CROP MICROBIOMES AND THE SEARCH FOR EFFECTIVE BIOCONTROL OF FUSARIUM GRAMINEARUM ON WHEAT Kristi Gdanetz MacCready By A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Plant Biology – Doctor of Philosophy 2018 ABSTRACT CROP MICROBIOMES AND THE SEARCH FOR EFFECTIVE BIOCONTROL OF FUSARIUM GRAMINEARUM ON WHEAT By Kristi Gdanetz MacCready Manipulation of naturally occurring microbial communities to reduce plant diseases or increase crop yields requires a thorough understanding of interactions within the phytobiome, in particular, how microbial communities change as plants age, across plant species and organs, and under different land management regimes. Plants were sampled from a wheat-maize-soybean crop rotation site that implements four different land management strategies (conventional, no-till, reduced inputs, and organic). The fungal and bacterial communities of leaves, stems, and roots of wheat, maize, and soybean throughout the growing season were analyzed using fungal internal transcribed spacer and bacterial 16S rRNA gene amplicon sequencing. Analysis of sequence-based fungal communities has some limitations due to the unreliable phylogenetic resolution of DNA sequence alignments. To improve this deficiency, a tool that improved phylogenetic resolution was developed. This tool increases the number of operational taxonomic units which are identified at genus and species levels. Endophytes were isolated from the wheat plants used for microbial community analysis and tested for antagonistic activity toward the wheat pathogen Fusarium graminearum during wheat seedling and head infection. Endophytes on crops can be developed to manage disease, and endophyte-based biocontrols could solve current limitations in F. graminearum disease control. Additionally, functional analysis of F. graminearum secondary metabolite genes provides insight into the function of their gene products for this fungal pathogen. Microbial community structure is affected by various genetic factors of the host plant, environmental factors, and interactions with other organisms. Understanding community responses to these factors is necessary for targeted manipulation of communities to reduce plant disease. Copyright by KRISTI GDANETZ MACCREADY 2018 ACKNOWLEDGMENTS The author would like to thank committee members, Frances Trail, Gregory Bonito, C. Robin Buell, Terence Marsh; past and present members of the Trail lab including Beth Brisco- McCann, Brad Cavinder, Caitlyn Cubba, Cristina de Miguel Rojas, Kayla Fellows-Stefanko, Lori Imboden, Wonyong Kim, Claudia Petrucco, Ludmila Roze, Rebecca Shay, Usha Sikhakolli, Ben Smith, Ashlyn Sovereen, Tara Watkins, Wanessa Wight, and Chris Wright; and funding sources Michigan Wheat Program, MSU Plant Science Fellowship, MSU CNS Dissertation Continuation Fellowship, MSU-KBS Graduate Student Small Grant, A.L. Rodger’s Scholarship, and Everett Beneke Endowment Award. v TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES KEY TO ABBREVIATIONS viii ix xii 1 1 1 1 8 9 11 19 19 21 Introduction Emergence of Phytobiome Field Shifting Attitudes in Disease Management Endophytes Offer Plant Protection Previous Crop Microbiome Studies Microbes in Agriculture Fusarium graminearum Head Blight of Wheat Conclusions CHAPTER 1 LITERATURE REVIEW CHAPTER 2 THE WHEAT MICROBIOME UNDER FOUR MANAGEMENT STRATEGIES, AND POTENTIAL FOR ENDOPHYTES IN DISEASE PROTECTION CHAPTER 3 24 HUB TAXA AND CORE MICROBIAL COMMUNITIES OF ROTATION CROPS UNDER COMMON LAND MANAGEMENT STRATEGIES 24 24 25 27 34 69 Abstract Introduction Materials and Methods Results Discussion Abstract Source 23 23 23 23 CHAPTER 4 FUNGAL MICROBIOMES: A STRATEGY FOR IMPROVED TAXONOMIC RESOLUTION OF ENVIRONMENTAL ITS SEQUENCES APPENDICES APPENDIX A EVALUATION OF PROTECTIVE ENDOPHYTES AGAINST FUSARIUM GRAMINEARUM HEAD BLIGHT Abstract Source 75 75 76 77 78 75 78 vi APPENDIX B FUNCTIONAL ANALYSIS OF KNOCKOUTS OF TERPENE SYNTHASE GENES IN FUSARIUM GRAMINEARUM 87 87 REFERENCES 100 vii LIST OF TABLES Table 1-1. Highlights of plant-microbiome studies of wheat, maize, or soybean conducted to date. 14 Table 3-1. Collection dates of plant growth stages. 31 Table 3-2. Primer sequences of microbiome target loci. Table 3-3. Sequence processing statistics and OTU assignment statistics. Table 3-4. Samples without high-quality reads after sequence processing. 31 37 38 Table 3-5. Mean estimated Shannon diversity ± standard deviation. Superscripts indicate significant differences within a growth stage and management style, after Tukey's HSD. 52 Table 3-6. PERMANOVA of the communities presented in Figures 3-15, 3-16, and 3-17, considering all factors and their interactions. Table 3-7. Hub taxa of KBS LTER crops. 54 63 Table A-1. Protective endophyte strains used for head inoculations (Gdanetz and Trail 2017). 82 90 93 94 Table A-2. TPS gene IDs with greatest log-fold change in expression. Table A-3. Primers used to generate terpene synthase gene knockouts. Table A-4. Summary of Fusarium spp. sequence alignments. viii LIST OF FIGURES 40 36 35 33 32 Figure 3-1. Layout of management strategy replicates at Michigan State University KBS LTER field site. Colored plots are planted in a three-year wheat-maize-soybean rotation cycle. Modified from http://lter.kbs.msu.edu/research/annual-plot-maps/ Figure 3-2. Overview of plants and growth stages. For plant collection dates and formal growth stage names see Table 3-1. Figure 3-3. Shared and unique bacterial or fungal operational taxonomic units among all crops. Figure 3-4. Shared and unique operational taxonomic units across plant organs Figure 3-5. Class-level relative abundance of operational taxonomic units in bacterial communities across crop, organ, and growth stage. Figure 3-6. Class-level relative abundance of operational taxonomic units in bacterial communities across crop, organ, and management strategies. Figure 3-7. Class-level relative abundance of operational taxonomic units in bacterial communities across maize growth stage, organ, and management strategies. Figure 3-8. Class-level relative abundance of operational taxonomic units in bacterial communities across soybean growth stage, organ, and management strategies. Figure 3-9. Class-level relative abundance of operational taxonomic units in fungal communities across crop, organ, and growth stage. Figure 3-10. Class-level relative abundance of operational taxonomic units in fungal communities across crop, organ, and management strategies. Figure 3-11. Class-level relative abundance of operational taxonomic units in fungal communities across maize growth stage, organ, and management strategies. Figure 3-12. Class-level relative abundance of operational taxonomic units in fungal communities across soybean growth stage, organ, and management strategies. Figure 3-13. Global alpha diversity patterns calculated by Shannon Diversity Index (H’). All treatments Global alpha diversity patterns calculated by Shannon Diversity Index (H’). All treatments and replicates were pooled within growth stages for (A) fungi and (B) bacteria. Data are represented by 24 replicates from each crop-growth stage-organ combination. Center line of boxes represents median of samples. The upper and lower sides of the boxes represent the third and first quartiles, respectively. Whiskers represent ± 1.5 times the interquartile range. Data 47 48 41 42 43 45 46 ix 51 50 points beyond whiskers represent outliers. Analysis of variance and Tukey’s honest significant difference were used to test significance (P < 0.05). Statistical support is detailed in Table 3-5. Figure 3-14. Alpha diversity maize fungi (A), soybean fungi (B), maize bacteria (C), soybean bacteria (D) found on different plant organs across growth stages, and under four different management strategies, estimated by Shannon Diversity Index. T1=conventional, T2=no till, T3=low input, T4=organic. Reproductive growth stage indicates early ear development, and early pod fill for maize and soybean, respectively. Data are represented by six replicates from each stage-management-organ combination. Center line of boxes represents median of samples. The upper and lower sides of the boxes represent the third and first quartiles, respectively. Whiskers represent ± 1.5 times the interquartile range. Data points beyond whiskers represent outliers. Analysis of variance and Tukey’s honest significant difference were used to test significance (P < 0.05). Statistical support is detailed in Table 3-5. Figure 3-15. Effect of growth stage on beta diversity of fungal (A) and bacterial (B) communities on all crops. Points colored by crop species, shapes indicate plant organ. Non- metric multidimensional scaling (NMDS) calculated by Bray-Curtis distance. Statistical support is detailed in Table 3-6. Figure 3-16. Influence of management strategies on beta diversity of maize fungal (A), maize bacterial (B), soybean fungal (C), soybean bacterial (D) communities originating from each plant organ. NMDS calculated by Bray-Curtis distance. Statistical support is detailed in Table 3-6. 59 Figure 3-17. Effect of growth stage on beta diversity of maize fungal (A), maize bacterial (B), soybean fungal (C), and soybean bacterial (D) communities originating from each management style. NMDS calculated by Bray-Curtis distance. Statistical support is detailed in Table 3-6. 60 Figure 3-18. Co-occurrence network of taxa in wheat roots. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. Figure 3-19. Co-occurrence network of taxa in wheat phyllosphere. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. Figure 3-20. Co-occurrence network of taxa in maize (A) and soybean (B) roots. Bacterial operational taxonomic units (OTUs) are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. 66 64 65 58 x 67 83 68 Figure 3-21. Co-occurrence network of taxa in maize phyllosphere. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. Figure 3-22. Co-occurrence network of taxa in soybean phyllosphere. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. Figure A-1. Inoculation method and experimental design of protective endophytes. Figure A-2. Fusarium Head Blight disease incidence. Wheat inoculated with endophytes (No E.: endophyte-free control) and challenged with Fusarium graminearum. Error bars represent standard error of the mean (3 independent replicates of 8 plants). Strains marked with asterisk had significantly reduced disease compared to endophyte free control (p < 0.05). 84 Figure A-3. Relative gene expression of F. graminearum terpene synthase genes. Expression levels during perithecia development on straw (A), and during plant infection (B), DW and BW indicate barley cultivars. Expression data from Hallen et al. 2007, Hallen and Trail 2008, Guenther et al. 2009, Hallen-Adams et al. 2011b. Figure A-4. Disease incidence of wild-type (PH-1) and independent knockout strains of FGSC_12186 during wheat head infection (A), distribution of perithecia production on wheat straw (B), and in vitro perithecia production (C). Bars in panel A represents the average of ten or more inoculated plants, repeated three independent times. Bars in panel B represent three sets of inoculate stems, repeated three independent times. Bars in panel C represent the number of perithecia on two cores per plate, six independent plates. Errors bars indicate standard deviation of the mean. Analysis of variance and Tukey’s honest significant difference were used to test significance; asterisk indicates significance (p < 0.001). 97 Figure A-5. Disease incidence of wild-type (PH1) and knockout strains during seedling infection. Subset of seedlings were also inoculated with best protective endophytes. Each bar represents ten plants repeated two independent times. 2B = ∆12186.2B; 3B = ∆12186.3B; 37, 38, 40, 70 are fungal endophyte strains from Gdanetz and Trail, 2017. Figure A-6. Conidia growth curves of wild-type (PH-1) and knockout strains. Each point represents the average of three independent replicates, ± standard error of the mean. Analysis of variance and Tukey’s honest significant difference were used to test significance. PH-1 displayed significantly higher rate of conidia production than both knockout strains at 72 hours, and higher conidia production than ∆12186.2B at 96 hours (p < 0.05). Strain ∆12186.3B displayed significantly higher conidia production than PH-1 and ∆12186.2B at 120 hours (p < 0.05). Figure A-7. In vitro perithecium development. Squash mounts of F. graminearum perithecia at 120 hours post-induction of sexual development, showing mature asci of wild-type (A) and delayed development of knockout strain 12186.3B (B). 99 98 99 96 xi ANOVA ARISA BLAST CONSTAX DGGE FGDB ITS KBS LTER MIPS NCBI NMDS OTU KEY TO ABBREVIATIONS Analysis of variance Automated ribosome intergenic spacer analysis Basic Local Alignment Search Tool Consensus Taxonomy Tool Denaturing gradient gel electrophoresis Fusarium graminearum Genome Database Internal transcribed spacer Kellogg Biological Station Long-term ecological research Munich Information Center for Protein Sequences National Center for Biotechnology Information Non-metric multi-dimensional scaling Operational taxonomic unit PERMANOVA Permutational analysis of variance TPS T-RFLP Terpene synthase Terminal restriction fragment polymorphism Tukey’s HSD Tukey’s honest significant difference test xii CHAPTER 1 LITERATURE REVIEW Introduction Growing interest in the manipulation of naturally occurring microbial communities to reduce plant diseases or increase crop yields has caused the plant microbiome and phytobiome fields to blossom in recent years. Before manipulation of plant or soil microbial communities can be successfully implemented, we must have a thorough understanding of the composition of a healthy microbial community, in particular, how microbial communities change as plants age, how communities differ across plant species, genotype and organ, under different land management regimes, and across various climates. This review will focus on the origins of the plant microbiome and phytobiome fields, how phytobiome research will be useful for changing preferences in disease management, how an increasing understanding of distributions of plant microbiomes and endophytes on crops will be useful to manage disease, and how endophyte based biocontrols could solve current weaknesses in Fusarium graminearum Schwabe (teleomorph Gibberella zeae [Schwein.] Petch) disease control. Emergence of Phytobiome Field Historically, before the onset of low-cost high-throughput DNA sequencing, microbes and plants were studied almost exclusively in paired isolations (for example, a pathogen plus the plant host). Studies of these limited plant-pathogen systems are also restricted by what we can 1 reasonably culture and grow in a laboratory setting. In the 1990s and early 2000s, techniques such as automated ribosome