ArticlePDF Available

Brazil’s Samuel Dam: Lessons for Hydroelectric Development Policy and the Environment in Amazonia

Authors:

Abstract and Figures

Brazil's Samuel Dam, which formed a 540-km2 reservoir in the state of Rondônia in 1988, provides lessons for development decisions throughout Amazonia and in other tropical areas. The decision to build the dam was heavily influenced by its role in the political strategies of key decision makers. Samuel illustrates both impacts and benefits of electricity supply and the dilemmas facing decision makers regarding the various options for planned electricity generation. Environmental costs included flooding forest and stimulating illegal logging activity throughout western Amazonia because of an exception opened for Samuel in Brazil's prohibition of export of raw logs. Samuel emitted substantially more greenhouse gases than would have been emitted by generating the same amount of electricity from oil. Contamination of fish in the reservoir resulted from methylation of mercury present in the soil. Social costs of the dam included resettlement of 238 families of farmers; impacts on indigenous people were indirect. Mitigating measures included faunal rescue and creation of a forest reserve. The lessons of Samuel include the need to consider a full range of alternatives prior to making decisions in practice and the importance of adhering to the logical sequence of decision making, where information is gathered and compared prior to the decision. It also shows the need to maintain flexibility when the costs and benefits of different alternatives change significantly over the course of the project's planning and execution, as occurred at Samuel.
Content may be subject to copyright.
BRAZIL'S SAMUEL DAM: LESSONS FOR
HYDROELECTRIC DEVELOPMENT POLICY
AND THE ENVIRONMENT IN AMAZONIA
Running head: Brazil's Samuel Dam
Philip M. Fearnside
National Institute for Research
in the Amazon (INPA)
C.P. 478
69011-970 Manaus-Amazonas
BRAZIL
Fax: 55-92-642-8909
E-MAIL: PMFEARN@INPA.GOV.BR
Revised: 8 March 2004
Main text 8971 words
Environmental Management (Accepted: 14 April 2004)
1
ABSTRACT / Brazil’s Samuel Dam, which formed a 540-km2
reservoir in the state of Rondônia in 1988, provides lessons
for development decisions throughout Amazonia and in other
tropical areas. The decision to build the dam was heavily
influenced by its role in the political strategies of key
decision makers. Samuel illustrates both impacts and benefits
of electricity supply and the dilemmas facing decision makers
regarding the various options for planned electricity
generation. Environmental costs included flooding forest and
stimulating illegal logging activity throughout western
Amazonia because of an exception opened for Samuel in Brazil’s
prohibition of export of raw logs. Samuel emitted
substantially more greenhouse gases than would have been
emitted by generating the same amount of electricity from oil.
Contamination of fish in the reservoir resulted from
methylation of mercury present in the soil. Social costs of
the dam included resettlement of 238 families of farmers;
impacts on indigenous people were indirect. Mitigating
measures included faunal rescue and creation of a forest
reserve. The lessons of Samuel include the need to consider a
full range of alternatives prior to making decisions in
practice and the importance of adhering to the logical
sequence of decision making, where information is gathered and
compared prior to the decision. It also shows the need to
maintain flexibility when the costs and benefits of different
alternatives change significantly over the course of the
project’s planning and execution, as occurred at Samuel.
KEYWORDS:
Amazonia, Brazil, Dams, Hydroelectric dams, Reservoirs
Hydroelectric dam construction is one of the most
controversial activities affecting the path of development in
tropical countries and is a leading driver of environmental
and social problems. In Brazilian Amazonia (Figure 1), the
full list of 79 planned dams (regardless of the expected date
of construction) would flood approximately 3% of Brazil’s
Amazon forest directly (Brazil, ELETROBRÁS 1987, p. 150, see
Fearnside 1995). Decisions on future hydroelectric projects
unleash chains of events with impacts reaching far beyond the
immediate vicinity of the dams and reservoirs.
[Figure 1 here]
In May 2001 Brazil entered into an “energy crisis,”
beginning with uncontrolled blackouts in major cities such as
São Paulo and Rio de Janeiro, followed by a series of
emergency measures to reduce electricity consumption. The
“crisis” was a combined result of poor planning of electricity
generation infrastructure, inefficient domestic and industrial
use of electricity, government subsidy of energy-intensive
export products such as aluminum, and low rainfall in
hydroelectric catchments. Among the measures implemented is
an abbreviation of the environmental review process for new
hydroelectric dams and other energy-related infrastructure,
effective 18 May 2001 (see: Gazeta Mercantil 2001). This
2
bodes ill for avoiding environmental impacts as Brazil
proceeds with its ambitious dam-building program in Amazonia.
The new measures also add urgency to making maximal use of
the lessons that can be learned from past experience in the
region.
The Samuel Dam is located on the Jamari River at
Cachoeira de Samuel (8o 45’S, 63o 25’W), 52 km from the city of
Porto Velho, Rondônia (Figure 2). The reservoir was almost
all under tropical forest at the time it was flooded. The
15,280-km2 catchment (Brazil, ELETRONORTE nd [C. 1987]) is
relatively small, only 24 times larger than the area of the
reservoir itself. The streamflow of the Jamari River is
consequently limited, with an average annual flow of 366 m3/s
(Brazil, ANEEL 2003). This restricts the power produced to a
theoretical maximum of 76.0 MW of average generation if all
water were used under optimal conditions, considering use of
171 m3/s per turbine with a nominal capacity of 44.41 MW and a
power factor of 0.80 (Brazil, ELETRONORTE nd [C. 1987])]. The
dam has 216 MW of installed capacity.
[Figure 2 here]
A comparison of existing dams in Brazilian Amazonia is
instructive as an indication of the relative merits of Samuel
(Table 1). The power density (Watts of installed capacity per
m2 of reservoir surface) is a useful overall indicator of
environmental impact: the lower the density higher the impact.
Average residence time (days the average drop of water
remains in the reservoir) is related to water quality: the
longer the residence time the lower the water quality, with
low concentrations of oxygen and high concentrations of
methane. Shallow average depth is likewise an indication of
high impact.
[Table 1 here]
It should be remembered that dams in widely separated
locations (as in Table 1) are not competing options for the
role filled by Samuel as a source of power for Rondônia.
However, among proposed hydroelectric projects in Rondônia
Samuel compares poorly in terms of cost per kilowatt of
installed capacity, as well has having a low power density
(Table 2).
[Table 2 here]
While more hydroelectric options in distant locations
were not in direct competition with Samuel as a potential
solution to supplying power to Rondônia, indirectly they do
compete in two ways. First, the option of linking Rondônia to
the national power grid and supplying it from more distant
generating sites was entirely possible, even though the
distances involved represented a greater barrier to the
transmission technology that existed at the time of the
decision to build Samuel than would be the case today. The
second way that potential hydroelectric projects elsewhere
3
compete is by providing the option of investing funds in dams
with greater cost effectiveness and lower environmental
impacts, and continuing to supply electricity to Rondônia from
oil-burning thermoelectric plants. The situation facing
decision makers at the time Samuel was initiated was similar
to the case of the decision to build Balbina instead of (or in
addition to) the larger but more distant Cachoeira Porteira
Dam (Fearnside 1989a). In both cases, consideration was not
given to the option of using oil generation as a bridge to
supply power until a more attractive power source could be
developed.
The present paper examines the political context in which
the decision was made to construct the Samuel Dam, its
monetary, environmental and social costs and benefits, and
mitigating measures. The case of Samuel highlights the
existence of multiple impediments to decision making
proceeding in accord with the logical sequence of steps where
costs and benefits are estimated and compared prior to making
a decision on implanting the project and the decision is made
in the best interests of the area’s residents and their
descendents. Whether or not Samuel was a worthy project, the
decision-making process offers important lessons for pending
hydroelectric developments in Rondônia and in many locations
throughout Amazonia and the World.
Decision-Making
Political Context of the Decision
When construction began on Samuel in 1982, Brazil was
still ruled by a military dictatorship that restricted public
discussion of such subjects. However, an “abertura” (opening)
had been in progress since 1979 in preparation for an orderly
transition to democracy, and Rondônia figured prominently in
the plans of military leaders for achieving this transition
while maintaining their influence over the country’s
government. Rondônia was a federal territory traditionally
administered by the Army, while the other two Amazonian
territories were traditionally administered by the other
military branches (Roraima by the Air Force and Amapá by the
Navy). Jorge Teixeira, the military-appointed governor of
Rondônia, was an Army colonel, and was fully engaged in
preparing Rondônia for statehood. The World Bank’s
POLONOROESTE project was a key part of this strategy: paving
the BR-364 Highway and encouraging migration to Rondônia
provided political justification for the National Congress to
approve creation of a new state, while the almost complete
dependence of the recent arrivals on government largesse for
providing land, access roads, agricultural financing and other
services made the settlers likely to vote for candidates of
the political party supported by the military (the Social
Democratic Party: PDS). The agreement reached for granting
statehood in 1984 created the new state (thereby gaining three
seats in the Senate, as well as additional seats in the
Chamber of Deputies), while allowing the appointed governor to
remain in office for an additional four years without having
4
to stand for election (e.g., Isto É 1984).
At the time of Rondônia’s drive to achieve statehood the
minister of the interior was Mario Andreazza, whose ability to
implant massive public works had been amply demonstrated by
his role in building the Transamazon Highway in 1970 (see
Fearnside 1986a). Statehood for Rondônia was an important
goal for Andreazza, who hoped to be chosen as president of
Brazil through the indirect electoral collage that continued
to choose Brazilian presidents through 1984. Andreazza had
been preparing his candidacy through promotion of public works
ever since the 1960s, when, in his travels as minister of
transportation, he perceived the lasting popularity that
public works had brought to former president Jucelino
Kubitcheck (Branco 1984). Building the Samuel Dam, in addition
to the BR-364 Highway, was an essential part of this strategy.
The influence of key individuals on the decision process must
be recognized: as the head of the World Bank’s Latin America
and Caribbean division (Robert Skillings) remarked at the time
with respect to the Bank’s approval of POLONOROESTE, it was
“hard to say ‘no’ to Andreazza.”
Jorge Teixeira (the military-appointed governor of
Rondônia) was also a man whose personality influenced the
course of history in the region. His ability to get things
done was much appreciated at the World Bank, where he was
known as the only man who had ever convinced the Bank to
finance a cemetery (in this case when he was mayor of Manaus,
prior to being appointed governor of Rondônia). A former
volunteer paratrooper in the Vietnam War, his style was seen
as ideal for taming the social chaos of Rondônia, which has
often been compared to the 19th Century “wild west” of the
United States. His tireless promotion of the development of
Rondônia undoubtedly helped convince decision makers both in
Brasília and in the multilateral development banks to invest
in Rondônia, including the Samuel Dam, beyond what would be
justified solely on the basis of financial, social and
environmental costs and benefits.
Environmental Impact Assessment
The Samuel Dam was under construction before the
Environmental Impact Study (EIA) and Report of Impact on the
Environment (RIMA) became mandatory in Brazil on 23 January
1986. Public hearings, also instituted in 1986, were also not
required for projects already under construction.
Nevertheless, ELETRONORTE contracted a series of environmental
studies (to be discussed later). While the grandfather clause
exempting Samuel from the EIA and RIMA requirements was always
emphasized at the time, it is curious that, years later, the
websites of both ELETRONORTE (Brazil, ELETRONORTE nd [2003])
and the consulting firm responsible for the environmental
studies (Sondotecnica nd [2003]) refer to these studies as the
first EIA/RIMA for an ELETRONORTE dam.
One key aspect of the decision to build Samuel for which
no consideration is known to have been given is an assessment
5
of alternative means of energy supply, as well as any re-
assessment over the lengthy planning and construction process as
the available options changed. For example, should there have
been a transmission line from Cuiabá, connecting with the
national grid? The technology of long-distance power
transmission improved markedly in the years over which Samuel
was under construction (Pires and Vaccari 1986). Did the
existence of the Samuel project remove the impetus to build such
a line and provide a larger source of power to this part of
Amazonia? Another development over the period of construction
was the 1986 discovery of gas in Urucú, two years before the
Samuel reservoir began to fill. Current infrastructure plans
include both a gas pipeline to Urucú and a transmission line to
Cuiabá, as well as additional dams, meaning that the
environmental impacts of all of these projects may be provoked
in addition to those already caused by the Samuel Dam.
The Role of Research
The role of research at Samuel became a public issue in
1986 when an advisory report on the matter by Brazil’s
National Council for Scientific and Technological Development
(CNPq) was leaked to the press. The report, authored by Zeli
Kacowicz, accused Brazil’s National Institute for Research in
the Amazon (INPA) of producing “aseptic reports ... that don’t
even scratch the surface of predicting the environmental impacts
of the construction of [hydroelectric] plants..”, and concluded
that the uncritical reports were due to “the necessity of INPA
having to sign an agreement with ELETRONORTE in order to, from
the paltry resources passed to it by the company [ELETRONORTE],
have enough operating capital to pay its bills for electricity,
water and telephones” (Kacowicz 1985; Jornal do Comércio 1986a).
