, and many third
world peoples for the first time experienced on a large scale the miseries of
civilization. The evidence for this impact comes from physical
anthropological research, which shows decreasing stature and increasing
pathology rates and severity in the skeletons of the populations that came
under the rule of colonial states (Cohen and Armelagos 1984). At the same
time, the general health and nutrition of people in the industrializing nations
improved markedly at that time due to the economic opportunities provided by
their colonial expansions, especially in the improvement of agricultural
production with the introduction of new crops (McKeown 1976).

Tropical American Land Use Under Social Stratification Today

In the study of the history of tropical land use in the Americas, it is
important to recognize the specific impact of stratified social stratification on
land use in particular areas of the tropical lowlands.

In many areas of the floodplains, such as the Middle Orinoco, Maraj6
Island, and the Llanos of the Bolivian Amazon, social stratification is the
primary reason why land is being used for cattle and not for subsistence
cultivation. The land is held by large land-owners for whom ranching is the
most lucrative method of use possible, given their goal of low-capital
commercial production for cash sale in urban markets. Intensive subsistence
production, which was the primary type of land use before the conquest in
such areas, is not desirable for the land-owners, because that would require
much more labor and capital investment and would be much more difficult to
control. The fact that the produce of the land is scooped off to make money for
the rancher means that the land cannot support a large resident rural
population. Thus, the political system determines the land use system, which
determines the population density.

Like ranching, plantation agriculture and forest extraction for export
have been production systems tied to socioeconomic stratification. In the
Yucatan Peninsula in the 19th century, hispanic landlords and foreign
companies took over large areas of previously Mayan land and put native
people to work as slaves or serfs on the plantations producing sisal fiber for
export. In the Brazilian Amazon, large areas of floodplain land and the
archaeological refuse dumps of the prehistoric societies were put into
intensive sugar production for national and international markets (Moreira
Neto 1988; Smith 1879). The need for wood for fuel and construction led to
extensive deforestation in the areas around the plantations.

The commercial extractive forestry industries were also closely linked
to colonial socioeconomic stratification. Worldwide, evidence of the history of
intensive extractive industries shows that they have usually been developed to
satisfy the economic interests of chiefdoms and states linked to regional or
world markets (e.g., Moreira Neto 1988; Rodrigues 1988; Hoffman 1984). Prior
to the extension of outside interests in the tropics, the forests were usually
cultivated for household subsistence and foraged for basic materials and food,
to collect items for regional trade, and for recreation. Scholars who argue that
commercial extraction is the only safe use for the forests appear to be unaware
that the "virgin" forests they are referring to were under crop production
much longer than they have been under intensive extraction.

In sum, one impact of socioeconomic stratification on land use in the
American tropics has been to limit the number of people who can be supported
on the land. Thus, the pattern of cash cropping, extraction, and cattle raising
by the large holders is one of the most important reasons for low density rural
occupation in the American tropical lowlands.

Considerations of political economy and social stratification, in addition
to ecology, are also paramount in just and effective solutions to the problem of
accommodating the rural poor of Amazonia. There is no evidence that their
low-intensity, sparse occupation has been harmful to the environment. The
detectable environmental harm there has come from commercial land use by
large owners and state agencies. Traditional small holders of these areas have
not been the source of the problem. When land tenure and taxation laws are
adjusted to prevent their dispossession and exploitation to eliminate
preferential treatment for large holders, less destructive land use systems will
be facilitated.

Methodological Insights from the Historical Perspective

What all these temporal and regional contrasts in land use show is that
the historical and comparative sciences are needed for the evaluation of
theories about land use capability. Theories are crucial because without them
we cannot begin to understand the world or use scientific knowledge to plan.
But the origins of theories often lie in intuitions with strong socio-political
content and weak empirical basis. Our view of what is logical and plausible is
often quite wrong, and existing data are often inadequate to reflect the
complexity of the phenomena under study. Increasingly accurate
explanations come from critical research that goes back and forth between
theory and data. The theory defines what data are relevant, and further
critical data gathering can correct the theory.

Both experimental and longitudinal data are needed for critical theory
testing. Experimental investigation is not enough for several reasons. It is too
costly to apply to all the large areas under evaluation, and in fact has never
been applied at a high-resolution in the tropics. This means that the results of
tests on soils in one area have to be extrapolated to another. If the nature of
the soils in the other areas is not the same, then the results do not apply. In
addition, tests of soil are not direct evidence of the land's ability to sustain a
certain kind of use long-term, especially if scientists do not study the lower
soil horizons and bedrock, which are what determine the soils capacity for
productivity under long-term use. Even agricultural trials are not definitive.
Short-term evaluation of cultivation projects is often misleading, because it
takes time for the nature of the results to become clear.

Accordingly, rigorous tests of theories must include data on the patterns
and effects of land use over long time periods.: Such evidence can be definitive
in a way extrapolations and localized experiments cannot be. However, just as
soil studies cannot be applied in areas with significantly different soil quality,
we need to apply this information advisedly and consider how it relates to
different areas of the Amazon.

The Future of Natural and Human Resources in Tropical America

The historical perspective on land use in tropical Latin America has
some lessons for planning the use and preservation of resources there.
Because of their ecological differences, the future of the forests and floodlands
and the different problems of plants, animals, and humans, need to be
separately considered.

The Forests

As documented above, large areas of the tropics have much higher
resource potential than the stereotypical nutrient poor terra firm tropical
rain forests. In many of these higher-nutrient areas, in prehistoric and early
historic times, people removed the forest for various types of long term,
intensive land use. Thus the "virgin" forests in a number of areas of tropical
Latin America are actually of post-conquest origin, the result of the
abandonment of cultivated land when Indian populations were decimated and
dispersed by conquest. In addition, large areas now luxuriantly forested were
not forested for long periods in prehistoric times due to geomorphological

processes. What do these facts mean for interpretations of the evolution of
Amazonian biota and what are the development and planning implications?

First, it is important that almost all the areas where forests have
regenerated have soils with abundant sources of nutrients, contributed either
from continually decaying bedrock minerals or from exogenous sources of
nutrients such as recent alluvium or the deep refuse of large prehistoric
human population centers. Therefore, we do not know whether long-term
intensive cultivation and clear cutting would be possible and whether forests
would regenerate in this fashion if the soils had been low-nutrient tropical
soils. Presumably, the weaker soils could not withstand such intensive use and
could be permanently degraded and their forests destroyed.

Second, we do not yet know if these long periods of deforestation and
cultivation altered floristic composition because there have been few studies
of the development of post-Pleistocene forests in the tropics. There is a great
danger, then, in situations of large scale, uncontrolled deforestation that both
forests could be destroyed forever. As Christopher Uhl has shown (this
volume), the forest on the nutrient-rich varzea of Pari seems to survive
continual, low intensity timbering by small holders but undergoes substantial
damage when intensively harvested by large companies. Most large
development projects do not provide for the restoration of the environment at
the end of the project. This is the case even though most of them, like the
hydroelectric dams, clearly are not indefinitely sustainable but
characteristically become obsolete relatively soon, due to sedimentation. In
the future, such projects need to both limit damage levels and include reserves
to shelter the species that will be needed to recolonize the area when the
project is over.

It is often claimed in opposition to forest conservation efforts that forest
resources are too valuable not to harvest. However, if the forests are valuable
resources now, they will be even more so in the future, and obviously it makes
no economic sense to destroy by overuse a resource that will be needed in the

The Animals

Numerous contemporary studies have shown that intensive hunting in
the vicinity of large settlements seems to alter the density and availability of
some species, favoring the survival of smaller, more demographically robust
taxa in favor of larger ones (e.g., Redford and Robinson 1987; Ross 1978). Such
effects may have occurred prehistorically in the vicinities of large

The archaeological finds show that along major rivers, forest hunting
was rarely an important part of the food supply during the demographic highs
of the last 2000 years of the prehistoric occupation, when subsistence was
based on seed crop cultivation and fishing. Instead, game appears in many
cases to have been limited to use as ceremonial or elite food. Since some large
prehistoric sites with abundant small fish remains lack large land fauna
altogether, it is possible that such animals became extinct locally. However,
the larger species absent in the archaeological garbage of some areas were
present in these areas in early historic times (e.g., Anonymous 1906; Roosevelt
1990a), which suggests that the large prehistoric populations did not

extinguish the species permanently. Other than the changes in the vicinity of
large settlements, the major impact on the animal population in ancient times
is likely to have been natural ecological change: large changes in water
quality and level and the massive depositional events and deforestation
episodes that occurred in the late Pleistocene and Early Holocene.

The danger of current and future deforestation for animal populations
is that there will be too little undisturbed forest to act as reservoirs for
repopulation. During the prehistoric and historic periods of deforestation, the
animal species that now inhabit the forests must have taken refuge in
neighboring forests or remained in the narrow gallery forests and forest
islands. As Redford has suggested (this volume), it is doubtful that forests can
regenerate without the presence of the animal species that help germinate
their propagules by eating and excreting them.


The history of fishery production does not seem to support the idea that
present fishery resources in mainstream areas are being damaged by overuse.
Some ichthyologists have suggested that permanent local damage may be
occurring (Goulding 1980), but so far the regional data do not show evidence of
this. In fact, the production records suggest that seasonal limitations on
effective fishing the Amazon waters have long limited harvests to a fraction of
the sustained yield levels theoretically possible. When refrigerated boats,
roads, and refrigerated trucks became available in the Orinoco, yields sky-
rocketed and continued to rise over several decades (Roosevelt 1980).
Although a given stream or backwater lake may be fished out in a given
season, its reconnection with the enormous network of waterways in the rainy
season means that this harvest will not affect its next year's productivity. It is
more common for a stream or river lake's fauna to be completely extinguished
annually by the recession of water, than it is for it to be completely fished out.

These considerations suggest that the closed water systems of the tropics
will be the most vulnerable to overuse, because their nutrients and fauna are
not replenished every year as in the floodplains. In addition, the reproductive
and behavioral characteristics of the animals can be expected to affect their
vulnerability to extinction. In some areas of the Amazon and Orinoco, recent
intensive commercial exploitation of very large and slowly reproducing
species, such as the manatee and pirarucu has definitely diminished the
abundance and contracted the geographic range of such species. In addition,
a few highly favored species, such as the Tambaqui, may be affected by
intense commercial exploitation (Gery 1984). Much more harmful to most fish
species than overfishing is serious damage to their habitat from change in
water quality from damming or erosion by the trampling of cattle or
deforestation. These processes alter the clarity, chemistry, and temperature of
the water, and consequently change the suitability of the habitat for different
fish, eliminating some and encouraging others.

Theoretically, the large prehistoric human populations and
concentrated settlements and the appreciable deforestation for cultivation
would have had the potential to impact fishery resources locally. In addition,
the construction of cultivation earthworks over large areas of the Amazon
floodplains in prehistoric times must have greatly changed habitats for fish.

Nonetheless, the prehistoric representation of fish species in areas of the
floodplains subject to intensive fishing and habitat alteration in precolombian
times does not seem to show that more species were present then than are
present now. In fact, the pattern seems to be that prehistoric harvesting was
more intensive and emphasis was on fewer species than are fished today. The
larger, more vulnerable species such as manatee and pirarucu were captured
in ancient times, but made up a very small part of consumption. What
predominated in prehistoric garbage are the bones of small fish that make up
the bulk of the fish biomass but that are not particularly important in today's
intensive commercial fishing, whose proprietors consider them "trash" fish.
Prehistoric and recent intensive fishing may have produced changes in
species representation, but there is no evidence so far of prehistoric
extinction of species, and in recent times, only the large, behaviorally and
reproductively vulnerable 'species appear to have been harmed by over-
exploitation. The implications of the evidence is that the majority of riverine
fish species have survived intensive harvesting and can continue to be an
important source of food and income for Amazonian populations.


Cattle ranching has been an important kind of land use over time in
Amazonian floodplain savannas for several reasons. The seasonally dry
climate in many areas makes cultivation more risky than ranging in some
years, and free ranging cattle are less labor intensive than planting, weeding,
harvesting, and transporting crops. Also, the relatively low rural population
density in some areas at present makes more intensive production
unnecessary there and difficult to provide labor for. In the context for the
larger economy, cattle ranching is a lucrative business, requiring relatively
low levels of investment and bringing a high return because of the high
demand for meat and other bovine products in urban and national markets.

Commonly, land use capability studies in Latin America classify many
floodland areas as land ecologically unsuitable for cultivation and therefore
apt only for cattle ranching. It has been pointed out (Brochado 1980),
however, that this kind of classification does not mean that the land cannot be
cultivated, and the archaeology shows that these areas were intensively
cultivated in prehistoric times (Roosevelt 1980; 1989). It is classified as
unsuitable for cultivation because cultivation would require capital intensive
improvements, such as drainage, that are not appropriate for the given
economic and demographic situation. In addition, large cattle ranchers who
tend to be influential in local and national government in Latin America
vigorously oppose the conversion of ranch land to subsistence agriculture.

Much of the literature decrying cattle-ranching in the American
tropics refers to ranching in upland forest areas. The low nutrient, seasonally
dry forests admittedly are poor areas for ranching, which causes erosion of
topsoil and is not sustainable, due to the poor nutrient supply to the soil.
Actually, as Nigel Smith has pointed out, pasture is very difficult to maintain
in evergreen rainforest areas, because shrubs and small trees quickly
recolonize, eliminating edible grasses. It is not clear whether ranching is
sustainable in the high-nutrient tropical forests, but such land has been
ranched for over a hundred years in Central America and is being converted
to pasture relatively rapidly at present in the Amazon (e.g., Descola 1981, 1982).