intergenic spacer analysis (ARISA), terminal restriction fragment polymorphism (T-RFLP), or denaturing gradient gel electrophoresis (DGGE) were used in microbial ecology to glean information about community composition, structure, and function. ARISA was used to distinguish between microbial communities, with PCR-amplified intergenic spacers that were separated on polyacrylamide gels. Analysis of these gels assumed each member of the community had a different length intergenic spacer (Fisher and Triplett 1999). T-RFLP is a technique based on DNA extraction and fluorescently-labelled PCR amplification followed by an enzyme digestion that can provide information about community membership (Liesack et al. 2004). When applied to complex microbial communities, researchers gained insight into diversity, but T-RFLP often oversimplified the picture of a microbial community, thus was often complemented with analysis of clone libraries; which is very is laborious. DGGE also required DNA extraction and PCR amplification of 16S or 18S rRNA genes. PCR products were separated on a denaturing gel based on DNA fragment composition instead of size (Strathdee and Free 2013). All of these techniques are laborious and time intensive. They also can provide oversimplified pictures of the communities analyzed because they largely rely on distinguishing between sized- based band patterns. The invention of “next-generation” sequencing (such as 454-Roche pyrosequencing or Illumina sequencing, now more appropriately termed sequence-by-synthesis), quickly replaced previous microbial community analysis methods such as ARISA, T-RFLP, or DGGE because DNA marker gene sequencing generated larger volumes of data generated for equal or reduced sample preparation costs. The first paper using pyrosequencing of the 16S rRNA gene to analyze the composition of a microbial community was published in 2002 (Jonasson et al. 2002). This 2 technique was popularized by the publication of tools such as Dotur, Mothur, and the RDP Classifier (Schloss and Handelsman 2005, Schloss et al. 2009, Cole et al. 2013). Entire research initiatives have arisen from these new technologies – Human Microbiome Project (Turnbaugh et al. 2007), Microbiome of the Built Environment (Lax et al. 2014), Earth Microbiome (Gilbert et al. 2014), and most recently the Phytobiome Initiative (APS 2016). Reduced sequencing costs and consistently increasing dataset sizes generated in a single sequencing run, coupled with the increased ease-of-access to high-performance computer clusters, has enabled a culture-independent era of microbial ecology research. These technologies have given scientists the ability to identify thousands of “species” or operational taxonomic units (OTUs) from a small sample, such as one gram of soil. Sequencing of barcode or marker regions, such as the fungal internal transcribed spacer (ITS) region and bacterial 16S rRNA gene coupled with high-throughput sequencing technologies allows researchers to obtain an in-depth picture of the community structure. The numbers of studies using these techniques have exploded in recent years with 14,570 results for a “microbiome” topic search on Web of Science, with the oldest publication dated in 2002. The term microbiome was first defined as a “characteristic microbial community” (Whipps et al. 1988). In the early days of genomics and metagenomics, the use of “microbiome” was co- opted to refer to the microbial meta-genome (Lederberg 2004, Huss 2014). Metagenomics – the sequencing of total DNA from a sample, instead of one gene or a single species’ genome, allows bioinformaticians to reconstruct many genomes simultaneously. Using this technique, we can interpret or hypothesize functions and lifestyles of organisms that we have not yet isolated in the lab. This information can be used to culture some previously difficult-to-culture or unculturable species and use these organisms for experimental manipulation in the laboratory (Browne et al. 3 2016). There has been recent formal acknowledgement of the original definition of the term microbiome as usage of this term is becoming more common (Shade et al. 2012, Marchesi et al. 2015, Leach et al. 2017, Prescott et al. 2017, Schlatter et al. 2017). As with the terms microbiome and metagenome, there has been some confusion regarding “phytobiome.” Phytobiome does not equate to “plant microbiome,” it actually has a more holistic definition; incorporating environmental factors, pathogens, herbivores, plant genetics, and the interactions among all aforementioned items on plant health. Due to methodological development from the previous microbiome research initiatives, plant microbiome researchers are at an advantage and can skip some of the methodological development that comes with an emerging field. But plant microbiome researchers cannot skip the descriptive community steps. There has been considerable work conducted toward describing the microbial communities of rhizosphere and root-associated taxa of major crops and model plants. There has been less focus on identifying the communities associated with phyllosphere (aboveground plant-associated) communities. Work on model plant systems has shown that the microbes inhabiting the aboveground organs of plants influence plant health and nutrient cycling (Delmotte et al. 2009, Ryffel et al. 2016, Vogel et al. 2016) From the limited number of studies that have investigated microbes across multiple organs of the same plant – it appears that there are organ-specific plant microbiomes (reviewed by Vandenkoornhuyse et al. 2015). From the diversity captured by previous studies, a core plant microbiome (microbes shared across all plants) does not appear to be present, but a meta-analysis of all the plant microbiome studies conducted to date may reveal otherwise. Although there is not yet a clearly established metadata guidelines for plant microbiome work – it would be advantageous for the plant microbiome field to establish them. Such guidelines have been 4 established for the Human (Turnbaugh et al. 2007), and Earth (Gilbert et al. 2014) Microbiome Projects. Plant microbiome researchers should strongly consider incorporating Earth Microbiome Project metadata guidelines into their studies, including measurements such as altitude, biome, host taxonomy, and growth conditions which will aid in future meta-analyses of sequence data (Earth Microbiome guidelines are available at http://www.earthmicrobiome.org/protocols-and- standards/metadata-guide/). It has also been demonstrated that transport and storage methods affect quality of samples (U’Ren et al. 2014). Immediate freezing of samples best preserves plant and microbial DNA for extraction and future re-sampling. Incubation and transport in extraction buffer from field site to lab is the next best ideal method, but one method commonly used (drying and room temperature storage) leads to a significant drop in DNA quantity and subsequently OTU recovery (U’Ren et al. 2014). Studies on plants in urban or agricultural habitats lend themselves to easy adaptation of best-practice methods (it is practical to drive in a cooler of dry ice) but studies of plant microbiomes in remote habitats such as mountains, deserts, rainforests, or tundra, using such best- practices can be inhibitory. Without modifying current lab protocols – researchers can streamline communication across the phytobiome research field. For example, consistent growth stage sampling for descriptive analyses, such as during flowering; consensus on plant organ names, specifically, root-associated, rhizosphere, and rhizoplane, would make cross-study comparisons and microbial meta-analyses more straightforward. Although, these examples are straightforward when considering crop plants, they quickly become complicated when considering non-crop plants because of the sheer diversity in plant natural histories. Also, due to the diversity of plant tissue structure, no single extraction method can universally be used across all plants, for example, woody or fibrous tissues require more complex extraction methods. Furthermore, chemicals in 5 some leaves, including phenolic compounds, can complicate DNA, RNA, and metabolite extraction methods (Couch and Fritz 1990). But standardization of methods for major cash crop systems such as wheat (Triticum aestivum), maize (Zea mays), or soybean (Glycine max) would be more practical to establish. Several fungal barcode regions, 18S, SSU, LSU, were tested and the results published before a consensus was established to use the ITS region (which comprises the ITS1, 5.8S, and ITS2 segments) as the formal barcode marker for fungi (Kõljalg et al. 2013, Hibbett et al. 2016). Due to the short nature of Illumina reads, only one of the ITS segments is often used. Neither the ITS1 or ITS2 regions are superior and use of either of these markers yields similar pictures of fungal communities (Dentinger at al. 2011, Bazzicalupo et al. 2013, Blaalid et al. 2013). Other researchers use primers for the full length ITS region and analyze the reads without merging the forward and reverse reads. This strategy essentially generates two complementary datasets of the fungal community that are analyzed in parallel, one for ITS1 and one for ITS2 (Benucci et al 2016, Johansen et al. 2016). Across the diverse plant habitats where microbiome research occurs – tropical, boreal, deciduous, agricultural – adopting one best-practice method for sample collection and processing across such diverse habitats is not feasible, or even recommended. A better solution would be the adoption of standard methods within each sub-field. For example, researchers interested in crop microbiomes should adopt the following methods: (1) Immediate freezing of samples in the field, or immediate suspension of samples into DNA extraction buffer when freezing is inhibitory due to practicality (U’ren et al. 2014); (2) Freeze-drying tissue samples and subsequent room temperature storage under desiccant; (3) Use of peptide nucleic acid clamps (which bind to specific DNA and prevent PCR amplification) to reduce chloroplast contamination of sequence libraries 6 (Sakai and Ikenaga 2013). Use of ITS2 fungal primers to reduce bias against early-diverging fungal lineages that may be created with ITS1 primer sets (Toju et al. 2012, Bazzicalupo et al. 2013, Blaalid et al. 2013), or use of full length ITS primers that provide amplicons for both ITS regions (Benucci et al. 2016, Johansen et al. 2016). (4) Report of environmental conditions, such as weather, for the growing season and at the time of sampling; (5) Report of soil type, pH, and series of location sampled. The soil is often assumed to be the primary source of microbial inoculum, and Evans et al. (2016) have demonstrated that the diversity and composition of microbial inoculum is often linked to the diversity and composition of the final community. (6) Report disease history or symptoms on plant, where applicable; (7) Report land use history, e.g., the number of years farmed and types of crops farmed; (8) Current and previous crop management strategies used, including details regarding: fertilizers, pesticides, tillage, cover crops, and planting or harvesting equipment; (9) Clean tools between sample collections to prevent cross- contamination; (10) Collection of an appropriate number of plants from a field to obtain statistical power (Kelly et al. 2015); (11) Pool replicate PCRs for each DNA extraction and set-up libraries to achieve deep sequence coverage of each sample (Manter et al. 2010, Smith and Peay 2014); (12) Standardization of methods to isolate endophytic (within or between plant cells) or epiphytic (surface) microbes, preferably after a systematic analysis of various methods. If all plant microbiome researchers would adapt these points when designing and conducting their studies, it would greatly increase the frequency with which researchers could re-analyze and generate hypotheses about larger ecological plant microbiome patterns. 7 Shifting Attitudes in Disease Management Plant defense responses can be specific to one pathogen species, but plants also have generic defense systems that are triggered by conserved microbial cues such as flagellin and chitin (reviewed by Durrant and Dong, 2004). The increasing use of “-omics” technologies is changing our understanding from pairwise plant-pathogen or plant-symbiont interactions to more intricate and complex systems that show evidence of a shared evolutionary arms race. There is some evidence that plants can recruit other organisms to assist with defense against pathogen attacks. For example, plants emit volatile compounds under herbivore predation that can attract parasitoid wasps (De Moraes et al. 1998) or predatory nematodes of the herbivores (Rasmann et al. 2005). There are a few examples of specific inter-Kingdom interactions that aid the plant with defense, scientists also have numerous examples of non-specific interactions that cause reduced pathogen pressure in crop fields; the phenomenon of disease-suppressive soils. Disease suppressive soils are observed to inhibit pathogens across numerous years of monoculture (reviewed by Schlatter et al. 2017). There are two categories of suppressive soils: general and specific. General suppression occurs when an apparently diverse soil community causes a natural decline in the previously dominant pathogen population, and this is believed to be due to microbial competition. Specific suppression is believed to be due to antagonistic interactions between the pathogen and one or few other microbial species. The phenomenon of disease-suppressive soils has been observed in diverse crop systems such as take-all of wheat or barley, Phytophthora root rot of avocado, and crown gall of almonds; but the mechanisms behind suppressive soils are not yet understood (New and Kerr 1972, Baker and Cook 1974, Weller et al. 2002). There are limited examples where the suppressive-soil phenomenon is transferable. Transference is possible by inoculating the population of suppressive-soil to a conducive soil 8 (reviewed by Schlatter et al. 2017). It has been demonstrated that diverse microbial populations have suppressive qualities (Kinkel et al. 2011, Kinkel et. al 2012, Bakker et al. 2014). Additionally, complex plant communities, regardless of plant species, selected for soil microbes that were more effective at inhibiting Fusarium spp. (Essarioui et al. 2017). Microbial community analysis techniques are beginning to be applied in suppressive-soil systems to help elucidate the mechanisms – with the goal of transferring or inducing general suppression (Poudel et al. 2016). Researchers hope to analyze microbial community and big-data techniques to understand suppressive soils and begin to manipulate them for agricultural benefit. These techniques could also be