تأثيرات تنش خشكي بر صفات زراعي گندم (.Triticum aestivum L) با استفاده از فراتحليل

Effects of drought stress on agronomical traits of wheat (Triticum aestivum L.): A meta-analysis

تنش خشكي يكي از مهم‌ترين ريسك‌ها در توليد گندم در بسياري از مناطق جهان بوده و به دليل تأثيرات چندبعدي آن، يك نگراني عمده در تحقق امنيت غذايي به شمار مي‌رود. پژوهش حاضر به‌منظور بررسي تأثير تنش خشكي بر صفات زراعي گندم (.Triticum aestivum L) انجام شده است. بدين منظور با استفاده از روش فراتحليل، نتايج پژوهش‌هاي مختلف مورد تحليل آماري قرار گرفته است تا ديدگاهي كلي و قابل‌تعميم در مورد اثر اين تنش‌ بر صفات زراعي اين گياه در شرايط مختلف ارائه شود. براي ايجاد بانك اطلاعات مورداستفاده در فراتحليل، جستجوي بانك‌هاي اطلاعات علمي-پژوهشي مرتبط با اهداف تحقيق انجام شد و پس از بررسي و ارزيابي 120 پژوهش، 60 مقاله با موضوع بررسي صفات عملكرد دانه، عملكرد بيولوژيك، شاخص برداشت، وزن هزار دانه و تعداد دانه در سنبله در شرايط تنش خشكي انتخاب‌شده و داده‌هاي آن‌ها مورد تحليل قرار گرفت. در اين تحقيق، به‌منظور ارزيابي تفاوت بين تيمارها در مطالعات مختلف از شاخص‌ اندازه اثر كوهن استفاده شده است. نتايج نشان داد تأثير تنش خشكي در مورد تمامي صفات موردبررسي منفي و معني‌دار بوده است. تغييرات اندازه اثر متغير تنش خشكي بر صفات مزبور 1.9248 تا 4.4280 بوده است. نتايج نشان مي‌دهند كه در تمامي شرايط مطالعه شده، تنش خشكي بيشترين تأثير را بر عملكرد دانه داشته و ساير صفات شامل وزن هزار دانه، تعداد دانه در سنبله و شاخص برداشت در رتبه‌هاي بعدي قرار دارند. تحليل نتايج پژوهش‌ها ازنظر سوگيري و همگني نشان مي‌دهد كه پژوهش‌هاي مربوط به تأثير تنش خشكي بر عملكرد دانه، شاخص برداشت و عملكرد بيولوژيك، وزن هزار دانه و تعداد دانه در سنبله همگني پاييني دارند كه حاكي از سوگيري مطالعات است.

Introduction Drought stress is one of the major risks to wheat production in many parts of the world. Due to its multidimensional effects, drought stress become a known concern for food security. The purpose of this study was to evaluate the effect of drought stress on agronomic traits of wheat using meta-analysis. A large number of studies have been conducted worldwide regarding effect of drought stress on crops as a major risk in cereals production. The multi-dimentional effect of drought on agronomic traits of wheat are too complex to compare using traditional methods. Therefore, meta-analysis was used in the present study to identify general trends among numerous independent drought-related studies and combine the results into one measure known as the effect size. Meta-analysis of field trials data can provides an overall view of variables affecting crop traits and precision of experimental data. The purpose of this study was to evaluate the effect of drought stress on agronomic traits and evaluating reliability of studies using meta-analysis. Material and methods Using scientific database of agronomic researches, 120 articles were evaluated to build a database of experimental data. The study synthesized results from 60 papers and published researches on drought effects and the database was built based on keywords such as wheat, drought stress, yield and agronomic traits. All studies that met the following criteria were included in the database: (1) available data of trial and control plots, (2) reported data of drought effect on at least one of five traits(yield, harvest index, biological yield, 1000-grain weight and grain per spike of wheat). The effect size for each datasets was calculated using cohen's index. Cohen's methodology applies differences between trial and control data as a measure of effect size. As a common measure in meta-analysis, Standard Mean Difference(SMD) used to evaluate precision and bias of experimental data of studies. I-squared measure used as estimation of heterogeneity for results of studies. An I-squared of less than 50% is usually viewed as low heterogeneity and over 50% as high heterogeneity. Results and discussion The results of analysis demonstrated that effect of drought stress on all mentioned agronomical traits is significant and negative. Using Cohen's index, the calculated effect size of drought stress on varies from 1.9248 to 4.4280 among traits. Grain yield had highest effect size and harvest index showed the lowest among all traits. Lower effect size in harvest index shows tolerance of the trait and it's relative sustainability during stress period. Analysis of experiments showed a medium rate of heterogeneity in studies on grain yield, harvest index and biological yield and a higher heterogeneity in trials on 1000-grain weight and grain per spike. Conclusion Using meta-analysis in agronomic researches provides a framework to examine reliability and generalizability of experiments and individual studies. Drought stress affects many agronomic traits of wheat in different way in spite of statistical significance, results of many studies are not reliable and evident degrees of biases. Higher heterogeneity of experimental data shows a higher probability of bias and less trustable results in the examined researches and lower reliability of the sources for field application and generalization of the results.