Comprehensive and up-to-date vegetation degradation assessments addressing the dual challenges of climate variability and human actions in ecologically sensitive catchments are limited. Vegetation degradation assessment is crucial in establishing locally valid causal factors of vegetation demise and identifying areas at risk of ecosystem service loss. This study investigates human-induced vegetation loss within the greater uMngeni Catchment, South Africa, using a spatio-temporal approach that integrates land use transitions and the residual trend (RESTREND) technique over the period 2000–2023. The main objective is to identify human induced degradation hotspots to establish potential areas for ecosystem service loss, thereby highlighting the need for targeted restoration efforts. Land use/land cover (LULC) data were obtained from multispectral Landsat images, with land use classes generated using the maximum likelihood classification algorithm in ArcGIS 10.5. The contribution rates of precipitation and soil moisture variations were assessed using 250-m MODIS NDVI time-series data, CHIRPS rainfall data, and NASA datasets. Additionally, the intensity analysis approach quantified and analyzed transitions in LULC over the period. Intensity analysis results revealed a 25% decline in natural vegetation, particularly forests and grasslands, attributed to urban and cropland expansion. RESTREND analysis identified approximately 17.1 and 16.2% of the catchment area as significantly degraded, with severe degradation observed in rainfall- and moisture-adjusted models’ outputs, respectively. Key hotspots were identified in Durban Metro and Pietermaritzburg, where 32.7 and 29% of the municipality bounds were classified as severely degraded. These findings point out the necessity for restoration and sustainable landscape management to safeguard the catchment’s ecological functions.