Herbert O. R. Schubart, INPA director at the time the report
was leaked (but not at the time the 1980 contract was signed for
INPA’s work in Samuel), while objecting to the report’s
“alarmist” tone, confirmed the basic facts of the report and
stated that “in truth, in a period of crisis, ELETRONORTE used
INPA’s name to protect itself from criticisms that were being
made by the community” (Jornal do Comércio 1986b).
The arrangement whereby INPA collected raw data, which
were then used as the basis of reports drafted by a consulting
firm, had the result of facilitating uncritical reports while
still giving ELETRONORTE the advantage of making use of INPA’s
name to bolster the project’s credibility. As at Balbina and
Tucuruí, confidentiality clauses in the contract allowed
ELETRONORTE to veto publication or public presentation of any
inconvenient results (Fearnside 1989a, 2001a). This
combination is a formula for the problems divulged in the
“Kacowitz Report” on the Samuel Dam research.
Despite improvements in the environmental impact
assessment system, the relevance of INPA’s experience at
Samuel to environmental studies of contemporary infrastructure
projects is evident. The impact studies for the Tocantins-
Araguaia Waterway (FADESP 1996a,b), carried out by the Federal
University of Pará (UFPa), have been the subject of ongoing
6
criticisms and legal contestation (Switkes 2002; see
Fearnside 2001b). The criticisms of UFPa’s reports and
financial dependency on consulting contracts are almost
identical those made in the “Kacowicz Report” on INPA’s work at
Samuel a decade earlier.
The Role of the World Bank
Before construction on Samuel began and almost a decade
before it was completed, Robert Goodland (1980), then head of
the World Bank’s miniscule (three-person) environment unit, had
singled out Samuel as an example of a dam with extraordinarily
high environmental impact relative to the power it would
generate. However, environmental matters carried relatively
little weight in World Bank decisions at the time, and Goodland
was the only professional ecologist on the Bank’s staff, in
contrast to some 3000 economists.
Samuel was closely associated with the World Bank-
financed POLONOROESTE project, known in Bank parlance as the
“Northwest Brazil Development Pole” (World Bank 1981). The
project reconstructed and paved the BR-364 Highway in 1982,
deliberately opening Rondônia to a flood of migrants from the
state of Paraná. This ranks as one of the World Bank’s greatest
environmental embarrassments, and led directly to the creation
of the Environment Department within the World Bank in May 1987
(Holden 1987), less than 48 hours after an exposé of the project
was aired on the 60-minutes television program in the United
States. POLONOROESTE caused a great increase in deforestation
and severe impacts on indigenous peoples (Fearnside 1986b,
1987a,b, 1989b, Schwartzman 1986). In announcing the creation
of the Environment Department, World Bank president Barber
Conable described POLONOROESTE as “a sobering example of an
environmentally sound effort which went wrong” (Holden 1987).
While Samuel was under construction, World Bank guilt for
POLONOROESTE was apparent, and the PLANAFLORO project was
financed in an attempt to undo some of the damage from the
earlier loan. Samuel would not have been needed were it not for
the flood of migrants brought by POLONOROESTE, leading to
anguished discussions at the Bank over Samuel and its impacts.
All of the state of Rondônia is considered to be in the area of
influence of POLONOROESTE.
Although Samuel was not financed as a separate World-Bank
“project,” in mid-1986 the World Bank approved a US$500 million
“sector loan” to supply imported equipment for the entire
electrical power sector of Brazil (e.g., Schwartzman and Melone
1987). Unlike “project loans,” individual projects within the
sector are not subject to environmental review in the case of
sector loans, thereby allowing funds from the World Bank to be
used at Samuel.
Context of Energy Development in Rondônia
Electricity Demand
7
Electricity is fundamental to modern life, both for
residential use and for most activities that provide
employment. Urbanization is intimately tied to electricity,
with urban areas providing much more universal access to
electricity to residents and attracting more electricity-
demanding commercial and industrial users. The availability
of residential electricity, and the possibility of the
employment in urban economic activities that depend on
electricity, are two of the primary reasons for movement of
population from rural to urban areas. In the 1970s and 1980s,
Rondônia was one of the most rapidly urbanizing areas in
Brazil; the population of Porto Velho grew at 7.64%/year from
1970 to 1991, more than quadrupling over this period (Browder
and Godfrey 1997, p. 127). Electricity use in Rondônia was
growing explosively prior to the decision to build Samuel,
having grown form 5.8 GWh in 1970 to 145 GWh in 1980 (Machado
and Souza 2003, p. 218). The precarious electricity supply
from diesel generators was recognized as a fundamental
limitation on Rondônia’s development (World Bank 1981).
Irregular electricity supply was one of the most frequent
complaints of sawmill owners at the time (personal
observation). A succession of industries in Rondônia has been
implanted to process forest and agricultural products in
Rondônia, at least as long as the productive resources last.
Timber was a major product in frontier areas throughout
Rondônia (Browder 1986), although sawmills abandon each
successive area as the supply of valuable wood is exhausted.
Cassiterite (tin) mining was important in the 1980s when the
price of tin was much higher than it is today. In some of the
already settled areas, milk production has become an important
industry that depends on local processing (Faminow 1998).
Soybean farming, a land use that is currently expanding
rapidly, is still primarily dependent on processing outside of
the state (Fearnside 2001b).
The social benefits of Samuel are significant, in that
the power is all consumed locally (Browder and Godfrey 1997,
pp. 326-329). This contrasts with dams such as Tucuruí, where
most of the electricity generated is used by multinational
aluminum companies. The industrial activities in Rondônia
have also been relatively energy sparing, at least when
compared with intensive uses like aluminum smelting. The
social context of energy development in Rondônia may change
radically in the near future if plans go forward to turn the
state into a major exporter of energy to the rest of Brazil.
Planned Electricity Generation
Jí-Paraná River dams
The Samuel Dam’s small generating capacity made the need
for other sources of electric power obvious from the inception
of the planning process for Samuel. Plans were laid for the
Ji-Paraná Dam on Rondônia's Ji-Paraná (Machado) River at one
of three sites selected for dams on that river. The Ji-Paraná
Dam would create a 957 km2 reservoir (Brazil, ELETRONORTE
8
1987), and would flood 107 km2 (6%) of the Lourdes Indigenous
Reserve of the Gavião and Arara tribes, plus 37.7 km2 (1.4%) of
the Jaru Biological Reserve (Brazil, ELETROBRÁS 1986, p.
6.23). Some of the earlier plans had called for flooding as
much as 60% of this reserve (Brazil, ELETRONORTE 1987). Because
the World Bank financed these reserves under the POLONOROESTE
program, lending funds to finance their flooding under a
proposed loan for building the Ji-Paraná Dam was described as
“pure folly” by the US executive director of the World Bank in
an unsuccessful attempt to block approval of the first
Brazilian power sector loan in 1986 (Foster 1986).
Preparations for the Ji-Paraná Dam were halted in 1989,
supposedly because generation of power from natural gas was
just about to begin. It now appears unlikely that the Ji-
Paraná Dam will be built because the real estate cost would be
too high, given that the land that would be flooded is almost
all in a settlement area. However, reactivation of
preparations for building this dam has recently been
recommended in a report sponsored by the World Bank-funded
PLANAFLORO project (Bartholo Jr. and Bursztyn 1999, pp. 160-
164). This report also recommends resuming the viability
studies for the other two dams identified on the Jí-Paraná
River.
Madeira River dams
The Madeira River, although only a tributary to the Amazon,
is one of the World’s great rivers, with a water flow equal to
that of the Yangzi River in China. In the 320-km stretch between
Guajará-Mirim and Porto Velho the river drops 60 m in elevation,
with an average flow of 20,000 m3/s. This creates the
opportunity for large hydroelectric dams, despite the problem
posed by the Madeira River’s extraordinarily high sediment
loads.
A pre-inventory report was completed for a dam at either
Cachoeira Teotônio or the adjacent Cachoeira Santo Antônio
(the preferred site is now Cachoeira Santo Antônio), 25 km
southwest of Porto Velho. Several plans were made, ranging
from 1000 to 8000 MW. The dam foreseen in the 2010 Plan would
have 3800 MW (Brazil, ELETROBRÁS 1987). The larger designs
include flooding into Bolivia, while the smaller ones only
flood in Brazil.
Dams on the Madeira River were seldom mentioned until
1997 and 1998, when the “Úmidas” plan was prepared in with
support from the PLANAFLORO project, to suggest directions for
Rondônia’s development through 2020. Embedded in a lengthy
discussion of sustainable development, the plan’s most concrete
proposal was to make Rondônia into an exporter of electricity to
central-south Brazil (Bartholo Jr. and Bursztyn 1999, pp. 160-
164). A key part of this would be to accelerate work on
determining the feasibility of dams on the Madeira River,
especially the Santo Antônio Dam [8o, 48’ 52.4” S, 63o 53’ 41.3”
S]. Other recommendations were to expand the planned gas-
powered thermal plant in Porto Velho using natural gas from
Urucú, resume studies for hydroelectric dams on the Ji-Paraná
9
River, with a suggested division of the fall into more dams
than the currently planned three (with a total of 1295 MW of
installed capacity), and tap the 241 MW of inventoried potential
for small hydroelectric dams represented by 64 such dams in
Rondônia for which preliminary studies were done by
ELETROBRÁS/ELETRONORTE. In addition to exporting power to São
Paulo, the plan suggests attracting industries to Rondônia. The
Úmidas project is endorsed by the Rondônia state government and
by parts of the federal government. It is waiting for a
financer.
During Brazil’s “big blackout” (apagão) in 2001, with
electricity rationing in effect in most of the country, plans
for dams on the Madeira River suddenly became prominent in the
National Council for Energy Policy (CNPE). Plans were
considered for a 6300 MW configuration for the Santo Antônio
Dam and for the 4200 MW Jirau Dam [9o 15’ 47.9” S, 64o 43’
52.4” W] further upstream (Corrêa 2001).
In March 2003 the government announced plans by Furnas
Centrais Elétricas and the Odebrecht construction firm for
dams on the Madeira River (Jornal do Brasil 2003). The
installed capacities of the dams were revised downward to 3580
MW for the Santo Antônio Dam (of which 2185 would be firm
power) and 3900 MW for the Jirau Dam (of which 2285 would be
firm power); construction would begin in June 2005 and
generation would begin in 2009 and reach completion in 2012
(Machado 2003). Reservoirs would be relatively small: 138 and
110 km2, respectively (Machado 2003). The cost would be US$4
billion, not counting a transmission line linking the dams to
the national grid (Quintella 2003). An additional dam in
Bolivia (presumably the planned 1500 MW Esperanza Dam on the
Beni River [10o 35’11.5” S, 65o 35’ 53.4” W]) would be needed
to bring installed capacity to 11,000 MW (Monteiro 2003).
Flooding the rapids on the Madeira River and opening the
stretch to barge traffic, expected to carry 50 million tons of
soybeans annually, is a major argument for the dams (Machado
2003). Facilitation of soybean transport implies forest
losses in both Brazil and northern Bolivia (Fearnside 2001b).
A key attraction of the plan is also that it circumvents the
regulatory barriers that currently impede two other energy
projects due to judicial orders requiring substantial
improvements in the environmental impact assessments: the
Urucú-Porto Velho gas pipeline (A Crítica 2003) and the Belo
Monte Dam on the Xingu River (Pinto 2002). The Madeira dams
would both provide an alternative to the pipeline for
supplying Rondônia with electricity and contribute
hydroelectric energy to the national grid on a scale that
planners had expected to obtain quickly from Belo Monte.
Gas pipeline
Generation of electricity with gas from Urucu, located 500
km NW of Porto Velho, is a top priority under the Avança Brasil
program (Consórcio Brasiliana 1998). This program includes the
2000-2003 Pluriannual Plan, in addition to an indicative
10
planning horizon to the end of 2007. Avança Brasil planned
for investment of US$43 billion in Amazonia, of which US$20
billion would be for infrastructure with direct environmental
impacts (see: Carvalho and others 2001, Fearnside 2002a,
Laurance and others 2001). The present government’s 2004-2007
Pluriannual Plan includes the same projects announced under
Avança Brasil. The pipeline is likely to result in substantial
increases in deforestation, since migrants could be expected
follow the access roads (Laurance and others 2001, Fearnside
2002a, Gawora 1998). The gas pipeline would provide access to
the last large block of undisturbed forest in Brazil’s Amazon
Region, as deforestation has so far been almost entirely
excluded from the portion of the State of Amazonas west of the
Madeira River.
