Taking advantage of architectural diversity to improve control of splash dispersed disease in cultivar mixtures : a modelling study based on field experimentation
Résumé
Using cultivar mixtures, i.e. growing several cultivars with different resistance traits in a same field, allows protection of susceptible cultivars and fungicide use reduction. Protective mixture effect against splash dispersed diseases such as septoria tritici blotch is variable depending on environmental conditions. Understanding disease reduction mechanisms is critical for improvement of mixture design and efficiency. Major mechanisms are related to modification of disease dispersal patterns within the mixture. Increased distance between two susceptible plants reduces spore transfer, while the presence of resistant plants provides a barrier to spore dispersal. Common mixture design recommendations include choice of cultivars with similar architecture in order to constitute homogeneous canopies which should minimize competition and ease agronomical management. However, no information is available on the possible impact of mixing cultivars with contrasted architecture on disease dispersal. We used a mechanistic modelling approach based on experimental results to characterize splash dispersal patterns in wheat cultivar mixtures with either uniform (homogeneous canopies) or contrasted straw height (heterogeneous canopies). We then performed numerical experiments to study the impact of variations in location of susceptible and resistant tissues within the canopy. Dispersal events were simulated based on field observations. Chosen rain dispersal events occurred during the post-flowering phase of wheat cycle, where disease develops on top leaves and reduces yield. At this stage, canopy structure is fully developed and fixed. Architectural measurements of individual plants were used to reconstruct static 3D mockups of pure stands and mixed canopies. Mockups provided information on resistant and susceptible tissue location within the canopies and were used to compute splash dispersal. Model inputs also included leaf disease severity and rain characteristics measured during studied dispersal events. Simulated spore dispersal was consistent with spore fluxes measured in the field. Disease reduction mechanisms that are specific to heterogeneous canopies were identified. Upper leaves of the tall cultivar acted as umbrellas by intercepting raindrops and reducing splash on diseased leaves at the bottom of the canopy. On the other hand, contaminated splash droplets produced at the same location were trapped by leaves of the lowest cultivar, thus reducing contamination of upper leaves of the tallest cultivar. Modelling results allowed quantification and hierarchisation of dispersal processes occurring in heterogeneous mixed canopies and outlined the importance of the location of resistant and susceptible organs within the canopy.