Abstract

Oilseed rape is a crop in which genotypes strongly influence each other when grown at normal plant densities. Competition effects were investi-gated in two experiments with various mixtures of winter oilseed rape genotypes. Mixtures of 75 % F1 hybrids and 25 % homozygeous doubled haploid plants already produced the same yield as a stand of 100 % hybrid plants. In subsequent generations of synthetic cultivars, the contribution of indivdual components to the yield of the synthetic cultivar changed from generation to generation in an unpredictable way.

Introduction

Rapeseed is a crop, in which genotypes can strongly influence each other within a heterogenic population by competition. This is of importance in two respects. Like in many other crops it is discussed to mix cultivars to improve yield and yield stability. Moreover, many present cultivars do not consist of only one genotype; synthetic cultivars are established from a number of various componentes, and in hybrid cultivars, a low percentage of parent lines may be occuring.

Mixing effects have been studied in winter rapeseed earlier with somewhat unequivocal results. In several studies (Léon and Diepenbrock 1987, Léon 1991), two-component mixtures outyielded the cultivars in pure stands by 4 to 6 % and showed a superior yield stability (Léon 1991). However, Grabiec and Krzymanski (1984) observed in four-component mixtures in the average a 3 % lower yield than in the pure lines.

The objective of the present study was to investigate competition in two very different types of winter oilseed rape populations. In the first experiment we investigated mixtures between doubled haploid lines (DH) and their F1 hybrids in various mixing ratios; in the second experiment we analysed different generations of synthetic populations established by three DH lines each.

Materials and methods

Doubled haploid (DH) lines were used which were not preselected for yield performance. The material consisted of three groups:

• Set A with 3 DH lines derived from the cultivar 'Jupiter',
• Set B with 3 DH lines derived from the breeding line 'WW 933',
• Set C with 3 DH lines derived from three different populations.

Experiment 1: The DH lines of sets B and C were crossed in all possible combinations by hand pollination. Field tests were performed with mixtures from DH lines and F1 hybrids in various proportions. DH lines and F1 hybrids of each set were represented as a mixture of the three lines or hybrids, respectively. Field tests were performed with two replications at 7 environments: Svalöv, Sweden (1996), Göttingen, Germany (1996), Hohenlieth, Germany (1995, 1996), Teendorf, Germany (1995, 1996), and La Rochell, France (1995). Plot size varied between 10 m² and 16 m².

Experiment 2: For all three sets, the lines were mixed in equal proportions and several generations of synthetic cultivars (SYN-1, SYN-2, SYN-3) were produced by random pollination under isolation. Yield tests were performed at two locations (Svalöv and Landskrona, both Sweden) in one year. After harvest, the genotypes were seperated by starch gel electorphoresis of isozymes (Becker et al. 1992), and the contribution of each genotype to plot grain yield was determined.

Results and discussion

Experiment 1: As expected, the heterosis value was clearly higher in population C made up of unrelated lines, than in population B (Fig. 1). But it is also noteworthy that a considerable amount of heterosis was observed in population B, too. Thus it can be concluded that in rapeseed heterosis can play a decisive role for yield increase even in genetically confined population cultivars. But it is particularly interesting to note that in both populations the relation between hybrid fraction and yield is not linear. Already with a portion of 50% to 75% hybrid plants the resulting yields equal those of a pure hybrid stand. Probably at usual plant densities the hybrids in such mixed stands suppress the plants of the inbred lines completely. Therefore the fact that allogamy is only partially expressed in rapeseed does not hinder the utilization of heterosis in synthetic cultivars of winter rape to be as efficient as in completely cross fertilizing crop species.

Fig. 1: Grain yield of mixtures from DH lines and their F₁ hybrids

The results presented in Figure 1 are also important for hybrid breeding. Evidently it is not required for the full exploitation of the yield potential of hybrid cultivars that the "hybridity" of the certified seed is 100%. A certain amount of parental lines in the seed does not necessarily reduce the final yields.

Experiment2: Within the synthetic populations, the genotypic composition of the various generations was analysed by isozyme markers and the relative contribution of the various genotypes to grain yield was determined (Table 1). Although in the synthetics A and B the components were derived from one initial population and thus were rather similar genetically, considerable differences were stated regarding the composition of the SYN-populations. In particular, the fraction of the hybrid plants was very small in the early generations. This result can be explained only by the fact that under the environmental conditions of southern Sweden the rate of cross fertilization is very low in winter rape. The rather severe shifts in composition of the synthetic populations may also provide an explanation of unexpected changes in yield developments during the consecutive SYN-generations. These changes are the most important difficulty when producing synthetic cultivars in winter oilseed rape.

Table 1: Genotypic composition (%) of synthetic populations derived from three DH lines in winter oilseed rape

^a F₁ and all other possible types of crosses

Acknowledgments

The DH lines were provided by Hilleshög. We are grateful to Ann-Sofie Fält, Inga Mathiasson and in particular Britta Karlsson for their excellent technical assistance. For performing field tests thanks are due to the breeder colleagues in the companies of Svalöf Weibull, Norddeutsche Pflanzenzucht Hans-Georg-Lembke, Semundo, and Plant Genetic Systems. The work was supported financially by the Swedish Council for Forestry and Agricultural Research (SJFR).

References

1. Becker, H.C., Damgaard, C., and Karlsson, B., 1992: Environmental variation for outcrossing rate in rapeseed (Brassica napus L.). Theor.Appl.Genet. 84:303-306.

2. Grabiec, B., and Krzymanski, J., 1984: Attempts of heterosis uns to improve winter rape yield in Poland. GCIRC Bulletin 1984(1):14-16.

3. Léon, J., 1991: Heterosis and mixing effects in winter oilseed rape. Crop Sci. 31:281-284.

4. Léon, J., and Diepenbrock, W., 1987: Yielding ability of pure stands and equal proportion blends of Rapeseed (Brassica napus L.) with double-low quality. J.Agr.Crop Sci. 159-82-89.