Figure 1.
Rarefaction, richness, and diversity analyses of 16s amplicon data.
Approximately 166 bacterial OTUs were detected through amplicon sequencing. Various community richness estimators consistently predicted the presence of over 300 OTUs in association with the A. glabripennis gut and, in agreement with this observation, the rarefaction curve failed to reach saturation. This indicates that additional OTUs would likely be detected with additional amplicon sequencing.
Figure 2.
Maximum likelihood analysis of representative sequences from operational taxonomic unit analysis (OTU) of bacterial 16S rRNA amplicons.
Representative sequences from each bacterial OTU were aligned with MEGA 4.0 and phylogenetic analysis using was performed using GARLI 2.0 (500 bootstrap pseudoreplicates and TIM1+I+G evolutionary model). Nodes were collapsed and labeled by taxonomic class. Number of OTUs and percentage of amplicons assigned to each class are labeled. OTUs that could not be assigned to class level by RDP were omitted from the analysis.
Figure 3.
Rarefaction, richness, and diversity analyses of 18S amplicon data.
Seven fungal OTUs were detected through amplicon sequencing. While rarefaction begins to approach saturation, richness estimates predict the presence of at least 11 fungal OTUs indicating that additional sampling may be necessary. This scenario is likely since additional 18S rRNAs from fungal taxa not detected in the 18S amplicons were detected in the shotgun reads (e.g., Fusarium spp.).
Figure 4.
Distribution of SEED assignments generated by MG-RAST.
Reads assigned to 28 SEED subsystems were detected in the A. glabripennis larval midgut metagenome. The most dominant subsystems found in association with this microbial community included clustering based subsystems, carbohydrate metabolism, and amino acid and derivatives metabolism.
Figure 5.
Hierarchical cluster analysis based on Pfam annotations of herbivore related metagenomes.
Agglomerative hierarchical cluster analysis based on a compositional Euclidean distance matrix was conducted using Pfam annotations from various herbivore related metagenomes. Three distinct clusters representing different herbivore biome-types are highlighted and labeled. These include herbivore gut communities, fungal gallery communities associated with phoem/xylem feeding insects and communities associated with insects feeding in heartwood.
Figure 6.
Principal components analysis (PCA) of Pfam domains from herbivore-related metagenomes.
Principal components analysis was conducted to plot samples in multidimensional space. Groupings detected in agglomerative cluster analysis are preserved (Mantel test, p< 0.0001) and are color-coded by groups identified in the dendrogram. Monte Carlo Permutation Procedure (n=1000 iterations): p<0.0001 for PCA 1 and PCA2. DP: Dendroctonous ponderosae DF: Dendroctonous frontalis CR: Costa Rica, FL: Florida.
Figure 7.
Distribution of glycoside hydrolase families found in the A. glabripennis gut metagenome.
Reads assigned to 36 glycoside hydrolase families were detected in the gut microbiome. The most dominant families were GH 1 and 3, while GH families 11, 45, 46, 61, and 71 were present in very low abundances.
Figure 8.
Xylose utilization pathway present in the A. glabripennis gut community.
Xylose released from hemicellulose can be converted into D-xylulose-5-phosphate and eventually into acetaldehyde. Acetaldehyde can be either converted into ethanol by alcohol dehydrogenase or into acetate by acetaldehyde dehydrogenase. These reactions are likely catalyzed by lactic acid bacteria or yeasts associated with the A. glabripennis gut.