Variability and bias in microbiome metagenomic sequencing: an interlaboratory study comparing experimental protocols.

Several studies have documented the significant impact of methodological choices in microbiome analyses. The myriad of methodological options available complicate the replication of results and generally limit the comparability of findings between independent studies that use differing techniques and measurement pipelines. Here we describe the Mosaic Standards Challenge (MSC), an international interlaboratory study designed to assess the impact of methodological variables on the results. The MSC did not prescribe methods but rather asked participating labs to analyze 7 shared reference samples (5 × human stool samples and 2 × mock communities) using their standard laboratory methods. To capture the array of methodological variables, each participating lab completed a metadata reporting sheet that included 100 different questions regarding the details of their protocol. The goal of this study was to survey the methodological landscape for microbiome metagenomic sequencing (MGS) analyses and the impact of methodological decisions on metagenomic sequencing results. A total of 44 labs participated in the MSC by submitting results (16S or WGS) along with accompanying metadata; thirty 16S rRNA gene amplicon datasets and 14 WGS datasets were collected. The inclusion of two types of reference materials (human stool and mock communities) enabled analysis of both MGS measurement variability between different protocols using the biologically-relevant stool samples, and MGS bias with respect to ground truth values using the DNA mixtures. Owing to the compositional nature of MGS measurements, analyses were conducted on the ratio of Firmicutes: Bacteroidetes allowing us to directly apply common statistical methods. The resulting analysis demonstrated that protocol choices have significant effects, including both bias of the MGS measurement associated with a particular methodological choices, as well as effects on measurement robustness as observed through the spread of results between labs making similar methodological choices. In the analysis of the DNA mock communities, MGS measurement bias was observed even when there was general consensus among the participating laboratories. This study was the result of a collaborative effort that included academic, commercial, and government labs. In addition to highlighting the impact of different methodological decisions on MGS result comparability, this work also provides insights for consideration in future microbiome measurement study design.

Manipulating atmospheric CO2 concentration induces shifts in wheat leaf and spike microbiomes and in Fusarium pathogen communities.

Changing atmospheric composition represents a source of uncertainty in our assessment of future disease risks, particularly in the context of mycotoxin producing fungal pathogens which are predicted to be more problematic with climate change. To address this uncertainty, we profiled microbiomes associated with wheat plants grown under ambient vs. elevated atmospheric carbon dioxide concentration [CO2] in a field setting over 2 years. We also compared the dynamics of naturally infecting versus artificially introduced Fusarium spp. We found that the well-known temporal dynamics of plant-associated microbiomes were affected by [CO2]. The abundances of many amplicon sequence variants significantly differed in response to [CO2], often in an interactive manner with date of sample collection or with tissue type. In addition, we found evidence that two strains within Fusarium - an important group of mycotoxin producing fungal pathogens of plants - responded to changes in [CO2]. The two sequence variants mapped to different phylogenetic subgroups within the genus Fusarium, and had differential [CO2] responses. This work informs our understanding of how plant-associated microbiomes and pathogens may respond to changing atmospheric compositions.

Minimal impacts on the wheat microbiome when Trichoderma gamsii T6085 is applied as a biocontrol agent to manage fusarium head blight disease.

Fusarium head blight (FHB) is a major fungal disease that causes severe yield and quality loss in wheat. Biological control can be integrated with other management strategies to control FHB. For this purpose, Trichoderma gamsii strain T6085 is a potential biocontrol agent to limit the infection of F. graminearum and F. culmorum in wheat. However, the possible impacts of T. gamsii T6085 on the broader microbiome associated with the wheat plant are not currently understood. Therefore, we identified bacteria and fungi associated with different wheat tissues, including assessment of their relative abundances and dynamics in response to the application of T6085 and over time, using amplicon sequencing. Residues of the prior year's wheat crop and the current year's wheat spikes were collected at multiple time points, and kernel samples were collected at harvest. DNA was extracted from the collected wheat tissues, and amplicon sequencing was performed to profile microbiomes using 16S v4 rRNA amplicons for bacteria and ITS2 amplicons for fungi. Quantitative PCR was performed to evaluate the absolute abundances of F. graminearum and T. gamsii in different wheat tissues. Disease progression was tracked visually during the growing season, revealing that FHB severity and incidence were significantly reduced when T6085 was applied to wheat spikes at anthesis. However, treatment with T6085 did not lessen the F. graminearum abundance in wheat spikes or kernels. There were substantial changes in F. graminearum abundance over time; in crop residue, pathogen abundance was highest at the initial time point and declined over time, while in wheat spikes, pathogen abundance increased significantly over time. The predominant bacterial taxa in wheat spikes and kernels were Pseudomonas, Enterobacter, and Pantoea, while Alternaria and Fusarium were the dominant fungal groups. Although the microbiome structure changed substantially over time, there were no community-scale rearrangements due to the T6085 treatment. The work suggests several other taxa that could be explored as potential biocontrol agents to integrate with T6085 treatment. However, the timing and the type of T6085 application need to be improved to give more advantages for T6085 to colonize and reduce the F. graminearum inoculum in the field.

