INTRODUCTION Maize is the primary staple food in many developing countries in the world. It is a versatile crop with wider genetic variability and able to grow successfully throughout the world covering tropical, subtropical and temperate agro-climatic conditions (Amiruzzaman et al., 2010). In North western Himalayan region, it is the most important crop of Kharif season. It occupies 290.0 thousands hectare with a production of 760.0 metric tonnes and productivity is high (25.70 q/ha) as compared to the national average of 22.46 q/ha in kharif season (Anonymous, 2019), yet there is considerable scope for increasing the productivity further with the use of quality seeds of recommended varieties/hybrids. The main target of maize breeding programme is to increase the yield using commercial exploitation of high yielding maize hybrids. The selection of parents and breeding strategies for the successful maize hybrid production will be facilitated by heterotic groupings of parental lines. Therefore, information on heterotic groupings of maize germplasm is essential for hybrid breeding program (Kumar et al., 2019, Chandel et al., 2019, Eisele et al., 2021). A Set of lines deriving from a common origin and displaying similar combining ability when crossed with lines from different origins is defined as a heterotic group. After development of inbred lines from known or unknown sources, breeders need to make thousands of crosses and evaluate grain yield in resulting F1 plants in replicated field experiments. Assigning lines to heterotic groups would avoid the development and evaluation of crosses that should be discarded, allowing maximum heterosis to be exploited by crossing inbred lines belonging to different heterotic groups (Mousa et al., 2021). Heterotic effects of the maize lines and their allocation into well-known heterotic groups is the secret for the success of a maize breeding programme, which would give utmost exploitation of heterosis. The classification of inbred lines into heterotic groups is therefore of very high importance in hybrid maize breeding. Melchinger and Gumber (1998) described a heterotic group “as a group of related or unrelated genotypes from the same or different populations, which display similar combining ability and heterotic response when crossed with genotypes from other genetically distinct germplasm groups”. Two major methods of heterotic group classification are mainly used in breeding programme: In the traditional method, breeders assign the germplasm into the different heterotic groups based on the estimates of the combining ability patterns obtained using information from testcross trials (Fan et al., 2001; 2004). The second method utilizes molecular markers to compute genetic similarity or genetic distance to assign maize inbred lines to different heterotic groups (Barata and Carena 2006). The present investigation was aimed to characterize the maize germplasm into different heterotic groups and evaluation of single cross hybrids in the sub-temperate region of north western Himalayas. MATERIAL AND METHODS A. Experimental material The experimental material comprised of 30 maize inbred lines crossed with two diverse male parents viz., BAJIM-08-26 (T1) and BAJIM-08-27 (T2) from two different pools during Kharif season in experimental field at CSK Himachal Pradesh Krishi Vishva Vidyalaya, Hill Agricultural Research & Extension Center, Bajaura, Kullu, (H.P.) India (Table 1). Sixty crosses along with parents and two standard checks viz., Bio-9544 and Palam Sankar Makka-2 were evaluated in randomized block design (RBD) with two replications during Kharif, 2016. Observations were recorded on ten randomly selected plants per treatment per replication for the traits viz., plant height (cm), cob placement (cm), cob length (cm), cob girth (cm), kernel rows per cob and kernels per row and were used for statistical analysis. However, days to 50 per cent tasseling, days to 50 per cent silking, days to 75 per cent brown husk, grain yield (q/ha), 1000 grain weight (g) and biological yield (q/ha) were recorded on plot basis. B. Statistical Analysis Analysis of variance for