The limitations characteristic of the poor-soil tropical rain forests do
not apply to ranching on the hydromorphic soils of the Amazon and Orinoco
floodplains. In fact, in such areas as the Orinoco Llanos, Maraj6 Island, and
the Llanos de Mojos of Bolivia, cattle ranching is a venerable industry, dating
back to the mission period. The primary limitation to higher stocking rates in
such areas is water: too little of it in the dry season and too much in the rainy
season (references summarized in Roosevelt 1980; R. Cardoso, personal
communication). Although particular locales may lack one nutrient or
another, there is no overall lack of nutrients. Indeed, both land and water in
the ranching areas tend to be eutropic, ie. suffering from an excess of
nutrient elements. Thus, although the land is often overgrazed, the lack of
forage is due to lack of water, not due to deficiencies of nutrients in the soil

Like clear cultivation, cattle ranching causes erosion of soil and
sedimentation in rivers, because vegetation is removed and soil is trampled. In
addition, the massive deposition of bovine excreta greatly affects water
quality. The cattle also tend to reduce the native fauna and destroy gardens,
and the overgrazing eliminates edible grasses in favor of inedible ones.
Floodplain ranching could be much less harmful ecologically and more
productive if overgrazing were eliminated and fencing put in.

The large floodplain areas of the lowlands which have been used for
cattle since soon after the conquest are one of the important areas open to
alternative, higher intensity agricultural land-use in the future, when the
demand for crops increases. Conversion of this land, however, will require
significant changes in land-holding, demography, and capitalization of land
use in the area.


An important resource in both forests and floodplains in the American
tropics is land cultivable for subsistence and export.

As mentioned above, substantial areas of the Amazon forests and
floodplains were intensively cultivated with crops such as corn and cotton for
long periods in ancient times. Presumably, they could in the future be
returned to intensive crop cultivation. The cropping patterns and the
extensive mounding and drainage systems of the ancient people are obvious
models for future agricultural development. Currently, in forest and savanna
floodland near large urban centers in Amazonia, modest sized family
companies have successfully carried out commercial farming by planting
combinations of annual and tree crops. In particular, the ethnic Japanese of
the Bel6m area have been very successful in such farming for many decades,
furnishing vegetables and spices for the city markets. In addition, coca is
being cultivated intensively and profitably as an illicit cash crop in areas of
the Upper Amazon forests where dispersed Indian communities were
previously doing slash-and-burn cultivation of manioc, confirming the
archaeological evidence that the land is suitable for intensive crop
production. In the Guianas, people of Dutch ancestry have developed
successful permanent rice farming on the extensive floodlands that were
farmed with maize by the ancient moundbuilders.

Both the indigenous farming of prehistoric times and successful
contemporary projects provide models for rational economic development of
farming in Amazonia in the future.

Indigenous Peoples

What of the Indian populations and their territories? I have shown
that some of the land where Indians still live was used much more intensively
before the conquest than after, which implies that it might be used more
intensively in the future without permanent damage. However, the fact that
some land in Indian territory is capable of intensive cultivation does not mean
that it should be put under intensive cultivation, except if the Indians wish to
do so. The value of keeping some of the land out of intensive use is that it is
presently a refuge both for native peoples and for biota. The history of land
use in Amazonia makes quite clear that when land is settled and put to
intensive use by Euroamericans, Indians cannot survive culturally or

It is fashionable in American science to be more concerned with nature
than with humanity, and the media sometimes seems more concerned with
ecological destruction than with human misery and death. Presidential
scientific commissions and national committees on global change often do not
include social scientists, although, ultimately, global change is a problem with
obvious human dimensions, whether as causes or as impacts. However,
prehistory and paleontology show that although the forest and rich animal
populations seem to have survived enormous insults of both natural and
human origin, many human populations have not survived. Perhaps, then, we
need to preserve the forest for humanity's sake.

Ensuring that Indians retain their ancestral territories without
interference may also be the best way to preserve the environment. Present
Indian land usage is quite compatible with long-term preservation of both the
forest and the animal populations. In fact, keeping the land as Indian
preserves is a good way to make sure that plants and animals will have the
territory they need to survive. Some ecologists get exercized when Indians use
the natural resources, but there is as yet no evidence that use by Indians has
caused long-term irreversible damage to the environment. Their presence
may be an insurance policy for the preservation of the land.

Dangers and Potentials for Future Development

In the past we assumed that the prehistoric Indians did not impact the
environment, because we saw them living in small dispersed communities and
subsisting by long-fallow swiddening and foraging. However, the large
population concentrations of ancient times, the extensive deforestation for
cultivation, and the intensive use of wild resources undoubtedly had the
potential to impact the environment substantially. Nevertheless, the forests
and fauna of these areas survived to the extent that ecologists and
anthropologists have universally assumed them to be virgin, although, as the
archaeology shows, many are not virgin.

Although land-use by prehistoric Indian societies definitely had the
potential to damage the environment, countries with ties to urban, national,

and the world markets have incomparably greater potential for ecological
harm. The large, permanent population concentrations of urban civilization,
the large scale commercial lumbering, the pollution from manufacturing and
mining operations, and the destructive, ineffective hydroelectric dams, all
have a much greater ability to impact local ecology than indigenous
agriculture and foraging.

In addition, it is clear that permanent urban and suburban settlement
permanently destroys the natural habitats. Thus, the urban resettlement of
large numbers of the Brazilian poor in Amazonia has a substantial potential to
lead to permanent environmental harm. If there is no provision for
returning the land to forest, then obviously regeneration cannot occur.
However, due to the rural depopulation that followed the conquest, large
forested areas of Amazonia outside of the urban centers may be less impacted
ecologically than they were in ancient times.

One of the most significant potential land uses in Latin America is the
conversion of present pasture land to subsistence and export farming. The
large areas of ranch land that now serve to enrich large holders and serve the
urban market could be put to cultivation by cooperatives or small holders.
There would be many political barriers to this, but no ecological ones. As the
archaeology shows, the land was so used for many hundreds of years before
the conquest. Used for cultivation, the land would be able to support many
more people than it does today and probably more than in prehistory, when
people lacked the technical, genetic, and transportation aids available for
agriculture today.

Cattle ranching has had a severe impact on the soils, vegetation, and
wildlife, and increased rural settlement and subsistence farming would also
have an imp4.ct, although without free-ranging cattle, there would probably
be more forest on the land, thus more rain, less erosion, and more native
fauna. But since Mesticized and European people are already living in these
regions, and since the damage from ranching is a fait accompli, such
conversions would presumably not worsen the situation for the environment,
and it would benefit the populace by making it self-supporting. Whether this
kind of rural production would benefit Amazonian countries economically on
a national and international level, I must leave to scholars with experience in
that area. Some economic theories popular in the U.S. hold that an
impoverished, underemployed populace is necessary for fostering national
prosperity, but these theories are unproven and, some would say, unethical.

It may be that there will be no economic advantage to returning the
recently reforested land to intensive commercial cultivation. Present world
market factors may make the products of such use uneconomical. The use of
the land for extractive industries and forestry may be a better way for the
future, since extraction has less impact on the forest than intensive
subsistence farming and dense rural settlement, and many of the products are
salable. Since the majority of the land in industrial nations' urban zones has
already been deforested, and since large scale deforestation has adverse effects
on climate, there is an obvious advantage to preserving those tropical forests
that still exist.

What does this complex and interesting historical picture of resource
use mean for our role as scholars? Some of us may decide that facilitating the
development of nation states is our goal, in which case we may participate as
consultants, to try to develop land use of the type that is practicable and
effective for such organizations. Others of us may decide to use our knowledge
to help develop ways for rural individuals to use the land to support
themselves better. This direction may be the best one for a peaceful and
productive future, since it would allow large numbers of people to be self-
supporting and self-respecting. Although helping nation states to develop
may be a good short-term goal, history shows that increasing hierarchy and
centralization does not necessarily lead to greater economic stability or
sustained growth. Most states have had very short lifespans compared to less
organized societies and are associated with disruptive demographic. and
ecological fluctuations. As political entities they may be needed in the short
run for the people of a region to compete successfully with other nations,
given that we live in an international world, with nations hoeing fields, so to
speak, outside their borders. However, the historic perspective does not
support the view that nation states are the most rational and efficient agents
of economic development. As scholars, we need to look closely and seriously at
rural peasants and indigenous peoples as effective agents for sustainable
development in the American tropics in the future.


Anderson, A.B. 1983. The biology of Orbignia martiana (Palmae): A tropical dry
forest dominant in Brazil. Ph.D. Dissertation, University of Florida,

Allen, W.L. 1968. A ceramic sequence from the Alto Pachitea. Peru: Some
implications for the development of tropical forest culture in South
America. Ph.D. Dissertation, University of Illinois, Urbana.

Anonymous. 1906. Noticias da Ilha Grande de Johanes dos rios e das igarapes
que ter na sua circumferencia, de alguns lagos que se ter descoberta de
alumas couzas curiosas. Viagens no Brazil 7. Institute Historico
Geografico Brasileiro 6(1): 294-301.

Athens, J.S. 1989. Pumpuentsa and the Pastaza Phase in southeastern Ecuador.
Nawpa Pacha 24: 1-29.

Balee, W. 1989. The culture of Amazonian forests. In D.A. Posey and W. Balee,
eds., Resource management in Amazonia: Indigenous and folk strategies.
Advances in Economic Botany No. 7. New York: New York Botanical
Garden, 1-21.

Bartlett, A. S. and E.S. Barghoorn. 1973. Phytogeographic history of the
Isthmus of Panama during the past 12,000 years (a history of vegetation,
climate, and sea level change). In A. Graham, ed., Vegetation and
vegetational history of Latin America. New York: Elsevier, 203-299.

Bennett, C.F. 1968. Human influences on the zoogeography of Panama. Ibero
Americana, No. 51. University of California Press.

Boomert, A. 1976. Precolombian raised fields in coastal Suriname. Proceedings
of the Sixth International Congress for the Study of the Pre-Colombian
Cultures of the Lesser Antilles. Guadeloupe, 1975, 134-144.

----- 1980. Hertenrits: An Arququinoid complex in Northwest Suriname (Part
1). Journal of Archaeology and Anthropology 3(2): 68-104.

Brenner, M., B. Leyden, and M.W. Binford. 1990. Recent sedimentary histories
of shallow lakes in the Guatemalan savannas. Journal of
Paleolimnology. In Press.

Brochado, J. 1980. The social ecology of the Marajoara culture. MA thesis,
University of Illinois, Urbana.

Bronson, B. 1966. Roots and the subsistence of the Ancient Maya. Southwestern
Journal of Anthropology 22: 251-279.

Brown, M. 1989. Shattered worlds remade: Reflections on Amazonia crisis cults.
In A.C. Roosevelt, ed., Rethinking Native Amazonia. Wenner-Gren
Symposium No. 109. Volume preparation.

Bush, M.B. and P.A. Colinvaux. 1988. A 7000-year pollen record from the
Amazon lowlands, Ecuador. Vegetatio 76: 141-154.

Bush, M., D. Pipemo, and P. Colinvaux. 1989. A 6000 year history of Amazonian
maize cultivation. Nature 340: 303-305.

Campbell, K.E., Jr., and C.D. Frailey. 1984. Holocene flooding and species
diversity in southwestern Amazonia. Ouaternary Research 21: 369-375.

Campbell, K.E., Jr., C.D. Frailey, and J. Arellano L. 1985. The geology of the Rio
Beni: Further evidence for Holocene flooding in Amazonia.
Contributions in Science 364: 1-18. Natural Museum of Los Angeles

Cardoso, R. 1985. Personal communication.

Carneiro, R.L. 1957. Subsistence and social structure: An ecological study of the
Kuikuru Indians. Ph.D. Dissertation, University of Michigan, Ann

----- 1961. Slash and burn cultivation among the Kuikuru and its implications
for cultural development in the Amazon basin. In J. Wilbert ed., The
evolution of horticultural systems in Native South America: Causes and
consequences. 47-67.

----- 1970. Hunting and hunting magic among the Amahuaca of the Peruvian
montafia. Ethnology 9: 331-341.

Chagnon, N. 1968. Yanomamo: the fierce people. New York: Holt, Rinehart and

----- 1988. Life histories, blood revenge, and warfare in a tribal population.
Science 239: 985-992.

Coe, M. 1966. The Maya. New York: Praeger Publishers.

Cohen, M. and g. Armelagos, eds. 1984. Paleopathology at the origins of
agriculture. Orlando: Academic Press.

Colinvaux, P. 1990. Personal communication.

Cooke, R.G. 1972. The archaeology of the western Code province of Panama.
Ph.D. Dissertation, London University. 2 volumes.

----- 1975. Excavaciones arqueol6gicas en el Sitio AG-3 (Sitio Sierra), distrito
de Aguadulce, Code, Panama. Preliminary report on the 1975
excavations, Direcci6n de Patrimonio Hist6rico, Panama.

Cowgill, U.M. 1961. Soil fertility and the ancient Maya. Transactions of the
Connecticut Academy of Arts and Sciences 42: 1-56.

Denevan, W. 1966. An aboriginal cultural geography of the Llanos de Moios de
Bolivia. Ibero-Americana No. 48. Berkeley: University of California

Derby, 0. 1877. Contribuicao para a geologia do Baixo Amazonas. Archivos do
Museu Nacional 2: 77-104. Rio de Janeiro.

Descola, P. 1981. From scattered to nucleated settlement: A process of
socioeconomic change among the Achuar. In N.E. Whitten Jr., ed.,
Cultural transformations and ethnicity in Modern Ecuador. Urbana:
University of Illinois Press, 615-646.

----- 1982. Territorial adjustments among the Achuar of Ecuador. Human
societies and ecosystems 21(2). Social Science Information. London:
Sage, 301-320.

----- 1989. Homeostasis as a cultural system. in A.C. Roosevelt, ed., Rethinking
Native Amazonia. Wenner-Gren Symposium No. 109. Volume in

Enge, K. I. and S. Whiteford. 1989. Keepers of water and earth: Mexican rural
social organization and irrigation. Austin: University of Texas Press.

Erickson, C. 1980. Sistemas agricolas prehispAnicos en los Llanos de Mojos.
American Indigena 40(4): 731-755.

