Citation
Vectors of dispossession : an assessment of land titling and deforestation in eastern Nicaragua

Material Information

Title:
Vectors of dispossession : an assessment of land titling and deforestation in eastern Nicaragua
Creator:
Rivera, Isaac Javier
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Geography and Environmental Sciences, CU Denver
Degree Disciplines:
Applied geography and geospatial sciences
Committee Chair:
Simon, Gregory
Committee Members:
Bryan, Joe
Moreno, Rafael

Notes

Abstract:
Awas Tingni marks the site of the historic Inter-American Court of Human Rights (IACHR) decision that issued legal recognition of property rights for the indigenous Mayagna (Sumo) community located in North Eastern Nicaragua. The titling of Awas Tingni marks a historical moment within international law that recognized that indigenous peoples had rights to property, and were therefore entitled to the human rights therein. Landownership and the right to private property was celebrated as a fundamental achievement in the progression of indigenous peoples’ rights by the Bretton Woods system that argued that private property was the best mechanism to defend indigenous lands and human rights. Titled officially in 2008, Awas Tingni, and North Eastern Nicaragua more broadly has faced an explosion of conflict within newly titled spaces, resulting in spatially specific forest loss. This study asks, what is the relationship between land titling processes and forest cover change? What is the spatial and temporal distribution of forest loss within titled Awas Tingni, Nicaragua? Using the open source Global Forest Cover (GFC) dataset, this study employs remote sensing analysis to distinguish the spatiality of forest loss between the years 2000 – 2016. Additionally, this study maps before, during, and after title forest loss for Awas Tingni and Eastern Nicaragua more broadly. Finally, the results of this study suggest that the legal administrative properties of land titles do not ensure that forests are protected; rather they reveal that the formalization of land titles only extends state governance and its priorities to newly titled and recognized spaces, warranting further research at the nexus of political ecology and political geography.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Isaac Javier Rivera. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
VECTORS OF DISPOSSESSION: AN ASSESSMENT OF LAND TITLING AND DEFORESTATION IN EASTERN NICARAGUA
by
Isaac Javier Rivera
BA., Geography, University of Colorado, 2013
A thesis submitted to the Faculty of the Graduate School of the
University of Colorado in partial fulfillment of the requirements for the degree of
Master of Arts
Applied Geography and Geospatial Science Department of Geography and Environmental Science
2018


©2018
ISAAC JAVIER RIVERA ALL RIGHTS RESERVED
ii


This thesis for the Master of Arts in Applied Geography and Geospatial Sciences by
Isaac Javier Rivera
Has been approved for the
Department of Geography and Environmental Science
By
Gregory Simon, Chair Joe Bryan Rafael Moreno
Date: May 12, 2018


Vectors of Dispossession: An Assessment on Land Titling and Deforestation in Eastern
Nicaragua
Thesis directed by Associate Professor Gregory Simon ABSTRACT
Awas Tingni marks the site of the historic Inter-American Court of Human Rights (IACHR) decision that issued legal recognition of property rights for the indigenous Mayagna (Sumo) community located in North Eastern Nicaragua. The titling of Awas Tingni marks a historical moment within international law that recognized that indigenous peoples had rights to property, and were therefore entitled to the human rights therein. Landownership and the right to private property was celebrated as a fundamental achievement in the progression of indigenous peoples’ rights by the Bretton Woods system that argued that private property was the best mechanism to defend indigenous lands and human rights. Titled officially in 2008, Awas Tingni, and North Eastern Nicaragua more broadly has faced an explosion of conflict within newly titled spaces, resulting in spatially specific forest loss. This study asks, what is the relationship between land titling processes and forest cover change? What is the spatial and temporal distribution of forest loss within titled Awas Tingni, Nicaragua? Using the open source Global Forest Cover (GFC) dataset, this study employs remote sensing analysis to distinguish the spatiality of forest loss between the years 2000 - 2016. Additionally, this study maps before, during, and after title forest loss for Awas Tingni and Eastern Nicaragua more broadly. Finally, the results of this study suggest that the legal administrative properties of land titles do not ensure that forests are protected; rather they reveal that the formalization of land titles only extends state governance and its priorities to newly titled and recognized spaces, warranting further research at the nexus of political ecology and political geography.
IV


The form and content of this abstract are approved. I recommend its publication. Approved: Gregory Simon
v


This thesis is dedicated to the past and future students of color in the Department of Geography and Environmental Science, University of Colorado Denver. I hope to make them proud.
Si, se puede!
VI


ACKNOWLEDGMENTS
I want to thank Glenn Morris and Joe Bryan for their guidance these last few years. It is because of them that I am still in the academy, and pursuing a PhD.
I would not be in the academy if it were not for their support.
VII


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION AND LITERATURE REVIEW 1
Awas Tingni and the Territorial Turn...................................1
Mapping Forest Loss....................................................5
On Vectors.............................................................7
Open Source Remote Sensing.............................................9
What this study is not................................................10
II. METHODS...............................................................14
What is the relationship between land titles and forest cover change?.14
Global Forest Cover Datasets..........................................15
Land Title Dataset....................................................16
Algorithmic GIS.......................................................17
III. RESULTS AND INTERPRETATION 18
Results...............................................................18
Limitations...........................................................27
IV. NEXT STEPS............................................................28
REFERENCES.................................................................35
viii


LIST OF FIGURES
FIGURE
1.5 Study Area ........................................................................12
1.5 Forest Type .......................................................................13
2.4 Algorithm Flow Chart..............................................................17
3.1 2001 Forest Loss ..................................................................20
3.1 2007 Forest Loss ..................................................................21
3.1 2008 Forest Loss ..................................................................22
3.1 2010 Forest Loss ..................................................................23
3.1 2016 Forest Loss ..................................................................24
3.1 Pre-Title Forest Loss..............................................................25
3.1 Post-Title Forest Loss ............................................................26
3.1 Trend Line.........................................................................27
IX


X


CHAPTER 1
INTRODUCTION AND LITERATURE REVIEW
Vector n: A representation of the world using points, lines, and polygons (ESRI GIS Dictionary)
Dispossession n: Dispossession is often described as capitalism’s “original sin” - the moment when land is enclosed and made property, displacing those who previously inhabited it and making of them a class of people who depend on selling their labor for survival. As Marx made clear, it was a condition to think with and from as a means of gaining insight into the workings of capital. The idea has been widely critiqued and rethought, but the problem it poses endures (Joe Bryan, Dispossession Syllabus, 2015).
1.1 Awas Tingni and the Territorial Turn
Eastern Nicaragua consists of what indigenous Miskito insurgents’ call the autonomous regions, or more insistently, ‘Indian land.’ Divided between the North Caribbean Autonomous Region (RAAN) and the South Caribbean Coast Autonomous Region (RACS), the regions are composed of numerous overlapping land claims that have resulted in a myriad of political conflicts between the Nicaraguan state, land administrators, Afro-decedents and indigenous peoples, and logging corporations (Bryan, 2012; Joel Wainwright & Bryan, 2009). Conflict over land, particularly after the defeat of the Sandinista government in the 1990’s to U.S. backed presidential candidate resulted in a scramble over rights to territory, particularly as indigenous peoples returned home from war time status. In many cases, from refugee camps located throughout Nicaraguan cities. Led by Miskito insurgents who ran on an anti-colonial platform to secure rights to territory, land titling became a significant tool in which to territorialize1 their wishes for greater land and resource autonomy, and racial equality, though not without its
11 take David Delaney’s articulation that common territory invokes an assumption that territory promotes peace by defining clear ‘sovereign’ boundaries.
1


limitations, or implications on both land and peoples (Bryan, 2007; Nietschmann, 1995; Wainwright & Bryan, 2009). This historical conjuncture for legal recognition in Awas Tingni is only a small part of a larger trend in Latin America that adopted neoliberal reforms as a tool for advocating indigenous peoples’ collective rights to land and resources (Larson, Cronkleton, Barry, & Pacheco, 2008; Offen, 2003). Termed the “territorial turn,” the World Bank has heavily advocated for indigenous peoples’ and Afro-decedents to obtain collective rights to property, and therefore foster legal governance over land (Bryan, 2012). Territory in the eyes of the state implies a jurisdictional space, a space effectively denoting legal jurisdiction for the production of territory through the private property regimes (Delaney, 2005; Hale, 2005).
The Nicaraguan state, however, had refused to recognize the autonomy of the Miskito and Mayangna (Sumo) communities, insisting that the land belonged to the state, and was therefore under state jurisdiction, and therefore subject to state governance over lands and resources. Moreover, I contend, to understand the drivers of forest cover change in Eastern Nicaragua, one must first reckon with the spatial history of the region, as the historical material conditions of past and present space (Carter, 2010; Harvey, 1984; Mann, 2009). Though this study’s primary intellectual contribution to this literature is a remote sensing analysis for the years 2000 - 2016 to understand the materiality and geography of forest cover, what came before is just as important, and arguably more given the historical conjucture of land titling economic history, human rights and rights to property, and their implications on land cover change (Hecht, 2014; Wainwright, Jiang, & Liu, 2013; Wainwright, Jiang, & Liu, 2014). Spatial history in this context helps us understand the time and space dynamics of social environmental change, and in particular to leverage both economic and environmental data to illuminate the production of space (Simon, 2014).
2


The case ofAwas Tingni us. Nicaragua (2001) marks a significant historical moment in international law and indigenous peoples’ rights that continues to facilitate and foster a myriad of research programs dedicated at understanding modem property law, and human rights (Anaya, 2004; Anaya & Grossman, 2002; Anaya & Williams Jr, 2001). More importantly, the case continues to have wide repercussions for peoples living within Awas Tingni, and their natural resources (Blomley, 2003; Bryan, 201 l;Wainwright & Bryan, 2009). Beginning as a petition to the Inter-American Commission on Human Rights (IACHR), the case marked the first legal recognition of indigenous peoples’ rights to property, as a matter of basic human rights (Anaya & Williams Jr, 2001). The case rebuked the Nicaraguan state for not recognizing the property rights of the communities of Awas Tingni, and recommended as a result, that they establish a legal system to produce a state procedure for the granting of titles to community lands (Anaya & Grossman, 2002). Awas Tingni in the years prior to the case collaborated with both James Anaya and the World Wildlife Federation (WWF) for the development of maps to secure their territorial claims in the courts, but also to begin investigating the impact of illegal logging within the area (Wainwright & Bryan, 2009).
In collaborating with the WWF, creating a sustainable timber industry that recognizes indigenous property rights was instrumental to what came after. The Nicaraguan state sold a timber concession to a Korean-financed company called SOLC ARSA, in lands claimed by the communities, and located just south of the WWF assisted land claim called MADESNA. The IACHR court case thereafter ruled that the Nicaraguan state violated the property rights of Awas Tingni by granting a concession to SOLC ARSA. What is significant here in regards to understanding the materiality of land titles lies in how the Court distinguished property, and rights to territorial and legal possession of land (Wainwright & Bryan, 2009). The Court ruled
3


that ‘customary practices’ alone were enough to establish rights of possession, or in other words, through economic utility, land becomes legally titled and is therefore removed from a collective state to officially possessed and privatized land (De Soto, 2000; Locke, 1689; Joel Wainwright, 2009; Ybarra, 2009). As a result, the state of Nicaragua would still be the granter of titles as well as the administrative envoy that would be in charge of granting and legalizing demarcations. Qualitative and quantitative evidence in Honduras, however, has revealed increased unrest over competing claims, resulting in violence in some cases between the Miskito peoples and Icidino colonos (colonists), as well as signs of increased logging, suggesting that similar patterns may also materialize in North East Nicaragua (Mollett, 2011, 2013).
Though Awas Tingni won the Court case, it has only been described as a ‘moral victory’ in contemporary discourse, for the state still has the ‘real’ sovereign power here (Wainwright & Bryan, 2009). It continues to view Awas Tingni as an extension of the state, rather than Sumo territory. Joel Wainwright and Joe Bryan note that the power relations within the recognition of land have only increased the territorial governance of the state that amounts to what they term an ‘aporia\ to explain the contradictions and limitations of land titling (Wainwright & Bryan, 2009). Land titles when mapped help the state align indigenous livelihoods and claims with state institutions, as well as state political economic policies. Since Awas Tingni was titled in 2008, conflict, deforestation, and the precariousness of peoples and claims have only increased. International Law, and international recognition in particular, has only increased the necessity of indigenous peoples to prepare for what continues to be a constant negotiation of rights with the nation-state. Moreover, this study contributes to literature in economic history, international law, and political ecology. Or more specifically, this study is located where political ecology meets political geography. When land titles and their genealogies meet the materiality of land cover
4


