A graphical abstract for assessing meteorological, agricultural, and hydrological drought characteristics, trends, propagation in a semi-arid catchment showing input data, methods and results. Semiarid catchments in Tanzania are vital for livestock grazing but highly vulnerable to recurrent drought. This study characterizes meteorological, agricultural, and hydrological droughts (1984–2022) across 3-, 6-, and 12-month timescales, using the standardized precipitation index (SPI), standardized streamflow index (SSI), and standardized soil moisture index (SSMI), respectively. Normalized Difference Vegetation Index (NDVI) anomalies were used as a proxy for pasture condition. Results indicate a unimodal rainfall regime (November–April) and persistent dry-season water deficits (June–October) driven by high evapotranspiration. Short-term droughts (SPI-3 and SSMI-3) occur frequently (0.9–1.8 events year−1), whereas SSI-3 are less frequent but longer-lasting (duration: 5–16 months). Long-term trends are weak overall, although a significant decline in streamflow is observed at the catchment outlet. Propagation is strongly timescale-dependent: short-term rainfall deficits rapidly translate into soil moisture anomalies in lowlands (SPI-3 to SSMI-3, lag 0–1 month, r ≈ 0.5–0.6), while SSI reflects cumulative multiseason deficits. NDVI responds most strongly to SPI-3 and SSMI-6 (lag 0–2 months; r up to 0.48), indicating rapid pasture sensitivity. These findings highlight a narrow early-warning window for forage decline and support adaptive strategies including timely destocking, rotational grazing, and improved dry-season water management to enhance resilience under increasing drought variability.
Graphical abstract showing Usangu Catchment map, data analysis workflow for rainfall, streamflow, and livestock, key results on hydrological trends, and conclusions on dry-season flow recovery after livestock eviction. This study examines climatic and hydrological responses to the 2006 livestock eviction in Tanzania's Usangu Catchment using 1984–2022 records of livestock numbers, rainfall, temperature, evapotranspiration (ET), and streamflow. Trends were assessed using Mann-Kendall and Sen's slope estimators, regime shifts with the Pettitt test, and rainfall -streamflow relationships through correlation and regression. Climate analysis revealed significant warming in both highlands (0.015–0.035 °C/month) and lowlands (up to 0.043 °C/month), with ET increasing significantly in highlands but showing weak and inconsistent trends in lowlands. Rainfall remained largely stable, except for a localized significant decline at Kimani after 2009. Prior to eviction, streamflow exhibited significant dry-season declines, baseflow instability, and weakened rainfall -streamflow linkages, particularly at the Usangu outlet. Following eviction, hydrological conditions improved, with enhanced low flows, recovery of high-flow indices, and reversal of annual streamflow trends at the outlet from decline (−76.97 Mm³/year) to increase (+31.06 Mm³/year). Rainfall -streamflow sensitivity remained weak to moderate but showed localized improvements in Mbarali and Upper Great Ruaha. These findings demonstrate that livestock removal contributed to partial restoration of dry-season flows. This recovery underscores the need for sustained protection of wetlands and regulation of water abstraction to secure long term hydrological resilience.