Farabee, W. 1922. Indian tribes of eastern Peru. Papers of the Peabody vMuseum
of American Ethnology and Archaeology Vol. 10. Cambridge: Harvard

Farriss, N.M. 1984. Maya society under colonial rule: The collective enterprise
of survival. Princeton: Princeton University Press.

Fejos, P. 1943. Ethnography of the Yagua. Viking Fund Publications in
Anthropology No. 1. New York.

Ferreira Penna, D.D. 1876. Breve n6ticia sobre os ceramics do Pari. Archives do
Museu Nacional 2: 47-67. Rio de Janeiro.

Flowers, N. 1973. Seasonal factors in subsistence, nutrition, and child growth
in a Central Brazilian Indian community. In R.B. Hames and W.T.
Vickers, eds., Adaptive responses of native Amazonians. Studies in
Anthropology. New York: Academic Press, 357-390.

Gery J. 1984. The fishes of the Amazon. In H. Sioli, ed., The Amazon: Limnologv
and landscape ecology of a mighty tropical river and its basin.
Dordrecht: Junk Publishers, 353-370.

Goulding, M. 1980. The fishes and the forest: Explorations in Amazonian
natural history. Berkeley: University of California Press.

Gross, D. 1975. Protein capture and cultural development in the Amazon basin.
American Anthropologist 77: 526-549.

Hammond, N. 1982. Ancient Maya civilization. New Brunswick: Rutgers
University Press.

Harner, M. 1973. The Jivaro: People of the sacred waterfall. Anchor books.
Garden City: Anchor Press/Doubleday.

Harrison, P.D. and B.L. Turner, II. 1978. Prehispanic Mava agriculture.
Albuquerque: University of New Mexico Press.

Heriarte M. de. 1964. Descricam do estado do Maranham, Para, Corupa, rios das
Amazonas. Faksimile--Ausgabe aus den MSS 5880 und 5879 der
Osterreichischen National-Bibliothek, Vien. Graz, Austria: Academische
Druke und Verlagsanstalt.

Hem, W.M. 1977. High fertility in a Peruvian Indian village. Human Ecology
5(40): 355-368.

Herrera F. 1989. The history of the Cuna Indians. Term Paper, Graduate
Seminar on Amazon Prehistory.

Herrera R. 1985. Nutrient cycling in Amazonian forests. In G.T. Prance and T.E.
Lovejoy, eds., Key environments: Amazonia. Oxford: Pergamon Press, 95-

Higham, C. 1989. The archaeology of mainland Southeast Asia. Cambridge:
Cambridge University Press.

Hill, K. and A. Magdalena Hurtado. 1989. Hunter-gatherers of the New World.
American Scientist. September-October: 436-444.

Hoffman, C.L. 1984. Punan foragers in the trading networks of southeast Asia.
In C. Schrire, ed., Past and present in hunter gatherer studies, 123-150.

Irion, G. 1984a. Sedimentation and sediments of Amazonian rivers and
evolution of the Amazonian landscape since Pliocene times. In H. Sioli,
ed., The Amazon: Limnology and landscape ecology of a mighty tropical
river and its basin. Dordrecht: Junk Publishers, 201-214.

----- 1984b. Clay minerals of Amazonian soils. In H. Sioli, ed., The Amazon:
Limnology and landscape ecology of a mighty tropical river and its
basin, 537-580.

Johnson, A. and T. Earle. 1987. The evolution of human societies from foraging
group to agrarian state. Stanford: Stanford University Press.

Johnson, G. 1978. Information sources and the development of decision-
making organizations. In C. Redman, M. Berman, E. Curtis, W.

Langhorne, N. Veraggi, and J. Wanser, eds., Social archaeology: Beyond
subsistence and dating. New York: Academic Press, 87-112.

Jones, J.C. 1984. Native peoples of lowland Bolivia. In M. Schmink and C.H.
Woods, eds., Frontier expansion in Amazonia. Gainesville: University of
Florida Press, 62-82.

Koch-Gruneberg T. 1967. Zwei iahre hunter den Indianern. Reisen in Nordwest
Brasilien 1903-1905. Graz: Akademische Druck- u. Verlagsanstalt.

Lathrap, D.W. 1970. The upper Amazon. New York: Praeger Publishers.

Leeds, A. 1988. Personal communication.

Lees, S. 1973. Sociopolitical aspects of canal irrigation in the valley of Oaxaca.
The Valley of Oaxaca I. University of Michigan Museum of Anthropology
Memoir No. 6. Ann Arbor.

Lopes-Ramos, E. 1983. Geologia de Mexico. Edicion Escolar.

Lundell, C. 1937. The vegetation of the Peten. Publication 478. Washington, D.C.:
The Carnegie Institution.

Marcus, J. 1982. The plant world of the sixteenth- and seventeenth-century
Lowland Maya. In K. Flannery, ed., Maya subsistence. New York:
Academic Press, 239-273.

McKeown, T. 1976. The modem rise of population. New York: Academic Press.

Meggers, B.J. 1954. Environmental limitations on the development of culture.
American Anthropologist 56: 801-824.

----- 1971. Amazonia: Man and nature in a counterfeit paradise. Chicago:

----- 1984. The indigenous peoples of Amazonia, their cultures, land use
patterns, and effects on the landscape and biota. In H. Sioli, ed., The
Amazon: Limnology and landscape ecology of a mighty tropical river
and its basin. Dordrecht: Dr. W. Junk Publishers, 627-648.

----- 1985. Aboriginal adaptation to Amazonia. In G.T. Prance and T.E. Lovejoy,
eds., Key environments: Amazonia. Oxford: Pergamon Press, 307-327.

Meggers, B.J., E.S. Ayensu, and W.D. Duckworth, eds. 1973. Tropical forest
ecosystems in Africa and South America: A comparative review.
Washington, D.C.: Smithsonian Institution.

Meggers, B.J. and C. Evans. 1957. Archaeological investigations at the mouth of
the Amazon. Bureau of American Ethnology, Bulletin 167. Washington,
D.C.: Smithsonian Institution.

Monteiro de Noronha, J. 1862. Roteiro da viagem da cidade do Pard ate as dltimas
col6nias do sertao da Provincia escripto na villa de Barcellos pelo
Vigirio geral do Rio Negro o Padre Jose Monteiro de Noronha no anno de
1768. Para: Typographia da Santos & Irmaos, 1-77.

Moreira Neto, C. de A. 1988. Indios da Amaz6nia. de maioria a minoria (1750-
1850). Petropolis: Editora Vozes.

Nimuendaju, C.U. 1949. Os Tapai6. Boletim do Museu Paraense Emilio Goeldi 10:

Pelto, G.H. and P.J. Pelto. 1983. Diet and delocalization: Dietary changes since
1750. Journal of Interdisciplinary History XVI (2): 507-528.

Petri, S. and V.J. Fulfaro. 1988. Geologia do Brasil. Slo Paulo: Editora da
Universidade de Slo Paulo.

Porras, P. 1987. Investigaciones arqueol6gicas a la Faldas del Sangay. Provincia
Morona Santiago: Tradici6n Upano. Quito: Artes GrAficas Senal.

Porro, A. 1981. Os Omagua da Amazonas. Demografia e padres de povoamento
no seculo XVII. In Contribuic6es a antropologia em homenagem ao
Professor Egon Schaden. Museu Paulista Serie Ensaios 4. Sao Paulo:
Universidade de Slo Paulo, 207-231.

----- 1983-84. Os Solim6es ou Jurimiguas. Territ6rio, migraqces e com6rcio
intertribal. Revista do Museu Paulista 29: 23-38.

----- 1989. Social organization and power in the Amazon floodplain. In A.C.
Roosevelt, ed., Rethinking native Amazonia. Volume in preparation.

Posey D.A. 1987. Contact before contact: Typology of post-colombian interaction
with the northern Kayapo of the Amazon basin. Boletim do Museu
Paraense Emilio Goeldi 3(2): 135-154.

Prance, G.T. and T.E. Lovejoy, eds. 1985. Key environments: Amazonia. Oxford:
Pergamon Press.

Puhleston, D.C. 1968. Brosimun alicastrum as a subsistence alternative for the
classic Maya of the Central Southern lowlands. MA thesis, University of
Pennsylvania, Philadelphia.

Putzer, H. 1984. The geological evolution of the Amazon basin and its mineral
resources. In H. Sioli, ed., The Amazon: Limnology and landscape ecology
of a mighty tropical river and its basin. Dordrecht: Junk Publishers, 15-

Redford, K. and J.G. Robinson. 1987. A game of choice: Patterns of hunting in
the Neotropics. American Anthropologist 89: 650-667.

Ranere, A. and P. Hansell. 1978. Early subsistence patterns along the Pacific
coast of Panama. In B.L. Stark and B. Voorhies, eds., Prehistoric coastal
adaptations: The economy and ecology of native Middle America. New
York: Academic Press, 43-59.

Rodrigues, A.G. 1988. Amazonia: Modificag6es socias e culturais decorrentes
do process de ocupaqao humana (Sec. XVII ao XX). Boletim do Museu
Paraense Emilio Goeldi. Antropologia 4(1): 65-115.

Roosevelt A.C. 1979. The Goldsmith: the Code Style of Ancient Panama, In A.C.
Roosevelt and J.G.E. Smith, eds., The ancestors: Native artisans of the
Americas. New York: Museum of the American Indian and University of
Washington Press, 66-101.

----- 1984. Population, health, and the evolution of subsistence: conclusions.
In M. Cohen and G. Armelagos, eds., Paleopathology at the Origins of
Agriculture. Studies in Archaeology. New York: Academic Press, 559-583.

----- 1987. Chiefdoms in the Amazon and Orinoco. In R.D. Drennan and C.
Uribe, eds., Chiefdoms in the Americas. Lanham, Md.: University Press of
America, 153-185.

----- 1989. Resource management in Amazonia before the conquest: Beyond
ethnographic projection. In D. Posey and W. Balee, eds., Advances in
Economic Botany 7. New York: New York Botanical Garden, 30-61.

----- 1990a. Moundbuilders of the Amazon: Geophysical archaeology on
Marai6 Island. Brazil. Orlando: Academic Press. In press.

----- 1990b. The excavations at Corozal: Stratigraphy and ceramic seriation.
Yale University Publications in Anthropology No. 82. In press.

----- 1990c. Ecological determinism in theories of social evolution in
Amazonia. In W. Neves, ed., Origins. adaptations, and biological diversity
of native Amazonians. Beldm: Museu Paraense Emilio Goeldi. In press.

Roosevelt, A.C., R.A. Housley, M. Inazio, S. Maranca, and R. Johnson. 1990. The
earliest pottery in the Americas 7000 B.P. from a shellmound in the
Brazilian Amazon. In preparation.

Ross, E.B. 1978. Food taboos, diet, and hunting strategy: The adaptation to
animals in Amazon cultural ecology, Current Anthropology 19(1): 1-36.

Ross E.B., and J.B. Ross. 1980. Amazon warfare. Science 207: 590-591.

Roth, W.E. 1924. An introductory study of the arts. crafts, and customs of the
Guiana Indians. Thirty-eighth Annual Report of the Bureau of
American Ethnology, 1916-1917. Washington, D.C.: Smithsonian

Salvador, M.L. 1976. the clothing arts of the Cuna of San Bias. In N.H. Graburn,
ed., Ethnic and tourist arts: Cultural expressions from the fourth world.
Berkeley: University of California Press, 165-182.

Sanders, W. and B. Price. 1968. Mesoamerica: The evolution of a civilization.
New York: Random House.

Schomburgk, R. 1922-1923. Travels in British Guiana, 1840-1844. Georgetown:
Daily Chronicle.

Schubert, C., H.O. Briceno and P. Fritz. 1986. Paleoenvironmental aspects of the
Caroni-Paragua river basin (Southeastern Venezuela). Interciencia
11(6) 278-289.

Sioli, H. 1973. Recent human activities in the Brazilian Amazon and their
ecological effects. In B.J. Meggers, E.S. Ayensu and W.D. Duckworth, eds.,
Tropical forest ecosystems in Africa and South America: A comparative
review. Washington, D.C.: Smithsonian Institution, 321-334.

Sioli, H., ed., 1984. The Amazon: Limnology and landscape ecology of a mighty
tropical river and its basin. Dordrecht: Dr. W. Junk Publishers.

Smith, H.H. 1879. Brazil. the Amazons and the coast. New York: Charles
Scribners Sons.

Smole, W.J. 1976. The Yanoama Indians: A cultural geography. Austin:
University of Texas Press.

Soil Survey Staff. 1975. Soil Taxonomy: A basic system of soil classification for
making and interpreting soil surveys. Soil Conservation Service,
Agricultural Handbook No. 436. Washington, D.C.: United States
Department of Agriculture.

Steward, J. 1949. South American cultures: an interpretive summary. In
Handbook of South American Indians. Vol. 5. The comparative ethnology
of South American Indians. Washington, D.C.: Smithsonian Institution,

Stocks, A. 1984. Indian policy in eastern Peru. In M. Schmink and C.H. Woods,
eds., Frontier expansion in Amazonia. Gainesville: University of Florida
Press, 33-61.

Stout, D.B. 1947. San Blas Cuna acculturation: An introduction. Viking Fund
Publications in Anthropology No. 9. New York.

van der Merwe, N., A.C. Roosevelt and J. Vogel. 1981. Isotopic evidence for
prehistoric subsistence change at Parmana, Venezuela. Nature 292: 536-

Vickers, W.T. 1984. Indian policy in Amazonian Ecuador. In M. Schmink and
C.H. Woods, eds., Frontier expansion in Amazonia. Gainesville: University
of Florida Press, 8-32.

West, R., ed. 1964. Handbook of Middle American Indians. Vol. 1. Natural
environment and early cultures. Austin: University of Texas Press.

White, C. and H. Schwarcz. 1989. Ancient Maya diet: As inferred from isotopic
and elemental analysis of human bone. Journal of Archaeological
Science 16(5): 451-474.

