Experimental and simulated soil mineral N dynamics for long-term tillage systems in northern France

Soil N mineralization was quantified in two long-term experiments in northern France, in which no-till (NT) and conventional tillage (CT) had been differentiated for 33 years (Site 1) and 12 years (Site 2). Both sites had the same soil type but differed in crop rotation. N mineralization kinetics were assessed in situ in bare soil in both systems for 254 days (Site 1) and 555 days (Site 2) by taking frequent measurements of water and nitrate contents from soil layers and using the LIXIM calculation model. The N mineralization potential was also determined in soil samples incubated under controlled laboratory conditions. Small or nonsignificant differences in water and nitrate contents between NT and CT were apparent within the soil profiles on both sites. Net mineralization did not differ significantly between sites or tillage treatments. The amount of N mineralized from August 2003 to April 2004 was 67 +/- 10 kg N ha(-1) on Site 1 and 74 +/- 5 kg N ha(-1) on Site 2, and 161 +/- 6 kg N ha(-1) from August 2003 to February 2005 on Site 2. The kinetics of N mineralization versus normalized time (equivalent time at constant temperature of 15 degrees C and water content at field capacity) were linear during the shorter period (254 days corresponding to 120 normalized days). The slope (N mineralization rate) did not differ significantly between treatments and sites, and the average rate was 0.57 +/- 0.05 kg N ha(-1) nd(-1). The kinetics were non-linear on Site 2 over the longer period (555 days corresponding to 350 normalized days). They could be fitted to an exponential model with a slope at the origin of 0.62 kg N ha(-1) nd(-1) . The N mineralization kinetics obtained in laboratory incubations for 120-150 normalized days were also almost linear with no significant differences between treatments. Assuming that mineralization took place in the ploughed layer (in CT) or over the same soil mass (in NT) they were in good agreement with the kinetics determined in situ on both sites. The calculated water drainage below the sampled profile was slightly greater in NT due to lower evaporation. The calculated leached N was slightly higher in NT than CT on Site 1, but did not differ between treatments on Site 2. It is concluded that N mineralization and leaching in NT and CT were similar, despite large differences in N distribution within the soil profile and a slight difference in organic N stock. (C) 2006 Elsevier B.V. All rights reserved.