applied to different populations of microbes, such as endophytes to protect against disease. Endophytes Offer Plant Protection The term endophyte is used to describe microorganisms that spend the majority or entirety of their life cycle living within a host plant. This classification is not taxon-specific, it is determined by organism lifestyle (Rodriguez et al. 2009). There is some debate about whether this classification is a valid lifestyle descriptor (e.g., saprotroph, necrotroph) or if it is better suited as a descriptor for a growth stage (van Overbeek and Saikkonen 2016). It has been proposed that endophytes may be latent pathogens or saprotrophs (van Kan et al. 2014), but it appears they instead are close relatives of pathogens (reviewed by Porras-Alfaro and Bayman 2010). Endophytic fungi have been documented to benefit their plant hosts in diverse conditions. There are several examples of endophytic fungi that do benefit the plant; such as providing heat, drought, or salt tolerance (Macia-Vicente et al. 2008, Rodriguez et al. 2008, reviewed by Rodriguez et al. 2009, Hubbard et al. 2012, Murphy et al. 2014, Mousa et al. 2016). They can improve salt and heat tolerance in wild grasses (Rodriguez et al. 2008). Epichloë endophytes prevent colonization of 9 smuts in grasses (Vignale et al. 2013). Endophytes can increase polyketide production of their plant hosts (Chagas et al. 2013). Seed-borne diseases decreased in barley when inoculated with endophytes (Murphy et al. 2014). In wheat, improved germination rates have been attributed to endophytes (Hubbard et al. 2012), as have protective effects against Stagonospora infection (Sieber et al. 1988). Recently, bacterial endophytes have shown disease and mycotoxin protective abilities in millet (Mousa et al. 2016). There are numerous biocontrol products on the market that prevent pathogen colonization, for example, BASF manufactures a Bacillus sp. seed coating for cotton that has anti-fungal properties, Serifel® (BASF, Florham Park, NJ, USA) and a Bacillus amyloiquefaciences strain from Valent to control nematodes (Valent, Liberytville, IL, USA). There are also biologically- based products that protect plants against abiotic stress, including the Epivio™ line from Syngenta (Syngenta, Basel, Switzerland). All products from Indigo Agriculture (Boston, MA, USA), and the mycorrhizal inoculant MycoGold (MycoGold, Cincinnati, OH, USA) provide plant growth- promotion. Biocontrol and plant-growth promoting products can have limited efficacy and may be effective for a limited range of pests or growing regions. When the mechanism of a biocontrol strain is known, it is usually only active for a narrow range of pathogens (reviewed by Fravel 2005). Evolutionary pressures or environmental factors can overcome the parasitic ability of a biocontrol strain. Implementing combinations of multiple biocontrol strains is sometimes more effective than inoculation with single strains (Xu et al. 2011). The first biocontrol strain, Agrobacterium radiobacter strain K84, was registered in the United States in 1979 for control of crown gall (Fravel 2005). Historically, identification and testing of biocontrol strains was time consuming. The current direction of the plant-microbiome field – use of high-throughput 10 technologies to analyze microbial communities and the interactions between community members, and the subsequent use of those data to identify important taxa in microbial networks – will likely provide more robust and sustainable control products (Poudel et al. 2016, Schlatter et al. 2017). Superior biocontrol products can be developed if the deployment of the strains is conducted in a manner which reduces competition or environmental stress. Implementation of endophytic biocontrol strains which can be vertically transmitted to the next generation of seed would reduce stress on these microbes from environmental factors or competition from soil and rhizosphere microbes and may make for more robust products on the market. Vertical transmission has been demonstrated previously in wheat (Huang et al. 2016), and should improve product-viability, but was not investigated in the strains used in this work. Vertical transmission of endophytic bacteria has been demonstrated in maize, soybean, pepper, and wheat (Mitter et al. 2017). Also, investigations of endophyte pathogenicity on non-host crops to prevent interference with rotations or neighboring fields is required to ensure efficacy of such biologically-based treatments. Previous Crop Microbiome Studies Of the three crop species (wheat, maize, soybean) studied here, wheat may have the most well-studied microbial community in terms of diversity of plant compartments and microbial groups. Across these crops there has been a general bias toward studying prokaryotes and conducting community analysis with 16S-based methods (Table 1-1). Investigations of fungal and eukaryotic microbial communities were less common. There are examples of culture-based and non-sequence-based methods (such as DGGE) that investigated the eukaryotic communities of maize (Saravanakumar et al. 2017) and soybean (Hamid et al. 2017), but only a few high- throughput sequence-based studies analyzed the fungal communities associated with these crop 11 plants, and these studies only investigated the rhizosphere fungal communities (Table 1-1). In addition to the bias against fungal communities, there is also a bias against analysis of phyllosphere microbial communities. Poor recovery of relatively low-abundance microorganisms, such as on a leaf surface, may be partly the cause, especially when community analysis methods available before next-generation sequencing were used. To our knowledge, only one previous study investigated maize leaf communities using high-throughput amplicon sequencing (Johnston- Monje et al. 2016), others used T-RFLP or ARISA (Balint-Kurti et al. 2010, Nettles et al. 2016). Similarly in soybean, there were fewer studies of leaf-associated microbial communities of field plants using amplicon sequencing (Copeland et al. 2015) or other techniques (Nettles et al. 2016), when compared with the number of root or root-associated studies. Thanks to the deep sequence coverage available with current methods, it is possible to recover and identify previously low- abundance and unknown organisms. Culture-based methods and other historic techniques, have yielded a preliminary picture of microbial community composition on crop plants, with the increased amounts of data generated with next-generation sequence methods, we know that culture-based and other techniques have only captured a small glimpse of microbial communities. There is a large gap in our knowledge regarding the detailed composition of microbial communities associated with various organs of crop plants. Filling in these gaps will provide a valuable resource for microbiome engineering to improve agricultural outputs. Despite the limitations of current and previous methods, and gaps in our understanding of community composition, we have a general understanding of community composition in specific scenarios. Previous studies have indicated that soil microbial communities differ when compared across management strategy (Berthrong et al. 2013, Xue et al. 2013, Hartmann et al. 2014) but the plant-associated or phyllosphere communities show less difference (Coleman-Derr et al. 2015, 12 Karlsson et al. 2017). The plant species in a rotation series appear to have an effect (Hartmann et al. 2014). Land-use history also influences the community, but with time, communities of restored landscapes become more similar to native communities (Jangid et al. 2011). 13 Crop wheat wheat wheat wheat wheat wheat wheat wheat soil soil soil soil soil soil soil organic (NOFERT) organic (BIODYN) organic (BIOORG) conventional (CONMIN) conventional (CONFYM) conventional yes yes yes yes yes no no Geneb 16S 16S 16S, ITS 16S, ITS 16S, ITS 16S, ITS 16S, ITS 16S 16S 16S 16S 16S 16S 16S 454 454 454 454 Techniquec Reference Zhao et al. 454 2016 Zhao et al. 2016 Hartmann et al. 2014 Hartmann et al. 2014 Hartmann et al. 2014 Hartmann et al. 2014 Hartmann et al. 2014 Chávez- Romero et al. 2016 Chávez- Romero et al. 2016 Jiménez-Bueno et al. 2016 Yin et al. 2017 454 454 454 454 454 454 TRFLP, 454 Donn et al. Yin et al. 2017 Bouffaud et al. 2014 2014 Li et al. 2012 454 microarray GeoChip Table 1-1. Highlights of plant-microbiome studies of wheat, maize, or soybean conducted to date. Organ/ Compartment soil Managementa conventional organic Rotation rice, wheat rice, wheat wheat soil reduced inputs wheat wheat wheat wheat wheat wheat soil rotation soil, rhizosphere no-till soil, rhizosphere soil, rhizosphere conventional greenhouse wheat, maize winter, spring wheat wheat, fallow no soil, rhizosphere no-till no soil, rhizosphere rice, wheat 14 Table 1-1 (cont’d) wheat soil, rhizosphere greenhouse no wheat wheat wheat wheat wheat wheat wheat wheat wheat wheat wheat wheat wheat wheat soil, rhizosphere, root soil, rhizosphere, root soil, rhizosphere, root rhizosphere organic integrated conventional growth chamber wheat, clover + forage potato, wheat, bean, wheat wheat, oil seed rape, wheat no rhizosphere conventional rhizosphere, root no-till rhizosphere, root conventional root soil, rhizosphere, leaf, endospheres soil, rhizosphere, leaf, endospheres soil, rhizosphere, root, leaf, endosphere leaf leaf seed conventional greenhouse greenhouse growth chamber organic, conventional reduced inputs greenhouse no no chickpea, pea, lentil, wheat no wheat, bean no no no no no 15 16S bacteria, fungi, oomycetes bacteria, fungi, oomycetes bacteria, fungi, oomycetes 18S 16S 16S ITS, 18S ITS 16S, ITS 16S, ITS 16S ITS ITS 454 culture culture culture cloning Illumina 454 454 cloning Illumina Illumina 454 454 454 16S, ITS Illumina Ofek et al. 2013 Lenc et al. 2014 Lenc et al. 2014 Lenc et al. 2014 Smit et al. 1999 Mahoney et al. 2017 Rascovan et al. 2016 Borrell et al. 2017 Kwaśna et al. 2010 Granzow et al. 2017 Granzow et al. 2017 Liu et al. 2017 Karlsson et al. 2017 Sapkota et al. 2017 Huang et al. 2016 Table 1-1 (cont’d) maize soil soil soil soil soil soil soil soil soil soil soil soil soil maize maize maize maize maize maize maize maize maize maize maize maize maize maize maize conventional integrated organic conventional organic no-till conventional rotation no-till conventional rotation rotation no no no no no no no yes no no yes wheat, maize 16S 16S metagenomic metagenomic metagenomic metagenomic metagenomic metagenomic 16S 16S 16S 16S 16S PFLA PFLA PFLA Ion Torrent Ion Torrent 454 454 454 454 454 454 454 454 454 454 454 Lazcano et al. 2013 Lazcano et al. 2013 Lazcano et al. 2013 Lupatini et al. 2017 Lupatini et al. 2017 Souza et al. 2013 Souza et al. 2013 Souza et al. 2013 Souza et al. 2015 Souza et al. 2015 Souza et al. 2015 Duncan et al. 2016 Jiménez-Bueno et al. 2016 Ofek et al. 2013 Peiffer et al. 2013 Peiffer et al. 2013 soil, rhizosphere greenhouse soil, rhizosphere conventional soil, rhizosphere organic no no no 16 Table 1-1 (cont’d) maize soil, rhizosphere greenhouse rhizosphere rhizosphere rhizosphere rhizosphere soil, rhizosphere, root root root, rhizosphere, leaf episphere conventional greenhouse greenhouse greenhouse growth chamber conventional greenhouse leaf endosphere conventional leaf episphere leaf seed soil soil soil no-till maize maize maize maize maize maize maize maize maize maize maize soybean soybean soybean soybean soybean soybean no no no no no no no no no no no no no no no 16S 16S, ITS 16S, ITS 16S 16S 16S 18S 16S ITS 16S 16S 16S 16S ITS 16S microarray ARISA 454 454 454 Illumina Bouffaud et al. 2014 Nettles et al. 2016 Saravanakumar et al. 2017 Zhu et al. 2016 Zhu et al. 2016 Niu et al. 2017 Illumina Gosling et al. 2013 Johnston- Monje et al. 2016 Nettles et al. 2016 Balint-Kurti et al. 2010 Mashiane et al. 2017 cloning Liu et al. 2012 454, DGGE Babujia et al. Illumina 2016 Han et al. 2017 Hernandez et al. 2017 Liu et al. 2015 Lu et al. 2017 Mendes et al. 2014 TRFLP ARISA TRFLP Illumina Illumina soil soil, rhizosphere soil, rhizosphere reduced inputs greenhouse maize, soybean no no fungi, nematodes DGGE Illumina 16S metagenomic 454 17 Table 1-1 (cont’d) soybean rhizosphere soybean rhizosphere soybean soybean soybean soybean soybean soil, rhizosphere, root soil, root rhizosphere, root root root soybean soil, leaf conventional greenhouse greenhouse greenhouse no-till conventional various soybean leaf endosphere conventional no no no no no no yes no no 16S, ITS 16S, ITS 16S bacteria 16S 18S ARISA Illumina Illumina TRFLP 454 TRFLP oomycetes (ITS) Illumina 16S ITS Illumina ARISA Nettles et al. 2016 Hamid et al. 2017 Xiao et al. 2017 Sun et al. 2017 Rascovan et al. 2016 Gosling et al. 2013 Rojas et al. 2017 Copeland et al. 2015 Nettles et al. 2016 ablank cells due to missing information in publications. b16S: prokaryotic marker gene, ITS: fungal marker gene, 18S: eukaryotic marker gene, blank cells did not use DNA or sequence-based technique. c454: 454-Roche pyrosequencing, Illumina: Illumina MiSeq. 