Grapevine wood microbiome analysis identifies key fungal pathogens and potential interactions with the bacterial community implicated in grapevine trunk disease appearance

BACKGROUND: Grapevine trunk diseases (GTDs) is a disease complex caused by wood pathogenic fungi belonging to genera like Phaeomoniella, Phaeoacremonium, Fomitiporia, Eutypa and members of the family Botryosphaeriaceae. However, the co-occurrence of these fungi in symptomatic and asymptomatic vines at equivalent abundances has questioned their role in GTDs. Hence, we still lack a good understanding of the fungi involved in GTDs, their interactions and the factors controlling their assemblage in vines. We determined the fungal and bacterial microbiome in wood tissues of asymptomatic and symptomatic vines of three main Greek cultivars (Agiorgitiko, Xinomavro, Vidiano), each cultivated in geographically distinct viticultural zones, using amplicon sequencing. RESULTS: We noted that cultivar/biogeography (lumped factor) was the strongest determinant of the wood fungal microbiome (p < 0.001, 22.7%), while GTD symptoms condition had a weaker but still significant effect (p < 0.001, 3.5%), being prominent only in the cultivar Xinomavro. Several fungal Amplicon Sequence Variants (ASVs), reported as GTD-associated pathogens like Kalmusia variispora, Fomitiporia spp., and Phaemoniella chlamydosporα (most dominant in our study), were positively correlated with symptomatic vines in a cultivar/viticultural zone dependent manner. Random Forest analysis pointed to P. chlamydosporα, K. variispora, A. alternata and Cladosporium sp., as highly accurate predictors of symptomatic vines (0% error rate). The wood bacterial microbiome showed similar patterns, with biogeography/cultivar being the main determinant (p < 0.001, 25.5%) of its composition, followed by the GTD status of vines (p < 0.001, 5.2%). Differential abundance analysis revealed a universal positive correlation (p < 0.001) of Bacillus and Streptomyces ASVs with asymptomatic vines. Network analysis identified a significant negative co-occurrence network between these bacterial genera and Phaemoniella, Phaeoacrominum and Seimatosporium. These results point to a plant beneficial interaction between Bacillus/Streptomyces and GTD pathogens. CONCLUSIONS: Our study (a) provides evidence that GTD symptomatic plants support a wood fungal microbiome, showing cultivar and biogeography-dependent patterns, that could be used as a proxy to distinguish between healthy and diseased vines, (b) points to strong interactions between the bacterial and fungal wood microbiome in asymptomatic vines that should be further pursued in the quest for discovery of novel biocontrol agents. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s40793-021-00390-1.