| A diallel cross analysis of grain number and grain
weight under different environments in wheat P. N. NARULA I. A. R. I., Regional Station, Pusa, Bihar, India Plant breeders are unanimous over the fact that yield is a very complex character and gains by selection on the basis of yield alone are difficult to achieve. This point has been further emphasized by the classical work of Grafius (1956) who put forward the geometrical logic for working on the basis of yield components. Extensive studies carried out by Mc Neal (1960), Kronstad & Foote (1964), Sharma & Knott (1964), Paroda and Joshi (1970) and Randhawa and Gill (1978) have shown that grain weight and grain number/ear can effectively be used for component breeding. The precise knowledge of the genetic parameters governing various component characters is useful in making decisions with regards to appropriate breeding system to be adopted to achieve the objectives. The present paper embodies results of genetical studies on grain number and grain weight under three experimental conditions under which wheat is commonly grown in India. Materials and Methods Nine wheat variaties of Indian and Mexican origin, namely NP 4, NP 852, NP 884, E 4845, E 5477, Sonora 64, Kalyansona, Safed Lerma and Sonalika having many contrasting characters, were crossed in all possible combinations excluding reciprocals. Both parents and their 36 F1S were grown under timely sown high fertility (E1), timely sown low fertility (E2) and late sown, moderate fertility (E3) levels at I.A.R.I., Regional Station, Pusa. Sowing was done in a randomized block design with three replications by dibbling the seeds 15 cms apart in rows 2.5 m long and 30 cms apart. Two rows were alloted to each parent and F1 per replication. Five plants per row were selected at random for recording data. Grain number per ear was recorded from the main tiller where-as 250 seeds were counted out of the bulk seed of the ten selected plants/replication. The analysis of variance for combining ability and estimation of various effects was done following the technique of Griffing (1956) for system 2, Model 1 and the genetic parameters were estimated as described by Hayman (1954). Results and Discussion Analysis of variance (Table 1) indicated highly significant differences due to progenies for the two measured traits under all the three environments. The variation among parents, F1 hybrids and parents vs hybrids comparison, were highly significant, showing thereby that considerable genetic diversity was present in the experimental material. Grain number/ear Kalyansona, in general recorded the maximum (74-61) grain number/ear and was followed by Safed Lerma, Sonora 64 and E 4845. NP 4 had the minimum (45-35) grains/ear under all the environments. Rest of the parents were intermediate. Amongst the F1S, there were only two crosses in E1, none under E2 and as many as four in E3 which were significantly superior to the best parent. The analysis of combining ability revealed that both gca and sca mean squares were highly significant under all the three experimental conditions, indicating the importance of both additive and non-additive gene effects in the control of grain number/ear. The present results are in close agreement with those of Singh et al. (1969), Paroda & Joshi (1970), Rehman (1979), and Thombre et al. (1983). The estimates of gca effects and mean value of the parents under the three environments are presented in Table 2. Cultivars, Kalyansona, Sonora 64, Safed Lerma and E 4845, in that order, were the best general combiners over environments, whereas all other varieties showed negative gca effects. It is to be noted that the value of the parents per se was in fairly close agreement with the one obtained by gca estimates. Thus the selection of the parents based on parental values, which has been the common practice is reasonably good. |
| --> Next |