change. Land title genealogies in this case opens up space for thinking about how power and knowledge of private property discourse converge in space, and influences how land change science might as a result interpret land cover change (Munroe, McSweeney, Olson, & Mansfield,
2014) .
1.2 Mapping Forest Loss
This study examines the relationship of land titling and deforestation in Eastern Nicaragua by mapping the spatiality and temporality of forest loss. Literature on mapping forest loss, and land cover change more broadly has been facilitated by an explosion in interest in remote sensing science to capture land cover change over long periods of time. Remote sensing for the past 20 years has received considerable attention by ecologists, forest monitors, and conservationists for its utility to capture objective and scientific measurements of forest cover change (Aide et al., 2013; Hansen et al., 2013; Haywood, Alfonsetti, Ortmann, & Takawo,
2015) . For the last 40 years, forest clearing scholarship has centered mostly on biotic and climatic dimensions of forest cover change, that have included carbon stocks, precipitation measurement, and arctic ice measurement (Foley et al., 2005; Hansen, Stehman, & Potapov, 2010). With increasingly improving sensor capabilities, remote sensing technology continues to reach new breakthroughs in resolution that can now classify tree cover species, to facilitating the U.S. military in a vast network of global surveillance, and counter-insurgency (Wainwright,
2016) . However, though scientists and the public at large continue to receive considerably improved and more easily assessable remote sensing data through government institutions and open data platforms, consideration of the human components of forest cover change, and the conditions that facilitate that change, have typically fallen outside of remote sensing literature, and have fallen instead within the domain of agrarian change, political ecology, and research on
5


the human dimensions of climate change (Hecht, 2014; Turner & Robbins, 2008; J Wainwright et al., 2013).
Mapping forest loss using remote sensing analysis has been used by NGO’s, non-profits, governments, and communities, to produce studies for understanding where forest change is occurring, and what it means for ecosystems, economies, and peoples (Aide et al., 2013; McSweeney et al., 2014; Wegmann, Leutner, & Dech, 2016; Wei et al., 2018). Only a few studies have attended to the regional and political economic dimensions that drive the material conditions for forest loss in Latin America, making questions of integration between the economic and physical all the more urgent (Hecht, 2014; Lambin et al., 2001; McSweeney et al., 2014; J Wainwright et al., 2013; Ybarra, 2017). This study draws from Joel Wainwright et al. (2014) mixed method approach, which integrates both remote sensing analysis with regional political economic policies that influences farming choices and methods, logging practices, and land use change more broadly (Joel Wainwright et al., 2014). Regional political economy in this case includes not just the economic policies that help distinguish and produce the space for logging (e.g., land titles, legal permits, state economic rationale), but also the road networks involved that were used to help create the material conditions for logging (Laurance, Goosem, & Laurance, 2009; Lefebvre, 2003; T. Mitchell, 2007; Newbold et al., 2015). Logging in Eastern Nicaragua in particular, has taken the form of legal logging by both state and international corporations, but also an unintended influx of illegal black markets that have developed at the same time (Joel Wainwright & Bryan, 2009). The focus of this study is to understand the spatial distribution of forest loss in its totality in relation to land titles, rather than develop a classification scheme to understand the particular logging concessions involved, or the ability of remote sensing to capture that change or not. This study draws upon open source remote sensing
6


datasets, and in particular, the open source Hansen Global Forest Cover dataset (Hansen et al., 2013).
1.3 On Vectors
Vectors, in a purely GIS sense are defined as points, lines, and polygons, which make up the material-components of world representation. In this study, vectors imply specifically to land titles, roads, over-lapping claims represented as polygons, colono sites, and bounded spaces of the dispossessed, or desmovilizados who have attempted to reclaim space. Additionally, this study attempts to open up space to think about vectors beyond their naturalized state on maps, for a reading of vector representation outside of contemporary political economic discourse, and outside there seemingly natural or neutral appearance. Moreover, I am particularly interested in the process of how vectors become both naturalized within maps, but also in contemporary discourse more broadly, but more precisely with how such vectors—points, lines, and polygons, and their meanings come to be both produced and managed (Crampton & Krygier, 2005; Pickles, 2004).
From land titles, to roads, each has a unique history, and a particular vectorization process in the social imaginary that render them visible or invisible (K. Mitchell & Elwood, 2016). Vectorization within the tradition of the critique of political economy can also be understood, I argue, at the intersection of economic geography and land change science (Munroe et al., 2014). Understanding the process and formalization of land titles, throughout space and time can be understood as a process of vectorization. Therefore, recognizing the material consequences on land and peoples through vectorization can help reveal how space is produced, and therefore subject space to enclosures for particular commodities, but also ontological
7


understandings of that space before private property (Goeman, 2013; Mezzadra & Neilson, 2013; Offen, 2003). Dispossession in this case is both the material and im-material understandings of land cover, or in this particular case, the literal loss of a forest.
Drawing from McKenzie Wark’s reading on vectorialism, and historical materialism more broadly, I contend for a reading of land titles that denaturalizes its intended political economic, and human rights assumptions within contemporary economic discourses (Harvey, 1984; Mann, 2009; Marx, 1867; Joel Wainwright & Bryan, 2009; Wark, 2004). More specifically, I argue for a reading of vectors that calls into question the ‘worlding’ material properties of land titles, the material properties of land titles being judicial recognition, spatial territorialtilization, and state extensions for governing through land titles (Joel Wainwright,
2009; Joel Wainwright & Bryan, 2009; Ybarra, 2009). Land titles when mapped in space connote a particular reading of space that was not there previously before, making the act of mapping, much like the process of formalizing land titles, a powerful tool in which no explanation is given to their convergence in space, or their material implications on it. Vectorization as a result is fundamentally connected to the process of mapping, particularly through GIS infrastructure, and through the formalization process of land titles that the state relies on to draw private property lines (Dwyer, 2015; Mezzadra & Neilson, 2013; Wilson,
2017).
Though this study will not particularly engage with the history of land titles per se, nor develop an analytic for thinking about vectors, nor assess their political economic implications on regional political geography, this study is intended to begin an albeit small contribution within literature dealing with the material implications of land titles, and private property more broadly by assessing forest loss within these spaces. Moreover, by pushing vectors beyond their
8


naturalized and neutral appearance on maps, I aim to open up space for thinking about vectors in much the same way that a cartographer or GISer might think about vector data. But further, by linking their GIS epistemology to liberal economic philosophy and their philosophers in the past and present moment, that have argued that private property, in this case the bounded vector land title, are intimately purposed for the protection of both land and human rights. I critique these hegemonic and naturalized assumptions through a critique of political economy. This merger of epistemology and economic philosophy, particularly the critique of political economy given the quantitative and qualitative evidence provided within this study, can be understood, I argue, as vectors of dispossession.
1.4 Open Source Remote Sensing
With remote sensing technology becoming ever more accessible, and more useful in regards to understanding land cover change, and capturing land cover trends over time, researchers in agrarian studies and political ecology have begun to integrate their scientific methods with promising results (Wegmann et al., 2016). This study draws from the Hansen et al (2013) dataset on publically available global forest cover for integration of forest cover change data, with political economy, or in this particular case—land titles. The global forest cover dataset at the time of its release was hailed as a fundamental step into democratizing remote sensing research, and environmental science more broadly, for its publically and freely available data, made possible by the Landsat program at NASA and the USGS (Hansen et al., 2013). It has become the staple dataset used by forest monitors, non-profits, NGO’s, and governments to assess forest cover. Though with limitations in regards to understanding particular land use
9


change at a finer scale of analysis, research has confirmed its accuracy at regional and global scales of analysis (Gonzalez et al., 2016; MacDicken, 2015).
Open source remote sensing, however, does not bridge the gap in the uneven geography of knowledge production on localized environmental expertise, it has, however, helped expand its usage across academic and non-academic circles, resulting in increased usage and demand (Elwood, 2008; Haraway, 1988; Lave, 2015). The Landsat program represents four decades worth of high resolution images acquired from Landsat satellites. This particular study uses the global forest cover dataset from Hansen et al (2013) between the years (2000 - 2016), drawing specifically from images acquired from Landsat 7-ETM+ and Landsat 8 OLI.
For greater transparency, replication, and integration of datasets, this study employs algorithmic GIS science to facilitate a multi-step analysis of both Landsat and land title datasets. More specifically, this study generates an algorithm for conducting remote sensing analysis of Landsat 7 and 8 images from the years (2000-2016) for the quantification of forest loss for before, during, and after title. Literature in open source remote sensing is increasingly drawn to the potential of algorithmic GIS science to conduct both complex analysis across different data types and data structures, and high memory demand inquiries, in a short amount of time, while at the same time providing new avenues for the public sharing of base code for interested institutions and peoples (Christophe & Inglada, 2009). The utility in using algorithms in my view is centered on the ability to share base code with communities that may find it useful for their own forest/title assessments, as well as the ability to produce both datasets and code for the replication of a similar study in another space. This algorithm, which I call vectors of dispossession will be uploaded to Github for public use.
10


1.5 What this study is not
This study’s main analytical and intellectual contribution is to help distinguish the spatiality and temporality of forest loss within titled spaces in Eastern Nicaragua, or more specifically, Awas Tingni. This study is not a forest type classification study, nor an analysis of the difference between Landsat 7 and Landsat 8 satellite resolution capture benefits, nor an assessment of the utility of open source remote sensing to conduct automated remote sensing analysis. Though the study could be viewed in relations to these not’s, its purpose rather is to add inquiry and empirical data that helps us understand the political ecology, and political geography of land tenure formalization schemes across Nicaragua, and other geographies around the world, but particularly as the World Bank continues to insist on that land tenure; private property; as the answer to unsustainability and human right’s violations (Dwyer, 2013, 2015; Watts, 2003). The recent debates in the Proceedings of the National Academy of Sciences further reflect the need for inquiry at the intersections between land tenure, forest cover change, and human rights (Blackman, Corral, Lima, & Asner, 2017; Robinson, Holland, & Naughton-Treves, 2017). It is our responsibility as geographers to make sense of these competing discourses that continue to foster world-changing polities, resulting in human-economic policies that have not addressed their aporias and contradictions within (Massey, 2004).
11


*(S)
Indigenous Territories (Titled) StudyArea During Title
2008
Forest Loss Year
| No Loss 2001 2002 | l 2003
] 2004 â–¡ 2005 | 2006 | 2007 | 2008 ] 2009
l 2010 I I 2011
m 2012
2013 | 2014 2015 â–  2016
(Figure 1. Study Area. Source, Author)
Forest Loss 2001 - 2016 data derived from Hansen et al. (2013), titled data provided by the National Commission of Titles and Demarcation of Nicaragua. Title data was reformatted for algorithm usage.
12


REGIONES ECOLbGICAS, ECOSISTEMAS Y BIODIVERSIDAD
(Figure 2. Forest Type. Source, The National Governance and Reconciliation Center of Nicaragua)
Figure 2 depicts bio-diversity, eco-systems, and ecological regions for the entirety of Nicaragua. Forest type in Eastern Nicaragua mostly consists of open forest, broad leaf forest, and pine savanna.
13


CHAPTER 2
METHODS AND METHODOLOGIES
2.1 What is the relationship between land titles andforest cover change?
Spatiality and Temporality
The purpose of this study is to better understand the spatiality and temporality of forest loss in North Eastern Nicaragua. Spatiality is of particular importance due to the spatial extent of land titles both locally in Awas Tingni, but also in regards to the more regionally specific trends that political economic policy facilitates, particularly as it relates to land management. Spatiality in this case is mapped at both regional and local scales. Regional scale in this case includes all land titles in the RAAN that run up to the border with Honduras in the North, and the Atlantic Ocean in the East. Local scale in this case focuses particularly on the Awas Tingni claim. Focusing at the scale of the local allows for a better understanding of how governance takes shape over space. In addition to mapping forest loss and land titles, this study will also map local roads, overlapping claims, and colonos, or in other words, the vectors undergoing inquiry in this study.
14


