Forecasting Yield and Tuber Size of Processing Potatoes in South Africa Using the LINTUL-Potato-DSS Model

Abstract

The LINTUL-Potato-DSS model uses the linear relationship between radiation intercepted by the crop and radiation-use efficiency (RUE), to calculate dry matter production. The model was developed into a yield forecasting system for processing potatoes based on long-term and actual weather and crop data. The model outcome (Attainable yield, Yatt) was compared to actual yields (Yact) of a summer crop in South Africa and the ratio Yact to Yatt was used for forecasting yield in winter crops. Results showed that accurate forecasts (<20% variation between the actual and forecasted values) could be produced already early in the growing season, and that for the cultivar Innovator, actual and forecasted yields were well correlated (r = 0.797). Forecasted and observed yields at harvest were not significantly different at the 5% level, P = 0.637 (t test). Forecasts of tuber number using LINTUL-Potato-DSS were not accurate in the present study and further research is needed on this aspect. It is concluded that the model is a valuable management tool that can be used to produce accurate forecasts of tuber yield from as early as 8 weeks before the final harvest. Since the model was tested with only one cultivar grown in three different growing regions of South Africa, further evaluation using different cultivars and localities is recommended.