Relative Performance Of Wheat Cultivars And Their Mixtures: Yield And Stripe Rust Severity

As a means of crop culture of using within-species crop heterogeneity to increase productivity, cultivar mixtures have been studied in many aspects in Europe and North America; however, a study of the subject is made a little in China, among them much is mainly focused on disease aspects of a few of small grain crops. The Province of Yunnan in the country is located in the North Temperate Zone and the sub-tropical zone, and in the province the area of wheat culture is about 55.9 thousand hectares and wheat stripe rust has tended to be increasingly severe in recent years. The first and main objective of this study was to compare yield performance of wheat cultivars and their mixtures. Its second objective was to evaluate effect of those mixtures on severity of stripe rust in wheat in natural infection.Based on the main objective of the study, multi-site and–treatment experiments on six wheat mixtures combined using seven cultivars-Fan19, Yin11-12, Chuanmai107, Jingmai10, Qingchun55, 46548-3 and An96-8, in which 2-7 components were included and equal proportions used, ware conducted in Qujing, Yunnan in 2003-2004, 2004-2005 and 2005-2006, respectively. A comparison between mixtures and their components was made by evaluation of mixture yields relative both to their component means (the results called effect on yield in the present study) and to yields of their higher- or highest-yielding components and lower- or lowest-yielding components.The results showed that in the absence of disease and under less severely diseased conditions, mixture yield exceeded their component means by 3.9% on average. In all 36 comparisons, frequencies of positive, zero and negative mixture effect on yield ware 69.4%, 8.3% and 22.3%, respectively, and the frequencies of the first two items ware summed up to 77.7%. For positive effects, 40.0% of comparative results ware significant (P<0.05); and in contrast, only 12.5% of comparative results ware significant (P<0.05)for negative effects. Thus, yield advantages of mixtures tended to be greater than disadvantages.Mixture effect on yield changed with individual environments (experimental years×sites×treatments) in its nature and magnitude, especially in the latter. There ware five out of the six mixtures found to be the maximum or minimum effects on yield in fertile soil. There was no significant relationship between mixture effect on yield and mixture component number: r=0.719(r0.05=0.811). Because strictly equally-proportional mixtures were used in the study, for mixture composition, factors influencing mixture effect on yield should be mainly due to component diversity.Although mixtures sometimes out-yielded their higher- or highest-yielding components, the differences between them ware seldom significant. In all 36 comparisons, there was only one mixture composed of Fan19 and Yin11-12 out-yielding its higher-yielding component significantly(P<0.05)in one of six environments, which accounting for 2.8% of the total comparisons. Yields of mixtures significantly inferior to those of their lower- or lowest-yielding components ware not found in the study.Mixtures'having an advantage over their components in yield stability was not demonstrated in the study using both measures of coefficient of variation and regression analysis, though yields of mixtures ware not any less stable than their components.Disease index of stripe rust in mixtures decreased to some degree, especially in the experiment conducted in 2004-2005, when the disease was relatively more severe than in 2003-2004. To sum up these two years'results, the reduction in disease index of stripe rust in mixtures relative to their component means was 57.70%, ranging from 37.2% to 72.2%. Based on data in 2004-2005, analysis indicated that there was a trend towards greater mixture effect on disease from increased component number; the main reason for this was that in mixtures, increase in component number substantially decreased the proportion of the component with the greatest disease index in pure stands so reduced its plant number.It was evident in the study that mixture effect on yield was not always closely correlated to mixture effect on stripe rust in wheatBased on the above results, considering that wheat cultivar mixtures did not damage stability of yield, that their yields had an advantage over their component means and that they reduced disease index of stripe rust, using cultivar mixtures combined using existing and new cultivars through field experiments would be of practical value to wheat production in Yunnan and could be a valuable supplement to growing alone pure line varieties in other wheat production areas in the country.