2.2 Global Forest Cover Datasets
Landsat Datasets
Landsat data was acquired from version 1.4 of the Global Forest Change dataset for the years 2000-2016. Landsat 7 (Enhanced Thematic Mapper Plus) ETM+ data is composed of a composite of yearly images between the years 2000-2013. Landsat 8 Operational Land Imager (OLM) and Thermal Infrared Sensor (TIRS) is composed of a composite of yearly images between the years 2013-2016. Landsat 7 ETM+ operates at a 30-meter spatial resolution. Landsat 8 OLM/TIRS also operates at a 30-meter spatial resolution but has an additional two bands that give it added seasonal precision and cloud removal properties for added clarity. The TIRS sensor is particularly important because of its ability to capture additional images than the Landsat 7 ETM+, resulting in a higher probability that images will be cloud free, resulting in less distorted and higher quality images.
The Global Forest Change dataset includes data the was reprocessed post 2011 adding to improved loss measurements, improved training data for calibrating the loss model, improved per sensor quality assessment models to filter input data, and improved input spectral features for building and applying the loss model. Data comparisons between the years 2001-2016 are therefore limited given quality differences across the dataset (Hansen et al., 2013). The dataset, however, is still useful in that it provides us an initial window for understanding the spatiality and temporality of forest loss. The dataset’s most useful contribution involves a prepackaged and classified dataset that is ready for user analysis. Data preparation was insignificant, for the
15


exception of clipping Landsat images to desired study area. The dataset is divided into 10x10 degree tiles, consisting of 8 bit values that have a spatial resolution of 1 arc-second per pixel, or about 30 meters per pixel at the equator. Forest loss during the period 2000-2016 is defined as a stand-replacement disturbance, or a change from a forest to non-forest state. Data is encoded as 0 for (no loss), values with a range of 1-16 correspond with the forest loss years for the years 2001-2016, respectively.
2.3 Land Title Dataset
Marcos Williamson of the Comision Nacional de Demarcacion y Titulacion, or CONADETI, compiled all land title GIS data using data from the center. The land title dataset consists of 29 polygons for all titled indigenous territories in Eastern Nicaragua. Date of title for all territories was formatted to only account for year of loss, rather than exact date of title. This allowed process allowed for easier integration with array data structures, for a more structured algorithm. Territories throughout the dataset are represented as polygons, and exist in this dataset as a shape file, or .shp file. All attribute data, including name of territories, and date of title remained intact post-processing.
2.4 Algorithmic GIS
Algorithmic GIS in recent years has grown due to demand for its utility to conduct highly effective, multi-layered analysis, and producing automation schemes that can save considerable amount of labor time to produce the same result. For future scale up of analysis, replication, and
16


sharing of particular functions, an algorithm was built to address the core research question of this study: what is the relationship between forest loss and land titles?
The inputs for this algorithm are as follows: 1) Hansen Global Forest Cover Raster derived from version 1.4 of dataset and 2) Land Title dataset derived the National Commission of Demarcation and Titling, of Nicaragua. Following inputs, the data was extracted by mask, clipped by raster, and converted into an array. An array is a data-structure that allows data to be converted into an index for easier processing, and easier outputs. Unprocessed Array is then converted into a subset of arrays that correspond to each year of titling and for all land title polygons. The subset is then converted into a list of arrays of ‘before’ titling, ‘during’ titling, and ‘after’ titling. Forest area loss is then calculated by the equation (counts * cell size) as a list of arrays. Arrays are then converted into raster’s for mapping visualization. Forrest loss area calculations are then exported into a final csv file for each year loss for all titled polygons through an open source process called ‘panda’s data-frame.’ Panda’s data-frames allow for easier indexing of arrays for easier export to csv.
17


Algorithm Flow Chart
Figure 3. Algorithm Flow Chart. Source, Author
This chart is intended to simplify and map out the multi-step process for conducting the core analytical contribution of this study. Additionally, it is also intended to demystify and de-fetishize algorithmic GIScience more broadly. Doing so in my view could greatly increase the usage and approachability of programming by those who may doubt their abilities to do GIS programming. Though far from perfect, this flow chart is designed to help conceptualize the logic behind the algorithm.
18


CHAPTER 3
RESULTS AND INTERPRETATION
Algorithm deliverables including all rasters of forest loss, mosaics of forest loss representing forest loss for before, during, and after title, were used to construct Figures 3-9. Cartography, and the mapping of forest loss in relation to land titles, overlapping claims, roads, and colonos were all created outside the algorithm.
19


Rivers
Awas Tingm Claim
A Colonos
Miskito Excombatants
Colectivo El Pro
Colectivo Tungla
Overlapping Community Claims
10 Comunidades
â–¡ Francia sirpi | Sta Clara | Desmovilizados
Forest Loss
Figure 4. 2001 Forest Loss. Source, Author
Figure 4 reveals very light forest loss throughout the Awas Tingni claim. Most of the forest loss is located just south east of the claim.
20


Rivers
Colonos
Huricane Felix Path
Colectivo El Pro
10 Comunidades
Awas Tingni Claim
Miskito Excombatants
Colectivo Tungla
Overlapping Community Claims
â–¡ Francia sirpi
â–¡ Sta Clara
| Desmovilizados
Forest Loss
Figure 5. Forest Loss 2007. Source, Author
Hurricane Felix hit Awas Tingni, Nicaragua on September 7, 2007. This figure reveals heavy forest loss clustering just South of the Hurricane Felix path. Due to increased cloud cover, and a faulty mirror in Landsat 7, forest loss distribution is distorted, and of less quality than Figure 3. This map should be read to understand the general spatial trend in forest loss for this year. All images for each year represent the sum amount of forest loss for any particular year.
21


Rivers
2008
A Colonos
â–¡ Indigenous Territories (Titled) Miskito Excombatants
| Colectivo El Pro Colectivo Tungla
Overlapping Community Claims
10 Comunidades Esperanza
â–¡ Francia sirpi
â–¡ Sta Clara
â–¡ Desmovilizados Forest Loss
Figure 6. 2008 Forest Loss. Source, Author
Figure 6 depicts very light forest loss. This is also the first official year that Awas Tingni was titled. This is also the first year after the devastation of Hurricane Felix.
22


Roads
Rivers
A Colonos
| Indigenous Territories (Titled) Miskito Excombatants Colectivo El Pro Colectivo Tungla
Overlapping Community Claims
10 Comunidades
Esperanza
___] Francia sirpi
â–¡ Sta Clara
| Desmovilizados
2010
Figure 7. 2010 Forest Loss. Source, Author
Figure 7 reveals forest loss clustering and a new peak in forest loss for the region. The clustering, however, is located primarily due South of titled Awas Tingni.
23


Roads
Rivers
Colonos
| Indigenous Territories (Titled)
Miskito Excombatants
Colectivo El Pro
Colectivo Tungla
Overlapping Community Claims
10 Comumdades
â–¡ Francia sirpi
___j Sta Clara
| Desmovilizados
Forest Loss
Figure 8. 2016 Forest Loss. Source, Author
Figure 8 reveals another peak in forest loss. Forest loss clustering in this case is located within titled Awas Tingni and just South of the title. Forest loss clustering in the South is in proximity to road networks, indicating that there may be correlation with forest loss and road network expansion. Within the title, forest loss clustering is primarily located east of the Sta Clara claim Clustering of forest loss within the Miskito Excombatants claim, and the site of Colonos, also reveal peaks of forest loss.
24


Roads
Rivers
No Loss
2002
2004
2005
2006
2007
Colectivo El Pro
10 Comunidades
â–¡ Indigenous Territories (Titled)
A Colonos Forest Loss
Miskito Excombatants
Colectivo Tungla
Overlapping Community Claims
Esperanza
â–¡ Francia sirpi
â–¡ Sta Clara
â–¡ Oesmovilizados
Figure 9. Forest Loss Pre-Title. Source, Author
Figure 9 reveals forest loss clustering primarily West of the Colectivo El Pro. Clustering however should be taken with a degree of caution given distorted images coming from the years 2005 and 2007.
25


| Indigenous Territories (Titled)
------ Roads
Rivers
2013
2014
2015
A Colonos Forest Loss
Miskito Excombatants Colectivo El Pro Colectivo Tungla
Overlapping Community Claims
10 Comunidades
Francia sirpi | Sta Clara | Desmovilizados
Figure 10. Forest Loss Post-Title. Source, Author
Figure 10 reveals forest loss after title to be primarily located just east of the Colectivo El Pro. Whereas pre-title forest loss is primarily located in the West. Road networks, and overlapping community claims are also primarily located in the mid to east sections of the Awas Tingni.
26


FOREST LOSS (MEETERS SQ)
Awas Tingni Claim Forest Loss (2001 -2016)
70000000 60000000 50000000 40000000 30000000 20000000 10000000 0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
LOSS YEAR
Figure 11 reveals a positive trend in forest loss throughout the Awas Tingni Claim over time. We see peaks in 2007, 2014, and 2016. Peak in 2007 is associated with Hurricane Felix. Peaks in 2014, and 2016 will be followed up with spatially specific qualitative data during the “next steps” of this study.
27


3.1 Results
In summary, the relationship between land titling and forest loss was shown to have a positive relationship. The brief history of the territorial turn, the algorithm as built, and the maps that came after the algorithm process, are all but smaller components that could each on their own feed into their own master’s thesis, or even dissertation’s. The utility of geography in this regard is that it is expansive enough to weave seemingly separate epistemological and ontological frameworks, while each having their own distinct histories that are woven that materialize in space. This study attempts to thread two seemingly separate and radically different data worlds onto themselves; that of forest cover and land title history for an improved understanding of the legal and political limitations of land titles to defend both land and human rights.
From the spatial analysis, there is a clear indication that forest loss begins very minimally in Awas Tingni in 2001 (Figure 3), following with rapid increase in 2007 (Figure 4). The rapid increase in forest loss is likely due to a Category 5 Hurricane Felix, which made a direct hit on Awas Tingni on September 7, 2007. In 2008, the first official year of titled Awas Tingni, we see most of the deforestation just south of the boundary (Figure 5). In 2010, we see another spike in deforestation, though most of that deforestation is just south of Awas Tingni (Figure 6). In 2016, we see a major spike throughout Eastern Awas Tingni (Figure 7). Much of the forest loss is located just east of the overlapping claims, and closer to the road networks that were mostly built after 2011. Both Miskito Excombatants claims, with Awas Tingni also show a significant amount of forest loss in 2016. The overall trend points towards a gradual increase in forest loss with peaks in both 2007, 2011, and 2016 for both Awas Tingni, and surrounding titled and untitled spaces. Given that most roads in this network were built after 2011, there may be a
28


correlation with road development, and forest loss, particularly as regional markets for logging concessions have increased, along with illegal black markets at the same time.
Forest loss for pre-title (Figure 8) reveals clustering of forest loss in proximity to colonos in western Awas Tingni. Clustering of forest loss, however, is slightly distorted due to less input data into the Hansen dataset. This is due to Landsat 7 ETM+ mirror failure that caused significant images to be loss. Forest loss for post-title (Figure 9) reveals spikes in forest clustering both within Awas Tingni and just south of the titled boundaries. Clustering of forest loss is particularly spiked in the years 2011, and 2016. Forest loss post-title is mostly concentrated in the mid to east boundaries of Awas Tingni, and particularly in proximity to newly built roads. Most of the roads within Awas Tingni were built after 2010. Due to lacking road data past 2012, it is likely that road networks have only expanded with time into various clustered forest loss locations.
29


3.2 Limitations
The reprocessed data from the Global Forest Cover dataset from 2011 onward, the use of Landsat 8 OLI data for 2013 onward, improved training data for calibrating the loss model, and improved input spectral features for building and applying the loss model for data 2011 onward, produced significant variability across the 2000-2016 dataset. Figures 3, 4, 5, and 6 reveal forest loss with less resolution clarity, and quality then figures with forest data from 2011 onward. Forest loss area calculations, as a result, are therefore unreliable in this case, and cannot be used to account for precise calculations of forest loss.
Additionally, maps of forest loss before 2011, display data that appear slightly pixelated, resulting in maps before 2011 to appear slightly course in some instances. Forest loss in 2007 in particular, which was also the year of Hurricane Felix, reveals clustered concentrations of forest loss, but at a lesser quality due to both increased cloud cover, and a faulty mirror in the Landsat 7 ETM+ sensor. Maps before 2011 should be read with some degree of caution. They are, however, still useful for understanding the spatiality of forest loss. Figure 3, 2001 loss in particular captures forest loss to a finer degree than forest loss in 2007 (Figure 4). This can be attributed to the gaps of data of Landsat 7 images for the year 2007, which are attributed to a faulty mirror on the Landsat sensor.
30


