Traditional agricultural production in frequently cropped soils can lead to soil structural damage, a decrease in soil quality, increased erosion, and aggravation of nitrogen (N), phosphorus (P) and pesticide losses into rivers and lakes. Aiming at reducing these problems, many countries have introduced the practice of no-till. In southern China, rice farming activities involves deep plowing in spring, but in summer, cultivation involves no-till or shallow plowing. No-till means that the land is not used as intensively and direct sowing or planting crops is the preferred method of production. China from the end of 1970s began to use the no-till approach in paddy field research; in southern China in the 1980s the natural no-till method became popular to improve the environment as a whole and, as well as promoting the sustainable development of paddy field ecosystems. Summer precipitation and paddy field surface water with high N and P concentrations can be the source of the loss of a large quantity of nutrients from runoff from the rice fields, which has the potential to be a source of pollution and can affect the water quality of the local river and lakes.
In summer a different farming mode involving changing the paddy field surface water is used, which can reduce N and P losses. The dynamic changes in N and P concentrations and their loss potentials and reduction effectiveness in paddy field surface water, under different stagnating times in three tillage models such as no cultivation, shallow plowing and deep plowing were investigated. This was done using a paddy simulation microzone experiment. Results showed that: (1) Deep plowing was favorable for the fixation of fertilizer, by the soil but total N and NH⁺₄-N in the water showed a gradual decrease with time. The microbial environment in the soil of non-cultivated and shallow plowing scenarios was favorable for nitrification, to which NO^-₃-N was released rapidly into the water as it was poorly adsorbed by the soil. The total P (TP) and dissolved P (DP) concentrations in the water of the non-cultivated and deep plowing systems were relatively high within 1-5 d, and the TP and DP concentration (when the discharge of the water from the three tillage regimes was delayed) was relatively low after 5 days. (2) The absolute TN losses from the water of the different tillage models were low after the water was left for 5 days. Non-cultivation, shallow plowing and deep plowing reduced the TN loses by 59.6%-65.7%, 70.2%-88.2% and 65.2%-77.3%, respectively. Total-N potential losses into the water of the deep plowing regime were relatively minor but were the main form of N lost in the water of all three farming scenarios. (3)The absolute losses of TP in the water of the non-cultivated regime were the highest while they were was lowest in the water of the shallow plowing system. After holding the water for 5 days, the TP lost (as calculated by the three tillage models) was reduced in the range of 54.7%-67.8%, 63.0%-85.1% and 52.5%-88.0%, respectively. The relative potential losses of TP in water of the shallow plowing regime were lower than others. Likewise, the relative form of N and P lost was different in three tillage models and showed variations over time. Thus, when discharged after 5 days' delay, the amount of N and P in the water lost from the paddies can be decreased effectively, which will significantly reduce the effect of paddy water discharge as an agricultural non-point source of pollution.