Diversity Estimate and Selection Study of Sugarcane Hybrids Suitable for Agro -Climatic Condition of the North-Western Zone of Country

Abstract

The whole work was carried out under three separate investigations in two parts. The nine agronomical characters, viz. germination percentage (Germi. %), number of tillers per clump (NT/C), number of mill able canes per clump (NMC/C), cane stalk height in cm (CSH), cane stalk girth in cm (CSG), leaf length in cm (LL), leaf breadth in cm (LB), field brix percentage (Brix%) and cane yield per clump in kg (CY/C) were used in the analyses for both the parts. The first part genetic study is divided into two portions. They are (A) diversity estimate and (B) inheritance of yield and yield components using North Carolina Design I (NCD-I). In portion (A) diversity estimate was studied in NCD- 1 derived from ten crosses of sugarcane (Saccharum officinarum L.) genotypes following mixed model for components of variation, coefficient of variability, heritability, genetic advance and genetic advance expressed as percentage of mean. Data were collected in two consecutive years namely 2008-2009 and 2009- 2010 under three different locations viz. BSRI -Ishurdi at Pabna, Horian at Rajshahi and RSRS at Thakurgoan districts which are 'in the North-Western zone of Bangladesh. All the characters represented a remarkable range of variation, between them CSH was the highest. The analysis of variance showed that all the items and their interactions were significant when tested against within error. The maximum genotypic (ci G) and phenotypic (cr2 p) variation was observed for CSH and the lowest for CSG. Phenotypic coefficient of variability (PCV) was higher than genotypic and others coefficient of variabilities for all the characters except LL. Where for LL, PCV was lower than YCV and LxYCV. The difference between phenotypic and genotypic variation was greater in magnitude for all the characters indicated the location (environment) had considerable effect on these characters. All the traits exhibited low heritability (h2 b), genetic advance (GA) and genetic advance expressed as percentage of mean (GA%). Among these are characters, Genni. % Exhibited moderately higher h2 b and GA% and, CSH exhibited moderately higher GA. In portion (B) the analysis of variance of NCD-I was done for the determination of additive (DR) and dominance (HR) component, degree of dominance, heritability, genetic advance and genetic advance expressed as percent of mean. The analysis of variance showed that in maximum cases block (B) item was non-significant and crosses (C), males and females in males’ items were significant. Significance of males and females in males in different set indicates substantial contribution of males and females to the variation among NCD-1 crosses. Various interaction items such as blocks x crosses, blocks x males and blocks x females in males were significant for the characters NMC/C, LL and LB and non-significant for rest of the six characters when tested against within crosses. In maximum cases additive (DR) component was negative and lower than dominance (HR) component. The negative estimation of additive component may be due to large sampling error or lack of random mating. The magnitude of additive component (DR) was higher than that of dominance component (HR) only for CY/C indicated that DR was more important than HR for CY/C. Higher magnitude of dominance variance in maximum cases indicated that they were governed by non-additive gene action. The preponderance of non-additive gene action for these characters indicated that improvement of these characters could be possible through heterosis breeding. Most of the characters showed over dominance (ratio > 1) and rest of them showed partial dominance (ratio < 1), which indicated dominance effects towards decreasing parents. Heritability in narrow sense (h\) was higher than broad sense (h2 b) only for LL but in other cases it is lower than broad sense. Heterosis over mid-parents was found to be significant for the character CSG. In case of CY/C the crosses f 2 x m 1, f 3 x m 2, f s x m 3,f 8 x m 4 and f 9 x m s revealed significant mid-parent heterosis. The results show that heterosis may be exploited for most of the characters as they showed preponderance of dominance gene. The second part of this study deals with the genotype x environment interaction. In this experiment the six environments were made with three locations in two consecutive years. The joint regression analysis of variance for nine characters showed that genotypic (G) and environment (E) items were highly significant for all the characters indicating that there were real differences existed between the genotypes and between the effects of different environments on the genotypes. The genotypes significantly interacted with the environment in the changing environment due to highly significant genotypes x environment interactions. In most of the cases both linear and non-linear regression was accounted for the G x E interaction and the heterogeneity of regression showed non-significance for all the characters. The genotype-environment (G x E) interaction was due to the slopes of non-linear relationship. The regression analysis of stability showed that the genotypes G1 and G3 for Germi. %; G2 for LL and Brix %; Gs for NMC/C; G6, G7 and G8 for LB; G9 for CSH and G4 and G9 for CY/C were the stable genotypes in the present materials. These genotypes might be considered as the most stable with the changing environments and could be used for future breeding programme. Regarding non-significant S2di and high values of bi, the genotypes G2 and Gs for NT/C, G2 and G3 for NMC/C, G9 for CSG and G2, G7 and G8 for CY/C were recommended only for favourable environments. The genotypes G4, G6, G9 and G10 for NMC/C, G7, Gs and G1o for CSH; G7 and Gs for CSG and G2 for LB were found poor adaptability to all environments because the regression coefficient is less than 1.0 and the S2di values were non-significant. Other genotypes were unpredictable due to their significant S 2di values for different characters.