Modelling soil organic matter changes in ley-arable rotations in sandy soils of Northeast Thailand

The effects of a 3-year ley-arable rotation (2 years under grass, legume or a grass-legume mixture, followed by 1 year under cassava) on the organic C content of upland sandy soils in Northeast Thailand were evaluated using four field trials (Khon Kaen, Chaiyaphum, Mahasarakham and Udon Thani). At all sites under the existing management system of continuous cassava, soil organic C decreased by 9-20% during the 3-year experiment. The carbon model developed at Rothamsted (RCM-26) predicted that soil organic C would decrease consistently under continuous cassava until it reached very small amounts (0.2-0.3%). This was the case for all sites, except Chaiyaphum, where soil organic C had already declined to 0.1%. Evaluating the rotations using RCM-26 showed that introducing a ley could increase soil organic C, except at Udon Thani where the organic C exceeded 0.47% before the start of the experiment. The data obtained with the rotations showed that RCM-26 could be used to predict the effects of changes in management on organic C content in the upland sandy soils of Northeast Thailand over the 3 years. To compensate for an underestimation of the effects of soil moisture deficit on the decomposition rates of organic matter, we propose that because of the prolonged dry season (from December to the following May) soil moisture deficit and the rate-modifying factor for moisture (A(m)) should be estimated independently of crop cover. Furthermore, to accommodate the very slow decomposition during this time, A(m) should be allowed to decrease to 0.1. Adopting this simple modification provides a more realistic estimate of the organic C inputs needed to simulate the content of soil organic C measured before trials began.