Summary
The Human Pressure
Analysis examines the extent
of human
activities that lead to forest
degradation and land conversion in the
Brazilian Amazon. The analysis
compiles a comprehensive set of
existing datasets. While at a rough
coarse scale (1:1,000,000), the analysis
provides an indication of the dimensions
of the human pressure allowing users to
distill insights for land-use planning
and monitoring. The analysis integrates
data for five
indicators of human activities:
deforestation,
urban zones,
agrarian
reform settlements,
forest fires,
and
mining.
At the time of the analysis no
comprehensive information on roads and
logging was available. Thus, the map of
human pressure did not factor in these
two important indicators. However,
understanding the crucial role of these
two factors, we did examine the
relationship among human pressure,
roads, and
logging using available information (Figure 1).
|
Figure 1 - Human
Pressure in the Brazilian Amazon. |
For the purpose of this study, human
pressure is classified under two
categories (Figure 2):
- Areas under
settlement - including areas
classified as deforested; urban
centers and lands allocated for
agrarian reform settlements. In this
areas there is a continuum of human
presence.
- Areas under
incipient human pressure - including
areas surrounding forest fires,
areas licensed for mining and
location of logging permits. In
areas under incipient human pressure
people may remain only temporarily,
but in some other cases, people will
likely settle in the future.
This differentiation not
only permits analysis of areas settled
or under pressure by adjacent
settlements, but of areas that could be
subjected to increasing pressure in the
future.
|
Figure 2 - Areas classified as under
Human Settlement include deforested
areas, urban zones, and areas of
Agrarian Reform Settlements. Areas
classified as under Incipient Human
Pressure include areas under human
influence based on the incidence of
fires, areas licensed for mining, and
locations where logging has been
authorized. |
Back to the Top
Study Area
The geographic area for this analysis
is the Brazilian Amazon (Figure 3),
which overlaps with two other geographic
areas, the Amazon Basin and the Legal
Amazon. The Amazon Basin
extends over 6.8
million square km through Venezuela,
Colombia, Ecuador, Peru, Bolivia,
Brazil, French Guiana, Guyana and
Suriname. Almost two thirds of the Basin
(4.1 million square km) lie within
Brazil's boundaries, this portion is
known as the Brazilian Amazon. The Legal
Amazon (black line) is a 5.1 million
square km administrative unit that
encompasses the Brazilan states of Acre,
Amazonas, Roraima, Amapá, Pará, Rondônia,
Mato Grosso, Tocantins and Maranhão.
Portions of the states of Maranhão,
Tocantins and Mato Grosso are outside of
the Amazon Basin.
|
Figure 3 - The
Amazon Basin (shaded), the Legal Amazon
(black line) and the Brazilian Amazon
(shaded area within the Legal Amazon). |
Back to
the Top
Input Datasets
Vegetation and
Deforestation
To create a baseline for
the study, the map of deforestation
produced by INPE in 2003 was
superimposed on the 1997 vegetation map
produced by IBGE. The deforestation map
shows deforestation as of 2001 and was
produced using three color composition
bands (red, infrared and mid infrared)
from a Landsat sensor at scale of
1:250,000. Deforested areas greater than
6.25 ha were digitized directly from the
computer screen using INPE's image
processing software (SPRING). Accuracy
of the process was estimated at 95%. On
the other side, the vegetation map shows
forested areas (i.e. dense, open and
transition forests) and other non-forest
native vegetation (i.e. scrublands and
savannas) (Figure 4).
|
Figure 4 - Forest Vegetation and
Deforestation. The baseline for the
Human Pressure analysis was a dataset
that combines vegetation (IBGE
1997) and deforestation as of 2001 (INPE
2003). |
Back to
the Top
Urban Zones
Using the IBGE 1999
dataset of populated places, 20km-wide
buffers were drawn around the location
points of municipal seats to estimate
the area under urban influence. The 20km
threshold is an arbitrary threshold as a
proxy to estimate human influence based
on field observations in areas adjacent
to urban centers where population is
growing and land uses are expanding
rapidly (Figure 5).
|
Figure 5 - Urban Zones. Point location
of municipal seats in the Legal Amazon
with surrounding 20-km buffer. |
Back to
the Top
Agrarian Reform Settlements
The analysis incorporates a map
produced by INCRA showing Agrarian
Reform Settlements established until
2002. Since the late 1970s, landless and
urban poor have pressured the government
for lands. As of 2002, 528,571 families
had been granted land holdings through
approximately 1,600 projects of Agrarian
Reform Settlement. Success of Agrarian
Reform Settlements is mixed (e.g.
Weiss 2002;
Da Silva 2001).
Many families are attracted by initial
earnings from subsidies (e.g. food
allowances, housing, and credit at
interest rates below market) and timber
sales. However, after the depletion of
timber resources, income from cattle
ranching and subsistence crops
decreases, partly because land use for
these activities tends to be extensive
and the family plots are relatively
small (50-100 ha). As a result, many
families abandon or sell their lots and
either seek new areas or migrate to
urban centers. Much of the land left
behind becomes consolidated in
large-scale properties, which tend to be
more efficient and profitable (Figure
6).
