Determining Irrigation Scheduling for Cotton and Peanut Using Cropping System Models

Irrigation scheduling is an important management practice for farmers who grow irrigated crops. Effective irrigation is possible by regular monitoring of soil water and crop development conditions in the field. However, this methodology requires frequent field visits and, consequently, it is time consuming. Computer simulation models can be an important aid for irrigation scheduling, as they integrate the soil-plant-atmosphere complex. The main objectives of this study were to determine the impact of different irrigation scheduling regimes on cotton and peanut growth and development and to evaluate the application of a crop simulation model as a tool for irrigation scheduling. Two experiments were conducted in four rainout shelters, located at the Griffin Campus of The University of Georgia, during 2005 and 2006. Cotton was grown in 2005 and peanut was grown in 2006. The CSM-CROPGRO-Cotton and CSM-CROPGRO-Peanut models were used to define the irrigation treatments by estimating the timing of irrigation and the amount of water to apply. The irrigation event was triggered when the actual soil water content in the effective root zone dropped below a specific threshold of the available water content (AWC) and then irrigation was applied until the soil water reached 100% of AWC. The irrigation treatments corresponded to 30%, 40%, 60% and 90% of the irrigation threshold (IT). The models require daily weather data, including maximum and minimum temperature, solar radiation and precipitation as input. Actual weather data were used until the current date and the daily weather data for the past 10 years were used to project until the end of the growing season. The cotton cultivar DP 555 BG/RR was planted on May 17, 2005 and the peanut cultivar Georgia Green was planted on May 22, 2006. For both experiments, the soil water content was monitored with Time-Domain-Reflectometry (TDR) and with a soil profile probe (PR2/6). The TDRs and PR2/6 probe were connected to data loggers. For both crops, growth analysis, including leaf area index (LAI), plant height and biomass, was conducted approximately every 18 days. Yield and yield components were obtained at final harvest. For cotton, the 30 and 40% IT treatments had significantly less number of bolls/m². The statistical analysis for lint yield showed that there were no significant differences between the 60 and 90% IT treatments. For cotton and peanut, low values of aboveground biomass at harvest and yield were found for the 30 and 40% IT treatments. The study showed that the dynamic crop growth models CSM-CROPGRO-Cotton and CSM-CROPGRO-Peanut can be used for irrigation scheduling, but a variable irrigation management depth needs to be incorporated in the model and a correct characterization of the soil properties is needed.