Abstract
Leaf appearance is often simulated using the phyllochron concept that assumes a linear relationship between leaf appearance rate (LAR) and temperature. A 2-year field experiment was conducted in winter wheat ( Triticum aestivum L.) to evaluate a non-linear model and two versions of the phyllochron approach to predict LAR, and to compare the differences in predictions of leaf appearance using daily, hourly, and maximum and minimum air temperatures. In the non-linear model, daily LAR (DLAR) was simulated based on the product of temperature and photoperiod response functions. The temperature function was a beta type distribution, while the photoperiod response was an exponential function. The non-linear model accurately predicted leaf appearance, and generally had a lower root mean square error (RMSE) than the two models based on the phyllochron over the two growing seasons. The mean square error was divided into systematic and unsystematic errors. For the first season, all models had a greater systematic than unsystematic error. For the second growing season, the phyllochron models had as large or larger systematic error than the non-linear model using the different temperatures. No significant improvement in the prediction of the Haun stage using hourly air temperature was found as compared to daily mean air temperature. In general, using the maximum and minimum temperatures provided the poorest predictions of LAR and resulted in the largest RMSE because these two temperatures do not represent the mean daily temperature.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have