Figure 6 - Area and
location of Agrarian Reform Settlements
as of 2002. |
Back to
the Top
Forest Fires
In the Brazilian Amazon,
forest fires are often associated with
the begining of human occupation
including land clearing and harvesting
of forest products. Using INPE's data,
the analysis incorporates evidence of
forest fires, as identified by satellite
(hotpixels), as a proxy indicator of
human pressure. INPE's dataset displays
1.1 square km "hot" pixels that are
registered as the AVHRR sensor passes
overhead and detects areas of heat that
are at least 50 meters. Although
hotpixels accurately detect incidence of
fires, they can be up to ~3 km off
because it is less precise to indicate a
fire at the pixel border (Setzer
and Malingreau 1996).
Hotpixels may indicate:
- areas that were cleared and
burned but not included in INPE's
deforestation map because of their
size (less than 6.25 ha);
- burned areas, such as logged
forests into which wildfires
subsequently escaped from nearby
pastures of shifting cultivation
plots.
Available data on hotpixels between
1996 and 2002 was collected for this
analysis. Hotpixels outside non-forested
areas (as classified by INPE in 2001)
were excluded, and hotpixels between
1996 and 1999 were designed as "old"
while those from 2000 to 2002 were
classified as "recent". A 10km buffer
around each hotpixel was drawn as a
coarse proxy to estimate human influence
around fires. This threshold correspond
to the 10km found to be the maximum
distance a hunter would search for the
most profitable prey from a given point
of access within the forest (Peres
and Terborgh 1995) (Figure 7).
Figure 7 -
Incidence of Forest Fires. Old and
recent hotpixels surrounded by 10-km
wide buffer. Data on incidence of fires
below 1 degree North does not exist
prior 1999. |
Back to
the Top
Mineral Reserves
and Areas Designated for Mineral
Prospection
The analysis of Human
Pressure incorporates a dataset produced
by the DNPM in 2001, showing areas
licensed for mineral prospection up to
1998 (Capobianco
2001). The dataset also includes
mineral reserves established by the
Brazilian Government to accomodate
part-time miners (wildcat miners) in the
western portion of the State of Pará.
The areas licensed for mineral
prospection are not necessarily active,
but they may become so if minerals of
interest are found (Figure 8).
Figure 8 - Areas
for licenses for mineral prospection up
to 1998. |
Back to
the Top
Authorized
Logging Operations
While timber harvesting
has been an important catalyst for human
settlement and deforestation, there is
no comprehensive map of areas impacted
by this land use in the Brazilian
Amazon. Existing datasets are
temporarily and spatially incomplete. To
identify parts of the Brazilian Amazon
that are potentially being logged,
Greenpeace Brazil mapped the location of
580 forestry operations in which logging
had been authorized as part of the
management plans registered with IBAMA
in 2000 (Figure 9).
Figure 9 - Location
of forestry operations in which logging
is authorized in management plans
registered with IBAMA in 2000. |
Back to
the Top |
References
Capobianco, J.P. R. Ed. 2001.
Biodiversidade na Amazônia Brasileiera:
avaliação e a ções prioritárias para a
conservação, uso sustentável e
repartição de benefícios. Estação
Liberdade and Instituto Socioambiental,
São Paulo, SP, Brasil.
Da Silva, J. G. 2001. Ainda
precisamos de reforma agrária no Brasil?
Ciência Hoje 170 (abril):
61-63.
Instituto Brasileiro de Geografia e
Estatística [IBGE]. 1997.
Diagnostico ambiental da Amazonia Legal.
Rio de Janeiro, RJ, Brasil.
Instituto Nacional de Pesquisas
Espaciais [INPE]. 2003.
Monitoramento da floresta amazônica
brasileira por satélite:
projeto Prodes. Instituto Nacional
de Pesquisas Espaciais. Ministry of
Science and Technology, Brasília, DF,
Brasil. Online at:
http://www.obt.inpe.br/prodes/
(3/10/05)
Peres, C. and J. Terborgh. 1995.
Amazonian nature reserves: an analysis
of the defensibility status of existing
conservation units and design criteria
for the future. Conservation Biology
1: 34-46.
Setzer, A. W. and J. P. Malingreau.
1996. AVHRR monitoring of vegetation
fires in the tropics: toward the
development of a global product. Pp.
25-39, in J. S. Levine, ed., Biomass
burning and global change: remote
sensing, modeling and inventory
development, and biomass burning in
Africa.
Weiss, J (Coordenador). 2002.
Proposta de ajuste de políticas agrárias
que estimulem o manejo florestal
sustentável na Amazônia Legal.
Ministério do Meio Ambiente, Brasília,
DF, Brasil.
Back to
the Top |
