Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
1
ECOLOGICAL FOOTPRINT OF ENERGY DEVELOPMENT IN EASTERN VENEZUELA’S HEAVY OIL BELT
By
CHRIS W. BAYNARD
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2008
3
To my wife Ana for her infinite patience, support and encouragement during this long process; to my children, Nico and Isabela, who will be glad I am no longer a student; and to my parents
Sylvia and Whaley for their help and encouragement
4
ACKNOWLEDGMENTS
I thank my wife, Ana; my children; and my parents for their support and patience. Jorge
Febles, Gary Harmon, Allen Tilley, Mark Workman and the Department of World Languages at
the University of North Florida gave me the flexibility to pursue my studies at the University of
Florida while working there. I thank Dr. Michael W. Binford at the University of Florida’s Dept
of Geography for help in overseeing this project. I also thank Drs. Jane Southworth, Tim Fik,
Julie Silva, Eric Keys, Grenville Barnes and Terry McCoy at the University of Florida for help in
data acquisition, analysis, ideas and support. Dr. David Lambert and Robert Richardson at the
University of North Florida offered help and suggestions with the GIS and remote sensing
components of this project. Dan Richard at the University of North Florida and Dr George
Casella at the University of Florida helped with statistical analysis. Matt Marsik, Ph.D.
Candidate at the University of Florida’s Dept of Geography also helped with remote sensing data
analysis, while Andrés Guhl proved to be a very helpful and congenial colleague.
I also thank Carlos Guerra and Rómulo Medina, at Operadora Cerro Negro in Venezuela
for their help in better understanding the dynamics of heavy oil production and for logistical
support. María Madalena Godoy Mello at Brazil’s National Institute for Space Research (INPE)
provided CBERS 2 satellite images. Representatives from ExxonMobil Venezuela, PDVSA,
Chevron Venezuela, Total, Shell Venezuela, BP Venezuela, ConocoPhillips USA, and Marathon
met and/or spoke with me about heavy oil operations in eastern Venezuela, provided gray
literature, reports and maps, and insights into corporate social responsibility practices in the
petroleum industry. ExxonMobil Venezuela also provided logistical support during fieldwork.
Thanks also to Alicia Moreau, Deud Dumith and José Acosta at the Instituto Geográfico de
Venezuela Simón Bolívar for their support and map data. Víctor and Rosella Pérez and Dr. Elisa
Pérez provided housing and logistical support. Finally, thanks to Tom Knode, Halliburton
5
Energy Services, and Mark Shemaria, Tidelands Oil Production Company, who served as SPE
E&P Environmental Safety Conference 2007 organizers and encouraged me to present my
research to the petroleum industry.
6
TABLE OF CONTENTS page
ACKNOWLEDGMENTS ...............................................................................................................4
LIST OF TABLES...........................................................................................................................9
LIST OF FIGURES .......................................................................................................................10
LIST OF ABBREVIATIONS........................................................................................................12
ABSTRACT...................................................................................................................................13
CHAPTER
1 LAND-COVER CHANGE AND NATURAL RESOURCE USE ........................................15
Introduction.............................................................................................................................15 Natural Resource Use, Land-use Land-Cover Change and Political Ecology .......................16 Renewable and Nonrenwable Resource Extraction................................................................18 Heavy Oil Production in Eastern Venezuela ..........................................................................19 Environmental Performance Variability in the Heavy Oil Belt..............................................20
2 LAND-USE AND LAND-COVER CHANGE RESULTING FROM RENEWABLE AND NONRENEWABLE RESOURCE EXTRACTION: A REVIEW ...............................22
Introduction.............................................................................................................................22 Background......................................................................................................................24 LUCC and Political Ecology (Theoretical Framework)..................................................24 Land-Cover Modification and Conversion......................................................................27
Specific Issues: Renewable Natural Resources ......................................................................30 Agriculture.......................................................................................................................30 Logging............................................................................................................................33
Type of logging ........................................................................................................33 Fire ...........................................................................................................................34 Other factors .............................................................................................................36
Specific Issues: Nonrenewable Natural Resources.................................................................37 Gold Mining ....................................................................................................................37
Land degradation......................................................................................................37 Regeneration.............................................................................................................39 Mercury pollution.....................................................................................................40 Solutions...................................................................................................................41
Oil Exploration and Production.......................................................................................42 Wildlife: caribou ......................................................................................................42 Seismic lines.............................................................................................................44
Roads ...............................................................................................................................44 Conclusion ..............................................................................................................................46
7
3 DO NATIONAL OIL COMPANIES PRODUCE GREATER ENVIRONMENTAL ALTERATIONS THAN MULTINATIONAL OIL COMPANIES? THE CASE OF VENEZUELA’S HEAVY OIL BELT ...................................................................................49
Remote Sensing, Geographic Information Systems (GIS) and Petroleum.............................52 Materials and Methods ...........................................................................................................54
Site Description ...............................................................................................................54 Data Description (Concessions and Control Groups) .....................................................55 Time Frame .....................................................................................................................56 Image Processing and GIS Procedures............................................................................58 Petroenergy Maps............................................................................................................59 Control Areas...................................................................................................................60 Normalized Difference Vegetation Index (NDVI)..........................................................60 Change-Detection: 1990 to 2000.....................................................................................60 GIS Techniques: the 2005 Petroscape.............................................................................61
Petroscape density ....................................................................................................62 Edge-effect zones .....................................................................................................63 Core-areas.................................................................................................................64 Agriculture ...............................................................................................................64
Rainfall Patterns ..............................................................................................................64 Results and Discussion ...........................................................................................................65
Vegetation Change ..........................................................................................................65 Petroscape........................................................................................................................66 Nonpetroscape .................................................................................................................68 Petro and Nonpetroscape.................................................................................................69
Conclusion ..............................................................................................................................71
4 SPATIAL ANALYSIS AND STATISTICAL CORRELATIONS AMONG PETROSCAPE FEATURES IN THE HEAVY OIL BELT ..................................................95
Introduction.............................................................................................................................95 Production in the Heavy Oil Belt ....................................................................................97 Land-Use Land-Cover Change........................................................................................98 Expected Findings ...........................................................................................................99
Methods ................................................................................................................................103 LUCC Related to the Petro and Nonpetroscape ............................................................104 The Coefficient of Determination .................................................................................105 Nonparametric Statistical Analysis ...............................................................................106
Results and Discussion .........................................................................................................108 Determining the R² Coefficient .....................................................................................108 Nonparametric Statistical Analysis ...............................................................................108
European and US results ........................................................................................109 Local results ...........................................................................................................109
Conclusion ............................................................................................................................111
8
5 LAND-COVER CHANGE AND NONRENEWABLE NATURAL RESOURCES ..........124
Summary...............................................................................................................................124 Future Work..........................................................................................................................127
LIST OF REFERENCES.............................................................................................................131
BIOGRAPHICAL SKETCH .......................................................................................................156
9
LIST OF TABLES
Table page 3-1 Venezuelan heavy oil belt strategic associations, company ownership by percentage,
syncrude (synthetic oil) production and upgraded quality 2005........................................93
3-2 Change-detection measures showing amounts of positive and negative change per site in km² and as a percentage of the site area. .................................................................93
3-3 2005 petroscape measures for the six study sites...............................................................94
3-4 2005 nonpetroscape measures for the six study sites.........................................................94
3-5 Overall rankings for land-cover change and petroscape features in the 6 sites. ................94
4-1 Venezuelan heavy oil belt strategic associations (concessions), company participation by percentage and upgraded syncrude quality............................................115
4-2 The R² coefficients for seven petroscape-related land-cover change measures and the three consortia of companies operating in the HOB........................................................116
4-3 The R² coefficients of determination between three types of company participation and 16 LUCC measures in the HOB................................................................................117
4-4 Kendall’s tau-b correlation coefficients showing the relationship between Local, European and US dominated concessions and seven important LUCC measures related to petroscape expansion .......................................................................................118
4-5 Kendall’s tau-b correlation coefficients and R² coefficients showing the relationship between concessions that were PDVSA-dominated, European-dominated, and US-dominated and important landscape-change measures ....................................................119
4-6 Petro and nonpetroscape data values for the four HOB concessions...............................122
10
LIST OF FIGURES
Figure page 3-1 Venezuela, the heavy oil belt and concessions, and surrounding countries 2005. ............73
3-2 Venezuela’s heavy oil belt, its four concessions: Sincor, Petrozuata, Ameriven, and Cerro Negro, and the upgrading/refinery plant called Jose located on the coast of Anzoátegui state, 2005.......................................................................................................74
3-3 Vegetation-cover change across all six sites (three concessions and three controls) between 1990 and 2000.. ...................................................................................................75
3-4 Increases in Normalized Difference Vegetation Index (NDVI) across all six sites between 1990 and 2000. ....................................................................................................76
3-5 NDVI losses and gains for all six sites between 1990 and 2000.. .....................................77
3-6 Gains and losses of the Normalized Difference Vegetation Index (NDVI) for the three concessions ...............................................................................................................78
3-7 Rainfall amounts (in mm) for Ciudad Bolívar for the October, November and December dry months preceding the 1990 and 2000 Landsat images used in this study...................................................................................................................................81
3-8 Decrease in NDVI between 1990 and 2000 overlain by 2005 petroscape features...........82
3-9 Edge-effect zones, or areas where significant ecological effects extend...........................83
3-10 Core-areas show size and distribution of the patches of land that remain after dissolving the edge-effect zones in figure 3-7.. .................................................................84
3-11 Linear nonpetroscape 2005.. ..............................................................................................85
3-12 Edge-effect nonpetroscape zones.......................................................................................86
3-13 Core-areas that remain after nonpetroscape features are dissolved from each site. ..........87
3-14 Petro and nonpetroscapes across all 6 sites........................................................................88
3-15 Agricultural patterns in the three concessions and the three control groups in 1990.. ......89
3-16 Agricultural patterns in the three concessions and the three control groups in 2000.. ......90
3-17 Amount of agricultural land (in km²) found in the three concessions and three control groups in 1990 and 2000....................................................................................................91
3-18 Future heavy oil development in Venezuela’s Heavy Oil Belt..........................................92
11
4-1 Venezuela’s heavy oil belt and the four current operations.............................................113
4-2 Oil company participation in the HOB in 2005 and under the new changes implemented in mid-2007. ...............................................................................................114
4-3 Potential E&P area within the heavy oil belt by 2030.. ...................................................115
12
LIST OF ABBREVIATIONS
BP British Petroleum
CVG Corporación Venezolana de Guayana
E&P Exploration and production: two key activities of the petroleum industry
FVI Fuel value index, a measure of the physical properties of wood as a fuel source
GIS Geographic information systems
GLCF Global Land Cover Facility, University of Maryland satellite image repository
GPS Global positioning system
HOB Heavy oil belt, the zone of heavy oil located in central and eastern Venezuela.
INPE Instituto Nacional de Pesquisas Espaciais, Brazil’s National Institute for Space Research
LUCC Land-use and land-cover change
M&E Management and Excellence, a sustainability ranking firm
MISR Multi-angle Imaging SpectroRadiometer
MNOC Multinational oil company
NDVI Normalized difference vegetation index
NOC National oil company
PDVSA Petróleos de Venezuela, the state-owned oil company
PSI Pacific sustainability index
13
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
ECOLOGICAL FOOTPRINT OF ENERGY DEVELOPMENT IN EASTERN VENEZUELA’S
HEAVY OIL BELT
By
Chris W. Baynard
August 2008 Chair: Michael W. Binford Major: Geography
Nearly all land-use and land-cover change (LUCC) studies concern conversions of land
related to two renewable resource activities, agriculture and forestry. LUCC related to
nonrenewable resource development is nonetheless another major agent of environmental change
that has been studied very little. This work examines specific land-use and land-cover changes
resulting from the production patterns of four petroleum operations between 1990 and 2005 in
eastern Venezuela’s heavy oil belt, an area containing some of the largest worldwide deposits of
heavy crude oil which only recently have been developed.
The four concessions consisted of the following consortia of public-private partnerships:
Sincor (PDVSA, Total, and Statoil); Petrozuata (PDVSA and ConocoPhillips); Ameriven
(PDVSA, ConocoPhillips and Chevron); and Cerro Negro (PDVSA, ExxonMobil, and British
Petroleum, or BP). Since mid-2007 the Venezuelan government has increased its role in these
operations and plans to expand production over the next two decades by establishing 27 new
concessions mostly with other state-owned oil companies. These activities will affect 18,000 km²
of dry tropical savannas and forest, a natural region that is already threatened.
This work incorporates remote sensing and GIS techniques to calculate indicators of
environmental alteration. The objective is to determine which operations produced the most and
14
least landscape disturbances based on six important landscape measures: amount of vegetation-
cover loss (based on a vegetation index); petroscape density; edge-effect zones; core areas;
number of rivers crossed; and well density. It also measures nonpetroscape features related to
agriculture and transportation. This work also uses statistical correlation analysis to determine
the relationship between observed LUCC and the type of public-private partnership.
Findings show that Petrozuata had the highest overall rank, meaning it had the highest
level of landscape disturbances. It was followed by Sincor, Cerro Negro and Ameriven.
Statistical results, however, do not show sufficient correlations between the type of company mix
and petroscape-related alterations. These findings contribute to emerging research using
geospatial technologies to monitor LUCC in heavy oil production zones and can be used to
assess environmental performance of petroleum operations, map evolving landscape changes in
current operations and plan for the establishment of new concessions.
15
CHAPTER 1 LAND-COVER CHANGE AND NATURAL RESOURCE USE
Introduction
This Alterations to land cover affect central components of earth system functions (Fujita
and Fox, 2005; Lambin et al., 2003; Read and Lam, 2002, Southworth, 2004; Soares-Filho et al.,
2004) and these changes are proceeding at a pace, magnitude and scale unprecedented in human
history (Jensen 2005). Understanding the factors that contribute to changes in land-cover and
land-use (LUCC) practices are of prime importance since these changes influence climate,
ecosystem goods and services, gross carbon fluxes and the vulnerability of people and places at
various spatial and temporal scales (Lambin et al., 2003; Fujita and Fox, 2005; Southworth,
2004; Asner et al., 2005; Rindfuss et al., 2007; Keys and McConnell, 2005).
Land-change science is a central component of geography since (in part) it examines
spatial patterns of human-environment interactions on the landscape. This complex system of
interactions (Veldkamp and Lambin, 2001) involves political, economic, cultural, social,
technological, institutional and environmental variables operating under various spatial and
social scales and combines natural and social science methods (Folke et al., 2007). As most
LUCC studies demonstrate, understanding social-ecological system dynamics involves teams of
researchers comprised of specialists in given disciplines (see: Lambin et al., 2001; Rindfuss et
al., 2007; Wassenaar et al., 2007; Gibbons et al., 2008; Soares-Filho et al., 2006; Caldas et al.,
2007, Pielke et al., 2007; Folke et al., 2007).
The complexity of understanding social-ecological system integrations (Folke et al., 2007),
creates two main problems. One, findings from a particular study tend to be case-specific; not
applicable to a wide range of situations and settings (Rindfuss et al., 2007; Keys and McConnell,
2005; Veldkamp and Lambin, 2001). Two, it is difficult to reconcile natural and social science
16
systems and methods, particularly across spatial scales (Veldkamp and Lambin, 2001)—what
Folke et al. (2007) refer to as the problem of “fit”.
Thus land-use researchers “often confine themselves to either a single process (e.g.
deforestation), or a single discipline (e.g. economic models)” (Veldkamp and Lambin, 2001, pp.
3). Since this project is the product of one researcher, I too confine myself to a single process:
land-cover change resulting from nonrenewable natural resource use in a dry-tropical forest zone
in eastern Venezuela. At the same time my field, geography, which itself is cross-disciplinary,
justifies that I draw here from land-use land cover-change science, political ecology approaches
and sustainability research.
Natural Resource Use, Land-use Land-Cover Change and Political Ecology
My interest lies not only in understanding what types of landscape changes result from
natural resource use, since they also lead to social alterations and subsequent land management
responses. I am also intrigued by the concept of sustainability—that is the interplay between
ecological, economic and socio-cultural dimensions taking place at various spatial and temporal
scales as well as organizational and institutional levels (Folke et al., 2007; Ellsworth, 2002). This
interplay “affects the chances that environmental and social conditions will indefinitely support
human security, well-being, and health” (McMichael et al., 2003, 1919).
Obviously some practices and institutions are more sustainable than others; that is some
natural resource use patterns (shaped by the nature of the resource itself, topography and the
rules governing them) create larger and more permanent changes on the landscape that in turn
cause important and long-term alterations such as habitat fragmentation; land degradation; land,
air and water pollution; soil erosion, increased sedimentation and biodiversity loss (Armenteras,
et al., 2006; Ferraz et al., 2006; Schneider and Dyer, 2006; McMichael et al., 2003; Dietz et al.,
2003; Cumming and Barnes, 2007). These shifts can lead to regional and global climate changes,
17
thus affecting earth system functions and the organisms that depend on them (Pielke et al., 2007;
Keys and McConnell, 2005).
What socioeconomic and socioecological factors most affect this capacity to create
sustainable natural resource use, or to at least reduce the negative environmental (and therefore)
social effects resulting from such practices? Secondly, what are some of the incentives for land
managers, including organizations, to behave in a more sustainable manner (Ostrom et al.,
1999)?
Political ecology theory proposes that human-environment interactions—as related to land
and natural resource use—consist of economic, political and environmental processes that
together produce particular landscapes (Robbins, 2004). In this context, humans react to
economic opportunities they perceive the natural system offers them in order to meet local
consumption needs or market demands. In essence it focuses on the political economy (the
relationship between politics and economics) of human-environment interactions. These
interactions, however, create competition, resistance and conflict among land managers to
access, use and exclude others from valuable land and natural resources—involving institutional
structures such as land and resource tenure, as well as social actors possessing various degrees of
power (Schmink and Wood, 1992, Folke et al., 2007; Wannasai and Shrestha, 2008; Rindfuss et
al., 2007; Simmons, 2002; Simmons et al., 2007; Ankersen and Barnes, 2004; Kennedy, 2003).
This competition for land and resources poses problems that require political solutions in terms
of policy and its implementation (Bryant and Bailey, 1997), underlining the importance that
political/governance structures have in this process. As Kennedy (2003, pp. 1861) notes on
sustainability: “The big question in the end is not whether science can help. Plainly it could.
18
Rather, it is whether scientific evidence can successfully overcome social, economic and political
resistance.”
Finally, human-environment interactions create landscape changes that may be temporary
or permanent; resulting in alterations or modifications (Veldkamp and Lambin, 2001) which can
then affect subsequent social adaptive behavior as well as ecosystem responses (Folke et al.,
2007).
Renewable and Nonrenwable Resource Extraction
Chapter 2 is a literature review of the environmental effects resulting from renewable and
nonrenewable natural resource extraction. This overview concerns LUCCs and secondary
outcomes that result from agriculture and logging (renewable natural resources) and gold mining
and petroleum exploration and production (E&P) (nonrenewables). A good deal of this literature
concentrates on tropical regions, particularly Amazonia (Soares-Filho et al., 2006; Flamenco-
Sandoval et al., 2007; Caldas et al., 2007; Nepstad et al., 1999) and tries to determine the most
important drivers that lead to these changes. These, however, tend to be scale specific
(Veldkamp and Lambin, 2001; Rindfuss et al., 2007; Keys and McConnell, 2005).
For example: at the local scale, social variables and accessibility to resources are the main
drivers of change; at the landscape scale, topography, climate and agricultural potential appear to
matter most; while at the national/regional level, climate, macroeconomic factors (such as
economic development, government policies and indebtedness, population pressures, market
accessibility) and demographic variables, such as household size and labor availability are most
important (Scrieciu, 2007; Veldkamp and Lambin, 2001; Caldas et al., 2007; Rindfuss et al.,
2007). Since the bulk of the LUCC literature addresses deforestation related to agricultural
expansion and logging, I found that studies concerning nonrenewable natural resources are
lagging despite the important and long-term landscape changes that result from their extraction
19
(Almeida-Filho, 2002; Peterson and Heemskerk, 2001). The work related to gold mining,
particularly small-scale mining (done by individuals and small groups) is a growing area of
research, nevertheless. On the other hand, investigations related to oil E&P are less abundant,
especially for Venezuela, a major oil producer and exporter to the US.
Heavy Oil Production in Eastern Venezuela
Venezuela makes an interesting case because it has very large resources of heavy oil; that
is degraded oil that requires more refining to produce a usable and valuable product
(Phizackerley and Scott, 1978; Jiayu and Jianyi, 1999; Meyer and Attanasi, 2003; Galarraga et
al., 2007). The technology is now viable, especially given the high price of oil worldwide, for the
Venezuelan government to expand its operations from the current 4 concessions to 27 in the next
two decades, potentially affecting 18,000 km² of territory across three states. This aggressive
development of heavy oil will have important implications on the regional dry tropical forest
ecosystem, which is already threatened (Fajardo et al., 2005).
Chapter 3 therefore addresses the lacunae in the literature by examining some general
LUCCs resulting from the four current operations and three control sites. I measure four general
landscape changes using GIS and remote sensing (satellite image analysis) techniques. They are:
1) land-cover loss, represented by positive and negative changes in the NDVI vegetation index 2)
petroscape (petroleum-infrastructure) density—the density of petroleum infrastructure features in
km/km²; 3) edge-effect zone, or the area over which significant ecological effects extend
outward from a road or petroleum infrastructure (Forman and Deblinger, 2000; Schneider et al.,
2003; Morton et al., 2004; Thomson et al., 2005); and 4) core areas, or the amount of intact land
mosaics that remain after the edge-effect zone has been taken into account (Morton et al., 2004;
Thomson et al., 2005; Griffith et al., 2002b). In this chapter I present an efficient low-cost way
for land managers and government officials to measure map and monitor LUCCs associated with
20
petroleum exploration and production using GIS methods and (often) readily available satellite
images.
Environmental Performance Variability in the Heavy Oil Belt
Petrozuata’s petroscape values leads to the question: do state-owned oil companies pay
less attention to the environmental effects created by oil operations and therefore create larger
and longer-lasting landscape alterations? Alternately: do multinational oil companies (MNOCs)
place greater importance on their environmental performance, and in particular European
companies? (Marica et al., 2008; Moser, 2001; Gouldson, 2006). If good environmental
performance leads to good economic results (Morse 2008), then why does variability exist
among four oil operations located very near each other, operating in the same landscape and
implementing the same business model? While land-use history is an important component
(Peterson and Heemskerk, 2001; Lambin et al., 2003; Flamenco-Sandoval, 2007; Cumming and
Barnes, 2007; Abizaid and Coomes, 2004) perhaps the mix of private-public partnerships that
exist among each of the four oil concessions also influences the outcome.
Chapter 4 examines the statistical relationship between oil-related LUCC and company
ownership. It divides the concessions into those dominated by 1) local-company (state); 2)
European; and 3) US interests. It expands the number and type of environmental-change
measures among all four HOB concessions (excluding control groups) and hypothesizes that
local-company dominated concessions will exhibit the most environmental alterations, followed
by US-dominated concessions, and that European-strong concessions will have the least changes.
Results show statistically significant associations between local-participation and petroscape
density and well density. No statistically significant correlations exist between US and European
participation and important petroscape measures. Finally, chapter 5 summarizes the results from
the literature review, the analysis of petroscape expansion using satellite images and GIS
21
techniques, and the statistical analysis of petroscape changes and the type of ownership among
the HOB concessions.
22
CHAPTER 2 LAND-USE AND LAND-COVER CHANGE RESULTING FROM RENEWABLE AND
NONRENEWABLE RESOURCE EXTRACTION: A REVIEW
Introduction
There is now essentially universal agreement across the disciplines involved in the emerging land change science that to understand the causes of land cover and use change, it is essential to incorporate human behavior in our theories, models, data, and analyses (Rindfuss et al., 2007, pp. 739).
A central way humans interact with the environment is through the extraction of natural
resources in order to meet the needs for internal consumption and external demand. As humans
compete with each other to access, control and use these resources while excluding or resisting
others, their actions are shaped by political and economic structures as well as by perceived
ecosystem opportunities and constraints operating at various spatial and temporal scales
(Schmink and Wood, 1992; Zimmerer and Bassett, 2003; Robbins, 2004; Bryant, 2001; Mambo
and Archer, 2007; Kennedy, 2003).
Natural resources, divided into renewable and nonrenewable, are a key component of
human-driven land-use land-cove change (LUCC)1. The distinction between the two types is that
renewables (or flow resources) are capable of being replaced by natural ecological processes or
sound management practices, while nonrenewables (or stock resources) are finite, exhaustible
(Harriss, 2006; Neumayer, 2000; Cutter and Renwick, 2004). The extraction of both types of
resources can result in land-cover modification and conversion and involves competition among
various actors wielding different levels of power (Schmink and Wood, 1992). At the same time,
these actors possess “ideas about ecology and political economy [that] actively shape human
perceptions and uses of nature (Bryant, 2001, pp. 162; emphasis in original).
1 Cutter and Renwick (2004, 4, 5) additionally differentiate between two other types of resources, perpetual, like the sun, and potential, which does not have value today but may in the future, such as solid trash and wastewater
23
The study of the interaction between environment, politics and society (Wolford, 2005)
constitutes a concise definition of political ecology. This subfield of geography (and other social
sciences) injects political considerations into traditional human-environment interaction and
global environmental-change science “by demanding that researchers look both at what changes
take place and why they occur” (Keys, 2005, pp. 229).
This paper responds to this demand by reviewing the LUCC and related literature and
synthesizing the general environmental changes resulting from two examples each of renewable
and nonrenewable natural resource use and some of the drivers behind these human-environment
interactions. Agriculture (land and soil) and logging (tree-cutting), as dominant drivers of
deforestation, are important examples of renewable natural resources extraction. Meanwhile
gold-mining and oil exploration and production (E&P) are two activities that can produce lasting
changes on the landscape even after operations have ceased. They represent examples of
nonrenewable natural resources use. The goal of this paper is to better understand the complex
dynamics behind these interactions by focusing on the political, economic and environmental
factors that help produce particular landscapes and the persistent changes that result once these
extractive activities cease. I link political ecology ideas to LUCC research and thus advance the
debate that land-change science, a complex study of the interaction of many variables, can first
be approached by focusing on the relationships between political, economic and environmental
factors.
This review can serve investigators, government officials, conservationists, industry and
others interested in better understanding the relationship between renewable and nonrenewable
resource extraction and the associated environmental changes. I limit the scope of this review by
concentrating on northern South America (Brazil, Venezuela, Guyana, Suriname and French
24
Guiana) when possible, given that a good deal of work exists on Amazonia and I am more
familiar with this region, particularly Venezuela.
Background
A review of the LUCC and similar literature shows a heavy focus on deforestation in
tropical regions, particularly in Amazonia’s humid forests (Soares-Filho et al., 2006; Asner et al.,
2005, 2004; Sebbenn et al., 2007; Flamenco-Sandoval et al., 2007; Vieira et al., 2003; Almeida-
Filho and Shimabukuro, 2004; Caldas et al., 2007; Espírito-Santo et al., 2005; Nepstad et al.,
1999). Meanwhile, dry tropical forests and savannas receive less attention even though they are
the dominant forest type in the tropics, have more productive soils, support much of the world’s
agriculture and human habitation (Southworth, 2004; Jepson, 2005; Fajardo et al., 2005; Daniels
et al., 2008; Trejo and Dirzo, 2000; Romero-Duque et al., 2007) and rank among the most
endangered terrestrial ecosystems (Lawrence et al., 2007; Trejo and Dirzo, 2000; Gordon et al.,
2004).
Nonrenewable natural resource extraction receives even less attention, since many of the
investigators concerned with deforestation study the alterations created by agriculture and
logging, rather than by gold mining and oil extraction despite the long-term changes these latter
activities create on forest cover (Peterson and Heemskerk, 2001; Almeida-Filho and
Shimabukuro, 2002; 2000). When the disturbances created by both types of resource extraction
are combined, they create cumulative effects that fragment the landscape, alter ecosystems,
reduce wildlife populations and create social risks for nearby inhabitants (Schneider et al., 2003)
LUCC and Political Ecology (Theoretical Framework)
Early LUCC research tried to model land-use change by measuring and predicting rates
(quantities) of change among biophysical attributes (Veldkamp and Lambin, 2001) while
ignoring social drivers. This work focused on learning what changes were taking place rather
25
than asking why they were occurring (Keys, 2005). Later efforts integrated socioeconomic
variables and the role of policies (Veldkamp and Lambin, 2001) resulting in case studies that
moved the inquiry beyond description toward an understanding of LUCC (Rindfuss et al., 2007).
But the resulting case studies lacked generalizability and sufficient detail to evaluate data and
methods (Veldkamp and Lambin, 2001; Rindfuss et al., 2007; Keys and McConnell, 2005),
leading Rindfuss et al. (2007, pp. 739) to recently state: “Indeed, the land change field itself does
not have agreement on what should be reported.”
This is reflected in the many processes examined that are believed to lead to LUCC. These
include complex interactions of economic, ecological, political, social, cultural, technological
and infrastructure factors operating at various temporal and spatial scales (Veldkamp and
Lambin, 2001; Hietel et al., 2007; Keys and McConnell, 2005; Turner et al., 2007). Not
surprisingly incorporating all these variables “is complicated by profound cultural differences
between social science and natural science disciplines” (Rindfuss et al., 2007, 739). Furthermore,
“not all causes of land-use change and all levels of organization are equally important. For any
given human-environment system, a limited number of causes are essential to predict the general
trend in land use” (Lambin et al., 2003, 222). The need to provide parsimonious explanations is
one reason that political ecology lends generalizability to studies concerning LUCC; after all
competition, conflict and landscape changes are central outcomes of human-environment
interaction concerning land and resources, particularly when they are scarce (Bryant and Bailey,
1997; Schmink and Wood, 1992; Rubenstein, 2004).
