Search

We are still challenged to fully understand to what extent controlling irrigation inputs is important for limiting phosphorus (P) losses from paddy soils and how this affects the contribution of dissolved P (DP), particulate P (PP), and colloidal P (Pcoll) fractions in P losses. For this investigation, an internet-based smart irrigation system was developed to precisely implement different irrigation scenarios––flooding irrigation (FI) and two water-saving irrigation (WSI) scenarios of intermittent (II) and controlled irrigation (CI). A simulation experiment was conducted and the concentrations of total P (TP), DP, PP, and Pcoll released to floodwater (FW) and leachate (LW) were determined during a rice season after the addition of different P forms (chemical fertilizer (CP), biogas slurry (BP), and manure (MP)) at a rate of 40 kg P·ha−1. The results revealed that the CI scenario effectively decreased P losses in FW (∼51%) and LW (∼63%) with a positive WSI ratio (39.1%), while II reduced it in FW by ∼11% and in LW by ∼27% with a WSI ratio of 11.0% compared with FI. The BP and MP additions caused a higher risk of P losses than CP under FI, with lower risk under WSI. DP largely contributed to P losses (FW: 30%–71%; LW: 42%–66%); however, more PP was released to FW under II (37%–54% of TP). The Pcoll (FW: 5%–23% and LW: 6%–24%) was more sensitive to release at initial periods after fertilization on account PP. This work provides a promising approach to timely control irrigation inputs, which could limit P losses from paddy soil; however, there remains a need to assess and generalize our initial evidence at field scales.