ارزيابي مدل‏هاي رگرسيون غير خطي براي پيش‏بيني سرعت به ساقه‏رفتن گندم(رقم تجن) در پاسخ به دما و طول‏روز

Evaluation of non-Linear regression models to predict stem elongation rate of wheat ((Tajan cultivar) in response to temperature and Photoperiod

The first step to quantify crop phenology is precise estimation of the parameters, which affect it. These parameters are mainly temperature and photoperiod. This study was conducted with eight sowing dates to quantify response of stem elongation rate (SER) to temperature and photoperiod in wheat (Tajan Cultivar). The regression models fitted to SER against temperature were flat, logistic, quadratic, cubic, dent-like, segmented and curve models. Meanwhile, quadratic, negative exponential and segmented models fitted to SER against photoperiod (these models were used as 21 combined models to describe SER as a function of photothermal day). Root mean square of errors (RMSE), coefficients of determination and regression coefficients of predicted values versus observed ones were used to find the most appropriate model. The results showed that segmented- segmented function was the best model to describe SER as a function of both temperature and photoperiod. Using this function, the base and optimum temperatures, critical photoperiod and photoperiod sensitivity coefficient were estimated as 4.14 °C, 28 °C, 14 h and 0.17, respectively. Study the effects of temperature and photoperiod on stem elongation phase indicated that Tajan cultivar has a quantitative or facultative long-day (LDP) response to photoperiod. Biological days required to pass from emergence to stem elongation and emergence to physiological maturity were estimated as 7.66 and 50.31, respectively.