Yield Potential of Bread Wheat Hybrids Produced by Genesis

Experts predict that the demand for wheat will increase by 40% in the next 20 years. Meeting this increased demand constitutes an enormous challenge, especially because yield levels of wheat may have reached a plateau in recent years. To increase wheat yield levels and meet the growing world demand for wheat, breeders need to exploit new technologies such as those now used to facilitate the development of hybrid wheat. Although wheat hybrid production has been attempted in the past, it was for the most part abandoned due to the difficulties inherent to its development. However, interest in hybrid wheat was renewed in the 1990s, due to the availability of new, more effective chemical hybridizing agents (CHAs).

Compared to cytoplasmic male sterility systems, effective CHAs permit the production of large numbers of hybrids from very diverse germplasm in a very short period of time. Genesis, a CHA produced by the Monsanto Company, induces male sterility in female lines when applied at the appropriate growth stage. It was approved by the US Environmental Protection Agency in 1997. That same year, CIMMYT re-started its hybrid wheat program in collaboration with Monsanto.

Heterosis may give hybrids a yield advantage over "normal" wheats. Yield advantages as high as 10-17% over the leading check cultivars have been reported for hybrid wheat in Italy and the UK. In China, a yield advantage of up to 30% has been reported for hybrids produced using CHAs. Other advantages have also been reported, such as hybrids' higher yield stability compared to pure lines.

The objectives of this study were to identify spring bread wheat hybrids with high yield potential under the irrigated conditions of the Yaqui Valley in northwestern Mexico.

Materials and Methods

During the 1996-97 growing season, advanced CIMMYT bread wheat lines were crossed as females to five high yielding male lines through the use of Genesis. Hybrids were produced by crossing the same set of female lines to different male lines in a factorial mating design that allows calculating the lines' combining abilities for different traits. Female and male variation provides information about the general combining ability of female and male lines, respectively, whereas female by male interaction provides information about specific combining ability for hybrid combination.

A total of 148 hybrids were included in two separate yield trials during the 1997-98 growing season. One trial included hybrids that adhered to a complete factorial design (11 female lines with five male lines). Some combinations did not yield enough seed to meet the requirements of a factorial design. Therefore, all those hybrids were included in the second experiment. The first year hybrids were evaluated for yield potential, height, bread-making quality, and leaf rust reaction, and the best 42 hybrids were promoted to the second year of evaluation in the 1998-99 growing season.

All yield trials were conducted under fully irrigated conditions in Ciudad Obregon, Sonora, Mexico. The experimental design was a latinized alpha lattice. Plot size was 4.8 m², including two beds, 0.8 m wide and 3 m long. Three and two replications were evaluated during the 1997-98 and 1998-99 growing seasons, respectively. Two commercial varieties (Bacanora and Rayon) were included as checks in the trials. All female and male parental lines were also included in the trials to calculate heterosis. Seeding density for hybrids and parental lines was calculated based on 200 viable seeds m^-2. Each check was planted at two different densities: the farmer's density (100 kg ha^-1) and the hybrid density.

Grain yield was adjusted to 12% moisture content. Adjusted entry means for yield were calculated for each experiment, and then the data were combined over two years as a randomized complete block design.

Hybrids that yielded significantly more than the commercial cultivar Rayon are presented in Table 1. The yield advantage of these hybrids over Rayon ranged from 13 to 17% when data were combined over two years. However, the difference between these hybrids and their mid-parent yield values was not significant. Mid-parent heterosis for the highest yielding hybrid was 6% (Figure1).

Overall general combining ability (GCA) effects were more important than specific combining ability (SCA) effects for yield (Table 2). These results indicated that additive gene effects were more important than non-additive gene effects in the variation expressed among these hybrid combinations. It is possible to obtain hybrids with high mid-parent heterosis for yield. For example, hybrids that included line 3 as a male parent had the highest mid-parent heterosis when compared to hybrids with male parent 1 and/or 4 (Figure 2). However, absolute yield levels of these hybrids were not always as high as those of the check cultivar and the other hybrids. The yield of male parent 3 was the lowest (6.3 t/ha) of all entries in the trial, resulting in lower mid-parent values. These findings indicated that low yielding lines should not be selected as parents, since additive gene effects are more important than non-additive gene effects for yield.

When yield levels of hybrids were compared to those of their high-, low-, and mid-parents, a significant positive correlation was found; the highest correlation coefficient was observed with mid-parent yield (Figure 3). In other words, the higher the mid-parent value, the higher the hybrid yield.

The difference between the yield levels of the highest yielding hybrid and its high-parent was not significant; however, the yield advantage over the leading cultivar was high in both years. This high yielding parental line also gave the highest GCA effects when used either as female or male parent in hybrid combinations. Because of its high susceptibility to leaf rust, this line cannot be grown commercially. However, leaf rust susceptibility can be overcome by using the major genes present in the complementary parent. For this reason, we continue to use this line in our hybrid program.

Conclusions

These preliminary results indicate that there is heterosis for grain yield in CIMMYT wheat lines under irrigated conditions. Hybrids that yielded better than leading cultivars were obtained.

Based on the positive correlation between yields of hybrids and their mid-parent values and relatively large, significant additive GCA effects, it can be concluded that among the germplasm used in this study the highest yielding advanced lines will produce the highest yielding hybrids.

We would like to acknowledge the full technical support and partial financial support of the Monsanto Company for this work.