Costs of Samuel
Direct Monetary Costs
The Samuel Dam was budgeted at US$835.97 million (Lobato
1993). Construction began in March 1982, and ELETRONORTE
expected to have all 5 turbines installed by 1990. Successive
delays due to budget restrictions undoubtedly increased the
actual costs. The first turbine was installed on 24 July 1989
and the last on 2 August 1996.
A variety of problems emerged during the construction
process, requiring additional expenses. One was the appearance
of “canalicos,” or small channels or cavities formed by
termites in the earth under the dam (Jury 1989). The problem
was solved by construction of an upstream blanket of concrete
to lengthen the percolation path (Cadman 1989). The problem of
“canalicos” also occurred at Tucuruí.
No figures have ever been released for the final cost of
the dam and its transmission lines. Under the optimistic
assumption that the dam cost the originally budgeted US$835.97
million, cost was US$3870 per kilowatt installed or
approximately the same as at Balbina, which is also on a small
river in a flat region inappropriate for hydroelectric
development (Fearnside 1989a). Considering a round figure of
US$1 billion for the construction cost at Balbina, that dam cost
was US$4000 per kilowatt of installed capacity. For comparison,
Tucuruí cost US$675/kilowatt and Itaipú US$1206/kilowatt (Veja
1987, p. 30).
As at Balbina and Tucuruí, a special steel was used in the
turbines, adding to construction costs but avoiding expensive
repairs of corrosion that the acid water causes with more
commonly used steels. The special steel was used because the
Curuá-Una Dam had suffered major repairs and lost generating
time because of corrosion of the turbines (Brazil,
ELETROBRÁS/CEPEL 1983). Samuel has had no problems with
corrosion of turbines.
Opportunity Costs
11
The cost of a dam is usually stated in terms of the
money spent to build the infrastructure and carry out
necessary preparatory tasks such as viability studies and
resettlement. However, the opportunity cost of sacrificing
the land use that would have occupied the site had it not been
used for a reservoir should also be part of the decision when
a dam-building project is initiated. In this case, the area
was covered by tropical forest, which has substantial
unrewarded value for its environmental services, in addition
to its value as a source of material products (Fearnside
1999b, 2000). However, Samuel was located in one of the areas
with the highest rates of deforestation anywhere on Earth; at
the time construction began in 1982 the population of Rondônia
was growing exponentially at 16% per year and deforested areas
were expanding at over 29% per year (both values refer to the
1975-1985 period; see Fearnside 1989b, p. 8), corresponding to
doubling times of only 4.3 and 2.4 years, respectively. Today,
had the area not been used for a reservoir, it would probably
be a landscape dominated by degraded cattle pasture, as in the
neighboring settlement areas. The opportunity cost of forest
loss could therefore be considered to be much less at Samuel
than in cases like Balbina, where forest was flooded in an
area that would almost certainly not have been cleared in the
absence of the dam. However, the migrants who would have
settled in the Samuel area probably found land elsewhere in
Rondônia or in other Amazonian states, where they probably
cleared approximately the same amount of forest. Viewed in
this way, the full area of forest flooded by Samuel must be
considered as a cost.
An opportunity cost of money also applies to decisions of
this type. Samuel was an expensive means of supplying energy
to Rondônia, and encouraging population migration to Rondônia
was an expensive choice as a response to the social problems
caused by mechanization of agriculture and consolidation of
land holdings in Paraná (Fearnside 1986b, 1987a). Money might
have been used in other ways, creating greater social
benefits. In addition to monetary expenses, the environmental
cost of using Rondônia as a safety valve for the land-tenure
problems of Paraná was also tremendous.
The inefficient use of money contributes indirectly to
one of Amazonia’s greatest problems: stimulation of
deforestation for land speculation. During the 1980s while
Samuel was being built, investors speculated massively in land
purchases as a means of turning Brazil's astronomical
inflation to their advantage (a motivation for deforestation
that decreased in relative importance from 1994 onwards after
the Plano Real economic package slowed the rate of inflation).
This inflation was, in part, fueled by ill-conceived projects
that injected cash into the economy without producing a
corresponding flow of products for consumers to buy with the
money. Examples include inefficient dams and marginally-
productive ranches in Amazonia. Speculators deforest in the
land they buy as a means of protecting it from loss to
invading squatters or to government expropriation for agrarian
reform; they plant cattle pasture which, despite its low
12
productivity, is the cheapest means of occupying a large
area (Fearnside 1993).
Environmental Costs
Forest Loss
The Samuel Dam has 0.40 Watts of installed capacity per m2
of reservoir area (Table 1), as compared to the average for
the 100 × 103 km2 of dams planned of 1 Watt/m2, also a very low
value (Rosa and others 1996, p. 134). At the best dam sites
in the region the power density can exceed 10 Watts/m2, but
values this high are often misleading because they fail to
include the impacts of less-favorable upstream dams used to
regulate streamflow and increase the installed capacity that
can be used effectively (Fearnside 1996).
Some confusion exists over the correct area of the Samuel
reservoir. According to the staff at the dam, the reservoir
area is 540 km2 at the normal operating level of 87 m above
msl. ELETRONORTE publications prior to filling the reservoir
gave the area at this elevation as 645 km2, while a LANDSAT
measurement by INPE from 1989 images indicated 465 km2 (see
Fearnside 1995, p. 11). However, the reservoir only finished
filling in July 1989, and difficulty in distinguishing dry
land from moribund forest in shallow water may account for the
difference. The 540-km2 area at the 87-m water level will be
used in the present paper. The area at the maximum water
level of 87.4 m is 586 km2, and at the historic minimum of 72 m
it is 135 km2 (based on adjusted areas from Brazil, ELETRONORTE
nd [C. 1987]).
At Samuel, 420 km2 of forest was lost, after deducting
from the reservoir’s 540-km2 total area the 29-km2 riverbed
area (calculated from Brazil, ELETRONORTE nd [C. 1986], see
Fearnside 1995, p. 11) and 91 km2 of previous clearing
(Fearnside 1995, p. 11). Because the reservoir is in a
relatively flat area, 57 km of dikes were built to confine the
lateral expansion of the reservoir and thereby increase the
head that could be generated without flooding a still larger
area.
Original plans would have resulted in further forest loss,
as they called for construction of a second dam upstream of
Samuel at Monte Cristo, 8 km downstream of the town of Ariquemes
(Brazil, ELETRONORTE. nd [C. 1985]). Regulation of the river’s
flow by this additional 243 km2 reservoir would increase the
firm power at Samuel from 60 to 70 MW, and the two dams together
would have a combined firm power of 95 MW. However, the
advanced state of settlement in the Ariquemes area now makes it
unlikely that expropriating the land for the Monte Cristo
reservoir would be politically feasible.
Logging loophole
Since 1965, Brazil has prohibited the export of raw logs,
thereby forcing logging companies to at least do a minimal
13
amount of the sawing in Brazil and contribute to creating
employment. However, a special exception was opened in this
prohibition to allow logs from Samuel to be exported (Nogueira
1988). From 1987 through 1989 a continuous chain of barges
arrived in the port of Itacoatiara with logs for loading on
ships, and one ship loaded with logs departed every two weeks
(Figure 3).
[Figure 3 here]
Sedimentation
Silt from mining of cassiterite (tin ore) is a large source
of sediments in the drainage basins affected. One negative
effect of the increase could be more rapid sedimentation of the
Samuel reservoir. One mining operation (Mineração Oriente Novo,
owned by the Paranapanema mining group) released large amounts
of sediment into the Rio Preto (a tributary in the Samuel
catchment) until it was stopped in 1986 by a federal court
order. Other operations in the Samuel catchment, such as the
BRASCAN mines, store their fines behind small retaining dams.
Cassiterite mining is now minimal due to the low price of tin.
Soil erosion is another major source of sediments (Graham
1986). Since much of the catchment is settlement areas,
deforestation for agriculture and ranching is widespread,
resulting in greater soil loss. A study of sediments in Paca
Lake (on the Jamari River 6 km upstream from its confluence with
the Madeira) used 210Pb chronological techniques to demonstrate
an order-of-magnitude increase in sedimentation rates since 1961
due to soil erosion in settlement areas and cassiterite mining
(Forsberg and others 1989).
Aquatic Ecosystems
Aquatic ecosystems in the section of the river now
occupied by the reservoir were completely altered. The
river’s sinuous course wound 255 km through the reservoir, now
134 km long, or 122 km in a straight line. Conversion of a
running-water (lotic) system to a still-water (lentic) one
inevitably involves loss of many species of fish and other
organisms, and relative increases in the abundance of others.
This is especially true when, as in the case of a reservoir
like Samuel, water at the bottom of the reservoir is anoxic
over much of the year. Samuel has an average turnover time of
0.4 years (Rosa and others 1997, p. 44); this is an unusually
long period for the average drop of water to remain in the
reservoir, and is more than twice the 0.14 year turnover time
of Tucuruí (which is also considered long). Upstream of the
Samuel reservoir the interruption of the annual fish migration
(the “piracema” or spawning run) can be expected to alter the
composition of species inhabiting the river. Prior to closing
the dam, 86 fish species were collected in the area in March
and April 1986 (dos Santos 1986). The Jamari River proper has
186 species of plankton out of a total of 210 in the Jamari
basin (Mera 1985, pp. 6, 9).
14
Most of the fish and fishing are in the upper reaches
of the reservoir. The reservoir as a whole does not have many
fish, although it did have an explosion of some fish
populations in the early years. Existence of the BR-364
Highway along one bank of the reservoir makes unauthorized
fishing easy. A fishermen’s cooperative was founded but later
disbanded.
The town of Itapoã do Oeste is pressing ELETRONORTE to
stock fish in the Samuel reservoir (de Oliveira 2001).
However, stocking is often inefficient as a means of fish
production because, unlike aquaculture in small ponds, most of
the fingerlings released in a reservoir are never recaptured.
The same investment in promoting aquaculture could result in
greater fish production.
As in other Amazonian reservoirs, the population of
macrophytes (water weeds) exploded in the early years.
ELETRONORTE measured the macrophyte areas in 1996 and 1997,
but decided that macrophyte areas since then have not been
large enough to justify continuing the measurements. LANDSAT
satellite measurements of macrophyte areas (in the dry season)
indicate 48% of the reservoir surface covered in 1989, falling
exponentially to 0.08% in 1998 (de Lima 2002, p. 47). The
main floating macrophytes are Salvinia, Eichhornia and
Oricularia (the last of which is an indicator of low-fertility
water). In the shallow areas a rooted weed known as
“poligano” (in the family Poliganaceae) is common.
For downstream ecosystems, lowering of the oxygen content
of the water is the change with the greatest effect. Water use
at Samuel is the worst possible for downstream water quality,
as virtually all water is passed through the turbines (the
spillway has not been used since 1996). Fortunately, there
were not many riverside residents along the downstream
stretch, settlement being limited to isolated families. The
stretch of river below the dam that is entirely dependent on
water released by Samuel is relatively short, as the Rio
Candeias comes into the Jamari 42 km downstream. Water
quality therefore improves below this point. The Rio
Candeias, with an average flow of 315 m3/s (April 1976-March
1996: Brazil, ANEEL nd [C. 1999]), or 90% of the Jamari’s
flow, is a big enough river to substantially improve water
quality, at least in the months when flow is high (i.e.,
except for August-November). In addition, one very small
river, the Rio Novo enters the Jamari 3 km downstream of
Samuel.
Water Table
Stress in trees bordering the reservoir is visible on
LANDSAT images, where the false colors of the area around the
reservoir indicate a swath of vegetation that is neither the
healthy forest nor the dead trees in the reservoir proper.
Water table alteration is the most likely explanation. Raising
of the water table is also one of the main complaints of the
nearby town of Itapoã do Oeste, where streets become mud holes
15
and where a series of canals has been built in an attempt to
drain the excess water (de Oliveira 2001).
Greenhouse-Gas Emissions
Greenhouse-gas emissions come from carbon dioxide released
by decay of dead trees that project above the water surface in
the reservoir, and from methane produced by decay under anoxic
conditions at the bottom of the reservoir (Fearnside 2002b,
2004). Some methane is released by the reservoir surface
through bubbling and diffusion, but much larger amounts are
released from water as it emerges from the turbines or from the
spillway. Most of the carbon in the methane comes from
decomposition of soft plant material, such as macrophytes and
the green vegetation that grows in the drawdown zone when
exposed and is regularly flooded when the water level
subsequently rises.
Parameters for calculation of greenhouse-gas emissions are
given in Table 3. Emissions are estimated in Table 4 for 1990
(the base year for national inventories of greenhouse gases
under the United Nations Framework Convention on Climate
Change), and for 2000 (after emissions had adjusted to levels
that are likely to remain stable over the long term). In 1990,
Samuel emitted 11.6 times more greenhouse gases than would
have been emitted from oil; in subsequent years these
emissions declined, but remained 2.6 times greater than the
fossil-fuel alternative in 2000.