CHAPTER 4
NEXT STEPS
4.1 Future Remote Sensing Analysis
What comes next in this study depends on ‘our’ ability to address the seemingly disconnected geographies of algorithmic GIS science, remote sensing, and political economy.
For algorithmic GIS science and remote sensing, and for the algorithm built for this particular study, its success will depend on how the non-trained GISer, or non-coder can insert updated remote sensing datasets onto its base code. Version 2.0 of the Hansen dataset is currently in the works, and should be released at the end of the year, giving this study a clear follow up in both accessibility, and adjustment using higher quality datasets. Moreover, updating the forest loss dataset will be essential for conducting accurate forest loss counts across the entirety of the dataset. Additionally, finding a clean cloud free World View data set image would allow for the digitizing of the most current road networks. Having updated land cover data, forest cover data, and road data will greatly improve future results, and the ability to use such data in conversation with the qualitative.
With improved datasets, remote sensing analysis can begin to identify the particular forest types that are being loss. With forest loss being broadly coded as “any forest disturbance” throughout the Hansen et al. dataset, identifying what exactly those disturbances are can better improve our understanding of the particular commodities involved in land use change. Through a series of classification schemes of forest and land cover type, future remote sensing analysis could greatly benefit from taking a greater account of the heterogeneity of land cover in Awas Tingni and Eastern Nicaragua more broadly. More so, identifying particular commodities can
31


also feed into the broader interconnections between political economy and local governance of land, resulting in new mechanisms ways to understand the political ecology of land titling systems.
4.2 Genealogies of Knowledge Production
Furthermore, addressing questions regarding ethics in knowledge production on this frontier can benefit from an approach that emphasizes a co-production of citizen science for conducting future analysis (Hacker, 2013; Lave, 2015). For instance, breaking free from the global North to global South flows of knowledge production that so much of open source and participatory mapping facilitates can further benefit from the co-production approaches of citizen science for addressing not just gaps in particular GIS methods, but also enabling a ground truth team to further strengthen the claims implied by land cover maps depicting forest loss on indigenous lands (Livingstone, 1995, 2010; Spivak, 1999). But also to strengthen a process in which the community, rather than the ‘researcher’ at a University in the global North interprets the meaning and implications of that forest loss (Massey, 2004; Spivak, 1999). What does it mean for indigenous peoples to see a forest stripped from its former, perhaps original standing? What are the ontological implications for indigenous peoples’ lives beyond the ecological ripple effect across commoditized ecologies of forest loss?
In short, I believe that integrating both ethnographies of forest lost and political economy could be quite useful. Additionally, integrating the results from this study with additional qualitative data from Awas Tingni, and the broader surrounding areas can further sharpen the claims and results made in this study. The qualitative data conducted by Joe Bryan in particular, when in conversation with the maps presented in this study will be able to produce yet new
32


geographies of understanding regarding the role of the state, maps, and the law in facilitating land cover change (Bryan, 2007). Questions within the traditions of political ecology, and political geography will as a result be enriched by the integration of qualitative and quantitative data, resulting in new ways to understand the political economy, legal geography, and the drivers of land cover change more broadly.
4.3 Vectors of Dispossession
Moreover, in drawing from historical materialist analytics that engage with questions of indigenous ways of knowing and remote sensing analysis, I propose a future study to further develop the idea and implications for thinking through the concept of vectors of dispossession. In short, I propose to develop an analytic that draws from historical materialism to better understand the spatial, political, and material dimensions of land titling. I call this merger of epistemological and ontological frameworks—vectors of dispossession. Doing so, in my view can better reveal the totalizing colonization of everyday life to address both inequities in methodologies, but also in how we ultimately understand the material and ontological dimensions of dispossession (Lefebvre, 2004; Monte-Mor, 2014). Drawing from literature in critical GIS, historical materialism, and decolonial studies, I will attempt to build upon this study’s success, and limitations, for pushing geography and geographers to reckon with how ‘we’ do geography (Coulthard, 2014; Mariategui, 1971; Springer, 2017; Webber, 2011; Wilson, 2017). In the words of Simon Springer, “what goes on in geography is, after all, up to us” (Springer, 2017). Connecting seemingly unrelated and distinct literatures that engage radically different epistemologies can produce new spaces for thinking about questions of dispossession.
33


Vectors of Dispossession are not particular to Eastern Nicaragua by any means, rather, they are the material analytics of colonized everyday life.
34


REFERENCES
Aide, T. M., Clark, M. L., Grau, H. R., Lopez- Carr, D., Levy, M. A., Redo, D., . . . Muniz, M.
(2013). Deforestation and reforestation of Latin America and the Caribbean (2001-2010). Biotropica, 45(2), 262-271.
Anaya, S. J. (2004). International Human Rights and Indigenous Peoples: The move toward the multicultural state. Ariz. J. Int'l & Comp. L., 21, 13.
Anaya, S. J., & Grossman, C. (2002). The Case of Awas Tingni v. Nicaragua: A Step in the International Law of Indigenous Peoples. Ariz. J. Int'l & Comp. L., 19, 1.
Anaya, S. J., & Williams Jr, R. A. (2001). The protection of indigenous peoples' rights over
lands and natural resources under the Inter-American human rights system. Harv. Hum. Rts. J., 14,33.
Blackman, A., Corral, L., Lima, E. S., & Asner, G. P. (2017). Titling indigenous communities protects forests in the Peruvian Amazon. Proceedings of the National Academy of Sciences, 201603290.
Blomley, N. (2003). Law, property, and the geography of violence: The frontier, the survey, and the grid. Annals of the Association of American Geographers, 93(1), 121-141.
Bryan, J. (2007). Map or Be Mapped.
Bryan, J. (2011). Walking the line: participatory mapping, indigenous rights, and neoliberalism. Geoforum, 42(1), 40-50.
Bryan, J. (2012). Rethinking territory: social justice and neoliberalism in Latin America’s territorial turn. Geography Compass, 6(4), 215-226.
Carter, P. (2010). The road to Botany Bay: An exploration of landscape and history. U of Minnesota Press.
Christophe, E., & Inglada, J. (2009). Open source remote sensing: Increasing the usability of cutting-edge algorithms. IEEE Geoscience and Remote Sensing Newsletter, 35(5), 9-15.
Coulthard, G. S. (2014). Red skin, white masks: Rejecting the colonial politics of recognition: Taylor & Francis.
Crampton, J. W., & Krygier, J. (2005). An introduction to critical cartography. ACME: An International Journal for Critical Geographies, 4(1), 11-33.
De Soto, H. (2000). The mystery of capital: Why capitalism succeeds in the West and fails everywhere else: New York: Basic Books.
Delaney, D. (2005). Entering the territory of territory. Territory: A Short Introduction, 1-33.
Dwyer, M. B. (2013). Building the politics machine: tools for ‘resolving’the global land grab. Development and Change, 44(2), 309-333.
Dwyer, M. B. (2015). The formalization fix? Land titling, land concessions and the politics of spatial transparency in Cambodia. The Journal of Peasant Studies, 42(5), 903-928.
Elwood, S. (2008). Volunteered geographic information: future research directions motivated by critical, participatory, and feminist GIS. GeoJournal, 72(3-4), 173-183.
Foley, J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R., . . . Gibbs, H. K. (2005). Global consequences of land use. Science, 309(5134), 570-574.
Goeman, M. (2013). Mark my words: Native women mapping our nations: U of Minnesota Press.
Gonzalez, A., Cardinale, B. J., Allington, G. R., Byrnes, J., Arthur Endsley, K., Brown, D. G., . .
. Loreau, M. (2016). Estimating local biodiversity change: a critique of papers claiming no net loss of local diversity. Ecology, 97(8), 1949-1960.
Hacker, K. (2013). Community-based participatory research: Sage Publications.
35


Hale, C. R. (2005). Neoliberal multiculturalism. PoLAR: political and legal anthropology review, 25(1), 10-19.
Hansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S., Tyukavina, A., . . . Loveland, T. (2013). High-resolution global maps of 21st-century forest cover change. Science, 342(6160), 850-853.
Hansen, M. C., Stehman, S. V., & Potapov, P. V. (2010). Quantification of global gross forest cover loss. Proceedings of the National Academy of Sciences, 107(19), 8650-8655.
Haraway, D. (1988). Situated knowledges: The science question in feminism and the privilege of partial perspective. Feminist studies, 575-599.
Harvey, D. (1984). On the history and present condition of geography: an historical materialist manifesto. The Professional Geographer, 36(1), 1-11.
Haywood, A., Alfonsetti, A., Ortmann, A., & Takawo, D. (2015). Improving national
greenhouse gas inventories for forestry and land use change using open-source software. Paper presented at the Geoscience and Remote Sensing Symposium (IGARSS), 2015 IEEE International.
Hecht, S. B. (2014). Forests lost and found in tropical Latin America: the woodland ‘green revolution \ Journal of Peasant Studies, 41(5), 877-909.
Lambin, E. F., Turner, B. L., Geist, H. J., Agbola, S. B., Angelsen, A., Bruce, J. W., . . . Folke,
C. (2001). The causes of land-use and land-cover change: moving beyond the myths. Global Environmental Change, 11(A), 261-269.
Larson, A. M., Cronkleton, P., Barry, D., & Pacheco, P. (2008). Tenure rights and beyond:
community access to forest resources in Latin America'. Center for international forestry research (CIFOR).
Laurance, W. F., Goosem, M., & Laurance, S. G. (2009). Impacts of roads and linear clearings on tropical forests. Trends in ecology & evolution, 24(12), 659-669.
Lave, R. (2015). The Future of Environmental Expertise. Annals of the Association of American Geographers, 105(2), 244-252.
Lefebvre, H. (2003). Space and the State. State/space: A reader, 84-100.
Lefebvre, H. (2004). The production of space, 1991. The City Cultures Reader(3).
Livingstone, D. N. (1995). The spaces of knowledge: contributions towards a historical
geography of science. Environment and Planning D: Society and space, 13(1), 5-34.
Livingstone, D. N. (2010). Putting science in its place: geographies of scientific knowledge: University of Chicago Press.
Locke, J. (1689). Second Treatise. Toronto: University of Toronto Press.
MacDicken, K. G. (2015). Global forest resources assessment 2015: What, why and how? Forest Ecology and Management, 352, 3-8.
Mann, G. (2009). Should political ecology be Marxist? A case for Gramsci’s historical materialism. Geoforum, 40(3), 335-344.
Mariategui, J. C. (1971). Seven interpretative essays on Peruvian reality '. University of Texas Press.
Marx, K. (1867). Capital, volume I: Harmondsworth: Penguin/New Left Review.
Massey, D. (2004). Geographies of responsibility. Geografiska Annaler: Series B, Human Geography, 86(1), 5-18.
McSweeney, K., Nielsen, E. A., Taylor, M. J., Wrathall, D. J., Pearson, Z., Wang, O., & Plumb, S. T. (2014). Drug policy as conservation policy: narco-deforestation. Science,
343(6110), 489-490.
36


Mezzadra, S., & Neilson, B. (2013). Border as Method, or, the Multiplication of Labor.
Mitchell, K., & Elwood, S. (2016). Counter-Mapping for Social Justice. Politics, Citizenship and Rights, 207-223.
Mitchell, T. (2007). The properties of markets. Do economists make markets, 244-275.
Mollett, S. (2011). Racial narratives: Miskito and colono land struggles in the Honduran Mosquitia. Cultural Geographies, 75(1), 43-62.
Mollett, S. (2013). Mapping deception: The politics of mapping Miskito and Garifuna space in Honduras. Annals of the Association of American Geographers, 103(5), 1227-1241.
Monte-Mor, R. (2014). Extended urbanization and settlement patterns in Brazil: an
environmental approach. Implosions/explosions: towards a study of planetary urbanization. Berlin: Jovis, 109-120.
Munroe, D. K., McSweeney, K., Olson, J. L., & Mansfield, B. (2014). Using economic geography to reinvigorate land-change science. Geoforum, 52, 12-21.
Newbold, T., Hudson, L. N., Hill, S. L., Contu, S., Lysenko, I., Senior, R. A., . . . Collen, B.
(2015). Global effects of land use on local terrestrial biodiversity. Nature, 520(1545), 45.
Nietschmann, B. (1995). Defending the Miskito reefs with maps and GPS: mapping with sail, scuba, and satellite. Cultural survival quarterly, 75(4), 34-37.
Offen, K. H. (2003). Narrating place and identity, or mapping Miskitu land claims in northeastern Nicaragua. Human organization, 62(4), 382-392.
Pickles, J. (2004). A history of spaces: Cartographic reason, mapping, and the geo-coded world: Psychology Press.
Povinelli, E. A. (2002). The cunning of recognition: Indigenous alterities and the making of Australian multiculturalism\ Duke University Press.
Robinson, B. E., Holland, M. B., & Naughton-Treves, L. (2017). Community land titles alone will not protect forests. Proceedings of the National Academy of Sciences, 201707787.
Simon, G. L. (2014). Vulnerability-in-production: a spatial history of nature, affluence, and fire in Oakland, California. Annals of the Association of American Geographers, 104(6), 1199-1221.
Spivak, G. (1999). A critique of postcolonial reason: toward a critique of the vanishing present: Cambridge, MA: Harvard University Press.
Springer, S. (2017). Earth writing. GeoHumanities, 1-19.
Turner, B. L., & Robbins, P. (2008). Land-change science and political ecology: Similarities, differences, and implications for sustainability science. Annual review of environment and resources, 33, 295-316.
Wainwright, J. (2009). ‘The first duties of persons living in a civilized community’: the Maya, the Church, and the colonial state in southern Belize. Journal of Historical Geography, 35(3), 428-450.
Wainwright, J. (2011). Decolonizing development: colonial power and the Maya (Vol. 36): John Wiley & Sons.
Wainwright, J., & Bryan, J. (2009). Cartography, territory, property: postcolonial reflections on indigenous counter-mapping in Nicaragua and Belize. Cultural Geographies, 16(2), 153-178.
Wainwright, J., Jiang, S., & Liu, D. (2013). Deforestation and the world-as-representation: the Maya forest of Southern Belize. Land Change Science, Political Ecology, and Sustainability: Synergies and Divergences, 169-190.
37


