COS 84-6 - Employing improved quantitative inference to describe genotypic and environmental variation in the endophytic fungal communities of coastal Douglas-fir needles

Abstract: Employing improved quantitative inference to describe genotypic and environmental variation in the endophytic fungal communities of coastal Douglas-fir needles

Kyle A. Gervers1, Joseph W. Spatafora1,2, Posy E. Busby1,2

1Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR, 97333, USA

2Co-last authors

The ecological patterns and processes associated with communities of endophytic fungi that cryptically infect and inhabit the living tissues of plants remain relatively uncharacterized. Additionally, despite the fact that high-throughput amplicon sequencing (HTAS) of infected host tissues has become the most efficient and thorough method for describing endophytic fungal communities, most published microbial ecology studies to date convert taxon read counts to relative abundances of reads to account for the fact that sequencing depth is randomly or undesirably distributed across samples. However, without sufficient a priori information, changes in the relative abundances of fungal taxa are difficult to interpret, as an increase in the relative abundance of a hypothetical Fungus A can be achieved by an increase in Fungus A’s absolute abundance or a decrease in the absolute abundance of Fungus B, C, and/or D. Recent innovations in organismal and synthetic DNA spike-in approaches to HTAS have emerged to partially resolve this analytical shortcoming. Using the same HTAS process to generate an internal reference sequence in each sample, these approaches calculate the load of each taxon as a ratio of taxon reads to reference reads, facilitating the description of directional fold-changes of taxa across samples. We report on an adaptation of the recently described host-associated microbial PCR method (hamPCR) to quantify fungal load in the Coastal Douglas-fir (Pseudotsuga menziesii var. menziesii; PSME) needle mycobiome. We show that load metrics derived from sequencing data were strongly correlated with concentrations of fungal gDNA serially diluted with microbe-free PSME gDNA (P value < 0.001, leave-one-out cross-validated R2 = 0.762), demonstrating the quantitative utility of the load metric quantified in our system. Additionally, a hierarchical clustering comparison of relative abundance- and load-based compositions obtained from environmentally-acquired PSME needles revealed that load-based compositions could resolve more groups for the same set of samples. Using this method, we are comparing the load-based compositions of PSME needles sampled from trees as part of a reciprocally-designed common garden study testing the roles of genotype and environment in structuring the Douglas-fir needle mycobiome.