18 Microbes in Agriculture Microorganisms, including archaea, bacteria, fungi, protists, and others, are ubiquitous in natural, agricultural, and built environments. Microbes are responsible for nutrient cycling, carbon and nitrogen cycling, organic matter degradation, and other ecological processes. Agriculturally relevant microbes are often pathogens such as F. graminearum, but microbes can also play a beneficial role in plant health and crop productivity. Microorganisms can interact with their plant hosts in numerous ways, ranging from beneficial or commensal to pathogenic or parasitic. For example, rhizobia and mycorrhizal fungi exchange nutrients such as nitrogen, carbon, and phosphorus, in addition to micronutrients during their symbioses (Bhuvaneswari et al. 1980, van der Heijden et al. 1998, Smith and Read 2008, Lira et al. 2015). Commercial strains of rhizobia and mycorrhizal fungi are available for purchase as growth-promoting supplements on crops, for example rhizobia-coated seeds sold by Smith Seed Services (Halsey, OR, USA), and Mykos spores from mycorrhizal fungi sold by Xtreme Gardening (Gilroy, CA, USA). There are also a range of neutral plant-microbe interactions, endophytic fungi grow asymptomatically between cells of a plant, and do not always provide a clear benefit to the plant hosts. There are beneficial or protective endophytes, as discussed above. Growing concerns about sustainability of modern agricultural systems has driven biocontrol research for both growth-promoting and pathogen control applications. Biocontrol products could fill a void in current control methods for certain pathogens, such as F. graminearum, for which no effective chemically based control is available. Fusarium graminearum Head Blight of Wheat F. graminearum is a filamentous fungal pathogen of cereal crops. Head Blight of wheat, also known as Head Scab, was first reported in the United States in 1891 (Arthur 1891). Fungi 19 have historically been given unique Latin binomials for each phase of growth, the anamorphic (asexual) and the telomorphic (sexual) growth stages. Use of the telomorphic name, Gibberella zeae, was recently dropped during the ‘one fungus, one name’ initiative (Taylor 2011, Wingfield et al. 2012). Head Blight of wheat is actually a disease complex and be caused by simultaneous or independent infections of F. graminearum and several other Fusarium spp. (Fusarium culmorum, Fusarium avenacum) or Microdochium species, but in the United States, this disease is most commonly caused by F. graminearum (Ioos et al. 2005, Xu et al. 2005, Siou et al. 2015). Previously, distinct populations of F. graminearum were believed to cause root and head infections. But due to genetic evidence, the F. graminearum clade was split into two species, F. graminearum and Fusarium pseudograminearum (Aoki and O’Donnell 1999). F. pseudograminearum is soil-borne, causing root and crown infections primarily in Australia (Aoki and O’Donnell 1999, Burgess et al. 2001). F. graminearum can cause seedling damping-off in fields with high inoculum levels and is the primary pathogen responsible for crown infections in the United States (Sutton 1982, Wiese 1987). Fusarium Head Blight is an economically devastating disease, causing an estimated $2-3 billion yield loss in the last several decades (McMullen et al. 1997, 2012). Although Head Blight can have such a severe impact, there is not one completely efficient control strategy. Head Blight is currently controlled with an integrated approach, there are no strongly resistant plants available, and fungicides are only moderately effective (Wegulo et al. 2015). Resistance mechanisms include prevention of initial infection, or prevention of spread throughout florets (Schroeder and Christensen 1963, Cook 1981, Mesterhazy 1995). Fungicides are usually sprayed at flowering to coincide with the time of F. graminearum ascospore release. Plants infected with F. graminearum 20 may still able to complete their life cycle, but the seed can be shriveled and depleted and may not be fit for food or feed due to mycotoxin contamination. F. graminearum produces an arsenal of mycotoxins, which are small molecules detrimental to health. Deoxynivalenol (DON), the primary mycotoxin produced by F. graminearum causes gastrointestinal distress (reviewed by Sobrova et al. 2010). Zearalenone, another mycotoxin, binds to the mammalian estrogen receptor and can interfere with animal development (Miksicek 1994, reviewed by Desjardins and Proctor 2007). Aurofusarin is a red-pigmented secondary metabolite produced by F. graminearum that has been linked to problems with egg development in the poultry industry and has weak antimicrobial activity (Mendentsev et al. 1993, Dvorska et al. 2001). Contamination of malting barley with F. graminearum interferes with malting process used during beer production. Hydrophobins from the fungus cause a chemical reaction colloquially termed “gushing” due to the explosive effect it can generate when beer is released from a container (Schwarz 2017). Conclusions Traditional microbiology and ecology have provided scientists with insight into the structure of microbial communities across various natural habitats. With the development of culture-independent microbial ecology techniques, scientists are able to understand the dynamics of microbial communities with ever-increasing detail. Although our understanding of microbial communities of natural habitats is continually improving, there are some gaps in our knowledge that the work presented in this dissertation aims to fill: the structure of fungal and bacterial communities across multiple plant organs and how it varies across growth stages, and land management strategies (Chapter 2-3). One of the intended applications of crop microbial 21 community analysis is the manipulation of these communities to reduce plant disease or increase crop productivity. I describe use of my findings from the wheat microbiome to initiate control of Fusarium Head Blight with beneficial endophyte strains (Chapter 2, Appendix A). I have initiated a functional analysis of F. graminearum terpene synthase genes to provide insight into the function of terpenes as signals for community manipulation and self-defense molecules for this fungal pathogen (Appendix B). Community structure is affected by various genetic and environmental factors, and understanding these factors is necessary for targeted manipulation of communities to reduce plant pathogen presence. 22 CHAPTER 2 THE WHEAT MICROBIOME UNDER FOUR MANAGEMENT STRATEGIES, AND POTENTIAL FOR ENDOPHYTES IN DISEASE PROTECTION Abstract Manipulating plant-associated microbes to reduce disease or improve crop yields requires a thorough understanding of interactions within the phytobiome. Plants were sampled from(cid:1)a wheat-maize-soybean crop rotation site that implements four different crop management strategies. We analyzed the fungal and bacterial communities of leaves, stems, and roots of wheat throughout the growing season using fungal internal transcribed spacer and bacterial 16S rRNA gene amplicon sequencing. The most prevalent operational taxonomic units were shared across all samples, although levels of the low-abundance operational taxonomic units varied. Endophytes were isolated from plants and tested for antagonistic activity toward the wheat pathogen Fusarium graminearum. Antagonistic strains were assessed for plant protective activity in seedling assays. Our results suggest that microbial communities were strongly affected by plant organ and plant age and may be influenced by management strategy. Source For a full text of this work go to: Gdanetz K, Trail F. 2017. The wheat microbiome under four management strategies, and potential for endophytes in disease protection. Phytobiomes. 1:158- 168. https://apsjournals.apsnet.org/doi/full/10.1094/PBIOMES-05-17-0023-R 23 CHAPTER 3 HUB TAXA AND CORE MICROBIAL COMMUNITIES OF ROTATION CROPS UNDER COMMON LAND MANAGEMENT STRATEGIES Abstract Manipulation of naturally occurring microbial communities to reduce plant diseases or increase crop yields requires a thorough understanding of interactions within the phytobiome. This paper presents the first in-depth comparative analysis of fungal and bacterial microbial communities in these economically important row crops. Plants were sampled from a long-term wheat-maize-soybean crop rotation research site under four different land management strategies. The fungal and bacterial communities of leaves, stems, and roots of all crops throughout the growing season were analyzed using fungal internal transcribed spacer and bacterial 16S rRNA gene amplicon sequencing. Our analysis showed that microbial communities differed across growth stages and plant organs of each host; this effect was most pronounced in the bacterial communities of wheat and maize aboveground organs. Roots consistently had the most unique bacterial OTUs. Fungal OTU composition across leaves and stems of wheat and soybean were similar, but the most notable changes in fungal OTU composition of the microbial communities were between roots of maize and soybean. Network analysis identified important members of the microbial communities of the three crops. Understanding the microbial community structure across various organs of crop plants and identifying important taxa will provide a critical resource for researchers for future experimental manipulations with plant-associated communities. 24 Introduction Microorganisms are ubiquitous in natural environments, including nearly all plant organs and tissues. These microbes have diverse functions, for example, rhizobia and mycorrhizae form symbioses with roots and provide nutritional advantages to their plant hosts (Bhuvaneswari et al., 1980, ven der Heijden et al. 1998). Phyllosphere microbes can fix carbon (Abanda-Nkpatt et al. 2006, Vorholt 2012) and stimulate plant defenses (Vogel et al. 2016). Plant pathogens are another guild of microorganisms that infect plant organs; root pathogens cause rots, foliar pathogens infect and destroy healthy leaf tissues, and many pathogens can infect the developing seed. The emergence of the plant-microbiome and phytobiome fields of study are causing plant pathologists to take a more integrated approach toward researching plant diseases; for example, biotic and abiotic factors are considered together, and microbial pathogens are viewed as part of a community instead of in isolation (Vorholt 2012, Mueller and Sachs 2015, Berg et al. 2016, Ellis 2017, Leach et al. 2017). Part of this research initiative requires identification of plant-associated microbial communities, or microbiomes. There is growing interest in the plant pathology field to treat plant diseases via manipulation of whole microbial communities. Before experimental manipulation of communities can occur, the core communities, microbes shared across habitats or niches, on various crop species or various plant organs must be identified, and better understood in terms of functional interactions between the host plant and the microbiome. The three most commonly grown crops in the United States are maize, soybean, and wheat with approximately 88, 83, and 47 million respective acres planted (USDA 2016). This paper presents the first in-depth comparative analysis of fungal and bacterial microbial communities in these economically important row crops. 25 The Michigan State University W. K. Kellogg Biological Station (KBS) Long-Term Ecological Research (LTER) site has been managed under the current strategies and crop rotations for more than 25 years (Robertson 2015). It is an ideal research site to test effects of management strategies on microbial communities since the different strategies are laid out in a replicated and randomized field. The wheat community data used here (Gdanetz and Trail 2017) have been previously investigated. In this study, we analyzed the fungal and bacterial microbial communities, with amplicon sequencing of maize and soybean, grown under four common management strategies. We also analyzed the changes in microbial communities at three growth stages for each crop, over the course of one three-year rotation cycle. The KBS-LTER site hosts a multi-year three-crop rotation series of wheat, maize, and soybean. Wheat is planted prior to maize, to avoid high disease pressure on the wheat crop since maize is also a host for wheat head blight pathogen, Fusarium graminearum, and the maize crop residue remains in the soil for two years, hosting pathogen propagules. Soybean follows maize, providing an opportunity for the rhizobia to replenish nitrogen in the soil. Prior to use as a research field site, the area was farmed privately. The crop rotation site is divided into plots containing the rotation series under four common crop management strategies; conventional, no-till, low inputs, and organic (Robertson, 2015). The KBS-LTER site provides a nearly ideal site to study the effects of agriculture on crop-associated microbial communities because of the long-term treatments and random replicate plot design. The goal of this multi-year study was to establish a detailed profile of a healthy microbial community on economically important crops. 26 Materials and Methods Site Description and Sample Collection The study site is located at the Kellogg Biological Station Long-Term Ecological Research (KBS LTER) site in Hickory Corners, Michigan, USA (42.411085, -85.377078). Annual precipitation for this site is 1,005 mm, with about half falling as snow (NCDC 1980-2010 climate normals for the Gull Lake Biological Station, https://lter.kbs.msu.edu/research/site-description- and-maps/general-description/climate-normals/). The annual precipitation during the study was 1177, 932, and 1153 mm, during wheat, maize, and soybean growing years, respectively. The KBS-LTER crop rotation site has been planted in the wheat-maize-soybean rotation series since 1993, and is organized in randomized, replicated plots under four land management strategies with six replicate plots for each strategy: T1 = conventional till, T2 = no-till, T3 = reduced chemical inputs with alfalfa cover crop, T4 = organic with alfalfa cover crop (Figure 3-1). Detailed management information can be found at https://lter.kbs.msu.edu/research/. Wheat seeds of the same cultivar, Pioneer 25R39 Soft Red Winter Wheat (Pioneer High-Bred International, Inc., Johnston, IA, USA), were planted in all plots, but wheat seeds for plots T1-T3 were coated with Gaucho fungicides (Bayer Crop Science, Research Triangle Park, NC, USA). A commercial maize hybrid cultivar, Dekalb DKC52-59 Corn Hybrid (Monsanto Company, St. Louis, MO, USA), was planted in plots T1-T3, and an organic-approved cultivar, Blue River Hybrids 25M75 Organic Corn (Blue River Organic Seed, Ames, IA, USA), was planted in T4 plots. A commercial soybean cultivar, Pioneer P22T69R (Roundup Ready®) Soybean Seed (Pioneer High-Bred International, Inc., Johnston, IA, USA), was planted in T1-T3 plots, and Viking Organic Soybean Seed, Variety 0.2265 (Albert Lea Seed, Albert Lea, MN, USA), was planted in T4 plots. Plants were collected at three analogous developmental stages for each crop: late vegetative growth, flowering, and early 27 seed/ear/pod development (Table 3-1, Figure 3-2). At each developmental stage, three intact plants were removed from each of the 24 plots. Fine and thick roots, and above-ground tissues from each plant were placed in separate sterile sample collections bags (Nasco Whirl-Pak®, Fort Atkinson, WI, USA) and maintained on ice during transport. Roots were rinsed to remove loosely attached soil, and plants were stored at -80°C, then lyophilized. Lyophilized plant tissue was stored at room temperature under a desiccant. Sample Processing Lyophilized plant tissue was ground with a Retsch Oscillating Mill M400 (Verder Scientific, Newtown, PA, USA) and DNA was extracted using the Mag-Bind® Plant DNA Plus Kit (Omega Bio-tek, Norcross, GA, USA) following the manufacturer’s protocols with the KingFisher™ Flex (ThermoFisher Scientific, Waltham, MA, USA). Fungal ITS2 and bacterial 16S rRNA gene libraries were generated as described previously and sequenced using Illumina MiSeq 2x250 bp chemistry (Table 3-2; Gdanetz and Trail 2017). Maize microbiome sequences are available with the National Center for Biotechnology Information (NCBI) Small Reach Archive project accession numbers for wheat, maize and soybean are SRP102192, SRP102245, and SRP120500, respectively. Sequence Processing and Community Analysis Forward and reverse read pairs from each library were merged with USEARCH (version v8.1.1861; Edgar 