Modification of Deoxynivalenol by a Fungal Laccase Paired with Redox Mediator TEMPO.

Mycotoxins such as deoxynivalenol introduce a health risk to the food supply and are costly to manage or avoid. Technologies for reducing or eliminating the toxicity of deoxynivalenol could be useful in a variety of processes, such as in preserving the value as animal feed of byproducts of ethanol production. We characterized transformation products of deoxynivalenol that were formed by the combination of a fungal laccase paired with the chemical mediator 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO), using chromatography, mass spectrometry, and nuclear magnetic resonance spectroscopy. Alcohol groups at the C3 and C15 positions of deoxynivalenol were oxidized to ketones, and the chemical mediator became covalently linked to the C4 position. Conditions experienced during gas chromatography led to the dissociation of TEMPO, forming 3,15-diketodeoxynivalenol. Understanding the range of possible modifications to deoxynivalenol and other trichothecenes is a necessary step toward effective remediation of contaminated grain.

Use of the volatile trichodiene to reduce Fusarium head blight and trichothecene contamination in wheat.

Fusarium graminearum is the primary cause of Fusarium head blight (FHB), one of the most economically important diseases of wheat worldwide. FHB reduces yield and contaminates grain with the trichothecene mycotoxin deoxynivalenol (DON), which poses a risk to plant, human and animal health. The first committed step in trichothecene biosynthesis is formation of trichodiene (TD). The volatile nature of TD suggests that it could be a useful intra or interspecies signalling molecule, but little is known about the potential signalling role of TD during F. graminearum-wheat interactions. Previous work using a transgenic Trichoderma harzianum strain engineered to emit TD (Th + TRI5) indicated that TD can function as a signal that can modulate pathogen virulence and host plant resistance. Herein, we demonstrate that Th + TRI5 has enhanced biocontrol activity against F. graminearum and reduced DON contamination by 66% and 70% in a moderately resistant and a susceptible cultivar, respectively. While Th + TRI5 volatiles significantly influenced the expression of the pathogenesis-related 1 (PR1) gene, the effect was dependent on cultivar. Th + TRI5 volatiles strongly reduced DON production in F. graminearum plate cultures and downregulated the expression of TRI genes. Finally, we confirm that TD fumigation reduced DON accumulation in a detached wheat head assay.

Genetic elucidation of interconnected antibiotic pathways mediating maize innate immunity.

Specialized metabolites constitute key layers of immunity that underlie disease resistance in crops; however, challenges in resolving pathways limit our understanding of the functions and applications of these metabolites. In maize (Zea mays), the inducible accumulation of acidic terpenoids is increasingly considered to be a defence mechanism that contributes to disease resistance. Here, to understand maize antibiotic biosynthesis, we integrated association mapping, pan-genome multi-omic correlations, enzyme structure-function studies and targeted mutagenesis. We define ten genes in three zealexin (Zx) gene clusters that encode four sesquiterpene synthases and six cytochrome P450 proteins that collectively drive the production of diverse antibiotic cocktails. Quadruple mutants in which the ability to produce zealexins (ZXs) is blocked exhibit a broad-spectrum loss of disease resistance. Genetic redundancies ensuring pathway resiliency to single null mutations are combined with enzyme substrate promiscuity, creating a biosynthetic hourglass pathway that uses diverse substrates and in vivo combinatorial chemistry to yield complex antibiotic blends. The elucidated genetic basis of biochemical phenotypes that underlie disease resistance demonstrates a predominant maize defence pathway and informs innovative strategies for transferring chemical immunity between crops.