Whitehead, N.L. 1988. Lords of the Tiger Spirit: A History of the Caribs in
Colonial Venezuela and Guyana. Dordrecht: Foris Publications.

----- 1989. The ancient Amerindian polities of the lower Orinoco, Amazon, and
Guyana coast. A preliminary analysis of their passage from antiquity to
extinction. In A.C. Roosevelt, ed., Rethinking native Amazonia. Volume
in preparation.

Whitten, N.E. 1976. Sacha Runa: Ethnicity and adaptation of Ecuadorian jungle
Ouichua. Urbana: University of Illinois Press.

Willey, G.R. 1982 Maya archaeology. Science 215: 260-267.

Wiseman, F.M. 1978. Agricultural and historical ecology of the Maya lowlands.
In P.D. Harrison and b. L. Turner, II, eds., Prehispanic Mava agriculture.
Albuquerque: University of New Mexico Press, 63-116.

World Bank. 1975. Integrating women into development. Washington, D.C.:
World Bank.

Wust, I. 1987. A pesquisa arqueologica e etnoarqueologica na parte central do
territorio Bororo. Revista de Antropologia No. 30. Universidade de Sao

----- 1989. The eastern Bororo seen from an archaeological perspective. In
A.C. Roosevelt, ed., Rethinking native Amazonia. Volume in preparation.







Ecological communities are spatially heterogeneous and temporally
dynamic (Sousa 1984). Their structure is an epiphenomenon emerging from
the individualistic response of species to both local physical and biotic
conditions. That is, community structure emerges as a response of the local
species pool to the type and amount of available resources, and to the way in
which resources are allocated among species according to intra- and
interspecific interactions (Brown 1981, Maurer 1987). Processes shaping these
two constraints take place at different spatial and temporal scales, and occur at
different levels of biological organization, ranging from the individual to the
ecosystem level (Diamond & Case 1986, Ricklefs 1987), including interactions
between levels of organization (Allen & Starr 1982). Within this framework,
factors affecting the type and amount of available resources ought to be
considered as ultimate factors behind ecological changes that shape
community structure.

Prehistoric human populations are potential ultimate factors. Humans
have relied on wild resources for a long time before exploiting domesticated
plants and animals. Further, while relying on cultivars, humans have
continued modifying the landscape (Birks et. al. 1988). In fact, the history of
resource use by humans overlaps with the history of the ecological
communities. In Chile for instance, human populations arrived by 11,000 BP,
while modern communities are younger than 4,000 BP (Montane 1976, Heusser
1983, Villagran 1990). Since their arrival into a given area, humans have
interacted with syntopic species, being potentially capable of modifying both
the nature of the resource spectrum and the way in which resources are
exploited and allocated by all species, including humans themselves (e.g.,
Kohler and Matthews 1988). In fact, ecological communities may have a
component of human interaction more pervasive than usually recognized
(Russell 1989).

The nature and extent of prehistoric human impact on contemporary
communities is usually acknowledged in terms of direct effects and major
changes, such as deforestation and land conversion from forest to pasture and
croplands (Delcourt 1987, Russell 1989). Subtle, indirect effects of human
subsistence activities are seldom recognized (Russell 1989), despite the fact
that indirect effects (sensu Miller and Kerfoot 1987) may be the prevalent type
of interaction shaping community structure (Roughgarden and Diamond

Our aim in this paper is to suggest that prehistoric human populations
should be considered an ultimate factor in shaping community structure via
indirect effects which span across different levels of ecological organization.
We attempt to show that long-term landscape use and modification, largely due
to land-clearing for agricultural purposes, could be responsible for several
ecological traits, such as the body size of organisms, their population
abundance, species richness and guild structure as well as the spatial
variability observed in community structure.

To this end, we will focus on the prehistoric human exploitation of the
Andean mountains of central Chile. We plan to elaborate on the effects
triggered by a change in the economy of the inhabitants of central Chile, the

adoption of horticulture-agriculture. Associated with this lifestyle is an
increase in land cleared. The reduction of the original vegetation cover is a
major and significant ecological change. In fact, the original vegetation of
central Chile seems to have been a dense woodland rather than the
contemporary scrubland (Rundel 1981). Here, we will explore potential
indirect and subtle changes induced by progressive land-clearing.

Three limitations and one disclaimer must be clarified. Consistent with
our area of expertise, we will center our analysis on the Andean mountains of
the Maipo River region near Santiago, hoping it will be representative of
central Chile. Despite the lack of systematic archaeological and
anthropological studies in the Andean region of central Chile, this region is
comparatively well known (Niemeyer 1958, Weisner and Weisner 1964,
Andwanter 1969, Madrid 1977, Stehberg 1978, 1980a, 1980b, Stehberg and
Dillehay 1988, Stehberg and Fox 1977, Saavedra et. al. 1988, Duran and Planella
1989, Falabella and Stehberg 1989). Unless otherwise credited, the analysis is
based on our unpublished work in this region. Second, we will narrow our
analysis to the potential effects of land-clearing associated with horticulture-
agriculture, the "best-known" period in the prehistory of the region (e.g.,
Duran and Planella 1989, Falabella and Stehberg 1989). Therefore, we will
compare ecological features of individuals, populations, and communities prior
to and following the advent of agriculture. Third, due to our expertise and
available information (Simonetti and Cornejo 1987), we will focus on ecological
changes occurring in small mammals, as a proxy for a true community
analysis. Finally, although we use the term "effect of land-clearing"
extensively, what should be read is "presumed effect."


Central Chile has been inhabited by human populations since 1100 BP
(Montane 1976). From then on, wild plants and animals may have been either
directly affected by human activities such as hunting and harvesting, or
indirectly affected through habitat modifications (e.g., Simonetti 1984).
However, the nature and extent of the prehistoric human influence upon
community structure is a matter of debate. Some ecologists have assumed that
the biota was already depauperate at the time of European contact, due to the
exploitation and alteration of the landscape following the advent of native
agriculture (Miller 1980). Others have assumed that native Amerindians had
little impact on vegetation prior to the contact (Elizalde 1970), and that
significant ecological changes due to economic activities occurred only
following European contact (e.g., Cunill 1971, Aschmann & Bahre 1977, Fuentes
& Hajek 1979). Further, it is also debated whether human populations used the
landscape in similar fashion at any given time, implying that if prehistoric
human populations had an impact, this was equivalent across the landscape
(Madrid 1977, Stehberg 1980a).

The presence of human populations in the Andean mountains of central
Chile dates back to 5500 BP, although older occupations are feasible given the
scanty number of dated sites and samples. In fact, there are undated
archaeological remains which are stratigraphically older than those dated
5500 BP (Saavedra et. al. 1988). Therefore, at least since the Archaic, groups
with different cultural backgrounds have inhabited the mountains, using the
landscape under different temporal-spatial patterns. In fact, at no time did

humans exploit the mountain resources and settle in this region in a single
fashion. The use of the mountains apparently was unrelated to broad
environmental features common to the complete region, such as altitudinal
vegetation changes. Occupation, however, varied according to site-specific
characteristics of each watershed (see Stehberg 1980a, Cornejo 1987).

During the Archaic Period (5000 BP 2000 BP), bands of hunter-
gatherers inhabited the mountains on a seasonal basis, relying on camelids,
possibly guanacos (Lama guanicoe) and, to a lesser extent, rodent consumption
(Stehberg & Fox 1977, Simonetti & Cornejo 1990). Abiotic resources, such as
obsidians and other volcanic rocks were also searched for and exploited in the
mountains of the Maipo River region. Sites were occupied seasonally
according to elevation, but in a given watershed, humans exploited biotic
resources over the complete altitudinal sequence, ranging from the flat valley
bottoms to the higher plateaus (700 to 3,000 masl). Fire, either unintentionally
propagated from campsites or set to drive game, was a common anthropogenic
disturbance that might have contributed to opening the original woodland
(Aschmann 1978, Heusser 1983).

During the Early Agricultural Period (2000 BP 1400 BP), the mountains
were differentially used. For instance, while at the El Manzano watershed,
rock shelters were used only as camping grounds while in transit, at El Yeso
watershed a base camp was established. In both areas, though, sites were used
also as cemeteries, suggesting a more permanent use. Subsistence was
presumably mixed, relying on wild mammals but some horticulture was also
practiced at low elevation sites along with the exploitation of domesticated
camelids (Andwanter 1969, Stehberg 1978, L. Cornejo unpublished). Reduction
of the woodlands may have increased due to continued wood harvesting, fires
for driving game or clearing land patches for horticulture.

By 1400 BP 500 BP significant changes took place regarding the
occupation of the mountains at almost each watershed within the region. At El
Manzano, for instance, a large permanent camp was established in the valley
bottom while the upper elevation rockshelters previously exploited on a
seasonal or occasional basis were abandoned after 3,500 years of use. However,
human populations continued to occupy other rockshelters located in adjacent
watersheds, such Los Maitenes, either as occasional or seasonal camping sites
(Niemeyer 1958, Madrid 1977). By this time, the large and flat valley bottom of
the Maipo River supported a large human population, attested to by the
number of cemeteries (five in less than seven km along the river course
[Madrid 1977]). Consequently, more land may have been cleared, converting
the woodlands into a more open vegetation.

Human subsistence turned into agriculture, but some groups may have
continued hunting and gathering (Keller 1952, Madrid 1977, Stehberg 1980a),
which could explain the differential pattern of landscape use. Then, some
sites, particularly rockshelters, continued being occupied by nomadic hunting
bands. Some of these groups used camping sites physically close to groups
practicing agriculture, but in other cases these groups excluded themselves
spatially. It should be mentioned that nomadic bands of hunters were present
at the time of the Spanish conquest and survived until the eighteenth century
(Stehberg 1980a).

The case of the El Manzano setting is noteworthy, as it may reveal the
establishment of some sort of human territories after the advent of agriculture
(Dyson-Hudson and Smith 1978). Prior to the full establishment of agriculture,
the watershed was used intensively albeit seasonally at every elevation.
However, with the advent of agriculture, neither the group located at the
valley bottom nor any other discernable cultural group used rockshelters
located in this area. This pattern suggests that when people started to rely
more heavily on cultivars, they precluded the entry of other groups to the

Consequently, the mountains would have been exploited in a mosaic
fashion by this time, with groups modifying the landscape through land-
clearing for agricultural purposes while others could be triggering ecological
changes through the direct exploitation of wild species.

Agriculture must have involved land-clearing, but the exploitation of
wild species may have not. Then, prehistoric impacts upon the distribution
and abundance of local species may have been different throughout the
landscape at any given time. This pattern of landscape exploitation may be
reflected in the patterning of ecological communities. In fact, at least
contemporary communities in the Andean mountains near Santiago are
characterized by a mosaic of patches differing in composition, cover and
dynamics, attributes presumably related to present-day resource uses,
neglecting the eventual role of long-term human-resource interaction
(Fuentes et. al. 1986, Simonetti and Falabella 1986).

The environmental alterations caused by native agriculture in terms of
land-clearing and associated disturbed lands may have contributed to the
structure of ecological communities in subtle ways, particularly if humans
were using the landscape in a mosaic fashion. In the following section we
present some cases that shed light on this point.


The landscape alteration of subsistence activities may modify both the
nature and quantity of resources available for wild species. Here, we will
explore the potential effects of land-clearing for the structure of mammalian
assemblages at the individual, population, and community levels. As
mentioned, we will emphasize the indirect effects taking place within and
between adjacent levels of organization.

Individual level. At the individual level, body size is an important ecological
variable (Peters 1983). Within a lineage, larger individuals are usually
dominant in intraspecific encounters, spend less energy on maintenance per
unit of body mass, and tend to be more efficient at extracting energy from food
resources (Peters 1983, Brown and Maurer 1986). These attributes confer
higher fitness to larger individuals. Therefore, an unstable escalation of body
size is expected among individuals facing a reduction in the amount of
resources available (Maynard Smith and Brown 1986).

Species associated with or restricted to the original vegetation will face
a depressed profile of resources after land-clearing for agriculture.
Generalist species, however, will encounter not a reduction in carrying
capacity, but a modified profile of resources. Therefore, body size of
organisms associated with the original vegetation should be larger following
the advent of native agriculture and land-clearing, but no change should be
evident among generalist species.

Such is the case of two rodents (Simonetti and Saavedra 1990).
Currently, Octodon bridges is restricted to dense woodlands in central Chile,
while Oryzomys longicaudatus is a habitat generalist (Simonetti 1989). Land
clearing by prehistoric people can be related to a reduction in the carrying
capacity for -. bridges but not for Q. longicaudatus. Consequently, the body
size of Q. bridges ought to be smaller prior to the advent of native agriculture
while the body size of Q. longicaudatus ought to remain constant. In fact, body
size of Q. bridges, expressed by the length of the toothrow, increases from 9.8
+/- 0.1 mm among individuals that lived prior to the advent of agriculture
(aged 4,000 BP and older), to 10.2 +/- 0.08 mm in individuals that lived later on
(1,300 BP and younger). In contrast, body size of Q. longicaudatus did not
change significantly between individuals inhabiting unaltered or human-
modified woodlands: 4.01 +/- 0.01 mm versus 3.95 +/- 0.05 mm (Simonetti and
Saavedra 1990).

That is, with the advent of native agriculture, rodents faced a modified
landscape that affected them differentially, triggering changes in body size.
Given that body size affects the total amount of energy channeled by an
individual organism and their population (Brown and Maurer 1986), human
activities are affecting the way in which resources and energy are allocated
among coexisting species (Brown 1981). In this regard, human activities could
be considered the ultimate factor shaping body size of habitat specialists and a
modifier of the energy flux among species within a community.

Population level. At the population level, land-clearing had a different,
although coupled, impact with that occurring at the individual level. The
reduction in vegetation cover modified not only body size, but also the
population numbers of species closely associated with it, leading to their local

Large-bodied species had lower population densities than smaller-sized
species. Low population size in turn carries a higher risk of extinction (Pimm
et. al. 1988). Therefore, human activities are increasing the chances of
extinction of those species tied to the original vegetation, not only by reducing
the total amount of resources available, but also due to an increase in the
likelihood of extinction due to demographic stochasticity.