2010, Edgar et al. 2011, Edgar and Flyvbjerg 2015). Low quality reads, (fastq expected error set to 1.0), and read pairs without mates were discarded (read length of 250 bp for 16S and 380 bp for ITS sequences). The 16S and ITS2 libraries from all three crops were concatenated into one large dataset for each barcode region. Sequences were processed as described previously (Gdanetz and Trail 2017). Briefly, the USEARCH pipeline was implemented 28 for chimera filtering and OTU assignment, with the cluster threshold set to 97% similarity. The Ribosomal Database Project Classifier was used for taxonomic assignment, with the 16S and the UNITE ITS training sets for bacteria and fungi, respectively, customized to include the plant hosts (Wang et al. 2007, Cole et al. 2013, Deshpande et al. 2016). OTUs matching plants, mitochondria, chloroplasts, or unidentified at Kingdom level were discarded. Samples were filtered to include OTUs that occurred in at least five samples and this trimmed dataset was used for all downstream analyses. Venn diagrams and tables for core taxa analysis were generated using the gplots package (Warnes et al. 2016). Before generating barplots, taxa were merged at Class level. Taxa that were present in less than 3% of the samples were removed. Relative abundances of taxa were calculated as a percentage of total sequences in each sample. Alpha diversity statistics (abundance transformation, observed OTUs, and Shannon’s Index) were calculated with the Phyloseq package (McMurdie and Holmes 2013) in the R statistical computing environment (version 3.3.3; R Core Team 2016). Non-rarefied data were used to calculate the Shannon Diversity Index (H’), an alpha diversity metric that measures species diversity within a sample (McMurdie and Holmes 2014), and analysis of variance (ANOVA) and Tukey’s honest significant difference test (HSD) were used to determine significance. Non-metric multi-dimensional scaling (NMDS) ordination analysis was conducted using Bray-Curtis distance values, a measure of species diversity between communities. Permutational analysis of variance (PERMANOVA) tests were completed using the adonis function in the vegan R package (Oksanen et al. 2016). Heterogeneity of the variances was calculated using the betadispersion function of the vegan package. Figures were generated with the ggplot2 package (Wickham 2009). Microbial hubs, species that influence the presence of others in an environment, were identified as highly connected nodes, with ten or more connections. 29 Fungal and bacterial OTU tables were concatenated before calculating the co-occurrence network using the SpiecEasi package in R (Kurtz et al. 2015). Network statistics and network plots were generated using Cytoscape (version 3.5.1, http://www.cytoscape.org). 30 Table 3-1. Collection dates of plant growth stages. Crop Wheat Wheat Wheat Maize Stage Namea Age 30 45 83 V7 Vegetative (C1) Boot/flowering (C2) Early seed development (C3) Vegetative (C1) Flowering (C2) Early ear development (C3) Vegetative (C1) Flowering (C2) Early pod development (C3) Managementb Collection Date 1-May-2013 T1-T4 30-May-2013 T1-T4 T1-T4 5-July-2013 24-Jun-2014 T1, T2 T3 01-Jul-2014 09-Jul-2014 T4 16-Jul-2014 T1, T2 23-Jul-2014 T3 T4 06-Aug-2014 20-Jul-2014 T1, T2 06-Aug-2014 T3 T4 19-Aug-2014 19-Jun-2015 T1-T3 29-Jun-2015 T4 T1-T3 15-Jul-2015 24-Jul-2015 T4 T1-T4 01-Sep-2015 Maize VT Maize R1 Soybean V2 Soybean R1 Soybean R6 aWheat growth stages were rated on Zadoks scale (Zadoks et al. 1947). bT1: conventional, T2: no-till, T3: reduced inputs, T4: organic. Table 3-2. Primer sequences of microbiome target loci. Primer 515f 806r ITS3_KYO4 ITS4_KYO3 ITS1F ITS4 Sequence 5' GTG CCA GCM GCC GCG GTAA 3' 5' TAA TCT WTG GGV HCA TCA GG 3' 5' GAT GAA GAA CGY AGY RAA 3' 5' CTB TTV CCK CTT CAC TCG 3' 5' CTT GGT CAT TTA GAG GAA GTA A 3' 5' TCC TCC GCT TAT TGA TAT GC 3' References 16S v4 PCR primers from (Kozich et al. 2013). ITS2 PCR primers from (Toju et al. 2012). Full length ITS primers (White et al. 1990, Bruns & Gardes 1993). 31 Figure 3-1. Layout of management strategy replicates at Michigan State University KBS LTER field site. Colored plots are planted in a three-year wheat-maize-soybean rotation cycle. Modified from http://lter.kbs.msu.edu/research/annual-plot-maps/. 32 Figure 3-2. Overview of plants and growth stages. For plant collection dates and formal growth stage names see Table 3-1. 33 Results Observed and Core Taxa We analyzed bacterial and fungal communities of maize and soybean grown under four different management strategies over a two-year period. After filtering and quality control, there were 14,395,302 and 27,341,137 high quality reads for fungi and bacteria, respectively, with means of 21,245 fungal and 42,166 bacterial reads per sample. No fungal sequences from soybean T4 stems and roots passed quality control. 7,728 fungal and 13,770 bacterial OTUs were identified from the combined data of all crops (Table 3-3). A total of 4,739 fungal OTUs and 8,942 bacterial OTUs were identified after trimming and filtering. Of the predicted bacterial and fungal OTUs, fewer OTUs were unique to soybean plants (Figure 3-3). Roots consistently had the most unique bacterial OTUs, with higher numbers of bacterial OTUs unique to roots than shared among organs in maize and soybean (Figure 3-4). 34 Figure 3-3. Shared and unique bacterial or fungal operational taxonomic units among all crops. 35 Figure 3-4. Shared and unique operational taxonomic units across plant organs. 36 Maize Soybean 16S library Wheat Maize Soybean 27,341,137 14,047,252 8,746,925 4,546,960 272,532 1,801,827 1,067,099 462,196 13,770 13,387 Wheat ITS library 14,395,302 4,538,082 6,998,733 2,760,569 3,794,174 1,451,825 1,268,377 1,055,034 Table 3-3. Sequence processing statistics and OTU assignment statistics. Step in Pipeline Merged read pairs Reads after dereplication Total OTUs OTUs after chimera removal 7,574 7,728 37 Samples with zero high-quality readsa aC-number indicates growth stage, T-number indicates treatment, R-number indicates replicate plot; L, S, R indicate leaf, stem, or root samples. Growth stage descriptions can be found in Table 3-1. Table 3-4. Samples without high-quality reads after sequence processing. Sequence Library Wheat ITS Maize ITS Soybean ITS Wheat 16S C1,T2,R6,L C1,T4,R3,S C3,T3,R6,R C3,T4,R6,R C1,T2,R6,R C2,T3,R2,S C3,T4,R1,R C3,T4,R6,S C2,T3,R4,S C3,T4,R1,S C3,T4,R2,R C3,T4,R2,S C3,T4,R3,R C3,T4,R3,S C3,T4,R4,R C3,T4,R4,S C3,T4,R5,R C3,T4,R5,S C3,T4,R6,R C3,T4,R6,S 38 Within a crop species, OTUs were generally conserved across growth stages and organs. Highly-abundant bacterial OTUs, belonging to Bacilli, Alphaproteobacteria, and Gammaproteobacteria, were generally shared across all crops (Figure 3-5). The phyllosphere organs, leaves and stems, of each crop had similar bacterial taxa at the Class level when compared with that crop’s roots (Figure 3-5, 3-6). Maize roots contained the highest number of unique bacterial taxa compared to other maize organs and other crops (Figure 3-4). OTUs found in maize leaf and stem samples were consistent across all management types at the oldest growth stage (Figure 3-7). Soybean root samples were dominated by Alphaproteobacteria. Bacterial communities also changed with growth stage; this effect was most pronounced in wheat and maize aboveground organs (Figure 3-5, 3-7). Both soybean leaf and stem samples had many unique taxa at the youngest and oldest growth stages (Figure 3-8). Although bacterial communities were similar across all management strategies within a crop at the Class level (Figure 3-6), we observed differences among crops and organs (Figure 3-5). OTUs found in maize stems were more consistent between no-till and low input samples. Sphingobacteria were not abundant under any managements, except maize organic (Figure 3-7). Unique taxa (Actinobacteria, Flavobacteriia, Opitutae) were observed in no-till and low input soybean samples at the vegetative stage (Figure 3-8). 39 Figure 3-5. Class-level relative abundance of operational taxonomic units in bacterial communities across crop, organ, and growth stage. 40 Figure 3-6. Class-level relative abundance of operational taxonomic units in bacterial communities across crop, organ, and management strategies. 41 Figure 3-7. Class-level relative abundance of operational taxonomic units in bacterial communities across maize growth stage, organ, and management strategies. 42 Figure 3-8. Class-level relative abundance of operational taxonomic units in bacterial communities across soybean growth stage, organ, and management strategies. 43 Fungal OTUs were more uniformly represented across crops than bacteria. The Dothideomycetes and Sordariomycetes were the most abundant fungal classes observed (Figure 3-9). Fungal OTU composition of leaves and stems of wheat and soybean were similar, but the most notable changes in fungal OTU composition of the microbial communities were between roots of maize and soybean (Figure 3-9). Root samples of all crops had a higher abundance of Agaricomycetes than phyllosphere organs (Figure 3-9, 3-10). Tremellomycetes were observed in phyllosphere organs on maize, but in no other samples (Figure 3-9, 3-11). Differences in some of the low-abundance taxa were observed across all growth stages (Figure 3-9). Fungal community members changed across growth stages on maize (Figure 3-11). Taxa more common and abundant in soybean, specifically Glomeromycetes and Orbiliomycetes, were observed only at vegetative and flowering growth stages in soybean roots (Figure 3-9, 3-12). As observed for bacteria, fungal communities were similar across all management strategies within each crop (Figure 3-10, 3-11, 3-12). 44 Figure 3-9. Class-level relative abundance of operational taxonomic units in fungal communities across crop, organ, and growth stage. 45 Figure 3-10. Class-level relative abundance of operational taxonomic units in fungal communities across crop, organ, and management strategies. 46 Figure 3-11. Class-level relative abundance of operational taxonomic units in fungal communities across maize growth stage, organ, and management strategies. 47 Figure 3-12. Class-level relative abundance of operational taxonomic units in fungal communities across soybean growth stage, organ, and management strategies. 48 Community Diversity Bacterial H’ of roots was generally significantly higher than H’ of leaves and stems, but there was not a significant difference in H’ across organs of fungal communities (Figure 3-13, Table 3-5). The mean fungal diversity of maize roots was lower than aboveground organs at the vegetative stage but was significantly lower only when compared with leaves from conventional plots. (Figure 3-14A, Table 3-5). Whereas fungal diversity in soybean roots decreased as plants aged, this trend was only significant when compared to leaves from low input plots during the flowering stage (Figure 3-14B, Table 3-5). Bacterial H’ of maize roots was higher than aboveground organs at all growth stages and under all management strategies, this was a significant increase except in T1 leaves and T1-T3 stems at the vegetative stage (Figure 3-14C, Table 3-5). The bacterial diversity of soybean roots was more variable than maize, but also decreased across the growing season (Figure 3-14D). Soybean leaves and stems during pod development from all management strategies had significantly higher H’ compared to roots, except stems from T4 (Table 3-5). 49 Figure 3-13. Global alpha diversity patterns calculated by Shannon Diversity Index (H’). All treatments and replicates were pooled within growth stages for (A) fungi and (B) bacteria. Data are represented by 24 replicates from each crop-growth stage-organ combination. Center line of boxes represents median of samples. The upper and lower sides of the boxes represent the third and first quartiles, respectively. Whiskers represent ± 1.5 times the interquartile range. Data points beyond whiskers represent outliers. Analysis of variance and Tukey’s honest significant difference were used to test significance (P < 0.05). Statistical support is detailed in Table 3-5. 50 Figure 3-14. Alpha diversity maize fungi (A), soybean fungi (B), maize bacteria (C), soybean bacteria (D) found on different plant organs across growth stages, and under four different management strategies, estimated by Shannon Diversity Index. T1=conventional, T2=no till, T3=low input, T4=organic. Reproductive growth stage indicates early ear development, and early pod fill for maize and soybean, respectively. Data are represented by six replicates from each stage-management-organ combination. Center line of boxes represents median of samples. The upper and lower sides of the boxes represent the third and first quartiles, respectively. Whiskers represent ± 1.5 times the interquartile range. Data points beyond whiskers represent outliers. Analysis of variance and Tukey’s honest significant difference were used to test significance (P < 0.05). Statistical support is detailed in Table 3-5. 51 V. F. S. V. F. S. L S R L S R L S R L S R L S R L S R T1 T3 T4 1.86±6.17a 2.45±0.35 2.06±0.22 2.42±0.34 1.99±0.32 2.20±0.66 2.44±0.16 2.62±0.49 3.10±0.30 2.30±1.06 2.96±0.48 3.18±0.27 2.38±0.30 2.27±0.26 2.22±0.31 1.47±0.22 2.15±0.33 2.34±0.52 2.24±0.50 2.12±0.20 2.49±0.64 2.44±0.73 2.68±0.24 2.33±0.54 1.66±0.14 1.59±0.04 1.57±0.28 0.95±0.36 1.89±0.47 1.76±0.34 1.69±0.36 1.59±0.57 2.37±0.28 2.44±0.45 2.28±0.38 1.79±1.16 Fungi T2 T3 T1 T4 2.57±0.24 2.72±0.07 2.18±0.12 2.17±0.64 1.89±0.51 2.15±0.61 1.71±0.24 2.32±0.14 2.72±0.86 2.75±0.20 3.09±0.32 2.08±0.90 2.18±0.22 2.30±0.32 1.55±0.19a 2.30±0.19 2.47±0.21 2.30±0.33 1.68±0.20 2.30±0.18 2.51±0.41 2.54±0.39 2.21±0.32 2.15±0.42 1.96±0.15 1.81±0.59 1.99±0.13 1.61±0.73 2.43±0.34 2.54±0.37 2.00±0.16 2.21±0.76 1.75±0.96 2.32±0.47 Soybean T1 3.42±1.03 3.21±0.52 4.88±1.52 2.97±0.73a 3.38±1.13a 5.20±0.21 3.14±0.17a 3.36±0.75a 5.02±0.49 T1 2.44±0.46 3.17±0.29 2.10±1.09 2.74±0.13 3.04±0.35 2.30±1.24 2.37±0.79a 3.05±0.41a 0.73±0.31 Bacteria T2 2.66±0.58a 2.58±0.57 4.64±1.61 3.59±0.38a 3.84±1.03a 4.95±0.79 3.44±1.15a 3.08±0.50a 5.44±0.50 T3 3.09±0.37a 3.80±0.69 5.46±0.82 3.41±1.19a 3.87±0.65a 5.54±0.50 3.20±0.92a 2.83±0.54a 5.24±0.48 T4 3.01±0.66a 3.34±0.42a 5.55±0.37 2.79±0.31a 4.33±0.59a 4.80±0.42 3.05±0.18a 2.86±0.26a 5.44±0.35 Bacteria T2 2.48±0.36 2.88±0.11a 1.37±0.66 2.48±0.15 2.97±0.11 2.34±0.84 2.24±0.61a 3.10±0.27a 0.93±0.26 T4 T3 2.77±0.34 2.66±0.31 2.97±0.17 3.22±0.17 3.85±1.20 3.13±1.27 2.00±0.17a 2.32±0.39 2.80±0.44 2.99±0.34 4.06±0.50 3.39±1.62 2.34±0.49a 2.70±0.19a 2.74±0.50a 1.64±0.70b 1.02±0.52 1.37±0.40 Table 3-5. Mean estimated Shannon diversity ± standard deviation. Superscripts indicate significant differences within a growth stage and management style, after Tukey's HSD. Samplec Fungi T2 Maize asignificantly different from roots (P<0.05). bsignificantly different from leaves (P<0.001). cV.