Soil metabolic factors, not pathogens, drive replant disease: insights from a 20-year soil legacy experiment

Abstract Background Replant diseases (RDs) often stem from nutrient deficiency, allelopathy, or pathogen accumulation. However, the RDs of certain crops are long-lasting and their causes remain unknown. We examined Panax quinquefolius RD in a space-for-time soil sequence representing crop rotation restoration over 1, 10, and 20 years using multiomics and bioassays. Results Compared with the soils with no ginseng cultivation history, we found 110 potential influencing factors (not including pathogens) related to RD in the 1-year rotation soil, 53 of which remained unrestored after 20 years. Soil pH and the levels of organic nutrients (amino acids, carbohydrates, and alditols), allelopathic-promoting metabolites (phenolic acids, amines, pyridines, etc.), and beneficial bacteria (Sphingomonas, Burkholderia-Caballeronia-Paraburkholderia, and Terrabacter) and fungi (Acremonium, Penicillium, and Naganishia) decreased, while the levels of allelopathic inhibitory factors (pyruvic and fatty acids) increased. The overall soil metabolism was downregulated. We confirmed that decreased organic nutrient levels and increased levels of allelochemical inhibitory substances impaired P. quinquefolius growth. Network analysis showed that soil metabolic factors rather than microbial factors were dominant. Conclusions Through a partial least squares structural equation model, we found that the overall changes in nutrient levels and metabolic and microbial factors contributed to short-term RD (1–10 year) persistence, whereas long-term RD (after 20 years) primarily resulted from disordered soil metabolite levels and pathways. This research will help deepen our understanding of the relationship between RD and potential changes in the factors influencing RD that are caused by soil legacy effects of valuable plants and provide theoretical guidance for effective soil quality improvement.