[Tables 3 & 4 here]
Social Costs
Resettlement
A total of 238 families were resettled from within the
submergence area (Munasinghe 1988, p. 5). The 50 km of the
BR-364 highway that was flooded represented the main source of
this population displacement (Brazil, ELETRONORTE nd [2003]).
Those displaced from the reservoir area were moved to the Rio
Preto do Candeias project (Brazil, ELETRONORTE nd [C. 1989]).
Of the family heads, 10% had a declared occupation as rubber
tappers (Brazil, ELETRONORTE 1990, p. 43).
In addition, 20 families from Cachoeira de Samuel (the
dam site, which was a bathing place for weekend visitors from
Porto Velho) were moved in 1984 to Vila Candeias on the BR-364
Highway roadside 20 km from Porto Velho. Six years later,
these families had either disappeared or were no longer
distinguishable as a community (Brazil, ELETRONORTE 1990).
Formation of the reservoir cut off road transportation to
part of a settlement area established by the National
Institute for Colonization and Agrarian Reform (INCRA).
ELETRONORTE has provided a ferry service linking this area to
the road system, but the service has been unreliable and is a
16
point of friction between ELETRONORTE and the surrounding
population (de Oliveira 2001).
Indigenous Peoples
No indigenous people were flooded by Samuel. However, the
dam may have caused impacts on the Uru-Eu-Uau-Uau tribe, which
inhabits the headwaters of the Jamari River, by cutting off fish
migration and by contributing to the attraction of additional
population to Rondônia, further increasing pressure on
indigenous areas (Leonel 1987, p. 30). Proximity to the
Karitiána indigenous area was considered a threat to the
Karipúna tribe, which had a population of only 175 individuals
(Koifman 2001, p. 417).
Health Impacts
Mosquitoes
Creation of a reservoir like Samuel can provide breeding
places for disease vectors such as the anopheline mosquitoes
that transmit malaria. Samuel is located between Ariquemes
and Porto Velho, which are known for having the highest
incidences of malaria in the World (Almeida and Rodrigues
1996, Ellis and others 1988). Although the very high incidence
of malaria has been a feature of this area since long before
the construction of Samuel, the presence of the reservoir
probably exacerbates the situation. At the dam site, up to
21.8 anopheline bites were counted per person per hour (March
1987), with monthly means up to 9.0 bites/person/hour (Tadei
1987, p. 6).
In addition to anopheline mosquitoes, Samuel resulted in
an explosion of mosquitoes of the genus Culex (Luz 1994). For
example, in September 2001 the swarms of these mosquitoes over
the reservoir were so dense that they forced suspension of
early-morning data collection on methane emissions (de Lima
2002, p. 43). Culex can transmit filaria (elephantiasis), but
the parasite has not yet appeared in Rondônia; it is present
in French Guiana and Surinam and may eventually spread to
Brazilian Amazonia.
Mercury contamination
One of the impacts of hydroelectric dams in Amazonia is
release of mercury from the soil in its toxic (methyl mercury)
form. Although goldmining is not a problem in the Samuel
catchment, the soils flooded by the reservoir contain mercury
from natural sources. This is because Amazonian soils are
millions of years old and have been gradually accumulating
mercury from deposition in rain and dust from volcanic eruptions
and other sources around the world. The anoxic conditions at
the bottom of a reservoir provide the environment needed for
methylation of mercury, which increases in concentration by
about ten fold with each link in the food chain from plankton to
fish to people who eat the fish. The concentration appears to
follow a pattern of increasing over several years, following by
17
a decrease, but differences among reservoirs and the limited
number of available measurements impede firm conclusions.
At Samuel, the only measurements available were done in
late 1991 (two years after filling the reservoir), and indicated
a total mercury concentration of 0.33 mg/kg fresh weight of
fish in tucunaré (Cichla ocellaris and C. temensis) (Malm and
others 1995). Tucunaré is a predatory fish that makes up most
of the commercial catch in Amazonian reservoirs, including
Samuel. An estimated 80% of the total mercury contained in
the fish is in the methyl (poisonous) form (Kashima and others
2001). The maximum concentration of total mercury in fish
considered safe for human consumption in Brazil was 0.5 mg/kg
fresh weight until 1998, when the criterion was revised upward
to 1.0 mg/kg fresh weight. The question of what safe levels
should be is a matter of controversy (Kaiser 2000). The World
Health Organization (WHO 1976) standard of 0.5 mg/kg fresh
weight is based on the assumption that a 70 kg human would
consume 60 g fresh weight of fish per day, but fish
consumption of approximately 200 g daily by those who live
beside Amazonian rivers and reservoirs indicates that the
levels of mercury in fish would have to be much lower
[approximately 0.15 mg/kg fresh weight] to meet the same
safety standard (Weisser 2001, p. 5). This is not to say that
riverside residents should forego eating fish, as the negative
impacts of poor nutrition could outweigh those of the mercury
(Cleary 1996).
High levels of mercury in fish were found at the Tucuruí
Dam, where total mercury in reached 1.1 mg/kg fresh weight of
fish in tucunaré six years after the reservoir was filled
(Porvari 1995). However, a measurement made 16 years after
filling the Tucuruí reservoir found a mean of 0.3 mg/kg fresh
weight in tucunaré (Santos and others 2001). The decline in
mercury concentration at Tucuruí is considered a matter of
good luck, as some reservoirs can maintain high levels for up
to 30 years (Olaf Malm, personal communication 2003).
The mercury in fish is reflected in concentrations in the
hair of people who eat them, as in the case of the high
concentrations at Tucuruí six years after filling (Leino and
Lodenius 1995). A unique data set at the Balbina reservoir
allows tracking of the history of mercury contamination there
over time. Changes in the concentration of mercury in women’s
hair were dated by sectioning hair samples from long-haired
women, revealing that mercury levels were low before flooding
the reservoir, followed by an abrupt rise with reservoir
filling, and a drop after concentration reached a peak 11.2
years after filling the reservoir (Weisser 2001, p. 37). The
drop may have been caused by a decline in the concentration in
the fish, and by the confounding effect of the fish catch from
the reservoir having diminished as a result of declining
fertility of the water, forcing the residents to eat chicken,
pond-raised fish and beef rather than fish from the reservoir
(Bruce R. Forsberg, personal communication 2001). At Balbina,
the concentration of total mercury in tucunaré was
approximately 0.34 mg/kg fresh weight in 1996, 8.4 years after
18
filling the reservoir (Kehrig and others 1998). This is a
weighted average methylmercury concentration between the two
species of tucunaré in 17 specimens is 0.27 mg/kg fresh weight
(Kehrig and others 1998), which is equivalent to 0.31 mg/kg
fresh weight of total mercury if methylmercury is assumed to
be 80% of total mercury (Kashima and others 2001). The hair
sample series at Balbina indicates a doubling of mercury
concentrations in the hair of fish-eating women between the
year of the measurement in fish (1996) and the peak
concentration in hair in 1999 (Weisser 2001, p. 37).
Comparisons among studies and reservoirs are complicated by a
significant positive correlation between the length of a fish
and the mercury concentration in its flesh in Tucunaré
(Weisser 2001). Mercury levels in tucunaré in Balbina more
than doubled between 1992 and 1997 for fish of any given
length (Weisser 2001, p. 44). A rise and drop in mercury
contamination similar to that at Balbina is likely to have
occurred at Samuel, but time-series data for Samuel are
unavailable.
Mitigating Measures
Forest Reserve
The presence of the Samuel Dam offered the opportunity
for creation of a protected area by the federal government’s
Special Secretariat of the Environment (SEMA), which has since
been incorporated into the Brazilian Institute for the
Environment and Renewable Natural Resources (IBAMA). Paulo
Nogueira Neto, who directed SEMA, was a master at capitalizing
on opportunities to create protected areas, as exemplified by
the ecological station he created at Samuel (Fearnside 2003).
The Samuel Ecological Station was created on the edge of the
reservoir in 1989. Approximately 4700 ha of the 20,854 ha
area was flooded (Brazil, ELETRONORTE 1990, p. 60). In 2002,
ecological stations were re-designated as “ecological
reserves” under Brazil’s new National System of Conservation
Areas (SNUC).
Faunal Rescue
A faunal rescue operation carried out as the reservoir was
filling attempted to collect animals stranded in treetops and
relocate them to nearby forested areas such as the forest
reserve associated with Samuel (de Sá 1992). Of 16,000 animals
rescued, 2854 were released in the reserve, while the remainder
were either sent to research institutions (11,417) or sacrificed
for museum collections or research (1,729). Release of the
animals, in truth, does not reduce the death toll to animals
because the individuals placed in forests elsewhere enter into
competition with individuals that are already there, and the
population as a whole soon reverts to the level before the
introductions (see Gribel 1993). At Samuel, primate biomass in
the adjacent ecological reserve was estimated at 154 ± 65 kg/km2
in 1988 prior to filling the reservoir, increasing to 255 kg/km2
in 1990, and returning to 153 ± 81 kg/km2 in 1991 (de Sá 1995).
Bird biomass in the reserve also increased when flooding
19
occurred, followed by a decline from 1990 onwards. Similar
patterns were seen with agoutis (Dasyprocta fuliginosa) and
brocket deer (Mazama spp) (de Sá 1995).
An unusual amount of information on wildlife and the effect
of the faunal rescue program is available for Samuel, thanks to
the efforts of Rosa de Sá, who walked 1224 km of transects in
the 1989-1991 period to survey mammal and bird populations in
the reserve adjacent to the reservoir and in a control area in
forest downstream of the dam (de Sá 1995). While recognizing
the improvements that ELETRONORTE made in its procedures since
the faunal rescue five years earlier at Tucuruí, the study
indicates that the basic problem of the approach, namely that it
is ineffective and very expensive as a conservation measure,
remains fundamentally unchanged. At Tucuruí, all animals
captured were released near the reservoir, where their survival
was problematic not only because of competition with the animal
populations already inhabiting the forests at the release points
but also because these forests were rapidly being cleared (there
were no protected areas). At Samuel, many of the animals
captured were donated to research institutions rather than being
released (a practice initiated at Balbina), and the creation of
a protected area adjacent to the reservoir was an important
improvement. At Tucuruí, the rescue operation cost UD$30
million, or US$134.80 per animal rescued (about half of which
were arthropods); the cost was US$280 per individual if only
birds and mammals are considered (Johns 1986 cited by de Sá
1995, p. 7). Cost figures have not been released for the rescue
operation at Samuel, but the fact that it is very expensive is
inescapable. In the words of de Sá (1995, p. 110), “rescue
operations have become a public relations strategy used by power
companies to appease public opinion.”
From the point of view of maintaining biodiversity,
investment in protecting existing forests would have much
greater return than faunal rescue operations of this type, but
this kind of investment has a lower public-relations value to
the companies. The recommendations of de Sá (1995, pp. 110-111)
are that rescue operations be confined to species falling into
one or more of the following categories: 1.) species classified
as “endangered” or “vulnerable,” either by the World
Conservation Union (IUCN) or by other criteria [at Samuel
species classified as vulnerable by IUCN included giant
anteaters (Mirmecophaga tridactyla), giant armadillos
(Priodontes maximus), and spider monkeys (Ateles paniscus)], 2.)
species unable to escape flooding, 3.) species that could be
used for research (such as snakes, scorpions and spiders from
Samuel that were used for development or production of vaccines
and other medicines), and 4.) species that could be used to re-
establish depleted populations elsewhere. Animals should only
be released in areas that have been previously studied and found
to have depleted populations (as through hunting). Other
investments, such as creation of conservation units and
investment in professional conservation staff, must receive
priority over the brief but photogenic activity of capturing and
releasing wildlife.
20
Conclusions: The Lessons of Samuel
The Samuel Dam illustrates a variety of aspects of the
decision-making process that have impeded the choice of
development options based on a prior evaluation of costs and
benefits, including environmental and social consequences.
Making a rational decision on whether a given development
project should be implemented depends on an accurate and
unbiased assessment of both sides of this balance. The Samuel
Dam has significant environmental and social costs, but also
real benefits. Some significant impacts were unforeseen due
to limited knowledge at the time of the decision to build
Samuel, especially in the cases of greenhouse-gas emissions
and mercury contamination. The political role of the dam
meant that the timing of decisions was incompatible with the
logical sequence whereby information is collected on the
various options, comparisons are made, and finally a decision
is reached.
