Artisinal gold mining activities and forest loss in upper Mazaruni Territory, Guyana, between 1986 and 2013

Abstract

Artisanal and small-scale gold mining activities (ASM) have grown significantly with a spike in markets over the past few decades. However, the spatial extent and environmental impacts of these activities are poorly known. Some recent studies suggest the need for appropriate scientific techniques to estimate these impacts. The study area of Jawalla village in Guyana’s Upper Mazaruni River Basin was selected for the following reasons: 1) documented growth in ASM-based production, 2) the country’s leading role in REDD+ forest protection, and 3) extremely low reported rates of forest clearance. The primary objectives of this study were to quantify the extent and total area of ASM, and to analyze and estimate the forest loss caused by these activities from 1986-2011. The changes in artisanal gold mining activity over a 25-year time period were studied based on available and suitable Landsat data with acceptable low cloud cover and a RapidEye image of 2011 with 5m resolution. Maximum-Likelihood supervised classification and knowledge engineer classifier were applied to the Landsat images. Object-based image analysis was used to segment and classify the RapidEye image. Post-classification comparisons were performed to estimate the changes in landuse and landcover. Accuracy of the classified images of 2010 and 2011 was then assessed against the RapidEye. High-Resolution Global Maps of Forest Change was also used to verify forest changes between 2000 and 2010. The results indicate that deforestation from gold mining increased significantly between 1986 and 2011. The findings also highlighted the advantage of using high-resolution images and object based image classification technique to obtain accurate estimates.