Keys and McConnell (2005) analyzed LUCC trends related to agricultural practices
worldwide by reviewing 91 case studies on this subject and found that the most common drivers
were economic (market), demographic, policy (political) and institutions (property regimes).
26
This does not imply that natural environmental variations, such as climate changes/extremes,
droughts, floods and insect outbreaks do not affect this process (Binford et al., 1997; Hodell et al
1995; Gasse et al., 1990), rather the basic premise is that people’s response to economic
opportunities, as meditated by political and market institutions—and environmental
characteristics (such as climate, topography and soils) drives LUCC at various scales (Lambin et
al., 2001, 2003; Zimmerer and Bassett, 2003; Keys and McConnell, 2005; Rindfuss et al., 2007).
Furthermore, in some cases such as floods, these “natural” disasters have social causes (filling
drains with refuse, inadequate maintenance, using drainage areas for housing and agriculture)
and they affect social groups unequally (Pelling, 2003).
As scale varies so do important drivers. Veldkamp and Lambin (2001) note that: 1) at the
local level social factors and accessibility are important; 2) at the landscape level topography and
agroclimatic variables matter most; while 3) at the national/regional level the important ones are
climatic, macroeconomic and demographic. Thus regarding a large region like Amazonia,
Armenteras et al. (2006, pp. 354) conclude that “deforestation is primarily determined by human
population, accessibility, land use and land tenure issues.”
Meanwhile, Robbins (2004, 2003a) promotes moving away from the binary notion that
environmental change results in destruction of nature, or that it’s socially created2, and instead
accept that nature is produced by human and non-human actors. This aspect of production
demonstrates that landscapes are never static; they are being altered, degraded, converted or
regenerated by human and natural processes continuously. At the same time political power, the
distribution of resources, and perceptions of what types of resources are scarce and valuable, are
not static either (Ellsworth, 2002).
2 The social creation of nature argument “is defined, delimited, and even physically reconstituted by different societies, often in order to serve specific, and usually dominant, social interests” (Castree 2001, 3).
27
Political ecology and land-change science pursue the same goal: understanding human-
environment interaction by studying and modeling how and why people interact with land and
resources and the outcomes of those interactions for both the social and natural world. One
difference, however, is that political ecology tends to focus on (contested versions of) human-
environment interactions related to particular resources such as gold or rubber (Schmink and
Wood, 1992; Salisbury and Schmink, 2007), while land change science treats “the environment
in terms of its array of ecosystem (environmental) goods and services” (Turner et al., 2007, pp.
20666). Thus integrating research that joins social, natural, with GIS and remote sensing science
(Turner et al., 2001) results in a land-system science whereby: 1) the pixel becomes the finest
grain of spatial specificity (Tuner et al., 2007); 2) political and economic forces regulate human
behavior and interactions with the environment; and 3) where the aim is to “better inform
environmental policy” (Redman et al., 2004, 161, 162). This fusion is an outgrowth of works
such as People and Pixels (Liverman et al., 1998), which attempted to span the gap between
remote sensing and social science research. Interestingly, even in traditional land-change science,
the political is ever present. As McCusker and Weiner (2003, pp. 197, 201) point out, “the use of
geospatial technologies to study landscape change has led to differing, and sometimes
contradictory, representations of nature.” This is because “the satellite image is… a social agent
of environmental change, serving as an ally in disputes over the nature of the landscape”
(Robbins, 2003b, 197).
Land-Cover Modification and Conversion
Land use is never static (Lambin et al., 2003); it “is a dynamic canvas on which human and natural systems interact” (Parker et al., 2003, pp. 314).
Land-use decisions comprise the single most important issue in environmental
management since they define the kinds of resources we use and the extent of environmental
28
alterations (Cutter and Renwick, 2004). First, land-cover is defined “by the attributes of the
earth’s land and immediate subsurface, including biota, soil, topography, surface and
groundwater, and human structures” (Lambin et al., 2003, 213). It “refers to the type of material
present on the landscape (e.g., water, sand, crops, forest, wetland, human-made materials such as
asphalt)” (Jensen, 2005, pp. 340). Land use refers to the “purposes for which humans exploit
land cover” (Lambin et al., 2003, pp. 216), or to “what people do on the land surface (e.g.,
agriculture, commerce, settlement)” (Jensen 2005, pp. 340).
Some alterations create expansive changes and endure a longer period than others,
resulting in land-cover conversion rather than modification. Conversion involves the shifting of
one land-cover type to an entirely different one, such as occurs in large-scale agriculture, logging
by clear-cutting of trees, urbanization and deforestation (Southworth, 2004; Lambin et al., 2003;
Daniels et al., 2008). Conversion is also associated with nonrenewable resource use, such as
mining and oil and gas extraction since these actions tend to create degraded areas as well as
large-scale permanent changes on the landscape that are often followed by the establishment of
urban infrastructure (Schneider et al., 2003; Almeida-Filho and Shimabukuro, 2002, 2000).
Roads, bridges, pipelines, tank farms, electric poles and power lines further affect the
surrounding setting transforming the landscape into an urban or periurban form. Another term for
conversion is land transformation, which “refers to radical changes in land use and cover, usually
over the long term” (Turner et al., 2007, pp. 20666).
Modification, on the other hand, refers to changes in the type of cover without changing its
overall classification (Lambin et al., 2003). Here, the land-cover remains in the same category
and is exemplified by selective tree cutting, thinning of forests, degraded grasslands, or changes
in crop patterns and input use (Southworth, 2004; Southworth et al., 2004; Daniels et al., 2008;
29
Veldkamp and Lambin, 2001). However, it can also refer to vegetation regeneration (land-cover
gain) such as successional forest growth and recovery on abandoned lands following natural and
social disturbances such as seismic exploration and certain gold-mining activities (Lee and
Boutin, 2006; Peterson and Heemskerk, 2001; de Figuerôa et al., 2006; Romero-Duque et al.,
2007; Hartter et al., 2008; MacFarlane, 2003), as well as the establishment of tree plantations,
also known as afforestation. Thus, not all land-cover changes involve degradation or conversion.
Yet even afforestation involves logging and converting tropical forest at times in order to
establish tree plantations. In southeastern Venezuela, for example, over 600,000 hectares of dry
tropical forest have been converted to Caribbean pine (Pinus caribaea var. hondurensis) in the
last 40 years (Barnola and Cedeño, 2000). Across the globe, parts of Indonesia’s rainforests are
being converted for oil palm production, a pattern expected to double in the next two-to-three
decades (Sandker et al., 2007). Furthermore, managed forests and plantations “tend to support a
lower biodiversity than unmanaged forests” (Timoney and Lee 2001, pp. 400).
Two other terms used in the LUCC literature related to land-use are intensification and
extensification. Both activities involve shifts in land-use patterns. Whereas intensification refers
to increasing the value of product per land unit, extensification refers to the opposite (McCusker
and Wiener, 2003). Intensification would involve planting more corn plants per hectare and
adding fertilizer in order to produce more, while extensification would involve opening up more
land to agriculture to increase production. Agricultural intensification also involves altering
plants and animals so that humans “supplant natural processes of nutrient and biological
regeneration” (Keys and McConnell, 2005, 321).
Intensification leads to other problems as well: 1) hybridization of local seed varieties, the
spread of monoculture and possible eradication on non-economically valuable plants; 2)
30
hydrological changes resulting from capturing and diverting rivers and streams; 3) introducing
high levels of agrochemicals (including pesticides and herbicides); 4) loss of nutrients such as
carbon; and 5) erosion caused by heightened (intense) soil use and increased mechanization
(Keys and McConnell, 2005; Donald et al., 2006). Land managers commonly use intensification
strategies when agricultural policies provide incentives, note Donald et al. (2006), and these
practices tend to reduce biodiversity. Through this perspective policy instruments are causing
this problem, and removing economic incentives (altering these policies) would help halt this
loss (Donald et al., 2006).
Another important outcome, aside from the alterations caused by both renewable and
nonrenewable resource extraction, is the establishment of settlements, towns and cities spurred
by the expansion of goods and service industries catering to the influx of workers attracted to a
natural resource boom (Schmink and Wood, 1992, Consiglio et al., 2006). These settlements
further urbanize the surrounding landscape and permanently alter the land cover (Lambin et al
2001) even after the boom ends. Lambin (1999, pp. 191-192) concludes that “Land-cover
modifications are generally more prevalent than land-cover conversions,” meaning that land-use
practices tend to keep the land cover in the same classification, rather than completely change it.
Specific Issues: Renewable Natural Resources
Agriculture
Globally the conversion of forests consumes around 13 million hectares per year (FAO,
2006) and the key driver of this change is the expansion of agriculture (FAO, 2006; Lambin et
al., 2003; Daniels et al., 2008). This holds true in Amazonia where growing cash crops (such as
soy, cotton) and raising cattle are the predominant large-scale agricultural activities leading to
deforestation (Schmink and Wood, 1992; Espírito-Santo et al., 2005; Soares-Filho et al., 2004;
Brandão and Souza, 2006; Ferraz et al., 2006; Flamenco-Sandoval, 2007). Deforestation is the
31
removal of vegetative cover beyond a minimal threshold—set at 10% by the United Nations
Food and Agriculture Organization (FAO, 2006), while agriculture is defined as “the cultivation
of domesticated crops and the raising of domesticated animals… to produce food, feed, drink and
fiber” (Jordan-Bychkov and Domosh 2003, pp. 441, 261). This now includes biofuels as
substitutes for fossil fuels (Eickhout et al., 2007; Demirbas and Demirbas, 2007). As Lambin et
al., note (2003, pp. 208), agricultural activity “has expanded into forests, savannas, and steppes
in all parts of the world” resulting in broad areas of land-conversion.
The agricultural pattern in some of Brazil’s forests and savannas is to first convert land to
cattle pastures and then to fields for growing soybeans and other crops as the agricultural frontier
advances (Ferraz et al., 2006; Nepstad et al., 2006). Rindfuss et al. (2007) identify the
commonalities in frontier land-use change as:
• New people with their own social organization and technologies enter an occupied but relatively natural landscape.
• Natural population increase and in-migration create population pressure.
• Settlement patterns at the village and household level alter the landscape.
• As these settlements become linked to national transportation systems they grow.
• Agriculture intensifies and expands in order to sustain a growing population. Cash crops are introduced and these local markets are linked to those at the regional, national and global level. This also brings additional LUCCs.
• Next, spatial patterns change, vegetation decreases, and the habitat is fragmented.
• Finally, environmental constraints and social factors such as state policy and institutional arrangements such as land tenure lead to changes in land use and settlement patterns.
By contrast since shifting agricultural plots tend to serve the household level they: 1) affect
small areas; 2) do not result in complete vegetation removal; and 3) tend to be abandoned after
three to five years unless population pressures are too great (Bassett and Zuéli, 2004; Fölster,
1995; Salazar, 1995; Peterson and Heemskerk, 2001; Lawrence et al., 2007). Several factors can
32
influence cultivation and fallow periods. Environmental variables include soil types and
condition, topography and rainfall patterns; while social factors include government policies and
livelihood strategies (Hartter et al., 2008; Donald et al., 2006; Keys and McConnell, 2005). As
abandoned fields regenerate to secondary forest they replenish nutrients and organic matter, as
well as remove and fix carbon from the atmosphere (Daniels et al., 2008; Kupfer et al., 2004;
Fearnside et al., 2007), highlighting an important point: reforestation is also an outcome of
LUCC (Dalle et al., 2006; Espírito-Santo et al., 2005; Lambin et al., 2003; Vieria et al., 2003;
Keys and McConnell, 2005). However, the “demands of an increasing human population are
causing the fallow period to be shortened” (Vieira et al., 2003, pp. 471), reducing the regrowth of
secondary forests which serve as important ecosystem reservoirs and which are important to
forest-based economies (Hartter et al., 2008).
While extended-use patterns are becoming more prevalent in tropical deforestation
frontiers as a response to population increase (Lawrence et al., 2007) colonizers also contribute
to deforestation by clearing plots for their homes and agricultural fields and as a way to increase
land value. Schmink and Wood (1992) found that some colonization schemes involved migrants
occupying a plot, clearing out the trees and then selling it as one of the few ways to earn money
in this region. At times these plots are sold to farmers with larger holdings that in turn
consolidate land for agroindustrial purposes (Schmink and Wood, 1992). For the most part,
however, shifting cultivation results in land-cover modification, while commercial crop and
cattle raising results in land conversion.
Concerning dry tropical forests, other factors such as overgrazing by cattle and goats can
lead to intense alterations, resulting in shrubland and grassland conversion which limits the
regeneration capacity of these forests (Trejo and Dirzo, 2000). The traditional solution has been
33
to reduce animal stock size, but this simplistic view, notes Turner (2003, pp. 163) overlooks “the
linkages between political economy, management, and ecology…[because] livestock populations
are also economic entities—they are managed by their owners as producers of milk or traction,
as commodities, and as stores of wealth.” This highlights the message by Rindfuss et al. (2007)
that LUCC studies must consider the human perspective.
Logging
Logging is a second important driver of LUCC, but less significant than agricultural
expansion (crops and cattle) in terms of area affected and effects on faunal diversity (Dunn,
2004, pp. 215). This is perhaps due to the regeneration of many abandoned logged sites, such as
in the Chocó rainforests of western Colombia, where only 20 to 30% of logged forests are
converted to agricultural use (Sierra et al., 2003). Sierra et al. (2003) note that this proportion is
lower than what has been reported in other regions such as Amazonia and attribute this lower
conversion rate to land tenure policies that allow local communities to legally exclude migrant
farmers. This highlights the importance of land tenure security in LUCC (Wannasai and
Shrestha, 2008), or as Anderies et al. (2004, pp. 4, 6) put it, “the rules devised to constrain
actions of agents” form the link between resource users and public infrastructure providers.
Thus, failure to “take into account the diversity of land tenure systems within” an area affects the
success of land management projects (Mambo and Archer, 2007, pp. 383).
Type of logging
The type of logging and location also matter. Selective logging (cutting down a few
marketable species) or reduced-impact logging (RIL) practices (extracting valuable trees while
trying to maintain forest integrity and meeting legal restrictions) (Asner et al., 2004; Zarin et al.,
2007) result in the thinning of forests. These actions lead to land-cover modification rather than
conversion since they cause fewer disturbances to the landscape and to faunal diversity (Dunn,
34
2004). Even forests that have been logged by conventional methods and experience severely
altered structure and species composition may still possess a “relatively undisturbed seed bank
and sprouting stumps… [that] contribute to fast regrowth” (Mesquita 2000, pp. 132). Yet this
holds true as long as the forest cover is not removed by bulldozers or heavy machinery. Boletta
et al. (2006, pp. 112, 113) noticed that when logging patterns in the xerophytic forests of the
Argentine Chaco changed from selective deforestation during the 1970s and 1980s to the use of
bulldozers in the mid-1990s, these new actions resulted in “the total loss of forest, including
seed-bearing trees.” In this scenario the pattern became one of: 1) unselective deforestation, 2)
clearing land with heavy machinery, 3) burning the vegetation that remained, 4) plowing the soil,
and 5) planting different crops (Boletta et al., 2006). Romero-Duque et al. (2007) also found that
dry tropical forest land cleared by bulldozers for home building in Mexico and later abandoned
also exhibited greatly diminished regeneration rates.
On the other hand, even when selective (not indiscriminate) logging takes place, the
resulting land-cover changes can be more pronounced than first thought. Asner et al. (2005) and
Nepstad et al. (1999) have observed that when the amount of selective logging occurring in
Amazonia is quantified via remote sensing techniques the true figure of deforestation doubles—
perhaps shifting the LUCC classification from modification to conversion.
Fire
The underestimation of true logging figures (by more than 50%) also applies to fires that
burn large areas of forest every year in Amazonia but that are not included in standard
deforestation measures (Nepstad et al., 1999; Soares-Filho et al., 2006). Nepstad et al. (2001)
believe that fires pose the greatest threat to Amazon forests. However, logging and fires are
related; logged forests are more vulnerable to fires than nonlogged forests (Asner et al., 2004).
Nepstad et al. (1999, pp. 505) explain: “Logging also increases forest flammability by reducing
35
forest leaf canopy coverage by 14-50%, allowing sunlight to penetrate the forest floor, where it
dries out the organic debris created by the logging.” Fires also reduce forest diversity (Gordon et
al., pp. 2004) by compromising living trees intentionally left on logged sites for conservation of
future wood potential, note Gibbons et al. (2008).
By damaging the stems, fires spur invertebrate and fungal decay which can lead to tree
collapse (Gibbons et al., 2008). Additionally, fires “ignited on agricultural lands can penetrate
logged forests, killing 10-80% of the living biomass and greatly increasing the vulnerability of
these forests to future burning” (Nepstad et al., 1999, pp. 505). Forest fires may grow larger than
normal during unusually dry seasons or droughts brought on by El Niño Southern Oscillation
(ENSO) (Nepstad et al., 1999, Whitlock et al., 2006; Siegert and Hoffmann, 2000; Tareq et al.,
2005). Finally, biomass burning during the dry season worldwide contributes to global
atmospheric pollution (Holzinger et al., 2001; Almeida-Filho and Shimabukuro, 2004). A major
obstacle to changing this behavior, note Nepstad et al. (2001, pp. 396) is that anthropogenic fires
are “deeply imbedded in the culture and economic logic of four million rural Amazonians trying
to survive or grow wealthy on the agricultural frontier.” It’s an inexpensive way for farmers to
clear a field, fertilize it and claim ownership (Nepstad et al., 2001).
Fires are also a part of the natural disturbance processes affecting forests, and total fire
suppression can lead to the encroachment of shrubs and rangeland loss (Timoney and Lee, 2001).
When fires do grip these old-growth forests they can cause large-scale damage. Therefore fire
suppression poses a greater threat than logging in some areas such as the dry forests of the
western US (Egan, 2007). In Alberta’s boreal forests, fires play a crucial role in maintaining the
complex mosaic that make up the forest (Dyer et al., 2001). Alternately, agricultural landscapes
affected by or maintained by fires, such as cattle pastures, can shift to secondary succession
36
forests if abandoned (Gordon et al., 2004) and can be covered by regenerated vegetation in 1 to 5
years (Nepstad et al., 1999).
Other factors
Another important limiting factor is the cutting cycle. For logging practices to be
sustainable, reduced-impact approaches must be accompanied by long cutting cycles of 60 years
or more, though 30 years or less is the norm (Kammesheidt et al., 2001; Zarin et al., 2007). The
loss of phosphorus during cycles of shifting cultivation in tropical forests can lead to slower
regeneration rates and therefore introducing phosphorus fertilizer to plots under production could
slow deforestation by reducing the need to cut new plots (Lawrence et al., 2007). Interestingly,
the majority of ground damage related to a wide range of logging practices is the creation of skid
trails, or paths where logs are dragged across the forest floor until they reach loading trucks
(Asner et al., 2004). These create even more lasting alterations than logging roads (Asner et al.,
2004).
Finally, fuelwood spurs a great deal of biomass removal around the world in order to
create energy for cooking, heating and boiling water (Masera et al., 2006). Since biomass
represents 50% to 90% of primary energy consumption in developing countries (Berrueta et al.,
2008), its removal also comprises an important driver of deforestation (Ramos et al., 2008). In
Swaziland and other areas, observes Wheldon (1990, pp. 83), “fuelwood demand exceeds
sustainable supply from accessible land.”
Interestingly, Ramos et al. (2008) found that physical properties of biomass as a fuel
source, such as the fuel value index (FVI) affected local preference for cutting and burning
specific types of fuelwood in northeastern Brazil. The authors’ recommendations pointed to
establishing political and social efforts toward “stimulating the cultivation of these trees in
fuelwood plantations and in the home-gardens within the community” (Ramos et al., 2008, 7).
37
Another important recommendation is adopting more efficient wood-burning stoves since they
not only reduce fuelwood consumption but also decrease negative health effects resulting from
prolonged inhalation of smoke from the cooking fires—which tends to affect women and
children the most (Wheldon, 1990; Berrueta et al., 2008; Ekici et al., 2005).
In summary, the results of large-scale deforestation include habitat fragmentation, loss of
biodiversity, shrinking of forest remnants, soil erosion, increased sedimentation, liquidation of
old growth forests and the species that depend on them; introduction of exotic biota, wetland
drainage, pollution, loss of carbon sinks and subsequent increased CO2 emissions (Ferraz et al.,
2006; Leslie et al., 2001; Miranda et al., 1998; Southworth, 2004; Timoney and Lee 2001;
Armenteras, et al., 2006). As Zarin et al. (2007, 922) put it: “Standing forests, regardless of their
long-term management potential, benefit society through ecosystem services not furnished by
agriculture or cattle ranching.”
Specific Issues: Nonrenewable Natural Resources
Gold Mining
Compared to renewable natural resources, mineral mining receives less attention in the
LUCC and related literature despite the serious environmental and social alterations associated
with theses activities (Parrotta et al., 2001; Peterson and Heemskerk, 2001, Almeida-Filho and
Shimabukuro, 2002; Leger 1991). Gold mining, as opposed to agriculture and logging, creates
two effects that produce long-lasting environmental and social problems: mercury pollution and
land degradation (Hilson, 2002; Almeida-Filho and Shimabukuro, 2002).
Land degradation
Gold is extracted two main ways. The first one involves large-scale commercial mining of
land-based hard-rock deposits that require advanced technology and resources (often from
foreign companies) and result in open-pit mines that exemplify large-scale land-use changes in
38
terms of cleared land, roads for transporting the material, tailing ponds (which hold waste
material and water containing high levels of mercury or cyanide), and growth of settlements that
house the miners and the accompanying service industry (Veiga and Hinton, 2002; Schmink and
Wood, 1992; Hilson and Monhemius, 2006; Lacerda and Marins, 1997; Castro and Sánchez,
2003).
Despite the visual impression caused by large open pit mines, the second and more
common method is performed by small-scale artisanal miners (10-15 million worldwide) who
both excavate surface soil deposits and perform fluvial mining by searching for flakes, pebbles
and nuggets along riparian streams in tropical Asia, Africa and South America (Veiga et al.,
2006; Lacerda, 1997; Hilson, 2002; Heemskerk, 2001; Malm, 1998; Peterson and Heemskerk,
2001; Schmink and Wood, 1992; Vieira, 2006). Small-scale miners often employ extraction
methods that entirely remove vegetation and degrade land, and incorporate mercury (and
cyanide) to amalgamate gold. Combined these two actions create thousands of polluted sites
across 55 countries (Veiga and Hinton, 2002; Veiga et al., 2006). In Ghana, “small-scale gold
mining activities cause a disproportionately large percentage of environmental degradation”
(Hilson, 2002, pp. 70). The driving force: “gold is most commonly mined on a small scale
because of its propensity to generate wealth quickly” (Hilson, 2002, pp. 58).
Peterson and Heemskerk (2001) describe the process: 1) about five miners clear an 80 x 80
m forest plot. They cut, remove and burn most of the trees while those that are difficult to
remove are left standing. 2) They then use a high-pressure diesel-powered water hose to remove
the topsoil layer and then the gold bearing sand and clay layer. 3) They pump the slurry
containing gold into a sluice box where they introduce mercury or cyanide to adhere to the gold.
39
4) They keep the gold clumps and release the tailings (slurry waste) into the adjacent forest or
into an abandoned mining pit.
Alternately, miners operating on watercourses 1) dredge river bottoms for 20 to 30 hour
phases; 2) sieve the sediments; 3) place the selected material in barrels; 4) amalgamate the
material with mercury or cyanide by hand or with mechanical stirrers until all gold is found; 5)
release the mercury-rich residues back into the river (Lacerda, 1998; Hilson, 2002; Kligerman et
al., 2001). During this process a lot of mercury also vaporizes into the atmosphere and oil from
the dredge equipment and motors enters the water, negatively affecting plankton and aquatic
fauna (Lacerda, 1998; Kligerman et al., 2001; Lacerda and Marins, 1997).
Regeneration
Not surprisingly, pressure washing river banks with hydraulic pumps, digging large open
pits, and dredging riverbeds stimulates soil erosion and stream sedimentation (affecting
hydrological patterns, reducing fish diversity and their community structure, and threatening
hydroelectric production); leads to land degradation, habitat destruction and the loss of wildlife;
causes microclimatic changes from the removal of vegetation; and creates potential new
problems from the remaining contaminated tailing ponds (Mol and Ouboter, 2004; Heemskerk,
2001; Almeida-Filho and Shimabukuro, 2002, 2000; Peterson and Heemskerk, 2001; Lacerda,
1998; Hilson, 2002; Kumah, 2006; Kligerman et al., 2001). In places like Ghana, miners remove
the vegetation, dig for gold and move on, leaving a scarred landscape composed of pits and
trenches that is not suitable for other uses (Hilson, 2002). In many cases, notes Kitula (2006, pp.
409) “both agricultural and grazing lands have been destroyed,” which affects local livelihoods.
The complete removal of most vegetation and roots also removes the seed bank, thus, like
areas that are converted to agriculture via the use of bulldozers (Boletta et al., 2006; Romero-
Duque et al., 2007), mining practices impede regeneration on abandoned plots (Peterson and
40
Heemskerk, 2001). When regeneration does occur, the rates vary over space and time (Almeida-
Filho and Shimabukuro, 2002). The edges of abandoned degraded areas regenerate faster than
the central zone where damage is more intense; indicating that recovery of the entire degraded
area takes much longer than anticipated, up to 20 years or more (Almeida-Filho and
Shimabukuro, 2002).
Rodrigues et al. (2004) observe that conserving forest fragments along these degraded
areas is crucial for allowing natural regeneration of shrub and tree species which act as a seed
source and animal refuge. Yet two other factors can also speed the regeneration process: gold
exhaustion and law enforcement since they encourage small-scale miners to abandon the area
and companies to close the large mines (Almeida-Filho and Shimabukuro, 2002; Castro and
Sánchez, 2003). The first is an outcome of geology or of limited tools and technology. The
second one points to the importance that resource tenure or good environmental governance has.
Concerning LUCC, heavily utilized areas result in land-cover conversion, while the lighter
impacted areas reflect modification. Finally, restoration projects by mining companies involve
afforestation “to rapidly establish tree cover on reclaimed mine sites and therefore facilitate
natural forest succession,” but often fast-growing exotic species are planted (Parrotta and
Knowles, 2001, 220), rather than native trees, altering ecosystem composition.
Mercury pollution
Though its use is illegal in many countries mercury (and cyanide—mainly used in
industrial applications) is widely implemented to help recover gold from the rocks, soils or river
sediments because it is effective, cheap and easy to use (Hilson, 2002; Mol and Ouboter, 2004;
Lacerda, 1997; Hilson et al., 2007; Donato et al., 2007; Sandoval et al., 2006, 416; Hilson and
Monhemius, 2006; Veiga et al., 2006). However, after it is utilized the mercury is often
discharged into local streams where it enters the food chain. There it is metabolized from an
41
inorganic form to methyl-mercury and biomagnified. By the time humans eat large (carnivorous)
fish it has become highly toxic (Hilson, 2002; Sandoval et al., 2006; Mela et al., 2007; Pinheiro
et al., 2008; Bose-O’Reilly et al., 2008; da Costa et al., 2008). Mercury also pollutes drinking
water and poisons workers and residents who handle the metal or inhale the fumes during the
burning process that separates the mercury from the gold (Lima et al., 2008; Sandoval et al.,
2006; Almeida –Filho and Shimabukuro, 2000, 2002; Reed, 2002; Castro and Sánchez, 2003;
Drake et al., 2001). These vapors are re-released when the gold is burned again in gold shops or
smelters, affecting areas within 1 km of the emission source (Veiga and Hinton, 2002; Bastos et
al., 2004)—not just the mining areas. Also noteworthy: if terrestrial animal populations are
completely driven away in mining areas (Heemskerk, 2001), then miners and local populations
are more prone to increase their fish intake, thus exacerbating their levels of toxic methyl-
mercury.
Solutions
Some ways to address the environmental problems associated with mining include: 1)
using biological treatments to take up metal from wastewaters (Moore et al., 2008; 2), including
almond shells (Estevinho et al., 2008); training small-scale miners on how to efficiently use
mercury (Castro and Sánchez, 2003; Hilson, 2002); 3) formalize (legalize) small-scale miners in
countries such as Ghana in order to regulate production so that greater transparency and control
will limit environmental degradation and improve local livelihoods (Hilson et al., 2007); 4)
introduce alternatives to mercury and cyanide as leach reagents for gold mining (Hilson and
Monhemius, 2006; Vieira, 2006); 5) use retorts to help capture a lot of the lost mercury (up to
95%) (Drake et al., 2001; Veiga and Hinton, 2002; Hilson, 2002); 6) establish decontamination
(recovery) plans between the government and mining companies within a specific time period
(Castro and Sánchez, 2003); 7) reduce gold consumption (demand) with awareness campaigns;
42
and finally 8) spur European Union countries—as the largest producers of mercury—to regulate,
tax and control its export (Veiga et al., 2006).
Oil Exploration and Production.
Though less expansive than the LUCC literature related to gold mining, research on
landscape changes and oil exploration and production appears to be the purview of petroleum
geologists, biologists and others interested in environmental management. The first group has
been using geographic information systems (GIS) and remote sensing techniques in their work
for over 30 years, primarily for geologic exploration (Jackson et al., 2002). Today geospatial
applications in the petroleum industry continue to focus on exploration but also on detecting and
modeling spills; implementing least-cost analysis for pipeline constructions; reducing operational
costs by managing spatial information; environmental baselining3; impact assessment and
analyzing settlement patterns (Gaddy, 2003; Jackson et al 2002; Janks and Prelat, 1994, 1995;
Osejo et al., 2004; Musinsky et al., 1998; Pochettino, 2001; Satapathy et al., 2008; Almeida-
Filho et al., 2002; Almeida-Filho, 2002; Zhang et al., 2007; Jaggernauth, 2003).