Enhanced Resistance to Fusarium graminearum in Transgenic Arabidopsis Plants Expressing a Modified Plant Thionin.

The fungal pathogen Fusarium graminearum causes Fusarium head blight (FHB) on wheat, barley, and other grains. FHB results in yield reductions and contaminates grain with trichothecene mycotoxins, which threaten food safety and food security. Innovative mechanisms for controlling FHB are urgently needed. We have previously shown that transgenic tobacco and citrus plants expressing a modified thionin (Mthionin) exhibited enhanced resistance toward several bacterial pathogens. The aim of this study was to investigate whether overexpression of Mthionin could be similarly efficacious against F. graminearum, and whether transgenic expression of Mthionin impacts the plant microbiome. Transgenic Arabidopsis plants expressing Mthionin were generated and confirmed. When challenged with F. graminearum, Mthionin-expressing plants showed less disease and fungal biomass in both leaves and inflorescences compared with control plants. When infiltrated into leaves, macroconidia of F. graminearum germinated at lower rates and produced less hyphal growth in Arabidopsis leaves expressing Mthionin. Moreover, marker genes related to defense signaling pathways were expressed at significantly higher levels after F. graminearum infection in Mthionin transgenic Arabidopsis plants. However, Mthionin expression did not appreciably alter the overall microbiome associated with transgenic plants grown under controlled conditions; across leaves and roots of Mthionin-expressing and control transgenic plants, only a few bacterial and fungal taxa differed, and differences between Mthionin transformants were of similar magnitude compared with control plants. In sum, our data indicate that Mthionin is a promising candidate to produce transgenic crops for reducing FHB severity and ultimately mycotoxin contamination.

Microbial Correlates of Fusarium Load and Deoxynivalenol Content in Individual Wheat Kernels.

Characteristics or constituents of plant-associated microbiomes may assist in constraining disease development. To investigate this possibility for the wheat-Fusarium head blight pathosystem, we assessed seed weight, pathogen load, deoxynivalenol content, and microbiome profiles for individual wheat kernels collected over 2 years from a disease-conducive environment. We found that the microbiomes of individual, hulled wheat kernels consist of dozens to greater than a hundred bacterial taxa and up to several dozen fungal taxa, and that year-to-year variation in microbiome structure was large. Measures of microbial community diversity were negatively correlated with measures of disease severity, and had significant power to explain variation in pathogen load among seeds. Several operational taxonomic units belonging to the genus Sphingomonas demonstrated particularly strong negative relationships with pathogen load. This study illuminates the composition of microbiomes associated with wheat kernels under disease-conducive field conditions, and suggests relationships between microbiome characteristics and Fusarium head blight that warrant further study.

Bacterial endophyte antagonism toward a fungal pathogen in vitro does not predict protection in live plant tissue.

Endophytic microbiota are potentially useful plant symbionts for conferring biotic or abiotic stress tolerance. Common approaches to identify putatively beneficial functions of endophytes rely on lab-based assays. However, if functional roles are context-dependent, lab-based assessments may not accurately represent functional outcomes under variable field conditions. Our objective was to test whether antagonism by bacterial endophytes towards a plant pathogen in vitro would be predictive of disease outcomes in live plant tissue. We challenged Fusarium graminearum, a fungal pathogen of wheat, against bacterial endophytes isolated from wheat plants in two in vitro assays. A subset of isolates, with in vitro antagonistic activity ranging from weak to strong, was selected for testing in live plant tissue (detached wheat heads). Assays were performed under different temperature and/or carbon dioxide conditions to test environmental dependency in the plant-endophyte-pathogen interactions. The two in vitro assays produced contrasting measures of pathogen inhibition, and neither predicted pathogen load reductions in the detached wheat head assay. Additionally, outcomes were environment-dependent and varied among bacterial isolates. Thus, endophytic impacts on plant performance cannot be easily inferred from simplified in vitro assays, and environmental gradients should be incorporated into future testing of microbial interactions in plant hosts.

Bulk soil bacterial community structure and function respond to long-term organic and conventional agricultural management.

