Urban expansion and loss of prime agricultural land in Sub-Saharan Africa: a challenge to soil conservation and food security

Abstract

Urbanisation often involves the conversion of various land use and land cover (LULC) classes including agricultural land to urban uses and leads to loss of soil diversity. Unfortunately, there is limited literature on soils lost in Sub-Saharan African (SSA) cities due to lack of detailed soil information yet this region is the fastest growing globally. This study attempts to bridge this gap by using remote sensing data and digital soil mapping (DSM) techniques to assess the rate of agricultural land conversion, determine the appropriate pre-urban soil prediction model and estimate the extent of loss of soils diversity and urban and peri-urban agriculture (UPA) using urban centres in Uganda as case studies. Multi-temporal LULC classification of Landsat ETM+ and TM images showed that built-up area in Kampala expanded 8 times between 1989 and 2015 and 5 times in Mbarara between 2002 and 2016 as a result of the conversion of savannah, wetlands and systematic targeting of agricultural land. DSM techniques involving legacy soil data and soil observations were used to predict the pre-urban soil patterns by modelling the relationships between observed soil classes and environmental covariates using random forests (RF), Multinomial Logistic Regression (MLR) and Boosted Regression Trees (BT) algorithms. The overall prediction accuracy was over 70% producing soil maps at 30m resolution. The soil diversity loss was determined by overlaying RF soil map with the multi-temporal LULC maps. The results show that 74% of the soils lost were in high and medium quality class for agricultural production. Moreover, the average size of farms under UPA decreased from about 1.9 acres in 2002 to about 0.5 acres in 2015. This study has revealed that DSM techniques and remote sensing can be useful in quantifying the loss of soil diversity to urbanisation and provides quantitative evidence that rapid urbanisation could lead to loss of good soils and increase food insecurity in SSA cities.