=vegetative, F.=flowering, S.=seed development, L=leaf, S=stem, R=root 52 NMDS plots generated from Bray-Curtis distances were used to visualize beta diversity of microbial communities. PERMANOVA was used to determine if centroids of clusters were significantly different. Analysis of homogeneity of variances was used to determine if differences among centroids may be due to high variances across clusters. Global analysis revealed clusters for each crop species and segregation was clear for aboveground and belowground samples in fungi (Figure 3-15A). At the vegetative growth stage, wheat bacterial communities formed a clear cluster, while maize and soybean communities shared a similar pattern; however, this pattern was lost in the later growth stages (Figure 3-15B). Within maize (Figure 3-16A-B) and soybean (Figure 3-16C-B), there was clustering of fungal and bacterial communities by plant organs. Centroids were significant, but variances were also significantly different (P < 0.001, Table 3-6). Within each growth stage, there was not clear clustering of fungal and bacterial communities by management strategy; although, root samples formed clusters separate from aboveground organs (Figure 3-17). Management had significant centroids (P < 0.01); significantly different variances were observed for fungal communities (P < 0.05) but were not significant for bacterial communities (Table 3-6). 53 Table 3-6. PERMANOVA of the communities presented in Figures 3-15, 3-16, and 3-17, considering all factors and their interactions. All Crops Factora Crop DoF, Res 2, 509 Organ Management Growth Stage Crop:Organ Crop:Management Organ:Management Crop:Growth Stage Organ:Growth Stage Management:Growth Stage Crop:Organ: Management Crop:Organ: Growth Stage Crop:Management: Growth Stage Organ:Management: Growth Stage Crop:Organ: Management: Growth Stage Global Fungi F 114.24 9 R2 0.176 63.804 0.098 8.561 0.020 19.809 0.030 22.692 0.070 5.596 0.026 2.599 0.012 13.611 0.042 5.008 0.015 2.502 0.012 P 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 2, 509 3, 509 2, 509 4, 509 6, 509 6, 509 4, 509 4, 509 6, 509 Bacteria DoF, Res R2 2, 499 86.883 0.146 F 2, 499 55.909 0.094 3, 499 2.814 0.007 2, 499 14.659 0.025 4, 499 29.333 0.098 6, 499 2.524 0.013 6, 499 2.105 0.011 4, 499 13.943 0.047 4, 499 4.707 0.016 2.103 0.011 6, 499 P 0.001 *** 0.001 *** 0.001 *** 0.001 *** 0.001 *** 0.001 *** 0.001 *** 0.001 *** 0.001 *** 0.001 *** 12, 499 1.970 0.020 0.001 *** 8, 499 4.911 0.033 0.001 *** 12, 499 1.797 0.018 0.001 *** *** *** *** *** *** *** *** *** *** *** *** *** *** 12, 509 2.356 0.022 0.001 8, 509 4.707 0.029 0.001 12, 509 2.174 0.020 0.001 12, 509 1.356 0.013 0.006 ** 12, 499 1.530 0.015 0.001 *** 22, 509 1.463 0.025 0.001 *** 24, 499 1.456 0.029 0.001 *** 54 Table 3-6 (cont’d) Maize Factora Growth Stage Organ Mgtb Organ:Mgt Growth Stage:Organ Growth Stage:Mgt Growth Stage:Organ:Mgt Growth Stage Mgt Growth Stage:Mgt Growth Stage Mgt Growth Stage:Mgt Growth Stage Mgt Growth Stage:Mgt Growth Stage Organ Growth Stage:Organ Growth Stage Organ Growth Stage:Organ DoF, Res 2, 162 2, 162 3, 162 6, 162 4, 162 6, 162 12, 162 2, 55 3, 55 6, 55 2, 52 6, 52 2, 55 3, 55 6, 55 2, 45 2, 45 4, 45 2, 45 2, 45 4, 45 Global Leaf Stem Roots T1 T2 Fungi F R2 P ** 10.940 0.082 0.001 21.514 0.161 0.001 *** 2.613 0.035 0.001 2.312 0.039 0.001 3.425 0.051 0.001 1.775 0.048 0.002 1.574 0.052 0.002 20.524 0.307 0.001 *** 5.969 0.134 0.001 3.308 0.148 0.001 7.978 0.197 0.001 *** 1.467 0.109 0.011 1.923 0.050 0.019 * 2.967 0.116 0.001 1.453 0.114 0.016 * 5.505 0.121 0.001 ** 12.775 0.282 0.001 2.290 0.101 0.002 3.575 0.084 0.002 12.786 0.300 0.001 *** 1.877 0.088 0.003 55 DoF, Res 2, 161 2, 161 3, 161 6, 161 4, 161 2, 53 3, 53 6, 53 2, 54 3, 54 2, 54 3, 54 6, 54 2, 44 2, 44 2, 44 2, 44 4, 44 Bacteria F R2 P 3.429 0.027 0.001 25.316 0.199 0.001 1.766 0.026 0.004 1.769 0.039 0.001 2.270 0.036 0.001 3.457 0.091 0.001 2.224 0.088 0.001 1.579 0.125 0.001 2.832 0.081 0.001 1.558 0.067 0.003 2.017 0.056 0.001 2.225 0.093 0.001 1.183 0.099 0.021 1.929 0.057 0.008 7.786 0.228 0.001 1.942 0.055 0.005 8.056 0.229 0.001 1.557 0.089 0.005 *** ** *** * *** Table 3-6 (cont’d) T3 Growth Stage Organ Growth Stage:Organ Growth Stage Organ Growth Stage:Organ Factora Growth Stage Organ Mgt Organ:Mgt Growth Stage:Organ Growth Stage:Mgt Growth Stage:Organ:Mgt Growth Stage Mgt Growth Stage:Mgt Growth Stage Mgt Growth Stage:Mgt Growth Stage Mgt Growth Stage:Mgt T4 Global Leaf Stem Roots 2, 34 2, 34 4, 34 2, 38 2, 38 4, 38 6.717 0.151 0.001 15.603 0.351 0.001 2.557 0.115 0.003 5.269 0.126 0.001 12.330 0.295 0.001 2.566 0.123 0.001 2, 35 2, 35 2, 38 2, 28 4, 38 Soybean Fungi F R2 DoF, Res P 2, 173 8.889 0.061 0.001 32.577 0.222 0.001 *** 2, 173 ** 3, 173 8.115 0.101 0.001 5.059 0.060 0.001 6, 173 4, 173 4.070 0.056 0.001 1.805 0.045 0.001 2.088 0.041 0.001 12, 173 DoF, Res 2, 169 2, 169 3, 169 6, 169 4, 169 6, 169 10, 169 2, 60 3, 60 6, 60 2, 55 3, 55 5, 55 2, 54 3, 54 5, 54 * 28.047 0.319 0.001 12.998 0.222 0.001 3.451 0.118 0.001 9.723 0.154 0.001 14.542 0.346 0.001 1.585 0.063 0.022 6.436 0.145 0.001 *** 3.513 0.119 0.001 2.261 0.127 0.001 2, 56 3, 56 2, 58 3, 58 6, 58 2, 58 3, 58 6, 58 56 1.714 0.060 0.021 6.895 0.241 0.001 2.295 0.064 0.005 10.632 0.298 0.001 1.856 0.104 0.004 Bacteria F R2 P 3.429 0.027 0.001 25.316 0.199 0.001 1.766 0.026 0.005 2.835 0.039 0.002 2.270 0.036 0.001 1.992 0.055 0.001 7.866 0.188 0.001 1.757 0.063 0.006 8.439 0.180 0.001 1.644 0.053 0.004 2.288 0.147 0.001 6.896 0.124 0.001 7.536 0.203 0.001 2.837 0.153 0.001 * *** *** * *** *** Table 3-6 (cont’d) T1 T2 T3 T4 Growth Stage Organ Growth Stage:Organ Growth Stage Organ Growth Stage:Organ Growth Stage Organ Growth Stage:Organ Growth Stage Organ Growth Stage:Organ 2, 45 2, 45 4, 45 2, 45 2, 45 4, 45 2, 44 2, 44 4, 44 2, 35 2, 35 6.207 0.111 0.001 20.694 0.369 0.001 *** 3.349 0.119 0.001 * 8.095 0.128 0.001 23.571 0.373 0.001 4.529 0.143 0.001 8.908 0.127 0.001 28.901 0.412 0.001 *** 5.176 0.148 0.001 5.295 0.124 0.001 18.135 0.425 0.001 2, 44 2, 44 4, 44 2, 43 2, 43 4, 43 2, 45 2, 45 4, 45 2, 41 2, 41 4, 41 4.428 0.074 0.004 28.616 0.481 0.001 2.221 0.075 0.013 3.969 0.064 0.003 31.774 0.514 0.001 2.287 0.074 0.012 3.654 0.076 0.001 17.464 0.365 0.001 2.112 0.088 0.009 6.343 0.130 0.001 12.638 0.260 0.001 4.589 0.189 0.001 *** *** ** aOnly values significant at * (P ≤ 0.10), ** (P ≤ 0.05), *** (P ≤ 0.01) shown. bMgt = management 57 Figure 3-15. Effect of growth stage on beta diversity of fungal (A) and bacterial (B) communities on all crops. Points colored by crop species, shapes indicate plant organ. Non-metric multidimensional scaling (NMDS) calculated by Bray-Curtis distance. Statistical support is detailed in Table 3-6. 58 Figure 3-16. Influence of management strategies on beta diversity of maize fungal (A), maize bacterial (B), soybean fungal (C), soybean bacterial (D) communities originating from each plant organ. NMDS calculated by Bray-Curtis distance. Statistical support is detailed in Table 3-6. 59 Figure 3-17. Effect of growth stage on beta diversity of maize fungal (A), maize bacterial (B), soybean fungal (C), and soybean bacterial (D) communities originating from each management style. NMDS calculated by Bray-Curtis distance. Statistical support is detailed in Table 3-6. 60 Microbial Networks Microbial hubs, species that influence the presence of others in an environment, were defined as highly connected nodes with ten or more connections and 35 hubs were discerned across all crops (Table 3-7). Of the hub taxa, 12 bacterial hubs and no fungal hubs were shared between wheat roots and phyllosphere (Figure 3-18, 3-19). All hub taxa in wheat roots had at least one negative correlation with a non-hub taxon. The two hubs Dongia sp. and Cytophaga sp. were negatively correlated with each other in wheat roots (Figure 3-18). Chalatospora sp. was the only hub taxon whose presence was not correlated with the other hub taxa in wheat roots (Figure 3-18). Nine of the hub taxa in the wheat phyllosphere were correlated with other hub taxa, and Pedobacter sp. co-occurred with seven of them (Figure 3-19). Of the wheat phyllosphere hub species, Sedimentibacter sp., Mucilaginibacter sp., and Ferruginibacter sp. did not have any negative correlations (Figure 3-19). The hub taxon in maize roots, Pseudoduganella sp., was predominantly correlated with other bacterial species, but negatively influenced the presence of two fungal taxa; Brachyphoris sp. and Rhexocercosporidium sp. (Figure 3-20A). Glomeromycota, a group of fungi that form mycorrhizal symbioses with plants, formed a small isolated network not connected with the main maize root network (Figure 3-20A). Hub taxa in the maize phyllosphere were all identified as fungi except one, and no negative correlations containing hubs were identified (Figure 3-21). Devosia sp. was identified as a hub taxon in both wheat and maize phyllospheres. We did not identify hub taxa that were shared across maize or soybean organs. Only one hub taxon was observed in each soybean root (Stagonospora pseudovitensis; Figure 3-20B) and phyllosphere (Bensingtonia subrosea; Figure 3-22) networks. Interestingly, the hub taxa in soybean roots was positively correlated with fungi and was negatively correlated with bacteria (Figure 3-20B). Closely related 61 taxa, such as several nodule-forming bacterial OTUs (Bradyrhizobium, Mesorhizobium, Neorhizobium) frequently occurred together in the networks. Although not identified as a hub, Fusarium sp. was present in the co-occurrence networks. The ITS barcode marker cannot be used to definitively identify Fusarium sp. down to genus, we do not know if the Fusarium OTUs recovered by microbiome sequencing are plant pathogens. Fusarium and/or Gibberella OTUs were found in all networks except maize roots. All bacterial hub taxa, except OTU 148, were shared across all crops. OTU 148 was not found in soybean. Twenty-one OTUs were identified as hub taxa on wheat; seven of these shared hub taxa status across the root and phyllosphere networks. All 21 OTUs can be found in leaves, stems, and roots of wheat. This phenomenon was observed for maize and soybean networks as well. Although hub taxa may not share hub status across crops or plant organs, these taxa are common OTUs found across plant organs. 62 Table 3-7. Hub taxa of KBS LTER crops. Crop Root Chalastospora ellipsoideaa Bradyrhizobium sp. Cytophaga sp. Dongia sp. Dyadobacter sp. Ferruginibacter sp. Massilia sp. Pedobacter sp. Polaromonas sp. Ralstonia sp. Salinibacterium sp. Sedimentibacter sp. Sphingomonas sp. Variovorax sp. Pseudoduganella sp. Phyllosphere Devosia sp. Discula destructivaa Dyadobacter sp. Ferruginibacter sp. Flavobacterium sp. Methylophilus sp. Mucilaginibacter sp. Pedobacter sp. Polaromonas sp. Rhizobium sp. Rhodopseudomonas sp. Salinibacterium sp. Sedimentibacter sp. Schizophyllum communea Discula destructivaa Neofavolus alveolarisa Ophiognomonia sogonoviia Devosia sp. Wheat Maize Soybean Stagonospora pseudovitensisa Bensingtonia subroseaa afungal OTU 63 Figure 3-18. Co-occurrence network of taxa in wheat roots. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. 64 Figure 3-19. Co-occurrence network of taxa in wheat phyllosphere. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. 65 Figure 3-20. Co-occurrence network of taxa in maize (A) and soybean (B) roots. Bacterial operational taxonomic units (OTUs) are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. 66 Figure 3-21. Co-occurrence network of taxa in maize phyllosphere. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. 67 Figure 3-22. Co-occurrence network of taxa in soybean phyllosphere. Bacterial OTUs are represented by diamond shapes, fungal OTUs are represented by circles, and nodes are colored by class. Solid lines indicate positive correlation, and dashed lines indicate a negative correlation between OTUs. Node size indicates degree of connectivity. 68 Discussion Here we present the first in-depth comparative analysis of fungal and bacterial microbial communities in the economically important row crops, wheat, maize, and soybean. The study demonstrates that microbial communities differ across three growth stages and organs of one plant host, and between plant hosts. Although the plant hosts studied here share some microbial taxa, the communities are distinct between each crop species, as determined by core taxa analysis and diversity metrics. Contrary to expectations, no large differences in plant microbial diversity among land management strategies were observed. Previous studies that explored differences in microbial communities among management strategies (such as conventional vs. organic management) analyzed rhizospheric soil communities or conducted comparisons of fields that were large distances apart (for example Bernard et al. 2012, Peiffer et al. 2013, Hartmann et al. 2014, Souza et al. 2015). Studies which surveyed plant microbiomes of the same genotypes across multiple locations found that geography and environment have a stronger influence on microbial community than management and plant genotype (Finkel et al. 2011, Peiffer et al. 2013, Copeland et al. 2015, Chen et al. 2016). The microbial reservoir of the soil may be the biggest influencing factor on the final microbial community composition. Analyses of diverse plant taxa showed that plant species and genotype (cultivar) influenced root microbial communities. Based on previous studies and the present study, land management affects soil and rhizosphere microbial community composition but appears to have a weaker impact on plant or phyllosphere microbial community composition (Bálint et al. 2013, Bonito et al. 2014, Coleman-Derr et al. 2015, Sapkota et al. 2015, Naylor et al. 2017). 69 The present study did not use fungicides directly on the crops, and T1-T3 managements used fungicide-treated seed. We did not have controls for these processes to assess the direct effect of fungicides on the microbiome. Previous studies have demonstrated that fungicide seed treatments affect phyllosphere microbiomes of crops tested during the vegetative growth phase (Nettles et al. 2016). It is not known if this observation would persist to reproductive growth stages. Fungicides have also been known to affect mycorrhizal associations and reproduction (Marx and Rowan 1981). However, Plante (2017) argues that crop management strategies, such as applications of fertilizers and pesticides, are not perturbation events in the traditional ecological sense that should alter the microbiota found on plants. Plante implies the application of pesticides may be a chronic stress for the native microbiota instead of a perturbation event (2017). However, the data presented in the current study indicates that fungicides may be secondary influencers of microbial community composition, with geography and genotype as the primary influencing factors. The microbial community composition of the wheat, maize, and soybean growth at our study site is distinct for each host crop and plant organ. Some of the most abundant OTUs found on wheat, maize, and soybean (Dothideomycetes, Sordariomycetes, Alphaproteobacteria, Gammaproteobacteria) are also found as some of the most abundant taxa in previous studies (Rascovan et al. 2016, Gdanetz and Trail 2017, Karlsson et al. 2017). Studies that analyze microbial communities across multiple organs or tissues on the same plant often show that there are distinct communities across these niches (for example, Ottesen et al. 2013, Coleman-Derr et al. 2015). Plant-microbe interactions can influence a plant’s susceptibility or resistance to pathogens by stimulating plant defense responses (reviewed by Durrant and Dong 2004). There is a bias in the literature toward analyzing soil or rhizosphere microbial communities (Table 1-1). 