Understanding how soil microbiomes respond to management is essential to maximizing soil health. We contrasted microbiomes in bulk soil under long-term organic and conventional management in a grain production setting. Management category significantly impacted the relative abundances of 17% of the most abundant taxa. Both conventional and organic management favored particular taxa, but these effects were not reflected in summary richness and diversity indices. Management systems also lead to differences in soil edaphic properties, including pH and nutrient status; this may have been the mechanism by which change in the prokaryote community was enacted. Community change between years of sampling was less pronounced, with only 6 taxa differentially abundant among years. Management category also impacted the abundance of functional genes related to the production and consumption of greenhouse gases. Particulate methane monooxygenase genes were more frequent in soil under organic management, while soluble methane monooxygenase genes were more frequent in soil under conventional management in 1 of 2 years. Nitrous oxide reductase genes were significantly less abundant in soils under second-year alfalfa than in soils under corn. This work highlights the ability of agricultural management to enact broad rearrangements to the structure of bulk soil bacterial communities.

Managing for Multifunctionality in Perennial Grain Crops.

Plant breeders are increasing yields and improving agronomic traits in several perennial grain crops, the first of which is now being incorporated into commercial food products. Integration strategies and management guidelines are needed to optimize production of these new crops, which differ substantially from both annual grain crops and perennial forages. To offset relatively low grain yields, perennial grain cropping systems should be multifunctional. Growing perennial grains for several years to regenerate soil health before rotating to annual crops and growing perennial grains on sloped land and ecologically sensitive areas to reduce soil erosion and nutrient losses are two strategies that can provide ecosystem services and support multifunctionality. Several perennial cereals can be used to produce both grain and forage, and these dual-purpose crops can be intercropped with legumes for additional benefits. Highly diverse perennial grain polycultures can further enhance ecosystem services, but increased management complexity might limit their adoption.

A fungal mock community control for amplicon sequencing experiments.

Microbial ecology has been profoundly advanced by the ability to profile complex microbial communities by sequencing of marker genes amplified from environmental samples. However, inclusion of appropriate controls is vital to revealing the limitations and biases of this technique. "Mock community" samples, in which the composition and relative abundances of community members are known, are particularly valuable for guiding library preparation and data processing decisions. I generated a set of three mock communities using 19 different fungal taxa and demonstrate their utility by contrasting amplicon sequencing data obtained for the same communities under modifications to PCR conditions during library preparation. Increasing the number of PCR cycles elevated rates of chimera formation, and of errors in the final data set. Extension time during PCR had little impact on chimera formation, error rate or observed community structure. Polymerase fidelity impacted error rates significantly. Despite a high error rate, a master mix optimized to minimize amplification bias yielded profiles that were most similar to the true community structure. Bias against particular taxa differed among ITS1 vs. ITS2 loci. Preclustering nearly identical reads substantially reduced error rates, but did not improve similarity to the expected community structure. Inaccuracies in amplicon sequence-based estimates of fungal community structure were associated with amplification bias and size selection processes, as well as variable culling rates among reads from different taxa. In some cases, the numerically dominant taxon was completely absent from final data sets, highlighting the need for further methodological improvements to avoid biased observations of community profiles.

The Potential for Cereal Rye Cover Crops to Host Corn Seedling Pathogens.

Cover cropping is a prevalent conservation practice that offers substantial benefits to soil and water quality. However, winter cereal cover crops preceding corn may diminish beneficial rotation effects because two grass species are grown in succession. Here, we show that rye cover crops host pathogens capable of causing corn seedling disease. We isolated Fusarium graminearum, F. oxysporum, Pythium sylvaticum, and P. torulosum from roots of rye and demonstrate their pathogenicity on corn seedlings. Over 2 years, we quantified the densities of these organisms in rye roots from several field experiments and at various intervals of time after rye cover crops were terminated. Pathogen load in rye roots differed among fields and among years for particular fields. Each of the four pathogen species increased in density over time on roots of herbicide-terminated rye in at least one field site, suggesting the broad potential for rye cover crops to elevate corn seedling pathogen densities. The radicles of corn seedlings planted following a rye cover crop had higher pathogen densities compared with seedlings following a winter fallow. Management practices that limit seedling disease may be required to allow corn yields to respond positively to improvements in soil quality brought about by cover cropping.

Plant community richness and microbial interactions structure bacterial communities in soil.

