Land use/cover changes and carbon stock assessment in Mukura sub-county Ngora district, Eastern Uganda

Abstract

Like other countries in the Sub-Saharan Africa, Uganda and Mukura in particular are not an exception to land use/cover changes from bush land to farm land. Mukura subcounty in Eastern Uganda is one of the areas where there is a fast transformation of bush land and green areas to built-up areas and subsistence agriculture. Therefore, there is need to understand the changes which are essential for engineering, geographic and planning process. This study aimed at determining land use/cover changes and carbon stock assessment in Mukura Subcounty-Ngora District Eastern Uganda. Specifically, the study intended to; determine the historical and current land use /cover changes for the years 1984, 2004 and 2015 using a series of Landsat TM/ETM/OLIC satellite imagery. The images were interpreted using a Parametric Rule of Maximum Likelihood method using Supervised Classification technique in ERDAS IMAGIN software (2014). Secondly this study determined the most important drivers of land use/cover changes over the last 31 years; and by this the study utilized household survey questionnaires to obtain socio economic data. A total of one hundred twenty (120) household head respondents out of 5711 households in the subcounty were randomly selected for interviewing. Soil carbon stock variations in the selected major land use/cover types (subsistence farmlands, built up areas, bush lands and wetlands) were measured through soil sampling at depth intervals of 0 – 15, 15 – 30 and 30 – 45cm making a total of eighty-four (84) soil samples that were analyzed in the laboratory using standard procedures. The soil laboratory data was analyzed using descriptive and quantitative statistics including analysis of variance (ANOVA) implemented within the GENSTAT software package (VSN International 2011). Determined the total carbon stock associated with land use/cover change. Classified images were subjected to NDVI to determine AGB. This was calculated as: NDVI = (IR – R)/ (IR + R); Where IR and R represent the remotely sensed response in the near-infrared (TM4) and red waveband (TM3), respectively. The total carbon stock was calculated by adding AGB and soil organic carbon in tonnes. The study findings indicated that subsistence farming, built up areas and wetlands were the most dominant land use/cover types; with subsistence farming posting the highest increase in land acquisition by 10.7 sq km between 1984 and 2004 and 16.7 sq km between 2004 and 2015. The wetlands decreased by 2 sq km between 1984 and 2004 and by 8.4 sq km between 2004 and 2015. The studied land use/cover types showed insignificant variation (p=0.0000). Soil organic carbon stocks significantly increased in relation with sampled soil depths in all the land use/cover types. As per the land use/cover modifiers, increased human population growth pressure and high household poverty levels were the major significant drivers influencing land use/cover change (p>0.05) This study showed that most of the land use/cover types were drastically getting converted to subsistence farming which was significantly lowering variations in the amount of soil organic carbon (SOC) stock with depth due to low root density factor.