| Summary: | Simulation models based on plant physiology are used to predict growth and yield of crops. Such models are important because they can be used to pre-evaluate treatments, thus, improving the effectivity of agricultural research and reducing the cost of field experiments. For efficiency purpose, crop models need to be calibrated and validated before using them. The objective of this study was to evaluate the performance of an existing crop model in predicting the yield of East Africa Highland Banana (EAHB) under deficit irrigation and different irrigation intervals. The model, CROPWAT 8.0, was calibrated, evaluated and applied for banana crop water requirements and estimation of EAHB yield. For calibration of CROPWAT 8.0, monthly climatic data (temperature, relative humidity, wind speed, sunshine hours and rainfall), crop and soil data are were used. Climatic data were provided by the New_LocClim software which is the local climate estimator of FAO, effective rain was set to zero because the experiment was conducted under a rain shelter. Three irrigation levels (IL) (80%, 90% and 100% of Evapotranspiration) were combined with three levels of irrigation intervals (D) (4, 6 and 8 days in a randomized complete block design (RCBD) with three replications. To evaluate the model for yield estimation, the observed yield was compared with the corresponding simulated values by CROPWAT 8.0 using mean squared deviation (MSD), Nash and Sutcliffe model efficiency (NSE), coefficient of determination (R2) and paired t-test. The predicted banana yield (39.1 ± 2.66 t ha-1) from the calibrated model was very close to the observed yield (38.4 ± 2.37 t ha-1 (p≥0.05, R2 = 0.82 and an NSE of 0.81. MSD analysis showed that the models prediction was more accurate at 8 or 6 days irrigation intervals than 4 days irrigation interval. The calibrated CROPWAT 8.0 model can be used efficiently to predict the yield of East Africa Highland Banana. [Fundam Appl Agric 2020; 5(3.000): 344-352]
|
|---|
