(go to KOMUGI Home) (go to WIS List) (go to NO.72 Contents)


Material and Method

The present study was carried out on 45 varieties consisting of 36 diverse varieties and four biblends (equal parts) of durum wheat and five commercial bread wheat varieties. All the genotypes were grown in randomized block design consisting of three replications in 3 artificially created environments as given below :
1) Normal sown, high fertility and irrigated (E1)
2) Medium fertility and restricted irrigations (E2)
3) Rainfed and low fertility (E3).

The data were recorded for traits as shown in
Table 1. Stability parameters of 45 varieties were computed using the model proposed by Eberhart and Russell (1966).


Results and Discussion

The introduction of dwarf wheat varieties led to the striking increase in the wheat yields all over the world. However, in developing countries limited input conditions put considerable constraints on the full exploitation of genetic potential of the new wheat varieties. In developing countries like India there is large area of rainfed wheat where annual fertilizer consumption per unit area is also very low. Under such situation it becomes imperative to test the wheat genotypes under rainfed, restricted irrigation/fertility conditions to identify their predictable response to inputs and irrigation levels. Hence G x E interaction would be of great interest if it is estimated under aforesaid conditions which may be more realistic to wheat growing situations. Eberhart and Russell (1966) suggested an approach to identify stable and responsive genotypes. According to their model, a stable genotype is characterized by average response (unit regression) and least deviation accompanied by high mean.

Highly significant variances due to varieties revealed the presence of genetic variability in the material included in this study for all the traits studied (
Table 1). The linear component of G x E interaction was significant for plant height, tiller number, biological yield, grain yield, and harvest index. Therefore, prediction for these traits appeared to be feasible. This also shows the existence of additive type of gene-action for these traits. Significance of pooled deviation for all these traits showed that the varieties differed considerably with respect to their response to different environments for these traits. Singh and Singh (1980) and Jatasara and Paroda (1980) observed similar estimates of linear proportion of G x E interaction for grain yield in bread wheat. Kaltsikes and Larter (1970) observed the same for grain yield, plant height and days to maturity in durum wheat. Nanda et al (1983), in bread wheat, found larger predictable portion for plant height and number of spikelets par spike.

Seven best varieties based on linear (b) and non-linear (S-2d) components of G x E interaction and high mean values are listed in
Table 2. Varieties CC 530, JU 72, R 6009 gave high grain yield, b value above average and S-2d non-significant. Therefore, these varieties were suitable for favourable environment. However, WH 822, R 6011, HI 8078, CPAN 6038 and R 6032 produced average grain yield, b value near unity and S-2d non-significant hence were stable and responsive. Amongst aestivum wheat varieties C 306 was desirable because of its average grain yield, stable and responsive behaviour. Similar results have been reported by Chowdhury and Paroda (1983) for C 306 variety. HD 2009 was also found to be average performer and stable. WH 157 was suitable for favourable environment since it had above average linear regression.

<--Back |-->Next

(go to KOMUGI Home) (go to WIS List) (go to NO.72 Contents)