Meta-analysis of the effect of salinity stress on grain yield, yield components and physiological traits in rice (Oryza sativa L.)

Introduction
Rice is often very sensitive to salinity in both seedling and reproductive stages. Excess salt and mineral salts in the soil or irrigation water can cause great damage to the plant, disrupting the natural growth of the plant and limiting the actual performance potential of rice varieties. Many researches have been conducted on the effect of salinity stress on rice. Although each research separately has a special value, the results of these researches are different and it will be difficult to summarize them separately in order to obtaine a general and comprehensive result. In the meta-analysis method, statistical analysis of a large number of researches is performed and by evaluating the overlap of the results of smaller studies, a general summary is obtained from the findings of all studies. Therefore, the results of the meta-analysis method has more validity. The purpose of the current study was to meta-analyze the results of the researches conducted on the effect of salinity stress on rice.
Materials and methods
To obtain the necessary data for the present study, the Persian keywords “rice”, “salinity stress”, “grain yield”, “yield components”, “NaCl”, “photosynthetic pigments”, “sodium content”, “potassium content”, “vegetative growth stage” and “reproductive growth stage” or their English equivalents were searched from the agricultural research information banks, such as Magiran, SID, Elsevier and Google. Salinity stress was also considered at two levels including mild stress (4 dS.m^-1) and severe stress (8 dS.m^-1). Out of the 50 reviewed studies on the effect of salinity stress on growth, yield, yield components, content of elements and photosynthetic characteristics of rice, 38 articles were selected and their data were extracted. Comprehensive meta-analysis (CMA) software was used for statistical analysis of the extracted data and drawing the relevant graphs.
Research findings
The results of the homogeneity test statistics of data extracted from previous studies (Q, I² and Tau²) showed that the heterogeneity between the studies for grain yield and panicle number per plant under severe salinity stress (8 dS.m^-1) and plant height under mild stress (4 dS.m^-1) was low and non-significant, therefore the fixed model was used in the meta-analysis. However, random model was used for other traits in both salinity stress conditions due to the heterogeneity between studies. Overall effect size for grain yield, plant height, panicles per plant, filled grains per panicle and potassium content were on the left side of the zero axis under both severe and mild stress conditions, indicating a significant reduction of these traits under salinity stress. But, the overall effect size for sodium content in both severe and mild salinity stresses were on the right side of the zero axis, which indicated the increasing effect of salinity stress on sodium content. The results of the overall effect size showed a decrease in chlorophylls a and b under both salinity stress conditions and carotenoids in mild stress. The results showed that grain yield with a decrease of 72.8 and 118.9 percent at the stress levels of 4 and 8 dS.m^-1, respectively, was the most sensitive and vulnerable trait to salinity stress. Also, the decrease in the number of panicles per plant was 30.42 and 35.29 percent, and in the number of filled grains per panicle was 25.16 and 57.84 percent at salinity levels of 4 and 8 dS.m^-1, respectively. This finding showed that the decrease in the number of panicles per plant at mild salinity stress was more than the number of filled grains per panicle, while with the increasing of salinity level to 8 dS.m^-1 (sever stress), the number of filled grains per panicle was more affected and showed more sensitivity to high levels of salinity stress compared to the number of panicles per plant.
Conclusion
The results of this study showed that increasing the number of fertile panicles and the number of filled grains per panicle, higher levels of potassium, lower levels of sodium, reducing the ratio of Na⁺/K⁺, maintaining the amount of chlorophyll pigments at the optimal level and preventing their destruction, increasing proline accumulation and activity of catalase and guaiacol peroxidase enzymes, prevention of cell membrane destruction and reduction of ion leakage, were suitable criteria for assessing salinity tolerance and identifying rice tolerant genotypes. The varieties Gharib Siah Reihani, Xudao9, Nagdong, Pokkali, Zayandehroud and Seng-e Tarom as well as a mutant line of Nemat variety were salinity tolerant genetic resources and can be used in breeding programs of salinity tolerance in rice.