Optimization of Application Levels of Irrigation Water, Superabsorbent Polymer and Cattle Manure Using Response-Surface Methodology: A Study on Sesame (Sesamum indicum L.) in an Ecological Cropping System

IntroductionDue to climate change, increasing and maintaining the current level of production in low rainfall conditions in semi-arid regions is an important challenge. On the other hand, agriculture in these areas is often of low productivity due to low water use efficiency. Considering the fact that Iran is one of the challenging arid and semi-arid regions of the world, so it faces the problem of water shortage and precipitation. Therefore, in such circumstances, finding eco-friendly solutions to increase water use efficiency to achieve sustainable agricultural goals seems necessary. In recent years, many efforts have been made in this field, in this regard, limited irrigation and the use of superabsorbent into the soil as two practical strategies for water conservation and optimal use have been considered.Optimizing the effective factors in production and management of water use in the farm could saves limited water resources and protects the soil, moreover, can increase yields. Extensive research has been done on the effect of different levels of irrigation, superabsorbent and fertilizer on different crops, but in the field of simultaneous optimization of these factors using statistical techniques, there is little or no information, so this study aims to optimize consumption of irrigation water, superabsorbent and cattle manure were designed and conducted in low-input sesame cultivation using surface-response methodology and Box-Behnken design.Materials and MethodsThe experiment was conducted as a Box-Behnken design. Experimental factors including a combination of different levels of irrigation water, superabsorbent and cattle manure. The total number of treatments required for this experiment was 15 treatments including 12 factorial combinations and 3 replications of the central point. The Box-Behnken design is essentially applicable and analyzed with one replication, but to fit the level of response equations, it is necessary to repeat the central points that represent the average level of high and low levels of each of the experimental factors. Using the Box-Behnken design, it is possible that most information from the minimum executive operation would be obtained through the distribution of trial points in the treatments. The values of these factors were determined by using software due to low and high levels of irrigation water (1500 and 3000 m3 ha-1), superabsorbent (0 and 160 kg ha-1) and cattle manure (0 and 30 ton ha-1) using software.In the Box-Behnken method, the response variable (y) is estimated by Equation 1.Where y is a dependent variable and according to sesame seed yield, biological yield, leaf area index, leaf dry weight, number of pods per plant, number of seeds per pods, 1000-seeds weight, relative water content and seed oil percentage were calculated separately; Xi is the independent variable, XiXj i the interaction of the independent variable i and the independent variable j, Xi2 the second power of the independent variable i and βi to βii are the coefficients of the equation. After obtaining the simulation results, using calculations and statistical methods, a quadratic polynomial is obtained which expresses the response rate (yield) as a function of input variables. Finally, after optimizing the obtained relationship and eliminating ineffective sentences, using statistical tests and criteria such as F test value, Lack of Fit test, Pvalue and R2 (coefficient of determination), the final relationship for predicting yield and other response variables is calculated for the present study (Equation 2).The obtained relationship is valid only within the limits defined for the input parameters and has no predictive power outside this range. In this equation, y: is a dependent variable already defined for Eq. 4. X1 is the independent variable of irrigation water, X2 is the superabsorbent, and X3 is the independent variable of manure, a1 to a9 are the equation coefficients. Optimal amounts of irrigation water, superabsorbent and manure were determined according to the possibility of maximum seed yield. Finally, the estimated values were compared with the observed data and the validity of the regression models was evaluated using the root mean square error (RMSE).Results and DiscussionIn general, considering the significant effect of linear component of manure from regression model and biological yield as well as linear effect of irrigation and manure on the number of seed per pod, it can be concluded that the use of manure ultimately increased seed yield. Regarding the quality yield of sesame (seed oil percentage), the significance of the linear component of the superabsorbent effect can guarantee the quality yield of sesame. The high significant (p≤0.01) effect of the second order (full quadratic model) component of manure on 1000-seeds weight also indicates the effectiveness of manure on yield components and finally seed yield. Optimization was performed with three scenarios. First, considering all three factors, irrigation, superabsorbent and manure were done. Optimization was done to produce the highest seed yield, in which case the highest seed yield (4541 kg ha-1) was obtained with 3,000 m3 ha-1 of irrigation water, without superabsorbent and with 30 t ha-1 of manure. In the next step, to investigate the role of superabsorbent in water saving, optimization was performed with half irrigation water, 100 kg ha-1 of superabsorbent and with no application of manure, which resulted in the production of 3380 kg ha-1 of seed. In the third case, the results of optimization for irrigation water in the range of 1500 to 2250 m3 ha-1 (limited irrigation), with no use of superabsorbent and the amount of manure from zero to 30 t ha-1, showed that applying 2250 m3 ha-1 of water and 30 t ha-1 of manure could be resulted in a seed yield of 4186 kg ha-1. The highest amount of irrigation water productivity, equal to 2.2 kg seed per m3 of water, was obtained from 2250 m3 of irrigation water. Third scenario compared with the first scenario shows a reduction of 750 m3 in the volume of irrigation water that resulted in only an 8% reduction in seed yield (4186 vs. 4541 kg of seed), therefore, the third scenario potentially could be chosen by the farmers. Depends on the level of availability of water resources, the balance of economic value of water against seed yield, and other environmental and management options, if the application of 30 t ha-1 of manure to achieve the stable seed yield is not economical for the farmer compared with the application of 100 kg ha-1 of superabsorbent, we can recommend the second scenario (1500 m3 of water plus 100 kg of superabsorbent, with no manure) that will result in the seed yield by 3380 kg ha-1. The difference in seed yield in this scenario compared with the third scenario is 806 kg of seed, so the farmer must take into account all economic and managerial conditions to select the appropriate scenario. In general, the results of this study showed that using eco-friendly inputs, it is possible to produce stable sesame in an arid and semi-arid region and achieve a yield beyond of the conventional high-input systems.