Determination of optimum concentration of three antioxidant enzymes for increased drought tolerance in Mung Bean (Vigna radiata L.) using genetic algorithm

The free radicals amounts are increased due to drought stress and subsequently they harm plant. The drought stress-resulted free radicals are scavenged by changing in activity of antioxidants like catalase, superoxide distmutase (SOD) and guaiacol peroxidase (GP) (regressors or independent variables; Xs). This reaction is known as resistance to (protection against) oxidative stress. This protection is reflected in traits like growth and grain production (biomass and harvest index, respectively; dependent or response variables; Ys). By maximizing Ys in relation to Xs, the combination of Xs with possible highest Ys can be obtained. This optimization comes from the relations of Ys with Xs under drought stress conditions. Due to having more than one Ys here, the relation tends to be multivariate and highly complex, especially when there is no strong positive relation (correlation greater than .95) between Ys variables. In such situations, the genetic algorithm can overcome the complexity. The output of mentioned algorithm can be used by plant breeders. In another words, breeders can increase drought tolerance by genetic manipulation of plant for optimum activity of Xs. This experiment was aimed to do the mentioned optimization for Mung Bean.
The experiment was carried out in pots located in an open filed to increase the accuracy and possibility of generalizing the results to field results. Pots had 5 kg capacity in which 5 seeds of line VC1973a were planted. For thinning, 3 seedlings were removed, and left 2 ones. The filed capacity was determined using weight method. Treatment levels were 4 levels of low irrigation including irrigation at 80% (control), 65% 50%, and 35% of field capacity. At maturity stage, the harvest index and biomass (biological yield) (Ys) were measured. The concentration of antioxidant enzymes catalase, SOD, and GP (Xs) were determined at flowering stage. For maximizing the function, first partial desirability function was determined. Then general desirability function was calculated. The value of Xs for which the highest amount of Ys is attainable was obtained on the basis of genetic algorithm and using MATLAB software.
The results indicated that a function with 7 components including the main and interactive effects of Xs could predict the relation of Ys with Xs well (Adjusted R2>0.97). The standardized regression coefficient was positive for catatlase which reveals that the drought resistance (harvest index and biomass) enhances with increasing the activity of this enzyme. Like catalase, the effect of GP was additive on drought tolerance, but considering its standardized regression coefficient, this enzyme had a lower effect than catalase. Due to negative standardized regression coefficient for SOD, it could be concluded that the drought tolerance doesn’t enhance with increasing activity of SOD.
The optimized concentration of catalase, SOD, and GP was 0.956 μmol H2O2 g−1 FW, 24.23 AU g−1 FW, and 21.23 AU g−1 FW, respectively for possible maximum drought tolerance. It should be mentioned that these optimized concentrations were all among observed concentrations. Moreover, reports indicate that there is genetic diversity in activity of antioxidant enzymes (prerequisite to carry out breeding programs for attaining the optimized activity) for Mung Bean.