Selection for High Grain Yield and Determination of Heritable Variation in Biomass Yield, Grain Yield and Harvest Index in Different Plant Height Classes of Breeding Populations of Spring Wheat (Triticum Aestivum L.).

Abstract

High grain yield is a prerequisite in the release of wheat genotypes intended for production under high input and capital-intensive cultures such as is practiced in Zimbabwe. Zimbabwe experiences shortages of wheat grain due to limited irrigation facilities and also due to the tropical nature of its climate, which confines wheat production to the dry, winter months, under costly irrigation regimes.

Breeding programmes serving wheat production systems which target high grain yields, such as is required in Zimbabwe should aim to generate elite wheat germplasm which carries high frequency of high yielding genotypes. Such breeding programmes should conduct efficient variety screening techniques during the promotion of advanced breeding material into replicated variety evaluation stages. This should ensure maintenance of high grain yield performance in the subsequent advanced variety testing for the purposes of variety release.

It is not feasible to evaluate a very large number of wheat lines in statistically designed experiments, under the constraints of time, resources and space. Thus only a limited number of accessions are, in time, advanced into preliminary statistically designed field experimentation with the rest being discarded. The historical dilemma facing wheat breeders has been the identification of reliable selection criteria for high grain yield performance during early breeding generations (McGinnis and Shebeski, 1968; Knott, 1972), and later in the screening of homozygous lines for advanced testing (Fischer and Kertesz, 1976; Ellison, Latter and Anttonen, 1985; Sharma, 1993). Wheat scientists have hitherto failed to identify simple plant characteristics which relate closely to grain yield potential at field crop densities. Reasons for this vary from the low additive genetic variances relative to the environmental and error variances that these simple plant traits are associated with (Fischer and Kertesz, 1976.), to the genotype x environment interaction and the presence of allometric relationships (Grafius, Thomas and Banard, 1976). The need to establish reliable criteria to screen homozygous wheat lines prior to replicated trial testing should continue to be addressed. One solution suggested has been the use of small (micro) or hill plots for early yield testing (Frey, 1965; Jensen and Robson, 1969; Briggs and Shebeski, 1971; Ellison, et al., 1985). Such plots attempt to simulate to some degree, the competitive environment of a crop, given the usual limitation on seed during early testing (Fischer and Kertesz, 1976).

The utility of morpho-physiological traits; biomass, grain mass and harvest index of single row or hill plots as indicators of yielding ability has produced varying results. Fischer and Kertesz (1976) found a strong phenotypic correlation between microplot and field crop grain yield of spring wheat in Northwest Mexico. The relationships of biomass and harvest index of microplots with grain yield of large plots were not included in their investigation. Relationships of the three morpho-physiological traits; biomass, grain mass and harvest index of thinly seeded plots with grain yield in commercial stands were evaluated for hard red winter wheat in the Southern Great Plains of USA by Sharma and Smith (1987). Grain mass followed by harvest index of thinly seeded plots was strongly correlated with grain yield in field crop stands. Biomass of thinly seeded plots was only moderately correlated with grain yield of commercial stands. This investigation did not however discriminate the effectiveness of the three selection criteria over different plant structure. In a later study by Kramer, Van ooijen and Spitters (1982), moderate to strong genetic correlation between grain yield of microplots of various sizes and mean grain yields in replicated trials of spring wheat in Wageningen, The Netherlands, were obtained.

Strong associations were encountered at wide inter-row spacing than at narrow-row spacing, suggesting that the widening of inter-row spacing of micro plots may be advantageous. Again the other two selection criteria, biomass and harvest index of microplots, were not examined in their study and the influence of plant stature on the relationships of microplot grain yield and large plot grain yield, were not ascertained.

Sharma (1993) used short stature spring wheat genotypes and correlation between biomass of microplots at the F3 generation and grain yields of corresponding genotypes in replicated trials in the following year in Rampur, Nepal, under a production culture similar to that practiced in Zimbabwe. The correlated response was higher at a high productivity system. Sharma’s 1993 study showed a negative relationship between biomass and harvest index. Grain mass and harvest index of microplots as selection criteria for high grain yield performance were not evaluated.

The evolution of the influence of plant stature on the relationship between biomass and harvest index of spaced plants with grain yield performance in replicated trials was attempted by McVetty and Evans (1980) in either tall or short stature groups of winter wheat cultivars. Harvest index of spaced plants was strongly correlated with grain yield among tall genotypes but biomass of spaced plants was effective among the short statured genotypes. Grain mass of spaced plants was not included as a selection criterion. More over the selection units were not in the form of microplots, but were spaced plants. It is evident that previous studies have not systematically and concurrently evaluated the effectiveness of the three morpho-physiological traits, biomass, grain mass (earmass) and harvest index (HI) of microplots, as selection criteria for high field crop grain yield performance for the purpose of establishing their relative selection efficiency. Plant stature appears to influence the nature and degree of correlated responses of grain yield to the different auxiliary traits in question. Hence, it is important to discriminate the relative selection efficiencies of the three criteria by plant height strata.

It is expected in the study that if the confounding effects of plant stature may be removed by stratifying populations into uniform plant height strata at the advanced stages of segregation, hill plot characters such as biomass, grain mass or HI, may be more reliable in their indications of grain yield potential of wheat genotypes. Further more, if the single ear progenies are propagated in wide rows and in ideal conditions of growth, characterized by uniform soils, uniform crop establishment and uniform moisture status, it is assumed that environmental component of variation in the expression of the three single ear-to-row progeny attributes, would be minimised.

Efficient utility of auxiliary traits as selection criteria for grain yield requires information on the mode of inheritance and inter-relationships of the selection traits and the responding trait in the actual environmental conditions intended for their application. Thus inheritance and mutual association studies on grain yield and the auxiliary traits used in this study should be conducted under the wheat production conditions of Zimbabwe.

The hypotheses tested in this study were:

1. The effectiveness of indirect selection for grain yield of wheat (Triticum aestivum L.) using auxiliary triats; biomass, earmass and harvest index of single-ear progeny plots is similar in the three selection criteria and is not influenced by height of the plant.

2. The mode of inheritance of grain yield of spring wheat and those of its related traits; biomass, earmass, HI, number of productive tillers and plant height under the irrigated and heavily fertilized culture in Zimbabwe, are predominantly additive in expression and are environmentally stable.

The above hypotheses were tested by the following objectives:

1. To measure the direct selection response of grain yield of spring wheat genotypes in replicated trials at two locations.

2. To measure and discriminate the relative selection efficiencies of single-ear progeny traits; biomass, earmass and HI in the indirect selection for grain yield of wheat over different plant height strata.

3. To ascertain correlated responses to selection among the selection traits by measuring phenotypic and genotypic correlations among them.

4. To study the inheritance of grain yield and the related traits; biomass, earmass, HI, number of productive tillers and plant height, under Zimbabwean irrigated conditions over two locations.