INTRODUCTION Wheat (Triticum aestivum L. 2n = 42) is an important cereal crop and is the staple food for more than 2.5 billion people all over the world (CRP, 2018). Globally it provides 20% of dietary calories and protein, thus contributing considerably to food security (Shiferaw et al., 2013). Worldwide wheat is the most widely cultivated than any other crop (Curtis and Halford, 2014) on an area of 219 million hectares with an annual production of 760.92 million metric tons (FAO, 2020). However, the key challenges to wheat production which leads to a decreased yield are the low annual rate (0.9%) of yield increase (Ray et al., 2013), stagnating yields (Ray et al., 2012) and the impacts of biotic stresses, abiotic stresses and climate change (Wheeler and von Braun 2013). Hence, to accelerate wheat breeding for higher yield potential, lesser genetic vulnerable, stress resistance and climate resilience, it is important to diversify the wheat germplasm resources. Determination of genetic variability is useful for crop improvement to boost productivity in more efficient way along with enhanced resistance under different stress conditions (Khodadadi et al., 2011). Thus, genetic improvement of crops including wheat relies on the availability of diverse and superior genotypes. It has, however, been frequently asserted that modern and intensive monoculture of prominent cultivars led to narrowing of genetic variability in crops. In recent years, this possibly led to a genetic vulnerability of crops to biotic and abiotic stresses (Vellve 1993; Russell et al., 2000; Fu et al., 2005). Thus, it is vital in plant breeding programs to broaden its genetic background of crops to maintain sufficient genetic diversity to allow production of new and diverse varieties suitable for cultivation under a variety of biotic and abiotic stress conditions (Singh et al., 2016; Zampieri et al., 2017). Grain yield is a complex trait, highly influenced by many environmental factors and genotype and environment interactions. In plant breeding programme, direct selection for yield could be misleading. A successful selection depends upon the knowledge on the genetic variability and association of agro-morphological traits with grain yield. So there is an acute need to assess the variability amongst various germplasm resources in different environments. Adaptation of variable wheat genotypes at diverse agro-ecological conditions of the northern hills zone under Multi-environment Trials (METs) is useful, as performance of genotypes over range of environments are considered to measure the criteria for genetic stability. The main objective of breeders is to develop varieties that give stable productivity under variable environmental conditions. The alteration in yield is the result of genotype × environment interaction. Thus, testing is done by assessing the variable wheat genotypes under different environmental conditions, to analyse the stability of a genotype, which gives the clear cut picture of the performance by variable genotypes at different locations. MATERIAL AND METHODS The experimental materials comprised of 35 diverse wheat genotypes including elite cultivars, advance lines, released varieties and three checks HS 542, HS 490 and HS 562. The experiments were carried out in randomized complete block design (RCBD) having three replications, at seven environments in five locations across different agroclimatic conditions of Northern Hills Zone viz., Research Sub Station Akrot (Una); Krishi Vigyan Kendra (KVK), Sundernagar; Hill Agricultural Research and Extension Centre (HAREC), Bajaura; Department of Organic Agriculture, College of Agriculture, CSKHPKV Palampur and Rice and Wheat Research Centre, Malan (RWRC Malan) during Rabi 2019-20. Each genotype was sown in three rows with plot size of 2.0 × 0.6 m2 with row to row spacing of 20 cm. The observations were recorded on five randomly selected competitive plants in each replication on various phonological and morpho-physiological traits namely: days to 50% flowering (DTF), days to 75% maturity (DTM), plant height (cm) (PH), tillers per plant (No.) (TPP), peduncle length (cm) (PL), grain yield per plant (g) (GY), biological yield per plant (g) (BY) and thousand kernel weight (g) (TKW). Statistical analysis. The Statistical analysis of the data for variability over the seven locations (environments) was done with the help of software WINDOSTAT Version 9.2 from Indostat Services, Hyderabad. Scott-Knott clustering algorithm was used to make comparison between treatment means (Scott and Knott, 1974). Analysis of variance was done as per Panse and Sukhatme (1984). The analysis of variance (ANOVA) was constructed based on the linear model given by Fisher (1954): Yij = m + gi + rj + eij Where, Yij is phenotypic observation of ith genotype grown in the jth replication; m is General population mean; gi is Effect of ith genotype; rj is the effect of jth replication and eij is error associated with ith genotype in jth replication. The phenotypic & genotypic components of variance (PCV & GCV), heritability in broad