mean data recorded was carried out as per suggested by Panse and Sukhatme (1985) to determine significant differences among genotypes. Combining ability analysis for grain yield was done according to Kempthorne (1957) and SAS statistical software was used for heterotic grouping of germplasm under study. RESULTS AND DISCUSSION A. Analysis of variance Analysis of variance for yield and yield contributing traits has been presented in Table 2. Significant differences among the genotypes were observed for all the characters except days to 50% pollenshed. The results of ANOVA revealed that there was significant variability among the genotypes under study. B. Mean performances of parents and crosses Seven crosses were found to have a mean yield significantly higher than the best check Bio-9544 for grain yield (Table 3). The top yielder crosses were L28 × T2, L15 × T2, L23 × T1 and L10 × T1. However, none of the parents exhibited the higher mean grain yield than the best check (Table 4). The range of mean values in parents for grain yield varied from 49.09 q/ha (L15) to 20.81 q/ha (L8). For 1000 grain weight, none of the parents showed a mean value for 1000 grain weight greater than the best check Palam Sankar Makka-2 (340 g). Only one cross combination L14 × T2 recorded higher 1000 grain weight than the Palam Sankar Makka-2 and the lowest was recorded for L2 × T2. Among the parents four lines viz., L3, L13, L17 and L28 recorded a mean value higher than the best check Bio-9544 for shelling percentage. Twenty four crosses exhibited a mean value for shelling percentage greater than Bio-9544. The highest shelling percentage of 87.68 per cent was recorded for L15 × T2 and the lowest of 78.60 per cent for L24 × T1. The mean value for rows per cob among the parents ranged from 9.8 (L25) to 16.6 (L22) and two lines L19, L22 exhibited a greater mean value than best check Bio-9544 (15.5). Thirteen crosses recorded a higher mean value for rows per cob than the Bio-9544. The highest value of 16.67 was observed for L4 × T2 and the lowest of 12.8 for L15 × T2 cross combination. For kernels per row, none of the parents exhibited the mean value for kernels per row higher than the best check Bio-9544 (40.50). Three crosses showed the value for kernels per row to be higher than Bio-9544. The maximum number of kernels per row 43.17 was recorded for L16 × T1 and the lowest of 28.34 for L23 × T2. The crosses that showed high number kernels per row were L16 × T1, L3 × T2 and L18 × T1. A range of 15.5 cm (L18) to 9.4 cm (L8) for cob length was observed among the parents. No parent had the average cob length more than the best check Palam Sankar Makka-2 (21.25 cm). Two crosses had a mean value for cob length higher than Palam Sankar Makka-2. The maximum cob length of 22.17 cm for the cross L25 × T1 and the lowest of 14.25 cm for cross L2 × T1 was observed. Some other crosses with high mean value for this character were L18 × T1 (21.34 cm) and L3 × T2 (21.26 cm). Six crosses showed the mean cob girth to be higher than the Palam Sankar Makka-2 (16.92 cm) some of these were L22 × T2 (17.92 cm) L19 × T2 (17.40 cm), L6 × T2 (17.25 cm) and L13 × T2 (17.25 cm). None of the parents showed higher value than best check Palam Sankar Makk-2. Out of sixty crosses only single cross L28 × T2 (329.78 q/ha) was observed to have a higher mean value for biological yield than the best check Bio-9544. The parents showed a range of 179.17 q/ha (L9) to 69.74 q/ha (L24) for this trait. Seventeen crosses recorded a higher mean value than Bio-9544 for this character. Among the crosses, L19 × T2 recorded the highest mean value of 0.62 whereas the lowest value of 0.30 was observed for the cross L27 × T2 for harvest index. No single parent had mean value of harvest index higher than that of the best check Bio-9544 (0.45). For days to 50 per cent pollen shed, five lines viz., L17, L26, L28, L5, L1 recorded mean value to be less than the best check Palam Sankar Makka-2 (61 days). Twenty six crosses had mean days for 50 per cent pollen shed less than the best check; some of these crosses were L26 × T2 (56.5 days), L4 × T1 (57 days), L4 × T2 (57 