The negative contribution of crop residues as a source of inoculum for plant diseases is well established. However, microbial ecologists have long reported positive effects of residues on the stability of agrosystems and conservation tillage practices have become increasingly widespread. Most studies have suggested that large microbial communities should be taken into account in plant disease management, but we know little about their ecological interaction with pathogens in the crop residue compartment. This review focuses on microbiomes associated with residues within the context of other microbial habitats in cereal-producing agroecosystems such as phyllosphere or rhizosphere. We connect residue microbiome with the survival of residue-borne fungal plant pathogens, thus combining knowledge in microbial ecology and epidemiology, two disciplines still not sufficiently connected. We provide an overview of the impact of residues on cereal disease epidemics and how dynamic interactions between microbial communities of non-buried residues during their degradation, along with soil and multitude of abiotic factors, can contribute to innovative disease management strategies, including next-generation microbiome-based biocontrol strategies. Starting from the classical but still relevant view of crop residues as a source of pathogen inoculum, we first consider possibilities for limiting the amount of residues on the soil surface to reduce the pathogen pressure. We then describe residues as a transient half-plant/half-soil compartment constituting a key fully fledged microbial ecosystem: in other words, an ecotone which deserves special attention. We focus on microbial communities, the changes in these communities over time and the factors influencing them. Finally, we discuss how the interactions between the microbial communities and the pathogens present on residues could be used: identification of keystone taxa and beneficial assemblages, then preservation of these taxa by adapted agronomic practices or development of synthetic communities, rather than the introduction of a single exogenous biocontrol species designed as a "treatment product".