Assessment of alternatives was almost completely lacking
in the case of Samuel. Among these alternatives was the
possibility of awaiting a more definitive solution to energy
supply for Rondônia, since the capacity of Samuel was soon
overwhelmed by demand anyway. The dam illustrates the dilemma
of whether to meet demand in a strictly incremental fashion,
even when the options (like Samuel) identified by this
approach are both financially expensive and of high impact
relative to their benefits, or if temporary solutions (such as
thermal generation) should be used until more attractive long-
term options can be implanted. These options included
supplying Rondônia through a transmission link to Brazil’s
national grid, exploitation of natural gas deposits in the
neighboring state of Amazonas, and larger dams elsewhere in
Rondônia. Of course, these options also have impacts that
would need to be compared, but the possibility of transmission
from the national grid would be especially favorable because
the transmission line route through Rondônia and Mato Grosso
is already deforested. Samuel illustrates the need for
flexibility if the balance of environmental impacts and
project benefits changes significantly during the course of
planning and construction, in this case due to improvement of
long-distance power transmission technology and the discovery
of natural gas. Many of the impacts of Samuel apply to
planned hydroelectric dams elsewhere in Amazonia and in other
tropical areas, and many of the decision-making challenges
posed by the dam are applicable to development projects
throughout the World.
Acknowledgments
The National Council of Scientific and Technological
Development (CNPq AI 523980/96-5; 350230/97-98; 465819/00-1;
470765/01-1) and the National Institute for Research in the
Amazon (INPA PPI 5-3150; 1-3160) provided financial support. I
thank the ELETRONORTE staff at Samuel for their patience during
my visits in 1987, 1995 and 2001. Reinaldo I. Barbosa and two
21
anonymous reviewers made useful comments on the manuscript. I
also thank the late Darrell Posey, who accompanied me at Samuel
in 1995 and encouraged me to write this paper.
Literature Cited
Almeida, A., and J. Rodrigues. 1996. Rondônia ecológica. Isto
É [São Paulo], 31 January 1996, Pages 90-95.
Bartholo Jr., R.S., and M. Bursztyn. 1999. Amazônia Sustentável:
Estratégia de Desenvolvimento Rondônia 2020. Instituto
Brasileiro do Meio Ambiente e dos Recursos Naturais
Renováveis (IBAMA), Brasília, DF, Brazil, 247 pp.
Branco, C. C. 1984. No caminho do Planalto: Transformada em
questão civil, a sucessão põe Andreazza e Maluf, crias de
Costa e Silva, contra Tancredo Neves, o contraparente de
Getúlio Vargas. Isto É [São Paulo]. 8 August 1984, Pages
18-19.
Brazil, ANEEL. 2001. Mapa.
http://www.aneel.gov.br/defaultmapa.html. Agência Nacional
de Energia Elétrica (ANEEL), Brasília, DF, Brazil.
Brazil, ANEEL. 2003. Agência Nacional de Energia Elétrica.
http://www.aneel.gov.br. Agência Nacional de Energia
Elétrica (ANEEL), Brasília, DF, Brazil.
Brazil, ANEEL. nd [C. 1999]. Sistemas Georeferenciados de
Interesse dos Setores de Energia e Hidrologia. Bacia 1.
Sistema de Estatísticas e Informações Hidrológicas (SIH),
Agência Nacional de Energia Elétrica (ANEEL), Brasília,
DF, Brazil. (CD ROM).
Brazil, ELETROBRÁS. 1986. Plano diretor para proteção e
melhoria do meio ambiente nas obras e serviços do setor
elétrico. ELETROBRÁS, Diretoria de Planejamento e
Engenharia, Departamento de Recursos Energéticos.
Brasília, DF, Brazil. (mimeographed). Irregular
pagination.
Brazil, ELETROBRÁS. 1987. Plano 2010: Relatório Geral. Plano
Nacional de Energia Elétrica 1987/2010 (Dezembro de 1987).
Centrais Elétricas do Brasil (ELETROBRÁS), Brasília, DF,
Brazil, 269 pp.
Brazil, ELETROBRÁS. 1993. Plano Nacional de Energia Elétrica
1993-2015: Plano 2015. Centrais Elétricas do Brasil
(ELETROBRÁS), Brasília, DF, Brazil.
Brazil, ELETROBRÁS/CEPEL. 1983. Relatório Técnico Final No.
963/83: Estudo Comparativo de Manutenção nas Usinas de
Curuá-Una e Moxotó. ELETROBRÁS/Centro de Pesquisas de
Energia Elétrica (CEPEL), Brasília, DF, Brazil, 48 pp.
Brazil, ELETRONORTE. 1987. UHE Ji-Paraná. Maio-87. Centrais
Elétricas do Norte do Brasil S.A. (ELETRONORTE), Brasília,
22
DF, Brazil, 8 pp.
Brazil, ELETRONORTE. 1990. Estação Ecológica de Samuel: Plano
de Manejo. Departamento de Estudos de Efeitos Ambientais,
Divisão de Ambiente Terrestre, Centrais Elétricas do Norte
do Brazil S.A. (ELETRONORTE), Brasília, DF, Brazil, 126
pp.
Brazil, ELETRONORTE. nd [C. 1985]. UHE Samuel. Centrais
Elétricas do Norte do Brasil S.A. (ELETRONORTE), Brasília,
DF, Brazil, 15 pp.
Brazil, ELETRONORTE. nd [C. 1986]. Reservatório da UHE Samuel:
Levantamento Planimêtrico. Centrais Elétricas do Norte do
Brasil S.A. (ELETRONORTE), Brasília, DF, Brazil. Map scale:
1:40,000.
Brazil, ELETRONORTE. nd [C. 1987]. UHE Samuel. Centrais
Elétricas do Norte do Brasil S.A. (ELETRONORTE), Brasília,
DF, Brazil, 22 pp.
Brazil, ELETRONORTE nd [C. 1989]. Enchimento do reservatório:
Usina Hidrelétrica de Samuel-Rondônia. Centrais Elétricas
do Norte do Brasil S.A. (ELETRONORTE), Brasília, DF,
Brazil. 16 pp.
Brazil, ELETRONORTE. nd [2003]. Usina Hidrelétrica Samuel.
http://www.eln.gov.br/meiosam1.htm. Centrais Elétricas do
Norte do Brazil S.A. (ELETRONORTE), Brasília, DF, Brazil.
Brazil, Furnas 2004. Usina de Manso - 210 MW. WEBFURNAS -
Furnas Centrais Elétricas S.A. http://www.furnas.com.br/
Browder, J. O. 1986. Logging the Rainforest: A Political Economy
of Timber Extraction and Unequal Exchange in the Brazilian
Amazon. Ph.D. dissertation in city and regional planning,
University of Pennsylvania, Philadelphia. 393 pp.
Browder, J. O., and B. J. Godfrey. 1997. Rainforest Cities:
Urbanization, Development, and Globalization of the
Brazilian Amazon. Columbia University Press, New York. 429
pp.
Cadman, J. D. 1989. Energy from the Amazon. Civil Engineering
December 1989:54-57.
Carvalho, G., A. C. Barros, P. Moutinho, and D. C. Nepstad.
2001. Sensitive development could protect Amazonia instead
of destroying it. Nature 409:131.
Cleary, D. 1996. Mercury contamination and health risk in the
Brazilian Amazon: An ethical dilemma. Revista do
Instituto de Medicina Tropical de São Paulo 38:247-248.
Coalição Rios Vivos. 1999. Barragens no Rio Tocantins.
http://www.riosvivos.org.br/materia.php?mat_id=177
23
Consórcio Brasiliana. 1998. Programa Brasil em Ação: Eixos
Nacionais de Integração e Desenvolvimento. Banco Nacional
de Desenvolvimento Econômico e Social (BNDES) PBA/CN-01/97.
Relatório Final do Marco Inicial, Consórcio Brasiliana,
Brasília, DF, Brazil, 3 vols.
Corrêa, M. 2001. “A aposta agora é na Amazônia: Inventário feito
pelo governo indica que se concentra no Norte 43% do
potencial energético.” Gazeta Mercantil [Brasília] 20 July
2001, p. A-4.
A Crítica [Manaus]. 2003. “MPF vê entraves para o gasoduto.”
19 March 2003
http://www.acritica.com.br/noticias/detalhe.asp?var_mater
ia_id=85571.
da Rosa, S. L., and Z. Cardoso. 1993. Estado de Tocantins:
Aportes para um desenvolvimento harmônico e autosustentado.
OCTA Consultaria e Planejamento S/C Ltda., São Paulo,
Brazil. 8 pp.
de Lima, I. B. T. 2002. Emissão de metano em reservatórios
hidrelétricos amazônicos através de leis de potência.
Ph.D dissertation in nuclear energy, Centro de Energia
Nuclear na Agricultura (CENA), Universidade de São Paulo,
Piracicaba, São Paulo, Brazil, 108 pp.
de Oliveira, R. J. M. 2001. Movimento Contra as Agressões
Sócio-Ambientais da UHE/Samuel “Projeto Reparar”: Os
peixes pedem passagem—os homens também. Prefeitura
Municipal de Itapuã do Oeste, Itapuã do Oeste, Rondônia,
Brazil, Irregular pagination.
de Sá, R. M. L. 1992. A view of hydroelectric dams in the
Amazon, with emphasis on the Samuel Dam, Rondônia. TCD
Newsletter 25:1-4. [Tropical Conservation and Development
Program, University of Florida, Gainesville, Florida].
de Sá, R. M. L. 1995. Effects of the Samuel Hydroelectric Dam on
Mammal and Bird Communities in a Heterogeneous Amazonian
Lowland Forest. Ph.D. dissertation in Forest Resources and
Conservation, University of Florida, Gainesville, Florida,
140 pp.
dos Santos, G. M. 1986. Estudos de Ecologia e Controle
Ambiental na Região do Reservatório da UHE Samuel,
Convênio ELN/CNPq/INPA de 01.07.82. Relatório Setorial,
Segmento Estimativa da Ictiofauna. Período Jan/Junho,
1986. Instituto Nacional de Pesquisas da Amazônia (INPA),
Manaus, Amazonas, Brazil, 38 pp.
Ellis, W. S., W. A. Allard, and L. McIntyre. 1988. Rondônia’s
settlers invade Brazil’s imperiled rain forest. National
Geographic December 1988:722-799.
FADESP (Fundação de Amparo e Desenvolvimento da Pesquisa).
1996a. Relatório de Estudos de Impacto Ambiental - EIA,
24
referente ao projeto de implantação da Hidrovia dos
rios Tocantins, Araguaia e Mortes. FADESP, Universidade
Federal do Pará (UFPA), Belém, Pará, Brazil, 7 vols.
FADESP (Fundação de Amparo e Desenvolvimento da Pesquisa).
1996b. Relatório de Estudos de Impacto Ambiental - EIA,
referente ao projeto de implantação da Hidrovia dos rios
Tocantins, Araguaia e das Mortes. FADESP, Universidade
Federal do Pará (UFPA), Belém, Pará, Brazil. 109 pp.
Faminow, M. D. 1998. Cattle, Deforestation and Development in
the Amazon: An Economic and Environmental Perspective. CAB
International, New York, 253 pp.
Fearnside, P. M. 1986a. Human Carrying Capacity of the Brazilian
Rainforest. Columbia University Press, New York, 293 pp.
Fearnside, P. M. 1986b. Settlement in Rondônia and the token
role of science and technology in Brazil's Amazonian
development planning. Interciencia 11:229-236.
Fearnside, P. M. 1986c. Spatial concentration of deforestation
in the Brazilian Amazon. Ambio 15:72-79.
Fearnside, P. M. 1987a. Deforestation and international
economic development projects in Brazilian Amazonia.
Conservation Biology 1:214-221.
Fearnside, P. M. 1987b. Causes of Deforestation in the
Brazilian Amazon. Pages 37-61 in R. F. Dickinson (ed.)
The Geophysiology of Amazonia: Vegetation and Climate
Interactions. John Wiley & Sons, New York, 526 pp.
Fearnside, P. M. 1989a. Brazil's Balbina Dam: Environment versus
the legacy of the pharaohs in Amazonia. Environmental
Management 13:401-423.
Fearnside, P. M. 1989b. A Ocupação Humana de Rondônia:
Impactos, Limites e Planejamento. (CNPq Relatórios de
Pesquisa No. 5). Conselho Nacional de Desenvolvimento
Científico e Tecnológico (CNPq), Brasília, DF, Brazil, 76
pp.
Fearnside, P. M. 1993. Deforestation in Brazilian Amazonia: The
effect of population and land tenure. Ambio 22:537-545.
Fearnside, P. M. 1995. Hydroelectric dams in the Brazilian
Amazon as sources of 'greenhouse' gases. Environmental
Conservation 22:7-19.
Fearnside, P. M. 1996. Hydroelectric dams in Brazilian
Amazonia: Response to Rosa, Schaeffer & dos Santos.
Environmental Conservation 23(2):105-108.
Fearnside, P. M. 1999a. Social Impacts of Brazil's Tucuruí
Dam. Environmental Management 24:485-495.