Wildlife: caribou
The second group of researchers studying oil (E&P) includes biologists concerned with the
landscape and habitat changes resulting from these activities. Many of these scientists are
primarily interested in avoidance and stress responses by wildlife to oil roads, wellpads and
activities—particularly caribou in Canada and Alaska (Nelleman and Cameron, 1996; Noel et al.,
2006; Schneider et al., 2007; Wolfe et al., 2000; Bradshaw et al., 1998; Harron, 2007; Dyer et al.,
2001). They have found, for example, that petroleum development in oilfields near Prudhoe Bay,
3 Baselining is the process of documenting landscape conditions in a given area to create a profile prior to exploration and production phases (Janks and Prelat 1994; Jackson et al., 2002). Over time, as activities increase in concession areas, the environmental and social alterations can be monitored through the analysis of air photos, satellite images and field site visits.
43
Alaska, displaced maternal female caribou by 4 km (Nelleman and Cameron, 1996). Schneider et
al. (2003) found that caribou in Alberta, Canada avoided seismic lines by 100 m, while for oil
roads the distance was 500 m. Harron (2003) disputes findings for seismic lines, pointing out that
caribou avoidance distances are only 40 m on each side.
Bradshaw et al. (1998) point out that disturbances to caribou caused by petroleum
exploration in Alberta can cause maternal females to expend more energy during winter months,
leading to undernutrition which delays parturition. Therefore a newborn calf has less time to
grow before the onset of the next winter, making it more prone to die from predation or
undernutrition (Bradshaw et al., 1998; Wolfe et al., 2000). Dyer et al. (2001) also studied caribou
avoidance behaviors to oil and gas activities by fitting 36 caribou with GPS (global positioning
system) collars. They recorded avoidance distances of 1,000 m for wells and 250 m for seismic
lines and roads, noticing that if caribou are displaced to less suitable habitats, their numbers may
go down, affecting demographic patterns (Dyer et al., 2001). Wolfe et al. (2000) found that
calving caribou avoided oil field infrastructures by 5-6 km, indicating that the effects of the
infrastructure footprint were much larger than thought. Furthermore, activities that displaced
reindeer and caribou affected subsistence and cultural resources for Arctic dwelling indigenous
groups (Wolfe et al., 2000). Since drilling permits allow for petroscape expansion, energy policy
affects local livelihood options.
Interestingly, as compared to agriculture, logging and gold mining, oil (and gas)
“developments are typically configured as point and linear disturbances scattered throughout
broader areas” that often encompasses 5-10% of the concession site (Wyoming GFD, 2004, pp.
5). Therefore, as opposed to clear-cutting forestry practices, the alterations created by oil
exploration and production may be harder to detect. Seismic lines are one exception.
44
Seismic lines
An important outcome of oil exploration is the creation of seismic lines (roads cut during
seismic phases of oil exploration). In addition to disturbing wildlife, seismic lines can create
permanent changes on the landscape that remain long after exploration activity has ceased. Lee
and Boutin (2006) examined the persistence of seismic lines on the Alberta boreal landscape
over 35 years and made some interesting findings:
• Wide seismic lines (5 to 8 m) cleared with bulldozers persist on the landscape for more than 35 years. Recovery to woody vegetation occurred in 8% of the cases and based on median recovery rates it would take 112 years for a near complete recovery. Lands cleared with bulldozers flatten and compact the terrain, altering hydrological patterns, reducing sites for seedling establishment, and converting the landscape to a wetter sedge-dominated community.
• The presence of seismic lines provided transportation corridors that in turn facilitate other industrial development, such as logging. These routes were subsequently used by the forestry industry to monitor regeneration rates and silvicultural practices. They also created routes ripe for 4-wheel-drive vehicles that hampered regeneration via tire track damage, water channelization, soil compaction and soil erosion on and next to the seismic lines.
• Changes to oil and gas exploration licenses (resource tenure/energy policy) combined with increased market prices spurred the creation of seismic lines in the 1970s. Today’s demand and tightening supplies is also prompting exploration and drilling. In fact a “crash programme production” is underway in Canada’s oil sands, partly to offset a peaking world oil production (Söderbergh et al., 2007, pp. 1931).
• The transfer of exploratory drilling licenses among oil companies is common; however seismic data remains proprietary information. This requires the new license holder to reopen the existing seismic lines and/or create new ones.
Roads
One commonality among all types of resources examined is the creation of roads that
result when land managers want to access trees, new plots of land, gold-rich areas or oil
reservoirs often in previously inaccessible areas. Roads, after all, are the largest human artifact
on earth (Forman et al., 2003) and a central catalyst for LUCC. In Amazonia 85% of
45
deforestation “is concentrated at up to 50 km from the main roads” (Brandão and Souza, 2006,
pp. 177).
Besides deforestation, road openings lead to the establishment of new agricultural
frontiers, particularly for soybeans and cattle ranching; increase fire advancement on forests;
prompt legal and illegal timber harvesting; and result in land speculation and colonization
(Brandão and Souza, 2006; Ferraz et al., 2006; Lele et al., 2000; Peres and Terborgh, 1995;
Soares-Filho et al., 2004; Schmink and Wood, 1992). In some cases “as soon as there was even
the prospect of a new road, migrant farmers would move into the area, taking de facto possession
of small plot ” (Schmink and Wood, 1992, pp. 79).
Most of these roads can be defined as local since their primary function is land access
(Forman et al., 2003). But roads also serve economic, political and social purposes at broader
scales. They stimulate the growth of commerce and increase market accessibility; they
consolidate territorial control; and tie land together for society (Brandão and Souza, 2006;
Ciccantel, 1999; Forman et al., 2003). The study of landscape changes resulting from the
creation, use and interaction of roads (and vehicles) and the surrounding environment is known
as road ecology (Forman et al., 2003), a growing subfield of landscape ecology, which
“emphasises the effect of spatial patterns of ecosystems on ecological processes” (Griffith et al.,
2002b, 1).
Concerning petroleum and roads, Musinsky et al. (1998) analyzed satellite images and
found that oil roads in national park in Guatemala were directly related to the establishment of
new settlements in the surrounding tropical rainforest. These actions were tied to political and
economic forces such as 1) inadequate park management; 2) poorly guarded roads and other
infrastructure that allowed easy access; 3) contradictory government policies and jurisdictions;
46
and 4) inefficient agricultural practices and rapid population growth (Musinsky et al., 1998).
While government policy solutions and enforcement should address some of these problems,
Musinsky et al. (1998) spurs energy companies to attend to oil roads by: 1) planning routes to
avoid habitat fragmentation and disruptions to hydrological cycles; 2) use existing roads and
waterways fully; 3) destroy non-necessary roads, bridges and entry points; 4) bury pipelines and
restore vegetation; and 5) monitor remaining roads and access points through field monitoring
and remote sensing techniques.
The same holds true for seismic roads; Lee and Boutin (2006) propose: 1) establishing
narrow seismic lines, less than 5 m; 2) using helicopters and hand cut trails when possible—as
opposed to using bulldozers; 3) reclaiming seismic lines so that they are not converted to
permanent paths and roads; and 4) implement best practices by sharing roads with other
extractive industries, such as forestry and logging, rather than creating new ones.
Conclusion
A survey of the LUCC and related literature shows a strong preference for work related to
natural resource use, particularly agriculture and logging as they relate to deforestation in
tropical regions. Nonrenewables receive much less attention, though studies related to gold-
mining are become more prevalent, particularly given the problems associated with mercury and
cyanide use. On the other hand, land-cover change research related to petroleum E&P is even
less abundant, even though these activities create large-scale and permanent changes on the
landscape and affect they way people respond in larger areas around the zones of direct
extraction.
Two zones containing some of the largest petroleum reserves in the world and undergoing
new exploration and production include the heavy oil sands in Canada and Venezuela’s heavy oil
belt. Alberta’s oil sand development has attracted a growing number of researchers, many of
47
whom examine how petroleum infrastructure affects wildlife, particularly caribou (Bradshaw et
al., 1998; Wolfe et al., 2000; Dyer et al., 2001). Others such as Schneider et al. (2003) address
forest disturbances created by oil E&P and provide a model of how to address this issue.
By contrast almost no work that I could find addresses this topic in Venezuela’s heavy oil
belt, despite the estimated 270 billion barrels of heavy crude located in central and eastern
Venezuela. The research concerning the heavy oil belt instead addresses drilling strategies
(Gipson et al., 2002), reservoir modeling (Tankersley and Waite, 2002), and heavy crude
chemical properties (Galarraga et al., 2007), but not land-cover changes despite the claim by
energy companies that they are pursuing sustainable business practices that minimize social and
environmental alterations (Kumah, 2006; Garvin et al., 2008; Mudd, 2007a, 2007b; Sandoval et
al., 2006).
Right now the Venezuelan government plans to expand the heavy oil belt operations from
four concessions to 31 in the next two decades potentially affecting 18,000 km² of dry tropical
forest and savanna. Quantification and certification efforts of oil in place have started in some of
these new blocks. Therefore urgent research is needed to better understand what types of
petroscapes (petroleum-related infrastructure landscapes) the four functioning concessions have
produced and which ones exhibit the least amounts of petroscape disturbances. This information
should be used during baseline studies (Kalacksa et al., 2008) for the upcoming concessions, if
possible before seismic activity begins, and then implemented during the design and production
phases. Introducing best business practices and lessons learned will reduce environmental and
therefore social changes (Gipson et al., 2002; Moser 2001). The application of geographic
information systems (GIS) and remote sensing techniques are very useful and efficient ways to
monitor LUCCs at the landscape level (Janks and Prelat, 1995, 1994; Musinsky et al., 1998;
48
Almeida-Filho, 2002; Almeida-Filho et al., 2002; Almeida-Filho and Shimabukuro, 2004, 2002,
2000; Gaddy, 2003). This is what now required in Venezuela’s heavy oil belt.
49
CHAPTER 3 DO NATIONAL OIL COMPANIES PRODUCE GREATER ENVIRONMENTAL ALTERATIONS THAN MULTINATIONAL OIL COMPANIES? THE CASE OF
VENEZUELA’S HEAVY OIL BELT
While energy-producing hydrocarbon resources dominate the global economy, their
mining and extraction create important land-use and land-cover changes (LUCC) that affect
climate, ecosystem goods and services and people’s vulnerability at different spatial and
temporal scales (Prakash and Gupta, 1999; Almeida-Filho and Shimabukuro, 2002, 2000;
Diamond, 2005; Schmidt and Glaesser, 1998; Lambin et al., 2003; Fujita and Fox, 2005;
Southworth, 2004). Concerning petroleum exploration and production (E&P), a global question
exists as to whether state-owed firms, also known as national oil companies (NOCs), can or do
perform as well as multinational oil corporations (MNOCs) in terms of balancing economic
efficiency, environmental and social outcomes.
This paper addresses this topic by examining the environmental outcomes in terms of
important LUCC measures related to petroleum E&P in eastern Venezuela’s heavy oil belt
(figure 3-1). This 54,000 km² zone contains over 1 trillion barrels of tar-like heavy crude oil, of
which 270 billion are considered recoverable using current technology (Gipson et al., 2002;
Talwani, 2002; Trebolle et al., 1993; Tankersley and Waite, 2002; Balke and Rosauer, 2002). In
this paper the local question is: does the local NOC perform as well as its MNOC partners in
terms of balancing environmental alterations associated with E&P at the landscape level? I
answer the question by examining the expansion of petroleum-related infrastructure features, or
the petroscape, in three of Venezuela’s four heavy oil belt operations between 1990 and 2005
using Landsat TM and ETM+ and CBERS satellite images and GIS techniques.
The concessions (see figure 3-2) and the amounts of company ownership involved are
shown in table 3-1. In terms of company participation, the breakdown is:
50
• Sincor: Local (NOC) 38%; European 62% • Petrozuata: Local 49.9%; US 50.1% • Ameriven: Local 30%; US 70% • Cerro Negro: Local 41.67%; US 41.67%; European 16.66%
Therefore, the local NOC, from here on referred to as Local, dominates Petrozuata;
European companies dominate Sincor; US companies dominate Ameriven and both Local and
US companies are tied in Cerro Negro (see table 3-1). Nevertheless, Cerro Negro is not included
in this analysis due to excessive cloud cover in the imagery data set corresponding to this
concession. Drawing from political ecology and organizational theory, as well as research from
oil and gas E&P alterations to landscape and wildlife, shows that energy policy and corporate
attitudes toward environmental planning and environmental regulations can affect LUCC in
terms of land degradation, landscape fragmentation, loss of biodiversity, river siltation, increased
pollution and greenhouse gas emissions (Bryant and Bailey, 1997; Nelleman and Cameron,
1996; Schneider et al., 2003; Morton et al., 2002; Morton et al., 2004; Lee and Boutin, 2006;
Lawrence, 2007; Schneider et al., 2003).
Through lessons learned, multinational oil companies (MNOCs) have increased their
efforts to minimize and monitor environmental and social impacts tied to E&P (Moser, 2001).
Several reasons exist for improving performance in health, safety, and environment (HSE) as
well as corporate social responsibility (CSR), or sustainable development initiatives. They
include preventing: costly environmental disasters, political protests, negative press coverage,
work stoppages, sabotage, project terminations, forced closings, and human rights and worker
abuses (Diamond, 2005; Lawrence, 2007; Anderson, 1994; Gladwin, 1977; Orlitzky et al., 2003).
Furthermore, implementing best business practices can increase employee and customer loyalty,
attract the attention and money of financial analysts, investors and lenders (Morhardt, 2002), can
offset actual or impending government regulation (Dashwood, 2007), improve a company’s
51
reputation, better address human rights (Kakabadse, 2007), improve community relations, and
help stabilize long-term costs (Reinhardt, 2000). As Shapiro (2000, pp. 63) notes: “Far from
being a soft issue grounded in emotion or ethics, sustainable development involves cold, rational
business logic.” Part of this logic involves identifying opportunities through sustainability
integration, rather than traditionally responding to regulations, risk reduction and cost controls
(Wilcoxon and Cramer, 2002). It also promotes voluntary adoption and reporting of sustainable
business practices defined by such groups as the International Petroleum Industry Environmental
Conservation Association (IPIECA). Furthermore, European companies tend to be ahead of US
companies in terms of implementing sustainability practices into their business plan (Marica et
al., 2008; Moser, 2001).
On the other hand, NOCs often tend to lack the experience and culture of minimizing
environmental changes particularly since (until recently) they do not compete internationally for
contracts and pursue their business as usual approach domestically (Diamond, 2005). When they
do participate in the global economy they must implement the most advanced technology and
methods in terms of their inputs to successfully compete (Porter and van der Linde, 2000), which
involves costly and sustained investments. A second challenge is that most developing countries
face major challenges and lack resources for planning and for managing natural systems and
environmental quality (Gladwin, 1977). Third, many leaders in the developing world, whose
countries depend on the production and export of primary products such as beef, bananas,
timber, cocoa, copper or petroleum, and who don’t have alternative sources of national income
“perpetuate practices that contribute to environmental degradation” (Bryant and Bailey, 1997,
pp.59). Finally, when the production of a natural resource comprises a country’s primary
industry and staple export product and is strongly controlled by the federal executive office, as is
52
petroleum in Venezuela, then other government bodies charged with the oversight of these
activities, such as the ministry of the environment, may not have the capacity or will to enforce
compliance. Based on the findings from this literature, I propose three hypotheses.
• Hypothesis 1: The Local-company dominated concession, Petrozuata, will exhibit the greatest amount of LUCC.
• Hypothesis 2: The US-company dominated concession, Ameriven, will display a smaller amount of LUCC.
• Hypothesis 3: The European-dominated concession, Sincor, will show the least amount of LUCC.
I test these hypotheses by measuring and ranking four important landscape ecology metrics
related to oil E&P: vegetation-cover change, petroscape density, edge-effect zones, and core-
areas using remote sensing and GIS methods (Griffith et al., 2002b; Morton et al., 2002;
Thomson et al., 2004; Morton et al., 2004) to determine how each performed.
Remote Sensing, Geographic Information Systems (GIS) and Petroleum
Remote sensing (satellite image analysis) and geographic information systems techniques
have proven useful and cost-effective prospecting tools for oil and gas deposits for over 30 years
(Almeida-Filho, 2002; Jackson et al., 2002). These methods have been applied to detecting
hydrocarbon seepages in northeastern Brazil (Almeida-Filho et al., 2002; Almeida-Filho et al.,
1999) and the Netherlands (Noomen et al., 2006); to identify oil-bearing sands in inner
Mongolia, China (Zhang et al., 2007); and to optimize costs of field infrastructure facilities in
western Argentina (Pochettino and Kovacs, 2001). A few studies address the role of oil E&P on
landscape-change. These include Janks and Prelat (1994; 1995) who use remote sensing and
promote these techniques to: 1) assess the health of vegetation in and around oil fields; 2)
identify local environmental effects of oil drilling and waste management; 3) optimize the
placement of infrastructure features such as exploratory wells, roadways, camps and pipelines;
53
and 4) monitor the progress of remediation efforts on abandoned well sites. Meanwhile
Musinksy et al. (1998) applied these technologies to determine the relationship between the
construction of oil roads and subsequent deforestation in Guatemala’s Laguna del Tigre National
Park. They found “a direct relationship between the construction of access roads for oil
exploration and production, and the settlement into undisturbed tropical rainforest ecosystems”
(Musinsky et al., 1998, pp. 1). Jackson et al. (2002) and Gaddy (2003) also note how remote
sensing technology has important business applications for the petroleum industry. These include
monitoring environmental change, managing existing field operations; evaluating land property
before leasing or abandonment; and least-cost analysis for pipeline construction (Jackson et al.,
2002).
Concerning spills and pollution, Kwarteng (1998, 1999) and Ud din et al. (2008) used
remote sensing to monitor contaminated oil lakes and the recovery of surrounding vegetation in
Kuwait; while Chust and Sagarminaga (2007) used remote sensing to discriminate oil slicks in
Venezuela’s Lake Maracaibo. Morton et al. (2002) used GIS analysis to estimate the economic
viability of undertaking oil and gas E&P activities in protected roadless areas of 6 western US
states. Thomson et al. (2005) conducted a spatial analysis of the gasscape (gas related
infrastructure features) in western Wyoming to determine the potential effects on 4 wildlife
species. Morton et al. (2004) implemented remote sensing and GIS analysis to estimate E&P-
driven changes to land-use and land-cover such as habitat fragmentation and loss of terrestrial
wildlife and aquatic species in the US Rocky Mountains. The Wilderness Society (2006) also
used remote sensing and GIS methods to measure habitat fragmentation and the effects on
wildlife in Wyoming. They, along with Morton et al. (2002) and Morton et al. (2004) conclude
that the expansion of E&P activities will cause permanent landscape changes that will reduce
54
wildlife diversity and numbers; will negatively affect local economies that depend on tourism
and hunting; and will lead to further habitat fragmentation through the expansion of new roads
and more people to wilderness areas. More importantly, they estimate that the recovery amounts
of oil and gas will be quite small, in some cases supplying US consumption needs 24 days (for
oil) and 9 to 11 weeks for gas (Morton et al., 2002).
This paper builds on the LUCC literature related to energy development by focusing on
Venezuela’s heavy oil belt, an underrepresented topic about an area that is about to expand its
concessions by more than 600% in the next 22 years in order to help double production. New
operations will mostly involve other NOCs from countries such as Brazil, Iran, Cuba, Chile,
China, India, Uruguay, Russia, Argentina and Ecuador, rather than MNOCs (PDVSA, 2008c,
2007a, 2007c, 2007e, 2007g, 2006a, 2006b; Petroguía, 2006-2007). These new partnerships will
create their own E&P landscapes with greater or lesser alterations than currently exist. Better
understanding how the four original operations managed their petroscapes can prove useful for
planning subsequent new ones.
Materials and Methods
Site Description
The heavy oil belt (HOB) is a 54,000km² stratigraphic trap that contains over one trillion
barrels of heavy and extra heavy crude oil of which 270 billion are considered recoverable using
current technologies (Gipson et al., 2002; Talwani, 2002; Trebolle et al., 1993; Tankersley and
Waite, 2002; Balke and Rosauer, 2002; Boza and Romero, 2001; Briceno et al., 2002). Located
north of the Orinoco River in central and eastern Venezuela, the HOB is subdivided into four
separate E&P zones spanning the states of Guárico, Anzoátegui, Monagas and a small portion of
Delta Amacuro. From west to east the E&P zones are Machete, Zuata, Hamaca and Cerro Negro
(Balke and Rosauer, 2002; Kopper et al., 2001). Recently, they have been renamed to Boyacá,
55
Junín, Ayacucho and Carabobo, respectively (PDVSA, 2005a, 2005b), though much of the
literature uses the old names when referring to these zones (see figure 3-2).
The four current operations are located in the southern llanos, or plains, of Anzoátegui
state located above the Orinoco River. This area is marked by sandy soils with low natural
fertility, grasslands, low and dispersed drought tolerant vegetation, gallery forests along the
Orinoco River tributaries, and a sparse population comprised of small scattered settlements
(Hernández et al., 1999; Dumith, 2004). Average annual temperature is 26ºC with mean diurnal
and seasonal fluctuations of 9.5º C and 3º C respectively (Chacón-Moreno 2004). A four-to-six
month dry season affects the phenological variation of this seasonal savanna ecosystem where
the environmental conditions are related to the amount of water availability (deficit or excess) in
the soil during the year (Chacón-Moreno 2004).
The important economic activities in these llanos include small-scale fishing along the
many Orinoco-tributaries, small and large-scale farming (soy, cotton, sorghum, corn, rice), cattle
ranching (since the 1548—White, 1956), petroleum production (since the 1930s), Caribbean pine
plantations (since the 1960s), and employment in the oil fields and pine plantations (Dumith,
2004; IADB, 1997; Parra, 2007; Mauricio et al., 2005; Hernández et al., 1999; Talwani, 2001).
All these activities help produce a heterogeneous landscape. The terrain is mainly flat, with a
slope of less than 2% (Castel et al., 2002). Elevations across the concessions range from near
zero to 120 meters above sea level (m.a.s.l) along the banks of Orinoco River tributaries, to 121
to 250 m.a.s.l. in the northern part of the HOB.
Data Description (Concessions and Control Groups)
The data used in this chapter include LUCC measures for 3 concessions and for 3 control group
areas quantified from satellite images of the study area and through GIS techniques. Researchers
such as (Morton et al., 2002; Morton et al., 2004; Schneider et al., 2003; Thomson et al., 2005;
56
and Griffith et al., 2002b) have found that the following landscape-measures provide important
indicators of ecological disturbance related to road and energy development:
• Vegetation-cover change—refers to how vegetation land-cover has changed between time periods. Here I examine the conversion of vegetation using the Normalized Difference Vegetation Index (NDVI), an algorithm that measures the amount and condition of vegetation in a satellite image (Jensen, 1996), has been widely used in vegetation-change studies (Mambo and Archer, 2006; Tarnavsky et al., 2008) and is particularly related to the expansion of the petroscape (Musinsky et al., 1998). This index also detects gains in vegetation cover. In order to detect changes between 1990 and 2000 I performed image differencing of NDVI images of the study area, a common technique in change-detection studies (Mambo and Archer, 2006) that involved subtracting the 1990 from the 2000 images.
• Road density—is a rough but useful measure that takes the total number of petroscape linear features measured in kilometers (km) and divided by the area in km²—in this case the size of the concession. It is used to assess the potential impact of roads and infrastructures on local environments (Forman et al., 2003). Because it’s a density measure, the length of linear features is somewhat normalized by dividing this figure by the concession size.
• Edge-effect zones—related to petroscape density, these zones include the area of significant ecological effects extending out from petroscape features (Forman and Deblinger, 2000). They are created by using a 600 meter (m) buffer around the linear petroscape features. I selected this distance based on work by Forman and Deblinger (2003) regarding roads, Schneider et al. (2003) and Dyer et al. (2001) in reference to energy roads in Alberta, Canada, and Morton et al. (2004), Thomson et al. (2005) and the Wilderness Society (2006) regarding LUCC and energy development in the western US.
• Core areas—which are the intact habitat areas that remain after the concession has been fragmented by petroleum production activities.
Time Frame
The objective was to measure the petroscape-related changes in land-cover between 1990,
2000 and 2005 and determine if the Local-dominated concession exhibited greater amounts of
change than American-company and European-company dominated sites. The year 1990
corresponds to the period before E&P began in the HOB concessions (Talwani, 2002), officially
known as strategic associations, and following a business model that involves production,
upgrading and commercializing heavy oil (Fipke and Celli, 2008.) The steps involve include:
57
• Extract crude from the ground using cold production. This means not adding gas or steam since reservoir temperature is warm enough for crude to be pumped to the surface.
• Once at the surface the crude becomes tar-like so a diluent, or a high-grade oil product, is added to improve mobility and make it more transportable.
• It is piped to a central processing facility (in the concession) where gas, water and salt are removed. The produced water is injected underground.
• The crude is then piped 200 km to the coastal processing plant called Jose, located west of the town of Barcelona (see figure 3-2).
• At Jose it is “upgraded” which involves the removal of nickel, sulfur, vanadium, liquid petroleum gas, petroleum coke and heavy gas oil to produce a valuable higher-grade oil, called syncrude (synthetic crude). During this process the diluent is recuperated and sent via pipeline back to field operations where it is reused.
• Some of the syncrude, is sent via ships to refineries in the US Gulf coast and to Germany. The highest grade syncrude, however, is sold on the open market.
(Oil company representative, 2005a; Guerra, 2005; Talwani, 2002; Robles, 2001; Kopper et al.,
2001; Pina-Acuna and Ferreira, 2004; Briceno et al 2003; Neff and Hagemann, 2007; Gipson et
al., 2002).
The baseline year, 1990, predates the authorization by the Venezuelan National Congress
to establish the concessions (Talwani, 2002). However, oil E&P has been part of the economic
activities in the eastern Venezuela basin since the early 1900s and in the area known as the HOB
since 1935 (Talwani, 2002; Rodriguez et al., 1997; Boza and Romero, 2001; Gonzalez and
Reina, 1994; Corfield, 1948). Between 1978 and 1983 the state-company, PDVSA, carried out
an intensive exploration and evaluation program in the HOB and in 1990 began horizontal
drilling projects in areas near the present concessions of Petrozuata, Ameriven (known as Bare),
Cerro Negro and other areas (Boza and Romero, 2001; Gonzalez and Reina, 1994; Rodriguez et
al., 1997). Exploration is still ongoing (Galarraga et al., 2007) and therefore E&P patterns are
likely to be discernible in satellite images of the study area apart from the four HOB concessions.
58
The next date, 2000, corresponds to early-to-mid production phases when operations were
underway, while 2005 includes full production, when the upgrader/refinery located on the coast
was operating. In addition to examining changes in the four concessions, I also examined
changes in three control areas, which were randomly selected polygons of a uniform size to
match the average size of all four concessions, 575 km². This was done to test whether the
infrastructure expansion changes occurring in the concessions were related to petroleum E&P
actions or other activities such as agriculture (crops and cattle).
Image Processing and GIS Procedures
I utilized the following images obtained from the Global Land Cover Facility (GLCF) at
the University of Maryland and through Brazilian National Institute for Space Research (INPE):
• Landsat TM path 2, row 54, acquired 19 April, 1990 • Landsat ETM+ path 2, row 54, acquired 30 April 2000 • CBERS CCD2 path 179, row 90, acquired 12 May 2005 • CBERS CCD2 path 180, row 90, acquired 20 February 2005 • CBERS CCD2 path 180, row 91, acquired 25 January 2005 The first two Landsat scenes (path 2 row 54) encompassed the three western concessions, Sincor,
Petrozuata and Ameriven and the dates correspond to the end of the dry season. These same
concessions lay across two CBERS images, path 180 row 90 and path 180 row 91, whose dates
correspond to the middle of the dry season. I mosaicked and feathered these two scenes to
produce an image encompassing the three concessions Sincor, Petrozuata and Ameriven.
The CBERS imagery was captured two-to-three months apart from the anniversary
Landsat imagery, making it difficult to compare changes during different times of the year4
(Jensen, 2005). More importantly, due to excessive noise inherent in these particular scenes5,
4 Daniels et al., 2008 indicate that image calibration corrects for differences due to non-anniversary date images.
5 Performing a Fourier transformation followed by an inverse Fourier transformation (Jensen 2005) still did not resolve some of the problems such as striping.
59
such as missing lines of data and some corrupted bands, I did not use them in the change-
detection process but did for the subsequent measures: petroscape density, edge-effect zone and
core-areas.
All images were processed with ERDAS Imagine 9.1. I used a Landsat ETM+ 2000 WRS2
path 2 row 54 scene rectified to the Universal Transverse Mercator (UTM) 20 N projection,
WGS-84 datum, as the base image for image-to-image registration of the Landsat TM 1990
image for the same path and row 6. Subsequently, I used the rectified 2000 Landsat TM 15m
panchromatic band to rectify the CBERS 2 2005 imagery because the improved spatial
resolution aided in detecting petroscape features. As with the 1990 TM scenes, a nearest
neighbor resampling algorithm was used with a root mean square error of less than 15 m for each
point (Southworth et al., 2004). The 20-m resolution on the CBERS 2 scenes was kept in the
rectified output. An overlay function then confirmed that all the scenes overlapped exactly across
the three image dates (Southworth et al., 2004). Waypoints and routes collected in the field, as
well as road networks, oil wells and production facilities were also examined to confirm proper
overlap.