The reduction in vegetation cover modifies the type of habitat available
for different species according to their preferences. Therefore, species
associated with open and disturbed landscapes ought to become dominant or
more abundant over time, compared to species associated with the native
vegetation (Simonetti 1989a). Prehistoric human activities may then be
modifying population numbers.

This seems to be the case for Octodon species. The current abundance of
. degus relative to Q_. lunatus is a function of shrub cover (Glanz 1977).
Octodon degus is more abundant in open, disturbed shrublands, while both Q.
lunatus and Q. bridges are restricted to dense woodlands, the presumed
original vegetation of central Chile (Glanz 1977, Simonetti 1988). The
abundance of 0. degus therefore is expected to increase relative to other
Octodon species through time, as vegetation would have been progressively
reduced and altered. In fact, the abundance of Octodon degus is positively
associated with time, while the relative abundance of Q. bridgesi-lunatus
decreases among Octodon remains from El Manzano and La Batea rockshelters,
both at the El Manzano watershed (Simonetti 1989a). The decrease in
abundance dates to 1,500 BP, a time in which native agriculture was already
established in central Chile (Falabella and Steberg 1989).

Octodon degus is the presumed most common small mammal in central
Chile nowadays (Simonetti 1988 and references therein). Such an abundance
could therefore be associated with the historical degradation of the native
vegetation (Simonetti 1986a, 1989a). Prehistoric human populations then
could be considered an ultimate factor shaping the abundance profile of
rodent species.

Community level. The effect of prehistoric human activities upon populations
also has consequences at the community level. On the one hand, it modifies
the number of coexisting species. That is, humans affect the way in which
available resources can be utilized and allocated (Brown 1981). On the other
hand, changes in the abundance of small mammals modifies the resource
spectrum for local predators (Simonetti 1988).

In terms of species richness, at least regarding caviomorph rodents,
richness is lower in recent times. Two local extinctions seem to have occurred.
Aconaemvs fuscus and Octodon bridges have all but disappeared from the
Andes near Santiago. The last record of Octodon bridges coincides with the
advent of native agriculture. As for A.. fuscus, it disappeared only in the most
recent strata (Simonetti and Cornejo, unpublished). This pattern suggests that
although prehistoric human populations might have driven _. bridges to
extinction, their influence was not pervasive enough to lead to extinction
other rodent species, which tend to be habitat generalists, and possibly then
more resilient to habitat changes (Simonetti 1989a). That is, prehistoric
humans might have altered only partially the rodent species pool.

Regarding community organization, humans may have altered the guild
structure of local predators. Assuming that predators have behaved similarly
through time, we can speculate regarding the impact of human degradation of
the original vegetation upon guild structure among vertebrate predators. At
present, the assemblage of local predators includes a carnivorous guild,
formed by three hawks (Buteo polyosoma, Geranoaetus melanoleucus, and
Parabuteo unicinctus) together with a fox (Dusicyon culpaeus). This guild
emerges from the opportunistic reliance of predators upon the consumption of
Q. degus as a primary prey, cuing on its abundance and size (Jaksic et. al. 1981,
Jaksic 1983). As long as the guild structure of predatory species emerges as an
opportunistic response to prey supply levels, these guilds can be considered as

a human-induced community trait (Simonetti 1988). Variables affecting the
proximate factor shaping predator diets and, by extension, guild affiliation,
ought to be considered ultimate agents determining guild structure.
Therefore, as long as prey supply levels--the abundance of ... degus--depends
on the degree of human alteration of the original vegetation, the predatory
guild emerges as an indirect effect of human subsistence activities (Simonetti
1988, 1989b; see also Simonetti 1986b).


Prehistoric human populations have been occupying the Andean
mountains of central Chile for at least 5,500 years. A pervasive effect of this
occupation has been a reduction of the original vegetation. This landscape
modification seems to have changed several ecological attributes of the
individual organisms, populations and communities of the area. These
changes were triggered by the process of land-clearing related to the adoption
and expansion of agriculture as a subsistence strategy among prehistoric
human populations. Accordingly, ecological communities may have been
already shaped in subtle and indirect ways by human activities prior to the
European contact (Miller 1980). Furthermore, these changes may have
occurred in a patchy fashion across the landscape following the different
types of occupation that coexisted at the regional level (Stehberg 1980a). As
such, humans have indeed the potential to indirectly influence community
structure by depressing the resource base and the local species pool,
triggering a series of linked changes that move across different levels of
ecological organization.


We are indebted to K. H. Redford and S. E. Sanderson for inviting us to
the Conference. This work has been supported by FONDECYT, most recently
through grant 871-89. Additional support has been provided by Universidad de
Chile, grant DTI 2596-8934.


Allen, T. F. H. and T. B. Starr. 1982. Hierarchy: perspectives for
ecological complexity. Chicago: University of Chicago Press.

Andwanter, M. 1969. Casa de piedra "Las Quiscas, La Dehesa, Comuna
de Las Condes, Prov. de Santiago, Chile. Actas, V Congreso
National de Arqueologfa Chilena 1: 295-306.

Aschmann, H. 1978. "Aboriginal Use of Fire." In H. A. Mooney and C. E.
Conrad (tech. coord.), Environmental Consequences of Fire and
Fuel Management in Mediterranean Ecosystems. USDA
Forest Service, General Technical Report WO-3, Washington, D.C.

Aschmann, H. and C. Bahre. 1977. "Man's Impact on the Wild
Landscape." In H. A. Mooney, ed., Convergent evolution in Chile
and California. Pennsylvania: Dowden, Hutchinson & Ross.

Birks, H. H., H. J. B. Birks, P. E. Kaland, and D. Moe, eds. 1988. The
Cultural Landscape -- Past. Present and Future. Cambridge:
Cambridge University Press.

Brown, J. H. 1981. "Two Decades of Homage to Santa Rosalia: Toward a
General Theory of Diversity." American Zoologist 21: 877-888.

Brown, J. H. and B. A. Maurer. 1986. "Body Size, Ecological Dominance
and Cope's Rule." Nature 324: 248-250.

Cornejo, L. 1987. "Site Catchment Analysis o Analisis de Area de
Cobertura?" Clava 3: 27-31.

Cunill, P. 1971. "Factores en la Destrucci6n del Paisaje Chileno:
Recolecci6n, Caza y Tala Coloniales." Informaciones
Geogrificas Chilenas 20: 235-264.

Delcourt, H. R. 1987. "The Impact of Prehistoric Agriculture and Land
Occupation on Natural Vegetation." Trends in Ecology and
Evolution 2: 39-44.

Delcourt, P.A. and H. R. Delcourt. 1987. Long-term Forest Dynamics of
the Temperate Zone. New York: Springer-Verlag.

Diamond, J. and T. J. Case, eds. 1986. Community Ecology. New York:
Harper & Row.

Duran, E. and M. T. Planella. 1989. "Consolidaci6n Agroalfarera:
Zona Central (900 a 1470 d.C.). In J. Hidalgo, V. Schiappacasse,
H. Niemeyer, C. Aldunate and I. Solimano, eds., Culturas de
Chile: prehistoria. Santiago: Editorial Andres Bello, 313-27.

Dyson-Hudson, R. and E. A. Smith. 1978. "Human Territoriality: An
Ecological Reassessment." American Anthropologist 80: 21-41.

Elizalde, R. 1970. La Sobreviviencia de Chile. 2a. ed., Ministerio de
Agriculture, Santiago.

Falabella, F. and R. Stebergh. 1989. "Los Inicios del Desarollo
Agricola y Alfarero: zona central (300 a.c. a 900 d.c.).
In Hidalgo, Schiappacasse, Niemeyer, Aldunate and Solimano,
eds., 295-311.

Fuentes, E. R., A. J. Hoffman, A. Poianig and M. C. Alliende. 1986.
"Vegetation Change in Large Clearings: Patterns in the
Chilean Matorral." Oecologia (Berlin) 68: 358-366.

Fuentes, E. R. and E. R. Hajek. 1979. "Patterns of Landscape
Modification in Relation to Agricultural Practice in Central
Chile." Environmental Conservation 6: 265-271.

Glanz, W. E. 1977. Comparative Ecology of Small Mammal Communities
in California and Chile. Unpublished doctoral dissertation,
University of California, Chile.

Heusser, C. J. 1983. "Quaternary Pollen Record from Laguna de Tagua,
Chile." Science 219: 1429-1432.

Jaksic, F. M. 1983. "The Trophic Structure of Sympatric Assemblages
of Diurnal and Nocturnal Birds of Prey." American Midland
Naturalist 109: 152-162.

Jaksic, F. M., H. W. Greene and J. L. Yanez. 1981. "The Guild Structure
of a Community of Predatory Vertebrates in Central Chile."
Oecologia (Berlin) 49: 21-28.

Janzen, D. H. 1988. "Tropical Ecological and Biocultural Restoration."
Science 239: 243-244.

Keller, C. 1952. Introduction to J. T. Medina, Los Aborfgenes de
Chile. Santiago: Fondo Hist6rico y Bibliografico Jose Toribio,

Koehler, T. A. and M. H. Matthews. 1988. "Long-term Anasazi Land
Use and Forest Reduction: A Case Study from Southwest Colorado."
American Antiquity 53: 537-564.

Madrid, J. 1977. Ocupaci6n Indigena en el Valle Superior del Rio
Maipo. Tesis de Licenciatura, Universidad de Chile, Santiago.

Maurer, B. A. 1987. "Scaling of Biological Community Structure: A
Systems Approach to Community Structure." Journal of
Theoretical Biology 127: 97-110.

Maynard Smith, J. and R. L. W. Brown. 1986. "Competition and Body
Size." Theoretical Population Biology 30: 166-179.

Miller, S. 1980. Human Influences on the Distribution and Abundance
of Wild Chilean Mammals: Prehistoric Present. Unpublished
doctoral dissertation, University of Washington, Seattle.

Miller, T. E. and W. C. Kerfoot. 1987. "Redefining Indirect Effects."
In W. C. Kerfoot and A. Sih, eds., Predation: Direct and Indirect
Impacts on Aquatic Communities. Hanover: University Press of
New England.

Montane, J. 1976. "El Paleoindio en Chile." Actas, XLII Congreso
International de Americanistas 3: 492-497.

Niemeyer, H. 1958. "Ocupaci6n Indigena en el Rio Colorado, Afluente
del Maipo." Revista Universitaria 43: 117-122.

Peters, R. H. 1983. The Ecological Implications of Body Size. Cambridge:
Cambridge University Press.

Pimm, S. L., H. L. Jones, and J. Diamond. 1988. "On the Risk of
Extinction." American Naturalist 132: 757-785.

Roughgarden, J. and J. Diamond. 1986. "Overview: The Role of Species
Interactions in Community Ecology." In Diamond and Case, eds.,
Community Ecology. New York: Harper & Row, 333-343.

Ricklefs, R. E. 1987. "Community Diversity: Relative Roles of Local and
Regional Processes." Science 235: 167-71.

Rundel, P. W. 1981. "The Matorral Zone of Central Chile." In F. Di Castri,
D. W. Goodall and R. L. Specht, eds., Mediterranean Type
Shrublands. Amsterdam: Elsevier, 175-201.

Russell, E. W. B. 1989. "Meeting Reviews: Human Impact on Neotropical
Communities." Bulletin of the Ecological Society of America
70: 61-62.

Saavedra, M., L. Corejo and F. Amello. 1988. Arqueologfa de la
Precordillera de Chile Central. Restimenes, XI Congreso
Nacional de Arqueologfa Chilena, Santiago, 33.

Simonetti, J. A. 1984. "Late Pleistocene Extinctions in Chile: A
Blitzkrieg?" Revista Chilena de Historia Natural 57: 107-10.

----- 1986a. Microhabitat Use by Small Mammals in Central
Chile. Unpublished doctoral dissertation, University of
Washington, Seattle.

----- 1986b. "Human-induced Dietary Shift in Dusicyon
culpaeus. Mammalia 50: 406-408.

----- 1988. "The Carnivorous Predatory Guild of Central
Chile: A Human-Induced Community Trait?" Revista Chilena
de Historia Natural 61: 23-25.

----- 1989a. "Small Mammals as Paleoenvironmental
Indicators: Validation for Species of Central Chile." Revista
Chilena de Historia Natural 62: 109-14.

----- 1989b. "Predator Diets, Guild Structure and Human
Disturbance: A Rebuttal to Meserve's Criticisms." Revista
Chilena de Historia Natural 62: 13-17.

Simonetti, J. A. and L. E. Corejo. 1987. "Bibliografia Zooarqueol6gica
Chilena." Medio Ambiente 8: 113-20.

----- 1990. "Archaeological Evidence of Rodent Consumption in
Central Chile." Submitted to Latin American Antiquity.

Simonetti, J. A. and F. Falabella. 1986. "Interacci6n Hombre-Recurso:
La Necesidad de un Enfoque Temporal." Actas, 29. Encuentro
Cientifico sobre el Medio Ambiente 3: 462-67.

Simonetti, J. A. and B. Saavedra. 1990. "Body Size of Small Mammals
in Human-Modified Landscapes." Manuscript.

Sousa, W. P. 1984. "The Role of Disturbance in Natural Communities."
Annual Review of Ecology and Systematics 15: 353-92.

Stehberg, R. 1978. "El Cementerio Alfarero Temprano de Chacayes,
Interior del Caj6n del Maipo, Datado en 430 Adios d.C. Actas,
IV Congress de Arqueologia Argentina: 301-20.

----- 1980a. "Aproximaci6n Metodol6gica al Estudio del Poblamiento
Humano de los Andes de Santiago (Chile)." Boletin del Museo
Nacional de Historia Natural (Chile) 37: 43-60.

----- 1980b. "Ocupaciones Prehispanicas en El Arrayin, con
Especial Referencia al Alero de Novillo Muerto." Boletin
del Museo Nacional de Historia Natural (Chile) 37: 43-60.

