Breeding for BYDV Tolerance in Wheat as a Basis for a Multiple Stress Tolerance Strategy

摘要

There are major theoretical and practical differences related to tolerance vs. resistance breeding. The correlation between disease resistance sensu stricto and protection against crop losses is not what the layman would expect.It seems entirely logical that a resistance gene reducing virus levels should also protect against losses. Yet tolerance genes simply reduce economic loss factors, regardless of virus titers. Resistance sensu stricto (reflected in reduced ELISA titers) always leads to reduced virus spread in the field, but not necessarily to smaller losses in individually infected plants. In studies of the Barley yellow dwarf virus (BYDV), visual symptom data, somewhat based on loss of chlorophyll, were much less reliable in wheat than in other species. Moreover, in wheat, susceptibility sensu stricto (reflected in high ELISA titers) is not related to a higher probability of serious losses. Species certainly differ in the manner in which virus titers correlate with symptoms and damage. Correlation between ELISA titers, symptomatology and losses may be found in bulk plant population samples, in cases where infection occurred at an early stage and under growing conditions that increased damage. Yet in species like breadwheat, durum wheat, oats and triticale, the correlation between symptoms and damage might become significant only under the most severe epidemic conditions; in average conditions, up to 25% economicloss can occur in the presence of minimal or near invisible symptoms. In fact, durum wheat and bread wheat often give counterintuitive research results which appear at face value like a violation of the Koch postulate. In these species, reliable tolerance data necessitate a study for many years and at many contrasting sites, using quantitative data on many parameters. This is because growing conditions represent an important variable that can modify the losses caused by barley yellow dwarf virus (BYDV)infection, and this interaction is serious in the genus Triticum. Inideal growing conditions, some amount of phloem damage may have very little real effect on crop yield, hi the presence of secondary stresses, the deleterious effects induced by BYDV maybe increased a lot. hi cases where secondary stresses are mild, BYDV-induced losses may also be minor, and this reduces the difference between tolerant and sensitive lines. Our current understanding of tolerance mechanisms is that diverse aspects ofbiotic and abiotic stress resistance can also improve virus tolerance, and that the tolerance mechanisms of one line or cultivar can be quite different from the mechanisms present in another. This has theoretical and practical consequences. Due to the complexnature of tolerance and to its interaction with environmental stress, tolerance data obtained at one site may correlate poorly with data from another site. The practical caveat is that BYDV information on bread wheat and durum wheat must be seen as muchmore site-specific and year-specific that the data on rusts for example, and that many test sites and/or years of data are needed for a valid evaluation of tolerance. A practical application of the current understanding of tolerance is that within breeding programs, BYDV can be used to increase the overall effects of secondary stresses in wheat and oats. Since yield stability and yield potential are both related to the capacity of plants to mitigate stress-induced losses, BYDV is used to make the response to secondary stress easier to evaluate. Instead of being simply a disease, BYDV is now one of the most essential tools in germplasm enhancement and cereal breeding, in order to identify stability of yield and quality, and also water and nutrient useefficiency.