Agriculture-driven land transformation: Predicting future land use changes in Makoni District, Zimbabwe using landsat data and cellular automata

Abstract

Predicting future land use and land cover (LULC) change is essential for sustainable development planning, environmental monitoring, and agricultural policy formulation. Yet in Zimbabwe, few studies quantify agriculture-driven land transformation at district scale or assess its long-term ecological implications. This study analysed historical LULC dynamics and modelled future changes in Makoni District from 2000 to 2040. Using Landsat imagery (2000, 2010, 2020) classified with a Random Forest algorithm, and a CA-ANN model incorporating elevation and slope, we quantified past trends and simulated future scenarios. Results show continuous cropland expansion alongside substantial forest and grassland decline, with projections indicating further natural vegetation loss and shrinking water bodies by 2040. These findings provide new evidence that agricultural intensification is driving systematic landscape fragmentation and emerging hydrological stress. The study highlights priority areas for intervention and offers insights supporting sustainable agriculture, land restoration, and climate-resilient planning.