Wainwright, J., Jiang, S., & Liu, D. (2014). The Maya forest of Southern Belize. Land Change Science. Political Ecology and Sustainability: Synergies and Divergences, 169.
Wainwright, J. D. (2016). The US Military and Human Geography: Reflections on Our
Conjuncture. Annals of the American Association of Geographers, 106(3), 513-520.
Wark, M. (2004). A hacker manifesto.
Watts, M. (2003). Development and governmentality. Singapore Journal of Tropical Geography, 24(1), 6-34.
Webber, J. R. (2011). Red October: left-indigenous struggles in modern Bolivia (Vol. 29): Brill.
Wegmann, M., Leutner, B., & Dech, S. (2016). Remote sensing and GISfor ecologists: using open source software: Pelagic Publishing Ltd.
Wei, X., Li, Q., Zhang, M., Giles- Hansen, K., Liu, W., Fan, H., . . . Liu, S. (2018). Vegetation cover—another dominant factor in determining global water resources in forested regions. Global change biology, 24(2), 786-795.
Wilson, M. W. (2017). New Lines: critical GIS and the trouble of the map.
Ybarra, M. (2009). Violent visions of an ownership society: The land administration project in Peten, Guatemala. Land Use Policy, 26(1), 44-54.
Ybarra, M. (2017). Green wars: conservation and decolonization in the Maya forest: Univ of California Press.
38


Full Text

PAGE 1

VECTORS OF DISP OSSESSION : AN ASSESSMENT OF LAND TITLING AND DEFORESTATION IN EASTERN NICARAGUA by Isaac Javier Rivera BA., Geograph y, University of Colorado , 2013 A thesis submitted to the Faculty of the Graduate School of the Unive rsity of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Applied Geography and Geospatial Science Department of Geography and Environmental Science 2018

PAGE 2

ii © 2018 ISAAC JAVIER RIVERA ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Master of Arts in Applied Geography and Geospatial Sciences by Isaac Javier Rivera Has been approved for the Department of Geography and Environmental Science By Gregory Simon, Chair Joe Bry an Rafael Moreno Date: May 12, 2018

PAGE 4

iv Vectors of Dispossession : An Assessment on Land Titling and Deforestation in Eastern Nicaragua T hesis directed by Associate Professor Gregory Simon ABSTRACT Awas Tingni marks the site of the historic Inter American Court of Human Rights (IACHR) decision that issued lega l recognition of property rights for the indigenous Mayagna (S umo) community located in North Eastern Nicaragua. The t itling of Awas Tingni marks a historical m oment within international law that recognized that indigenous peoples had rights to property, and were therefore entitl ed to the human rights therein. Landownership and the right to private property was celebrated as a fundamental achievement in the progr rights by the Bretton Woods system that argued that private property was the best mechanism to defend indigenous land s and human rights . Titled officially in 2008, Awas Tingni, and North Eastern Nicaragua more broadly has faced an e xplosion of conflict within newly titled spaces , resulting in spatially specific forest loss . This study asks, what is the relationship between land titling processes and forest cover change? What is the spatial and temporal distribution of forest loss wit hin titled Awas Tingni, Nicaragua? Using the open source Global Forest Cover (GF C) dataset, this study employs remote sensing analysis to distinguish the spatiality of forest loss between the years 2000 2016. Additionally , this study maps before, during, and after title forest loss for Awa s Ti n gni and Eastern Nicaragua more broadly . Finally , the results of this study suggest that the legal administrative properties of land titles do not ensure that forests are protected; rather they reveal that the formal ization of land titles only ext ends state governance and its priorities to newly titled and recognized spaces, warranting further research at the nexus of political ecology and political geography.

PAGE 5

v The form and content of this abstract are approved. I rec ommend its publication. Approved: Gregory Simon

PAGE 6

vi This thesis is dedicated to the past and future students of color in the Department of Geography and Environmental Science, University of Colorado Denver. I hope to make them pro ud. Sí, se puede !

PAGE 7

vii ACK NOWLEDGMENTS I want to thank Glenn Morris and Joe Bryan for their guidance these last few years. It is because of them that I am still in the academy , and pursuing a PhD . I would not be in the academy if it were not for their support.

PAGE 8

viii TABLE OF CONTENTS CHAPTER I. INTRODUCTION AND LITERATURE REVIEW ................................ ............................ 1 Awas Tingni and the Territorial Turn ................................ ................................ ..................... 1 Mapping Forest Loss ................................ ................................ ................................ ............... 5 On Vectors ................................ ................................ ................................ .............................. 7 Open Source Remote Sensing ................................ ................................ ................................ . 9 What this study is not ................................ ................................ ................................ ............ 10 II. METHODS ................................ ................................ ................................ ............................ 14 What is the relationship between land titles and forest cover change ? ................................ .. 14 Global Forest Cover Datasets ................................ ................................ ................................ 15 Land Title Dataset ................................ ................................ ................................ .................. 16 Algorithmic GIS ................................ ................................ ................................ ..................... 17 III. RESULTS AND INTERPRETATION ................................ ................................ .............. 18 Results ................................ ................................ ................................ ................................ .. 18 Limitations ................................ ................................ ................................ ........................... 27 IV . NEXT STEPS ................................ ................................ ................................ ....................... 28 REFERENCES ................................ ................................ ................................ ............................. 35

PAGE 9

ix LIST OF FIGURES FIGURE 1.5 Study Area ................................ ................................ ................................ .............................. 12 1.5 Forest Type ................................ ................................ ................................ ............................. 13 2.4 Algorithm Flow Chart ................................ ................................ ................................ ............. 17 3.1 2001 Forest Loss ................................ ................................ ................................ ..................... 2 0 3.1 2007 Forest Loss ................................ ................................ ................................ ..................... 21 3.1 2008 Forest Loss ................................ ................................ ................................ ..................... 22 3.1 2010 Forest Loss ................................ ................................ ................................ ..................... 2 3 3.1 2016 Forest Loss ................................ ................................ ................................ ..................... 24 3.1 Pre Title Forest Loss ................................ ................................ ................................ ............... 25 3.1 Post Title Forest Loss ................................ ................................ ................................ ............. 26 3.1 Trend Line ................................ ................................ ................................ ............................... 27

PAGE 10

x

PAGE 11

1 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW V ector n: A representation of the world using points, lines, and polygons (ESRI GIS Dictionary) D ispossession n : Dispossession is often the moment when land is enclosed and made property, displacing those who previously inhabited it and making of them a class of people who depend on selling thei r labor for survival. As Marx made clear, it was a condition to think with and from as a means of gaining insight into the workings of capital. The idea has been widely critiqued and rethought, but the problem it poses endures (Joe Bryan, Dispossession Syl labus, 2015). 1.1 Awas Tingni and the Territorial Turn the autonomous regions , or more insistently Divided between the No rth Caribbean Autonomous Region (RAAN) and t he South Caribbean Coast Autonomous Region (RACS), the regions are composed of numerous overlapping land claims that have result ed in a myriad of political conflict s between the Nicaraguan state, land administrators, Afro decedents and indigenous peoples , and logging corporations (Bryan, 2012; Joel Wainwright & Bryan, 2009) . Conflict over land, particularly after the defeat of the Sandi to U.S. backed presidential candidate r esulted in a scramb le over rights to territory , particularly as indigenous peoples returned home from war time status . In many cases, from refugee camps l ocated throughout Nica raguan cities. Led by Miskito insurgents who ran on a n anti colonial platform to secure rights to territory, land titling became a significant tool in which to territorialize 1 their wishes for greater land and resource autonomy , and racial equality, though not without its 1

PAGE 12

2 limitations , or implications on both land and peoples (Bryan , 2007; Nietschmann, 1995; Wainwright & Bryan, 2009) . This historical con juncture for legal recognition in Awas Tingni is only a small pa rt of a larger trend in Latin America that adopted neoliberal reforms as a tool for (Larson, Cronkleton, Barry, & Pacheco, 2008; Offen, 2003) has heavily advocated for indigenous peoples and Afro decedents to obtain collective rights to property , and therefore foster legal governance over land (Bryan, 2012) . Territory in the eyes of the state implies a jurisdictional space, a space ef fecti vely denoting legal jurisdiction for the production of territory through the private property regimes (Delaney, 2005; Hale, 2005) . The Nicaraguan state, however, had refused to recognize the autonomy of the Miskito and Mayangna (S umo) communities, insisting that the land belonged to the state, and was therefore under state jurisdiction , and therefore subject to state governance over lands and resources . Moreover, I contend , to understand the drivers of forest cover change in Eastern Nicaragua , one mus t first reckon with the spa tial hi story of the region, as the h istor ical material conditions of past and present space (Carter, 2010; Harvey, 1984; Mann, 2009) . Thou gh this contribution to this literature is a remote sensing analysis for the years 2000 2016 to understand the materiality and geography of forest cover , what ca me before is just as important, and arguably more given the his torical conjucture of land titling economic history , human rights and rights to property , and their implications on land cover change (Hecht, 2014; Wainw right, Jiang, & Liu, 2013; Wainwright, Jiang, & Liu, 2014) . Spatial history in this context helps us understand the time and space dynamics of social environmental change, an d in particular to leverage both economic and environmental data to illuminate the production of space (Simon, 2014) .

PAGE 13

3 The case of Awas Tin gni vs. Nicaragua (2001) marks a significant historical moment in international law and indigenous peoples rights that continues to facilitate and foster a myriad of research programs dedicated at understanding modern property law, and human rights (Anaya, 2004; Anaya & Grossman, 2002; Anaya & Williams Jr, 2001) . More importantly, the case continues t o have wide repercussions for peoples living within Awas Tingni, and their natural resources (B lomley, 2003; Bryan, 2011; Wainwright & Bryan, 2009) . Beginning as a petition to the Inter American Com mission on Human Rights (IACHR), the case marked the first legal recognition of indigenous peoples rights to property , as a matter o f basic human rights (Anaya & Williams Jr, 2001) . The case rebuked the Nicaraguan state for not recognizing the property rights of the communities of Awas Tingni, and r ecommended as a result, that they establis h a legal system to produce a state procedure for the granting of titles to community lands (Anaya & Grossman, 2002) . Awas Tingni in the y ears p rior to the case collaborated with both James Anaya and the World Wildlife Federation (WWF) for the development of maps to secure their territorial claims in the courts , but also to begin investigating the impact of illegal logging within the area ( Wainwright & Bryan, 2009) . In collaborat ing with the WWF, creating a sustainable timber industry that recognizes indigenous property rights was instrumental to what came after . The Nicaraguan state sold a timber concession to a Korean finance d company called SOLCARSA, i n land s claimed by the com munities, and located just s outh of the WWF assisted land claim called MADESNA. The IACHR court case thereafter ruled that the Nicaraguan state violated the property rights of Awas Tingni by granting a concession to SOLCARSA. What is significant here in re gards to understanding the materiality of land title s lies in how the Court disti nguished property, and rights to territorial and legal possession of land ( Wainwright & Bryan, 2009) . The Court ruled

PAGE 14

4 alone were enough to establish rights of possession, or in other words, thro ugh economic utility, land becomes leg ally titled and is therefore removed from a collective state to officially possessed and privatized land (De Soto, 2000; Locke, 1689; Joel Wainwright, 2009; Ybarra, 2009) . As a result, the state of Nicaragua would st ill be the granter of titles as well as the administrative envoy that would be in charge of granting and legalizing demarcations. Qualitative and quantitative evidence in Honduras , however, has revealed increased unr est over competing claims, resulting in violence in some cases between the Miskito peoples and ladino colonos (colonists) , as well as signs of increased logging , suggesting that similar patterns may also materialize in North East Nicaragua (Mollett, 2011, 2013) . Though Awas Tingni won the C ourt case, it has only in contemporary discourse , for the state still has the real sovereign power here ( Wainwright & Bryan, 2009) . I t c ontinues to view Awas Tingni as an exten sion of the state, rather than S umo territory . Joel Wainwright and Joe Bryan note that the power relations within the recognition of land have only increased the terr itorial governance of the state that amount s to wh at they term an aporia he contradictions and limitations of land titling ( Wainwright & Bryan, 2009) . Land titles when mapped help the sta te align indigenous livelihoods and claims with state instituti ons, as well as state political economic policies. Since Awas Tingni was titled in 2008 , conflict, deforestation, and the precariousness of peoples and claims have only increased. International Law, and international recognition in particular, has only increased the necessity of indigenous peoples to prepare for what continues to be a constant negotiation of rights with the nation state. Moreover, this study contributes to literature in economic history, international law, and political ecology. Or more specifically, this study is located where political eco logy meets political geography. When land ti t les and their genealogies meet the materiality of land cover

