MSc

ABSTRACT
The research work entitled ‘Comparative analysis of different stability models on superior cultures of paddy (Oryza sativa)’ was carried out at College of Agriculture, Vellayani during 2018-2020. The objective was to compare various linear and nonlinear stability models to identify stable superior cultures of paddy and to study the clustering pattern of cultures over the years. Secondary data from various performance evaluation trials conducted on superior cultures of paddy (13 cultures) over the years (2015-2018) at RARS, Pattambi was used for the analysis. The five characters included in the study were plant height (cm), number of panicles per plant, straw yield (t ha-1), grain yield (t ha-1) and 100 grain weight (g).
Bartlett’s Chi square test was used for testing the homogeneity of the error variance over the years for all the characters as an initial step. Among the five characters only straw yield has shown heterogeneity in error variance. Further, pooled analysis was done to test G X E interactions based on all the characters. The results of pooled analysis confirmed a significant G X E interaction for grain yield, straw yield and plant height revealing that genotypes responded differentially to environment.
The stability of 13 paddy cultures over the years was done using Eberhart and Russell’s model, Perkins and Jinks model, Freeman and Perkins model and AMMI model. ANOVA and stability parameters viz regression coefficient and deviation from regression for each model was estimated. The stable genotypes identified in Eberhart and Russell’s model based on regression coefficient (b^Ei) and deviation from regression (S_di^2 (E)) were Cul2 followed by Cul5 and Cul15. In case of Perkins and Jinks model the stable genotypes determined on the basis of the regression coefficient (B_i) were Cul2 and Cul5. However, Cul10 followed by Cul2 and Cul5 were found to be the stable genotype under Freeman and Perkins model with respect to the stability parameter- regression coefficient (b_Fi). AMMI model incorporates Principal Components Analysis (PCA) for GEI and based on the results of this model cultures Cul2, Cul5, Cul15 and Cul9 were identified as the stable genotypes across different environments particularly in first and third environments. Env 1 and Env 2 showed opposite characteristics while Env 2 was comparatively more stable than other two but was less yielding. One among the AMMI Model selection indices named as AMMI based Selection Index (ASTABi) was used to rank the genotypes to obtain the stable genotypes. This also resulted in obtaining the most stable genotypes as Cul2.
Comparison of the four stability models was carried out using Kendall’s coefficient of concordance revealed no similarity among the ranking of parameters of these four models. Spearman’s rank correlation coefficient was determined for the pair wise comparison of fur models. The correlation matrix and correlogram was also obtained. A perfect positive rank correlation was observed between the parameters of Eberhart and Russell model and Perkins and Jinks model suggesting the similarity of the parameters under these two models. Whereas the rank correlation between Eberhart and Russell and Freeman and Perkins with AMMI model were non-significant indicated the deviation of the results in AMMI model from other two models. However, AMMI model was found to be the best since complete enumeration, summarization, and pattern of GEI interaction was made possible only in this model. The most stable genotype based on the entire four models was Cul2 followed by Cul5, Cul9 and Cul15.
The hierarchical cluster analysis using euclidean distance as similarity measure and average linkage as clustering method was performed using the grain yield over the three years (2015-2018). This result of the study also emphasised that the highly stable genotypes identified under different stability models were clustered together in a single cluster (Cluster II). The intra cluster and inter cluster distance measure revealed that there was high genetic divergence between the clusters in which the stable genotypes are included. Cluster analysis was also performed for different years based on all the characters under study. The clustering pattern in the year 2015-16 and 2017-18 was found to be almost similar in nature since the genotypes enclosed in the clusters were nearly same. In the year 2016-17 the clustering pattern was found to be different from other two years. This shows the influence of environment in the performance of the genotype. The clusters in which high genetic divergence was found in the year 2015-16 which was similar to that of the clusters in 2017-18. The cluster mean for the five characters under study showed extensive difference. Cluster II had recorded highest mean for most of the character (straw yield (t ha-1), grain yield (t ha-1) and 100 grain wt (g)). Therefore, hybridization between the selected genotypes from the divergent clusters is essential to judicously combine all the targeted traits.
Among the different stability models studied Eberhart and Russells and Perkins and Jinks models provided almost similar stable cultures which was highly related to the cultures selected on the basis of AMMI model. Moreover, the different stability cultures identified were put together in one cluster in cluster analysis further confirmed the superiority of the stable genotypes over the others.