ارزیابی توابع کاهش جذب آب کاهو برگی (Red Salad Bowl) تحت تنش خشکی در شرایط گلخانه‌ای

مدل‎های جذب آب اگر بتوانند مسیر حرکت آب و واکنش گیاه به تنش را به‌طور صحیح پیش‌بینی کنند، ابزاری مفید در بهینه‎سازی مصرف آب محسوب می‎گردند. هدف از تحقیق حاضر بررسی مدل‎های جذب آب کاهوبرگی تحت تنش خشکی می‎باشد. برای این منظور آزمایشی در گلخانه دانشکده کشاورزی دانشگاه شهید چمران اهواز، در پاییز 1396 انجام شد. آزمایش با سه تیمار آبی 60، 80 و 100 درصد نیاز آبی گیاه در سه تکرار و در قالب یک طرح کاملاً تصادفی انجام گرفت. برای تعیین زمان آبیاری از روش وزنی استفاده گردید. تعرق نسبی با استفاده از تغییرات رطوبت روزانه خاک محاسبه و مدل‌های کاهش جذب آب فدس و همکاران (1978)، ون گنوختن (1987)، دایرکسن و همکاران (1993) و همایی (1999) مورد ارزیابی قرار گرفتند. از شاخص‌های آماری بیشینه خطای نسبی، ضریب تعیین، ریشه میانگین مربعات خطا، کارایی مدل‌سازی و ضریب جرم باقیمانده استفاده شد. نتایج نشان داد، مدل همایی (1999) نسبت به سایر مدل‎ها برازش بهتری ارائه می‎دهد (0.96=r^2 و 9.14=rmse). ضمن اینکه مدل‌های فدس و همکاران (1978) با 0.43=r^2 و 16.46=rmse، ون گنوختن (1987) با 0.51=r^2 و 8.62=rmse و دایرکسن و همکاران (1993) با 0.48=r^2 و 12.5=rmse، برازش نزدیکی نسبت به هم داشتند.

Evaluation of Water Uptake Reduction Functions of Lettuce (Red Salad Bowl) under Drought Stress Conditions

Introduction: when the loss of water from the leaves by transpiration process exceeds the water in the root zone, water stress occurs. If water uptake reduction functions can predict the flow of water to the roots properly, with no need to field measurements, it is possible to determine the time of irrigation for maximum yield with the help of chemical and physical properties of water and soil and specific plant parameters. From a conceptual point of view, two main approaches of water uptake modeling exist, which differ in the way they predict the volumetric rate of root water uptake. The microscopic models describe the water flow towards the single root. Macroscopic models are based on the principles of energy and mass transfer and described water uptake by the entire root area, regardless of the impact of individual roots. In general, macroscopic models have been considered in plant growth and soilplantatmosphere models. The most important models of macroscopic water uptake can be referred to the model of Feddes et al. (1978), the VanGenuchten model (1987), the Dirksen et al. (1993) and Homaee (1999). Saraei Tabrizi et al. (2015) with an evaluation of the water uptake reduction functions under water stress conditions on basil plant showed that the Homaee (1999) model was more suitable than other models. The purpose of the present study is to evaluate the four macroscopic water uptake reduction functions of Feddes et al., (1978), Van Genuchten (1987), Dirksen et al., (1993) and Homaee (1999) in order to predict water uptake by lettuce root and determining the most suitable model to predict the reduction of water uptake of lettuce under water stress conditions.  ;Materials and Methods: The experiment was conducted in a completely randomized design with three replications in 2017 in the research greenhouse of Agriculture Faculty, Shahid Chamran University of Ahvaz. The experiment consisted of irrigation water at three levels (I1:100%, I2:80% and I3:60% of crop water requirement). The soil texture was medium. The cultivation was indirect (seedling) in pots of diameter 22 and height 30 cm. For this purpose, 9 pots were used. The plant 's growth period was about 70 days. The irrigation was done by the manual method and by the graduated bushel. During the growing season, treatments were irrigated fifteen times. The volume of water used for treatments I3, I2 and I1 were respectively 792, 1055 and 1320 )m3.ha1(. The weighted method was used to determine the time of irrigation and the soil moisture characteristic curve was used to measure the matric potential. 12 pots were considered as destructive ones for measuring plant weight as it was not possible to measure the weight of plants per day. In order to evaluate the water uptake reduction functions, relative transpiration was plotted against the absolute value of matric potential and the best model was determined by fitting them, to the measured data. For this purpose, the statistical indicators of the Maximum Error (ME), coefficient of determination (R2), Root Mean Square Error (RMSE), modeling efficiency (EF) and Coefficient of Residual Mass (CRM) were used.;Results and Discussion: Based on the results, the model of Homaee (1999) and the model of Van Gennuchten (1987) had the best fit on the whole range of measured data, respectively. Then, the model of Dirksen et al. (1993) and, the model of   Feddes et al. (1978) were   ranked. In addition, in models that |h| was smaller than 8000 cm, like Feddes et al. (1978), Van Gennuchten (1987), Dirksen et al. (1993) showed good fit and proximity to each other. Babaazadeh et al. (2017) in studying the effect of salinity and drought stress on the uptake of root water of   basil, concluded that the Homaee (1999) model had the best agreement with experimental data and increasing drought stress reduced the potential of water uptake by roots. Also Saraei Tabrizi et al. (2015) concluded that the Homaee (1999) model had the best fit with the measured data and the results were in accordance with the results of this study. Based on the results of this study, for simulation of water uptake, the models of Feddes et al. (1978) and Dirksen et al. (1993) are slightly overestimated and Van Gennuchten (1987) and Homaee (1999) models have slightly underestimated. Homaee (1999) in his research was conducted that in treatment of 70% water requirement supply for alfalfa, Feddes et al. (1978) model was overestimated and other models were underestimated which are close to the results. Homaee (1999) model was more consistent compared to other water uptake models because of considering two thresholds for the model.;Conclusion: According to the results, Homaee (1999) model was better than other models (RMSE=9.14 and )‎. The results of the models of Feddes et al. (1978) with R2 =0.43 and RMSE = 16.46, Van Gennuchten (1987) with R2 = 0.51 and RMSE = 8.62 and Dirksen et al. (1993) with R2 = 0.48 and RMSE = 12.5 were closely related to each other.