An outlook on wheat health in Europe from a network of field experiments

Wheat disease management in Europe is mainly based on the use of fungicides and the cultivation of resistant cultivars. Improving disease management implies the formal comparison of disease management methods in terms of both crop health and yield levels (attainable yield, actual yield), thus enabling an assessment of yield losses and yield gains. Such an assessment is not available for wheat in Europe. The objective of the analysis reported here is to provide an overview of wheat health and yield performance in field experiments in Europe. Data from field experiments in six European countries (Belgium, France, Germany, Italy, Norway, and Sweden) conducted between 2013 and 2017 were analysed to that aim. Relationships between multiple disease levels, yield, level of cultivar resistance, level of fungicide protection, and weather patterns were assessed. The analyses included 73 field experiments, corresponding to a total of 447 [fungicide protection level x cultivar] combinations. Analyses across the six countries led to ranking the importance of foliar wheat diseases as follows, in decreasing order: leaf blotch (septoria tritici blotch, septoria nodorum blotch, and tan spot), leaf rust, yellow rust, and powdery mildew. Fusarium head blight was observed in France and Italy, and stem rust was sporadically observed in Italy. Disease patterns, crop inputs (fertiliser, fungicides), and yields widely varied within and across countries. Disease levels were affected by the level of fungicide use, by cultivar resistance, as well as by weather patterns. While this analysis enables a better documentation of the status of wheat health in Europe, it also highlights the critical need for policies in Europe enabling a more judicious use of pesticides. First, common standards for field experiments are needed (experimental designs and protocols; disease assessment procedures and scales; references, including reference-susceptible cultivars); second, assessments in farmers' fields – and not in research stations – are necessary; and third, there is a need to use available process-based crop models to estimate attainable yields, and so, yield losses.