马莲河下游流域是我国陇东典型的黄土高原沟壑区，在气候和土地利用变化背景下研究其产水量的时空特征和变化响应，对区域生态系统可持续发展具有重要意义。基于InVEST （Integrated Valuation of Ecosystem Services and Trade-offs）模型定量评估1990年、2000年、2010年和2020年4个时期马莲河下游流域产水量的时空格局和变化，运用地理探测器分析影响因子对产水功能空间分异的影响程度。结果表明：（1） 1990—2020年，马莲河下游流域总产水量整体呈现先减后增再减的变化趋势，2020年总产水量相比1990年减少了5.9×10⁷ m³，减少率为25.43%。不同时期产水量在空间上呈现南部和边缘地带高、北部和中心地带低的分布特征。（2） 不同土地利用类型产水能力大小排序依次为：城镇用地>未利用地>耕地>低覆盖度草地>高覆盖度草地>灌木林>有林地>水域。（3） 产水量与降水量呈现明显的正相关，与实际蒸散发量和海拔之间存在负相关。降水和实际蒸散发是决定产水量空间分布和变化的主导因素，q值分别为0.616~0.735和0.517~0.653。研究成果可为陇东黄土高原沟壑区水土资源开发、利用和管理提供科学支撑。

This study examines water yield patterns in the lower reaches of the Malian River Basin, which are typical loess plateau gully areas in eastern Gansu Province, China. These areas are crucial for regional ecosystem health and understanding the temporal and spatial characteristics and response of water yield considering climate and land-use changes. Employing the Integrated Valuation of Ecosystem Services and Trade-offs (InVEST) model, we quantitatively assess the spatial and temporal patterns of water yields across four years: 1990, 2000, 2010, and 2020. Spatial heterogeneity in water yield across the lower reaches of the Malian River Basin was assessed using geographic detection techniques. Analysis revealed a fluctuating trend in total water yield from 1990 to 2020, characterized by an initial decrease, followed by an increase, and finally another decrease. Compared to 1990, total water yield in 2020 declined by 5.9×10⁷ m³ (25.43% reduction). Furthermore, spatial analysis revealed a distinct pattern in water yield distribution across the basin. Higher water yield was observed in the southern and marginal areas, whereas lower yield characterized the northern and central regions. Land-use type significantly influenced water yield capacity. Ranked from highest to lowest, the order was as follows: construction land>bare land>agricultural land>low-coverage grassland>high-coverage grassland>shrub land>forest land>open water. Moreover, a significant positive correlation was identified between water yield and precipitation, suggesting that precipitation plays a key role in water production. Conversely, a negative correlation emerged between actual evapotranspiration and altitude. Our analysis identified precipitation and actual evapotranspiration as the primary drivers of spatial variations and temporal changes in water yield, with q values ranging from 0.616 to 0.735 and 0.517 to 0.653, respectively. These findings provide valuable scientific evidence to support the development, utilization, and management of soil and water resources in the loess plateau of eastern Gansu.