25
Fearnside, P. M. 1999b. Biodiversity as an environmental
service in Brazil's Amazonian forests: Risks, value and
conservation. Environmental Conservation 26:305-321.
Fearnside, P. M. 2000. Environmental services as a strategy for
sustainable development in rural Amazonia. Pages 154-185
in C. Cavalcanti (ed.) The Environment, Sustainable
Development and Public Policies: Sustainability in
Brazil. Edward Elgar, Cheltenham, UK, 219 pp.
Fearnside, P. M. 2001a. Environmental impacts of Brazil's
Tucuruí Dam: Unlearned lessons for hydroelectric
development in Amazonia. Environmental Management 27:377-
396.
Fearnside, P. M. 2001b. Soybean cultivation as a threat to the
environment in Brazil. Environmental Conservation 28:23-38.
Fearnside, P. M. 2002a. Avança Brasil: Environmental and
social consequences of Brazil’s planned infrastructure in
Amazonia. Environmental Management 30:748-763.
Fearnside, P. M. 2002b. Greenhouse gas emissions from a
hydroelectric reservoir (Brazil’s Tucuruí Dam) and the
energy policy implications. Water, Air and Soil Pollution
133:69-96.
Fearnside, P. M. 2003. Conservation policy in Brazilian
Amazonia: Understanding the dilemmas. World Development
31(5):757-779.
Fearnside, P. M. 2004. Greenhouse gas emissions from
hydroelectric dams: Controversies provide a springboard
for rethinking a supposedly “clean” energy source.
Climatic Change (in press).
Fearnside, P. M., and R. I. Barbosa. 1996. Political benefits
as barriers to assessment of environmental costs in
Brazil's Amazonian development planning: The example of
the Jatapu Dam in Roraima. Environmental Management
20(5):615-630.
Forsberg, B., J. M. Godoy, R. Victoria, and L. A. Martinelli.
1989. Development and erosion in the Brazilian Amazon: A
geochronological case study. GeoJournal 19:402-405.
Foster, H. 1986. Statement by Hugh W. Foster, U.S. Alternative
Executive Director to the Board of Executive Directors.
Brazil: Electric Power Sector Loan. Written statement to
the World Bank Board of Executive Directors dated 19 June
1986. 4 Pages (Reproduced in: Information Packet: Brazilian
Power Sector Loan, Environmental Defense Fund, Washington,
DC).
Galy-Lacaux, C., R. Delmas, C. Jambert, J.-F. Dumestre, L.
Labroue, S. Richard, and P. Gosse. 1997. Gaseous emissions
and oxygen consumption in hydroelectric dams: A case study
26
in French Guyana. Global Biogeochemical Cycles 11:471-
483.
Galy-Lacaux, C., R. Delmas, J. Kouadio, S. Richard, and P.
Gosse. 1999. Long-term greenhouse gas emissions from
hydroelectric reservoirs in tropical forest regions. Global
Biogeochemical Cycles 13(2):503-517.
Gawora, D. 1998. Produção de petróleo e de gás natural no Estado
do Amazonas: Projeto Petrobrás de gás natural de Urucú.
Lateinamerika Dokumentionsstelle Brasilien, Universität
Gesamthochschule Kassel, Kassel, Germany, 23 pp.
Gazeta Mercantil [Brasília]. 2001.“Energia: MP fixa prazos
para licenças ambientais.” 15 May 2001.
(http://www.gazetamercantil.com.br).
Goodland, R. J. A. 1980. Environmental ranking of Amazonian
development projects in Brazil. Environmental Conservation
7:9-26.
Graham, D. H. Jr. 1986. The Samuel Dam: Land Use, Soil Erosion,
and Sedimentation in Amazonia. Masters thesis in
geography, University of Florida, Gainesville, Florida, 75
pp.
Gribel, R. 1993. Os mamíferos silvestres e as grandes
barragens na Amazônia. Pages 125-140 in E.J.G. Ferreira,
G. M. Santos, E. L. M. Leão, and L. A. Oliveira (eds.)
Bases Científicas para Estratégias de Preservação e
Desenvolvimento da Amazônia, Vol. 2. Instituto Nacional
de Pesquisas da Amazônia (INPA), Manaus, Amazonas,
Brazil, 435 pp.
Holden, C. 1987. World Bank launches new environment policy.
Science 236:769.
Isto É. [São Paulo]. 1984. “Truques sem futuro: Constituinte e
ministros são temperos para tomar palatável a prorrogação
de Figueiredo.” 16 May 1984, Pages 14-17.
Johns, A. D. 1986. Effects of Habitat Disturbance on
Rainforest Wildlife in Brazilian Amazonia. Final Report.
World Wildlife Fund U.S., Washington, DC.
Jornal do Brasil [Rio de Janeiro]. 2003. “Novas usinas serão
licitadas.” 10 March 2003.
http://jbonline.terra.com.br/papel/economia/2003/03/10/jo
reco20030310008.html
Jornal do Comércio [Manaus]. 1986a. “Em defesa do INPA”. 1
April 1986, p. 2.
Jornal do Comércio [Manaus]. 1986b. “Schubart acha parecer
sobre o Inpa alarmista.” 1 April 1986, Page 13.
Junk, W. J., and J. A. S. de Mello. 1987. Impactos ecológicos
27
das represas hidrelétricas na bacia amazônica brasileira.
Pages 367-385 in G. Kohlhepp, and A. Schrader (eds.)
Homem e Natureza na Amazônia. Tübinger Geographische
Studien 95 (Tübinger Beiträge zur Geographischen
Lateinamerika-Forschung 3). Geographisches Institut,
Universität Tübingen, Tübingen, Germany, 507 pp.
Jury, W. J. 1989. Considerações sobre a ocorrência de
canalículos em locais de barragem na região amazônica.
Masters thesis in geotechnica,Universidade de São Paulo
em São Carlos, São Carlos, São Paulo, Brazil, 74 pp.
http://ged.capes.gov.br/AgTest/silverstream/pages/pgRelat
orioTesesCompleto.html?ANOBASE=1989&IDPROGRAMA=3300204500
9P3&SEQUENCIAL=5
Kacowicz, Z. 1985. Convênio ELETRONORTE/CNPq/INPA. Relatórios
semestrais dos projetos “Estudos de ecologia e controle
ambiental nas regiões dos reservatórios UHEs de Balbina e
Samuel: Comentários. Conselho Nacional de Desenvolvimento
Científico e Tecnológico (CNPq), Brasília, DF, Brazil, 3
pp.
Kaiser, J. 2000. Mercury report backs strict rules. Science
289: 371-372.
Kashima, Y., H. Akagi, Y. Kinjo, O. Malm, J. R. D. Guimarães,
F. Branches, and R. Doi. 2001. Selenium and mercury
concentrations in fish from the lower Tapajos River and
the Balbina Reservoir, Brazilian Amazon. Page 280 in 6th
International Conference on Mercury as a Global Pollutant
(ICMGP). Oct. 15-19, 2001, Minamata, Japan. ICMGP,
Minamata, Japan, 392 pp.
Kehring, H. A., O. Malm, H. Akagi, J. R .D. Guimarães, and J.
P. M. Torres. 1998. Methylmercury in fish and hair
samples from the Balbina Reservoir, Brazilian Amazon.
Environmental Research 77:84-90.
Koifman, S. 2001. Geração e transmissão da energia
elétrica:Impacto sobre os povos indígenas no Brasil.
Cadernos da Saúde Pública 17:314-423.
Laurance, W. F., M. A. Cochrane, S. Bergen, P. M. Fearnside, P.
Delamônica, C. Barber, S. D’Angelo, and T. Fernandes.
2001. The Future of the Brazilian Amazon. Science
291:438-439.
Leino, T., and M. Lodenius. 1995. Human hair mercury levels in
Tucuruí area, state of Pará, Brazil. The Science of the
Total Environment 175:119-125.
Leonel, M. 1987. Contribuição à integração dos componentes
ambientais à avaliação do POLONOROESTE: Hidrelétricas e BR-
429. POLONOROESTE, Porto Velho and Fundação Instituto de
Pesquisas Econômicas (FIPE), Rio de Janeiro, RJ, Brazil, 61
pp.
28
Lobato, E. 1993. "Odebrecht faz maioria das obras do setor
elétrico." Folha de São Paulo, 5 May 1993, Page 1-7.
Luz, S. L. B. 1994. Estudo sobre os culicíneos (diptera:
culicidae: culicinae) crepusculares e noturnos, vetores
potenciais de patógenos de importância medica e
veterinária na Estação Ecologica de Samuel - Rondônia.
Masters thesis in veterinary medicine, Universidade
Federal Rural do Rio de Janeiro, Seropédica, Rio de
Janeiro, Brazil, 94 pp
http://ged.capes.gov.br/AgTest/silverstream/pages/pgRelat
orioTesesCompleto.html?ANOBASE=1994&IDPROGRAMA=3100201300
3P3&SEQUENCIAL=20
Machado, J. A. da C., and R. C. R. Souza. 2003. Fatores
determinantes da construção de hidrelétricas na Amazônia:
Bases para a exigência de indenização. Pages 209-251 in L.
E. Aragón, and M. Clüsener-Godt (eds.), Problemática do Uso
Local e Global da Água da Amazônia. Núcleo de Altos Estudos
Amazônicos (NAEA), Universidade Federal do Pará (UFPA),
Belém, Pará, Brazil. 504 pp.
Machado, O. 2003. “Hidrelétricas no rio Madeira somarão US$ 4
bilhões em investimentos.” Canal Energia 12 March 2003.
http://www.canalenergia.com.br/cenergia/calandra.nsf/0/B0
172016E1B04E2203256CE70059B3E8?OpenDocument.
Malm, O. M. B. Castro, W. R. Bastos, F. J. P. Branches, J. R.
D. Guimarães, C. E. Zuffo, and W. C. Pfeiffer. 1995. An
assessment of Mercury pollution in different goldmining
areas, Amazon Brazil. The Science of Total Environment
175:127-140.
Mera, P. A. S. 1985. Estudos de Ecologia e Controle Ambiental
na Região do Reservatório da UHE Samuel, Convênio
ELN/CNPq/INPA de 01.07.82. Relatório Semestral, Segmento
Zooplancton/Fitoplancton, Período Julho/Dezembro, 1985.
Instituto Nacional de Pesquisas da Amazônia (INPA),
Manaus, Amazonas, Brazil, 157 pp.
Mineração Taboca S/A, and Perfil S/A. 1990. Estudo de Impacto
Ambiental. [UHE Pitinga 2]. Mineração Taboca S/A, Manaus,
Amazonas, Brazil, Irregular pagination.
Monteiro, R. R. 2003. “Furnas quer Itaipu amazônica.”
Jornal do Brasil [Rio de Janeiro]. 12 March 2003.
http://jbonline.terra.com.br/papel/economia/2003/03/12/jo
reco20030312010.html
Munasinghe, M. 1988. Brazil: Proposed second power sector
loan: Social and environmental aspects: Background note.
World Bank. Memorandum dated 19 Sept. 1988 to
participants in briefing session on the Brazilian Second
Electric Power Sector Loan. [available from Bank
Information Center, Washington, DC], 9 pp.
Nogueira, W. 1988. “China importa madeira em toras” Amazonas
29
em Tempo, 9 March 1988, Section 1, Page 8.
Pinto, L. F. 2002. “A derrota de Belo Monte.” O Estado de
S.Paulo 12 November 2002. http:///www.estadao.com.br.
Pires, F. B., and F. Vaccari. 1986. Alta-tensão por um fio.
Ciência Hoje 4(23):49-53.
Porvari, P. 1995. Mercury levels of fish in Tucuruí
hydroelectric reservoir and river Mojú in Amazonian, in
the state of Pará, Brazil. The Science of the Total
Environment 175:109-117.
Quintella, C. 2003. “Furnas apresenta projeto de hidrelétricas
para geração de 7,3 mil MW.” Jornal do Brasil [Rio de
Janeiro], 12 March 2003.
http://jbonline.terra.com.br/pepel/economia/2003/03/12/jore
co20030312010.html.
Revilla Cardenas, J. D. 1986. Estudos de ecologia e controle
ambiental na região do reservatório da UHE de Samuel.
Convênio: ELN/MCT/CNPQ/INPA de 01.07.82. Relatório
Setorial, Segmento: Estimativa da Fitomassa. Período julho-
dezembro 1986. Instituto Nacional de Pesquisas da Amazônia
(INPA), Manaus, Amazonas, Brazil, 194 pp.
Revilla Cardenas, J. D., and I. L. do Amaral. 1986. Estudos de
Ecologia e Controle Ambiental na Região do Reservatório da
UHE de Samuel: Convênio ELN/CNPq/INPA, de 01.07.82.
