Quantifying deforestation rates in a cocaine source region in Bolivia: A study to verify rates of land-use change under pro-coca policies

Author: Edmond Nagombi

Nagombi, Edmond, 2017 Quantifying deforestation rates in a cocaine source region in Bolivia: A study to verify rates of land-use change under pro-coca policies, Flinders University, School of the Environment

This electronic version is made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact copyright@flinders.edu.au with the details.


This thesis focuses on humid tropical deforestation in one of the most important coca growing regions in central Bolivia, Chapare. On the one hand, the cultivation of coca leaves which is the source of cocaine paste, their processing and narcotrafficking are global, social and legal issues that lead to violence, corruption, and instability in foreign policy frameworks. On the other hand, their cultivation has local environmental and economic repercussions that are not always negative.

The project is grounded in geospatial science and the methodology comprises of three main sections: image pre-processing, land-use and land-cover classification aimed at forest/non-forest mapping, and the generation of forest and non-forest statistics for individual farms in four communities in Chapare. Landsat 4 & 5 (TM) (2011) and Landsat 8 (OLI) satellite (2015 and 2016) image data were used. Pre-processing steps covered verification of geometric and radiometric parameters, image mosaicking and, for the 2015 data, pan sharpening. Thus, the images were classified using unsupervised classification to map major land-use and land-cover types from 2015 imagery. These were verified with reference to field data collected in 2015 that was made available to this research project. Forest and non-forest classification was carried out for the 2011, 2015 and 2016 image data. All geospatial analyses were done in ERDAS Imagine 2015 and ArcGIS version 10.4.

The results of the three classified images for 2011, 2015 and 2016 shows overall accuracies of 84.51%, 87.84% and 98.42% respectively. Comparing the four different communities investigated in detail in the study area, the areas in the Community I and Community II shows regrowth of forest areas while areas in Community IV shows increased rate of deforestation. Community I with an average of 20 ha of land parcels and a standard deviation of 3 hectares has a current average regrowth rate of 204.73 ha/year which is 2 ha/year per farm at the end of the study period in 2016. Community II with mix grazing and farming land parcels indicated by its size of 50 ha (50 hectare land parcels are meant for pastures) have a regrowth of 786 ha/year which is 8.46 ha/year per farm in 2016. Community III and IV has a clearance rate of 25.18 ha/year which is 0.42 ha/year per farm and 79.56 ha/year which 1.99 ha/year per farm.

However, thereis 50% decrease in deforestation rate for the communities mapped in detail compared to statistics generated for the 1980s and 1990s. This is indicative of the Bolivian government’s pro-coca policies and legislation since they were first elected in 2006, and their increasing influence since c. 2000. The quantitative data reaffirms the hypothesis that weak anti-coca polices lead to less deforestation in Chapare.

A key recommendation arising from the use of pan-sharpening technique used in this study shows improvement of the quality of the image and its spatial resolution in the 2015 classification discriminating individual forest degradation that needs to be explored. That is recommended as an advantage that must be employed on newly available sources of high spatial resolution imagery, to more accurately map and monitor fluctuations in the rates of deforestation in different communities and different farmers in each community under policies that either promote or discourage coca cultivation in future researches.

Keywords: deforestation, deforestation and land use, land use and land cover change, anti-narcotics policies, pro-coca policies, Chapare, Bolivia, coca trade, coca source region

Subject: Environmental Science thesis

Thesis type: Masters
Completed: 2017
School: School of the Environment
Supervisor: Prof Andrew Millington