Plant species, plant community diversity and microbial interactions can significantly impact soil microbial communities, yet there are few data on the interactive effects of plant species and plant community diversity on soil bacterial communities. We hypothesized that plant species and plant community diversity affect soil bacterial communities by setting the context in which bacterial interactions occur. Specifically, we examined soil bacterial community composition and diversity in relation to plant "host" species, plant community richness, bacterial antagonists, and soil edaphic characteristics. Soil bacterial communities associated with four different prairie plant species (Andropogon gerardii, Schizachyrium scoparium, Lespedeza capitata, and' Lupinus perennis) grown in plant communities of increasing species richness (1, 4, 8, and 16 species) were sequenced. Additionally, soils were evaluated for populations of antagonistic bacteria and edaphic characteristics. Plant species effects on soil bacterial community composition were small and depended on plant community richness. In contrast, increasing plant community richness significantly altered soil bacterial community composition and was negatively correlated with bacterial diversity. Concentrations of soil carbon, organic matter, nitrogen, phosphorus, and potassium were similarly negatively correlated with bacterial diversity, whereas the proportion of antagonistic bacteria was positively correlated with soil bacterial diversity. Results suggest that plant species influences on soil bacterial communities depend on plant community diversity and are mediated through the effects of plant-derived resources on antagonistic soil microbes.

Estimating beta diversity for under-sampled communities using the variably weighted Odum dissimilarity index and OTUshuff.

MOTIVATION: In profiling the composition and structure of complex microbial communities via high throughput amplicon sequencing, a very low proportion of community members are typically sampled. As a result of this incomplete sampling, estimates of dissimilarity between communities are often inflated, an issue we term pseudo ß-diversity. RESULTS: We present a set of tools to identify and correct for the presence of pseudo ß-diversity in contrasts between microbial communities. The variably weighted Odum dissimilarity (DwOdum) allows for down-weighting the influence of either abundant or rare taxa in calculating a measure of similarity between two communities. We show that down-weighting the influence of rare taxa can be used to minimize pseudo ß-diversity arising from incomplete sampling. Down-weighting the influence of abundant taxa can increase the sensitivity of hypothesis testing. OTUshuff is an associated test for identifying the presence of pseudo ß-diversity in pairwise community contrasts. AVAILABILITY AND IMPLEMENTATION: A Perl script for calculating the DwOdum score from a taxon abundance table and performing pairwise contrasts with OTUshuff can be obtained at http://www.ars.usda.gov/services/software/software.htm?modecode=30-12-10-00. CONTACT: daniel.manter@ars.usda.gov SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Diffuse symbioses: roles of plant-plant, plant-microbe and microbe-microbe interactions in structuring the soil microbiome.

A conceptual model emphasizing direct host-microbe interactions has dominated work on host-associated microbiomes. To understand plant-microbiome associations, however, broader influences on microbiome composition and functioning must be incorporated, such as those arising from plant-plant and microbe-microbe interactions. We sampled soil microbiomes associated with target plant species (Andropogon gerardii, Schizachyrium scoparium, Lespedeza capitata, Lupinus perennis) grown in communities varying in plant richness (1-, 4-, 8- or 16-species). We assessed Streptomyces antagonistic activity and analysed bacterial and Streptomyces populations via 454 pyrosequencing. Host plant species and plant richness treatments altered networks of coassociation among bacterial taxa, suggesting the potential for host plant effects on the soil microbiome to include changes in microbial interaction dynamics and, consequently, co-evolution. Taxa that were coassociated in the rhizosphere of a given host plant species often showed consistent correlations between operational taxonomic unit (OTU) relative abundance and Streptomyces antagonistic activity, in the rhizosphere of that host. However, in the rhizosphere of a different host plant species, the same OTUs showed no consistency, or a different pattern of responsiveness to such biotic habitat characteristics. The diversity and richness of bacterial and Streptomyces communities exhibited distinct relationships with biotic and abiotic soil characteristics. The rhizosphere soil microbiome is influenced by a complex and nested array of factors at varying spatial scales, including plant community, plant host, soil edaphics and microbial taxon and community characteristics.

Subinhibitory antibiotic concentrations mediate nutrient use and competition among soil streptomyces.

