عنوان مقاله [English]
With respect to this issue that Iran is located in semi-arid conditions and limited by water resources, so water conservation in agricultural systems plays main role to increase production and determination of water optimum amount is first step to gain this purpose. Nitrogen is one of the main effective factors on quantity and quality of crops. According to the studies, only 40-60% of nitrogen fertilizers is used by crops and this value decreases with increasing of fertilizer application. There is complicated interaction between amount of irrigation water and nitrogen fertilizer, thus it is necessary to consider optimum level of them simultaneously. To obtain acceptable economical yield and reducing environmental pollutions, used inputs in farms should be applied as optimum with respect to expected target. One of the important methods to gain optimum level of inputs is response-surface method. There is no study to investigate usage of this method for inputs optimization in sugar beet. Therefore, the purpose of the study was optimizing of nitrogen fertilizer and irrigation in sugar beet via the response-surface method by using a central composite design.
Material and Methods
We used available data and information from studies which had been accomplished about nitrogen fertilizer and irrigation in Hamedan, Iran to determine optimum levels of these treatments. So needed treatments were designed based on high and low levels of nitrogen fertilizer (0 and 240 Kg.ha-1) and irrigation (8000 and 14000 m3.ha-1) by Minitab software ver.16 as central composite design (CCD). CCD is one of the response-surface methods and the number of treatments in this design is calculated by equation of 2k + 2k + r, where k is the studied factors and r is number of replication for central point. Number of replication for central point under two factors has been reported as 5, thus for central composite design with two factors, 13 treatments is needed. To fit data, regression equation was used and evaluated based on regression variance analysis. In general, the full quadratic polynomial equation was tested to determine the significance of the model and the components of the model. RMSE, ME, R2 indexes and 1:1 line were used to judge the difference between simulated and observed data.
Results and Discussion
ANOVA results showed that regression model was significant to estimate all dependent variables based on F test. Correlation coefficient of dependent variables including root yield, sugar and sugar white, water use efficiency and nitrogen use efficiency determined as higher than 96%. It implies that the high proportion of the variability for these traits was explained by the fitted regression model. According to the lower values of RMSE than 10 and higher values of ME than 0.89, it could be concluded that the model had acceptable and suitable results to estimate studied traits in sugar beet. The results of t-test to compare fitted regression with line 1:1 illustrated that slope and intercept values in fitted and 1:1 line had no significant difference. The results showed that root, sugar and white sugar yield were increased by increasing nitrogen fertilizer under all levels of irrigation. Response-surface curve of α-amino nitrogen as affected by irrigation and nitrogen fertilizer indicated that α-amino was elevated by increasing nitrogen fertilizer application. As data, water use efficiency decreased by water consumption. In the other hand, nitrogen use efficiency was decreased by applying nitrogen fertilizer under all levels of irrigation. Optimum range of treatments were obtained as 9500-12000 m3.ha-1 for irrigation and 110-130 Kg.ha-1 for nitrogen fertilizer treatment based on overlaid plot method. The results of treatments optimization by using analytical solution method illustrated that applying 133 Kg.ha-1 and 10667 m3.ha-1 were suggested as optimum amounts of treatments. Based on these optimum levels of treatments, root yield, sugar and white sugar yield, α-amino, water use efficiency and nitrogen use efficiency were estimated as 80.1 ton.ha-1, 14.94 ton.ha-1, 12.49 ton.ha-1, 2.56 meq.100 g-1, 1.39 Kg sugar.m-3 and 74.24 Kg sugar. Kg-1, respectively.
As result, to optimize treatments including nitrogen fertilizer and irrigation, response-surface method had acceptable adequate to predict variables in sugar beet based on statistical indexes. Optimum value of nitrogen fertilizer and irrigation were predicted as 133 Kg.ha-1 and 10667 m3.ha-1, respectively by using analytical solution. Therefore, the results indicate that the application of optimum values can reduce environmental hazards and produced acceptable sugar yield.