Ecological inference and contaminant detection from fungal microbiome data with q2-fungal-traits
Fungi are key members of microbial communities, yet microbiome surveys often lack trait-based information required for ecologically meaningful interpretation of mycobiome data. To demonstrate the value of fungal trait-based phenotyping in microbiome research, we re-analyzed N=3,221 samples across four case studies spanning human, agricultural, and environmental systems. In human cancer and vineyard datasets, trait-based analysis detected fungi producing macroscopic fruiting bodies, likely introduced via airborne spore dispersal, indicating widespread contributions of transient or contaminant fungi that can confound interpretation of sequencing data from tumor biopsies and grape berries. In sourdough fermentations, filamentous fungi were highly abundant alongside traditionally-recognized yeast and occupied distinct ecological niches. In forest soils, increasing habitat disturbance was associated with increased prevalence and abundance of plant pathogens, and a marked decline in ectomycorrhizal and lichenized fungi. These changes were accompanied by a shift toward large-spored taxa in urban soils, consistent with enhanced stress tolerance. To facilitate broader adoption of fungal phenotyping in microbiome studies, we introduce q2-fungal-traits, a QIIME2 plugin for automated integration of fungal taxonomy derived from marker-gene or shotgun metagenome sequencing surveys with ecological and functional trait data. The plugin assigns lifestyle-related traits and spore size estimates through hierarchical taxonomic matching and integrates directly into standard microbiome workflows. Our case studies demonstrate that integrating trait-based ecology with mycobiota datasets can generate novel findings and testable hypotheses, enabling inference of the functional (ir)relevance of community constituents. Our work contributes to bridging the gap between descriptive community profiling and functional ecology in microbiome research.