Genotype – Environment Interaction of Segregating and Non Segregating Generation of Four Tomato Crosses for Yield and Yield Attributing Characters Reecha T Das 1 and Pranab Talukdar 2 Corresponding authors- e-mail : talukdar_pranab@yahoo.co.in Published by the Indian Society of Genetics, Biotechnology Research and Development, Biotech Bhawan 5 E Nikhil Estate, DPS Road, Shastripuram, Agra 282007 Online management by www.isgbrd.co.in Indian Res. J. Genet. & Biotech 8(1) : 18 – 22 (2016) Abstract A study was conducted during 2013-2014 to assess the stability of genotypes for yield and yield attributing characters using P 1 , P 2 , F 1 , F 2 , B 1 and B 2 generations obtained each of the four tomato crosses viz. Cross-I (H7997 x CLN 1621 E), Cross- II (H7997 x BL 337), Cross - III (H7997 x Nagcarlan) and Cross- IV (H7997 x CLN 2366A) in four test environments. The highly significant environmental variance for almost all the characters suggested considerable difference among environments and their predominant effect on characters. The variance due to genotype was significant for most of the traits under study. Further, the hybrids H7997 x CLN 1621E, H7997 x Nagcarlan and H7997 x CLN2366A with high mean and non significant regression coefficient and deviations mean squares exhibited average stability for fruit yield per plant. Key words : tomato, yield, yield attributing characters, stability analysis. Introduction Gene expression is subject to modification by the environment; therefore, genotypic expression of the phenotype is environmentally dependent (Kang, 7). The development of new cultivars involves breeding of cultivars with desired characteristics and the stability of these traits in target environments. Inconsistent genotypic responses to environmental factors such as temperature, soil moisture, soil type, or fertility level from location to location and year to year are a function of genotype - environment (GE) interactions. The present investigation with six generations obtained from four crosses was undertaken to study the G- E interaction with the objective of obtaining suitable varieties which could perform well over a spectrum of environment and also to identify suitable types suited to particular environment. Linear regression slope as a measure of stability was considered by Finlay and Wilkinson (3). They suggested that a genotype was maximum stable, when its mean performance was high and regression of its performance over the environmental mean approached zero. Eberhart and Russell (4) observed that the deviation from regression, which is the non linear parameter, should also be taken into consideration along with the linear parameter, i.e. the regression coefficient while examining a variety for phenotypic stability. They observed that an ideal variety should possess regression coefficient equal to unity (b i =1). This variety would have average response to the changes in environments. Regression value larger than unity indicates the sensitivity of the variety to the changes in environmental condition. Such a variety is termed as ‘below average stable’ and performs much better than its inherent potentially in high yielding environmental conditions, but the performance is poor in stress condition. Regression values less than unity signifies the insensitivity of the variety to changes in the environment and such an ‘above average stable’ variety is suitable specifically for stress environments. They further suggested that, a genotype should exhibit the least deviation from regression (S 2 di ), to be stable one. The variance due to deviation from regression coefficient is primarily due to the uncontrollable causes and depends on the environment (Bains and Gupta, 1). In most of the studies on regression analysis of genotype x environment interaction, a linear relationship between genotype-environment interaction and environmental index has been reported (Freeman, 3). 1,2 Department of Plant Breeding and Genetics, Assam Agricultural University, Jorhat 785013, India