Relatório Setorial. Segmento Minhocucus, Período
Janeiro/Junho 1986. Instituto Nacional de Pesquisas da
Amazônia (INPA), Manaus, Amazonas, Brazil, 5 pp.
Rosa, L. P., M. A. dos Santos, J. G. Tundisi, and B. M. Sikar.
1997. Measurements of greenhouse gas emissions in Samuel,
Tucuruí and Balbina Dams. Pages 41-55 in L. P. Rosa, and M.
A. dos Santos (eds.), Hydropower Plants and Greenhouse Gas
Emissions. Coordenação dos Programas de Pós-Graduação em
Engenharia da Universidade Federal do Rio de Janeiro
(COPPE/UFRJ), Rio de Janeiro, RJ, Brazil, 111 pp.
Rosa, L. P., R. Schaeffer, and M. A. dos Santos. 1996. Emissões
de metano e dióxido de carbono de hidrelétricas na Amazônia
comparadas às termelétricas equivalentes. Cadernos de
Energia No. 9:109-157.
Santos, H. S. B., O. Malm, and H. A. Kehrig. 2001. Mercury
contamination in Cichla temensis (tucunaré) from Tucuruí
Reservoir, Brazilian Amazon. Page 136 in 6th
International Conference on Mercury as a Global Pollutant
(ICMGP). Oct. 15-19, 2001, Minamata, Japan. ICMGP,
Minamata, Japan, 392 pp.
Schimel, D., and 75 others. 1996. Radiative forcing of climate
change. Pages 65-131 in J. T. Houghton, L. G. Meira Filho,
B. A. Callander, N. Harris, A. Kattenberg, and K. Maskell
(eds.), Climate Change 1995: The Science of Climate Change.
30
Cambridge University Press, Cambridge, U.K. 572 pp.
Schwartzman, S. 1986. Bankrolling Disasters: International
Development Banks and the Global Environment. Sierra Club,
Washington, DC, 32 pp.
Schwartzman, S., and M. Melone. 1987. The multilateral
development banks and the Brazilian electrical energy
sector. Environmental Defense Fund (EDF), Washington, DC,
29 pp. + attachments.
Sondotecnica. nd [2003]. Usina hidrelétrica Samuel.
http://www.sondotecnica.com.br/ambiente_br.html.
Sondotecnica, SA, Rio de Janeiro, RJ, Brazil.
Switkes, G. 2002. Brazilian government pushes ahead with plans
for huge dam in Amazon. World Rivers Review 17(3):12-13.
Tadei, V. P. 1987. Estudos de Ecologia e Controle Ambiental na
Região do Reservatório da UHE Samuel, Convênio
ELN/CNPq/INPA de 13-01-87, Relatório de Avanço e
Acompanhamento Semestral, Segmento Doenças Endêmicas-
Vetores da Malária, Período Janeiro-Julho, 1987.
Instituto Nacional de Pesquisas da Amazônia (INPA),
Manaus, Amazonas, Brazil, 28 pp.
Tundisi, J. G., M. A. Santos, and C. F. S. Menezes. 2003.
Tucuruí reservoir and power plant: Management experience
brief.
http://www.worldlakes.org/uploads/Tucurui_15Jun03.pdf.
van de Vate, J. F. 1996. Hydropower: Potential sources of
greenhouse gases. In Assessment of Greenhouse Gas Emissions
from the Full Energy Chain for Hydropower, Nuclear Power
and Other Energy Sources. Papers Presented at an IAEA
Advisory Group Meeting Jointly Organized by Hydro-Québec
and the International Atomic Energy Agency, Hydro-Québec
Headquarters, Montréal (Canada) 12-14 March 1996.
International Atomic Energy Agency (IAEA), Geneva,
Switzerland. (I1-AG-832.4), Irregular pagination.
Veja. [São Paulo]. 1987. "Um jogo de faraós e empreiteiros." 20
May 1987, Pages 28-37.
Weisser, S. C. 2001. Investigation of the history of mercury
contamination in the Balbina Reservoir, Amazon, Brazil.
Masters thesis in environmental toxicology, Universität
Konstanz, Konstanz, Germany, 66 pp.
WHO (World Health Organization). 1976. Environmental Health
Criteria 1: Mercury. WHO, Geneva, Switzerland.
World Bank. 1981. Brazil: Integrated Development of the
Northwest Frontier. Latin American and Caribbean Regional
Office, International Bank for Reconstruction and
Development (IBRD), The World Bank, Washington, DC. 101 pp.
31
FIGURE LEGENDS
Figure 1 – Brazil’s Legal Amazon region with locations mentioned
in the text.
Figure 2 –- The Samuel reservoir and the state of Rondônia.
Figure 3 - Logs on a barge in Itacoatiara (February 1988)
awaiting loading on ships bound for China. The exception
to Brazil’s prohibition of exporting raw logs opened for
logs from the Samuel Reservoir was reportedly used as cover
for export of logs coming from all over western Amazonia.
32
COVER PHOTOGRAPH CAPTION
Flooded trees in the Samuel reservoir, Rondônia, Brazil (March
2001).
Table 1. Comparison of Environmental Indicators of Existing Dams in Brazilian Amazonia
Dam
Predominent
vegetation
Year
filled River State
Official
Area
(km2)
Installed
capacity
(MW)
Power
density
(W/m2)
Average
depth
(m)
Residence
time
(days) References
Samuel Forest 1988 Jamari Rondônia 540 216 0.40 8.4 143.3
Balbina Forest 1987 Uatumã Amazonas 2,360 250 0.11 4.8 200.4 (a)
Tucurui-I Forest 1984 Tocantins Pará 2,430 3,960 1.63 20.2 51.3 (b, c)
Curuá-Una Forest 1977 Curuá-Una Pará 78 40 0.74 6.1 30 (d, e)
Jatapu Forest 1994 Jatapu Roraima 45 5 0.11 4.4 39.3 (f)
Luis
Eduardo
Magalhães
(Lajeado) Cerrado 2000 Tocantins Tocantins 630 900 1.43 (g)
Manso Cerrado 2000 Manso
Mato
Grosso 387 210 0.54 19.1 502.6 (h)
Coaracy-
Nunes Forest 1975 Araguari Amapá 23 68 2.96 (e)
Pitinga Forest 1984/1990 Pitinga Amazonas 54/81 10/23 0.19/0.28 3.5/5.6 25.7/60.6 (i)
aFearnside 1989a
bFearnside 1999a
cFearnside 2001a
dJunk and de Mello 1987
eTundisi and others 2003
fFearnside and Barbosa 1996
gda Rosa and Cardoso 1993, Coalição Rios Vivos 1999
hBrazil, Furnas 2004
iMineração Taboca S/A and Perfil S/A. 1990
34
Table 2. Comparison of proposed dams in Rondônia
Construction cost
Dam River
Installed
Capacity
(MW)
Reservoir
area
Power
density
Cost
base
date
Cost without
transmission
(US$million)
Unit
cost
(US$/kW) Reference
Samuel Jamari 216 540 0.40 Dec. 91 965 4580.2 (a)
Barão de
Melgaço
Barão de
Melgaço 105 Jun. 86 316 2960.8 (a)
Tabajara Ji-Paraná 725 Jun. 85 721.8 995.3 (a)
Ji-Paraná Ji-Paraná 512 957 0.54 Dec. 91 812.3 1563.5 (b)
Monte Cristo Jamari 58.4 243 0.24 Apr. 78 74.7 1279.3 (c)
Santo Antônio Madeira 3580 138 25.94 Jun. 05 (d)
4000 534.8 (d)
Jirau Madeira 3900 110 35.45 Jun. 05 (d)
aBrazil, ELETROBRÁS 1993, Vol. 2.
bBrazil, ELETRONORTE 1987
cBrazil, ELETRONORTE nd [C. 1985]
dMachado 2003
36
Table 3 – Parameters for greenhouse-gas emissions
Parameter Value Source
ABOVE-WATER DECAY
Area of forest flooded 420 km2 See text.
Average above-ground biomass 425 t/ha Revilla Cardenas 1986,
Revilla Cardenas and Amaral 1986
SURFACE EMISSIONS
Mean surface emission 69.7 mg CH4/m2/day de Lima 2002, p. 90
TURBINES
CH4 concentration at 6.0 mg CH4/liter Measured in March 1989
a depth of 30 m by José Tundisi (Rosa and others 1997,
p. 42)
Average streamflow 366 m3/second Brazil, ANEEL 2003
37
Water use per turbine 171 m3/s Brazil ELETRONORTE nd [C. 1987]
Mean depth to turbine 28 m Assumed equal to normal
intake in 1990 operating level
Mean CH4 concentration 7.5 mg CH4/liter Seasonal cycle adjustment: Fearnside 2002a
based on Galy-Lacaux and others 1997, 1999
Mean reservoir area in 2000 239 km2 (a)
Mean depth to turbine 24 m Assumptions similar to those for 1990
intake in 2000
Mean CH4 concentration at 5.4 mg CH4/liter Same adjustment as for 1990
turbine intake in 2000
Percentage of CH4 60% Assumption of relation to
released at turbines 89% release at Petit Saut
(Galy-Lacaux and others
1997; see Fearnside 2002a)
SPILLWAY
38
Mean depth to spillway 14 m Assumed equal to normal
intake in 1990 operating level
Mean CH4 concentration at 6.4 mg CH4/liter Same adjustment as for turbines
spillway intake in 1990
Percentage of CH4 released 60% Assumption
at spillway
FOSSIL FUEL DISPLACEMENT
Emission of thermal generation 806.1 g CO2 gas Mean of seven studies (range 686-949 g)
equivalent/kWh reviewed by van de Vate (1996)
generated from oil
Transmission loss 3% Brazil, ELETRONORTE nd [C. 1987]
Generation per turbine at normal 44.41 MW/turbine Brazil, ELETRONORTE nd [C. 1987]
operating level
Turbine installation dates: 24 July 1989,
39
27 March 1990,
10 December 1994,
13 October 1995,
2 August 1996
Generation in 1990 605,220 MWh Calculated from water flow and the dates
of turbine installation
Generation in 2000 533,856 MWh Brazil, ANEEL 2001
Global warming potential of CH4 21 t CO2 gas Schimel and others 1996, p. 121 [adopted
equivalent/t CH4 gas by the Kyoto Protocol for the 2008-2012
first commitment period]
-------------------------------------------------------------------------------------------------------
(a) Area that corresponds to the midpoint between the maximum and minimum reservoir volumes (based on
Brazil, ELETRONORTE nd [C. 1987], adjusted proportionally to a full area of 540 km2.
Table 4. Annual emissions of greenhouse gases at Samuel
Emission source
Annual emission
(million t CO2-
equivalent C)
1990 2000
Above water decay 1.13 0.06
Surface 0.08 0.03
Turbines 0.24 0.19
Spillway 0.04 0
Total hydroelectric emission 1.50 0.29
Fossil fuel C displaced - 0.13 - 0.11
Net hydroelectric emission 1.37 0.18
41
Figure 1
42
Figure 2
43
Fig. 3 [low-resolution black & white version. Use high-resolution
black & white for print; color for on-line]
44
Cover photograph [low-resolution black & white version]
... Conversely, Anopheles mosquitos (the vectors of malaria) diminished in abundance after completion of the Tucuruí Dam (Tadei et al. 1991). At the Samuel Dam (in Brazil's state of Rondônia) Culex mosquitos exploded dramatically and Anopheles mosquitos, which were already abundant before construction of the dam, are also believed to have increased (Fearnside 2005). (See Chapter 21 on health). ...
Chapter
Full-text available
This Report provides a comprehensive, objective, open, transparent, systematic, and rigorous scientific assessment of the state of the Amazon’s ecosystems, current trends, and their implications for the long-term well-being of the region, as well as opportunities and policy relevant options for conservation and sustainable development.
... Com relação a submersão da floresta nativa, segundo Fearnside (2005), essa é uma situação comum na Amazônia. O autor afirma que o principal impacto causado por Hidrelétricas na Amazônia brasileira é a inundação de florestas, o que causa danos incalculáveis a biodiversidade e a sociedade, principalmente, as diretamente afetadas e, na região do médio rio Araguari, é um dano ambiental que desencadeia diversos outros danos. ...
