A Multifactor analysis of fungal and bacterial community structure in the root microbiome of mature Populus deltoides trees.
Migun Shakya, Neil Gottel, Hector Castro, Zamin K. Yang, Lee Gunter, Jessy Labbé, Wellington Muchero, Gregory Bonito, Rytas Vilgalys, Gerald A. Tuskan, Mircea Podar, and Christopher W. Schadt
2013 October 16, PLoS ONE 8(10): e76382; DOI: 10.1371/journal.pone.0076382
Bacterial and fungal communities associated with plant roots are central to the host health, survival and growth. However, a robust understanding of the root-microbiome and the factors that drive host associated microbial community structure have remained elusive, especially in mature perennial plants from natural settings. Here, we investigated relationships of bacterial and fungal communities in the rhizosphere and root endosphere of the riparian tree species Populus deltoides, and the influence of soil parameters, environmental properties (host phenotype and aboveground environmental settings), host plant genotype (Simple Sequence Repeat (SSR) markers), season (Spring vs. Fall) and geographic setting (at scales from regional watersheds to local riparian zones) on microbial community structure. Each of the trees sampled displayed unique aspects to its associated community structure with high numbers of Operational Taxonomic Units (OTUs) specific to an individual trees (bacteria >90%, fungi >60%). Over the diverse conditions surveyed only a small number of OTUs were common to all samples within rhizosphere (35 bacterial and 4 fungal) and endosphere (1 bacterial and 1 fungal) microbiomes. As expected, Proteobacteria and Ascomycota were dominant in root communities (>50%) while other higher-level phylogenetic groups (Chytridiomycota, Acidobacteria) displayed greatly reduced abundance in endosphere compared to the rhizosphere. Variance partitioning partially explained differences in microbiome composition between all sampled roots on the basis of seasonal and soil properties (4% to 23%). While most variation remains unattributed, we observed significant differences in the microbiota between watersheds (Tennessee vs. North Carolina) and seasons (Spring vs. Fall). SSR markers clearly delineated two host populations associated with the samples taken in TN vs. NC, but overall host genotypic distances did not have a significant effect on corresponding communities that could be separated from other measured effects.
The first row of stacked bar represents the overall relative abundance across the entire data set, the second row represents endosphere vs. rhizosphere, third row represents the relative abundance in each watershed and the fourth row represents the relative abundance in May vs. September.
Each bar represents total variance partitioned into pure effect or interaction of two or all factors. Only variance proportions that were statistically significant (p≤0.05) are listed. Variables for host and soil properties were selected based on stepwise selection (forward and backward) to remove non-significant terms from the model. (A) Variance partitioning of bacterial community from endosphere and rhizosphere. (B) Variance partitioning of fungal community from endosphere and rhizosphere
Citation: Shakya M, Gottel N, Castro H, Yang ZK, Gunter L, et al. (2013) A Multifactor Analysis of Fungal and Bacterial Community Structure in the Root Microbiome of Mature Populus deltoides Trees. PLoS ONE 8(10): e76382. doi:10.1371/journal.pone.0076382