پژوهشنامه اصلاح گیاهان زراعی (Jun 2024)
Evaluation of Seed Yield Stability of Lentil Genotypes Based on REML/BLUP and Multi-Trait Stability Index (MTSI)
Abstract
Extended Abstract Background: Lentil is a popular legume crop in the Mediterranean region, widely grown for its nutritious seeds and improving soil fertility. Interest in legumes is increasing as a protein source to replace meat in the future. Identification of high-yield genotypes with adaptation to a wide range of environments is one of the major goals in crop and lentil breeding programs. Combining the best linear unbiased predictions (BLUP), additive main effects, and multiplicative interaction (AMMI) methods in multi-environment experiments and multi-trait stability selection (MTSI) helps to better evaluate plant genotypes and achieve more accurate results. Additive main effect and multiplicative interaction (AMMI) and BLUP are two methods for analyzing multi-environment trials. The linear mixed effects model (LMM) and the restricted maximum likelihood (REML) estimator methods are among the important methods that have been proposed to analyze the data of multi-environmental experiments. In this regard, the BLUPs obtained from the interaction of genotype and environment are performed with principal component analysis or single value analysis on the matrix. This method uses the stability index of the weighted average of absolute scores of the best unbiased linear forecasts (WAASB), the weighted average of the stability index of WAASB, and the dependent variable (WAASBY). Researchers have also proposed an MTSI based on factor analysis, in which grain yield, other traits, and the stability of each are simultaneously used to identify stable genotypes. This research aimed to identify stable and high-yielding lentil genotypes in autumn cultivation. Methods: To evaluate the seed yield stability of 12 lentil genotypes along with three check genotypes, including Kimia, Bileh Sawar, and local landrace, an experiment was conducted as a randomized complete block design with three replications at Agricultural Research Stations of Khorramabad (Lorestan), Zanjireh (Ilam), and Sararoud (Kermanshah) in three cropping years (2019-2022). Each plot consisted of four lines with a length of 4 m and a distance of 25 cm from each other. iIn addition to the usual crop care such as weeding and pest control, the desired traits and characteristics, such as the number of days to 50% flowering, plant height, and number of days to maturity, were measured during the growing season. Hundred-seed weights and the yield of each plot were measured after the maturity and harvesting of experiments. Combined analysis of variance (ANOVA) was performed using SAS software, and the average traits of the treatments were compared using the LSD test. For statistical analyses, the Metan Ver.1.9.0 (multi-environment trial analysis) package was used in the R software environment. To estimate stability quantities, singular value decomposition (SVD) was applied to the matrix of BLUPs obtained from genotype-by-environment interactions with an LMM. Variance components were estimated by the REML method. After analyzing the variance of the data, the stability parameters of WAASB and WAASBY (for simultaneous selection based on average performance and stability) were estimated using the eigenvalues obtained from the AMMI analysis on BLUP, and the best genotypes were selected with these two indicators. Genotypic stability values were obtained from the Harmonic Average of the Genotypic Values (HMGV) index. The compatibility of genotypes was evaluated based on the relative performance index of genotypic values (RPGV). The harmonic mean index and relative performance of genotypic value (HMRPGV) were used to simultaneously evaluate stability, compatibility, and seed yield. Results: The effect of environment, genotype, and genotype × environment interaction were significant on seed yield, plant height, days to flowering, days to maturity, seed filling period, seed filling ratio, seed yield formation rate, rainfall efficiency, and single seed weight. The genotype effect was significant on all traits, except for the seed-filling period. Based on the biplot analysis, genotypes 4, 6, 7, 9, and 10 had higher yield stability in addition to the highest seed yield. The Scree test showed that the first three principal components explained 45.41, 19.13, and 14.34% of the genotype × environment interaction variation obtained from BLUP for grain yield, respectively; in total, they justified 78.87% of the variation. Based on a weighted average of absolute scores of WAASB, genotypes 6, 10, and 12 were high-yielding and stable. Genotypes 1 and 10 were superior based on the (MTSI). The harmonic mean and HMRPGV introduced genotypes 10, 9, 4, and 12 as the genotypes that had high stability and compatibility in addition to high seed yields. Conclusion: Based on all the analyses, genotype 10 was the most stable genotype, which, in addition to seed yield, was superior to other genotypes in terms of the other measured traits and can be a candidate for introduction as a new cultivar.