Petroenergy Maps
After scanning paper petroenergy maps at high resolution (Petroguía, 2004-2005, 2006-
2007), 1) I rectified them to the first base image. Next I digitized the outlines for the Orinoco
heavy oil belt and each concession block to create vector polygons (in a GIS7). Then I overlaid
them over the Landsat 1990 and 2000 scenes and on the CBERS 2005 mosaicked scene to
provide the location of the concessions boundaries. Finally I used the concession polygons to
6 This was accomplished using Leica Geosystems ERDAS Imagine 9.1.
7 ESRI ArcMap 9.1 was used for GIS work and analysis in this project.
60
create 9 subset images—3 each for 1990, 2000 and 2005, representing the before-during-and-
after establishment of the heavy oil operations.
Control Areas
To further test whether the infrastructure expansion changes occurring in the concessions
were related to petroleum E&P actions as opposed to other activities such as agriculture, I
randomly generated polygons (using Hawth’s Analysis tools extension for ESRI ArcGIS 9.2).
Each one was equivalent to an average size association block, or 575km². Out of a possible 15,
three were selected, referred to here as control 02, 04 and 07 respectively. These three were
chosen because they had minimal cloud cover, fit within the smaller footprint created by the
CBERS mosaicked image (113km wide) overlain onto the Landsat ETM+ (185km wide) scene,
and did not overlap current HOB concessions.
Normalized Difference Vegetation Index (NDVI)
After masking out the clouds on the subset images I calculated the NDVI vegetation index
(mentioned above under data description) for the Landsat scenes:
Band 4 (NIR) – Band 3 (R) Band 4 (NIR) + Band 3 (R)
I chose NDVI because it analyzes continuous data, rather than fixed classifications; is better at
detecting land-cover modifications; and has been used in studies involving petroscape-related
LUCC because of the contrast against healthy biomass (Southworth, 2004; Southworth et al.,
2004; Musinsky et al., 1998; Kwarteng, 1998, 1999).
Change-Detection: 1990 to 2000
Here I examined the changes in vegetation-cover between 1990 and 2000 through image
differencing (subtracting the imagery of one date from another) of the NDVI images for 3
concessions and 3 control polygons. This is a commonly used technique in change-detection
61
(Mambo and Archer, 2007) where “the subtraction results in positive and negative values in
areas of radiance change and zero values in areas of no change” (Jensen, 2005, pp. 471). This
process of selecting positive, negative and zero change values is termed thresholding and “is
rather dependent on the analyst” (Mambo and Archer, 2007, pp. 382). After examining three
threshold levels, I selected 25% since it best brought out petroscape features in the images. These
change-detection maps in figures 3-3, 3-4 and 3-5 show negative and positive changes in the
NDVI for the concessions and control polygons. Negative changes were much greater than
positive changes for all six images and involved petroscape and nonpetroscape features such as
agricultural land and roads. Meanwhile gains in the vegetation index primarily followed the river
courses, where riparian or riverine forests grow (San José et al., 2001). Figure 3-6 graphs the
negative and positive change in NDVI for the three concessions and the three control groups.
Such broad changes across all six images suggest that not only petroscape and agricultural
activities were driving land-cover change, but that climatological factors were also involved.
Figure 3-7.shows how rainfall amounts in 1989 and 1999, which preceded the capture of
the 1990 and 2000 Landsat scenes used in this study, varied from average values. The 1989 dry
season was abnormally dry, while the 1999 dry season was unusually wet. These differences in
precipitation likely affected the spectral profiles of the dry tropical savanna landscape, not to
mention river levels.
GIS Techniques: the 2005 Petroscape
The 2005 CBERS subset images were used to map infrastructure features, consisting of the
linear petro and nonpetroscape attributes in km to test if the land-cover changes found between
1990 and 2000 were related to petroscape expansion. First, I digitized the discernible road
62
network and petroscapes in the six sites (using on-screen digitizing)8. Using ground control
points of well sites gathered in the field and provided by Carrera (August 2004, 2005) along with
the 20 m resolution from the CBERS imagery it was possible to distinguish the petroscape,
which included wells and wellpads, roads and central production facilities. Nonpetroscapes were
equally discernible. In instances where cloud cover or noise obstructed features I referred to the
Landsat 2000 scene, where in some cases petroscape features where already established.
Agriculture was certainly part of the landscape as early as 1979, based on Landsat scenes of the
study area.
Once all features were digitized I overlaid this vector data onto the 1990-2000 change-
detection maps to help determine their accuracy. By analyzing the digitized full petroleum
production pattern in the concession sites as it related to the change-detection maps it was
possible to determine if observed negative change followed (and would continue to follow) the
expansion of the petroscape established by 2005. For both Sincor and Petrozuata, the change-
detection maps did fit the pattern of petroscape expansion found in 2005. Thus, this threshold
was a good predictor of ultimate changes—see figure 3-8.
Petroscape density
This measure is calculated by dividing the total length of linear features in km, by the
concession size in km² and is commonly used in landscape change studies to determine the effect
of roads and other infrastructure features on the landscape (Forman et al., 2003; Morton et al.,
2002; Thomson et al., 2004; Morton et al., 2004; Schneider et al., 2003). Habitat fragmentation is
a particularly important outcome of road and infrastructure building. It’s defined “as a process
where a large habitat area is transformed into small patches, isolated among each other by a
8 Robert Richardson, GIS Specialist at the University of North Florida, helped with this task.
63
habitat matrix different to the original one” (Ferraz et al., 2006, pp. 460). Since greater
petroscape density results in increased habitat fragmentation this suggests that best practices
regarding infrastructure building and the associated ecological effects are not implemented
(Forman et al., 2003; Morton et al., 2004; Forman and Deblinger, 2000; Morton et al., 2004).
Edge-effect zones
These zones include areas extending out from infrastructure features that experience
significant ecological effects (Forman and Deblinger, 2003). I calculated these by creating a 600
m buffer around the linear petro and nonpetroscape. This buffer distance is based on work from
three groups. First, Forman and Deblinger (2000), who calculated average value of 600 m for
what they term road-effect zones affecting moose, white-tailed deer, forest and grassland birds,
and amphibians on a suburban highway in eastern Massachusetts. Second, investigators
interested in avoidance and stress responses by caribou to oil roads, wellpads and petroleum
activities in Alaska and Canada found avoidance zones ranging from 250 m to 4 km, with 600 m
as an average value (Nelleman and Cameron, 1996; Noel et al., 2006; Schneider et al., 2007;
Wolfe et al., 2000; Bradshaw et al., 1998; Harron, 2007; Dyer et al., 2001).
Third, Thomson et al. (2005, pp. 17), noted that in open landscapes in western Wyoming
mule deer avoidance zones may be close to 500 m and that these animals “showed no evidence
of acclimating to energy-related infrastructure.” Other large wildlife in the same area, such as
pronghorn exhibited a weak avoidance of areas within 965 m, while elk avoidance zones to roads
and active oil and gas wells was 1.9 km in the summer and almost 1 km in winter (Thomson et
al., 2005). In the case of sage-grouse (birds), hens that nested an average of 1.1 km from roads
were most successful at raising chicks, while those averaging 270 km from these features tended
to have broods that did not survive the first three weeks after hatching (Thomson et al., 2005).
64
Taking these figures into account and noting that deer and tapir and parrots live in the
Venezuelan llanos (Pérez, 2000; Gorzula, 1978; Marín et al., 2007; López-Fernández and
Winemiller, 2000; Veneklaas et al., 2005; Rosales et al., 1999; Padilla and Dowler, 1994) I
selected the 600 m buffer. Figure 3-9 shows the extent of the edge-effect zones for both the
concessions and the control areas.
Core-areas
Core-areas define land beyond a given distance or effect zone from transportation areas
(Thomson et al., 2005). Here, they refer to intact habitat areas that remain after the concession
has been fragmented by petroleum production activities (in km²). These patches are important for
maintaining wildlife diversity as well as ecosystem goods and services (Forman and Deblinger,
2000; Spellberger 1998; Jones et al., 2000; Vos and Chardon, 1998). Core-areas are essentially
the land patches that remain after the edge-effect petro and nonpetroscapes are dissolved from
the concession using GIS techniques—see figures 3-10 and 3-13.
Agriculture
In order to better gauge the role of agriculture in these llanos I also digitized the
agricultural areas within the concessions and control groups. I used spectral analysis and field-
gathered waypoints to find and classify these plots, as well as the 15 m panchromatic bands and
composite Landsat images for 1990 and 2000 to delineate these areas—see figures 3-15 through
3-17. Given season dynamics at times it was difficult to distinguish dry savanna vegetation from
cleared agricultural plots.
Rainfall Patterns
Finally, I examined rainfall records for the town of Ciudad Bolívar, the nearest
metropolitan area with relevant rainfall records and the same climatological regime as the study
area. Data showed significant rainfall variation between the months of October through
65
December 1989 and 19999. These correspond to part of the dry season preceding the acquisition
of the 1990 and 2000 Landsat images.
Results and Discussion
Vegetation Change
Image differencing of NDVI images (2000-1990) of three HOB concessions and three
control groups revealed a pattern whereby vegetation biomass decreased much more than it
increased across all six sites. Table 3-2 shows the amount of land in km² per site that exhibited
NDVI losses and gains, as well as the percentage of the site that was affected. Based on the
changes in NDVI, Ameriven showed the greatest loss of vegetation cover—with almost 50% of
its land affected, followed by the three controls, then Sincor and finally Petrozuata. Figures 3-3,
3-4, and 3-5 display the patterns of land-cover change based on the NDVI. Meanwhile, Sincor
had the greatest gains in vegetation-cover, followed by Sincor. Control 04 and 06 came next,
respectively. Ameriven and Control 07 showed the least gains. Excluding Sincor and Petrozuata,
agricultural activity appears to be a predominant feature in the remaining sites.
Since the Landsat images were acquired in April, at the end of the dry season, the NDVI
highlights the healthy biomass, which is strongly affected by variations in temperature and
rainfall (Funk and Brown 2006). This is relegated to the riparian forests and agricultural areas
not yet harvested. Furthermore, as the end of the dry season approaches, a lot of agricultural and
cattle land has been burned in anticipation of the rainy season—a long-time land management
tool (Chacón-Moreno, 2004; White, 1956). All these factors help explain why gains in healthy
biomass are sparse, particularly in Ameriven and the control sites.
9 Rainfall records for the dry months of January and February 1990 were not available.
66
The detected gains in NDVI are likely due to the difference in rainfall between 1990 and
2000. Based on rainfall records of Ciudad Bolivar, October through December1989 were
markedly drier than average, while the same time period in 1999 was wetter than average (see
figure 3-7). These months with complete data, correspond to the dry season prior to the capture
of the Landsat 1990 and 2000 scenes. 1999 was the wettest year in 68 years (1921-2002, for
years with complete 12 month records), while March 2000 was the wettest March in 68 years.
Such high levels of rainfall during the normally driest months of the year apparently had an
impact on the area rivers, leading to unseasonable flooding and vegetation growth along the
riparian forests in 2000. Thus with a month’s lag time, the observed April patterns of vegetation
growth along the rivers is probably due to excess rainfall.
Petroscape
When the 2005 petroscape is overlain on the change-detection maps, it covers the bulk of
Petrozuata and Sincor, but also extends into parts of Ameriven and control 07. This measure is
an important indicator of habitat fragmentation. In Sincor and Petrozuata the petroscape
primarily extends over the areas of negative NDVI. This suggests that the 25% threshold selected
(because it highlighted petroscape features) captured areas that indeed exhibited petroscape
expansion in 2005.
Controls 02 and 04 do not appear to have a petroscape. While Petrozuata has the longest
linear km petroscape and petroscape density, followed by Sincor, control 07 also has substantial
petroscape. This site is not an HOB concession and yet it has more linear features and density
than Ameriven. This confirms that oil E&P has been part of the land history in the HOB for
decades (Boza and Romero, 2001; Gonzalez, 1994; Rodriguez et al., 1997; Galarraga et al.,
2007), though it is distinct from the strategic association business model to which Sincor,
Petrozuata and Ameriven belong.
67
It is curious that Ameriven shows the least amount of petroscape even though its
development was approved in mid-1999, initial drilling commenced in 2000 and production
started in November 2001, only three years after Petrozuata (Gipson et al., 2002; Chevron,
2001). It reached commercial production in October 2004 with the completion of the upgrading
plant, and full capacity in the first quarter of 2005 (Chevron, 2006, 2005, 2004; oil company
representative, 2005e). Therefore the full petroscape would have been established by the time the
April 2005 CBERS image was acquired. This helps support the claim by Gipson et al. (2002)
that this concession has implemented best business practices including the reduction of field
processing facilities and pipelines, and through longer horizontal wells, fewer wells, wellpads
and related infrastructure.
Table 3-3 shows important petroscape values, including linear length, density, edge-effect zones
and core-areas. While Sincor has a similar petroscape length to Petrozuata, the petroscape
density values are quite different. This is due to the concession size, where Petrozuata is the
smallest. This means that for every km² there exists 1.13 km of linear petroscape features in
Petrozuata. Sincor is next, followed by control 07 and Ameriven (with none in controls 02 and
04). Morton et al. (2004) observe that in open landscapes with little surrounding vegetation (such
as many parts of the Venezuelan llanos) road avoidance by wildlife occurs at road density rates
of 0.8 km per km² or higher. Petrozuata is the only site with values above this threshold,
suggesting that this concession may have fewer wildlife densities than surrounding areas.
The edge-effect zones appear in figure 3-9 and represent a 600 m buffer extending from
the linear petroscape. This time the land occupied is calculated in km² and Sincor shows the
highest values followed by Petrozuata, Control 07 and Ameriven last. While this measure
provides a distance from which significant ecological effects extend (Forman and Deblinger,
68
2000), when size of the edge-effect per site is calculated then nearly 75% of Petrozuata’s
concession is affected. This compares with 62% for control 07, 53% of Sincor and almost 20%
of Ameriven.
Finally, the size and distribution of core-areas is an important measure because it shows
how many land patches remain intact after the petroscape edge-effect has been dissolved. Figure
3-10 displays this pattern. This time Ameriven had the largest core-area, occupying 82% of this
concession. Next came control 07, followed by Sincor and Petrozuata—with only 26% of its land
consisting of nonpetroscape patches. It is important to note that these core-areas do not represent
intact, roadless mosaics providing unaltered habitat for local wildlife. Clearly agricultural
activity and a network of roads exist in all the sites, as the next results indicate. This core-area
measure assumes that if only E&P activities are modifying or converting the land, then the
particular patterns selected by the consortia of each concession show that Ameriven has
potentially altered the least amount of habitat. This may be due to the implementation of best
environmental business practices.
Nonpetroscape
As table 3-4 and figures 3-11 through 3-13 illustrate (see following pages), five of the six
sites displayed nonpetroscape features—primarily non-oil roads and farm roads. Ameriven, with
486 km of linear nonpetroscape had the highest values and density (0.73), primarily related to the
pine plantations and agricultural lands on its site. The actual petroscape comprised a much
smaller amount of the overall infrastructure.
Control 07 followed and had a significant nonpetroscape density of 0.69. This was
followed by control 02, control 04 and Petrozuata. Sincor did not appear to have nonpetroscape
features. The edge-effect zones showed control 07 having the broadest expanse, followed very
closely by Ameriven. Control 04 came next; followed by control 02 and Sincor with the smallest
69
amount (again Sincor had zero petroscape). Based on these patterns control 07 has the most
habitat fragmentation since the infrastructure features extend throughout the site. Ameriven, with
similar values had larger areas not affected by infrastructure.
Finally, the core-areas map, figure 3-13, shows how control 07 has the fewest intact areas
at 220.6 km². Petrozuata followed, then Ameriven, control 04 and control 02. Control 02 is
located closest to the Orinoco River, an area that is sparsely populated (Dumith, 2004), which
helps explain how it has the largest core-areas of all sites (excluding Sincor). However, when the
overall size of core-areas is normalized per concession size, then Petrozuata has the largest
percentage of core-areas, 85%, followed by control 02, control 04, Ameriven and control 07.
Petro and Nonpetroscape
Figure 3-14 shows the overlay of the petro and nonpetroscape on each site. Controls 02
and 04 appear to have the least amount of landscape disturbances, while the rest have been
clearly altered by human activity. Out of these four, the alterations in Sincor seem exclusively
tied to petroleum E&P. Petrozuata’s LUCC is centered on petroleum activities agriculture is also
evident. Based on an examination of the 1990 and 2000 images, it predated oil activities in this
concession. Out of the HOB concessions, Ameriven had the smallest petroscape, but a significant
nonpetroscape related to agriculture, in particular Caribbean pine plantations. Part of the reason
for a decreased petroscape could be the implementation of lessons learned and best practices in
this, the newest concession (Gipson et al., 2002). On the other hand, as the newest operation
perhaps it has not fully explored and drilled the site allotted to it, although Chevron (2006) says
“In 2005 the project facility reached total design capacity of processing and upgrading.” Perhaps
as production declines, new wells will be drilled to maintain capacity, which would require
further changes in land-cover. A further explanation for the reduced petroscape pattern could be
the timber damage fees that oil companies have to pay to forestry companies for removing wood
70
as part of their operations (Schneider et al., 2003; Carrera 2005). This could be offset by the high
price of oil, though world prices did not consistently reach above $30 until April 2004 (EIA,
2008).
Petrozuata and Sincor show similar linear petroscape expansions. However Petrozuata’s
smaller size means that the density of petroscape features is the greatest at 1.13 km/km². This
was the first concession established and could in part explain the greater petroscape expansion.
Had Cerro Negro been included these results would perhaps differ. Rankings
As table 3-5 highlights, Petrozuata and control 07 tied and exhibited the most alterations
related to petroleum E&P between 1990 and 2005. Sincor followed, then Ameriven. Controls 02
and 04 showed no discernible petroscape. A higher number means more disturbances. Since
positive changes in NDVI and core-areas petroscape refer to reduced alterations, their rank
values were inverted so as not to confound results. After all, a larger core-area number meant
that more land remained undisturbed.
Based on these rankings, one of three hypotheses is supported. As expected, Petrozuata,
with the highest level of local-company participation had most petroscape-related disturbances
with a rank of 25. On the other hand, Sincor, with European-company dominance did not show
the least landscape alterations, ranking 22. Instead, Ameriven, with US-company dominance
showed the least, though it was expected to rank second, below Petrozuata. It ranked 21. Control
04 ranked 18 and control 02, 15. Nevertheless these two showed no petroscape, but since they
were included in the rankings, their zero values were assigned values of 1.5. Consequently
though they had no core-areas, when the values were inverted, instead of receiving ranks of 1.5,
they got 5.510.
10 Not inverting these rankings for Controls 02 and 04 did not change rankings.
71
Conclusion
When the important petroscape values are ranked (NDVI change, linear and petroscape
density, edge-effect zones and core-areas) Petrozuata exhibits the greatest changes, supporting
the hypothesis that it would. Sincor, where European companies dominate did not show the
smallest alterations as expected. Instead, Ameriven, where US companies have the greatest
presence showed the least alterations. Thus Sincor and Ameriven switched rankings. Perhaps
most surprising was finding that one of the control sites, 07, exhibited a substantial petroscape,
tying with Petrozuata. While this block is not one of the strategic associations to which the oil
concessions examined belong, it confirms that E&P is part of the land-use history in the HOB.
The area where control 07 is located is known as the Bare field, where more than 150 wells had
been drilled by 2002 (Gipson et al., 2002).
Though only three out of four concessions were examined in this paper, including three
control groups helped increase the sample and determine which types of economic activities are
most prevalent in this part of the eastern llanos. The methods used here provide an innovative
approach to studying landscape change related to nonrenewable resource extraction. Future work
could build on these findings by implementing these methods to establish baseline conditions
prior to subsequent E&P. This would help implement best practices during the planning phase in
order to reduce associated environmental changes. Subsequent efforts would continue to use
remote sensing and GIS techniques as a cost-effective way to monitor the expansion of
operations throughout the project’s life cycle and then during recuperation phases.
Finally, while these results are based on a very small sample, they do indicate that US-
company participation may be influential in the implementation of best business practices since
the Ameriven concession, which had a 70% US-company participation exhibited the least
petroscape alterations. Consequently, two US MNOCs recently stopped operating in Venezuela,
72
ExxonMobil and ConocoPhillips, while the Local company increased its participation to a
majority status in all four concessions. ConocoPhillips was a partner in Ameriven, while
ExxonMobil was a partner in Cerro Negro.
If E&P continues to expand as planned by the Venezuelan government and the Local
company, then the HOB stands to experience alterations to 18,000 km², or 33% of the HOB land
area, see figure 3-18. The types of business partners selected may likely affect the extent of
petroscape related alterations.
73
Figure 3-1. Venezuela, the heavy oil belt and concessions, and surrounding countries 2005
(Petroguía, 2006-2007; ESRI Geographic Network Services 2008). The yellow outline delimits the heavy oil belt, which is divided into four E&P zones, and spreads across the dry tropical savanna landscape. The red smaller polygons show the location of the four concessions.
74
Figure 3-2. Venezuela’s heavy oil belt, its four concessions: Sincor, Petrozuata, Ameriven, and Cerro Negro, and the upgrading/refinery plant called Jose located on the coast of Anzoátegui state, 2005. Though the HOB spans Guárico, Anzoátegui and Monagas states, the four concessions are all located in Anzoátegui (ESRI 2006; Petroguía, 2006-2007). From west to east the four heavy oil belt exploration and production zones, marked by the red divisions are: Machete, Zuata, Hamaca and Cerro Negro. However, they have now been renamed to Boyacá, Junín, Ayacucho and Carabobo (PDVSA, 2005a, 2005b; Kopper et al., 2001; Balke and Rosauer, 2002). Note: the HOB also reaches into the eastern state of Delta Amacuro, but its reach is so slight it is not included on this map or mentioned in the text.
75
Figure 3-3. Vegetation-cover change across all six sites (three concessions and three controls)
between 1990 and 2000. The areas in red indicate a decrease in NDVI by 25% or more. The dark background represents areas of some increase, unchanged, or some decrease. In Sincor and Petrozuata, the petroscape becomes discernible as lines (roads) and polygons, production facilities are established.
76
Figure 3-4. Increases in Normalized Difference Vegetation Index (NDVI) across all six sites between 1990 and 2000. The green areas represent gains in NDVI by 25% or more. The dark background represents areas of some increase, unchanged, or some decrease. Note that green areas are mostly riparian forests following river courses in the study area. These smaller changes in vegetation cover contrast with the large changes in figure 3-3.
77
Figure 3-5. NDVI losses and gains for all six sites between 1990 and 2000. Red indicates areas of NDVI increase, and green gains. The agricultural landscape becomes evident in control 07 and in Ameriven, where rectangular polygons represent pine plantations in the area.
78
A
B
Figure 3-6. Gains and losses of the Normalized Difference Vegetation Index (NDVI) for the three concessions: A) Sincor, B) Petrozuata, and C) Ameriven, as well as the three control groups: D) Control 02, E) Control 04, and F) Control 07.
81
0.00
20.00
40.00
60.00
80.00
100.00
120.00
Month
1989 13.50 39.30 17.30
Average 91.80 59.60 41.50
1999 99.00 72.00 94.00
October November December
Figure 3-7. Rainfall amounts (in mm) for Ciudad Bolívar for the October, November and December dry months preceding the 1990 and 2000 Landsat images used in this study. Note that all three months were drier in 1989 than average values and 1999 measures, helping explain the phenological differences in both images.
82
Figure 3-8. Decrease in NDVI between 1990 and 2000 overlain by 2005 petroscape features. The red lines show the extent of the petroscape in four sites. Controls 02 and 04 displayed no discernible petroscape. Petrozuata has the longest petroscape in linear km and density, followed by Sincor, control 07 and finally Ameriven.
83
Figure 3-9. Edge-effect zones, or areas where significant ecological effects extend (Forman and Deblinger, 2000), appear in red. Based on the petroscape linear features, these 600 m buffered zones indicate how much land of each site is potentially affected by infrastructure features which in turn have negative consequences for wildlife and ecosystems (Morton et al., 2004; Thomson et al., 2005; Forman and Deblinger, 2000). Controls 02 and 04 displayed no petroscape edge-effect zone. The white background areas show the negative changes in NDVI between 1990 and 2000. The areas in black represent little positive or negative change as well as areas of no change.
84
Figure 3-10. Core-areas, marked in green, show size and distribution of the patches of land that remain after dissolving the edge-effect zones in figure 3-7. Controls 02 and 04 displayed no petroscape core-areas.
85
Figure 3-11. Linear nonpetroscape 2005. The yellow lines represent roads related to agriculture and transportation not associated with petroleum activities. While Sincor appeared not to have a nonpetroscape, Petrozuata displayed the smallest one out of the remaining sites. This indicates that these two concessions are petroleum E&P zones. Controls 02 and 04 had some transportation and agricultural roads across their sites, but Ameriven and control 07 had the most. The checkerboard pattern near the center of Ameriven shows the 1-km blocks typical of the Caribbean pine plantations in this part of Venezuela.
86
Figure 3-12. Edge-effect nonpetroscape zones. Yellow lines indicate the extent of the edge-effect zones. With a 600 m buffer extending out from the linear nonpetroscape both Ameriven and control 07 infrastructure disturbances stretch across a significant amount of their plot. They are followed by control 04 and control 02. Petrozuata has the least and Sincor was not included because of no discernible nonpetroscape.
87
Figure 3-13. Core-areas that remain after nonpetroscape features are dissolved from each site. Overall controls 02 and 04 had the largest areas, followed by Ameriven, Petrozuata. Control 07 had the least.
88
Figure 3-14. Petro and nonpetroscapes across all 6 sites. Red indicates petro, yellow reflects nonpetro. The NDVI background has been cleared to make the two types of features more discernible. Control 07 exhibits high levels of both. Ameriven shows more nonpetro than petroscape features while Petrozuata shows a petroscape mostly.
89
Figure 3-15. Agricultural patterns in the three concessions and the three control groups in 1990. Both Ameriven and Control 07 had more than half their area dedicated to agriculture.
90
Figure 3-16. Agricultural patterns in the three concessions and the three control groups in 2000. Activity appears to have increased everywhere except in Ameriven—see figure 3.17.
91
0
50
100
150
200
250
300
350
400
450
Location
Km
2 of
Lan
d
AGR 1990 Amt Land Use Km2 1.52 21.27 402.78 35.47 28.39 252.28AGR 2000 Amt Land Use Km3 2.32 40.58 373.93 42.63 40.31 412.8
Sincor Petrozuata Ameriven Control 02 Control 04 Control 07
Figure 3-17. Amount of agricultural land (in km²) found in the three concessions and three control groups in 1990 and 2000. Note how Ameriven and Control 07 have large amounts of their territory dedicated to agriculture. In Ameriven’s case, Caribbean pine plantations extend through the concession.
93
Table 3-1. Venezuelan heavy oil belt strategic associations, company ownership by percentage, syncrude (synthetic oil) production and upgraded quality 2005 (Petroguía, 2005-2006; Talwani, 2002; Mommer, 2004).
Sincor Petrozuata Ameriven Cerro Negro
Company Ownership
PDVSA 38% Total (France) 47% Statoil (Norway) 15%
PDVSA 49.9% ConocoPhillips 50.1%
PDVSA 30% Chevron 30% ConocoPhillips 40%
PDVSA 41.67% ExxonMobil 41.67% BP(United Kingdom) 16.66%
Syncrude Production and Upgraded Quality
180,000 b/d From 8.5º to 32º API
104,000 b/d From 9º to 19-25º API
190,000 b/d From 7-10º to 26º API
105,000 b/d From 6º-10º to 16º API
Congressional Authorization
1993 1993 1997 1997
Early Production
2001 1999 2002 1999
Full Production (upgrader Start-up)
Mar 2002 Jan 2001 Oct 2004 Aug 2001
Table 3-2. Change-detection measures showing amounts of positive and negative change per site in km² and as a percentage of the site area.
Concession/ Site
Negative NDVI km²
Positive NDVI km²
% Veg Loss
% Veg Gain
Sincor 77.08 41.36 14.90 8.20 Petrozuata 39.07 21.00 12.99 6.95 Ameriven 309.90 2.19 46.57 0.33 Control 02 235.70 17.33 40.81 3.23 Control 04 208.95 23.12 36.15 3.86 Control 07 296.03 0.35 51.35 0.06
94
Table 3-3. 2005 petroscape measures for the six study sites.
Concession/ Site
Concession Area (km²)
Petroscape Linear 2005 (km)
Petroscape Density 2005 (k/km²)
Edge Effect Petroscape 2005 (km²)
% of Site Affected by Edge-Effect
Core Area Petroscape 2005 (km²)
% of Site Affected by Core Areas
Sincor 517.00 302.33 0.58 271.57 53. 00 245.80 48.00 Petrozuata 300.00 338.70 1.13 222.49 74.00 78.29 26.00 Ameriven 665.00 119.63 0.18 117.07 18.00 548.37 82.00 Control 02 575.55 0.00 0.00 0.00 0.00 0.00 0.00 Control 04 578.00 0.00 0.00 0.00 0.00 0.00 0.00 Control 07 576.50 153.86 0.27 164.80 62.00 411.70 72.00
Table 3-4. 2005 nonpetroscape measures for the six study sites.
