Generation mean analysis for some agronomic traits at two bread wheat crosses under two different moisture conditions

In Iran, as in most poor and developing countries, wheat is the most important food in the diet of the people. Therefore, the field of study and research on various aspects of this important and strategic crop is still open to relevant researchers and experts. Given that the achievement of yield potential in crops, including wheat, is influenced by genetics, environmental conditions, their interaction and the application of proper management in the optimal use of available resources, so the genetic capacity of the existing germplasm should be used to produce suitable breeding lines in order to make the most of natural and climatic resources. Therefore, screening genotypes, selecting the proper parents, and modifying them through conventional and new methods can still be helpful. Estimating genetic effects and studying the inheritance pattern of important agronomic traits and in general knowledge of germplasm genetic information and knowledge of the genetic system of the studied trait is one of the most useful tools in designing breeding methods to improve target traits. This study aimed to estimate the genetic model controlling important agronomic traits and also to estimate heterosis and heritability in two bread wheat cross under normal and terminal drought stress conditions using generation mean analysis method and multivariate regression analysis.

A field investigation materials consisted of basic generations obtained from the crosses of a local cultivar “Marvdasht” (female parent) with two cultivars “Sistan” and “Norstar” (male parents). Both the crosses and resultant progenies were performed and developed under field conditions. All the six generations derived from the above two crosses were sown under two distinct conditions, normal (non-stress) and terminal drought stress in a randomized complete blocks design with three replicates at the Research Farm of Razi University, Kermanshah, Iran during the next cropping season (2015-2016). Terminal drought stress was imposed in Mid-May; while non-stress plots were irrigated whenever required. Weeds were controlled manually when necessary. The rainfall in 2015-2016 was 653 mm. Data from 10 plants in each of P1, P2 and F1 generations, 30 plants in F2 and 15 plants in each of BC1 and BC2 were randomly recorded per replication. Statistical analyses were done using different methods and software.

Significant differences between the generations were found for most of the traits which indicated the presence of genetic variation in the plant populations and scope of improvement through breeding methods such as selection and expression of heterosis. The scaling and joint scaling tests revealed that the simple additive-dominance model was not sufficient to explain the genetic variation in the crosses for all the studied traits apart from the number of spike per plant, main stem diameter and number of spikelet per spike at Marvdasht × Sistan, and hundred kernel weight and peduncle length at Marvdasht × Norstar. It could be concluded that the inheritance of these traits is governed by epistasis gene action. The results of analysis of variance obtained by regression method showed that fixable genetic effects had the highest relative contribution of the generation sum of squares for most of the traits in both crosses and under both normal and drought stress conditions. High estimates of broad-sense heritability for hundred kernel weight (0.67) in normal conditions at Marvdasht × Sistan cross and plant height (normal: 0.76 and stress: 0.67) and peduncle length (normal: 0.72 and stress: 0.67) at the Marvdasht × Norstar cross shows that the effects of dominance have a greater role in the inheritance pattern of these traits and therefore it is possible to produce hybrid varieties to improve these traits.

The results of generation mean analysis showed that the type and action of genetic effects were variable in both crosses and for different traits and indicate the need to adopt a special breeding method to improve them. The presence of duplicate mode of gene interactions signified the involvement of epistatic effects for most of the traits. Based on the present investigation, it could be inferred that the genetic control model was somewhat similar for most of the traits under both conditions at both crosses, and it has not been largely affected by drought stress.