Though traditionally perceived as weapons, antibiotics are also hypothesized to act as microbial signals in natural habitats. However, while subinhibitory concentrations of antibiotics (SICA) are known to shift bacterial gene expression, specific hypotheses as to how SICA influence the ecology of natural populations are scarce. We explored whether antibiotic 'signals', or SICA, have the potential to alter nutrient utilization, niche overlap, and competitive species interactions among Streptomyces populations in soil. For nine diverse Streptomyces isolates, we evaluated nutrient utilization patterns on 95 different nutrient sources in the presence and absence of subinhibitory concentrations of five antibiotics. There were significant changes in nutrient use among Streptomyces isolates, including both increases and decreases in the capacity to use individual nutrients in the presence vs. in the absence of SICA. Isolates varied in their responses to SICA and antibiotics varied in their effects on isolates. Furthermore, for some isolate-isolate-antibiotic combinations, competition-free growth (growth for an isolate on all nutrients that were not utilized by a competing isolate), was increased in the presence of SICA, reducing the potential fitness cost of nutrient competition among those competitors. This suggests that antibiotics may provide a mechanism for bacteria to actively minimize niche overlap among competitors in soil. Thus, in contrast to antagonistic coevolutionary dynamics, antibiotics as signals may mediate coevolutionary displacement among coexisting Streptomyces, thereby hindering the emergence of antibiotic resistant phenotypes. These results contribute to our broad understanding of the ecology and evolutionary biology of antibiotics and microbial signals in nature.

Variations in diversity and richness of gut bacterial communities of termites (Reticulitermes flavipes) fed with grassy and woody plant substrates.

Diets shape the animal gut microbiota, although the relationships between diets and the structure of the gut microbial community are not yet well understood. The gut bacterial communities of Reticulitermes flavipes termites fed on four individual plant biomasses with different degrees of recalcitrance to biodegradation were investigated by 16S rRNA pyrosequencing analysis. The termite gut bacterial communities could be differentiated between grassy and woody diets, and among grassy diets (corn stover vs. sorghum). The majority of bacterial taxa were shared across all diets, but each diet significantly enriched some taxa. Interestingly, the diet of corn stover reduced gut bacterial richness and diversity compared to other diets, and this may be related to the lower recalcitrance of this biomass to degradation.

Root exudation of phytochemicals in Arabidopsis follows specific patterns that are developmentally programmed and correlate with soil microbial functions.

Plant roots constantly secrete compounds into the soil to interact with neighboring organisms presumably to gain certain functional advantages at different stages of development. Accordingly, it has been hypothesized that the phytochemical composition present in the root exudates changes over the course of the lifespan of a plant. Here, root exudates of in vitro grown Arabidopsis plants were collected at different developmental stages and analyzed using GC-MS. Principle component analysis revealed that the composition of root exudates varied at each developmental stage. Cumulative secretion levels of sugars and sugar alcohols were higher in early time points and decreased through development. In contrast, the cumulative secretion levels of amino acids and phenolics increased over time. The expression in roots of genes involved in biosynthesis and transportation of compounds represented in the root exudates were consistent with patterns of root exudation. Correlation analyses were performed of the in vitro root exudation patterns with the functional capacity of the rhizosphere microbiome to metabolize these compounds at different developmental stages of Arabidopsis grown in natural soils. Pyrosequencing of rhizosphere mRNA revealed strong correlations (p<0.05) between microbial functional genes involved in the metabolism of carbohydrates, amino acids and secondary metabolites with the corresponding compounds released by the roots at particular stages of plant development. In summary, our results suggest that the root exudation process of phytochemicals follows a developmental pattern that is genetically programmed.

Potential impact of soil microbiomes on the leaf metabolome and on herbivore feeding behavior.

It is known that environmental factors can affect the biosynthesis of leaf metabolites. Similarly, specific pairwise plant-microbe interactions modulate the plant's metabolome by stimulating production of phytoalexins and other defense-related compounds. However, there is no information about how different soil microbiomes could affect the plant growth and the leaf metabolome. We analyzed experimentally how diverse soil microbiomes applied to the roots of Arabidopsis thaliana were able to modulate plant growth and the leaf metabolome, as assessed by GC-MS analyses. Further, we determined the effects of soil microbiome-driven changes in leaf metabolomics on the feeding behavior of Trichopulsia ni larvae. Soil microbiomes differentially impacted plant growth patterns as well as leaf metabolome composition. Similarly, most microbiome-treated plants showed inhibition to larvae feeding, compared with unamended control plants. Pyrosequencing analysis was conducted to determine the soil microbial composition and diversity of the soils used in this study. Correlation analyses were performed to determine relationships between various factors (soil microbial taxa, leaf chemical components, plant growth patterns and insect feeding behavior) and revealed that leaf amino acid content was positively correlated with both microbiome composition and insect feeding behavior.