Stehberg, R. and T. D. Dillehay. 1988. "Prehistoric Human
Occupation in the Arid Chacabuco-Colina Ecotone in Central
Chile." Journal of Anthropological Archaeology 7: 136-62.

Stehberg, R. and K. Fox. 1977. "Excavaciones Arqueol6gicas en el
Alero Rocoso de Los Llanos, Interior del Arrayan, Provincia
de Santiago." Actas, VII Congreso de Arqueologia de Chile
1: 217-41.

Villagran, C. 1990. "History of the Temperate Forest of Chile
During the Late Glacial and the Holocene Periods."
International Workshop: Comparative Studies of North and

South American Temperate Ecosystems: Theory, Approaches,
and Limitations. Termas de Chillan, Chile.

Weisner, L. and R. Weisner. 1964. "Recolecci6n de Superficie de
La Dehesa de Lo Barnechea, Las Condes, Provincia de Santiago.
Actas, III Congreso Internacional de Arqueologfa Chilena
1: 183-5.








Tropical deforestation has become the subject of considerable debate
and concern. Experts disagree over how quickly tree-covered land near the
equator is being cleared (Lanly 1982) and deforestation's impacts on global
climate (Detwiler and Hall 1988). By contrast, the claim that land clearing
threatens the world's stock of "biological information" is less controversial.
Tropical forests' high biological diversity is indisputable; although they cover
less than 10 percent of the Earth's land surface, they contain approximately
half the world's plant and animal species (Wilson 1988).

Concern over the impacts of tropical deforestation in Africa, Asia, and
Latin America has begun to stimulate analysis of national policies
conditioning human interaction with natural environments in the developing
world. World Bank economists have found that subsidies and tax breaks
encourage land clearing in the Brazilian Amazon (Binswanger 1989; Mahar
1989). In addition, the argument is made that rural development policies
pursued in a number of countries displace the rural poor, who tend to resettle
in environmentally fragile hinterlands (Blaikie 1985). To date, however,
investigation of the social context of deforestation has not included much
analysis of the tenurial incentives at work along expanding agricultural
frontiers. This is a serious omission since, as Bromley and Cernea (1989) point
out, tangible environmental problems in developing countries are often a
manifestation of underlying institutional crisis.

This paper addresses that crisis, tropical deforestation in Ecuador
serving as a case study. To begin, the tenure regime facing those who live in
or use tree-covered land in that country is described. Next, four specific
institutional incentives for deforestation in Ecuador and other Latin American
countries are examined. First, the waste and misuse of forest resources is, in
part, a classic open access problem. Second, stipulating that deforestation is a
prerequisite for land tenure sets in motion a cycle of excessive land clearing
and erosive farming. Third, bureaucratically induced tenure insecurity
further diminishes private incentives to conserve natural resources. Fourth,
formal property law in Latin America induces the demise of indigenous
common property regimes, which have long provided a framework for
sustainable agriculture and forest conservation.

Based on this examination of these four institutional incentives, we
conclude this paper with discussion of policy reforms needed to ensure the
conservation of Latin America's tropical forests.


It is appropriate to begin an overview of any Latin American country's
institutional regime by recognizing that the state makes extensive claims on
the natural environment. In Ecuador, for example, subsurface resources are
government property. With passage of the 1972 Water Law, all water resources
were nationalized. Coastal wetlands are "national patrimonies." Similarly,
most of the country's tree-covered land is designated as "forest patrimony" or
national parks.

These claims far outstrip the government's capacity to manage
resources or even to ensure that its claims are honored by the public at large.
Weak management of Ecuador's public forests is a case in point. No rangers
work in the 2,000,000 ha of forest patrimony delimited in the northwestern
and northeastern parts of the country (MAG 1987) and, as of 1987, a mere two
administrators, 25 technicians, and 119 permanent and seasonal rangers had
been assigned to the 2,100,000 ha of parks in continental Ecuador (DINAF 1988).

A marked discrepancy between public sector claims on resources and
the government's capacity to control access to "its properties" can bring about
what Hardin (1968) calls a "tragedy of the commons." Recognizing the
potential for open access problems, Ecuador's government allows individuals
and firms to acquire public lands. However, ecosystem destruction is typically
a prerequisite for private tenure. Private parties interested in forest
management, for example, cannot acquire legal interests in tree-covered land,
timber concessions having been banned in 1982. Instead, the Ecuadorian
Institute for Agrarian Reform and Colonization (IERAC) has only adjudicated a
claim for private tenure in a frontier parcel if at least half of that parcel has
been cleared.

By no means does deforestation win an agricultural colonist formal
tenure quickly. IERAC requires a long time, often years, to adjudicate claims
for formal property rights. Delays are explained in part by administrative
constraints. IERAC's record-keeping system is extremely cumbersome and the
agency did not acquire its first computer until the late 1980s. In addition, the
complexity of formal property law draws out the adjudication process. As
Seligson (1984) points out, IERAC is obliged to execute ten separate procedures
during the course of settling a tenure claim.

To complete a description of institutional conditions in Ecuador's
tropical forests, one must consider the impacts of formal property law on the
country's indigenous inhabitants. Those impacts have been the subject of a
great deal of anthropological research. Macdonald (1981) reports, for example,
that the periodic fallowing scheme long practiced by the Amerindian
community of Pasu Urcu, in eastern Ecuador, was abandoned during the 1970s
after IERAC agents informed the community that fallow lands could be claimed
by agricultural colonists, who were 50 km away at the time. This and other
case studies suggest that Amerindians respond to tenurial incentives much as
do agricultural colonists. As a result, indigenous resource management
regimes are discarded.

Ecuador's property rights regime is entirely representative of
institutional conditions throughout Latin America. In every country with
extensive tropical forests, the public sector's claims on tree-covered land far
outstrip its ability to manage or control resources. Accordingly, open access
problems are chronic. Throughout the region, deforestation is a prerequisite
for formal tenure. Agricultural colonists in the Brazilian Amazon, for
example, obtain title in a forested parcel only by clearing a large part of it
(Mahar 1989). By the same token, tenure insecurity is a problem in most of
Latin America. IERAC's time-consuming adjudication procedures are followed
throughout the region by counterpart agencies established in the early 1960s
under the auspices of the Alliance for Progress, as de Soto (1989) has
documented vividly in a case study undertaken in Lima, Peru. Finally,

suppression of indigenous groups' tenurial arrangements is the norm in many
Latin American countries.

In the sections that follow, we describe how each of these elements of
the institutional order in Latin America contributes to depletive human
interaction with the natural environment.


Four specific institutional incentives for deforestation are relevant to
much of Latin America. The first of these is the problem of "open access." The
distinction between "open access" and "common property" is a celebrated one
in the resource economics literature, although one continually finds
confusion between the concepts at the level of both theory and policy. As
Ciriacy-Wantrup and Bishop (1975) clearly show, common property (res
communes in Roman law) and open access (res nullius) are two quite different
structures of property rights.

Often, what appears to the outside observer to be open access may
involve tacit cooperation by individual users according to a complex set of
rules specifying rights of joint use. This is common property. Empirically, it
is crucial to distinguish between open access and common property if
appropriate policy is to be formulated. Problems of open access arise from
unrestricted entry, whereas problems of common property result from tension
in the structure of joint use rights adopted by a particular village or group.
These tensions may arise from a variety of complex causes, including
population pressure, changes in technology, climate, or political forces. Too
often, these causes have been confused, and the problem ascribed simply to
the "Tragedy of the Commons," in which the misuse of resources is attributed
to the institution of common property itself.

A fundamental issue in much of the developing world is the degree to
which resource mismanagement has actually been caused by common
property arrangements. In the Sahel and southern Africa, for example,
serious misuse of resources has been alleged to be the direct result of
traditional common property institutions (see Hitchcock 1981; Picardi and
Seifert 1976; Glantz 1977). In response, Western economic consultants and
planners have called for the imposition of private property rights (Johnson
1972; Picardi 1974). Similarly motivated private property schemes have been
attempted throughout the developing world. Many, perhaps most, have failed
to stop overuse, and in many cases may have contributed to even more rapid
degradation of resources and increased inequality in already unequal
distributions of wealth. Not unlike the European experience with enclosure,
lands formerly held in common are often transferred to individuals (such as
high-ranking government bureaucrats) who can exercise influence in the
allocation of use rights. These individuals have then failed to manage these
resources effectively. Despite this record, such policies are often supported by
those who argue on theoretical grounds that individual incentives inevitably
lead common property to be mismanaged. Modern economists often refer to
this as the "free rider" problem. When applied to resource management, the
free rider problem leads to the conclusion that common property is not a
viable institutional alternative.

Yet common property may be as viable as private property on grounds
of both efficiency and equity. Rather than representing an atavistic
arrangement of rights which inevitably results in inefficient resource use,
much value may lie in existing common property institutions, as well as in
new institutional arrangements with common property characteristics. In
many cases, these institutions are well adapted to the particular resource
constraints facing villages and groups in developing countries. In this sense
they relate to work on institutional constraints and innovation developed by
Hayami and Ruttan (1985).

In Ecuador and throughout Latin America, state claims of property
rights are unenforceable and essentially meaningless. Furthermore, by
undercutting legitimate systems of common property, such claims lead to a
system of open access, to which the only apparent solution is often argued to
be "privatization." The result is that the possible advantages of common
property management are lost in the shuffle.


To avoid open access problems of the type described in the preceding
section, the governments of Ecuador and other Latin American countries
routinely transfer natural resources to private parties. The legal tradition
governing this transfer dates to the early days of the colonial era, the first
European settlers in the Andes having been ceded "idle" lands (tierras baldias)
only when they proposed to use those lands for crop or livestock production.
This tradition is deeply embedded in formal as well as informal property lav."
IERAC and counterpart agencies in neighboring countries have required that
idle forests be converted into "productive" cropland or pasture before
recognizing private rights in a colonized parcel. At the same time,
agricultural use rights are the central feature of informal tenure regimes
throughout Latin America.

Vesting private tenure in those who convert tropical forests into
agricultural land serves powerful political and economic interests. The
establishment of "live frontiers" is viewed in many countries as a way to
strengthen a nation's territorial claims in the Amazon Basin and other remote
areas (Landau 1980). In addition, the migration of the rural poor to the
frontier tends to vent social pressures that can lead to political conflict
(Blaikie 1985). However, where land clearing is a prerequisite for property
rights, a cycle of excessive clearing and soil depletion tends to be set in

The institutional underpinnings of this cycle can be clarified with the
aid of a model, developed by Southgate (1990), that describes the opportunity
costs of inputs allocated to deforestation and soil conservation (ND and NC,
respectively) as well as the returns to those same inputs. Opportunity costs, W,
are an increasing function of the sum of ND and NC:

W = W [ND + NC] W' > 0. (1)

The returns to land clearing, RD, comprise the present value of crops grown
on newly deforested land less the discounted opportunity cost of agricultural
inputs employed on that same land. In general, those returns are an
increasing and concave function of the extent of deforestation, which can be
represented as ND divided by the inputs needed to clear a unit of land, d:

RD = RD [ND/d] RD'> O RD" <0. (2)

The returns to erosion control, RC, comprise two parts. The first is the present
value of additional crop production that comes about because a higher level of
soil quality is maintained. The second part is the present value of any
persisting reduction in crop production costs. Each element of RC is an
increasing and concave function of the area where erosion control measures
are being applied, which can be represented as NC divided by the inputs
needed to control erosion from a unit of land, c:

RC= RC [Nc/d] RC'>0 RC" < 0. (3)

Since agricultural colonists are preempted from capturing non-
agricultural rents (e.g., the net returns to forest management), they try to
maximize the present value of additional crop production associated both with
soil conservation and land clearing less the opportunity costs of those two
activities. This is accomplished by satisfying two first order conditions:

W' = RC'/c and (4)

W' = RD'/d. (5)

Equation (4) indicates that a settler should increase inputs allocated to erosion
control, NC, up to the point where NC's marginal opportunity cost, W', equals
the marginal returns of those same inputs: RC'/c. A similar rule, expressed in
equation (5), governs a colonist's decisions regarding ND.

The inefficiency of the second guideline, equation (5), is obvious. When
the tenure regime prevents individuals from capturing non-agricultural
rents, agriculture's extensive margin is found where agricultural rents equal
zero. By contrast, the agricultural frontier is found where agricultural rents
equal non-agricultural rents whenever settlers are in a position to take into
account the returns and costs associated with non-agricultural land uses.

As settlers respond to institutional incentives by allocating too many
inputs to land clearing, a second inefficiency arises. With ND set too high,
inputs' current scarcity value rises. This, in turn, discourages erosion control.
The linkage between NC and ND is appreciated by referring to a four quadrant
diagram (Figure 1), the northeastern and southwestern quadrants of which
show RC'/c and RD"/d, respectively. Also indicated in the southwestern
quadrant is the difference between RD'/d and the marginal rental value of
tree-covered land, C/d. The marginal opportunity cost of inputs currently
allocated to soil conservation and land clearing, W', is shown in the
northwestern quadrant and the sum of NC* and ND* is represented in the
southeastern quadrant. Note that, if the tenure regime were to change so that
farmers could internalize the marginal rental value of tree-covered land, not

only would inputs allocated to deforestation decline, from ND* to ND*', but the
associated decrease in wages would induce an increase in conservation effort,
from NC* to NC*'.