sense (h²bs) and Genetic advance as (%) of mean were computed as suggested by Burton and De Vane (1953); Johnson et al. (1955). Phenotypic coefficient of variation (PCV %) = (σp/ x) × 100 Genotypic coefficient of variation (GCV %) = (σg/ x) × 100 Where, σp is the phenotypic standard deviation; σg is the genotypic standard deviation; σe is environmental standard deviation and x is the population mean. Heritability [h²bs%] = σ²g / (σ²g + σ²e) × 100 Where, σ²g is genotypic variance and σ²e is phenotypic variance Genetic advance (GA) = K × σp × h² (bs) GA% of mean = (GA/ x) × 100 Where, K is the selection differential at 5% selection intensity i.e. 2.06. RESULTS AND DISCUSSIONS Analysis of variance and mean comparison. The analysis of variance revealed (Table 1) significant difference in all the investigated traits in all environments. The seed yield per plant in pooled environments ranged from 4.59-5.84 with a mean value of 5.27 g and it is minimum in E4 (2.40g) and maximum in E6 (7.80g). In E2 seven genotypes viz., TAW 164, HPW 376, HPW 439, HPW 373, HPW 314 and BW 272 were significantly superior to the best check i.e., HS 542; in all other environments none of the genotype was significantly superior to the best check. Parameters of variability. The various parameters of variability viz., mean, range, phenotypic coefficient of variation (PCV) and genotypic coefficient of variation (GCV) along with heritability in broad sense (h2bs) and genetic advance (GA) expressed as percentage of mean for different traits are presented in Table 2. High PCV (>25%) was observed for peduncle length; moderate (15-25%) for tillers per plant, Thousand kernel weight whereas, it was low (<15%) for days to 50% flowering, days to 75% maturity, plant height, biological yield per plant and grain yield per plant. High GCV (>25%) was observed for peduncle length; whereas, it was low (<15%) for days to 50% flowering, days to 75% maturity, plant height, tillers per plant, thousand kernel weight, biological yield per plant and grain yield per plant. The lower value of GCV as compared to PCV for all the traits indicated that the effect of environment was negligible on the development of plants. The experiments revealed that broad sense heritability (h2bs) was high (>70%) for days to 50% flowering, days to 75% maturity and peduncle length; moderate (50-70%) for plant height whereas, it was low (<50%) for number of tillers, thousand kernel weight, biological yield per plant and grain yield per plant. Genetic advance expressed as percentage of mean was not observed to be high (>50%) for none of the trait; moderate (25-50%) for peduncle length and thousand kernel weight and low (>25%) for days to 50% flowering, days to 70% maturity, plant height, tillers per plant, biological yield per plant and grain yield per plant. The traits viz. peduncle length and thousand kernel weight showed moderate to high GCV, PCV, heritability and genetic advance as % of mean, showing crucial involvement of additive gene action in the expression of these traits. Scott-Knot Test Grouping have revealed that the eight genotypes 25, 31, 7, 16, 18, 19, 9 and 29 (i.e., HPW 314, HPW349, TAW 164, HPW 442, PW 1076, MT 19-18, TAW 172 and MLW 1449) are at par to each other and to the check varieties HS 542 HS 490 HS 562 and were significantly superior in pooled grain yield per plant over seven environments to the rest of genotypes. The analysis of variance showed the existence of adequate genetic variation at various agro-ecological zones of Himachal Pradesh, among the various wheat genotypes for all the traits studied. Jee et al. (2019) also found that the analysis of variance was highly significant between the genotypes under their study. It shows that the germplasm selected for the study have significant variability which is required by any breeding program. It was observed that value of PCV was slightly higher than the value of GCV. Similar results were observed by Singh et al. (2018) in wheat, Gautam et al. (2021) in chickpea and Shiva Kumar et al. (2021) in linseed indicating less effect of environment on the expression of traits studied. Moderate to high PCV and GCV was observed for peduncle length, tillers per plant, Thousand kernel weight. Singh et al. (2019); Morteza et al. (2018) recorded high GCV and PCV for biological yield per plot and harvest index. Heritability and genetic advance are two important selection parameters, these are useful in selection for various traits for genetic improvement of grain yield in wheat. Heritability broad sense (h2bs) and GA (%) of mean was observed moderate to high for days to 50% flowering, days to 75% maturity, plant height, peduncle length and thousand kernel weight. Singh et al. (2019) observed high heritability and genetic advance as 5% mean for traits like harvest index, number of grains per spike and biological yield per plot. Singh et al. (2018) also found high GCV and PCV along with high heritability and high genetic advance which indicated considerable inclusion of additive gene action and selection would be effective for these traits.