Background
In western France, rapeseed farmers face significant yield losses due to root parasitism caused by Phelipanche ramosa, a holoparasite plant of the Orobanchaceae family. Recently, a reduction of parasitic plant development has been observed in fields with history of severe infestation. In a same given pedoclimatic environment, this so-called soil-suppression of parasitism might presumably result from microbial mechanisms targeting P. ramosa. We tested this hypothesis on two soils from neighboring rapeseed fields, with similar physicochemical properties but contrasted parasitism, characterized as suppressive and conducive. We assessed these soils in a hydroponic co-cultivation system of P. ramosa and B. napus, and simultaneously sampled rhizosphere exudates weekly, as well as rhizosphere and rapeseed roots at three time points along parasite development. Comparisons were thus drawn between conducive and suppressive soils, both in untreated or gamma-sterilized conditions, regarding the effects of soil derived signaling metabolites on broomrape key early parasitic stages (i.e. germination and pre-haustorium development), late parasite development as well as soil bacterial and fungal structures.
Results
We demonstrate that the suppressive soil mitigates broomrape parasitism by reducing both parasite attachments and development, while causing tubercle necrosis. Activity assays on initial soils as well as co-cultivation rhizosphere exudates reveal that pre-attachment stages of broomrape are not affected by suppression. Root exudates, either they were collected from plants cultivated on suppressive or conducive soils, exerted similar contribution on the induction of broomrape seed germination and on haustoria formation. Thus, attachment and post-attachment stages are more likely to be suppressed by soil microbiota. Given the observed dissimilarities in microbial composition between the soils, we suggest that rather than bacteria, fungi play a greater part in parasitism suppression. Because suppression is not correlated with fungal diversity during co-cultivation, we suspect that in addition to general microbial interactions, specific groups of fungi participate in parasitism reduction. Further correlation network analyses (WGCNA) propose that three clusters of ASVs are indeed correlated with suppressive parasitic traits by their abundance. In these clusters, 7 ASVs are inversely correlated with parasitic attachment and one is positively correlated with the occurrence of necrotic tubercles, while being more abundant in the suppressive soil (DEseq2). These 7 ASVs are from Nectriaceae, Niessliaceae, Bartaliniaceae, Holtermanniaceae families as well as from Helotiales and Pleosporales orders, while the last ASV is identified as a Berkeleyomyces sp., a necrotrophic fungal genus known as a causal agent of black root rot.
Conclusion
This study explains the observed contrasted parasite development on two physiochemically similar soils by linking parasitic infestation to fungal dynamics in the rhizosphere. This finding extends our knowledge of disease suppressive soils to plant parasitism and proposes biological leads for potential biocontrol of broomrape.