The cycle of excessive deforestation and insufficient erosion control
that is set in motion by a frontier tenure regime is not necessarily decelerated
by a change in non-institutional incentives. Lower interest rates or higher
commodity prices, for example, enhance both the present value of crops
grown on deforested land and the present value of additional crop production
associated with erosion control. These impacts are represented by outward
displacement of the two functions, RC'/c and RD'/d, shown in the northeastern
and southwestern quadrants, respectively, of Figure 1. Responding to this
shift in incentives, settlers allocate more labor to soil conservation as well as

Whether NC* or ND* rise or fall in response to increased timber prices
depends entirely on the tenure regime. If non-agricultural rents (value C in
the model) can be captured, then the price increase discourages land clearing,
ND. As a result, the current opportunity cost of labor falls, which in turn
causes NC to rise. Under a frontier property regime, by contrast, settlers
selling logs removed from land to be used for agricultural production treat
timber values as a negative argument of land clearing costs. Consequently, an
increase in those values enhances RD/d, both absolutely and relative to RC/c.
As illustrated in Figure 2, this accelerates deforestation and discourages soil


Legal traditions governing the transfer of tierras baldias to private
parties are not the only way that property rights are attenuated in Latin
America. Bureaucratically induced tenure insecurity is also chronic
throughout the region. Furthermore, just as a cycle of excessive deforestation
and insufficient erosion control arises where agricultural use rights are a
major feature of the institutional order, insecure property rights contribute to
depletive human interaction with the natural environment.

In part, tenure insecurity discourages resource conservation by
reducing the chances that current land users will capture the long term
benefits of resource conservation. A typical situation arises when a farmer
has doubts about when or if his application for formal tenure will be accepted.
At least for the time being, that farmer will not adopt erosion control
measures, for example, that chance crop yields only after the passage of
several years.

Insecure property rights also contribute to resource degradation by
impeding access to formal credit. In most Latin American countries, public
sector development banks loan money only to those farmers whose land has
been adjudicated. Private banks, of course, do not accept non-judicated land as

collateral. Denied access to formal credit, farmers without clear title to their
land must rely on informal credit markets, in which interest rates are
considerably higher. This discourages the adoption of conservation measures
(e.g., a switch to agroforestry) that carry short term costs.

The results of empirical research carried out in a number of developing
countries bolster the argument that insecure property rights influence
decisions regarding the use and management of natural resources. Feder e tal.
(1988) have found a clear statistical relationship between the strength of
farmers' formal property rights and their willingness to invest in land
improvements. In Latin America, Southgate eLt.a. (forthcoming) have
identified a statistically significant linkage between deforestation and tenure
insecurity in eastern Ecuador's Amazonian lowlands. The latter finding bears
out the claim that agricultural colonists safeguard their tenuous legal claims
on land by using it continuously for crop and livestock production (Rudel

The problem of insecure property rights is chronic for Latin America's
small farmers. That same group also tends to be concentrated on
environmentally fragile lands. Accordingly, enhancing tenure security is an
essential element of strategies to conserve the region's renewable natural


Tenure insecurity is not a problem oniy for recent migrants to the
agricultural frontier. The property rights of indigenous forest dwellers are
also in jeopardy. As a result, the latter group is encouraged to participate in
the destruction of the habitat from which it has traditionally drawn

The assault on forest dwellers' tenure is often direct. The creation of
parks and military zones and other forms of resource nationalization renders
irrelevant the structure of rights and duties previously developed by the local
community. Similarly, recognizing private land claims while ignoring
communal claims, as several Latin American governments have done from
time to time, assures the demise of common property, which is the
predominant form of tenure in Latin America's tropical forests. Because it has
tended to obfuscate the distinction between open access resources and common
properties, the economic literature addressing the tragedy of the commons has
legitimized this policy approach.

More subtle forms of pressure are often applied against forest dwellers'
communal property arrangements. In many countries, registering a
communal claim requires more time, money, or legal expertise than
registering an individual claim. This is an important drawback for indigenous
groups, which have limited financial means as well as restricted access to legal
services. In addition, when governments state that land uses characteristic of
communal tenure regimes are "non-tenurable," those regimes tend to break

down. From an individual's standpoint, for example, the net benefits of
observing fallowing norms are seriously diminished by laws, such as those
that exist in much of Latin America, that make land "improvement" a
prerequisite for formal tenure. Because improvement has, in practice, been
equated with deforestation, forsaking encroachment on a fallow parcel carries
the risk that someone else will assert an individual claim on that same parcel.
Anthropological case studies, like the one carried out by Macdonald (1981),
show that indigenous forest dwellers respond to this risk by forsaking
traditional common property arrangements and becoming agents of

A variant of a model first developed by Schelling (1973) can be used to
understand this response. Described in that model are the benefits for an
individual of cooperating in a collective resource management scheme as well
as the private benefits of defecting from the scheme. The former equal the
individual's share of total net returns of the scheme captured by the
cooperating coalition. Generally, the private benefits of cooperating increase
as the size of the coalition increases. The benefits for an individual of
defecting, which are also a positive function of the number of agents who join
the cooperating coalition, consist of part or all of the benefits of the collective
scheme not captured by the coalition along with other net returns of
individual action.

If the private benefits of defecting exceed the private benefits of
cooperating regardless of the size of the cooperating coalition (i.e., if defection
is a universally dominant strategy), the game is a multi-person prisoner's
dilemma. In such a game, individuals must be coerced into adhering to a
mutually beneficial collective arrangement. Another possibility is depicted in
Figure 3, which describes the benefits to one individual of cooperating or
defecting in a strategically interdependent game with n + 1 players. If a
"minimum coalition" of n* or more players is assembled in that game, then
other individuals will freely choose to cooperate as well.

Setting deforestation as a prerequisite for property rights reduces the
private benefits of cooperating in traditional collective resource management
schemes. Within the context of the model, this event is represented by
downward displacement of the curve relating private benefits of cooperation
to the size of the cooperating coalition (see Figure 4). If those benefits fall,
stronger forms of coercion must be used to make individuals observe group
rules. Alternatively, it becomes more difficult to raise within a group the
minimum coalition needed to bring about universal voluntary agreement in
collective schemes. As illustrated in Figure 4, a decline in the private benefits
of cooperating results in an increase (from n* to n**) in the size of the
minimum coalition needed to effect a collective management scheme.


Institutional reform is always a politically charged undertaking.
Ideologues of the right, who have supreme confidence in the workings of the
marketplace, argue that all natural resources should be divided among private
holdings, the owners of which can be expected to develop their properties
efficiently. They distrust any deviation from a perfectly comprehensive
regime of private tenure, expecting that a tragedy of the commons will arise
wherever a resource is not privately owned. At the same time, ideologues of
the left doubt that market exchange of private interests in natural resources
can ever result in their being used wisely. Only government, they reason, is
capable of developing resources efficiently.

Neither perspective should be neglected. Most societies are very
comfortable with the idea of dividing agricultural land, for example, among
private holdings. Provided nonpoint pollution associated with agricultural
production does not result in major downstream costs, there is no strong
reason for government to interfere in private decisions regarding the use,
management, or exchange of agricultural holdings. By contrast, some
resources (e.g., the air we breathe) cannot be divided among private holdings.
Government must take primary responsibility for the conservation of such

Of course, the dogmatic right is loathe to acknowledge instances in
which tenurial arrangements favored by the left are suitable. Similarly, it is
difficult to convince the rigid left that decisions regarding the use and
management of many resources are best left to individual property owners
heeding price signals generated in unregulated, competitive market,
Furthermore, both extremes share an ideological blindspot. As Hayami (1988)
points out, neither the left nor the right has been prepared to admit the value
of "intermediate" tenurial arrangements: the institutions communities around
the world have long used to deal with "local externalities."

Local externalities are a universal feature of agriculture and natural
resource development. For example, one farmer's water use is bound to have a
direct effect on the welfare of his neighbors just as his welfare is greatly
affected by their water use. Economists are only now beginning to recognize
that game theory and other models can be used to explain why an individual
agent facing such a situation finds that his or her personal welfare is
enhanced by voluntarily cooperating in collective institutional arrangements
(e.g., a village level water rationing scheme) developed to address local
externality problems (Schelling 1973; Axelrod 1984).

As a consequence of growing interest in such arrangements, however,
the "menu" of tenurial solutions to third world environmental problems is
being expanded. It has always included the policy prescriptions of the right
(i.e., strengthening or establishing private property rights) as well as those of
the left (i.e., increasing public sector control of resources). Intermediate
approaches (i.e., reinforcing the institutions communities have long used to
resolve local externality problems) are now generally accepted as being
worthy of consideration as well.

More than anything else, applying the menu of tenurial solutions to
resource degradation problems requires hard-headed economic objectivity.
That is, all costs and advantages of different tenurial approaches to any
particular environmental issue must be carefully assessed. For example,
before deciding to draw on the strengths of a regime of private property
rights, the costs of establishing and administering such a regime, which can
be considerable (Runge 1986), need to be investigated. Similarly, heavy
reliance on community-level arrangements is a suitable approach to
environmental policy only when local externalities are truly important.
Finally, even when the impacts of resource degradation are broadly
distributed, government action is called for only if expected improvements in
environmental quality compare favorably to the costs of that action.


Axelrod, R. 1984. The Evolution of Cooperation. New York: Basic

Binswanger, H. 1989. "Brazilian Policies that Encourage Deforestation
in the Amazon." World Bank, Environment Department Working
Paper No. 16.

Blaikie, P. 1985. The Political Economy of Soil Erosion in Developing
Countries. London: Longman.

Bromley, D. and M. Cernea. 1989. "The management of Common
Property Natural Resources: Some Conceptual and Operational
Fallacies." World Bank, Discussion Paper No. 57.

Ciriacy-Wantrup, S. and R. Bishop. 1975. "Common Property as a
Natural Resource Policy." Natural Resources Journal 15: 713-27.

de Soto, H. 1989. The Other Path. New York: Harper and Row.

Detwiler, R. and C. Hall. 1988. "Tropical Forests and the Global
Carbon Cycle." Science 239: 42-47.

Direccion Nacional Forestal (DINAF). 1988. Plan de Accion Forestal
para el Ecuador: Diagnostico del Sector Forestal (draft). Quito:
Ministerio de Agricultura y Ganaderia (MAG).

Feder, G. T. Onchan, Y. Chalamwong, and C. Hongladarom. 1988.
Land Policies and Farm Productivity in Thailand. Baltimore:
Johns Hopkins University Press.

Glantz, M. (ed.). 1977. Desertification: Environmental Degradation
in and around Arid Lands. Boulder: Westview Press.

Hardin, G. 1968. "The Tragedy of the Commons." Science 168:

Hayami, Y. 1988. "Community, Market, and State" (Elmhirst
Memorial Lecture), XX International Conference of
Agricultural Economists, Buenos Aries.

Hayami, Y. and V. Ruttan. 1985. Agricultural Development: A
Global Perspective. Baltimore: Johns Hopkins University

Hitchcock, R. 1981. "Traditional Systems of Land Tenure and
Agrarian Reform in Botswana." Journal of African Law 24.

Johnson, O. 1972. "Economic Analysis, the Legal Framework, and
Land Tenure Systems." Journal of Law and Economics
15: 259-76.

Landau, G. 1980. "The Treaty for Amazonian Cooperation: A Bold
New Instrument for Development." Georgia Journal of
International and Comparative Law 10(3): 463-89.

Lanly, J. 1982. "Tropical Forest Resources" (Forestry Paper No. 30),
U.N. Food and Agriculture Organization.

Macdonald, T. 1981. "Indigenous Responses to an Expanding Frontier:
Jungle Quichua Economic Conversion to Cattle Ranching." In
N. Whitten, ed., Cultural Transformation and Ethnicity in Modern
Ecuador. Urbana: University of Illinois Press.

Mahar, D. 1989. "Government Policies and Deforestation in Brazil's
Amazon Region." World Bank.

Ministerio de Agricultura y Ganaderia (MAG). Informe Final de la
Limitacion del Patrimonio Forestal del Estado. Quito.

Picardi, A. 1974. "A Systems Analysis of Pastoralism in the West
African Sahel." In Framework for Evaluating Long Term
Strategies for the Development of the Sahel-Sudan Region,
Annex 5. Cambridge: Massachusetts Institute of
Technology Center for Policy Alternatives.

Picardi, A. and W. Siefert. 1976. "A Tragedy of the Commons in the
Sahel." Technology Review 78: 42-51.

Rudel, T. 1983. "Roads, Speculators, and Colonization in the Ecuadorian
Amazon." Human Ecology 11(4): 385-403.

Runge, C. 1986. "Common Property and Collective Action in
Economic Development." World Development 14: 623-35.

Schelling, T. 1973. "Hockey Helmets, Concealed Weapons, and
Daylight Savings: A Study of Binary Choices with
Externalities." Journal of Conflict Resolution 17: 381-428.

Seligson, M. 1984. "Land Tenure Security, Minifundization, and
Agrarian Development in Ecuador." Quito: U.S. Agency for
International Development.

Southgate, D. 1990. "The Causes of Land Degradation along
'Spontaneously' Expanding Agricultural Frontiers in the
Third World." Land Economics 66(1): 93-101.

Southgate, D., R. Sierra, and L. Brown. Forthcoming. "A
Statistical Analysis of the Causes of Deforestation in
Eastern Ecuador." World Development.

Wilson, E.. (ed.). 1988. Biodiversity. Washington: National
Academy Press.

DEPLETION, 1967-1985




This paper examines the influence of international and domestic
macroeconomic factors on the depletion of tropical forests from 1967 to
1985. Data from 45 tropical developing countries were used to evaluate the
impacts of export prices, income, international debt and exchange rate
adjustments on forest depletion. The effects of population, arable land
availability and food self-sufficiency were also evaluated.

The results of this study call attention to the important and often ignored
linkage between such macroeconomic variables as real exchange rate
movements and environmental degradation in many developing countries. The
results suggest that if international organizations are interested in preserving
whatever little is left of the world's tropical forests, they have to reexamine
their policies towards developing countries particularly with respect to
structural adjustment.

Analysis is focused on closed broadleaved forests which compose 97% of
closed tropical forests and are the main sources of commercial tropical timber.1
Since data on deforestation are not available for all years of the study period,
forest disturbance from industrial logging was used as the indicator of
deforestation. This is justified by the close correlation and the typically
sequential relationship between logging and deforestation.

