بررسی پایداری عملکرد دانه ژنوتیپ‌های گندم نان با استفاده از روش‌های تک متغیره ناپارامتری و مدل gge بای پلات

به منظور تعیین پایداری عملکرد دانه ژنوتیپ‌‌‌های گندم نان، آزمایشی با استفاده از 23 ژنوتیپ گندم نان به همراه دو رقم شاهد با استفاده از طرح بلوک‌‌‌های کامل تصادفی با سه تکرار و در دو محیط آبی (بدون تنش) و دیم (تنش) جمعا (شش محیط) در مزرعه تحقیقاتی و آزمایشگاه‌‌‌های گروه زراعت و اصلاح نباتات پردیس کشاورزی و منابع طبیعی دانشگاه رازی کرمانشاه به مدت سه سال زراعی (95، 96 و 97) اجرا شد. نتایج تجزیه واریانس مرکب نشان داد که بین مکان ها، سال ها، ژنوتیپ ها، اثرات متقابل سال × مکان، سال× ژنوتیپ، مکان × ژنوتیپ، سال × مکان × ژنوتیپ اختلاف معنی داری در سطح احتمال آماری 1% وجود داشت. بنابراین نتایج حاصل بیانگر این مطلب بود که ژنوتیپ های گندم مورد مطالعه در محیط های مورد بررسی واکنش های متفاوتی از خود نشان داده اند. از روش‌‌‌های ناپارامتری برای ژنوتیپ‌‌‌های پایدار استفاده شد. بر اساس معیارهای si(1) و si(2) ژنوتیپ‌‌‌های pishtaaz و wc4592 به عنوان پایدارترین ژنوتیپ ها معرفی شدند. همچنین معیارهای si(3) و si(6) ژنوتیپ‌‌‌های pishgam، wc5001 و pishtaaz را به عنوان پایدارترین ژنوتیپ ها معرفی نمودند. بر اساس معیارهای ناپارامتری npi(1)، npi(2)، npi(3) و npi(4) ژنوتیپ‌‌‌های pishgam و pishtaaz به عنوان پایدارترین ژنوتیپ ها انتخاب شدند. همچنین جهت بررسی اثر متقابل ژنوتیپ×محیط و شناسایی پایدارین ژنوتیپ ها از روش gge biplot استفاده شد. بررسی همزمان پایداری و عملکرد ژنوتیپ ها با استفاده از بای پلات مختصات محیط متوسط (aec) نشان دادکه ژنوتیپ‌‌‌های، wc5001، wc4840 و wc4582 به همراه دو رقم شاهد pishtaaz و pishgam علاوه بر عملکرد بالا دارای پایداری عملکرد بیشتری بودند. هم چنین این ژنوتیپ ها دارای نزدیک ترین فاصله نسبت به ژنوتیپ‌‌‌های ایده آل بودند.

Study of grain yield stability of bread wheat genotypes using non-parametric method and GGE biplot

IntroductionThe necessity of considering wheat production as the staple food of most people in the world reveals the urgent need to produce this strategic product. The most important aspect of producing advanced lines in addition to yield consideration is the stability of the studied traits, especially the stability of grain yield in different environments.Materials and methodsIn this study, 23 bread wheat genotypes with 2 cultivars as control during three cropping years at Razi University of Kermanshah Agricultural Research Field were tested by randomized complete block design with three replications in two irrigated (no stress) and rainfed environments (stress) was implemented. After determining the performance of each genotype, by first performing Bartlett test and proving homogeneity of variance, combined analysis of variance was performed assuming the effect of genotypes and environment (year and location) constant. Nonparametric univariate stability statistics based on Nasser and Huhn’s (1987) and Tennarasu’s (1995) criteria were used for selection of stable wheat genotypes. Next, the genotype effect + genotype×environment (GGE) biplot suggested by Yan et al. (2007) was used. Other analyzes were performed using SPSS 16 and Genstat 12 software.Results and discussionIn this analysis, Ftest was used to investigate the significant effects of variance components of grain yield based on the model (random effect of year and fixed effects of genotype and location). There was a significant difference between places, years, genotypes, interactions of year × place, year × genotype, place × genotype, year × place × genotype at the statistical probability level of 1%. Therefore, the results showed that the studied wheat genotypes showed different reactions in the studied environments. Also, the years and places studied had different effects on the performance of genotypes The Nonparametric statistics studied for selection of stable genotypes from the studied cultivars were evaluated based on the proposed criteria of Nasser and Hoon (Nasser and Huhn, 1987) and Thennarasu (1995). The results indicated that Si(1) usually had higher mathematical expectation and smaller variance than Si(2) in the Nasser and Huhn (1987) method, so the accuracy of Si(1) in selecting genotypes was higher. Stability can be far greater than Si(2) statistics. In this regard, Kaya and Taner (2003) have described the simplicity of calculating the Si(1) statistic as the reason for its preference over the Si(2) statistic. Graphical analysis was used to study the variety of cultivars, environments and the interaction of genotypes and environments. The results of GGE biplot showed that the first and second principal components accounted for 43.1% and 20.9%, respectively, of 64% of the total variation, indicating the relative validity of the biplot in justifying G + GE changes.ConclusionsOverall, a closer examination of the results of nonparametric statistics indicated that genotypes 3 and 8 (Vanguard) were identified as the most stable genotypes by the two statistics Si(1) and Si(2). Whereas, Si(3) and Si(6) statistics identified genotypes 15 (pioneer) and 13 as stable genotypes. According to NPi(1) statistics, genotype 12 was the most stable genotype according to NPi(2), NPi(3) and NPi(4) statistics. This suggests that the use of nonparametric methods by Tennarasu (1995) and Nasser and Huhn (1987) may not lead to the selection of high yielding stable genotypes Soughi et al., (2016). In a study by Abdulahi et al. (2007) on the stability of safflower seed yield, they stated that the statistics of Si(1), Si(2) and Si(3) actually represent a static concept of stability and dependence. They were not significant with mean performance. Therefore, the use of multivariate methods of sustainability decomposition that actually discusses the dynamic concept of sustainability can be important. Overall, the results of multivariate stability analysis showed that GGE Biplot is a suitable method for simultaneous selection of stability and yield of cultivars and lines. In this study, GGE biplot results showed that 20, 17, 15 (pioneer), 9, 6 and 20 genotypes with average yield were among the most stable genotypes in terms of grain yield among studied genotypes., 22 and 24 were identified as the most undesirable genotypes for stability and yield.AcknowledgementsWe would like to thank the colleagues of the Department of Agriculture and Plant Breeding, Razi University of Kermanshah, as well as the technicians involved in this research project, who ultimately contributed to providing the facilities needed to carry out this research.