Concession/ Site
Concession Area km2
Non petroscape Linear 2005 km
Non petroscape Density 2005 k/km2
Edge Effect Non petroscape 2005 km²
% of Site Affected by Edge-Effect
Core Area Non petroscape 2005 km2
% of Site Affected by Core-area
Sincor 517.38 0.00 0.00 0.00 0.00 0.00 0.00 Petrozuata 300.77 47.39 0.16 45.61 15.16 255.17 84.84 Ameriven 665.44 486.02 0.73 348.32 52.34 317.11 47.65 Control 02 577.55 168.26 0.29 148.62 25.73 428.93 74.27 Control 04 578.00 128.78 0.22 152.36 26.36 425.73 73.66 Control 07 576.50 397.48 0.69 355.91 61.74 220.60 38.27
Table 3-5. Overall rankings for land-cover change and petroscape features in the 6 sites. A higher number means greater overall landscape disturbances. Note: positive NDVI and core-area petroscape values were inverted to maintain rank consistency.
Concession/ Site
Negative NDVI km²
Positive NDVI km²
Petroscape Linear km²
Petroscape Density 2005
Edge Effect Petroscape 2005 km²
Core Area Petroscape 2005 km2
Overall Rank
Sincor 2 1 5 5 6 3 22 Petrozuata 1 3 6 6 5 4 25 Ameriven 6 5 3 3 3 1 21 Control 02 4 4 1.5 1.5 1.5 5.5 18 Control 04 3 2 1.5 1.5 1.5 5.5 15 Control 07 5 6 4 4 4 2 25
95
CHAPTER 4 SPATIAL ANALYSIS AND STATISTICAL CORRELATIONS AMONG PETROSCAPE
FEATURES IN THE HEAVY OIL BELT
Introduction
Worldwide, heavy and extra-heavy oil comprise about 40% to 50% of total petroleum
resources (Ascencio-Cendejas and Reyes-Venegas, 2006; Khelil, 2006). Compared to traditional
petroleum, this tar-like crude requires different financial and technological inputs to convert it to
usable petroleum products and until the late 1990s was not an attractive investment (particularly
in Venezuela). Today, increased demand and higher prices mean that exploration and production
(E&P) activities for heavy oil are not only viable but will expand in the next decade especially in
Canada, Mexico, Brazil and Venezuela11 (Ascencio-Cendejas and Reyes-Venegas, 2006). In
Canada, Alberta’s oil sands deposits total 95% of that country’s proven reserves (EIA, 2007a).
By 2020 Mexico’s heavy oil production is expected to increase to 50% of overall national
production (Ascencio-Cendejas and Reyes-Venegas, 2006).
In Venezuela current heavy oil production is centered in the eastern part of the country
known as the heavy oil belt (HOB), located on the plains, or llanos, north of the Orinoco River
(see figure 4-1). Current production is around 635,000 barrels per day12 (b/d) or about 21% of
Venezuela’ total production—which ranges from 2.8 to 3.6 million barrels per day (mb/d)
depending on the source (OPEC, 2007, EIA, 2007a; BP, 2007).
In the late 1990s Venezuela began four concessions—also known as strategic associations
in the HOB, comprised of public-private partnerships between seven companies. One local
company, state-owned Petróleos de Venezuela Sociedad Anónima, or PDVSA, had an ownership
11 In addition to these countries, other major deposits are found in the US, Trinidad, Madagascar, Albania, Rumania, and Russia and China (Phizackerley and Scott 1978; Jiayu and Jianyi 1999; Zhang et al., 2007).
12 An oil “barrel” is 42 US gallons, or 158.984 liters.
96
stake in each of the four concessions. The other partners were multinational oil companies
(MNOCs) from Europe and the US, each with different levels of participation and investment,
see table 4-1.
The companies operated until mid-2007, when Venezuela’s President Hugo Chávez
changed the contract terms 25 days after the National Congress passed a law allowing him to rule
by decree in certain matters (Gaceta Oficial, 2007). The new law (number 5,200) required the
HOB operations to reorganize. The local company, PDVSA, increased its role by becoming a
majority partner and decision maker. This reduced the role of the MNOCs and provided them
four months to agree to the new terms or cease operations (see figure 4-2). Two companies,
ExxonMobil and ConocoPhillips withdrew not only from the HOB, but from Venezuela
(ConocoPhillips, 200813). This study, concerns the time period 1990-2005, before these changes
took effect.
The global research question is: what are the environmental implications of any natural
resource development done by local institutions supported by governments, multi-national
companies subject to international scrutiny, or a consortia of locals and multinationals overseen
by governments? More specifically in this chapter the local research question is: does the
government make it easier for the local oil company to avoid environmental regulations as
evidenced in the amount of important land-cover changes measured in each concession?
I address the relationship between the observed land-use land-cover changes (LUCC)
occurring in the HOB, particularly between 2000 and 2005, and the three types of partnerships
defined by the role of participation among the concessions. While geological events explain the
presence of petroleum—which is often located in sedimentary basins such as Venezuela’s
13 ExxonMobil moved its case to international arbitration, resulting in PDVSA having $12 billion in assests frozen in the US. This ruling was lifted.
97
Orinoco basin, the two actors extracting oil are the state and multinational oil corporations
(MNOCs) because they have the resources, equipment, expertise and legal framework to provide
global energy supplies.
Production in the Heavy Oil Belt
As table 4-1 shows, the four concessions were comprised of seven companies. Their
extraction and production model was to:
• Extract heavy crude from HOB wells at reservoir temperature—this means that no heat or steam injected into the wells was necessary to pump the crude to the surface (due to its tropical location). This practice is now changing with the intention to increase recovery rates, since higher reservoir temperatures reduce viscosity. However, these practices also involve the use of solvents inside the reservoir which increases the potential for air and water pollution.
• Add a diluent at the wellhead. Once it reaches the surface this heavy crude is tar-like and requires “thinning” which is accomplished by adding high-grade petroleum which allows the extracted crude to flow freely through pipelines.
• Once diluted the heavy crude is sent to the central processing facility in the concession area where it is degassed, dewatered and desalted.
• From there it is sent via a 200 km pipeline to the coastal upgrading/refinery plant called Jose, where it is converted to syncrude, which is short for synthetic crude oil. This process removes the high levels of vanadium, nickel and sulfur and produces a valuable petroleum product.
• Finally, depending on the quality of the upgraded syncrude (see table 4-1), it is sent to offshore refineries (in Germany or the US), or is sold as is. Sincor produces the highest-quality syncrude, followed by Ameriven, Petrozuata and finally Cerro Negro.
(oil company representative, 2005a, 2005d; Guerra, 2005; PDVSA, 2008a, 2007e; 2005a;
Petroguía, 2006-2007; Pawlewicz ,1995; Tankersley and Waite, 2002, 4; Galarraga et al., 2007).
An important issue concerning LUCC in Venezuela concerns government plans to increase
overall production to 5.85 mb/d by 2012. Much of this growth is expected to come from the
HOB as 27 more operations are developed by 2030 (oil company representative, 2005a; PDVSA,
2007h) This time, however, the local state-company will have a majority role in each operation
98
and mainly rely on national oil company (NOC) partnerships from other countries (PDVSA,
2007a; 2006; 2005; Petroguía, 2006-2007). This combination of public-public partnerships will
produce particular E&P landscapes with possibly greater or lesser amounts of landscape
alterations than under the old partnerships.
Land-Use Land-Cover Change
Lambin (2001) observes that LUCC is a function of individual and social responses to
opportunities and constraints for new land uses created by markets and policies (Lambin, 2001).
In this context, positive incentives are the opportunities available while negative ones are the
constraints—operating in the framework of energy markets and energy policies. In the HOB the
positive incentives are to control, extract, produce, refine and sell very large amounts of
profitable heavy crude using proven technology in a mostly uniform seasonal savanna landscape
(Chacón-Moreno, 2007).
Some disincentives include barriers to entry because of the high capital costs, advanced
technology and expertise needed. These issues are less challenging for MNOCs than for local
companies. A main challenge for MNOCs, however, is working in a country where energy laws
are evolving, where previous contracts are changed, where capital and technological investments
may not be fully realized, and where the state now requires MNOCs to work under its leadership
and rules. In Venezuela this framework is shaped by a state-centered political economy funded
by petro-dollars and marked by strong state control and centralized planning; price controls;
fixed currency rates; increased social spending on the poor (the government’s political
supporters); increased state control of key industries such as telecommunications, electricity,
banking and steel; anti-privatization measures; increased state and reduced private property
rights; and complete control over the local oil company, PDVSA.
99
Such command-and-control governance strategies tend to discourage innovation in
behavior and technology, are less effective and economically inefficient (Dietz et al., 2003).
Furthermore, when local oil companies don’t operate internationally or compete for contracts in
other countries, they may not have an incentive to reduce social and environmental alterations
tied to E&P activities (Diamond, 2005). But does the level of local company participation
actually affect environmental performance when it comes to heavy oil exploration and
production in eastern Venezuela? What about the level of European or US company
participation? Referring to energy production in Alberta, Canada, Timoney and Lee (2001, pp.
388) observe that “economic activity leads to environmental degradation unless ecosystem-based
management is integrated into economic decision making” In this context, a higher level of local
company participation in Venezuela would correlate with greater levels of environmental
alteration.
Expected Findings
This study tries to answer the question: do governments make it easier for local companies
to avoid environmental regulation while MNOCs meet or exceed regulations as evidenced in the
amount of important land-cover changes measured in each concession? In terms of company
dominance the HOB concessions exhibit the following patterns: Ameriven 70% US; Sincor 62%
European; Petrozuata 49.9% Local; Cerro Negro tied—41.67% Local, 41.67% US.
Using statistical correlation analysis I test three hypotheses related to the three levels of
company participation and the LUCC measures examined in the previous chapter and expanded
in this chapter.
The hypotheses are:
• Local company-dominated concessions will exhibit the greatest changes in land-cover
• US company-dominated concessions will show lower levels of LUCC
100
• European company-dominated concession will exhibit the lowest levels of LUCC
These hypotheses are based on a review of petroleum geology, LUCC, organizational and
natural resource management literature particularly in relation to oil and gas E&P, field
interviews with oil company representatives and results from the previous chapter (Thomson et
al., 2005; Morton et al., 2004, 2002; Wilderness Society, 2006; Musinsky et al., 1998; Smith et
al., 2003; Moser, 2003; oil company representative, 2005a, b, c, d, e, f, g, h).
The first hypothesis is derived in part from the regional context, where the local oil
company is strongly controlled by the federal executive office; where political and economic
power is concentrated (McCoy and Fensom, 2006, 17; CIA, 2008). Such entrenching of political-
economic institutions often lacks checks and balances (Timoney and Lee, 2001). In Venezuela
petroleum has been the traditional export earner since the 1920s (when it overtook coffee) and
currently encompasses about 90% of export earnings, 50% of federal budget revenues and about
30% of gross domestic product (GDP) (CIA, 2008; EIA, 2007b; Trujillo, 2004; Torres, 2004;
Williams, 2006).
McCoy and Fensom (2006, pp. 8) note that in Venezuela “adherence to the rule of law is
low and the prevalence of corruption is high.” Also concerning Venezuela’s business
environment, Lopez-Claros (2006) states the following:
What is especially noteworthy about Venezuela is that there has been a notable deterioration in the quality of the business environment, with serious concerns about property rights, the independence of the judiciary, waste in government spending and overall levels of corruption. Venezuela is not using oil revenues to enhance the economy’s competitiveness such as through a major improvement in the educational system or in the country’s dilapidated infrastructure.
While at the regional level, Chacón-Moreno (2007, pp. 3-4) observes: “Government
policies are focused on increasing the use and exploitation of the Llanos areas in order to elevate
101
the quality of life of the Venezuelan people, but most of the time, these policies lack
environmental controls.”
Since petroleum is such a central component of the Venezuelan economy and crucial to
large domestic social programs and international energy agreements promoted by the current
administration (PDVSA, 2008b), activities that reduce production and therefore profits, such as
meeting environmental regulations, may not be fully enforced on local companies since
ecological concerns may not be part of the decision making process (Timoney and Lee, 2001).
For example, the Ministry of the Environment and Natural Resources (MARN) requires a
complete environmental impact assessment (EIA) report and a register of activities susceptible to
degrade the environment (RASDA) report for the establishment of new industrial facilities,
including oil operations (Sebastiani et al., 2001). However, Sebastiani et al. (2001, pp. 140)
reported that MARN “has adopted a conciliatory and not enforcing attitude once a facility is in
operation.” This facility was the upgrader for the Ameriven HOB concession (Sebastiani et al.,
2001).
Another issue concerns oversight of one government agency such as the Ministry of
Energy and Petroleum by another like MARN. This sets up a conflict of interest whereby one
government (environmental) agency is responsible for monitoring another powerful state-
controlled company whose earnings comprise key government and export revenues, help fund
sizeable social programs14 benefiting the President’s political base, finance international oil-
driven agreements benefiting its allies such as Ecuador, Bolivia, Argentina, Nicaragua and Cuba,
and pay for infrastructure projects.
14 These social programs include 17 Missions that foster education, health services, nutrition and microenterprises (MENPET 2007).
102
Wages earned by government employees may also be a contributing factor because “most
government officials are poorly paid, which encourages acceptance of bribes, and this is
particularly likely when they are responsible for managing natural resources with high financial
value” observe Smith et al. (2003, 68). Finally, in cases of negligence the judicial branch may
not have the capacity (will and/or desire) to bring legal action against a state-owned energy
company.
The second hypothesis, that US company-dominated concessions will show lower levels of
LUCC, is based on their technological know-how and experience in addressing environmental
changes related to E&P worldwide, as well as international scrutiny by governments, non-
government organizations (NGOs), sustainability rating agencies and investors such as the Dow
Jones sustainability world index (Dow Jones, 2007a; 2007b; Bruni, 2007), World Economic
Forum (WEF, 2005); the Global 100 (Global 100, 2008); Management and Excellence (M&E,
2008, 2007, 2006, 2005); the Pacific Sustainability Index (PSI, 2007) from Claremont McKenna
College; and the Environmental Sustainability Index (ESI, 2005) from Yale University and
Columbia University15. A traditional view in organizational theory proposes that a company’s
obligations extend only to its shareholders (Dashwood, 2007). This perspective has dominated in
the US but is beginning to change, and eventually environmental reporting may be as
commonplace as financial reporting is today (Morhardt, 2002).
The third hypothesis, that European company-dominated concession will exhibit the lowest
levels of LUCC, is also supported by their technological know-how as well as a longer history
and culture of sustainable business practices in Western Europe (Lawrence, 2007). In this region,
sustainability is a concept tied to a sense of balance between the market and the social sphere
15 In future work I will examine the statistical relationship of these indices with the observed LULCC in the HOB.
103
(Lawrence, 2007), which contrasts with the objective of maximizing profits every quarter. The
European Union, for example, has adopted stricter environmental standards and their policies are
based on the precautionary principle whereby insufficient or inconclusive scientific evidence is
considered when reaching policy decisions such as bans on questionably hazardous products
(Schapiro, 2007). Also, some of the sustainability indices mentioned above, very often rank
European companies higher in environmental and social performance than US companies, while
Venezuelan companies and government rank at the bottom (M&E, 2008, 2007, 2006, 2005;
Global 100, 2007, 2008; ESI, 2005).
Given the declining production from conventional wells; the difficulty in finding and
securing new oil sources; the importance of maintaining a positive company image for
shareholders, the general public and state governments; and the challenge of establishing new
business ventures in countries with increased state control over the energy industry; it is in the
interest of both US and European MNOCs to engage in sustainable practices that reduce
environmental and social changes related to their E&P operations. This concept of sustainability
refers to a company’s capacity to reduce environmental and therefore social disturbances
resulting from business operations while still managing a profitable enterprise and maintaining a
healthy ecosystem for future generations (Lovins et al., 2000). This same concept holds true for
local companies that want to compete internationally.
Methods
I first used GIS and remote sensing techniques outlined in the previous chapter to expand
the land-cover change findings to 16 measures (noted below). Then I used bivariate correlation
analysis to determine if a statistically significant relationships exists, as well as the strength of
the associations between the amounts of 1) local (PDVSA); 2) European; or 3) US participation
in the four HOB operations and the magnitude of resulting land-cover changes
104
LUCC Related to the Petro and Nonpetroscape
All these landscape-level changes for the HOB concessions were measured using remote
sensing and GIS techniques (Baynard 2007). They were:
• NDVI negative change—the index amount of vegetation in km² that showed a negative change based on the Normalized Difference Vegetation Index (NDVI) between 1990 and 2000. This index is a model that helps detect vegetation vigor and is often used as an indicator of biomass (Jensen, 2005; Fung and Siu, 2000; Mangiarotti et al., 2008).
• NDVI positive change—the index amount of vegetation cover that showed increases between 1990 and 2000 measured in km² and using the NDVI.
• Petroscape density—the linear amount of petroscape features in km divided by the concession size in km².
• Nonpetroscape density—the linear amount of agricultural and other nonpetroscape features in km/km².
• Allscape density—the combined density of petro and nonpetroscape features in km/km².
• Concession area—the size of the E&P zone in km².
• Edge-effect zone petroscape—the area extending 600m from petroscape features in which significant ecological effects are expected to extend in km². The 600 m buffer was selected after reviewing work by Forman and Deblinger (2003) regarding roads, Schneider et al. (2003) and Dyer et al. (2001) in reference to energy roads in Canada, and Morton et al. (2004), Thomson et al. (2005) and the Wilderness Society (2006) regarding LUCC and energy development in the western US.
• Edge-effect zone nonpetroscape—the area extending 600m from nonpetroscape features in which significant ecological effects are expected to extend in km² (Forman and Deblinger, 2003; Schneider et al 2003; Morton et al., 2004; Thomson et al., 2005; Wilderness Society, 2006).
• Core areas petroscape—the size in km² of intact areas that remain after the effects of petroscape features (edge-effect zones) are accounted for in each concession (Thomson et al., 2005; Morton et al., 2004; Wilderness Society, 2006; Morton et al., 2002)
• Core areas nonpetroscape—the size in km² of intact areas that remain after the effects of nonpetroscape features (edge-effect zones) are accounted for in each concession.
105
• Number of wells—the total number of wells and wellpads identified in the 2005 satellite images16 of the concession areas.
• Density of wells—the number of wells and wellpads per concession area identified in the 2005 satellite images.
• Wells edge-effect zone—the area extending 600m from the wells and wellpads/km² identified in the 2005 satellite images of each concession. The same buffer distance was selected as with edge-effect zone petroscape.
• Wells core-areas—the size in km² of intact areas that remain after the edge-effects of the 2005 wells and wellpads are accounted for in each concession.
• Number of rivers crossed—the sum of river crossings created by petro and nonpetroscape features in each concession. Since many infrastructure features reduce connectivity between rivers and their floodplains (Stein et al., 2002), a lower number implies fewer ecosystem disruptions.
• Riverscape density—the linear amount of rivers/km². These rivers were measured in each concession using Landsat 2000 images of the study region17.
The Coefficient of Determination
An examination of scatterplots provided an R² coefficient of determination showing how
well a regression line approximated real data points for the different types of company
participation. With a small sample size (n=4), coefficients below 0.70 are not considered
significant—and even those above 0.70 require further scrutiny, see table 4-3.
NDVI change (negative and positive) was measured between 1990 and 2000 for three of
the concessions; Cerro Negro was excluded. Most of the measures discussed here concern 2000
and 2005 (or both) because of data availability and the fact that HOB operations did not begin
until the late 1990s.
16 These subset images were created using Leica Geosystems Erdas Imagine 9.1 software from the China-Brazil Earth Resources Satellite (CBERS) CCD sensor with a 20 m spatial resolution.
17 The 2000 image date was selected since it coincided with high seasonal flooding (following an abnormally heavy rainy season—see chapter 3) and thus showed a full extent of area rivers.
106
Nonparametric Statistical Analysis
For the next step I selected a nonparametric test because the sample size was small (n=4)
and the data did not meet the assumptions of parametric inferential statistics (Ferguson et al.,
1999; Lehmkuhl, 1996). As with research regarding environmental and sustainability reporting,
note Morhardt et al. (2002), different types and intensities of environmental alterations by
different companies are best measured as categorical rather than continuous variables, and
therefore are inherently appropriate for nonparametric statistical analysis.
Kendall’s tau-b rank correlation coefficient measures a precise degree of association
between two sets of ranks (Burt and Barber, 1996), rather than their actual values. This addresses
differences in magnitude in the data which was necessary since the variables in my dataset were
measured in different units (e.g. area in km² of vegetation cover loss, percentage of company
participation and number of wellpads).
Like Pearson’s correlation coefficient Kendall’s tau-b has the following properties: 1) a
value of +1 means that the agreement between two ranks is perfect; 2) a value of -1 means the
disagreement between two ranks is perfect; 3) other arrangements have values that fall between
+1 and -1; and 4) a value of 0 means that the rankings are independent (Rumsey, 2007).
However, unlike Pearson’s, Kendall’s (and other nonparametric measures like Spearman’s rank
correlation) work with the median rather than the mean, making this statistic more flexible since
it’s not affected by outliers (Rumsey, 2007).
Kendall’s tau-b is calculated by comparing each pair of ranked observations. Rankings that
are in agreement, or ordinally correct, are considered concordant, while those that are not in
agreement are discordant (Burt and Barber, 1996). This measure compares all pairs of
observations, records the number of concordant and discordant pairs and uses the function:
107
• τ = NC– N D n(n-1)/2
(Burt and Barber, 1996).
Furthermore, Kendall’s tau-b (as opposed to tau-a) considers ties. Pett (1997, 268)
explains: “the range of possible values for this coefficient is smaller because the number of
concordant and discordant pairs will always be smaller than the total number of pairs
(concordant + discordant + ties).”
Since a correlation coefficient of 0 means the paired variables are independent, this is used
to set the null hypothesis. Here, H O: τ= 0, whereby the amount of local company participation
paired with each of the 16 ecological variables is independent (i.e., there is no relationship). The
alternate hypothesis is: HA: τ≠ 0, whereby the paired variables are dependent. The same
hypotheses hold for European-company participation and for US-company participation. Table
4-5 shows the matrix for Kendall correlation coefficients and significance levels for the three
types of company consortia and the 16 measures analyzed above for the R² correlation of
determination, as well as the R² scores.
The decision rule to determine significance with a sample of 4 is to compute a sampling
distribution containing a finite number of possible combinations between paired variables ranked
1 through 4 (Abdi, 2007). Excluding ties, the number is 24, which means that an exact
correspondence in rank would occur only 1 in 24 times, or .042, which is less than the
conventional .05 cutoff for significance (Abdi, 2007; Richard, 2008; Rumsey, 2007). In order to
use a test statistic of significance of Kendall’s tau, such as the Z score, the sample size n must ≥
10 (Burt and Barber, 1996; Abdi, 2007; Lehmkuhl, 1996).
108
Results and Discussion
Determining the R² Coefficient
Table 4-2 shows the R² coefficients of determination for seven important petroscape-
related LUCC measures, while table 4-3 displays all LUCC values. Local participation showed a
strong downward association (R² ≥0.9) with core-area petroscape for 2000 and 2005; a upward
relationship with petroscape density (R² ≥0.73); and an downward relationship with negative
NDVI (R² ≥0.76). This means that for Petrozuata, core-areas were smaller and petroscape
density was higher which mean larger alterations. However, negative-changes in vegetation-
cover were smaller indicating that local-company ownership may not be related to negative
changes in land cover, but rather US or European ownership, or other explanations.
European participation showed only one statistically significant correlation with positive
NDVI (R²=0.770). But how does a concession exhibit vegetation-cover gain over time if E&P
activities are increasing? In the case of Sincor it had the largest number for total length of rivers
for the four concessions, at 561.40 linear km. Thus when the rivers flooded during an abnormally
wet dry season, captured in the 30 April, 2000 Landsat ETM+ image (described in chapter 3), the
numerous riparian forests expanded, likely revealing the reported positive changes (gains) in
NDVI. US-company presence was correlated with larger core areas presumably because of lower
petroscape density in Ameriven.
Nonparametric Statistical Analysis
Table 4-4 shows the Kendall’s tau-b coefficients for seven important petroscape-related
LUCC measures, while table 4-5 displays all LUCC values.
109
European and US results
Results for Kendall’s tau show that no variables exhibited significant correlation with
European participation, thus I fail to reject the null hypothesis that European company
participation is independent of the LUCC variables examined.
Six landscape measures showed ± 1.0 correlation for US participation, but only three had
statistically significant levels. In this case they were all related to the nonpetroscape, such as the
roads created by crop agriculture and pine plantations. Thus, I fail to reject the null hypothesis
that US-participation is independent of any of the important landscape-change measures
necessary to monitor direct E&P activities during the life-cycle of a given concession. Therefore,
attention must turn to local-company participation.
Local results
Local participation is strongly positively correlated with petroscape density, well density,
number of wells and river density (1.0, p < 0.01). The first two suggest that best practices are not
implemented in terms of minimizing seismic and future roads or not closing off unnecessary
roads resulting in further alterations to wildlife and ecosystems (Lee and Boutin, 2006;
Wilderness Society, 2006; Schneider et al., 2003). It can also indicate that drilling practices focus
on quantity of wells, rather than employing long lateral wells (≥ 1,500 ) that can reach oil sands
from far away thus increasing production while reducing well count (Gipson et al., 2002). Best
practices, however, apply to more than wells. As Kharaka and Dorsey (2005, pp. 61) note, the
cumulative ground-surface disturbances from oil E&P can be high, since these activities involve:
“site clearance, construction of roads, tank batteries, brine pits and pipelines, and other
modifications necessary for the drilling of exploration and production wells and construction of
production facilities.”
110
The correlation between well density and Petrozuata may be related to being the first
concession to begin operations. It is reasonable that it has the most wells established. On the
other hand, given the coarse spatial resolution of 30 km and 20 km for the 2000 Landsat ETM+
and 2005 CBERS CC2 satellite images respectively, and the similar spectral profiles between dry
savanna vegetation and vegetation disturbed by E&P, it’s very possible that the number of wells
identified in the satellite images is inaccurate for all four concessions. This in turn would
complicate comparisons and therefore it is recommendable to use imagery with higher spatial
resolution (Tang et al., 2007).
While not a key LUCC measure in this paper, the relationship with river density was also
strong. Petrozuata had the highest river density (number of linear river kilometers per
concession) at 1.14, followed closely by Sincor at 1.09 (though Sincor had twice the number of
river crossings). Because the sites selected for Petrozuata and Sincor are crossed by many rivers
and tributaries, the potential exists to create numerous hydrological disturbances. The Local
chose Petrozuata to begin HOB operations because it believed this site held the best potential for
drilling success (oil company representative 2005b), though newer technology such as horizontal
and directional drilling can reduce landscape disturbances (Gipson et al., 2002). Table 4-6 shows
important petro and nonpetroscape measures used in this paper.
Local participation was very weakly correlated with edge-effect petroscape (0.333,
p=0.602), despite the strong relationship to petroscape. As with core-areas this lack of strong
correlation contradicts the relationship that petroscape density has on edge-effect zones and
subsequent core areas. The lack of negative correlation to nonpetroscape density values is also
surprising. Though Petrozuata had a low nonpetroscape, Cerro Negro, which has the second
111
highest local-company participation had the highest nonpetroscape levels of all concessions (2.11
for 2000 and 2.22 for 2005). Presumably this would influence results, though it didn’t.
Another unexpected finding was the lack of correlation to number of rivers crossed since
Petrozuata, at 25, had the second highest number (while Sincor had 52, Ameriven 1, and Cerro
Negro 0 in 2005).
Though Local participation was correlated with petroscape density and well density,
correlations to other important measures such as negative NDVI, core-area petroscape, edge-
effect petroscape and number of rivers crossed were weak or close to zero. Therefore I fail to
reject the null hypothesis that Local-company participation is independent of the important
changes in land-cover associated with petroscape expansion.
Conclusion
After examining the correlations between the three different consortia of oil companies
(Local, European and US) operating in the HOB and important LUCC measures for 2000 and
2005 I fail to reject the three hypotheses of statistical independence for all of them.
Only one, the Local company, with its largest interests in Petrozuata, exhibited statistically
significant associations with petroscape density and well density. These two measures represent
the amount of oil roads, wells and facilities digitized from satellite images at two time periods
(2000 and 20005). However, these two correlations in themselves were not enough to reject the
null hypothesis of statistical independence since additional significant correlations would have
been required given the small sample size. Higher resolution imagery would increase accuracy in
petroscape feature detection such as wells and wellpads, allowing for more accurate comparisons
across concessions.
Concerning Petrozuata, whether the relationship between Local participation and the
petroscape is attributed to a lack of government oversight requires more research and data. The
112
Local company is now a majority partner in the current HOB concessions and will legally remain
so during the establishment of 27 new concessions.
Finally, using the lowest and highest linear and density petroscape values can provide a
scenario of what future changes may look like as the HOB operations expand. Using 2005
petroscape values for Petrozuata and Ameriven reveals the following figures. If the 27 new
planned concessions are established over the next 12 years, each with an average concession size
of 671 km²-- which is larger than Ameriven at 665 km² (PDVSA, 2008c), then the potential
petroscape density per concession could range from 0.18 to 0.50, a 36% difference. Regionally,
this translates into linear petroscape features of 3,230 km vs. 9,145 km, a sizeable difference,
which combined with regional development plans for the area above the Orinoco could
transform this entire region within a few years.
113
Figure 4-1. Venezuela’s heavy oil belt and the four current operations. From west to east they are: Sincor, Petrozuata, Ameriven and Cerro Negro. The Jose upgrading complex, located on the coast, processes the HOB crude to produce syncrude, or synthetic crude.
114
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
Company
Perc
enta
ge o
f ow
ners
hip
2005 3.75 4.17 7.50 10.42 11.75 22.53 39.892007 2.43 4.17 7.50 0.00 7.58 0.00 78.34
Statoil BP Chevron Exxon Total Conoco PDVSA
Figure 4-2. Oil company participation in the HOB in 2005 and under the new changes implemented in mid-2007.