This paper has four sections. The rest of the first section discusses
tropical forest depletion and the role of government policy in encouraging this
process. Section II briefly reviews global economic developments from 1967 to
1985. Section III empirically examines the impact of these global economic
changes on the depletion of closed tropical broadleaved forests. The paper
concludes with a few brief comments.


There are different degrees of forest depletion depending on extent of
removal of tree cover. Forest depletion can be either qualitative or
quantitative2. The term "deforestation" pertains to quantitative forest

Qualitative depletion involves impoverishment of the forest through
marginal or pronounced disturbance without complete removal of forest cover.
Selective logging of commercially valuable wood species is an example of an
activity resulting in qualitative forest depletion. Although there is no
reduction in size of the forested area, the forest is degraded and, in many cases,
its survival seriously compromised after logging operations.

Quantitative depletion or deforestation, on the other hand, involves the
complete and permanent removal of all forest cover. Deforestation occurs
when forests are cleared for slash and burn cultivation, pasture or export crop
production. Typically, deforestation occurs in steps. Logging and other forms
of forest disturbance such as mining and road building on forested lands
eventually lead to complete removal of forest cover.

Thus, the distinction between qualitative forest depletion and
deforestation is often nebulous, separated only by time. This is particularly
true where high demand for land prevents the regeneration and recovery of
the forest following logging or any other type of extractive activity in forested

The Deforestation Process

Early estimates of the annual rate of tropical deforestation ranged from
an average of 1% to 2% with some countries, such as the Cote d'Ivoire,
experiencing deforestation rate as high as 6% per year (Myers, 1984). In 1980,
it was projected that forests in developing countries would decline by an
average of 3% to 6% per year in some cases, faster in others (Barney, 1980).
Recent data from satellite imagery show that deforestation in the tropics is a
more serious problem than it was thought to be. Actual deforestation in Brazil
and in India in 1980, for example, were three and ten times those of early FAO
estimates, respectively (Repetto, 1990).

Commercial logging is often the first step to deforestation. On average,
only about 5 to 35 cubic meters of merchantable wood are extracted per hectare
of tropical closed broadleaved forests' IFAO/UNEP, 1981). These small
commercial volumes, however, are associated with disproportionate damage to
the forest due to careless use of equipment and inefficient logging practices.

Typically, at least half of the remaining stock, including immature trees
of commercial value and harvestable stocks of less desirable varieties are
damaged beyond recovery (Repetto, 1990). More importantly, logging
operations leave behind access roads and trails that make logged over forests
vulnerable to permanent colonization and settlement by landless peasants,
ranchers and speculators3.

The demands of a growing population for food, fuel and housing coupled
with institutions that provide unequal access to income and existing arable
lands have encouraged the clearing of tropical forests in many developing
countries. As land rent4 increases with demand, the poorest and least
productive agricultural producers are pushed into the forest frontier by their
inability to successfully compete for cultivable lands. Logging roads serve as
major arteries leading these would-be settlers and cultivators into the forest.

The fragility of tropical forest soils and their general unsuitability for
agriculture are soon manifested in rapid soil nutrient depletion and
productivity decline5. Labor surpluses exacerbate the situation by pushing
marginal productivity and frontier agricultural income even lower. In order to
maintain output level, additional areas of forest often have to be cleared and

cultivated. In the absence of alternative employment opportunities created by
domestic economic growth, soil nutrient depletion, declining productivity and
labor surpluses conspire to encourage further deforestation.

Tropical deforestation has serious ecological and economic
consequences. Species extinction leading to decreased biodiversity, soil erosion
leading to siltation and flooding, microclimate as well as global climate changes
have all been linked to large scale forest removal6. These ecological changes
translate into real economic costs in terms of lost lives, property and
productivity7 in affected areas worldwide.

Government Policy and Tropical Forest Depletion

Ever since Malthus, population has occupied center stage in discussions
of issues of natural resource depletion. While population remains an
important consideration in tropical deforestation, it is increasingly argued that
institutional factors and economic conditions, both domestic and international,
have just as much, if not more, influence on the fate of tropical forests.8

In recent years, attention has focused on the role of government policies
in tropical forest depletion. There is growing consensus among analysts that
logging and forest conversion to agriculture are largely the result of
government policies9.

About 90% of all tropical forests are public forests under government
control and protection (FAO/UNEP, 1981). Governments derive revenues from
economic activity in these forests through user fees, royalties and taxes and,
therefore, benefit from some degree of forest exploitation or conversion to
more profitable types of land use.

Government policy encouraging or condoning the logging and clearing
of public forests can be theoretically viewed in an optimization framework
using a generalization of Faustmann's model. In this model, the government
acts as the forest manager and maximizes the present value of expected
perpetual returns per unit of forestland (Appendix A). The area of forest the
government allows to be logged or converted to agriculture is consistent with
this expected present value maximization.

Unfortunately, most governments have been myopic and ineffective
forest managers. Governments have either expressly or tacitly approved
commercial logging which have often led to deforestation. With very few
exceptions, governments have not captured the full value of forest resource
rents. Instead, they have allowed most of the rent to flow to timber
concessionaires and speculators who are typically linked to foreign

Even more unfortunately, many governments tended to view forests as
obstacles to development and have directly or indirectly encouraged forest
clearing (Guess, 1981). Landlessness exacerbated by urban-biased development
policies has induced population movements1 I and has encouraged forest
destruction through "spontaneous" or government-supported colonization of

the forest frontier. Where strong political interests militate against the
implementation of land reform, governments have justified forest colonization
and resettlement programs as means of achieving a better balance between
population and land resources while, at the same time, addressing distributional
concerns12. Frontier colonization programs have also been used to secure
contested territory along borders with other countries.

Using land tenure and other incentives, many governments have
encouraged forest clearing for agricultural production, notably, of exportable
crops. Unfortunately, government-supported production activities on cleared
forestlands have generally proven to be not only environmentally disastrous
but also economically nonviable. Forest clearing for cattle ranching in Latin
America readily comes to mind13.

Government policies, however, are not created in a vacuum. Although
developing country governments must ultimately bear responsibility for their
policies, the constraints imposed by global economic conditions and the
pressures exerted by international development agencies have had important
influence in shaping these policies. Tropical deforestation in response to
government policies can be better understood in this global economic context.
Indeed, conditions in the international market appears to have provided much
of the impetus for tropical forest depletion during the study period.

Many tropical developing countries, particularly those in Southeast Asia,
have relied on foreign exchange earnings from their forest products exports
(Appendix Table 1). Some countries have encouraged cultivation of
agricultural crops on forestlands as a means of generating foreign exchange
through exports, or, of saving foreign exchange through import substitution.
Increasing wood and agricultural commodity prices in the international
market encourage forest depletion by raising the opportunity cost of keeping
timber unharvested and land under forest cover.

However, falling international commodity prices do not necessarily
afford protection to tropical forests. As the regression results in section III
suggest, acute foreign exchange shortages and deteriorating currency values
have led to the "mining" of tropical forest resources even during periods of
depressed commodity prices.


This section discusses developments in the global economy which have
encouraged the exploitation of tropical forests during the study period. The
discussion emphasizes the role played by the International Monetary Fund
(IMF) and how its policies affected developing economies of the tropical world.

Collapse of the Bretton Woods System

The period from 1967 to 1971 witnessed the collapse of two important
pillars of the IMF -- the system of agreed par values and the gold convertibility

of the US dollar. These happened with the Nov. 1967 devaluation of the pound
sterling and the Aug. 15, 1971 suspension of the US dollar's gold convertibility
(De Vries, 1986).

The Nixon administration's decision to suspend the dollar's gold
convertibility, in the face of what then seemed to be intolerable current
account deficits, effectively ended the system of fixed exchange rates
established by the Bretton Woods Conference in 1944. Currency adjustments
worldwide after the collapse of the system of fixed exchange rates ushered in
an era of greater volatility in international prices.

Grain Shortages

In 1971, 1972 and 1973, poor harvest in each of the world's major grain
producing areas coupled with the reduction of reserve stocks in the US led to
worldwide grain shortages and raised prices in the international market
(Paulino, 1986). Speculation on primary commodities sent commodity prices in
1972-73 to levels not experienced since the Korean War in the early 1950s (De
Vries, 1986). The high prices had devastating effects in poorer countries
which of necessity had to import grain to feed their peoples (Blaxter, 1986). It
also raised concerns about food security and self-sufficiency in the developing
world14. Many countries adopted policy measures aimed at increasing food
security by expanding domestic food production. In many of these countries,
increased production meant expansion of agricultural areas at the expense of

First Oil Crisis

In March 1973, the system of fixed exchange rate officially gave way to a
system of floating exchange rates. The US dollar was again devalued during
this time. Until then, the world economy was practically on a dollar standard.
The devaluation of the US dollar created problems for the world economy and
was an important factor in the first oil crisis of 1973.

Oil producing countries had been feuding for years with powerful
multinational oil companies over the sharing of oil revenues. When the dollar
was devalued in 1973, oil producing countries demanded an increase in the
posted price of oil on which was based the royalties paid by oil companies.
Their contract had been denominated in US dollars and did not protect oil
producing countries from losses caused by currency devaluations (Penrose,
1976). In the midst of these negotiations, the Arab-Israeli war broke out and
Arab oil producers decided to use oil as a political weapon15.

The price of Saudi marker crude increased from $3.01 to $5.12 per barrel
in October 1973 to $11.65 in January 1974. These sharp increases in the price of
oil translated into trade deficits for oil importing countries and surpluses for
oil exporters. It reduced real incomes and growth in oil importing countries
and fueled inflationary pressures worldwide (Lin, 1981; Park, 1976).

The higher oil prices encouraged energy conservation efforts the world
over. In industrialized countries, consumers switched to coal, natural gas and
other energy sources. In most developing countries, consumers increasingly
turned to wood-based fuels which already supplied the major portion of energy
needs (Pereira et al., 1987).

In 1974-1975, the world experienced stagflation, a new phenomenon
marked by both inflation and recession. Despite the recession, some
commodities, notably sugar, coffee and tea enjoyed booming prices. The
commodity price boom peaked in 1976 and lasted through 1977.

The IMF's New Role

With the collapse of the fixed exchange rate system, it became necessary
for the IMF to redefine its function. During the early 1970's, the IMF began to
assume the function of surveillance of exchange rate policies in order to
contain inflation and excess international liquidity (De Vries, 1986).

During this period also, the IMF sought to increase its capacity to lend to
developing countries experiencing temporary balance of payments deficits.
This new role of the IMF was not unrelated to the massive OPEC oil surpluses
that had to be recycled in the international financial markets (Park, 1976).

Following the first oil crisis, the IMF cushioned non-oil developing
countries from increasing oil import bills through financing from the IMF's
Oil Facility. These countries were thus spared the necessity of painful
structural adjustments in their economies.

Credit Expansion

Following the lead of the IMF and the World Bank, other multilateral and
private lending institutions also increased lending to developing countries
after the first oil crisis. However, even before the excess liquidity created by
the oil price hikes, commercial banks had begun to turn their attention to
developing countries (Darity and Horn, 1988). Multilateral lending institutions
were also already beginning to respond to the clamor for increased
development lending during the late 1960s (FAO, 1970). The oversupply of
funds caused by OPEC surpluses simply prompted these lending institutions to
aggressively seek clients among developing countries during the mid-1970s.

Most commercial loans were channeled to newly industrializing
countries, such as Brazil and Mexico, and to middle income developing
countries. These countries were considered better credit risks and, hence,
received disproportionate shares of the loans from private commercial banks.
Interest on these loans were based on the London Inter-Bank Offer Rate
(LIBOR). Low income countries, on the other hand, relied mostly on
concessional lending from official multilateral sources.

With increased competition among lending institutions, even countries
with questionable creditworthiness were extended loans. During this period of

credit expansion, there were strong elements of "loan pushing"16. On
hindsight, many authors put part of the blame for the present debt crisis on
the institutions' overwillingness to lend during this period17.

Loans were made available for the finance of development projects in
agriculture, forestry, energy, transport and other economic sectors. In all
sectors, there was a marked preference for large-scale projects such as
multipurpose dams and agro-industrial complexes (Payer, 1982).
Transmigration programs were also financed out of these development loans.
In some countries, part of the borrowed funds lined the pockets of a
"kleptocratic" elite18.

Many loan-financed development projects have come under fire for the
social and environmental disruption they have caused (SAM, 1987). In many
cases, the projects displaced communities most of which were resettled in
forested zones. Natural forests were cleared in some countries to establish
plantations19 of fast-growing trees for pulp and paper or for fuel for
dendrothermal plants (Anderson and Huber, 1988). Road construction and
transmigration projects also destroyed vast areas of the Amazon forest (Mahar,
1989) and of forests in other regions.

IMF Conditionality

In 1976, the IMF drastically changed its policy. Further lending was
made conditional on structural adjustments undertaken by a borrowing
country to accommodate the higher oil prices. Signs of recovery from the
1974-75 recession in industrial courries, favorable prices for primary
commodity exports, and, concern over mounting developing country debt
prompted the policy change (De Vries, 1987).

In 1977, the IMF decided to use exchange rate devaluations as the major
tool to facilitate the adjustment. Devaluations are believed to achieve
reductions in real income by increasing the price of tradable goods, thereby,
contributing to decreased expenditures. They are also believed to cause relative
price changes that make exports and import substitutes more profitable and
imports more expensive. The resulting realignment in production and
increase in export earnings are supposed to correct a country's trade and
payments imbalance.

The conditionality was based on the premise that many debtor countries
have failed to completely adjust to global economic conditions because of
excessive aggregate demand and because of an allocation of expenditure that is
biased against exports (Fishlow, 1985). Developing country debtors were asked
by the IMF to restrain aggregate demand, control inflation and devalue their
currency. Demand restraint typically involved cut-backs in public outlays,
elimination of subsidies, tax increases and increased prices for public services.

These measures were meant to reduce fiscal deficits that contribute to
aggregate demand and domestic credit creation. Nominal wages were contained
to control inflation. High interest rate policies were used to encourage
domestic savings. These prescriptions forced upon developing countries have