Résumé
Varietal mixtures are a promising agro-ecological approach to stabilizing yields by reducing diseases. The effects of mixtures stem from modifications of epidemiological processes and plant–plant interactions, which are often underestimated and could explain some of the paradoxical observations that have been made in field trials. The role of plant–plant interactions in modifying the susceptibility of bread wheat and durum wheat to Septoria tritici blotch (STB) has not been determined. By producing full matrices of binary mixtures of varieties of each of these species in the absence of epidemics, this study aimed to determine the effect of such plant–plant interactions on STB symptoms—specifically area of necrosis (lesions) and production of pycnidia (spore-containing structures). We employed statistical modeling to compare the mean and variance of the phenotypes of focal plants across all mixture combinations of different genotypes versus the pure condition (one genotype only), and in each specific mixture versus pure conditions. The results demonstrated significant effects of plant–plant interactions on wheat susceptibility to STB. Notably, these interactions had specific rather than general effects, with some but not all genotypic combinations significantly influencing the susceptibility of the focal plant to STB. Furthermore, mixtures resulted in reduced necrosis with lower variance, but increased pycnidia formation. These results reinforce the need to consider specific plant–plant interactions for their contribution to trait means and variances in varietal mixtures.