TILLAGE EFFECTS ON PLANT EXTRACTABLE SOIL-WATER IN A SILTY CLAY VERTISOL IN SOUTHERN ITALY

This paper describes how the CERES-Wheat simulation model can be used to estimate tillage effects on soil water regimes of a silty clay soil in Foggia, Southern Italy. The four tillage treatments compared are t raditional mouldboard ploughing, ripper subsoiling, surface disc-harro wing and minimum tillage with rotary hoeing under continuous durum whe at cropping. For each tillage treatment the CERES-Wheat model was used to calculate the water balance for several layers in the root zone. T he water balance routine of the model estimates the water content with in saturation and the lower limit at any time, Inputs required by the model are some basic information about the site, weather, genetic para meters and management practices as well as some soil properties, such as albedo, bulk density, organic matter and N contents, The model was calibrated by estimating the genetic parameters for the minimum tillag e treatment in the season 1984-85. The same set of parameters was used for the subsequent validation procedure, Statistical tests proved tha t the match between measured and simulated soil water content values w as quite good, The simulation results also showed some differences amo ng different tillage treatments. The model predicted the lowest plant extractable soil water values and a different water content distributi on along the soil profile of the ripper subsoiling in comparison with the other tillage treatments. The soil water content was lower until 2 0-40 cm depth and higher at 40-60 cm depth in the ripper treatment as compared with the others. In deeper layers differences became non-sign ificant, This might be due to the cracks produced by the ripper throug h which rainfall infiltrated in deep layers.