Article
Full-text available
Este artigo tem como finalidade expor o diagnóstico e análise dos fatores que provocaram a conflitos socioambientais de comunitários que desenvolviam suas atividades socioeconômicas nas áreas de influência dos reservatórios das Usinas Hidrelétricas Coaracy Nunes, Ferreira Gomes e Cachoeira Caldeirão instaladas e em operação no médio rio Araguari, no município de Ferreira Gomes, no Estado do Amapá. os reservatórios dessas usinas influenciaram e influenciam direta e indiretamente o cotidiano da população de dois municípios: Ferreira Gomes e Porto Grande. O objetivo geral da pesquisa foi diagnosticar e analisar os principais fatores que causaram a conflitos socioambientais em famílias que habitam as comunidades no entorno dos reservatórios das três usinas hidrelétricas no médio rio Araguari. Metodologicamente, se utilizou pesquisa bibliográfica e em sítios da internet (Empresas e Prefeituras), além de pesquisas de campo, com a aplicação de formulários. Os principais resultados da pesquisa inferiram que os principais fatores que contribuíram para o surgimento de conflitos socioambientais dos comunitários emergem pela necessidade de condições mínimas de sobrevivência, como a escassez da pesca e a redução de área para o plantio de culturas de sobrevivência.
... Adentrando nos anos 1980 e considerando a dualidade apresentada, por um lado, pelo sucesso no provimento de energia para os complexos minero-metalúrgicos, principalmente em Tucuruí, e, por outro, pelos colossais problemas socioambientais causados pelos empreendimentos na região, com destaque ao absoluto desastre decorrente de Balbina, sem prejuízos de outros empreendimentos (FEARNSIDE, 1989(FEARNSIDE, , 2001(FEARNSIDE, , 2004(FEARNSIDE, , 2016, é possível notar a incorporação paulatina de um discurso ambiental oficial, por meio de conceitos como viabilidade socioambiental, usos múltiplos e inserção regional nas peças de planejamento setorial (LEMOS, 1999(LEMOS, , 2007. ...
... Global policy therefore attempts to address climate change through an integrated approach of eco-social economics where common metrics (finance) are used to valuate natural capital and ecosystem services (Cisneros-Montemayor et al., 2019). There is debate on the suitability of ecologic valuations with regard to social equity (Peskett et al., 2008;FPP, 2011;Barbesgaard, 2016) and ecologic concerns (Fearnside, 1995;Fearnside, 2005;Fearnside, 2016). However, when carefully implemented, such valuations can aid conservation agendas and help communicate the value of natural capital to policy makers (Cavanagh et al., 2016;Dasgupta, 2021). ...
Article
Full-text available
Cetaceans provide a range of ecosystem services of value to anthropogenic interests. Following the Common International Classification of Ecosystem Services (CICES) these are categorised as regulation and maintenance, provisioning, and cultural values. This study focuses on those of importance to climate change mitigation through regulation and maintenance. Under regulation and maintenance, cetaceans can store, transport, and influence stocks of carbon through: climate regulation through carbon sequestration, enhanced biodiversity and ecosystem potential, and enhanced primary productivity. ‘Climate regulation through carbon sequestration’ can be quantified as carbon fixation through living biomass and ‘whale-falls’. Cetacean populations store significant stocks of carbon in living biomass. After death, sinking whale carcasses, ‘whale-falls’, provide a significant transfer of biomass and nutrients to benthic sediments and support deep sea ecosystems. During their lifespan, cetaceans also disperse nutrients through feeding and excretion both horizontally through the ‘whale-conveyor’ and vertically through the ‘whale pump’. As nutrient limitations hinder phytoplankton growth, these processes can be quantified as the increased potential of phytoplankton carbon fixation from cetacean driven nutrient cycling. Enhanced biodiversity, ecosystem potential, and primary productivity can be quantified as carbon fixation through nutrient cycling. This study reviews the evaluative and valuative techniques used in cetacean ecosystem service research and adapts and applies them to the Joint Cetacean Protocol (JCP) data which details cetacean abundance and distribution in Europe. They are then reviewed with regards to their robustness, application in markets, and in decision making processes. Cetacean populations are estimated to contain 2 Mt C, cycle 60,000 t N yr-1, and impact carbon fluxes by as much as 22 Mt C yr-1 in the survey area. The values highlight key areas for cetacean conservation: the Outer Hebrides, west of south Wales, around the Isle of Man, to the east of England, and to the north-east of the Shetland Islands. There is, however, large uncertainty in the evaluative processes used; nutrient cycling models presented in this study don’t capture removal of excess nutrients, or the values of enhanced biodiversity and ecosystem potential. As such, they are not sufficiently robust to quantify market values but highlight key areas for future research on climate change mitigation through conservation. Key areas of future research include phytoplankton uptake rates of nitrogen and phosphorus in nutrient limited waters, quantification of ‘enhanced biodiversity and ecosystem potential’, and nutrient removal from coastal waters.
... Conversely, Anopheles mosquitos (the vectors of malaria) diminished in abundance after completion of the Tucuruí Dam (Tadei et al. 1991). At the Samuel Dam (in Brazil's state of Rondônia) Culex mosquitos exploded dramatically and Anopheles mosquitos, which were already abundant before construction of the dam, are also believed to have increased (Fearnside 2005). (See Chapter 21 on health). ...
Chapter
Full-text available
This Report provides a comprehensive, objective, open, transparent, systematic, and rigorous scientific assessment of the state of the Amazon’s ecosystems, current trends, and their implications for the long-term well-being of the region, as well as opportunities and policy relevant options for conservation and sustainable development.
Article
Full-text available
O uso da altimetria radar é uma forma de monitorar o nível de grandes rios e lagos da bacia Amazônica através de satélites que tem sido utilizada os últimos dez anos com grande eficiência. Embora estações fluviométricas convencionais possuam grande série histórica de dados, diversas áreas da planície amazônica ainda permanecem sem informações sobre nível do rio. Testes em áreas com estações terrenas têm sido realizados, porém ainda há muitas áreas da Amazônia para serem investigadas. Este artigo avalia o uso de dados altimétricos das missões ENVISAT e JASON-2 em dois rios da bacia Amazônica, o Rio Branco (tributário do rio Negro) e o Rio Jamari (tributário do rio Madeira), em cujas bacias há pouca informação altimétrica validada. O processamento dos dados altimétricos foi realizado através do programa Virtual ALtimetry Station (VALS). Dados fluviométricos de campo da rede hidrométrica da Agência Nacional de Águas (ANA) foram usados para identificar o ciclo hidrológico e avaliar os níveis de água derivados dos radares altimétricos nos rios analisados. Como resultado, encontrou-se elevada concordância entre as medições de campo e aquelas derivados dos radares altimétricos. As análises sugerem que dados do nível de água derivados da altimetria radar em rios represados possuem maior precisão em relação aos dados derivados de rios não represados.
Article
Este estudo analisou a fragmentação florestal por meio de métricas de ecologia da paisagem na bacia Hidrográfica do Rio Ji-Paraná, Estado de Rondônia. Utilizada a classificação do uso e ocupação do solo disponibilizada pelo MapBiomas. As métricas de ecologia da paisagem foram calculadas utilizando o software FRAGSTATS. A fragmentação florestal analisada consistiu em quatro classes de acordo com o tamanho dos fragmentos, sendo as classes: menor que 5 ha, entre 5 e 10 ha, entre 10 e 100 ha e maior que 100 ha, comparando os anos de 1985 e 2019. Ainda, no estudo verificou-se o NDVI anual médio para cada classe de tamanho de fragmento a partir de dados do Moderate Resolution Imaging Spectroradiometer com 500 metros de resolução espacial e 16 dias de revisita usando a plataforma Google Earth Engine. A análise da fragmentação da paisagem mostrou um aumento no número de fragmentos com tamanho menor que 5 ha de 1985 para 2019. Além disso, apresentou aumento no número de áreas nucleares disjuntas para todas as classes de fragmentos e mostrou um aumento da distância entre fragmentos maiores que 100 ha. Quanto ao NDVI, todos os tamanhos de fragmentos apresentaram diferenças significativas quando comparados aos fragmentos com tamanho maior que 100 ha, que apresentaram os maiores valores.
Article
Full-text available
Zoonotic spillover is a phenomenon characterized by the transfer of pathogens between different animal species. Most human emerging infectious diseases originate from non-human animals, and human-related environmental disturbances are the driving forces of the emergence of new human pathogens. Synthesizing the sequence of basic events involved in the emergence of new human pathogens is important for guiding the understanding, identification, and description of key aspects of human activities that can be changed to prevent new outbreaks, epidemics, and pandemics. This review synthesizes the connections between environmental disturbances and increased risk of spillover events based on the One Health perspective. Anthropogenic disturbances in the environment (e.g., deforestation, habitat fragmentation, biodiversity loss, wildlife exploitation) lead to changes in ecological niches, reduction of the dilution effect, increased contact between humans and other animals, changes in the incidence and load of pathogens in animal populations, and alterations in the abiotic factors of landscapes. These phenomena can increase the risk of spillover events and, potentially, facilitate new infectious disease outbreaks. Using Brazil as a study model, this review brings a discussion concerning anthropogenic activities in the Amazon region and their potential impacts on spillover risk and spread of emerging diseases in this region.
Book
Full-text available
Fearnside, P.M. (ed.) 2022. Destruição e Conservação da Floresta Amazônica, Vol. 1. Editora do INPA, Manaus. 356 pp. (no prelo).
Article
Full-text available
Este estudo analisou a fragmentação florestal por meio de métricas de ecologia da paisagem na bacia Hidrográfica do Rio Ji-Paraná, Estado de Rondônia. Utilizada a classificação do uso e ocupação do solo disponibilizada pelo MapBiomas. As métricas de ecologia da paisagem foram calculadas utilizando o software FRAGSTATS. A fragmentação florestal analisada consistiu em quatro classes de acordo com o tamanho dos fragmentos, sendo as classes: menor que 5 ha, entre 5 e 10 ha, entre 10 e 100 ha e maior que 100 ha, comparando os anos de 1985 e 2019. Ainda, no estudo verificou-se o NDVI anual médio para cada classe de tamanho de fragmento a partir de dados do Moderate Resolution Imaging Spectroradiometer com 500 metros de resolução espacial e 16 dias de revisita usando a plataforma Google Earth Engine. A análise da fragmentação da paisagem mostrou um aumento no número de fragmentos com tamanho menor que 5 ha de 1985 para 2019. Além disso, apresentou aumento no número de áreas nucleares disjuntas para todas as classes de fragmentos e mostrou um aumento da distância entre fragmentos maiores que 100 ha. Quanto ao NDVI, todos os tamanhos de fragmentos apresentaram diferenças significativas quando comparados aos fragmentos com tamanho maior que 100 ha, que apresentaram os maiores valores.
Book
Full-text available
Fearnside, P.M. 1986. Human Carrying Capacity of the Brazilian Rainforest. Columbia University Press, New York, NY, U.S.A. 293 pp. ISBN 0-231-06104-8
Book
Full-text available
Fearnside, P.M. 1989. Ocupação Humana de Rondônia: Impactos, Limites e Planejamento. Relatórios de Pesquisa No. 5, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Brasília, DF, Brazil. 76 pp.
Article
Full-text available
Over the 1978-1988 period in Brazilian Amazonia, forest was lost at a rate of 22 000 km2 yr-1 for 1989-1990 and 11 000 km2 yr-1 for 1990-1991. The reduction in the rate since 1987 has mostly been due to Brazil's economic recession rather than to any policy changes. The number of properties censused in each size class explains 74% of the variation in deforestation rate among the nine Amazonian states. Multiple regressions indicate that 30% of the clearing in 1991 can be attributed to small farmers, and the remaining 70% to either medium or large ranchers. The social cost of reducing deforestation rates would therefore be much less than is implied by frequent pronouncements that blame "poverty' for environmental problems in the region. -Author
Article
Full-text available
vol. 4 issue8 O " esquecimento de nomes próprios" na psicopatologia da vida cotidiana, de Sigmund Freud Avanços recentes na pesquisa em fixação biológica de nitrogênio no Brasil author index subject index articles search
Article
Full-text available
When roads open up, man moves with them. In Brazil, highway construction has been associated with migration which has in turn caused deforestation as the migrants seek to make their new homes habitable as well as make a living. Interestingly, once migrants have been installed for a time, deforestation slows as they devote more effort towards conserving their resources. Satellite photographs have been able to track some of these changes. -Author
Article
Full-text available
Methane, carbon dioxide, and hydrogen sulfide emissions from the hydroelectric dam of Petit Saut on the Sinnamary River in French Guyana have been measured over a 2 year period. Since the beginning of the reservoir filling (January 1994), 300 km2 of tropical forest have been submerged. Emissions of CH4 by diffusion and by bubbling into the atmosphere or by degassing of the water released into the river, as well as the stock of dissolved gases in the lake, and their temporal evolutions were determined. Maximum emissions of 800 t CH4 per day were reached in February 1995, corresponding to dissolved CH4 concentrations of 14 mg4L−1 in the water column. The biological oxidation of methane results in a strong oxygen consumption in lake and river waters. Total emissions of CH4 and CO2 from January 1994 to December 1995 were calculated from the whole data set, which also allows us to calculate the total carbon loss since reservoir filling. About 10% of the carbon stored in soil and vegetation was released in gaseous form within 2 years.