High-Throughput DNA Metabarcoding of Stem Sections from Trees with Cavities Describes Fungal Communities Associated with Variable Wood Decay, Position on Stem and Tree Species

The presence of unculturable fungi, variability in timing and frequency of fungal fruiting, hyper-rich fungal communities, and genetic and environmental variability explains the difficulty in adopting ideal sampling schemes and fungal identification approaches in studies of fungal communities in wood at variable stages of decay. Here, we use intensive within-tree sampling of five standing trees with cavities paired with high-throughput DNA metabarcoding, to study fungal communities in decayed and healthy wood of trees from two Populus species in British Columbia, Canada. The amplification of over 3000 fungal DNA sequence variants shows the presence of a hyper-rich wood fungal community that not only varied depending on PCR primers, tree species, tree stem portion and wood decay stage. but also included a large number of taxa unassignable to any known sub-kingdom taxonomic order based on published DNA sequences. Our data show that the use of two different primer sets greatly increases the power of the metabarcoding analysis. By testing three alternative models of fungal community composition, we identify the model that best explains fungal community by considering the position on the stem and distance from the cavity. We suggest this model may be used to design optimal sampling schemes to describe fungal communities in trees experiencing discrete decay pockets or cavities.