Attribution and driving force of nitrogen losses from the Taihu Lake Basin by the InVEST and GeoDetector models

Quantifying temporal and spatial changes in reactive nitrogen (Nr) losses from a watershed and exploring its main drivers are the key to watershed water quality improvements. Huge Nr losses continue to threaten the safety of the water environment in the Taihu Lake Basin (TLB). Here, the InVEST and GeoDetector models were combined to estimate Nr losses in the TLB from 1990 to 2020 and explore driving forces. Different scenarios for Nr losses were compared, showing that Nr loss peaked at 181.66 × 10³ t in 2000. The key factors affecting Nr loss are land use, followed by elevation, soil, and slope factors, and their mean q-values were 0.82, 0.52, 0.51, and 0.48, respectively. The scenario analysis revealed that Nr losses increased under the business-as-usual and economic development scenarios, while ecological conservation, increased nutrient use efficiency, and reduced nutrient application all contribute to a reduction in Nr losses. The findings provide a scientific reference for Nr loss control and future planning in the TLB.