RESUMO
Enhanced release of phosphorus (P) from soils with snowmelt flooding poses a threat of eutrophication to waterbodies in cold climatic regions. Reductions in P losses with various soil amendments has been reported, however effectiveness of MgSO4 has not been studied under snowmelt flooding. This study examined (a) the P release enhancement with flooding in relation to initial soil P status and (b) the effectiveness of MgSO4 at two rates in reducing P release to floodwater under simulated snowmelt flooding. Intact soil monoliths were collected from eight agricultural fields from Southern Manitoba, Canada. Unamended and MgSO4 surface-amended monoliths (2.5 and 5.0 Mg ha-1) in triplicates were pre-incubated for 7 days, then flooded and incubated (4 °C) for 56 days. Pore water and floodwater samples collected at 7-day intervals were analyzed for dissolved reactive P (DRP), pH, Ca, Mg, Fe and Mn. Redox potential (Eh) was measured on each day of sampling. Representative soil samples collected from each field were analyzed for Olsen and Mehlich 3-P. Simulated snowmelt flooding enhanced the mobility of soil P with approximately 1.2-1.6 -fold increase in pore water DRP concentration from 0 to 21 days after flooding. Mehlich-3 P content showed a strong relationship with the pore water DRP concentrations suggesting its potential as a predictor of P loss risk during prolonged flooding. Surface application of MgSO4 reduced the P release to pore water and floodwater. The 2.5 Mg ha-1 rate was more effective than the higher rate with a 21-75% reduction in average pore water DRP, across soils. Soil monoliths amended with MgSO4 maintained a higher Eh, and had greater pore water Ca and Mg concentrations, which may have reduced redox-induced P release and favored re-precipitation of P with Ca and Mg, thus decreasing DRP concentrations in pore water and floodwater.