115
Figure 4-3. Potential E&P area within the heavy oil belt by 2030. New concessions will occupy about 18,000 km², or 33% of the HOB land area (Petroguía, 2006-2007).
Table 4-1. Venezuelan heavy oil belt strategic associations (concessions), company participation by percentage and upgraded syncrude quality.
Concession Name Company Ownership Syncrude Production and Upgraded Quality
Sincor PDVSA 38% Total (France) 47% Statoil (Norway) 15%
180,000 b/d From 8.5º to 32º API
Petrozuata PDVSA 49.9% ConocoPhillips 50.1%
104,000 b/d From 9º to 19-25º API
Ameriven PDVSA 30% Chevron 30% ConocoPhillips 40%
190,000 b/d From 7-10º to 26º API
Cerro Negro PDVSA 41.67% ExxonMobil 41.67% BP(United Kingdom) 16.66%
105,000 b/d From 6º-10º to 16º API
116
Table 4-2. The R² coefficients for seven petroscape-related land-cover change measures and the three consortia of companies operating in the HOB. Some of the repeated categories represent values for the years 2000 and 2005.
PDVSA Participation
European Participation US Participation
Negative NDVI (1990-2000) km²
0.765 0.147 0.388 Positive NDVI (1990-2000) km²
0.143 0.77 0.952 Petroscape Density 2000 0.758 0 0.069 Petroscape Density 2005 0.73 0.003 0.034 Core Area Petroscape 2000
0.905 0.017 0.156 Core Area Petroscape 2005
0.923 0.021 0.171 Edge Effect Petroscape 2000
0.428 0.35 0.595 Edge Effect Petroscape 2005
0.122 0.377 0.502 Well Density 2005 0.895 0.084 0.001 Number of Rivers Crossed 2005 0.054 0.613 0.712
117
Table 4-3. The R² coefficients of determination between three types of company participation and 16 LUCC measures in the HOB.
PDVSA Participation
European Participation US Participation
Negative NDVI (1990-2000) km² 0.765 0.147 0.388 Positive NDVI (1990-2000) km² 0.143 0.77 0.952 Petroscape Density 2000 0.758 0 0.069 Petroscape Density 2005 0.73 0.003 0.034 Nonpetroscape Density 2000
0.02 0.096 0.121 Nonpetroscape Density 2005
0.011 0.094 0.111 Allscape Density 2000 0.09 0.14 0.083 Allscape Density 2005 0.144 0.159 0.084 Concession Area Km2 0.83 0.016 0.017 Core Area Petroscape 2000
0.905 0.017 0.156 Core Area Petroscape 2005
0.923 0.021 0.171 Core Area Nonpetroscape 2000 0.011 0.662 0.703 Core Area Nonpetroscape 2005 0.018 0.657 0.711 Edge Effect Petroscape 2000
0.428 0.35 0.595 Edge Effect Petroscape 2005
0.122 0.377 0.502 Edge Effect Nonpetroscape 2000 0.37 0.309 0.523 Edge Effect Nonpetroscape 2005 0.352 0.328 0.541 Number of Wells 2005 0.978 0.005 0.043 Well Density 2005 0.895 0.084 0.001 Wells Edge-effect 2005 Km²
0.496 0.004 0.066 Wells Core Area 2005 Km²
0.91 0.013 0.024
118
Table 4-3. Continued. Number of Rivers 2000 0.23 0.317 0.482 Number of Rivers 2005 0.054 0.613 0.712 River Density 0.251 0.142 0.265
Table 4-4. Kendall’s tau-b correlation coefficients showing the relationship between Local, European and US dominated concessions and seven important LUCC measures related to petroscape expansion. Sig refers to significance level while ** means correlations are significant at the 0.01 level (2-tailed).
PDVSA Participation
European Participation
US Participation
Correlation Coefficient -1.000 0.000 0.333 Negative NDVI
(1990-2000) km² N = 3 Sig. (2-tailed) 0.117 1.000 0.602
Correlation Coefficient 0.333 0.816 -1.000 Positive NDVI
(1990-2000) km² N = 3 Sig. (2-tailed) 0.602 0.221 0.117
Correlation Coefficient 1.000(**) 0.000 -0.333 Petroscape Density
2000 and 2005 Sig. (2-tailed) . 1.000 0.602 Correlation Coefficient -1.000 0.000 0.333 Core Area Petroscape
Km2 2000 and 2005 Sig. (2-tailed) 0.117 1.000 0.602 Correlation Coefficient 0.333 0.816 -1.000 Edge Effect Petroscape
Km2 2000 and 2005 Sig. (2-tailed) 0.602 0.221 0.117
Well Density 2005 Correlation Coefficient 1.000(**) 0.000 -0.333
Sig. (2-tailed) . 1.000 0.602 Number of Rivers Crossed 2000 and 2005
Correlation Coefficient 0.333 0.816 -1.000
Sig. (2-tailed) 0.602 0.221 0.117
119
Table 4-5. Kendall’s tau-b correlation coefficients and R² coefficients showing the relationship between concessions that were PDVSA-dominated, European-dominated, and US-dominated and important landscape-change measures. Sig refers to significance level while ** means correlations are significant at the 0.01 level (2-tailed).
PDVSA Participation
European Participation
US Participation
Correlation Coefficient 1.000 0.000 -0.333 PDVSA Participation
Sig. (2-tailed) . 1.000 0.602 R² 0.015 0.026
Correlation Coefficient 0.000 1.000 -0.816 European Participation
Sig. (2-tailed) 1.000 . 0.221 R² 0.015 0.923
Correlation Coefficient -0.333 -0.816 1.000 US Participation
Sig. (2-tailed) 0.602 0.221 . R² 0.026 0.923
Correlation Coefficient -1.000 0.000 0.333 Negative NDVI
(1990-2000) km² N = 3 Sig. (2-tailed) 0.117 1.000 0.602 R² .765 .147 .388
Correlation Coefficient 0.333 0.816 -1.000 Positive NDVI
(1990-2000) km² N = 3 Sig. (2-tailed) 0.602 0.221 0.117 R² 0.143 0.770 0.952
Correlation Coefficient 1.000(**) 0.000 -0.333 Petroscape Density
2000 and 2005 Sig. (2-tailed) . 1.000 0.602
2000 R² 0.758 0.000 0.069 2005 R² 0.730 0.003 0.034
Correlation Coefficient -0.333 -0.816 1.000(**) Nonpetroscape Density
2000 and 2005 Sig. (2-tailed) 0.602 0.221
2000 R² 0.02 0.096 0.121 2005 R² 0.011 0.094 0.111
Correlation Coefficient 0.333 -0.816 0.333 Allscape Density 2000
and 2005 Sig. (2-tailed) 0.602 0.221 0.602
2000 R² 0.09 0.140 0.083 2005 R² 0.144 0.159 0.084
120
Table 4-5. Continued. Correlation Coefficient -1.000 0.000 0.333 Concession Area km²
Sig. (2-tailed) 0.117 1.000 0.602 R² 0.830 0.016 0.017
Correlation Coefficient -1.000 0.000 0.333 Core Area Petroscape
Km2 2000 and 2005 Sig. (2-tailed) 0.117 1.000 0.602
2000 R² 0.905 0.017 0.156 2005 R² 0.923 0.021 0.171
Correlation Coefficient -0.333 -0.816 1.000(**) Core Area
Nonpetroscape Km2 2000 and 2005 Sig. (2-tailed) 0.602 0.221 2000 R² 0.011 0.662 0.703 2005 R² 0.018 0.657 0.711
Correlation Coefficient 0.333 0.816 -1.000 Edge Effect Petroscape
Km2 2000 and 2005 Sig. (2-tailed) 0.602 0.221 0.117
2000 R² 0.428 0.350 0.595 2005 R² 0.122 0.377 0.502
Correlation Coefficient -0.333 -0.816 1.000(**) Edge Effect
Nonpetroscape Km2 2000 and 2005 Sig. (2-tailed) 0.602 0.221 2000 R² 0.370 0.309 0.523 2005 R² 0.352 0.328 0.541
Correlation Coefficient 1.000(**) 0.000 -0.333 Number of Wells 2005
Sig. (2-tailed) . 1.000 0.602 R² 0.978 0.005 0.043 Well Density 2005 Correlation
Coefficient 1.000(**) 0.000 -0.333
Sig. (2-tailed) . 1.000 0.602 R² 0.895 0.084 0.001 Wells Edge-effect 2005 Km²
Correlation Coefficient 1.000(**) 0.000 -0.333
Sig. (2-tailed) . 1.000 0.602 R² 0.496 0.004 0.066 Wells Core Area 2005 Km²
Correlation Coefficient -1.000 0.000 0.333
Sig. (2-tailed) 0.117 1.000 0.602 R² 0.910 0.013 0.024 Number of Rivers Crossed 2000 and 2005
Correlation Coefficient 0.333 0.816 -1.000
Sig. (2-tailed) 0.602 0.221 0.117
121
Table 4-5. Continued. 2000 R² 0.23 0.317 0.482 2005 R² 0.054 0.613 0.712 River Density Correlation
Coefficient 1.000(**) 0.000 -0.333
Sig. (2-tailed) . 1.000 0.602 R² 0.251 0.142 0.265
122
Table 4-6. Petro and nonpetroscape data values for the four HOB concessions. Concession Name Sincor Petrozuata Ameriven Cerro Negro
Concession Area (area km²) 517.38 300.77 665.44 303.15 PDVSA Participation (%) 38.00 49.90 30.00 41.67
European Participation (%) 62.00 0.00 0.00 16.66 US Participation (%) 0.00 50.10 70.00 41.67
Negative NDVI 1990-2000 (km²) 77.55 39.00 312.55 0.00
Positive NDVI 1990-2000 (km²) 41.36 21.00 2.19 0.00
Petroscape Linear 2000 (km) 243.07 276.79 19.85 61.51
Nonpetroscape Linear 2000 (km) 0.00 39.19 523.69 578.19
Petroscape Density 2000 (km/km²) 0.47 0.92 0.03 0.20
Nonpetroscape Density 2000 km/km²) 0.00 0.13 0.79 1.91
Allscape Density 2000 (area km²) 0.47 1.05 0.82 2.11
Edge-effect Petroscape 2000 (km²) 217.87 193.60 26.04 96.58
Edge-effect Nonpetroscape 2000 (km²) 0.00 37.41 351.33 274.00
Core-area Petroscape 2000 (km²) 299.50 107.17 639.40 206.57
Core-area Nonpetroscape 2000 (km²) 0.00 263.36 314.11 29.15
Rivers Crossed 2000 (number) 22.00 18.00 0.00 0.00
Petroscape Linear 2005 (km) 302.33 338.70 119.63 90.56
Nonpetroscape Linear 2005 (km) 0.00 47.39 485.71 580.97
Petroscape Density 2005 (km/km²) 0.58 1.13 0.18 0.30
Nonpetroscape Density 2005 (km/km²) 0.00 0.16 0.73 1.91
Allscape Density 2005 (km/km²) 0.59 1.28 0.91 2.22
Edge-effect Petroscape 2005 (km²) 271.57 222.49 117.07 96.27
Edge-effect Nonpetroscape 2005 (km²) 0.00 45.61 348.32 274.00
123
Table 4-6. Continued.
Edge-effect Allscape 2005 (km²) 271.60 268.00 468.40 370.27
Core-area Petroscape 2005 (km²) 245.80 78.29 548.37 206.88
Core-area Nonpetroscape 2005 (km²) 0.00 255.17 317.11 29.15
Rivers Crossed 2005 (number) 52.00 25.00 1.00 0.00
Wells and Wellpads 2005 (number) 142.00 195.00 98.00 168.00
Well Density 2005 (wells/concession) 0.27 0.65 0.15 0.55
Well Edge-effect 2005 (km²) 111.15 134.25 86.05 81.46
Wells Core-areas 2005 (km²) 406.22 166.52 579.38 221.68 Rivers Length (km) 561.40 342.74 185.01 26.88
Riverscape Density (rivers/concession) 1.09 1.14 0.28 0.09
124
CHAPTER 5 LAND-COVER CHANGE AND NONRENEWABLE NATURAL RESOURCES
Summary
Land-use land-cover change involves the complex interaction of social and ecological
variables across various spatial and temporal scales. This work has explored human-environment
interactions in the context of natural resource extraction, since the use of these resources for local
needs or to meet external market demands creates different and important types of landscape
alterations. These perturbations can create short-lived or nonpermanent changes, known as
modifications, or they can lead to new land-cover classes known as conversion (Veldkamp and
Lambin, 2001; Lambin et al., 2003; Southworth, 2004; Daniels et al., 2008). Both of these results
imply a loss of land cover. However, not all natural resource use leads to land-cover loss. Forest
plots, logging sites and grazing pastures may revert back to secondary forest once these activities
cease (Kupfer et al., 2004, 510; Fearnside et al., 2007, 678). This transition may be fostered by
land tenure; policy changes and implementation; environmental feedbacks and climate change;
different market demands and varying household makeup. The planting of forests also leads to
land-cover gain, though tree plantations do not provide the same ecosystem services as do
natural forests (Parrotta and Knowles, 2001, 220).
Nonrenewable natural resource extraction, on the other hand, tends to create permanent
land-cover changes because of the nature of mining and drilling, the establishment of permanent
infrastructure needed for operations, as well as the waste sites required for the large volumes of
saline water produced (Kharaka and Dorsey, 2005, 61). Furthermore, once viable extractivist
activities take place, they tend to attract new settlers to the areas of direct interaction. These
settlers create additional LUCCs through the establishment of homes and agricultural plots; new
roads that create access to area resources, thus expanding LUCC; and the growth of service
125
industries that can eventually convert a small settlement into a town or city (Schmink and Wood,
1992, Consiglio et al., 2006).
Whereas agricultural activity in the tropics converts a great deal of forested land every year
for the establishment of soybean farms and cattle ranches or small-scale agricultural plots (FAO,
2006, Lambin et al., 2003; Wassenaar et al., 2007), logging and fires also contribute to
significant land conversion (Nepstad et al., 2001; Siegert and Hoffmann, 2000; Hammond et al.,
2007). Gold mining, on the other hand, tends to be pursued at the small-scale level (individuals,
or groups of five) (Peterson and Heemskerk, 2001) and thus creates smaller pockets change
though cumulatively they can create widespread degradation (Hilson, 2002). While large-scale
gold mining involves digging large open pits and establishing large tailing ponds that hold
contaminated waste material, small-scale mining can involve spray-washing land and riverine
deposits with motorized equipment that removes the vegetation, roots, soil and seeds, thus
degrading the landscape considerably and delaying regeneration for up to 20 years or more
(Almeida-Filho and Shimabukuro, 2002; Peterson and Heemskerk, 2001). On top of this,
mercury (and cyanide—in the large operations) is used to help recover gold from the extracted
soil and rocks. Thereafter it is released into the water, where it can convert to toxic
methylmercury. It can also contaminate miners and (gold) business owners through skin
contamination and inhalation of fumes when the mercury is burned away from the gold both in
the field and in gold processing shops (Veiga and Hinton, 2002, 19; Bastos et al., 2004; Mol et
al., 2001). The simple technology needed and the potential for monetary earnings, as compared
to other options in many developing countries, leads to thousands implementing these practices
in a single site (Hilson, 2002). Worldwide around six million miners operate in more than 40
126
countries (Veiga et al., 2006) and it’s a central driver of deforestation in Surimane (Heemskerk,
2001).
Oil exploration and production can be the most lucrative of all these activities, and given
the necessary knowledge, technology and capital, this activity is the purview of state
governments and multinational oil companies (MNOCs). Different management practices during
the exploration and production phase can create greater or lesser LUCCs which can result in
modifications or conversions lasting 35 years or more. State-owned oil companies, who often
don’t compete abroad, may not have the culture and history of including environmental
performance as part of their operations, while for MNOCs this is increasingly an important
aspect of day-to-day business (Diamond, 2005; Moser 2001; Consiglio et al., 2006).
Negative press coverage, pressure from employees and customers, complaints from area
residents affected by operations; the threat of legal action; the desire to be perceived as attractive
to stock market investors and customers; and the competition for dwindling opportunities can
help explain the importance of sustainability to MNOCs (Lawrence, 2007; Orlitzky et al., 2003;
Morhardt, 2002). Even so, the “cumulative impacts from these operations are high, because a
total of about 3.5 million oil and gas wells have been drilled to date in the United States, but
currently, only about 900,000 are in production (Kharaka and Dorsey, 2005, pp. 61-62).
Worldwide we would expect to see similar ratios.
In Venezuela, plans to greatly expand exploration and production of the vast deposits in
the heavy oil belt will potentially affect 18,000 km² of dry tropical savanna/forest. It is important
the government and the oil companies involved create baseline studies of the concession areas
prior to exploration. The innovative methods presented in this study integrate the use of GIS and
remote sensing techniques to measure landscape disturbances related to petroleum exploration
127
and production. This approach is particularly useful for monitoring and measuring current
landscape patterns and subsequent changes (Janks and Prelat, 1994; Griffith et al., 2002a),
though in personal conversations with industry representatives it appears that this methodology is
not widely used for these purposes. It is also important that environmental impact assessment
(EIA) studies be implemented throughout the life of the project, not just as a precondition to
begin drilling (Sebastiani et al., 2001). Finally, the implementation of best business practices,
such as minimizing the size of seismic lines and the equipment used to create them; closing and
recuperating routes that are no longer needed; implement long horizontal drilling; partner with
area industries to share roads and infrastructure; reduce the number of river crossings; and be
transparent in environmental reporting are important ways for oil companies to reduce
environmental alterations and become more sustainable (Musinsky et al., 1998; Schneider et al.,
2003; E&P Forum/UNEP, 1997; Morhardt, 2002; Gipson et al., 2002; Stein et al., 2002;
Musinsky et al., 1998).
One commonality among all four types of resources is accessibility. In order to access
these resources routes and trails must be built. Thus, roads are the catalyst that opens up new
landscapes for local and migrant land managers (Perz et al., 2007; Musinsky et al., 1998; Forman
et al., 2002; E&P Forum/UNEP, 1997; Consiglio et al., 2003; Soares-Filho et al., 2004). Once
the roads are created to access a particular resource (trees, gold, oil), then settlers move in and
establish homes, agricultural plots and cattle ranches; previously inaccessible forests are now
open to loggers and miners; and LUCC speeds up.
Future Work
As Morton et al. (2004, pp. 8) observe, “there are few studies that examine the exact size
and extent of the ecological footprint of energy development. Spatial analysis can help fill this
information gap.” Concerning petroleum E&P in the heavy oil belt, using additional satellite
128
images captured at different times of the year, such as the wet season, can help distinguish
landscape disturbances from dry tropical savanna. Field surveys of oil roads will permit seismic
and other roads to be more easily identified, which in turn can direct reclamation efforts and the
closing of nonessential or redundant roads (Wilderness Society, 2006; Musinsky et al., 1998; Lee
and Boutin, 2006). This will likely reduce habitat fragmentation, as well as the building of new
roads which increases access to land and resources and can lead to the building of settlements.
Other factors to consider include identifying rivers or tributaries that do not need to be crossed
and therefore avoid interruptions to flow patterns and their connectivity to floodplains (Stein et
al., 2002, 1); distinguishing between main-channel and floodplain habitats, and the “presence of
migratory barriers (e.g., waterfalls) that create predator-free environments” (Wantzen et al.,
2006, pp. 63).
Additional future work should consider other remote sensing techniques to identify (urban)
infrastructure features such as fuzzy-spectral mixture analysis (Tang et al., 2007), the normalized
difference built-up index (NDBI) used to monitor growth and distribution of urban built-up
areas, as well as built-up area calculation which involves subtracting the NDVI from the NDBI
(Jensen 2005). Almeida-Fihlo and Shimabukuro (2000) used radar images (JERS-1 SAR) to
detect areas disturbed by gold mining in northern Brazil. This method might prove useful for
studying petroscapes, particularly since many are located in tropical countries where dense cloud
cover on satellite imagery is often a problem.
Wildlife studies, for which several exist in Canada’s heavy oil sands, could help determine
more precise avoidance zones toward petroscape features by of local fauna such as the caiman,
tapir, Capuchin monkeys, capybara, bats, deer, manatee, reptiles such as the green anaconda,
parrots, amphibians and many fish (Pérez, 2000; Gorzula, 1978; Marín et al., 2007; López-
129
Fernández and Winemiller, 2000; Veneklaas et al., 2005; Rosales et al., 1999; Padilla and
Dowler, 1994) This is turn could be considered when establishing new infrastructures,
particularly in areas containing endangered species.
Therefore, adapting spatial technologies to the environmental impact assessment (EIA)
plan is a useful and cost-effective way of monitoring E&P-driven changes throughout the life-
cycle of oil operations, as well as aiding in policy formulation and decision making (Sebastiani et
al., 2001; Satapathy et al., 2008; Griffith et al., 2002a). While geospatial technologies may be
used in finding oil, drilling and production scenarios (Yatabe and Fabbri, 1986; Janks and Prelat,
1994, 1995; Almeida-Filho, 2002; Almeida-Filho et al., 2002; Zhang et al., 2007) these
technologies do not appear to be widely used to monitor “Above Ground Review” of oil-related
landscape changes (Consiglio et al., 2006, pp. 2). Some exceptions include Chust and
Sagarminaga (2007) who used data gathered from the Multi-angle Imaging SpectroRadiometer
(MISR) to detect oil spills in Venezuela’s Lake Maracaibo; Kwarteng (1998, 1999) who detected
contaminated vegetation surrounding oil lakes in Kuwait using data from Landsat TM, Indian
Remote Sensing Satellite (IRS) 1-D, and Spot satellite; and Ud-Din et al. (2008) who used
Landsat thermal data, (band 6) to map hydrocarbon polluted sites in Kuwait. These efforts show
a small but growing effort to use geospatial technologies to monitor petroleum related spills.
However, they do not focus on the establishment and extension of petroscape among regional
sites as a way of reducing landscape alterations.
On the other hand, work by Musinsky et al. (1998) correlated deforestation rates in a
Guatemalan national park with the development of an oil road. Thomson et al. (2005), Morton et
al. (2002); Morton et al. (2004); and the Wilderness Society (2006) embrace the use of GIS and
remote sensing techniques to monitor how E&P operations affect LUCC and use this information
130
to propose best business practices. Meanwhile, Schneider and Dyer (2006) do the same in
Canada’s Alberta oil sands. My project marks a first step to perform similar work in Venezuela,
especially in the important heavy oil belt region where alterations to regional ecosystems
stemming from heavy oil E&P will have important human-environment implications at the local
and regional scale, which in turn contributes to global environmental change.
131
LIST OF REFERENCES
Abdi, H., 2007. The Kendall rank correlation coefficient. In: Salkind, N. (Ed.), Encyclopedia of Measurement and Statistics. Sage Publications, Thousand Oaks, CA. Available at <http://www.utdallas.edu/~herve/Abdi-KendallCorrelation2007-pretty.pdf>.
Abizaid, C., Coomes, O. T., 2004. Land use and forest fallowing dynamics in seasonally dry tropical forests of the southern Yucatán Peninsula, Mexico. Land Use Policy 21, 71-84.
Almeida-Filho, R., 2002. Remote detection of hydrocarbon microseepage-induced soil alteration. International Journal of Remote Sensing 23 (18), 3523-3524.
Almeida-Filho, R., Miranda, F. P., Galvão, L. S., Freitas, C. C., 2002. Terrain characteristics of a tonal anomaly remotely detected in an area of hydrocarbon microseepage, Tucano Basin, north-eastern Brazil. International Journal of Remote Sensing 23 (18), 3893-3898.
Almeida-Filho, R., Shimabukuro, Y., 2004. Monitoring biomass burning in the Brazilian Amazon. International Journal of Remote Sensing 25 (24), 5537-5542.
Almeida-Filho, R., Shimabukuro, Y., 2002. Digital image processing of a Landsat TM time series for mapping and monitoring degraded areas caused by independent gold miners Roraima State, Brazilian Amazon. Remote Sensing of the Environment 79, (1), 42-50.
Almeida-Filho, R., Shimabukuro, Y., 2000. Detecting areas disturbed by gold mining activities through JERS-1 SAR images, Roraima State, Brazilian Amazon. International Journal of Remote Sensing 21 (17), 3357-3362.
Anderson, L. E. 1994. The Political Ecology of the Modern Peasant: Calculation and Community. Johns Hopkins University Press, Baltimore.
Ankersen, T., Barnes, G., 2004. Inside the polygon: emerging community tenure systems and forest resource extraction. In: Zarin, D. J., Alavalpati, R. J. J., Putz, F. E., Schmink, M. (Eds.), Working Forests in the Neotropics: Conservation Through Sustainable Management? Columbia University Press, New York, pp. 156-177.
Armenteras, D., Rudas, G., Rodriguez, N., Sua, S., Romero, M., 2006. Patterns and causes of deforestation in the Colombian Amazon. Ecological Indicators 6, 353-368.
Ascencio-Cendejas, F., Reyes-Venegas, O., 2006. Conceptual model design studies for a self-sustained and complete productive utilization of high-viscosity/extra heavy-oil onshore reservoir in Mexico. Society of Petroleum Engineers 104019-MS, 1-5..
Asner, G. P., Keller, M., Silva, J. N. M., 2004. Spatial and temporal dynamics of forsest canopy gaps following selective logging in the eastern Amazon. Global Change Biology 10, 765-783.
Asner, G. P., Knapp, D. E., Broadbent, E. N., Oliveira, P. J. C., Keller, M., Silva, J. N., 2005. Selective logging in the Brazilian Amazon. Science 310, 480-482.
132
Balke, S. C., Rosauer, M. S., 2002. Real time, real fast: drilling horizontal wells in a heavy oil environment. Society of Petroleum Engineers 78954-MS, 1-8.
Barnola, L. F., Cedeño, A., 2000. Inter-population differences in the essential oils of Pinus caribaea needles. Biochemical Systematics and Ecology 28, 923-931.
Bastos, W. R., Fonseca, M. de F., Pinto, F. N., Rebelo, M. de F., dos Santos, S. S., da Silveira, E. G., Torres, J. P. M., Malm, O., Pfeiffer, W. C., 2004. Mercury persitence in indoor environments in the Amazon region, Brazil. Environmental Research 96, 235-238.
Baynard, C., 2007. Mapping and Monitoring the Petroscape in Venezuela's Heavy oil belt with Landsat and CBERS satellite images. Society of Petroleum Engineers 106803, 1-14.
Berrueta, V. M., Edwards, R. D., Masera, O. R., 2008. Energy performance of wood-burning cookstoves in Michoacan, Mexico. Renewable Energy 33, 859-870.
Bevilacqua, M., Cárdenas, L., Flores, A. L., Hernández, L., Lares, E., Mansutti, A., Miranda, M., Ochoa, J., Rodríguez, M., Selig, E., 2002. The State of Venezuela’s Forests: A Case Study of the Guayana Region. Global Forest Watch, Caracas, Venezuela.
Boletta, P. E., Ravelo, A. C., Planchuelo, A. M., Grilli, M., 2006. Assessing deforestation in the Argentine Chaco. Forest Ecology and Management 228 (1-3), 108-114.
Bosco, M. L., Varrica, D., Dongarrà, G., 2005. Case study: inorganic pollutants associated with particulate matter fron an area near a petrochemical plant. Environmental Research 99, 18-30.
Bose-O’Reilly, S., Lettmeier, B., Roider, G., Siebert, U., Drasch, G., 2008. Mercury in breast milk—a health hazard for infants in gold mining areas? International Journal of Hygiene and Environmental Health In press. 1-9.
Boza, G. C. S., Romero, T. H., 2001. Orinoco oil belt well construction using ‘wells-in-series’ technology with horizontal and multilateral trajectory. Society of Petroleum Engineers 69710-MS, 1-8.
BP (British Petroleum)., 2007. Oil production. Statistical Review of World Energy 2007. Available at <http://www.bp.com/productlanding.do?categoryId=6848&contentId=7033471>.
Bradshaw, C. J. A., Boutin, S., Hebert, D. M., 1998. Energetic implications of disturbance caused by petroleum exploration to woodland caribou. Canadian Journal of Zoology 76, 1319-1324.
Briceno, M., Summers, L. E., Quijada, O. J., 2002. Horizontal wells pressure build up analysis in Orinoco heavy oil belt. Society of Petroleum Engineers 78969-MS. 1-13.
Bruni, T., 2007. Why our approach to sustainability is unsustainable. Journal of Petroleum Technology 59 (10), 156.
133
Bryant, R. L., 2001. Political ecology: a critical agenda for change? In: Castree, N. (Ed.), Social Nature: Theory, Practice, and Politics. Blackwell Publishers, Oxford, UK, pp. 151-169.
Bryant, R. L., Bailey, S., 1997. Third World Political Ecology. Routledge, London.
Burt, J. E., Barber, G. M., 1996. Elementary Statistics for Geographers. Guilford Press, New York.
Caldas, M., Walker, R., Arima, E., Perz, S., Aldrich, S., Simmons, C., 2007. Theorizing land cover and land use change: the peasant economy of Amazonian deforestation. Annals of the Association of American Geographers 97 (1), 86-110.
Carrera, E., 2005. Chief, Social and Environmental Unit, Masisa, personal communication.
Castree, N., 2001. Socializing nature: theory, practice, and politics. In: Castree, N. (Ed.), Social Nature: Theory, Practice, and Politics. Blackwell Publishers, Oxford, UK, pp. 1-21.
Castro, S. H., Sánchez, M., 2003. Environmental viewpoint on small-scale copper, gold and silver mining in Chile. Journal of Cleaner Production 11, 207-213.
