Genotype x enviroment interaction for quality parameters of irrigated spring wheat

English: The primary aim of this study was to examine the effect of environment, genotype and their interaction on bread-making quality characteristics of irrigated spring wheat cultivars. The wheat samples were obtained from trials conducted in irrigation areas by the Small Grain Institute (Breeding Department), during 1997 and 1998. The trials were conducted at six localities in 1997 and seven in 1998. Among the localities five were in cooler and two in warmer irrigation areas. For the cultivars nine were common in both years but two were added in 1997, and one in 1998 making a total of 11 and 10 respectively. The source of cultivars was Small Grain Institute (5) and Sensako (7). Trials were conducted at experimental stations or farms of collaborators where the soil and climate are representative of a specific area. A randomised block design with four replications was used. Interrow spacing was maintained at 0.17 m and planting rate varied according to 1000 kernel mass of a specific cultivar. This was to ensure that a uniform stand of plants per m-2 is achieved. Fertilizers were applied according to recommendations on the basis of individual soil analyses. Other management practices were performed as required and a net plot of six rows (5.1 m-2) was harvested. The samples were analysed for a total of 21 quality characteristics, important for bread-making. The statistical analysis of the results showed that both genotype, environment and their interaction had significant influence on quality characteristics. This implied presence of variations among genotypes, environmental conditions as well as interaction effects in their response to quality parameters. However, above all, environment was the most dominant factor, which contributed to the variation for' most of the characteristics. Protein content (particularly flour) closer to 12% (optimum) and with good quality was very important. This character together with higher grain filling to ensure optimum starch-protein interaction were necessary for cultivars, sites and interaction effects to show good quality characteristics. Generally climatic conditions particularly temperature showed interference with quality parameters especially in warmer areas. This is due to the fact that high temperature has an influence on protein composition and definitely the ratios of glutenin-ta-gliadin part of the protein. However, for cooler areas, the protein quality was probably not affected, but the higher grain filling realised at some of the sites caused protein content to be low. This is again due to the fact that in high potential environments, nitrogen (N) is first utilised for maximum plant growth and yield. Thereafter the excess N is then translocated to the grain for storage as protein. Considering the quality parameters themselves, some showed negative correlations with others. This indicates that improvement of one characteristic will have a negative influence on the other, and this is the cause of breeders delay in releasing new cultivars. By using the canonical variate analysis (cómponent of AMMI), it was possible to see which characteristics group together in discriminating among genotypes, locations and interaction effects. Therefore with this analysis it was possible to determine which characters need appropriate genotype, environment or interaction effects to get optimum values. From the study results we can recommend that, T4, which ranked lower for most of the quality parameters may be risky for very large commercial production. However, this genotype may be used as a donor parent for improvement of alveograph P/L ratio for which it showed good results. Other genotypes, which also showed less potential, include, SST65, SSTSS, SST876 and Palmiet. These genotypes should be grown under conditions where they are expected to perform well. An example is Palmiet at Loskop, Koedoeskop and BullHill where it showed positive environment interactions for loaf volume at 12% protein. Also with environment protein yield closer to optimum (not very low), Palmiet may perform intermediately for most of the parameters. SST55 and SST65 showed higher loaf volumes at 12% protein and may be used to improve this parameter. Nevertheless they were lower in other parameters particularly mixograph development time, mixograph point score and SDS-sedimentation. Kariega ranked lower for alveograph P/L ratio in both years and may need improvement for this parameter. However, this genotype together with SST38, SST57, SST822, SST825, Inia and Marico showed intermediate to higher rank for most of the characteristics. Therefore they may be reliable at most of the sites unless environment becomes less optimal or negative environment interactions occur. For the locations, Loskop was the site that showed the most contrast with other sites. This is due to the fact that this site had values for mixograph development time, and to a lesser extent mixograph point score above recommended values for both years. Also the high protein realised at this site implies management practices should be aimed at reducing fertility and thus nitrogen yield. Therefore at Loskop, Koedoeskop (warmer areas) and Barkly West, cultivar selection and timing of seeding are necessary to ensure that grain filling is not influenced or does not coincide with higher temperatures. The higher hectoliter and kernel masses at most of the sites except Loskop implies higher grain yields and this was the main cause of realised low protein content. Therefore for other sites in addition to cultivar selection, management practices to increase fertility and thus protein content are necessary particularly at Prieska. Also since leaching may lose fertility, there is a need to control the irrigation water. Genotype x environment interactions were observed for protein content and related parameters. Grain filling was also responsible for this interaction particularly for the parameters influenced by starch-protein interaction. However, Prieska, and to a lesser extent Douglas showed less environment interactions compared to other sites. Therefore at Loskop, Koedoeskop, Hopetown, BullHilI and Barkly West identification of genotypes with positive environment interactions will be necessary. The canonical variate analyses for genotypes showed that 10 parameters were mainly responsible for the variations among genotypes. Mixograph development time, mixograph point score, alveograph P/L ratio, alveograph strength, loaf volume at 12% protein, SDS-sedimentation, breakflour yield, kernel mass (TKM and SKCS-W) and SKCS-diameter were responsible for genotype variation. Therefore for all these parameters, identification of cultivars, which will give desirable values, is important. Breeders may give them priority when making selection in their programmes. The canonical variate analysis for environments showed more parameters, about 13 being influenced by the environment variations. However, most of them were associated with protein content and grain filling. Most of the characteristics, which discriminated the interaction effects, were those under genotype and environment. Therefore identification of cultivars, which will give good results, and management to ensure higher protein with good quality and higher grain filling, are necessary. Taking into account the grading of wheat Prieska and Douglas showed to be stable and may give a BSS grade for most of the cultivars. BuIIHilI, Hopetown, Barkly West and Koedoeskop had intermediate stability and gave BS1 and BS2 grade for most of the cultivars particularly in 1998 due to low protein content. Loskop was alone showing to be highly unstable as it gave BS2 in 1997 for all cultivars, and in 1998 they were BPS and BP1. However, for this site despite the grade being higher, quality characteristics may show low values. Therefore cultivar selection and management are important to ensure that the grades are maintained and quality characteristics to be closer to optimum. Also mixing the products, particularly from Loskop and to a lesser extent BuIIHilI, Barkly West, Hopetown and Koedoeskop when the environments are not optimal with those of Prieska and Douglas will improve the quality. The probable cultivars to be mixed are SST825, Inia, SST822, SST57, SST38 and to a lesser extent Marico and Kariega. Therefore there is a need for processing industries and particularly researchers and farmers to work closer together. This will help to identify which cultivars and management practices need to be implemented at certain locations in order to have good quality crops. Also, since farmers are risk takers, despite following all the recommendations, environmental deviation may cause their products to have reduced/poor quality. In such a case, the government and processing industries should have a programme of supporting such farmers.