days), L17 × T1 (57 days) and L27 × T1 (57.5 days). Five lines viz., L17, L26, L5, L28 and L1 showed mean days to 50 per cent silking less than the best check Palam Sankar Makka-2 (63 days). Twenty seven crosses had mean days to 50 per cent silking less than the best check Palam Sankar Makka-2. The mean days for this character ranged from 58.5 days (L17 × T1, L26 × T2) to 68.5 days (L2 × T2, L28 × T1). Among parents fifteen lines exhibited mean days to 75 per cent brown husk to be less than the best check Palam Sankar Makka-2 (98.5 days). The means days of parents for this trait ranged from 88.5 days (L17) to 96.5 days (L4). Thirty crosses exhibited the less number of mean days as that of the best check Palam Sankar Makka-2. The highest number of mean days of 106 days was observed in the cross L2 × T2 and lowest days of 92.5 days was for cross L17 × T1. The results obtained are in conformity with the findings of Suthamathi and Nallathambi (2015) for early maturity. All lines except L21 and both the testers T1 and T2 showed a mean plant height less than the best check Palam Sankar Makka-2 (169.17 cm). The mean plant height for parents ranged from 101.67 cm (L28) to 173.67 cm (L21). Twelve crosses exhibited mean plant height less than that of the best check Palam Sankar Makka-2. The least mean height among all the crosses was exhibited by the cross L7 × T2 (132.33 cm) and the highest mean height of 217.5 cm by cross L12 × T1. Twenty nine lines and both the testers T1 and T2 showed the mean height of cob placement to be less than the best check Palam Sankar Makka-2 (94.67 cm). Twenty nine crosses exhibited the mean height for cob placement less than that of the best check Palam Sankar Makka-2. It ranged from 55.67 cm (L7 × T2) to 125.69 cm (L3 × T1). Similar results were reported by Aminu et al. (2014); Talukder et al. (2016) with high negative GCA as desirable for plant height and cob placement. C. Heterotic grouping of germplasm Among thirty inbred lines, eleven lines exhibited positive and significant GCA effects and twelve lines exhibited negative and significant GCA effects for grain yield, out of these L28 (25.14) had the highest GCA effect followed by L12 (21.46), L10 (14.10), L16 (14.14) and L23 (10.92). Line L28 was the best general combiner and L25 the poorest general combiner (Table 5). Thirteen crosses out of total of sixty crosses recorded significantly positive SCA effects. These were L15 × T2, L7 × T1, L2 × T1, L28 × T2, L19 × T2, L3 × T1, L5 × T2, L23 × T1, L25 × T1, L14 × T1, L11 × T2, L10 × T1 and L18 × T1. The SCA effects ranged from 31.11 (L15 × T2) to -31.11 (L15 × T1) (Table 6). The maize inbred lines were grouped into different heterotic groups on the basis of SCA effects were analyzed for yield traits (Table 7). Germplasm lines showing positive SCA effects with T2 and negative with T1 were grouped into heterotic group A whereas the germplasm lines showing positive SCA effects with T1 and negative with T2 were assigned into group B. Fifteen lines viz., L1, L4, L5, L6, L9, L11, L13, L15, L17, L19, L21, L27, L28, L29 and L30 assigned to Group A whereas remaining fifteen lines namely; L2, L3, L7, L8, L10, L12, L14, L16, L18, L20, L22, L23, L24, L25 and L26 were assigned to Group B. However, the lines with positive GCA effects for yield are of practical importance to a breeder for developing high yielding hybrids. Keeping this aspect in view, the eight lines viz., L5, L6, L9, L11, L13, L15, L21 and L28 which showed a positive GCA effect and positive SCA effects with T2 and negative SCA effects with T1 are considered more productive in the heterotic group A whereas the eight lines viz., L3, L10, L12, L14, L16, L23, L24 and L26 which showed a positive GCA effect and positive SCA effects with T1 and negative SCA effects with T2 are more productive in heterotic group B (Table 7). Similar results with respect to the heterotic grouping of maize germplasm have been reported by several workers. Ejigu et al. (2017) assigned 16 lines and two testers in two heterotic groups. Elmyhun et al. (2020) also grouped the maize germplasm in heterotic groups on the basis of combining ability.