PAGE 15

5 change. Land title genealogies in this case open s up space for thinking about how power and kno wledge of private property discourse converge in space, and influences how land change science might as a resu lt interpret land cover change (Munroe, McSweeney, Olson, & Mansfield, 2014) . 1.2 Mapping Forest Loss This study examines the relationship of land titling and deforestation in Eastern Nicaragua by mapping the spatiality and temporality of forest loss. Literature on mapping forest loss, and land cover change more broadly has been facilitated by an explosion in interest in remote sensing science to capture land cover change over long periods of time . Remote sensing for the past 20 years has received considerable attention by ecologists, forest monitors, and conservationists for its utility to capture objective and scientific measurements of forest cover change (Aide et al., 2013; Hansen et al., 2013; Haywood, Alfonsetti, Ortmann, & Takawo, 2015) . For th e last 40 years, forest clearing scholarship has centered mostly on biotic and climatic dimensions of forest cover change, that have included carbon stocks, precipitation measurement, and arctic ice measurement (Foley et al., 2005; Hansen, Stehman, & Potapov, 2010) . With increasing ly improving sensor capabilities, remote sensing technology continues to reach new breakthroughs in resolution that can now classi fy tree cover species, to facilitating the U.S. military in a vast network of global surveillance, and counter insurgency (Wainwright, 2016) . However, thoug h scientists and the public at large continue to receive considerably improved and more easily assessable remote sensing data through government institutions and open data platforms, consideration of the human components of forest cover change, and the con ditions that facilitate that change, have typically fallen outside of remote sensing literature, and have fallen instead within the domain of agrari an change, political ecology, and research on

PAGE 16

6 the human dimensions of climate change (Hecht, 2014; Turner & Robbins, 2008; J Wainwright et al., 2013) . profits, governmen ts, and communities, to produce studies for understanding where forest c hange is occurring , and what it means for ecosystems, economies, an d peoples (Aide et al., 2013; McSweeney et al., 2014; Wegmann, Leutner, & Dech, 2016; Wei et al., 2018) . Only a few studies have attended to the re gional and politica l economic dimensions that drive the material conditions for forest loss in Latin America, making questions of integration between the economic and physical all the more urgent (Hecht, 2014; Lambi n et al., 2001; McSweeney et al., 2014; J Wainwright et al., 2013; Ybarra, 2017) . This study draws from Joel Wainwright et al. (2014) mixed method approach , which integrates b oth remote sensing analysis with regional political economic policies that influ enc es farming choices and methods , logging practices , and land use change more broadly (Joel Wainwright et al., 2014) . Regional political economy in this case includes not just the economic policies that help distinguish and produce the space for logging (e.g., land titles, legal permits , state economic rationale ) , but also the road networks involved that were used to hel p create the material conditions for logging (Laurance, Goosem, & Laurance, 2009; Lefebvre, 2003; T. Mitchell, 2007; Newbold et al., 2015) . Logging in Eastern Nicaragua in particular, has taken the form of legal lo gging by both state and internat ional corporations, but also an unintended influx of illegal black markets that have develo p ed at the same time (Joel Wainwright & Bryan, 2009) . The focus of this study is to understand the spatial distribution of forest los s in its totality in relation to land titles , rather than develop a classification scheme to understand the particular logging concess ions involved, or the ability of remote sensing to capture that change or not. This study draws upon open source remote sensing

PAGE 17

7 datasets, and in particular , the open source Hansen Global Forest Cover dataset (Hansen et al., 2013) . 1.3 On Vectors Vectors, in a purely GIS sense are defined as points, lines, and po lygons, which make up the material components of world representation. In this study, vectors imply specifically to land titles, roads, over lapping claims represented as polygons , colono sites, and bounded spaces of the dispossessed , or de s mov ilizados who have attempted to reclaim space . Additionally, this study attempts to open up space t o think about vectors beyond their naturalized state on map s , for a reading of vector representation outside of contemporary political economic discourse , and outside there seemingly natural or neutral appearance . Moreover, I am particularly interested in the process of how vectors become both naturalized within maps , but also in contempor ary discourse more broadly, but more precisely with how such vectors points, lines, and p olygons, and their meanings come to be both produced and managed (Crampton & Krygier, 2005; Pickles, 2004) . From land titles, to roads, each has a unique history, and a particular vectorization process in the social imaginary that render the m visible or inv isible (K. Mitchell & Elwood, 2016) . Vectorization within the tradition of the cri tique of political economy can also be understood, I argue, at the intersection of economic geography and land change science (Munroe et al., 2014) . Understanding the process and formalization of land tit les, throughout space and time can be understood as a process of vectorization . Therefore, r ecogn izing the material consequences on land and peoples through vectorization can help reveal how space is produced, and therefore subject space to enclosures for particular commodities, but also ontological

PAGE 18

8 understandings of that space before private property (Goeman, 2013; Mezzadra & Neilson, 2013; Offen, 2003) . Dispossession in this case is both the material and im material understandings of land cov er, or in this particular case, the literal loss of a forest . Drawing vectorialism , and historical materialism more broadly, I contend for a reading of land titles that denaturalizes its intended political economic, and hu ma n rights assumptions within contemporary economic discourse s (Harvey, 1984; Mann, 2009; Marx, 1867; Joel Wainwright & Bryan, 2009; Wark, 2004) . More specifically, I argue for a reading of vectors that calls into worlding propertie s of land t itles, t he material propertie s of land titles being judicial recognition, spatial territorialtilization , and state extensions for governing through land titles (J oel Wainwright, 2009; Joel Wainwright & Bryan, 2009; Ybarra, 2009) . Land titles when mapped in space connote a particular reading of space that was not there previously before, making the act of map ping , much like the process of formalizing land titles, a p owerful tool in which no explanation is given to the ir convergence in space, or their material implications on it. Vectorization as a result is fundamentally connected to the process of mapping, particularly through GIS infrastructure , and through the fo rmalization process of land titles that the state relies on to draw private property lines (Dwyer, 2015; Mezzadra & Neilson, 20 13; Wilson, 2017) . Though this study will not particularly engage with the history of land titles per se, nor develop an analytic for thinking about vectors, nor assess their political economic implications on regional political geography, this study is intended to begin an albeit small contribution within literature dealin g with the material implications of land titles, and private property more broadly by assessi ng forest loss within these spaces . Moreover, by pushing vectors beyond their

PAGE 19

9 naturalized an d neutral appearance on maps, I aim to open up space for thinking about vectors in much the same way that a cartographer or GISer might think about vector data . But further, by linking their GIS epistemology to liberal economic philosophy and their philoso phers in the past and present moment , that have argued that private property, in this case the bounded vector land title, are intimately purposed for the protection of both land and human rights. I critique these hegemonic and naturalized assumptions throu gh a critique of political economy. This merger of epistemology and economic philosophy , particularly the critique of political economy given the quantitative and qualitative evidence provided within this study , can be understood , I argue, as vectors of di spossession . 1.4 Open Source Remote Sensing With remote sensing technology becoming ever more accessible, and more useful in regards to understanding land cover ch ange, and capturing land cover trends over time, researchers in agrarian studies and politi cal ecology have begun to integrate their scientific methods with promising results (Wegmann et al., 2016) . This study draws from the Hansen et al (2013) dataset on publically available global forest cover for integration of forest cover change data, with political economy, or in this particular case land title s . The global forest cover dataset at the time of i ts release was hailed as a fundamental step into democratizing remote sensing research, and environmental science more broadly, for its publically and freely available data, made possible by the Landsat pr ogram at NASA and the USGS (Hansen et al., 2013) . It has become the staple dataset used by forest monitors, non assess forest cover. Though with limitations in regards to understanding particular l and use

PAGE 20

10 change at a finer scale of analysis, research has confirmed its accuracy at regional and global scales of analysis (Gonzalez et al., 2016; MacDicken, 2015) . Open source remote sensing, however, do es not bridge the gap in the uneven geography of knowledge production on localized environmental expertise, it has , however, helped expand its usage across academic and non academic circles , resulting in increased usage and demand (Elwood, 2008; Haraway, 1988; Lave, 2015) . The Landsat program represents four decades worth of high resolution images acquired from Landsat sat ellites. This particular study uses the global forest cover dataset from Hansen et al (2013) be tween the years (2000 2016), drawing specifically from images acquired from Landsat 7 ETM + and Landsat 8 OLI . For greater transparency, replication, and integration of datasets , this study employs algorithmic GIS science to facilitate a multi step analy sis of both Landsat and land title datasets . More s pecifically, this study generates an algorithm for conducting remote sensing analysis of Landsat 7 and 8 images from the years (2000 2016) for the quantification of forest loss for before, during, and afte r title. Literature in open source remote sensing is increasing ly drawn to the potential of al gorithmic GIS science to conduct both complex analysis across different data types and data structures, and high memory demand inquiries, in a short amount of tim e , while at the same time providing new avenues for the public sharing of base code for interested institutions and peoples (Christophe & Inglada, 2009) . The utility in using algorithms in my view is centered on the ability to share base code with communities that may find it useful for their own forest/title assessments , as well as the ability to pr oduce both datasets and code for the repl ication of a similar study in another space. This algorithm, which I call vectors of dispossession will be uploaded to Github for public use.

PAGE 21

11 1.5 What this study is not he spatiality and temporal ity of forest loss within titled spaces in Eastern Nicaragua, or more specifically, Awas Tingni. This study is not a forest type classification study, nor an analysis of the difference between La ndsat 7 and Landsat 8 satellite res olution capture benefits , nor an assessment of the utility of open source remot e sensing to conduct automated remote sensing analysis rpose rather is to add inquiry and empirical data tha t helps us understand the political ecology, and political geography of land tenure formalization schemes across Nicaragua, and other g eographies around the world, but particularly as the Wo rld Bank continues to insist on that land tenure; private property ; as the (Dwyer, 2013, 2015; Watts, 2003) . The recent debates in the Proceedings of the National Academy of Sciences furt her reflect the need for inquiry at t he intersections between land tenure, forest cover change, and human rights (Blackman, Corral, Lima, & Asner, 2017; Robinson, Holland, & Naughton Treves, 2017) . It is our responsibility as geographers to make sense of these competing discourses that co ntinue t o foster world changing polities , resulting in human economic policies that have not addressed the ir aporias and contradictions within (Massey, 2004) .

PAGE 22

12 (Figure 1. Study Area. Source, Author) F orest Loss 2 001 2016 data derived from Hansen et al. (2013), titled data provided by the National Commission of Titles and Demarcation of Nicaragua. Title data was reformatted for algorithm usage.

PAGE 23

13 (Figure 2. Forest Type. Source , The National Governance and Reconciliation Center of Nicaragua) Figure 2 depicts bio diversity , eco systems, and ecological regions for the entirety of Nicaragua. Forest type in Eastern Nicaragua mostly consists of open forest, broad leaf forest, and pine savanna .

PAGE 24

14 CHAPTER 2 METHODS AND METHODOLOGIES 2.1 What is the relationsh ip between land titles and forest cover change? Spatiality and Temporality The purpose of this study is to better understand the spatiality and temporality of forest loss in North Eastern Nicaragua. Spatiality is of particular importance due to the spatia l extent of land titles both locally in Awas Tingni, but also in regards to the more regionally specific trends that political economic policy facilitates, particularly as it relates to land management. Spatiality in this case is mapped at both regional an d local scales. Regional scale in this case includes all land titles in the RAAN that run up to the border with Honduras in the North, and the Atlantic Ocean in the East. Local scale in this case focuses particularly on the Awas Tingni claim. Focusing at t he scale of the local allows for a better understanding of how governance takes shape over spa ce. In addition to mapping forest loss and land titles, this study will also map local roads, overlapping claims, and colonos , or in other words, the vectors unde rgoing inquiry in this study.

