The soybean (Glycine max L.)-wheat (Triticum aestivum, L.) cropping system is the dominant cropping system on Vertisols of Madhya Pradesh in central India, but the yield of soybean has been decreased significantly over recent years due to biotic and abiotic factors. The yield of wheat following soybean in this region is also governed by the availability of irrigation water. So managing this cropping system in terms of nutrient and water in Vertisols is important for soybean and wheat production. The proposed research work was undertaken to analyze N release from different organic materials like rice (Oryza sativa, L.) and wheat straw, green manure (like Gliricidia) and farmyard manure (FYM), modeling N release using APSIM (Agricultural Production System Simulator). The APSIM model has been calibrated for first time for Indian FYM, and also tested for the soybean-wheat cropping system permitting prediction of N release from organic amendments varying in C/N ratios and biochemical properties.
A laboratory incubation experiment was carried out for 98 days at 30º C under aerobic conditions to study the effects of rice and wheat straw applied at 5 and 10 g kg-1 to a Vertisol soil in the presence or absence of additional N (as urea). The study showed an interactive effect between the rate of application of the residues and additional mineral N. Without additional N, the mineral N in soil was completely immobilised within two weeks, irrespective of the rate of application; initial net immobilisation was limited by the availability of N and was independent of residue rate. When additional N was supplied, initial net immobilisation was dependent on the rate of application of the residue.
We used the APSIM SoilN module to simulate N mineralisation from high C/N ratio crop residues, and compared the predictions with the observed data from our incubation study and other published data sets. Model performance was generally satisfactory with a modelling efficiency of 0.82. The major discrepancy between the modelled and the observed data was a tendency for the model to underestimate the initial rate of immobilisation.
Application of Gliricidia induced N mineralisation from the start of incubation period, with the amount of N mineralised increasing with rate of application. In contrast, application of FYM resulted in immobilisation of mineral N in soil, irrespective of the rate of application. The initial net immobilisation from FYM was limited by availability of N in the soil for the higher rate of application. The prediction of N mineralised from Gliricidia by the APSIM N module was better than FYM. For better understanding of the influence of FYM composition on the rate and extent of N mineralisation, it was necessary to parameterise iii the model based on the FYM composition. Water soluble C and N components were used to specify the conceptual pool FPOOL1 and the measured lignin content to specify the C content of FPOOL3. Using these values and an iterative approach to identifying an appropriate value for FPOOL2, a more accurate prediction was obtained. The performance of APSIM model in simulating the N mineralisation from different rates of Gliricidia and FYM was satisfactory once appropriately parameterised. The ability to include this type of simulation of soil processes in crop simulation modelling could provide viable options for optimizing N management strategies involving utilization of crop residues, green manures and FYM in cropping systems.
The APSIM model was parameterised for soybean (cv JS 335) and wheat (cv Sujata), the dominant cultivars used in the soybean-wheat cropping system in Madhya Pradesh. The model was validated for this system using data from a long-term fertiliser experiment involving use of organic and inorganic sources of N. The model prediction of grain yield and N uptake for soybean under inorganic and organic treatments was satisfactory. For wheat, the model predicted grain yield and N uptake well in all treatments. The discrepancy between the observed and predicted data was attributed to the effects of factors such as diseases and pests, which were not considered by the model. In wheat, the predicted crop yield was most strongly influenced by variation of amount of irrigation and N used during its growing season.
After the model was parameterised and validated, the use of the model in long-term simulation of grain yield of soybean and wheat, N (from organic, inorganic and integrated sources) and water management of soybean-wheat system was studied. Using the model, it was possible to assess the soil and crop management strategies, like optimum dates of sowing, irrigation amount and timing, and manipulating the application of irrigation water based on availability and FYM as N source. Simulation suggested that application of FYM to soybean resulted in greater wheat yield and smaller N losses in soybean-wheat system. Integrated use of FYM and inorganic fertiliser-N reduced the loss of N from the system. The simulation study also indicated the frequency and amount of FYM to be applied in order to maintain long-term yield sustainability of the system.