Tree identity shapes composition, substrate shifts dominance in deadwood fungi of an old-growth lowland beech forest

Wood-inhabiting fungal (WIF) communities are strongly structured by host tree species and environmental gradients. However, how WIF communities are structured among deadwood substrates when such drivers are minimized remains less clear, as do the implications for eDNA-based monitoring. We addressed these questions in a mono-dominant and unmanaged old Beech ( Fagus sylvatica ) forest, using long-read community profiling. Sampling followed a cross-fraction, tree-linked design that linked four deadwood substrate types to their source tree: logs, snags, attached fine woody debris (aFWD), and fallen fine woody debris (fFWD). Across three analytical scopes, we quantified (1) stand-scale substrate structuring, (2) within-substrate variation, and (3) directed cross-substrate transitions. Alpha diversity was comparable across all substrates, but total richness did not approach saturation. In contrast, dominant diversity (Hill q 2 ) saturated rapidly. Communities were strongly turnover-driven, indicating species replacement rather than nestedness. Tree identity explained a consistent fraction of compositional variance at the stand scale and remained the strongest predictor within coarse woody debris. In fine woody fractions, tree-identity effects weakened and substrate characteristics gained relative importance. Substrate differentiation was stronger in abundance-weighted than in presence–absence space, indicating that substrates primarily reorganize dominance within partially shared species pools. These results show that tree-level legacy persists across fragmented deadwood fractions, while substrate context modulates competitive outcomes. Given these patterns, long-read eDNA efficiently captures dominant and recurrent components of WIF communities, whereas recovery of total richness is unrealistic under typical sampling effort. Monitoring should therefore prioritize sampling many trees with pooled cores per tree and broad coverage of substrate types and characteristics, including decay stage, size, and depth gradients, rather than intensive within-object stratification.