PAGE 25

15 2.2 Global Forest Cover Datasets Landsat Datasets Landsat data was acquired from version 1.4 of the Global Forest Change dataset for the years 2000 2016. Landsat 7 (Enhanced Thematic Mapper Plus) ETM+ data is composed of a composite of yearly images between the years 2000 2013. Landsat 8 Operational Land Imager (OLM) and Thermal Infrared Sensor (TIRS) is composed of a composite of yearly images between the years 2013 2016. Landsat 7 ETM+ operates at a 30 meter spatial resolution . Landsat 8 OLM/TIRS also operates at a 30 meter spatial resolution but has an additional two bands that g ive it added seasonal precision and cloud removal propert ies for added clarity . The TIRS sensor is particularly important becaus e of its abi lity to capture additional images than the Landsat 7 ETM+, resulting in a higher probability that images will be cloud free, resulting in less distorted and higher quality images. The Global Forest Change dataset includes data the was reprocessed post 20 11 adding to improved loss measurements, i mproved training data for calibrating the loss model, i mproved per sensor quality assessment m odels to filter input data, and i mproved input spectral features for building and applying the loss model. Data co mparis ons between the years 2001 2016 are therefore limited given quality differences across the dataset (Hansen et al., 2013) . The dataset, however, is still useful in that it provides us an initial window for understanding the spatialit y classified dataset that is ready for user analysis. D ata preparation was in significant , for the

PAGE 26

16 exception o f clipping Landsat images to desired study area . Th e dataset is divided into 10x10 degree tiles, consisting of 8 bit values that have a spatial resolution of 1 arc second per pixel, or about 30 meters per pixel at the equator. Forest l oss during the period 2000 2016 is defined as a stand replacement distur bance, or a change from a forest to non forest state. Data is encoded as 0 for (no loss) , values with a range of 1 16 correspond with the forest loss years for the years 2001 2016, respectively. 2.3 Land Title Dataset Marcos Williamson of the Comisión Nacional de Demarcación y Titulación, or CONADETI, compiled all land title GIS data using data from the center . The land title dataset consists of 29 polygons for all titled indigenous territories in E astern Nicaragua. Date of title for all territories was formatted to only account for year of loss, rather than exact date of title . This allowed process allowed for easier integration with array data structures, for a more structured algorithm. Territories throughout the dataset are represented as polygons, and exist in this dataset a s a shape file, or .shp file. All attribute data, including name of territories, and date of title remained intact post processing. 2.4 Algorithmic GIS Algorithmic GIS in recent years has grown due to demand for its utility to conduct highly effectiv e , mul ti layered analysis, and producing automation schemes that can save considerable amount of labor time to produce the same result. For future scale up of analysis, replication, and

PAGE 27

17 sharing of particular functions, an algorithm was built to address the core research question of this study: w hat is the relationship between forest loss and land titles? The inputs for this algorithm are as follows: 1) Hansen Global Forest Cover Raster derived from version 1.4 of dataset and 2) Land Title dataset deriv ed the National Commission of Demarcation and Titling, of Nicaragua. Following inputs, the data was extracted by mask, clipped by raster, and converted into an array. An array is a data structure that allows data to be converted into an index for easier pr ocessing, and easier outputs. Unprocessed Array is then converted into a subset of arrays that correspond to each year of titling and for all land title polygon s arrays. Arrays are then converted into for mapping visualization. Forrest loss area calculations are then exported into a final csv file for ea ch year loss for all titled polygons s data frame data frames allow for easier indexing of arrays for easier export to csv.

PAGE 28

18 Figure 3 . Algorithm Flow Chart. Source, Author This chart is intended t o simplify and map out the multi step process for conducting the core analytical contribution of this study. Additionally, it is also intended to demystify and de fetishize algorithmic GIScience more broadly. Doing so in my view could greatly increase the usage and approachability of programming by those who may doubt their abilities to do GIS programming. Though far from perfect, this flow chart is designed to help conceptualize t he logic behind the algorithm.

PAGE 29

19 CHAPTER 3 RESULTS AND INTERPRETATION Algorithm deliverab les including all rasters of forest loss, mosaics of forest loss representing forest l oss for before, during, and after title, were used to c onstruct Figures 3 9. Cartogra phy, and the mapping of forest loss in relation to land titles, overlapping claims, roads, and colonos were all created outside the algorithm.

PAGE 30

20 Figu re 4 . 200 1 Forest Loss . Source, Author Figu re 4 reveals very light forest loss throughout the Awas Tingni claim. Most of t he forest loss is located just south e ast of the claim.

PAGE 31

21 Figure 5 . Forest Loss 2007. Source, Author Hurricane Felix hit Awas Tingni, Nicaragu a on September 7, 2007 . This figure reveals heavy forest loss clustering just South of the Hurricane Felix path. Due to increased cloud cover, and a faulty mirror in Landsat 7, forest lo ss distribution is distorted, and of less quality than Figure 3. This map should be read to understand the general spatial trend in forest loss for this year. All images for each year represent the sum amount of forest loss for any particular year.

PAGE 32

22 Figure 6 . 2008 Forest Loss. Source, Author Figure 6 depicts very light for est loss. This is also the first official year that Awas Tingni was titled. This is also the first year after the devastation of Hurricane Felix .

PAGE 33

23 Figure 7 . 2010 Forest Loss. Source, Author Figure 7 reveals forest loss clustering and a new peak in for est loss for the region. The clustering, however, is located primarily due South of titled Awas Tingni.

PAGE 34

24 Figure 8 . 2016 Forest Loss . Source, Author Figure 8 reveals another peak in forest loss. Forest loss clustering in this case is located within tit led Awas Tingni and just South of the title. Forest loss clustering in the South is in proximity to road networks, indicating that there may be correlation with forest loss and road network expansion. Within the title, forest loss clustering is primarily l ocated east of the Sta Clara claim. Clustering of forest loss within the Miskito Excombatants claim, and the site of Colonos, also reveal peaks of forest loss.

PAGE 35

25 Figure 9 . Forest Loss Pre Title. Source, Author Figure 9 reveals forest loss clustering prim arily West of the Colectivo El Pro. Clustering however should be taken with a degree of caution given distorted images coming from the years 2005 and 2007.

PAGE 36

26 Figure 10 . Forest Loss Post Title. Source, Author Figure 10 reveals forest loss after title to be primarily located just east of the Colectivo El Pro. Whereas pre title forest loss is primarily located in the West. Road networks, and overlapping community claims are also primarily located in the mid to east sections of the Awas Tingni.

PAGE 37

27 Figure 11 reveals a positive trend in forest loss throughout the Awas Tingni Claim over time . We see peaks in 2007, 2014, and 2016. Peak in 2007 is associated with Hurricane Felix. Peaks in 2014, and 2

PAGE 38

28 3.1 Res ults In summary, the relationship between land titling and forest loss was shown to have a positive relationship. The brief history of the territorial turn, the algorithm as built, and the maps that came after the algorithm process , are all but smaller co mponents that could each on their own feed into their own regard is that it is expansive enough to weave seemingly separate epistemological and ontological frameworks, while each hav ing their own disti nct histories that are woven that materialize in space. This study attempts to thread two seemingly separate and radically different data worlds onto themselves; that of forest cover and land title history for an improved understanding o f the legal and political limitations of land titles to defend both land and human rights. From the spatial analysis, there is a clear indication that forest loss begins very minimally in Awas Tingni in 2001 (Figure 3), following with rapid increase in 2 007 (Figure 4). The rapid increase in forest loss is likely due to a Category 5 Hurricane Felix, which made a direct hit on Awas Tingni on September 7, 2007. In 2008, the first official year of titled Awas Tingni, we see most of the deforestation just sout h of the boundary (Figure 5). In 2010, we see another spike in deforestation, though most of that deforestation is just south of Awas Tingni (Figure 6). In 2016, we see a major spike throughout Eastern Awas Tingni (Figure 7). Much of the forest loss is loc ated just east of the overlapping claims, and closer to the road networks that were mostly built after 2011. Both Miskito Excombatants claims, with Awas Tingni also show a significant amount of forest loss in 2016. The overall trend points towards a gradua l increase in forest loss with peaks in both 2007, 2011, and 2016 for both Awas Tingni, and surrounding titled and untitled spaces. Given that most roads in this network were built after 2011, there may be a

PAGE 39

29 correlation with road development, and forest lo ss, particularly as regional markets for logging concessions have increased, along with illegal black markets at the same time. Forest loss for pre title (Figure 8) reveals clustering of forest loss in proximity to colonos in western Awas Tingni. Clusteri ng of forest loss, however, is slightly distorted due to less input data into the Hansen dataset. This is due to Landsat 7 ETM+ mirror failure that caused significant images to be loss. Forest loss for post title (Figure 9) reveals spikes in forest cluster ing both within Awas Tingni and just south of the titled boundaries. Clustering of forest loss is particularly spiked in the years 2011, and 2016. Forest loss post title is mostly concentrated in the mid to east boundaries of Awas Tingni, and particularly in proximity to newly built roads. Most of the roads within Aw as Tingni were built after 2010. Due to lacking road data past 2012, it is likely that road networks have only expanded with time into various clustered forest loss locations.

PAGE 40

30 3.2 Limit ations The reprocessed data from the Global Forest Cover dataset from 2011 onward, the use of Landsat 8 OLI data for 2013 onward, improved training data for calibrating the loss model, and improved input spectral features for building and applying the lo ss model for data 2011 onward, produced significant variability across the 2000 2016 dataset. Figures 3, 4, 5, and 6 reveal forest loss with less resolution clarity, and quality then figures with forest data from 2011 onward. Forest loss area calculations, as a result, are therefore unreliable in this case, and cannot be used to account for precise calculations of forest loss. Additionally, maps of forest loss before 2011 , display data that appear slightly pixelated, resulting in maps before 2011 to appe ar slightly course in some instances. Forest loss in 2007 in particular, which was also the year of Hurricane Felix, reveals clustered concentrations of forest loss, but at a lesser quality due to both increased cloud cover, and a faulty mirror in the Land sat 7 ETM+ sensor. Maps before 2011 should be read with some degree of caut ion. They are, however, still useful for understanding the spatiality of forest loss. Figure 3, 2001 loss in particular captures forest loss to a finer degree than forest loss in 20 07 (Figure 4). This can be attributed to the gaps of data of Landsat 7 images for the year 2007 , which are attributed to a faulty mirror on the Landsat sensor .

PAGE 41

31 CHAPTER 4 NEXT STEPS 4.1 Future Remote Sensing Analysis What comes next in this study depends on our ability to address the seemingly disconnected geographies of algorithmic GIS science, remote sensing, and political economy. For algorithmic GIS science and remote sensing, and for the algorithm built for this particular study, its success will depend on how the non trained GISer, or non coder can insert updated remote sensing datasets onto its base code. Version 2.0 of the Hansen dataset is currently in the works, and should be released at the end of the year , giving this study a clear follow up in both acces sibility, and adjustment using higher quality datasets . Moreover, updating the forest loss dataset will be essential for conducting accurate forest loss counts across the entirety of the dataset. Additional ly, finding a clean cloud free World V iew data set image would allow for the digitizing of the most current road networks. Having updated land cover data, forest cover data, and road data will greatly improve future results, and the ability to use such data in conversation with the qualitative. With improved datasets, remote sensing analysis can begin to identify the particular forest types that are being loss. With forest loss being broadly throughout the Hansen et al. dataset , identify ing what exactly those disturbances are can better improve our understanding of the particular commodities involved in land use change. Through a series of classification schemes of forest and land cover type, future remote sensing analysis could greatly benefit from taking a greater account of the heterogeneity of land cover in Awas Tingni and Eastern Nicaragua more broadly. More so, iden tifying particular commodities can

PAGE 42

32 also feed into the broader interconnections between political econom y and local governance of land, resulting in new mechanisms ways to understand the political ecology of land titling systems. 4.2 Genealogies of Knowle dge Production Furthermore , addressing questions regarding ethics in knowledge production on this frontier can benefit from an approach that emphasizes a co producti on of citizen science for conducting future analysis (Hacker, 2013; Lave, 2015) . For instance, breaking free from the global North to global South flows of knowledge production that so much of open source and participatory mapping facilitates can further benefit from the co production approaches of citizen science for addressing not just gaps in particular GIS methods, but also enabling a ground truth team to further strengthen the claims implied by land cover maps depicting forest loss on indigenous lands (Livingstone, 1995, 2010; Spivak, 1999) . But also to strengthen a process in which the community, rat her than University in the global North interprets the meaning and implications of that forest loss (Massey, 2004; Spivak, 1999) . What does it m ean for indigenous peoples to see a forest stripped from its former, perhaps original standing? What are the ontological implications for indigenous peoples lives beyond the ecological ripple effect across commoditized ecologies of forest loss? In short, I believe that integrating both ethnographies of forest lost and political economy could be quite useful. Additionally, integrating the results from this study with additional qualitative data from Awas Tingni, and the broader surrounding areas c an further sharpen the claims and results made in this study. The qualitative data conducted by Joe Bryan in particular, when in conversation with the maps presented in this study will be able to p roduce yet new

PAGE 43

33 geographies of understanding regarding the role of the state, maps, and the law in facilitating land cover change (Bryan, 2007) . Questions within the traditions of political ecology, and political geography will as a result be enriched by the integration of qu alitative and quantitative data, resulting in new ways to understand the p olitical economy, legal geography, and the drivers of land cover change more broadly . 4.3 Vectors of Dispossession Moreover , in drawing from historical materialist analytics that engage with questions of indigenous ways of knowing and remote sensing analysis , I propose a future study to further develop the idea and implications for thinking through the concept of vectors of dispossession . In short, I pr opose to develop an analytic that draws from historica l materialism to better understand the spatial, political, and material dimensions of land titling. I call this merger of epistemological and ontological frameworks vectors of dispossession . Doing so, in my view can better reveal the totalizing colonizatio n of everyday life to address both inequities in methodologies, but also in how we ultimately understand the material and ontological dimensions of dispossession (Lefebvre, 2004; Monte Mor, 2014) . Drawing from literature in critical GIS, historical materialism, and decolonial studies, I will attempt to build upon this g geography and geographers to reckon with how (Coulthard, 2014; Mariátegui, 1971; Springer, 2017; Webber, 2011; Wilson, 2017) . (Springer, 2017) . Connecting seemingly unrelated and distinct literatures that engage radically different epistemologies can produce new spaces for thinking about questions of dispossession.