Chacón-Moreno, E. J., 2007. Ecological and spatial modeling: mapping ecosystems, landscape changes, and plant species distribution in Llanos del Orinoco, Venezuela. ITC Dissertation 139, Wageningen University. Available at <http://www.itc.nl/library/papers_2007/phd/chaconmoreno.pdf >.
Chacón-Moreno, E. J., 2004. Mapping savanna ecosystems of the llanos del Orinoco using multitemporal NOAA satellite imagery. International Journal of Applied Observation and Geoinformation 5, 41-53.
Chust, G., Sagarminaga, Y., 2007. The multi-angel view of MISR detects oil slicks under sun glitter conditions. Remote Sensing of Environment 107, 232-239.
CIA., 2008. Venezuela. The World Factbook. 6 Mar. Available at <https://www.cia.gov/library/publications/the-world-factbook/print/ve.html>
ConocoPhillips., 2007. Exploration and Production: South America, Venezuela. Available at <http://www.conocophillips.com/NR/rdonlyres/A4915FFA-625E-4D73-A227-B3AA0049A76A/0/fb_southamerica.pdf>.
Consiglio, M., Witchalls, B., Armstrong, K., Stampa, M., Madec, A., Abdullah, M., Depraz, S., Fraser, E., 2006. A guide to social impact assessment in the oil and gas industry. Society of Petroleum Engineers 98495-MS, 1-6.
Corfield, G. S., 1948. Recent activities in Venezuela’s petroleum industry. Economic Geography 24 (2), 114-118.
Cutter, S. L., Renwick, W. H., 2004. Exploitation, Conservation, Preservation: A Geographic Perspective on Natural Resource Use. 4th ed. John Wiley & Sons, Hoboken, NJ.
134
Cumming, G. S., Barnes, G., 2007. Characterizing land tenure dynamics by comparing spatial and temporal variation at multiple scales. Landscape and Urban Planning 83, 219-227.
Da Costa, G. M., dos Anjos, L. M., Souza, G. S., Gomes, B. D., Saito, C. A., Pinheiro, M. da C. N., Ventura, D. F., Filho, M. da Silva., Silveira, L. C. L., 2008. Mercury toxicity in Amazon gold miners: visual dysfunction assessed by retinal and cortical electrophysiology. Environmental Research In press, 1-10.
Dalle, S. P., de Blois, S., Caballero, J., Johns, T., 2006. Integrating analyses of land-use regulations, cultural perceptions and land-use/land-cover data for assessing the success of community-based conservation. Forest Ecology and Management 222, 370-383.
Daniels, A., Painter, K., Southworth, J., 2008. Milpa imprint on the tropical dry forest landscape in Yucatan, Mexico: remote sensing and field measurement of edge vegetation. Agriculture Ecosystems and Environment 123, 293-304.
Dashwood, H. S., 2007. Canadian mining companies and corporate social responsibility: weighing the impact of global norms. Canadian Journal of Political Science 40 (1), 129-156.
De Figuerôa, J. M., Pareyn, F. G. C., Araújo, E. de L., da Silva, C. E., dos Santos, V. F., Cutler, D. F., Baracat, A., Gasson, P., 2006. Effects of cutting regimes in the dry and wet season on survival and sprouting of woody species from the semi-arid caatinga of northeast Brazil. Forest Ecology and Management 229, 294-303.
Demirbas, A. H., Demirbas, I., 2007. Importance of rural bioenergy for developing countries. Energy Conversion and Management 48, 2386-2398.
Diamond, J., 2005. Collapse: How Societies Choose to Fail or Succeed. Viking, New York.
Diamond, J., 2003. The Last Americans: Environmental Collapse and the End of Civilization. Harper’s Magazine (June) 43.
Dietz, T., Ostrom, E., Stern, P. C., 2003. The struggle to govern the commons. Science 302, 1907-1912.
Donald, P. F., Sanderson, F. J., Burfield, I. J., van Bommel, F. P. J., 2006. Further evidence of continent-wide impacts of agricultural intensification on European farmland bird, 1990-2000. Agriculture Ecosystems and Environment 116, 189-196.
Donato, D. B., Nichols, O., Possingham, H., Moore, M., Ricci, P. F., Noller, B. N., 2007. A critical review of the effects of gold cyanide-bearing tailings solutions on wildlife. Environment International 33, 974-984.
Dow Jones and Company., 2007a. Dow Jones sustainability world index (DJSI World). Available at <http://www.sustainability-index.com/djsi_pdf/publications/Factsheets/SAM_IndexesMonthly_DJSIWorld.pdf>.
135
Dow Jones and Company., 2007b. . DJSI world 2007/2008. Available at <http://www.sustainability-index.com/djsi_pdf/djsi_world/Components/SAM_DJSIWorld_Components.pdf>.
Drake, P. L., Rojas, M., Reh, C. M., Mueller, C. A., Jenkins, F. M., 2001. Occupational exposure to airborne mercury during gold mining operations near El Callao, Venezuela. International archives of occupational and environmental health 74, 206-212.
Dumith, D., 2004. Instituto Geográfico de Venezuela Simón Bolívar, Caracas, Venezuela, personal communication.
Dunn, R. R., 2004. Managing the tropical landscape: a comparison of the effects of logging and forest conversion to agriculture on ants, birds, and Lepidoptera. Forest Ecology and Management 191, 215-224.
Dyer, S. J., O’Neill, J. P., Wasel, S. M., Boutin, S., 2001. Avoidance of industrial development by woodland caribou. The Journal of Wildlife Management 65 (3), 531-542.
E&P Forum/UNEP., 1997. Potential environmental impacts. In: Environmental management in oil and gas exploration and production. UNEP IE/PAC Technical Report 37 (= E&P Forum Report 2.72/254), 11-16. Available at <http://www.ogp.org.uk/pubs/254.pdf >.
Egan, D., 2007. Conserving and Restoring Old Growth in Frequent-fire Forests: Cycles of Disruption and Recovery. Ecology and Society 12 (2), 1-5. Available at <http://www.ecologyandsociety.org/vol12/iss2/art23/>.
EIA (Energy Information Administration)., 2008. World crude oil prices: total world 1978-2008. Available at <http://tonto.eia.doe.gov/dnav/pet/pet_pri_wco_k_w.htm>.
EIA (Energy Information Administration)., 2007a. November 2007 International Petroleum Monthly: World Crude Oil Production (Including Lease Condensate), Part D - Venezuela, Vietnam, Yemen, Other, World, OPEC-12, OPEC-11, Persian Gulf, and North Sea. Table 4.1d. Available at <http://www.eia.doe.gov/ipm/supply.html>.
EIA (Energy Information Administration)., 2007b. Country analysis briefs: Venezuela. US Department of Energy. Available at <http://www.eia.doe.gov/emeu/cabs/Venezuela/Full.html>.
Eickhout, B., van Meijl, H., Tabeau, A., van Rheenen, T., 2007. Economic and ecological consequences of four European land use scenarios. Land Use Policy 24, 562-575.
Ekici, A., Ekici, M., Kurtipek, E., Akin, A., Arslan, M., Kara, T., Apaydin, Z., Demir, S., 2005. Obstructive airway diseases in women exposed to biomass smoke. Environmental Research 99, 93-98.
136
Ellsworth, L., 2002. A place in the world: tenure security and community livelihoods, a literature review. Forest Trends. Available at <http://www.forest-trends.org/documents/publications/Place_World.pdf.pdf#search=%22A%20place%20in%20the%20world%3A%20tenure%20security%20and%20community%20livelihoods%2C%20a%20literature%20review%22>.
ESI., 2005. 2005 environmental sustainability index: benchmarking national environmental stewardship. Yale Center for Environmental Law and Policy, Yale University, and Center for International Earth Sciences Information Network, Columbia University. Available at <http://www.yale.edu/esi/ESI2005_Main_Report.pdf >.
Espírito-Santo, F. D. B.; Shimabukuro, Y.E. Kulpich, T. M., 2005. Mapping forest successional stages following deforestation in Brazilian Amazonia using multi-temporal Landsat images. International Journal of Remote Sensing 26 (3), 635-642.
ESRI., 2008. Geographic Network Services for ArcGIS 9.2. Redlands, CA.
Estevinho, B. N., Ribeiro, E., Alves, A., Santos, L., 2008. A preliminary feasibility study for pentachlorophenol column sorption by almond shell residues. Chemical Engineering Journal 136 (2-3), 188-194.
FAO., 2006. Global Forest Resources Assessment 2005: Progress towards sustainable forest management. (= FAO Forestry Paper 147). Rome, Italy. Available at <ftp://ftp.fao.org/docrep/fao/008/A0400E/A0400E00.pdf>.
FAO., 2005. Forest area and total area change. Food and Agriculture Organization of the United Nations. Available at <http://www.fao.org/forestry/site/32248/en>.
Fajardo, L., González, V., Nassar, J. M., Lacabana, P., Portillo, C. A., Carrasquel, F., 2005. Tropical dry forests of Venezuela: characterization and current conservation status. Biotropica 37 (4), 531-546.
Faminow, M.D., Vosti, S.A., 1998. Livestock-deforestation links: Policy issues in the western Brazilian Amazon. Food and Agriculture Organization of the United Nations. Rome, Italy. Available at <http://www.fao.org/wairdocs/lead/x6139e/x6139e00.htm>.
Fearnside, P. M., Barbosa, R. I., de Alencastro, P. M. L., 2007. Burning of secondary forest in Amazonia: biomass, burning efficiency and charcoal formation during land preparation for agriculture in Apiaú, Roraima, Brazil. Forest Ecology and Management 242, 678-687.
Ferguson, T., Genest, C., Hallin, M., 1999. Kendall’s tau for autocorrelation. Department of statistics papers. University of California, Los Angeles. Available at <http://repositories.cdlib.org/cgi/viewcontent.cgi?article=1183&context=uclastat>.
Ferraz, S. F. B., Capâo, L. M. S. A. C., Vettorazzi, C. A., 2006. Temporal scale and spatial resolution effects on Amazon forest fragmentation assessment in Rondônia. International Journal of Remote Sensing 27 (3), 459-472.
137
Fipke, S., Celli, A., 2008. The use of multilateral well designs for improved recovery in heavy oil reservoirs. Society of Petroleum Engineers 112638-MS. 1-11.
Flamenco-Sandoval, A., Ramos, M. M., and Masera, O. R., 2007. Assessing implications of land-use and land-cover dynamics for conservation of a highly diverse tropical rain forest. Biological Conservation 138, 131-145.
Folke, C., Pritchard, L. J., Berkes, F., Colding, J. Svedin, U., 2007. The problem of fit between ecosystems and institutions: ten years later. Ecology and Society 12 (1). Available at <http://www.ecologyandsociety.org/vol12/iss1/art30/>.
Forman, R. T. T., Sperling, D., Bissonnette, J. A., Clevenger, A. P., Cutshall, C. D., Dale, V. H., Fahrig, L., France, R., Goldman, C. R., Heanue, K., Jones, J. A., Swanson, F. J., Turrentine, T., Winter, T. C., 2002. Road Ecology: Science and Solutions. Island Press, Washington, DC.
Forman, R. T. T., Deblinger, R. D., 2000. The ecological road-effect zone of a Massachusetts (U.S.A.) suburban highway. Conservation Biology 14 (1), 36-46.
Forman, R. T. T., Alexander, L. E., 1998. Roads and their major ecological effects. Annual Review of Ecology and Systematics 29, 207-231+C2.
Fujita, S. T. Y., Fox, J., 2005. Resource use dynamics and land-cover change in Ang Nhai village and Phou Phanang National Reserve Forest, Lao PDR. Environmental Management 36 (3), 382-393.
Fung, T., Siu, W., 2000. Environmental quality and its changes, an analysis using NDVI. International Journal of Remote Sensing 21 (5), 1001-1024.
Funk, C. C., Brown, M. E., 2006. Intra-seasonal NDVI change projections in semi-arid Africa. Remote Sensing of Environment 101 (2), 249-256.
Gaceta Oficial de la República Bolivariana de Venezuela., 2007. Asamblea Nacional. Número 38.617. 1 Feb. Available at <http://www.tsj.gov.ve/gaceta/febrero/010207/010207-38617-01.html>.
Gaddy, D. E., 2003. Introduction to GIS for the Petroleum Industry. PennWell, Tulsa, OK.
Galarraga, F., Márquez, G., Reategui, K., Martínez, A., 2007. Comparative study of crude oils from the Machete area in the Eastern Venezuelan Basin by pyrolysis of asphaltenes. Journal of Analytical and Applied Pyrolysis 80, 289-296.
Garvin, T., McGee, T. K., Smoyer-Tomic, K. E., Aubynn, E. A., 2008. Community—company relations in gold mining in Ghana. Journal of Environmental Management In press, 1-16.
Gasse, F., Téhet, R., Durand, A., Gilbert, E., Fontes, J. C., 1990. The arid-humid transition in the Sahara and the Sahel during the last glaciation. Nature 346, 141-146.
138
Gibbons, P., Cunningham, R. B., Lindenmayer, D. B., 2008. What factors influence the collapse of trees retained on logged sites? A case-control study. Forest Ecology and Management 255, 62-67.
Gipson, L. J., Owen, R., Robertson, C. R., 2002. Hamaca heavy oil project: lessons learned and an evolving development strategy. Society of Petroleum Engineers 78990, 1-9.
Gladwin, T. N., 1977. Environment, Planning and the Multinational Corporation. JAI Press, Greenwich, Connecticut.
Gordon, J. E., Hawthorne, W. D., Reyes-García, A., Sandoval, G., Barrance, A. J., 2004. Assessing landscapes: a case study of tree and shrub diversity in the seasonally dry tropical forests of Oaxaca, Mexico and southern Honduras. Biological Conservation 117, 429-442.
Global 100., 2008. Most sustainable corporations in the world. Available at <http://www.global100.org/2008/index.asp>.
Global 100., 2007. Global 100 most sustainable corporations in the world. Available at <http://www.global100.org/2007/index.asp>.
Gonzalez, R., Reina, M., 1994. Producing extra heavy oil from the Orinoco Belt by electrical submersible pumping system- a pilot test. Society of Petroleum Engineers 26987-MS, 1-16.
Gorzula, S. J., 1978. An ecological study of Caiman crocodilus crocodilus inhabiting savanna laggons in the Venezuelan Guayana. Oecologia 35 (1), 21-34.
Gouldson, A., 2006. Do firms adopt lower standards in poorer areas? Corporate social responsibility and environmental justice in the EU and the US. Area 38 (4), 402-412.
Griffith, J. A. Martinko, E. A., Whistler, J. L. Price, K. P., 2002a. Preliminary Comparison of Landscape Pattern–Normalized Difference Vegetation Index (NDVI) Relationships to Central Plains Stream Conditions. Journal of Environmental Quality 31, 846-859.
Griffith, J. A., Trettin, C. C., O’Neill, R. V., 2002b. A landscape ecology approach to assessing development impacts in the tropics: a geothermal energy example in Hawaii. Singapore Journal of Tropical Geography 23 (1), 1-22.
Guerra, C. V., 2005. Public Affairs and Government Relations, Operadora Cerro Negro, S. A. Puerto la Cruz, Venezuela, personal communication.
Hammond, D. S., ter Steege, H., van der Borg, K., 2006. Upland soil charcoal in the wet tropical forests of central Guyana. Biotropica 39 (2), 153-160.
Harriss, C. L., 2006. Nonrenewable exhaustible resources and property taxation. 2006. American Journal of Economics and Sociology 65 (3), 693-699.
139
Harron, D. E., 2007. Assumptive error and overestimation of effects in wildlife model output. Ecology and Society 12 (1). Available at <http://www.ecologyandsociety.org/vol12/iss1/resp3/>.
Hartter, J., Lucas, C., Gaughan, A. E., Aranda, L. L., 2008. Detecting tropical dry forest succession in a shifting cultivation mosaic of the Yucatán Peninsula, Mexico. Applied Geography In press, 1-16.
Hawth’s Analysis Tools for ArcGIS. 2006. Available at <http://www.spatialecology.com/htools/tooldesc.php>.
Hernández, J. V., Ramos, C., Borjas, M., Jaffee, K., 1999. Growth of Atta laevigata (Hymenoptera formicidae) nests in pine plantations. The Florida Entomologist 82 (1), 97.103.
Heemskerk. M., 2003. Risk attitudes and mitigation among gold miners and others in the Suriname rainforest. Natural Resources Forum 27, 267-278.
Heemskerk. M., 2001. Do international commodity prices drive natural resource booms? An empirical analysis of small-scale gold mining in Suriname. Ecological Economics 39, 295-308.
Hietel, E., Waldhardt, R., Otte, A., 2007. Statistical modeling of land-cover changes based on key socio-economic indicators. Ecological Economics 62, 496-507.
Hilson, G., 2002. The environmental impact of small-scale gold mining in Ghana: identifying problems and possible solutions. The Geographical Journal 168 (1), 57-72.
Hilson, G., Hilson, C. J., Pardie, S., 2007. Improving awareness of mercury pollution in small-scale gold mining communities: challenges and ways forward in rural Ghana. Environmental Research 103, 275-287.
Hilson, G. Monhemius, A. J., 2006. Alternatives to cyanide in the gold mining industry: what prospects for the future? Journal of Cleaner Production 14, 1158-1167.
Hilson, G. Potter, C., 2003a. Structural adjustment and subsistence industry: artisanal gold mining in Ghana. Development and Change 36 (1), 103-131.
Hilson, G. Potter, C., 2003b. Why is illegal gold mining activity so ubiquitous in rural Ghana? African Development Review 15 (2-3), 237-270.
Hodell, D. A., Curtis, J. H., Brenner, M., 1995. Possible role of climate in the collapse of the Classic Maya civilization. Nature 375, 391-394.
IADB., 1997. Venezuela: the fruit of foresight. Inter American Development Bank. Available at <http://www.iadb.org/idbamerica/archive/stories/1997/eng/7g1.htm>.
140
Jackson, L. A., Rubin, T. D., Evans, M. E., Korenaga, G. L., Smart, D. R., Samuels, P. A., Hanrahan, F. E., 2002. Remote sensing for environmental baselining in the petroleum industry. Society of Petroleum Engineers 74164-MS, 1-7.
Jaggernauth, D. D. J. 2003. Using geographical information system (GIS) technology in state owned company. Society of Petroleum Engineers 81157-MS, 1-5.
Janks, J. S., Prelat, A. E., 1995. Multi-temporal analysis of oil field activities in south-central Oklahoma using Landsat Thematic Mapper, aerial photography and GIS. Society of Petroleum Engineers 29692-MS, 31-42.
Janks, J. S., Prelat, A. E., 1994. Environmental impact assessment using remote sensing technology: methodology and case studies. Society of Petroleum Engineers 27121, 313-321.
Jensen, J. R., 2005. Introductory Digital Image Processing: A Remote Sensing Perspective. 3rd ed. Pearson Prentice Hall Inc., Upper Saddle River, NJ.
Jepson, W., 2005. A disappearing biome? Reconsidering land-cover change in the Brazilian savanna. The Geographical Journal 171 (2), 99-111.
Jiayu, N. Jianyi, H., 1999. Formation and distribution of heavy oil and tar sands in China. Marine and Petroleum Geology 16 (1), 85-95.
Jordan-Bychkov, T. G. Domosh, M., 2003. The Human Mosaic: A Thematic Introduction to Geography. W. H. Freeman and Company, New York.
Kalacksa, M., Sanchez-Azofeifa, G. A., Rivard, B., Calvo-Alvarado, J. C., Quesada, M., 2008. Baseline assessment for environmental services payments from satellite imagery: a case study from Costa Rica and Mexico. Journal of Environmental Management In press, 1-12.
Kakabadse, A. P., 2007. Being responsible: boards are reexamining the bottom line. Leadership in Action 27 (1), 2-6.
Kharaka, Y. K., Dorsey, N. S., 2007. Environmental issues of petroleum exploration and production: introduction. Environmental Geosciences 12 (2), 61-63.
Khelil, C., 2006. Oil production capacity expansion. 12-13 Sep. Third OPEC International Seminar. Available at <http://www.opec.org/opecna/Speeches/2006/OPEC_Seminar/PDF/Chakib%20Khelil.pdf>.
Kitula, A. G. N., 2006. The environmental and socio-economic impacts of mining on local livelihoods in Tanzania: a case study of Geita District. 2006. Journal of Cleaner Production 14, 405-414.
Kennedy, D., 2003. Sustainability and the Commons. Science 302, 1861.
141
Keys, E., 2005. Book review: Political Ecology: An Integrative Approach to Geography and Environment-Development Studies. Zimmerer, K. S., Bassett, T. J. (Eds.), 2003. Annals of the Association of American Geographers 95 (1), 223-236.
Keys, E., McConnell, W. J., 2005. Global change and the intensification of agriculture in the tropics. Global Environmental Change 15, 320-337.
Kligerman, D. C., La Rovere, E. L., Costa, M. A., 2001. Management challenges on small-scale gold mining activities in Brazil. Environmental Research Section A 87, 181-198.
Kopper, R., Kupecz, J., Curtis, C., Cole, T., Dorn-Lopez, D., Copley, J., Munoz, A., Caicedo, V., 2001. Reservoir characteristics of the Orinoco heavy oil belt: Miocene Oficina Formation, Zuata field, eastern Venezuela basin. Society of Petroleum Engineers 69697-MS, 1-16.
Kupfer, J. A., Webbeking, A. L., Franklin, S. B., 2004. Forest fragmentation affects early successional patterns of shifting cultivation fields near Indian Church, Belize. Agriculture Ecosystems & Environment 103, 509-518.
Kumah, A., 2006. Sustainability and gold mining in the developing world. Journal of Cleaner Production 14, 315-323.
Kwarteng, A. Y., 1999. Remote sensing assessment of oil lakes and oil polluted surfaces at the Greater Burgan oil field, Kuwait. International Journal of Applied Earth Observation and Geoinformation 1 (1), 36-47.
Kwarteng, A. Y., 1998. Multitemporal remote sensing data analysis of Kuwait’s oil lakes. Environment International 24 (1/2), 121-137.
Lacerda, L. D., 1997. Global mercury emissions from gold and silver mining. Water, Air and Soil Pollution 97, 209-221.
Lacerda, L. D., Marins, R. V., 1997. Anthropogenic mercury emissions to the atmosphere in Brazil: The impact of gold mining. Journal of Geochemical Exploration 58, 223-229.
Lambin, E.F. 1999. Monitoring forest degradation in tropical regions by remote sensing: some methodological issues. Global Ecology and Biogeography. 9. 191-198.
Lambin, E.F., Geist, H. J., Lepers, E., 2003. Dynamics of land-use and land-cover change in tropical regions. Annual Review of Environmental Resources 28, 205-241.
Lambin, E. F., Turner, B. L., Geist, H. J., Agbola, S. B., Angelsen, A., Bruce, J. W., Coomes, O. T., Dirzo, R., Fischer, G., Folke, C., George, P. S., Homewood, K., Imbernon, J., Leemans, R., Li, X., Moran, E. F., Mortimore, M., Ramakrishnan, P. S., Richards, J. F., Skånes, H., Steffen, W., Stone, G. D., Svedin, U., Veldkamp, T. A., Vogel, C., Xu, J., 2001. The causes of land-use and land-cover change: moving beyond the myths. Global Environmental Change 11, 261-269.
142
Lawrence, D., D’Odorico, P., Diekmann, L., DeLonge, M., Das, R., Eaton, J., 2007. Ecological feedbacks following deforestation create the potential for a catastrophic ecosystem shift in tropical dry forest. Proceedings of the National Academy of Sciences 104 (52), 20696-20701.
Lawrence, R., 2007. Corporate social responsibility, supply-chains and Saami claims: tracing the political in the Finnish forestry industry. Geographical Research 45 (2), 167-176.
Layrisse, I., 1999. Heavy oil production in Venezuela: historical recap and scenarios for the next century. Society of Petroleum Engineers 53464-MS, 1-13.
Lee, P., Boutin, S., 2006. Persistence and development transition of wide seismic lines in the western boreal plains of Canada. Journal of Environmental Management 78, 240-250.
Leger, J. P., 1991. Trends and causes of fatalities in South African mines. Safety Science 14, 169-185.
Lehmkuhl, L. D., 1996. Nonparametric statistics: methods for analyzing data not meeting assumptions required for the application of parametric tests. Journal of Prosthetics and Orthotics. 8 (3), 105-113. Available at <http://www.oandp.org/jpo/library/1996_03_105.asp>.
Lele, U., Viana, V., Verissimo, A., Vosti, S., Perkins, K., Husain, S. A., 2000. Brazil Forests in the balance: challenges of conservation with development. World Bank Operations Evaluation Department. Available at <http://lnweb18.worldbank.org/oed/oeddoclib.nsf/b57456d58aba40e585256ad400736404/d985594af1f1dfb185256970007c1acc/$FILE/BrazilCS.pdf>.
Lima, L. R. P. de A., Bernardez, L. A., Barbosa, L. A. D., 2008. Characterization and treatment of artisanal gold mine tailings. Journal of Hazardous Materials 150, 747-753.
Lopez-Claros, A., 2006. Latin America’s economic competitiveness. World Economic Forum. Available at <http://www.weforum.org/en/events/ArchivedEvents/latinamerica/IssuesinDepth/LatinAmericaseconomic/Latin_Americas_economic>.
López-Fernández, H., Winemiller, K. O., 2000. A review of Venezuelan species of Hypophthalmus (Siluriformes: Pimelodidae). Ichthyology Exploration Freshwaters 11 (1), 35-46.
Lovins, A. B., Lovins, L. H., Hawken, P., 2000. A road map for natural capitalism. In: Harvard Business Review: on Business and the Environment. Harvard Business School Press, Boston.
Lugo, R. G., Eggenschwiler, M., Uebel, T., 2001. How fluid and rock properties affect production rates in a heavy-oil reservoir Cerro Negro, Venezuela. Society of Petroleum Engineers 69694-MS, 1-8.
143
M&E (Management and Excellence)., 2008. World’s most sustainable and ethical oil companies 2008. Available at <http://www.management-rating.com/archivo/Brochure%20Oil%20Study%202008L.pdf>.
M&E (Management and Excellence)., 2007. World’s most sustainable and ethical oil companies 2007. Available at <http://www.management-rating.com/archivo/Brochure%20Oil%20Study%202007l.pdf>.
M&E (Management and Excellence)., 2006. World’s most sustainable and ethical oil companies 2006. Available at <http://www.management-rating.com/archivo/Brochure%20Oil%20Study%202006-CHL.pdf>.
M&E (Management and Excellence)., 2005. Ethics in the oil industry. Available at <http://www.management-rating.com/archivo/Brochure%20Oil%20Study%202005.pdf>.
MacFarlane, A. K., 2003. Vegetation response to seismic lines: edge effects and on-line succession. University of Alberta, MS Thesis. Available at < http://www.borealcentre.ca/reports/seismic_thesis.pdf>.
Malm, O., 1998. Gold mining as a source of mercury exposure in the Brazilian Amazon. Environmental Research, Section A 77, 73-78.
Mambo, J., Archer, E., 2007. An assessment of land degradation in the Save catchment of Zimbabwe. Area 39 (3), 380-391.
Mangiarotti, S., Mazzega, P., Jarlan, L., Mougin, E., Baup, F., and Demarty, J, 2008. Evolutionary bi-objective optimization of a semi-arid vegetation dynamics model with NDVI and σ0 satellite data. Remote Sensing of Environment 112, 1365-1380.
Marica, S., Cetean, V., Lazaroiu, G., 2008. Unitary management and environmental performance by monitoring and protection of mineral resources for construction materials from Romania. Building and Environment 43, 1082-1090.
Marín, G., Bastidas, L., Muñoz, J., Oliveros, O., Navarro, R., 2007. Perfil ecológico de la avifauna de los llanos orientales de Venezuela en función de los impactos antrópicos. Interciencia 32 (6), 391-398. Available at <http://www.interciencia.org/v32_06/index.html>.
Masera, O., Ghilardi, A., Drigo, R., Trossero, M. A., 2006. WISDOM: a GIS-based supply demand mapping tool for fuelwood management. Biomass and Bioenergy 30, 618-637.
Mauricio, J., Vincent, L., Moret, A. Y., Quevedo, A., 2005. Comparación entre modalidades de muestreo en plantaciones de pino caribe en el oriente de Venezuela. Revista Forestal Venezolana 49 (1), 17-25. Available at <http://www.saber.ula.ve/db/ssaber/Edocs/pubelectronicas/revistaforestal/vol49-1/articulo2.pdf>.
144
McCoy, T. L., Fensom, M., 2006. 2006 Latin American business environment report. University of Florida, Latin American Business Environment Program. Available at <http://translationstudies.ufl.edu/publications/LABER2006.pdf>.
McCusker, B., Wiener, D., 2003. GIS representations of nature, political ecology, and the study of land use and land cover change in South Africa. In: Zimmerer, K. S., Bassett, T. J. (Eds.), Political Ecology: An Integrative Approach to Geography and Environment-Development Studies. Guilford Press, London, pp. 181-200.
McMichael, A. J., Butler, C. D., Folke, C., 2003. New visions for addressing sustainability. Science 302, 1919-1920.
Mela, M., Randi, M. A. F., Ventura, D. F., Carvalho, C. E. V., Pelletier, E., Ribeiro, C. A. O., 2007. Effects of dietary methylmercury on liver and kidney histology in the neotropical fish Hoplias malabaricus. Ecotoxicology and Environmental Safety 68, 426-435.
Meyer, R. F., Attanasi, E. D., 2003. Heavy oil and natural bitumen- strategic petroleum resources. (=USGS Fact Sheet 70-03). Available at < http://pubs.usgs.gov/fs/fs070-03/fs070-03.html>.
