IMPROVING THE PRECISION OF SOYBEAN VARIETY TRIALS USING TREND ANALYSIS MODELS
Abstract: Spatial variability is a common problem perhaps faces the agronomists and plant breeders in variety trials especially with large number of genotypes. Although, the blocking system using complete or incomplete block designs accounted for a proportion of this local heterogeneity as inter-block variability, a considerable amount of intra-block variability still remains unaccounted for, however, and this may lead to mask the differences between genotype means. To hold this undesirable part of variability, yield data from 33 genotypes of soybean and three released cultivars (Crawford, Giza 111 and Giza 35) were analyzed in a variety trial using randomized complete block design (RCBD), simple square lattice (6x6) and trend analysis. The field experiments were conducted at Sakha Agricultural Research Station, Kafr El-Sheikh Governorate, during 2009 and 2010 seasons. The coefficient of determination (R2) of the model, coefficient of variation (CV %), Relative efficiency (RE%), Type I error and Type II error were used as statistical criteria to investigate the usefulness of trend analysis over RCBD and lattice design model in accounting for spatial variability. Also, Pearson and Spearman rank correlation coefficients were computed to identify the effect of the used models on the ranks of the genotype means. The obtained estimates showed the superiority of trend analysis over RCBD in both seasons on the basis of lower values for each of CV%, Type I error and Type II error plus high values of R2 % of the model and RE%. Similar performance of the models of lattice design and trend analysis was only obtained in 2009 season while trend analysis model was more efficient in 2010 season. Highly significant correlation coefficients (Pearson and Spearman) were detected indicating considerable degree of similarity between the tested models in adjusting the genotype means for spatial variability. Finally, it could be concluded that trend analysis appeared to be an useful procedure to account for intra-block heterogeneity especially when the pattern of this variation is complex and in curvilinear form.
Publication year 2011
Pages 89-103
Organization Name
Author(s) from ARC
External authors (outside ARC)
    أكرم رشاد مرسى قسم بحوث المحاصيل البقولية - معهد بحوث المحاصيل الحقلية
Publication Type Researsh & Applied Activities
↑