Assessing the Effect of Urban Expansion on Forest Degradation in Huye District, Rwanda

Abstract

Urban expansion is among the principal drivers of land-use and land-cover (LULC) change in tropical secondary cities, converting forest and agricultural land into built-up area and eroding the ecosystem services on which surrounding populations depend. This study applied a mixed quantitative-geospatial and socio-institutional design to assess the effect of urban expansion on forest degradation in Huye District, a secondary city in Rwanda's Southern Province, over 2000–2025. Supervised classification of multi-temporal Landsat imagery (2000, 2015 and 2025; 30 m resolution) in ArcGIS 10.8 produced five LULC classes; post-classification change detection quantified area transitions, and the Normalized Difference Vegetation Index (NDVI) characterised changes in forest density and vigour. A household survey of 150 respondents, proportionally allocated across five urbanising sectors following Yamane's formula, together with 22 purposively selected key-informant interviews, supplied socio-economic and institutional context; Pearson correlation analysis in RStudio examined the association between built-up expansion and forest cover. The combined residential and urban-zone share increased from 39.15% in 2000 to 40.69% in 2025, while forest cover fell from 654.85 ha to 601 ha (-8.22%), a net loss of 53.85 ha. Across the three classified reference years (2000, 2015 and 2025), built-up share and forest-cover share were strongly inversely associated (r = -1.00, p = 0.002, n = 3). Change-detection matrices identified residential-to-urban-zone (27.2%) and forest-to-agriculture (19.3%) as the dominant land-cover transitions. Respondents attributed forest degradation primarily to settlement expansion (36.6%) and charcoal production (31.7%), and urban growth itself to population growth (39%) and rural-urban migration (26.8%); a majority (71.4%) considered enforcement of forest-protection regulations effective, though 14.3% reported persistent gaps. These patterns parallel evidence from Masaka Sector and other Rwandan secondary cities, indicating a broader national trajectory of urbanisation-driven forest loss that is nonetheless comparatively moderate in Huye. Because the correlation analysis rests on only three temporal observations, it should be read as a strong descriptive association rather than a fully powered inferential test; we recommend annual or near-annual multi-temporal imagery in future monitoring to enable more robust trend modelling. Strengthened land-use zoning enforcement, reforestation, and expanded access to alternative domestic energy sources are recommended to reconcile continued urban growth with forest conservation in Huye District.