70 Work on model systems shows that phyllosphere microbes can influence disease progression and nutrient cycling (Delmotte et al. 2009, Vorholt 2012, Agler et al. 2016, Ryffel et al. 2016, Vogel et al. 2016). We demonstrated that root-associated microbial communities are often distinct from those of the phyllosphere or aboveground organs. Although the rhizosphere has a higher abundance and diversity of plant-associated microbes compared to the phyllosphere (Ottesen et al. 2013, Coleman-Derr et al. 2015, Zarraonaindia et al. 2015), microbes inhabiting other plant surfaces may also be important to plant health. Neglecting aboveground plant surfaces, in favor of the soil and rhizosphere, during microbial community analyses will exclude microbes that have a life cycle which is not captured by a soil or root-associated phase, for example, important plant pathogens such as rust fungi. Our findings indicate that maize and soybean fungal diversity decreases across the growing season while bacterial diversity remains constant. Previous studies examining changes in plant- associated microbial communities over the course of a growing season have been inconclusive. Some studies showed increasing community diversity over time (Shade et al. 2013, Gdanetz and Trail, 2017) whereas others showed a decrease (Copeland et al. 2015, de Souza et al. 2016), or no significant change (Sapkota et al. 2017). The diversity patterns of the soybean fungal communities we examined reinforced the temporal patterns observed on maize from the KBS LTER study site, and in previous studies of soybean from other field sites (Copeland et al. 2015). Further studies are needed to examine the plant microbial community dynamics over time to determine if the observed differences throughout the growing season are driven by host plant or environmental factors. Some aspects of soil diversity, specifically the alpha-diversity in this study, appear to vary temporally across the seasons, but this may also be an artifact of seasonal changes such as moisture (Lauber et al. 2013). Further investigation should be conducted into the degree to which random 71 colonization by microbes affects the end community composition. Computer modelling of the plant microbiome suggests the initial colonizers influence final community structure (Evans et al. 2016). An understanding of the influence of the initial colonizers on final community composition will have implications for community manipulation in field crops and synthetic environments such as greenhouses or hydroponic systems. Community structure is dictated by environmental microbial diversity. However, there are certain members of microbial communities, hub taxa, which can strongly influence community composition. In the current study, hub taxa identified in the plant organ(s) of one crop did not share hub taxa status across all crops and plant organ(s). However, these OTUs are commonly occurring and are found across all crops and conditions. These organisms are targets for future studies with synthetic microbial communities. Most of the wheat microbial hubs identified in this study are bacterial OTUs. This may be due to low sequencing coverage from maize and soybean 16S rRNA gene libraries. Fungal hub taxa were also identified for all crops, with many fungal hubs in the maize phyllosphere, indicating that fungi are important members of plant microbial communities. Some of the fungal hub species identified in the current study are pathogens of other plant species, for example, Discula destructiva a hub species of both the wheat and maize phyllospheres, causes dogwood anthracnose. The presence of a non-host pathogen hub may prevent the infection of another pathogen, as demonstrated by Agler et al. (2016). Fusarium graminearum disease incidence was low during the study years, this may be due to the presence of hub taxa, specifically D. destructiva. However, some of the OTUs identified as hub species in the microbial co- occurrence networks may be assigned incorrect taxonomies due to incomplete reference databases, instead they may be closely related species. As observed in this study and as described by Berry 72 and Widder (2014), closely related OTUs such as Rhizobia, wood rot fungi, and mycorrhizal fungi frequently co-occur in network analysis. Taxonomic assignment of OTUs to species-level resolution is possible with the ITS region of the rRNA gene in some fungal lineages, but it is not possible for OTUs identified by the 16S rRNA gene. Further work on development and testing of network analysis methods (as discussed by Poudel et al. 2016, Schlatter et al. 2017), followed by experimental validation of microbial communities is needed. Microbial hub taxa are important members of the plant-associated microbiota that can influence final community composition and plant health (Agler et al. 2016). Identification of hub taxa and understanding their influence on final community composition will be important for growing plants in synthetic environments such as hydroponic systems or on the International Space Station (Foster et al. 2014). For the interest of microbial manipulation in agriculture, future work should determine if there is a degree of phylogenetic or functional conservation in microbial hub species. If function proved to be a more important characteristic, this may make selection and development of microbial cocktails to improve crop yields or reduce plant diseases, easier to develop. Analyses such as the hub taxa identified here will be a valuable resource for researchers wishing to experimentally manipulate these microbial communities. Future work in our lab will test the hub taxa identified here in synthetic community experimental manipulations with crop plants for protective outcomes. This study shows that multiple organs (root, leaf, and stem) on the same plant often have distinct microbial communities. The microbial reservoir of the soil and land management strategy are factors which can influence the plant-associated microbial communities; however, plants are still able to select for specific microbes. This selection may be in the form of root exudates that recruit microbes (Wu et al. 2015, Biere and Goverse 2016), or leaf chemistry that is hostile toward 73 germination of spores which land on a leaf surface. Detailed understanding of the composition of plant-associated microbial communities will allow growers to add in the influence of the microbiota when calculating pesticide or fertilizer applications in precision agriculture approaches (Tackenberg et al. 2016, Pineda et al. 2017). As we demonstrated here with inferred microbial networks, and as others have shown previously (Agler et al. 2016, Mark Welch et al. 2016), microorganisms interact with each other across Kingdoms. To better emulate natural microbial community composition, controlled experiments with synthetic plant microbial communities should include microbes from across the tree of life. Although not investigated in the current study Archaea, Protists, Oomycetes, and viruses can interact with other members of microbial communities – exciting research opportunities lie ahead of those attempting to understand these complex microbial communities composed of organisms from multiple Kingdoms across the tree of life (Anonymous 2017). We also want to state the value of culture-based work in this era of computational and “-omics” biology; researchers generating culture collections from the samples or field sites at which they conduct microbiome studies are building an important resource useful for experimental testing of the hypotheses or questions generated by microbial ecology work. 74 CHAPTER 4 FUNGAL MICROBIOMES: A STRATEGY FOR IMPROVED TAXONOMIC RESOLUTION OF ENVIRONMENTAL ITS SEQUENCES Abstract One of the most crucial steps in high-throughput sequence-based microbiome studies is the taxonomic assignment of sequences belonging to operational taxonomic units (OTUs). Without taxonomic classification, functional and biological information of microbial communities cannot be inferred or interpreted. The internal transcribed spacer (ITS) region of the ribosomal DNA is the conventional marker region for fungal community studies. While bioinformatics pipelines that cluster reads into OTUs have received much attention in the literature, less attention has been given to the taxonomic classification of these sequences, upon which biological inference is dependent. Here we compare how three common fungal OTU taxonomic assignment tools (RDP Classifier, UTAX, and SINTAX) handle ITS fungal sequence data. The classification power, defined as the proportion of assigned OTUs at a given taxonomic rank, varied among the classifiers. Classifiers were generally consistent (assignment of the same taxonomy to a given OTU) across datasets and ranks; a small number of OTUs were assigned unique classifications across programs. We developed CONSTAX (CONSensus TAXonomy), a Python tool that compares taxonomic classifications of the three programs and merges them into an improved consensus taxonomy. This tool also produces summary classification outputs that are useful for downstream analyses. Our results demonstrate that independent taxonomy assignment tools classify unique members of the 75 fungal community, and greater classification power is realized by generating consensus taxonomy of available classifiers with CONSTAX. Source For the full text of this work see: Gdanetz K, Benucci GMN, Vande Pol N, Bonito G. CONSTAX: a tool for improved taxonomic resolution of environmental fungal ITS sequences. BMC Bioinformatics. 18:538. https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-017-1952-x 76 APPENDICES 77 APPENDIX A EVALUATION OF PROTECTIVE ENDOPHYTES AGAINST FUSARIUM GRAMINEARUM HEAD BLIGHT 78 Introduction Fusarium Head Blight of cereals is a complex of diseases caused by F. graminearum (reviewed by Trail 2009, Xu and Nicholson 2009, Wegulo et al. 2015), and a consortium of species (Fusarium culmorum, Fusarium avenaceum, Fusarium poae, Microdochium majus, and Microdochium nivale). Multiple fungal species can be isolated from a single head (Ioos et al. 2005, Xu et al. 2005, Xu and Nicholson 2009). While wheat, barley, oats, and rye are all susceptible crops, Head Blight is best characterized in wheat and barley. Not only does this disease cause yield losses, but it also degrades the quality of harvested and stored grains resulting in the accumulation of mycotoxins (Brown and Proctor 2013). Currently, there are no completely resistant varieties of wheat or barley available. Partially resistant varieties in combination with foliar fungicides, is the most effective plant protection strategy against F. graminearum infection (Wegulo et al. 2015). Due to changing social pressures and the growing risk of fungicide resistance, there is an increasing interest in the manipulation of microbial communities to increase agricultural output and reduce disease (Chaparro et al. 2012, APS 2016, Busby et al. 2017, Schlatter et al. 2017). Manipulation of microbes associated with crops to improve agricultural outcomes is not a new concept. Scientists have known that microbes, such as rhizobia and mycorrhizae, are beneficial to plant growth for over a century (Young and Haukka 1996, Berch et al. 2004), and research on beneficial microorganisms has resulted in species and strains of microbes that have shown efficacy in productivity improvement or disease protection. For example, Alfa-Guard® to control Aspergillus infection of peanuts (Fravel 2005, Dorner and Lamb 2006), and Trichoderma harzianum (RootShield®, Trichodex®, Binab T®) to control various infections in crops (Fravel 2005). Additionally, there are biocontrol bacterial strains such as Bacillus subtilis to control various soil borne fungal diseases, and Pseudomonas 79 spp. to control diseases on various plants ranging from ornamentals, turf grass, to tree fruits (Fravel 2005). More recently, scientists have been attempting to identify microbial consortia that provide general suppression of plant and soil pathogens. However, there are limited stories of success with this strategy (reviewed by Schlatter et al. 2017). Difficulties may be due to competition with native microbes in the soil, as in the case of a biocontrol applied via microbial seed coatings; or the harsh environmental conditions such as UV light on leaf surfaces and rapidly changing moisture conditions (Lindow and Brandl 2003), for a foliar or spray-based biocontrol. Use of endophytic microbes with protective abilities may make for more effective control products on the market as there would be less exposure to harsh environmental conditions such as light and water stress (Lindow and Brandl 2003, Gdanetz and Trail 2017). Fusarium graminearum causes both flower and crown infections (Trail 2009, Wang et al. 2015). Fungal endophytes with protective abilities against seedling infection were tested in Chapter 2; here the protective ability of these strains during Head Blight initiation was tested. Materials and Methods Strains and Growth Conditions All fungal cultures were maintained on agar medium under 35% glycerol at -80ºC. Endophyte strains were isolated from field-grown wheat plants as described previously (Gdanetz and Trail 2017). A Michigan wild-type strain of F. graminearum, PH-1 (NRRL #31084, FGSC #9075; Trail and Common 2000), was used as pathogen inoculum. Endophytic fungal strains tested for protective ability by Gdanetz and Trail (2017) were used in the current study. Identity of endophytes was determined via sequencing of the ITS rRNA gene after DNA amplification with the primer pair ITS1F and ITS4, as described previously (White et al. 1990, Bruns and Gardes 80 1993, Gdanetz and Trail 2017). Fungal strains were grown on malt extract agar medium (2% malt extract (Difco, Detroit, MI, USA), 1.5% agar) for four days prior to plant inoculation. A 4-mm diameter borer was used to generate agar plugs of each fungal culture. The uncolonized portion of the plug was removed and the colonized part of each plug was cut into quarters. Plant Inoculation and Disease Rating Wheat plants of the F. graminearum-susceptible cultivar Wheaton were grown in a greenhouse (21-23ºC with 16 hours of supplemental lighting). Plants were inoculated with protective endophytes and/or the pathogen during flowering (Table A-1). Forceps were used to place the colonized agar blocks into the middle floret of each plant (Guenther and Trail 2005, Hallen-Adams et al. 2011b). Agar plugs of endophyte cultures and fresh malt extract agar plugs (control plants) were inserted into florets, then plants were placed into a misting chamber (misting every 15 minutes) for 24 hours to stimulate fungal growth. Then F. graminearum colonized agar plugs were added to the same florets containing the endophyte-colonized plugs and to control plant florets (Figure A-1). Plants were returned to the misting chamber for an additional 48 hours. Each plant was rated for Head Blight symptoms (necrosis, premature bleaching, bent awns) for 15 days post inoculation. Disease incidence, the number of infected florets per plant, was compared between endophyte-inoculated plants and the endophyte-free controls. Statistics on mean disease incidence were calculated in R Studio (R Core Team 2016) and plots were created with the ‘ggplot2’ package (Wickham, 2009). 