Mol, J. H., Ouboter, P. E., 2004. Downstream effects from erosion from small-scale gold mining on the instream habitat and fish community of a small neotropical rainforest stream. Conservation Biology 18 (1), 201-214.
Mommer, B., 2004. The value of extra-heavy crude oil from the Orinoco belt. Venezuelan Ministry of Energy and Mines. Available at <http://www-personal.umich.edu/~twod/venezuela/nwaeg_mar07/doktor_mommer_orimulsion.pdf>.
Moore, B. A., Duncan, J. R., Burgess, J. E., 2008. Fungal bioaccumulation of copper, nickel, gold and platinum. Minerals Engineering 21, 55-60.
Morhardt, J. E., 2002a. Clean, Green, and Read All Over: Ten Rules for Effective Corporate Environmental and Sustainability Reporting. ASQ Quality Press, Milwaukee, WI.
Morhardt, J. E., Baird, S., Freeman, K., 2002. Scoring corporate environmental and sustainability reports using GRI 2000, ISO 14031, and other criteria. Corporate Social Responsibility and Environmental Management 9, 215-233.
Morton, P., Weller, C., Thomson, J., M. Haefele, Culver, N., 2004.Drilling in the Rocky Mountains: how much and at what cost? The Wilderness Society, Washington DC, pp. 1-33.
Morton, P., Weller, C., Thomson, J., 2002. Energy and western wildlands: a GIS analysis of economically recoverable oil and gas. The Wilderness Society, Washington DC, pp. 1-31.
Morse, S., 2008. On the use of headline indices to link environmental quality and income at the level of the nation state. Applied Geography 28, 77-95.
145
Moser, T., 2001. MNCs and sustainable business practice: the case of the Colombian and Peruvian petroleum industries. World Development 29 (2), 291-309.
Mudd, G. M., 2007a. Global trends in gold mining: towards quantifying environmental and resource sustainability? Resources Policy 32, 42-56.
Mudd, G. M., 2007b. Gold mining in Australia: linking historical trends and environmental and resource sustainability. Environmental Science and Policy 10, 629-644.
Musinsky, J. N., Reining, C. C. S., Rosenfeld, A. B., Thomsen, J. B., 1998. An analysis of human settlement along the Xan oil road in Laguna del Tigre National Park, Guatemala. Society of Petroleum Engineers 46824-MS, 1-11.
Nelleman, C. Cameron, R. D., 1996. Effects of petroleum development on terrain preferences on calving caribou. Arctic 49 (1), 23-28.
Nepstad, D., Carvalho, G., Barros, A. C., Alencar, A., Capobianco, J. P., Bishop, J., Moutinho, P., Lefebvre, P., Silva, U. L. J., Prins, E., 2001. Road paving, fire regime feedbacks, and the future of Amazon forests. Forest Ecology and Management 154 (3), 395-407.
Nepstad, D. C., Veríssimo, A., Alencar, A., Nobre, C., Lima, E., Lefebvre, P., Schlesinger, P., Potter, C., Moutinho, P., Mendoza, E., Cochrane, M., Brooks, V., 1999. Large-scale impoverishment of Amazonian forests by logging and fire. Nature 398, 505-508.
Neumayer, E., 2000. Scarce or abundant? The economics of natural resource availability. Journal of Economic Surveys 14 (3), 307-335.
Noel, L. E., Parker, K. R., Cronin, M. A., 2006. Response to Joly et al. (2006), a reevaluation of caribou distribution near an oilfield road on Alaska’s North Slope. Wildlife Society Bulletin 34 (3), 870-873.
Nogueira, E. M., Fearnside, P. M., Nelson, B. W., França, M. B., 2007. Wood density in forests of Brazil’s ‘arc of deforestation’: implications for biomass and flux of carbon from land-use change in Amazonia. Forest Ecology and Management 248, 119-135.
Noomen, M, F., Skidmore, A. K., Van der Meer, F. D., Prins, H. H. T., 2006. Continuum removed band depth analysis for detecting the effects of natural gas, methane and ethane on maize reflectance. Remote Sensing of Environment 105, 262-270.
Oil Company Representative., 2007. European company operating in Venezuela, personal communication.
Oil Company Representative., 2005a. European company operating in Venezuela, personal communication.
Oil Company Representative., 2005b. European company operating in Venezuela, personal communication.
146
Oil Company Representative., 2005c. European company operating in Venezuela, personal communication.
Oil Company Representative., 2005d. US company operating in Venezuela, personal communication.
Oil Company Representative., 2005e. US company operating in Venezuela, personal communication.
Oil Company Representative., 2005f. US company operating in Venezuela, personal communication.
Oil Company Representative., 2005g. Venezuelan company, personal communication.
Oil Company Representative., 2005d. European company operating in Venezuela, personal communication.
OPEC., 2007. Annual Statistical Bulletin 2006. Available at <http://www.opec.org/library/Annual%20Statistical%20Bulletin/pdf/ASB2006.pdf>.
Orlitzky, M., Schmidt, F. L., Rynes, S. L., 2003. Corporate social and financial performance: a meta-analysis. Organization Studies 24 (3), 403-441.
Ostrom, E., Burger, J., Field, C. B., Norgaard, R. B., Policansky, D., 1999. Revisiting the commons: local lessons, global challenges. Science 284, 278-282.
Osuji, L. C., Onojake, C. M., 2006. Field reconnaissance and estimation of petroleum hydrocarbon and heavy metal contents of soil affected by the Ebocha-9 oil spillage in Niger Delta, Nigeria. Journal of Environmental Management 79, 133-139.
Pacific Sustainability Index (PSI)., 2007. Claremont McKenna College, Claremont, CA. Available at <http://www.roberts.cmc.edu/PSI/whatthescoresmean.asp>.
Padilla, M., Dowler, R. C., 1994. Tapirus terrestris. Mammalian Species 481, 1-8.
Parker, D. C., Manson, S. M., Janssen, M. A., Hoffmann, M. J., Deadman, P., 2002. Multi-Agent Systems for the Simulation of Land-Use and Land-Cover Change: A Review. Annals of the Association of American Geographers 93 (2), 314.
Parra, P., 2007. Farmer from Guárico State, Venezuela, personal communication.
Parrotta, J. A., Knowles, O. H., 2001. Restoring tropical forests on lands mined for bauxite: examples from the Brazilian Amazon. Ecological Engineering 17, 219-239.
Pawlewicz, M., 1995. Heavy-oil resources of the United Status. USGS. Available at <http://energy.cr.usgs.gov/1995OGData/HEAVYOIL.pdf >.
PDVSA., 2008a. Venezuela e Italia acuerdan proyectos en la Faja. 29 Feb. Available at <http://www.pdvsa.com>.
147
PDVSA., 2008b. Más de 600 mil barriles diarios producen mejoradores de la Faja Petrolífera del Orinoco. 11 Mar. Available at <http://www.pdvsa.com>.
PDVSA., 2008c. Petróleo para el pueblo. Available at <http://www.pdvsa.com/index.php?tpl=interface.sp/design/readmenuprinc2.tpl.html&newsid_temas=38>.
PDVSA., 2007a. Chávez y Kirchner consolidan integración latinoamericana en la faja petrolífera del Orinoco. 20 Feb. Available at <http://www.pdvsa.com/index.php?tpl=interface.sp/design/readsearch.tpl.html&newsid_obj_id=3538&newsid_temas=0>.
PDVSA., 2007b. Venezuela y Belarús confirman existencia de 26,5 millardos de barriles de crudo en el bloque Junín 1 de la Faja. Available at <http://www.pdvsa.com
PDVSA., 2007c. Faja petrolífera del Orinoco. Available at <http://www.pdvsa.com/index.php?tpl=interface.sp/design/readmenuprinc.tpl.html&newsid_temas=96>.
PDVSA., 2007d. Chile participará en faja petrolífera Orinoco. 18 Apr. Available at <http://www.pdvsa.com>.
PDVSA., 2007e. MENPET inciará proceso de selección de empresas en proyectos integrados de la Faja. 22 Dec. Available at <http://www.pdvsa.com>.
PDVSA., 2006a. PDVSA y Petropars iniciaron perforación en el Bloque Ayacucho 7 de la Faja Petrolífera del Orinoco. 18 Sep. Available at <http://www.pdvsa.com/index.php?tpl=interface.sp/design/readsearch.tpl.html&newsid_obj_id=3004&newsid_temas=0>.
PDVSA., 2006b. Aspiramos que Ecuador esté pronto en la Faja Petrolífera del Orinoco. 20 Dec. Available at <http://www.pdvsa.com>.
PDVSA., 2005a. ¿Qué es la Faja Petrolífera del Orinoco? 19 Aug. Available at <http://www.pdvsa.com>.
PDVSA., 2005b. Faja Petrolífera del Orinoco. Available at <http://www.pdvsa.com/interface.sp/database/fichero/doc/570/9.PDF>.
Pelling, M., 2003. Urban environmental risk in Guyana. In: Zimmerer, K. S., Bassett, T. J. (Eds.), Political Ecology: An Integrative Approach to Geography and Environment-Development Studies. Guilford Press, London, pp. 73-93.
Peres, C. A., Terborgh, J. W., 1995. Amazonian nature reserves: an analysis of the defensibility status of existing conservation units and design criteria for the future. Conservation Biology 9 (1), 34-46.
148
Perez, J. A., 2000. Primer informe de país para la convención sobre diversidad biológica. Ministerio del Ambiente y de los Recursos Naturales (MARN), Caracas, Venezuela.
Perz, S. G., Caldas, M. M., Arima, E., Walker, R. T., 2007. Unofficial road building in the Amazon: socioeconomic and biophysical explanations. Development and Change 38 (3), 529-551.
Peterson, G. D. Heemskerk, M., 2001. Deforestation and forest regeneration following small-scale gold mining in the Amazon: the case of Suriname. Environmental Conservation 28 (2), 117-126.
Petroguía., 2006-2007. Energy map of Venezuela. Editorial Los Barrosos, Caracas, Venezuela.
Petroguía., 2004-2005. Energy map of Venezuela. Editorial Los Barrosos, Caracas, Venezuela.
Pett, M. A., 1997. Nonparametric Statistics in Health Care Research: Statistics for Small Samples and Unusual Distributions. Sage Publications, Thousand Oaks, CA.
Phizackerley, P. H., Scott, L. O., 1978. Major tar-sand deposits of the world. In: Chilingarian, G. V., Yen, T. F. (Eds.), Bitumens, Asphalts and Tar Sands. Elsevier Publishing, Amsterdam.
Pielke, R. A. S., Adegoke, J., Beltrán-Przekurat, A., Hiemstra, C. A., Lin, J., Nair, U. S., Niyogi, D., Nobis, T. E., 2007. An overview of regional land-use and land-cover impacts on rainfall. Tellus B 59 (3), 587-601.
Pina-Acuna, G., Ferreira, D., 2004. Well-testing challenges of the Hamaca Project, Orinoco heavy oil belt, Venezuela. Society of Petroleum Engineers 86930-MS, 1-10.
Pinheiro, M. C. N., Macchi, B. M., Vieira, J. L. F., Oikawa, T., Amoras, W. W., Guimarães, G. A., Costa, C. A., Crespo-López, M. E., Herculano, A. M., Silveira, L. C. L., do Nascimento, J. L, M., 2008. Mercury exposure and antioxidant defenses in women: a comparative study in the Amazon. Environmental Research In press, 1-7.
Porter, M. E., van der Linde, C., 2000. Green and competitive: ending the stalemate. In: Harvard Business Review: on Business and the Environment. Harvard Business School Press, Boston.
Prakash A., Gupta, R. P., 1999. Surface fires in Jharia coalfield, India—their distribution and estimation of area and temperature from TM data. International Journal of Remote Sensing 20 (10), 1935-1946.
Ramos, M. A., de Medeiros, P. M., de Almeida, A. L. S., Feliciano, A. L. P., de Albuquerque, U. P., 2008. Can wood quality justify local preferences for firewood in an area of caatinga (dryland) vegetation? Biomass and Bioenergy In Press, 1-7.
Read, J.M. Lam, N. S., 2002. Spatial methods for characterising land cover and detecting land-cover changes for the tropics. International Journal of Remote Sensing 23 (12), 2457-2474.
149
Redman, C. L., Grove, M. J., Kuby, L, H., 2004. Integrating social science into the long-term ecological research (LTER) network: social dimensions of ecological change and ecological dimensions of social change. Ecosystems 7, 161-171.
Reinhardt, F. L., 2000. Bringing the environment down to earth. In: Harvard Business Review: on Business and the Environment. Harvard Business School Press, Boston.
Richard, D., 2008. Department of Psychology, University of North Florida, personal communication.
Rindfuss, R. R., Entwisle, B., Walsh, S. J., Mena, C. F., Erlien, C. M., Gray, C. L., 2007. Frontier land use change: synthesis, challenges, and next steps. Annals of the Association of American Geographers 97 (4), 739-754.
Robbins, P., 2004. Political Ecology: A Critical Introduction. Blackwell Publishing, Oxford, UK.
Robbins, P., 2003a. Political ecology in political geography. Political Geography 22, 641-645.
Robbins, P., 2003b. Fixed categories in a portable landscape. In: Zimmerer, K. S., Bassett, T. J. (Eds.), Political Ecology: An Integrative Approach to Geography and Environment-Development Studies. Guilford Press, London, pp. 181-200.
Robles, J., 2001. Application of advanced heavy-oil production technologies in the Orinoco heavy-oil-belt, Venezuela. Society of Petroleum Engineers 69848-MS, 1-11.
Rodrigues, R. R., Martins, S. V., de Barros, L. C., 2004. Tropical rain forest regeneration in an area degraded by mining in Mato Grosso State, Brazil. Forest Ecology and Management 190, 323-333.
Rodriguez, H. A., Vaca, P., Gonzalez, O., de Mirabal, M. C., 1997. Integrated study of a heavy oil reservoir in the Orinoco belt: a field case simulation. Society of Petroleum Engineers 38015-MS, 309-310.
Rodríguez, J. P., 2000. Impact of the Venezuelan economic crisis on wild populations of animals and plants. Biological Conservation 96 (2), 151-159.
Romero-Duque, L. P., Jaramillo, V. J., Pérez-Jiménez, A., 2007. Structure and diversity of secondary tropical dry forests in Mexico, differing in their prior land-use history. Forest Ecology and Management 253, 38-47.
Rosales, J. G., Petts, G., Salo, J., 1999. Riparian flooded forests of the Orinoco and Amazon basins: a comparative review. Biodiversity and Conservation 8, 551-586.
Rubenstein, S. L., 2004. Steps to a political ecology of Amazonia. Tipití 2 (2), 131-176.
Rumsey, D., 2007. Intermediate Statistics for Dummies. Wiley Publishing, Hoboken, NJ.
150
Salisbury, D. S., Schmink, M., 2007. Cows versus rubber: changing livelihoods among Amazonian extractivists. Geoforum In press, 1-17.
Sandker, M., Suwarno, A., Campbell, B. M., 2007. Will forests remain in the face of oil palm expansion? Simulating change in Malinau, Indonesia. Ecology and Society 12 (2), 1-18. Available at <http://www.ecologyandsociety.org/vol12/iss2/art37/>.
Sandoval, M. C., Veiga, M. M., Hinton, J., Sandner, S., 2006. Application of sustainable development concepts to an alluvial mineral extraction project in lower Caroni River, Venezuela. 2006. Journal of Cleaner Production 14, 415-426.
San José, J. J., Bracho, R., Nikonova, N., 2001. A comparative analysis of the flooding and fire effects on the energy exchange in a wetland community (Morichal) of the Orinoco Llanos. Journal of Hydrology 242, 228-254.
Satapathy, D. R., Katpatal, Y. B., Wate, S. R., 2008. Application of geospatial technologies for environmental impact assessment: and Indian scenario. International Journal of Remote Sensing 29 (2), 355-386.
Schapiro, M., 2007. Exposed: The Toxic Chemistry of Everyday Products and What’s at Stake for American Power. Chelsea Green Publishing, White River Junction, VT.
Schmidt, H. Glaesser, C., 1998. Multitemporal analysis of satellite data and their use in the monitoring of the environmental impacts of open cast lignite mining in areas in Eastern Germany. International Journal of Remote Sensing 19 (12), 2245-2260.
Schmink, M. Wood, C., 1992. Contested Frontiers in Amazonia. Columbia University Press, New York.
Schneider, R. Dyer, S., 2006. Death by a thousand cuts: Impacts of in situ oil sands development on Alberta’s boreal forest. Canadian Parks and Wilderness Society and The Pembina Institute. Available at <http://www.pembina.org/pdf/publications/1000-cuts.pdf>.
Schneider R. R., Stelfox, J. B., Boutin, S., Wasel, S., 2007. Scenarios are plausible stories about the future, not forecasts. Ecology and Society 12 (1), 1. Available at <http://www.ecologyandsociety.org/vol12/iss1/resp4>.
Schneider R. R., Stelfox, J. B., Boutin, S., Wasel, S., 2003. Managing the cumulative impacts of land uses in the western Canadian sedimentary basin: a modeling approach. Conservation Ecology 7 (1), 1-11. Available at <http://www.consecol.org/vol7/iss1/art8>.
Scrieciu, S. S., 2007. Can economic causes of tropical deforestation be indentified at a global level? Ecological Economics 62 (3-4), 603-612.
Sebastiani, M., Martín, E., Adrianza, D., Méndes, C., Villaró, M., Saud, Y., 2001. Linking impact assessment to an environmental management system. Case study: downstream upgrading petroleum plant in Venezuela. Environmental Impact Assessment Review 21 (2), 137-168.
151
Sebbenn, A. M., Degen, B., Azevedo, V. C. R., Silva, M. B., de Lacerda, A. E. B., Ciampi, A. Y., Kanashiro, M., Carneiro, F. d. S., Thompson, I., Loveless, M. D., 2007. Modelling the long-term impacts of selective logging on genetic diversity and demographic structure of four tropical tree species in the Amazon forest. Forest Ecology and Management In press, 1-15.
Shapiro, R. B. 2000. Cited in: Magretta, J. Growth through global sustainability: an interview with Monsanto’s CEO, Robert B. Shapiro. In: Harvard Business Review: on Business and the Environment. Harvard Business School Press, Boston.
Siegert, F., Hoffmann, A. A., 2000. The 1998 forest fires in east Kalimantan (Indonesia): a quantitative evaluation using high resolution, multitemporal ESR-2 SAR images and NOAA-AVHRR hotspot data. Remote Sensing of the Environment 72, 64-77.
Sierra, R., Tirado, M., Palacios, W., 2003. Forest-cover change from labor-and-capital-intensive commercial logging in the southern Chocó rainforests. The Professional Geographer 55 (4), 477-490.
Simmons, C. S., 2002. The local articulation of policy conflict: land use, environment, and Amerindian rights in eastern Amazonia. The Professional Geographer 54 (2), 241-258.
Simmons, C. S., Walker, R. T., Arima, E. Y., Aldrich, S. P., Caldas, M. M., 2007. The Amazon land war in the south of Pará. Annals of the Association of American Geographers 97 (3), 567-592.
Smith, R. J., Muir, R. D. J., Walpole, M. J., Balmford, A., Leader-Williams, N., 2003. Governance and the loss of biodiversity. Nature 426, 67-70.
Soares-Filho, B. S., Nepstad, D. C., Curran, L. M., Cerqueira, G. C., Garcia, R. A., Ramos, C. A., Voll, E., McDonald, A., Lefebvre, P., Schlesinger, P., 2006. Modelling conservation in the Amazon basin. Nature 440, 520-523.
Soares-Filho, B., Alencar, A., Nepstad D. Cerqueira, G., Diaz, M. V., Rivero, S. Solórzano, L., Voll, E., 2004. Simulating the response of land-cover changes to road paving and governance along a major Amazon highway: the Santarém-Cuiabá corridor. Global Change Biology 10 (5), 745-764.
Söderbergh, B., Robelius, F., Alekett, K., 2007. A crash programme scenario for the Canadian oil sands industry. Energy Policy 35, 1931-1947.
Southworth, J., 2004. An assessment of Landsat TM band 6 thermal data for analysing land cover in tropical dry forest regions. International Journal of Remote Sensing 25 (4), 689-706.
Southworth, J, Munroe, D., Nagrenda, H., 2004. Land cover change and landscape fragmentation—comparing the utility of continuous and discrete analyses for a western Honduras region. Agriculture, Ecosystems and the Environment 101, 185-205.
152
Spellberger, I. F., 1998. Ecological effects of roads and traffic: a literature review. Global Ecology and Biogeography Letters 7 (5), 317-333.
Stein, J. L., Stein, J. A., Nix, H. A., 2002. Spatial analysis of anthropogenic river disturbance at regional and continental scales: identifying the wild rivers of Australia. Landscape and Urban Planning 60, 1-25.
Talwani, M., 2002. The Orinoco heavy oil belt in Venezuela: Or heavy oil to the rescue? Rice University Department of Earth Science, Houston, TX. Available at <http://cohesion.rice.edu/naturalsciences/earthscience/research.cfm?doc_id=2819>.
Tang, J., Wang, L., Myint, S. W., 2007. Improving urban classification through fuzzy supervised classification and spectral mixture analysis. International Journal of Remote Sensing 28 (18), 4047-4063.
Tankersley, T. H., Waite, M. W., 2002. Reservoir modeling for horizontal exploitation of a giant heavy oil field- challenges and lessons learned. Society of Petroleum Engineers 78957-MS, 1-6.
Tarnavsky, E., Garrigues, S., Brown, M. E., 2008. Multiscale geostatistical analysis of AVHRR, SPOT-VGT, and MODIS global NDVI products. Remote Sensing of Environment 112, 539-549.
Thomson, J. L., Schaub, T. S., Culver, N. W., Aengst, P. C., 2005. Wildlife at a crossroads: energy development in western Wyoming. The Wilderness Society, Washington DC, pp. 1-40.
Timoney, K., Lee, P., 2001. Environmental management in resource-rich Alberta, Canada: first world jurisdiction, third world analogue? Journal of Environmental Management 63, 387-405.
Torres, I., 2004. The mineral industry of Venezuela. US Geological Survey Minerals Yearbook. Available at <http://minerals.usgs.gov/minerals/pubs/country/2004/vemyb04.pdf>.
Trebolle, R. L., Chalot, J. P., Colmenares, R., 1993. The Orinoco heavy oil-belt pilot projects and development strategy. Society of Petroleum Engineers 25798-MS, 207-216.
Trejo, I. Dirzo, R., 2000. Deforestation of seasonally dry tropical forest: a national and local analysis in Mexico. Biological Conservation 94, 133-142.
Treweek, J., Veitch, N., 1996. The potential application of GIS and remotely sensed data to the ecological assessment of proposed road network schemes. Global Ecology and Biogeography Letters 5 (4/5), 249-257.
Trujillo, A. J., 2004. Petro-state constraints on health policy: guidelines for workable reform in Venezuela. Health Policy 67, 39-55.
153
Turner, B. L. II., Lambin, E. F., Reenberg, A., 2007. The emergence of land change science for global environmental change and sustainability. Proceedings of the National Academy of Sciences 104 (52), 20666-20671. Available at <http://www.pnas.org/cgi/reprint/104/52/20666?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&fulltext=the+emergence+of+land+change+science&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT>.
Turner, B. L. II., Villar, S. C., Foster, D., Geoghegan, J., Keys, E., Klepeis, P., Lawrence D., Mendoza, P. M., Manson, S., Ogneva-Himmelberger, Y., Plotkin, A. B., Salicrup, D. P., Chowdhury, R. R., Savitsky, B., Schneider, L., Schmook, B., Vance, C., 2001. Deforestation in the southern Yucatán peninsular region: an integrative approach. Forest Ecology and Management 154, 353-370.
Turner, M., 2003. Environmental science and social causation in the analysis of Sahelian pastoralism. In: Zimmerer, K. S., Bassett, T. J. (Eds). Political Ecology: An Integrative Approach to Geography and Environment-Development Studies. Guilford Press, London, pp. 159-180.
Ud-Din, S., Al-Dousari, A., Literathy, P., 2008. Evidence of hydrocarbon contamination from the Burgan oil field, Kuwait—interpretations from thermal remote sensing data. Journal of Environmental Management 86, 605-615.
Urquiza-Haas, T., Dolman, P. M., Peres, C. A., 2007. Regional scale variation in forest structure and biomass in the Yucatan Peninsula, Mexico: effects of forest disturbance. Forest Ecology and Management 247, 80-90.
Veiga, M. M., Hinton, J. J., 2002. Abandoned artisanal gold mines in the Brazilian Amazon: a legacy of mercury pollution. Natural Resources Forum 26, 15-26.
Veiga, M. M., Maxson, P. A., Hylander, L. D., 2006. Origin and consumption of mercury in small-scale gold mining. Journal of Cleaner Production 14, 436-447.
Veldkamp, A. Lambin, E. F., 2001. Predicting land-use change. Agriculture, Ecosystems and Environment 85, 1-6.
Vieira, I. C. G., de Almeida, A. S., Davidson, E. A., Stone, T. A., de Carvalho, C. J. R., Guerrero, J. B., 2003. Classifying successional forests using Landsat spectral properties and ecological characteristics in eastern Amazonia. Remote Sensing of Environment 87, 470-481.
Vieira, R., 2006. Mercury-free gold mining technologies: possibilities for adoption in the Guianas. Journal of Cleaner Production 14, 448-454.
Veneklaas, E. J., Fajardo, A., Obregon, S., Lozano, J., 2005. Gallery forest types and their environmental correlates in a Colombian savanna landscape. Ecography 28, 236-252.
154
Wannasai, N., Shrestha, R. P., 2007. Role of land tenure security and farm household characteristics on land use change in the Prasae watershed, Thailand. Land Use Policy In press, 1-10.
Wantzen, K. M., Ramírez, A., Winemiller, K. O., 2006. New vistas in neotropical stream ecology—preface. Journal of the North American Benthological Society 25 (1), 61-65.
Wassenaar, T., Gerber, P., Verburgh, P. H., Rosales, M., Ibrahim, M., Steinfeld, H., 2007. Projecting land use changes in the Neotropics: the geography of pasture expansion into forest. Global Environmental Change 17, 86-104.
Wheldon, A. E., 1990. Energy in Swaziland: options for increased self-reliance at individual and national level. Solar and Wind Technology 7 (1), 83-88.
White, C. L., 1956. Cattle raising: a way of life in the Venezuelan llanos. The Scientific Monthly 83 (3), 122-129.
Wilcoxon, B. R., Cramer, D. F., 2002. Strategic sustainability integration: the next step in the journey. Society of Petroleum Engineers 74409-MS, 1-7.
Wilderness Society., 2006. Habitat fragmentation from roads: travel planning methods to safeguard Bureau of Land Management Lands. Science and Policy Brief. Washington DC, pp. 1-16.
Williams, P. A., 2006. Projections for the geopolitical economy of oil after war in Iraq. Futures 38 (9), 1074-1088.
Wolfe, S. A., Griffith, B., Wolfe, C. A. G., 2000. Response of reindeer and caribou to human activities. Polar Research 19 (1), 63-73.
Wolford, W., 2005. Book review: Political Ecology: A Critical Introduction. Robbins, P., 2004. Annals of the Association of American Geographers 95 (3), 692-726.
World Economic Forum (WEF)., 2005.Partnering for success: business perspectives on multistakeholder partnerships. Available at <http://www.weforum.org/pdf/ppp.pdf>.
Wyoming, GFD (Game and Fish Department)., 2004. Recommendations for development of oil and gas resources within crucial and important wildlife habitats. Available at <http://gf.state.wy.us/downloads/pdf/og.pdf>.
Yatabe, S. M., Fabbri, A. G., 1986. The application of remote sensing to Canadian petroleum exploration: promising yet unexploited. Computers and Geosciences 12 (4), 597-609.
Zarin, D. J., Schulze, M. D., Vidal, E., Lentini, M., 2007. Beyond reaping the first harvest: management objectives for timber production in the Brazilian Amazon. Conservation Biology 21 (4), 916-925.
155
Zhang, G. F., Shen, X. H., Zou, L. J., Li, C. J., Wang, Y. L., Lu, S. L., 2007. Detection of hydrocarbon bearing sand through remote sensing techniques in the western slope zone of Songliao basin, China. International Journal of Remote Sensing 28 (8), 1819-1833.
Zimmerer, K. S., and T. J. Bassett. (Eds.), 2003. Political Ecology: An Integrative Approach to Geography and Environment-Development Studies. Guilford Press, London.
156
BIOGRAPHICAL SKETCH
Chris W. Baynard was born in Maryland in 1966. Within a month his family and he moved
to Spain where they lived in Madrid for 9 years. Thereafter they moved back to the US and
settled in Charleston, SC. There, Chris attended the College of Charleston, earning a B.S. in
psychology and minoring in Spanish and communications. Afterward, he earned his M.A. in
Spanish from Louisiana State University (LSU) in Baton Rouge, and a minor concentration in
geography and anthropology. At LSU he met his soon-to-be wife Ana, from Venezuela, and they
were married in Baton Rouge and in Venezuela. After completing his M.A. Chris entered
academics as a Spanish instructor first at LSU, then at the University of North Florida, in
Jacksonville, FL. Convinced he wanted to continue with an academic career, Chris began his
doctoral program in geography at the University of Florida (UF), studying part-time while
working full-time and living in Jacksonville. At UF his research interests were land-use and
land-cover change related to nonrenewable natural resources, environmental business practices,
Latin America, and sustainable development in frontier regions.
Following completion of his PhD Chris will be joining the Department of Economics and
Geography at the University of North Florida as an assistant professor. Chris has been married to
Ana Baynard for 11 years and they have two children, Nico and Isabela.