Nonlinear response of network resilience to land-use intensity for water purification services in a large lake basin

With the intensification of agricultural land use in large lake basins, the role of water purification networks in reducing non-point source pollution has become increasingly critical. The networks' resilience offers a systematic framework for addressing water pollution risks. However, the nonlinear response of water purification network resilience to land-use intensity remains poorly understood. This study focused on agricultural nitrogen (N) application in the Dongting Lake (DTL) Basin, integrating the water purification deficit (N exposure) delivery pathway with population density to construct a spatially explicit water purification network. From the perspective of disturbance, resistance, and recovery, we identified the spatial clustering characteristics of water purification network resilience and revealed its nonlinear response to land-use intensity. The results indicated that the DTL surrounding area and the lower reaches of the Xiangjiang River Basin exhibited the highest N exposure (over 60 kg·person/ha²) and N delivery (over 10 kg/ha) in the water purification network. Regarding network resilience, from 1990 to 2020, sub-basins with high disturbance (over 0.80) initially clustered around the DTL surrounding area before shifting outward, suggesting an initial increase followed by a decline in the risk of water quality degradation in the DTL surrounding area. Overall, as agricultural N application increased, the growth rates of both recovery and resistance within the network gradually declined, while the growth rate of disturbance steadily increased. The total N application within the sub-basin itself and in upstream sub-basins was a key factor influencing the resilience of the water purification network. This study provides systematic insights into water quality management in large lake basins from a network resilience perspective, based on N delivery processes across sub-basins.