RESUMEN
Abundant evidence has demonstrated the feasibility of reducing phosphorus (P) input to face diminishing phosphate rock resources and deteriorating environmental quality in double-cropping paddy. However, the sustainability of reduced P input in the context of maintaining productivity and P efficient utilization is not yet clear. Herein, an 8-year (2013-2021) field-based database was built to explore the effects of reduced P input on rice productivity and the soil-plant P trade-off in double-cropping paddy. In the early and late rice seasons, compared with conventional P fertilization (early rice, 90 kg hm-2; late rice, 60 kg hm-2), the average yield of reduced 10 % P treatment increased by 4.3 % and 2.1 %, respectively; reduced 10-30 % P treatments increased average P use efficiency by 17.1-18.4 % and 14.0-17.2 %, decreased average total P runoff loss by 14.9-33.2 % and 20.8-36.4 %, and decreased average total P leaching loss by 18.5-49.0 % and 24.0-46.1 %, respectively. Compared with conventional fertilization, reduced P fertilizer input by 10 % significantly increased the content of the soil labile-P fraction while reducing that of the soil stable-P fraction. Soil ligand-P and exchangeable-P content decreased with the gradient reduction of P fertilizer input (10-30 %). The main predictors of the change in rice yield and plant P uptake were soil ligand-P and exchangeable-P content, respectively. The dominant predictor of both the P runoff loss and the P activation coefficient was the inorganic P content extracted by NaHCO3. These findings suggest that reduced P input by 10 % could maintain rice productivity and P use efficiency in the double-cropping paddy, and the transformations between soil P components and increases in P bioavailability may be the key drivers maintaining rice productivity and P utilization under the context of reduced P loading.