PAGE 44

34 Vectors of Dispossession are not particular to Eastern Nicaragua by any means, rather, they ar e the mat erial analytics of colonized everyday life.

PAGE 45

35 REFERENCES Aide, T. M., Clark, M. L., Grau, H. R., López Carr, D., Levy, M. A., Redo, D., . . . Muñiz, M. (2013). Deforestation and reforestation of Latin America and the Caribbean (2001 2010). Biotropica, 45 (2), 262 271. Anaya, S. J. (2004) . International Human Rights and Indigenous Peoples: The move toward the multicultural state. Ariz. J. Int'l & Comp. L., 21 , 13. Anaya, S. J., & Grossman, C. (2002). The Case of Awas Tingni v. Nicaragua: A Step in the International Law of Indigenous Peopl es. Ariz. J. Int'l & Comp. L., 19 , 1. Anaya, S. J., & Williams Jr, R. A. (2001). The protection of indigenous peoples' rights over lands and natural resources under the Inter American human rights system. Harv. Hum. Rts. J., 14 , 33. Blackman, A., Corral, L., Lima, E. S., & Asner, G. P. (2017). Titling indigenous communities protects forests in the Peruvian Amazon. Proceedings of the National Academy of Sciences , 201603290. Blomley, N. (2003). Law, property, and the geography of violence: The frontier, th e survey, and the grid. Annals of the Association of American Geographers, 93 (1), 121 141. Bryan, J. (2007). Map or Be Mapped. Bryan, J. (2011). Walking the line: participatory mapping, indigenous rights, and neoliberalism. Geoforum, 42 (1), 40 50. Bry territorial turn. Geography Compass, 6 (4), 215 226. Carter, P. (2010). The road to Botany Bay: An exploration of landscape and history : U of Minnesota Press. Christop he, E., & Inglada, J. (2009). Open source remote sensing: Increasing the usability of cutting edge algorithms. IEEE Geoscience and Remote Sensing Newsletter, 35 (5), 9 15. Coulthard, G. S. (2014). Red skin, white masks: Rejecting the colonial politics of r ecognition : Taylor & Francis. Crampton, J. W., & Krygier, J. (2005). An introduction to critical cartography. ACME: An International Journal for Critical Geographies, 4 (1), 11 33. De Soto, H. (2000). The mystery of capital: Why capitalism succeeds in the West and fails everywhere else: New York: Basic Books. Delaney, D. (2005). Entering the territory of territory. Territory: A Short Introduction , 1 33. Developme nt and Change, 44 (2), 309 333. Dwyer, M. B. (2015). The formalization fix? Land titling, land concessions and the politics of spatial transparency in Cambodia. The Journal of Peasant Studies, 42 (5), 903 928. Elwood, S. (2008). Volunteered geographic info rmation: future research directions motivated by critical, participatory, and feminist GIS. GeoJournal, 72 (3 4), 173 183. Foley, J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R., . . . Gibbs, H. K. (2005). Global consequences of land use. Science, 309 (5734), 570 574. Goeman, M. (2013). Mark my words: Native women mapping our nations : U of Minnesota Press. Gonzalez, A., Cardinale, B. J., Allington, G. R., Byrnes, J., Arthur Endsley, K., Brown, D. G., . . . Loreau, M. (2016). Estim ating local biodiversity change: a critique of papers claiming no net loss of local diversity. Ecology, 97 (8), 1949 1960. Hacker, K. (2013). Community based participatory research : Sage Publications.

PAGE 46

36 Hale, C. R. (2005). Neoliberal multiculturalism. PoLAR: political and legal anthropology review, 28 (1), 10 19. Hansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S., Tyukavina, A., . . . Loveland, T. (2013). High resolution global maps of 21st century forest cover change. Science, 342 (6160), 8 50 853. Hansen, M. C., Stehman, S. V., & Potapov, P. V. (2010). Quantification of global gross forest cover loss. Proceedings of the National Academy of Sciences, 107 (19), 8650 8655. Haraway, D. (1988). Situated knowledges: The science question in femini sm and the privilege of partial perspective. Feminist studies , 575 599. Harvey, D. (1984). On the history and present condition of geography: an historical materialist manifesto. The Professional Geographer, 36 (1), 1 11. Haywood, A., Alfonsetti, A., Ortm ann, A., & Takawo, D. (2015). Improving national greenhouse gas inventories for forestry and land use change using open source software. Paper presented at the Geoscience and Remote Sensing Symposium (IGARSS), 2015 IEEE International. Hecht, S. B. (2014). Journal of Peasant Studies, 41 (5), 877 909. Lambin, E. F., Turner, B. L., Geist, H. J., Agbola, S. B., Angelsen, A., Bruce, J. W., . . . Folke, C. (2001). The causes of lan d use and land cover change: moving beyond the myths. Global Environmental Change, 11 (4), 261 269. Larson, A. M., Cronkleton, P., Barry, D., & Pacheco, P. (2008). Tenure rights and beyond: community access to forest resources in Latin America : Center for international forestry research (CIFOR). Laurance, W. F., Goosem, M., & Laurance, S. G. (2009). Impacts of roads and linear clearings on tropical forests. Trends in ecology & evolution, 24 (12), 659 669. Lave, R. (2015). The Future of Environmental Experti se. Annals of the Association of American Geographers, 105 (2), 244 252. Lefebvre, H. (2003). Space and the State. State/space: A reader , 84 100. Lefebvre, H. (2004). The production of space, 1991. The City Cultures Reader (3). Livingstone, D. N. (1995). The spaces of knowledge: contributions towards a historical geography of science. Environment and Planning D: Society and space, 13 (1), 5 34. Livingstone, D. N. (2010). Putting science in its place: geographies of scientific knowledge : University of Chica go Press. Locke, J. (1689). Second Treatise . Toronto: University of Toronto Press. MacDicken, K. G. (2015). Global forest resources assessment 2015: What, why and how? Forest Ecology and Management, 352 , 3 8. Mann, G. (2009). Should political ecology be M materialism. Geoforum, 40 (3), 335 344. Mariátegui, J. C. (1971). Seven interpretative essays on Peruvian reality : University of Texas Press. Marx, K. (1867). Capital, volume I: Harmondsworth: Penguin/New Left Review . Massey, D. (2004). Geographies of responsibility. Geografiska Annaler: Series B, Human Geography, 86 (1), 5 18. McSweeney, K., Nielsen, E. A., Taylor, M. J., Wrathall, D. J., Pearson, Z., Wang, O., & Plumb, S. T. (2014). Drug policy as conservation polic y: narco deforestation. Science, 343 (6170), 489 490.

PAGE 47

37 Mezzadra, S., & Neilson, B. (2013). Border as Method, or, the Multiplication of Labor . Mitchell, K., & Elwood, S. (2016). Counter Mapping for Social Justice. Politics, Citizenship and Rights , 207 223. Mitchell, T. (2007). The properties of markets. Do economists make markets , 244 275. Mollett, S. (2011). Racial narratives: Miskito and colono land struggles in the Honduran Mosquitia. Cultural Geographies, 18 (1), 43 62. Mollett, S. (2013). Mapping decep tion: The politics of mapping Miskito and Garifuna space in Honduras. Annals of the Association of American Geographers, 103 (5), 1227 1241. Monte Mor, R. (2014). Extended urbanization and settlement patterns in Brazil: an environmental approach. Implosion s/explosions: towards a study of planetary urbanization. Berlin: Jovis , 109 120. Munroe, D. K., McSweeney, K., Olson, J. L., & Mansfield, B. (2014). Using economic geography to reinvigorate land change science. Geoforum, 52 , 12 21. Newbold, T., Hudson, L . N., Hill, S. L., Contu, S., Lysenko, I., Senior, R. A., . . . Collen, B. (2015). Global effects of land use on local terrestrial biodiversity. Nature, 520 (7545), 45. Nietschmann, B. (1995). Defending the Miskito reefs with maps and GPS: mapping with sai l, scuba, and satellite. Cultural survival quarterly, 18 (4), 34 37. Offen, K. H. (2003). Narrating place and identity, or mapping Miskitu land claims in northeastern Nicaragua. Human organization, 62 (4), 382 392. Pickles, J. (2004). A history of spaces: Cartographic reason, mapping, and the geo coded world : Psychology Press. Povinelli, E. A. (2002). The cunning of recognition: Indigenous alterities and the making of Australian multiculturalism : Duke University Press. Robinson, B. E., Holland, M. B., & Nau ghton Treves, L. (2017). Community land titles alone will not protect forests. Proceedings of the National Academy of Sciences , 201707787. Simon, G. L. (2014). Vulnerability in production: a spatial history of nature, affluence, and fire in Oakland, Calif ornia. Annals of the Association of American Geographers, 104 (6), 1199 1221. Spivak, G. (1999). A critique of postcolonial reason: toward a critique of the vanishing present: Cambridge, MA: Harvard University Press. Springer, S. (2017). Earth writing. Geo Humanities , 1 19. Turner, B. L., & Robbins, P. (2008). Land change science and political ecology: Similarities, differences, and implications for sustainability science. Annual review of environment and resources, 33 , 295 316. the Church, and the colonial state in southern Belize. Journal of Historical Geography, 35 (3), 428 450. Wainwright, J. (2011). Decolonizing development: colonial power and the Maya (Vol. 36): John Wiley & Sons. Wainwright, J., & Bryan, J. (2009). Cartography, territory, property: postcolonial reflections on indigenous counter mapping in Nicaragua and Belize. Cultural Geographies, 16 (2), 153 178. Wainwright, J., Jiang, S., & Liu, D. (2013 ). Deforestation and the world as representation: the Maya forest of Southern Belize. Land Change Science, Political Ecology, and Sustainability: Synergies and Divergences , 169 190.

PAGE 48

38 Wainwright, J., Jiang, S., & Liu, D. (2014). The Maya forest of Southern Belize. Land Change Science. Political Ecology and Sustainability: Synergies and Divergences, 169 . Wainwright, J. D. (2016). The US Military and Human Geography: Reflections on Our Conjuncture. Annals of the American Association of Geographers, 106 (3), 51 3 520. Wark, M. (2004). A hacker manifesto . Watts, M. (2003). Development and governmentality. Singapore Journal of Tropical Geography, 24 (1), 6 34. Webber, J. R. (2011). Red October: left indigenous struggles in modern Bolivia (Vol. 29): Brill. Wegmann, M., Leutner, B., & Dech, S. (2016). Remote sensing and GIS for ecologists: using open source software : Pelagic Publishing Ltd. Wei, X., Li, Q., Zhang, M., Giles Hansen, K., Liu, W., Fan, H., . . . Liu, S. (2018). Vegetation cover another dominant factor in determining global water resources in forested regions. Global change biology, 24 (2), 786 795. Wilson, M. W. (2017). New Lines: critical GIS and the trouble o f the map. Ybarra, M. (2009). Violent visions of an ownership society: The land administration project in Peten, Guatemala. Land Use Policy, 26 (1), 44 54. Ybarra, M. (2017). Green wars: conservation and decolonization in the Maya forest : Univ of Californ ia Press.