Abstract
Plants shape their surrounding soil, influencing subsequent plant growth in a phenomenon known as plant-soil feedback (PSF). This feedback is driven by chemical and microbial legacies. Here, we cultivated six crops from two functional groups, i.e., three grasses (Lolium, Triticum, and Zea) and three legumes (Glycine, Lens, and Medicago), to condition a living soil. Subsequently, the same species were sown as response plants on conspecific and heterospecific soils. We employed high-throughput sequencing in tandem with soil chemistry, including total organic matter, pH, total nitrogen, electrical conductivity, phosphorus, and macro and micro-nutrients. Our results showed that Glycine exhibited the strongest negative PSF, followed by Triticum and Zea, while Lolium displayed low feedback. Conversely, Lens demonstrated robust positive PSF, with Medicago exhibiting slight positive feedback. Soil chemistry significance indicated only higher Cl content in Triticum soil, while Lens displayed higher Zn and Mn contents. Microbial diversity exhibited no significant variations among the six soils. Although conditioning influenced the abundance of functionally important microbial phyla associated with each plant, no specificity was observed between the two functional groups. Moreover, each crop conditioned its soil with a substantial proportion of fungal pathogens. However, co-occurrence analysis revealed a strong negative correlation between all crop's biomass and fungal pathogens, except Glycine, which exhibited a strong negative correlation with mutualists such as Arthrobacter and Bacillus. This underscores the complexity of predicting PSFs, emphasizing the need for a comprehensive understanding of plant interactions with both pathogens and mutualists, rather than focusing solely on host-specific pathogens.