81 Table A-1. Protective endophyte strains used for head inoculations (Gdanetz and Trail 2017). Strain ID 11 30 34 36 37 38 40 44 45 51 57 59 70 88 ID by full length ITS locus Microdochium bolleyi Alternaria tenuissima Alternaria sp. Aspergillus niger Alternaria tenuissima Fusarium solani Fusarium sp. Fusarium sp. Penicillium reticulisporum Phoma sp. Phoma sp. Fusarium sp. Fusarium oxysporum Penicillium commune 82 Figure A-1. Inoculation method and experimental design of protective endophytes. 83 Figure A-2. Fusarium Head Blight disease incidence. Wheat inoculated with endophytes (No E.: endophyte-free control) and challenged with Fusarium graminearum. Error bars represent standard error of the mean (3 independent replicates of 8 plants). Strains marked with asterisk had significantly reduced disease compared to endophyte free control (p < 0.05). 84 Results and Discussion Endophyte strains that demonstrated protective ability against F. graminearum damping- off in a seedling assay (Gdanetz and Trail, 2017) were tested for protective ability against Head Blight infection. Endophyte strains that demonstrated protective ability in seedlings also demonstrated protective ability in wheat heads. Plants inoculated with strains 37 (Alternaria tenuissima), 40 (Fusarium sp.), and 70 (Fusarium oxysporum) had reduced disease when compared with endophyte-free plants (Figure A-2) in both mature plants and seedlings. Five additional strains showed protective ability in mature plants but not seedlings: 30 (Alternaria tenuissima), 34 (Alternaria sp.), 57 (Phoma sp.), 59 (Fusarium sp.), and 44 (Fusarium sp.). Strain 38 (Fusarium solani) was the only endophyte that caused reduced disease in seedlings, but not in mature plants. Changing consumer demands are creating a renewed interest in biocontrol organisms for agriculture. Several candidate strains that may be used as protective endophytes in wheat were identified (Gdanetz and Trail 2017), although, further work is required to identify the mechanism of action of these strains. It is not yet known if the fungi tested here protect by creating a physical barrier (colonizing and taking up space between plant cells which blocks F. graminearum growth); stimulating systemic or induced plant defenses which then indirectly prevents F. graminearum from colonizing the plant; or by directly reducing growth or pathogenicity of F. graminearum via hyphal parasitism or antibiotic production. Field implementation of a protective spray just prior to or during flowering (as when protective fungicides are applied to wheat) is possible, however the use of spray-based biocontrols is not ideal. Weather conditions could alter spray efficacy, and spray implementation generates 85 additional costs for growers (preparing solutions, and additional passes of equipment over the field). Most of the strains tested here were protective when inoculated directly into the floret. However, we do not yet know which protective mechanism, as described above, is used by these strains. Future work on the identification of the mechanism, in addition to mapping the spread of colonization through the plant (one or more plant organs) is required. Due to sensitivity of microbial strains to abiotic factors, and competition from other microbes, inoculations with protective microbial cocktails would increase chances of success. Siou et al. (2015) have demonstrated that competition between fungal species commonly found on infected heads can affect the disease outcomes and have reported that co-inoculation of heads with multiple strains does not increase mycotoxin disease risk (Siou et al. 2015). Future work with these strains should include seed inoculations with combinations of microbes (microbes that utilize different methods of protection) to increase likelihood of successful protection. Once successful combinations of protective strains have been established, field trials should be pursued. 86 FUNCTIONAL ANALYSIS OF KNOCKOUTS OF TERPENE SYNTHASE GENES IN FUSARIUM GRAMINEARUM APPENDIX B 87 Introduction The ascomycete fungus F. graminearum Schwabe causes Head Blight, also known as Head Scab, of cereals. Head Blight has caused devastating yield losses in the last several decades (McMullen et al. 1997, 2012). Each spring this soil-borne fungus fires ascospores from perithecia, which land on the plant and colonize from the head down through the stem. F. graminearum hyphae and perithecia-initials overwinter on crop debris remaining on the soil surface (Trail 2009). Due to soil conservation efforts, there has been an increase in the prevalence of no-till farming, however, this comes with a tradeoff – higher disease pressure from F. graminearum (McMullen et al. 2012). F. graminearum is well-known not only for causing yield losses in crops, but also because it can produce an array of secondary metabolites, including mycotoxins, that impact both the agriculture industry and human health. The genome of F. graminearum was sequenced in 2003 [www.broad.mit.edu/annotation/ genome/fusarium_graminearum/Home.html]. The first draft of the annotation encoded 51 secondary metabolite genes (Cuomo et al. 2007). Secondary metabolites are chemicals produced during cell growth that are not involved in essential cell functions such as glycolysis, respiration, or cell division. There are four classes of enzymes producing secondary metabolites commonly produced by the ascomycete fungi and synthesized by F. graminearum – non-ribosomal peptide synthases, polyketide synthases terpene synthases (TPS), and cytochrome P450s (reviewed by Tudzynski 2005). Some of the most notable secondary metabolites in F. graminearum are deoxynivalenol, nivalenol, and their derivatives. Deoxynivalenol and nivalenol are trichothecene mycotoxins, made by terpene synthases, that have been extensively studied and affect the health of humans and livestock (Bennett and Klich 2003, Rocha et al. 2005). Zearalenone, which is synthesized by a polyketide synthase, is an estrogen-mimic which also causes health issues for 88 livestock and humans (Miksicek 1994, reviewed by Desjardins and Proctor 2007), and has been shown to be toxic to some fungal species (Utermark and Karlovsky 2007). Aurofusarin is a red pigment produced by polyketide synthase enzymes that gives F. graminearum mycelium their characteristic pink color (Kroken et al. 2003). Deoxynivalenol and zearalenone have been shown to stimulate Alternaria mycotoxin production and fungal growth (Müller et al. 2014). Aspergillus fumigatus epipolythiodioxopiperazine secondary metabolites (disulfide-containing cyclic peptides) also have been shown to have anti-fungal activity (Patron et al. 2007, Coleman et al. 2011). We hypothesize that F. graminearum uses secondary metabolites to deter competitor microbes while overwintering on crop debris, as it has been demonstrated in other fungi that secondary metabolites can serve as antibiotic or self-defense molecules. Studies of F. graminearum growth while overwintering on crop debris would provide insight into how this fungus survives on dead plant tissue, a nutrient source which would be very attractive to, and under attack from saprotrophic microbes. An understanding of the cues that trigger production of F. graminearum secondary metabolites and their genetic regulation could provide insight into another possible control point in the life cycle of Head Blight. Materials and Methods Identification of Terpene Synthase Genes A recent re-annotation of the F. graminearum genome classified 17 previously unannotated genes as TPS (Sieber et al. 2014). Previously published Affychip (Affymetrix, Santa Clara, CA, USA) microarray data from F. graminearum colonized wheat plants and straw (Hallen et al. 2007, Hallen and Trail 2008, Guenther et al. 2009, Hallen-Adams et al. 2011b) was re-normalized and 89 analyzed using the ‘limma’ package in R (Ritchie et al. 2015, R Core Team 2016). Gene expression patterns across the 17 TPS genes were analyzed. Corresponding GeneChip (Affymetrix, Santa Clara, CA) probe IDs and gene IDs from the current and previous versions of the F. graminearum genome are available in Table A-2 (King et al. 2015). Table A-2. TPS gene IDs with greatest log-fold change in expression. FGDB Gene ID Current Gene ID FGSG_03066 FGSG_06784 FGSG_09381 FGSG_10933 FGSG_12186 FGRAMPH1_01G11973 FGRAMPH1_01G23179 FGRAMPH1_01G27225 FGRAMPH1_01G20835 FGRAMPH1_01G06541 Affy Chip ID fgd147-240_at fgd275-400_at fgd383-420_at fgd455-20_at fg02725_s_at, fgd136-30_at fgd418-70_at FGSG_17725 FGRAMPH1_01G07277 Homologous Gene Analysis Gene sequences were downloaded from the F. graminearum genome browser: Munich Information Center for Protein Sequences (MIPS) Fusarium graminearum Genome Database (FGDB) version 3.2 [http://mips.helmholtz-muenchen.de/genre/proj/FGDB/]. NCBI Basic Local Alignment Search Tool (BLAST) (Altschul et al. 1990) was used to identify homologous genes in other Fusarium spp. Protein alignments were created using MAFFT version 7.351b (Katoh et al. 2017) and statistics on the alignment were calculated using Geneious version 6 (Biomatters Inc., Newark, NJ). Phyre2 was used to identify structural and functional domains of the predicated proteins (Kelley et al. 2015). NCBI BLASTp was used to search for paralogs within the genome of F. graminearum. 90 Generation and Genetic Analysis of Transformants The wild-type F. graminearum strain PH-1 (NRRL #31084, FGSC #9075, Trail and Common 2000) was used for all genetic experiments. After confirmation of transformant strains, cultures were maintained as freezer stocks under 35% glycerol at -80°C. TPS Candidate genes were chosen based on log-fold changes in gene expression between the two infection stages analyzed. Primers used to generate and confirm fragments for putative transformants are listed in Table A-3. A split-marker strategy, which combined the flanking regions upstream and downstream of the target gene with one half of the hygromycin phosphatase (hph) gene (Carroll et al. 1994), was used to generate DNA constructs for deletion mutants of one of the six candidate genes (FGSC_12186). Homologous recombination, with hph as the selectable marker, was used to introduce the split-marker DNA fragments into F. graminearum protoplasts following the polyethylene glycol mediated transformation protocols of (Cavinder and Trail 2012). PCR screening by size shift was used for confirmation of putative transformants Phenotypes of two confirmed knockout strains of FGSC_12186 gene were determined as described below. To determine whether or not TPSs are involved in fungal self-defense, mutant cultures were challenged in vitro as described by Gdanetz and Trail (2017). Briefly, cultures were co-inoculated on malt extract agar medium plates and were scored into one of five interaction categories (Gdanetz and Trail 2017). Pathogenicity of mutants was assessed with wheat head infections; the middle floret was inoculated with 1 × 105 conidia ml-1, as described previously (Hallen-Adams et al. 2011a, 2011b). Pathogenicity was also tested in seedling infections as described by Baldwin et al. (2010) and modified by Gdanetz and Trail (2017). Briefly, wheat seedlings were grown in vermiculite containing F. graminearum inoculum. A subset of seedlings was inoculated with protective fungal endophytes; 37 (Alternaria tenuissima), 38 (Fusarium 91 solani), 40 (Fusarium sp.), and 70 (Fusarium oxysporum); as described in Gdanetz and Trail (2017). Perithecial development on wheat straw was induced by placing inoculated straw on the surface of moist sterile vermiculite, as described previously (Guenther et al. 2009). Perithecium formation on wheat straw was quantified as the percent nodes along the length of the stem with visible perithecia. In vitro perithecium production was induced on Carrot Agar medium (Klittich and Leslie 1988) by the addition of 900 ul of tween-60 as described previously (Trail and Common 2000). Quantity of perithecia generated on Carrot Agar medium were measured by counting the number of fruiting bodies on the surface of an agar plug taken with a 7-mm borer. Fruiting-body and ascospore development at 72, 96, and 120 hours post-tweening were measured as described previously (Cavinder et al. 2012). Conidia cultures were initiated with 1-ml of 1 × 105 conidia ml-1 into 100-ml of carboxymethyl cellulose medium (Cappellini and Peterson 1965). Conidia cultures were shaken at 225 rpm in a 25°C incubator, and conidia collected via filtering through three layers of sterile Miracloth (Millipore-Sigma, Burlington, MA, USA) before estimating concentration with a hemocytometer. Conidial production curves were calculated after measuring conidia concentrations at 48, 72, 96, and 120 hours. 92 PCR Product Length 909 588 1,867 1,217 966 Table A-3. Primers used to generate terpene synthase gene knockouts. Sequence 5' to 3' Primer 12186-L5 TCT TTT GGA TGT TTG GAT G 12186-L3 CGT CAG ATC GAT GGT AGT TGT CGT CGA CTC AAA GGC TGT GAA TTA TGT GTA G 12186-R5 ACA CTG GTG ACG GCT AAC CAG AAC TGT CAA ACA TAT TCG AGA TAT AAC CCC 12186-R3 GGG TGT GCT TCA TTT CAT Description aFP for flanking region upstream of 12186 bRP flanking region upstream of 12186 with 5' overhang for hph RP flanking region downstream of 12186 with 3' overhang for hph FP flanking region downstream of 12186 FP to amplify hph from pCB1004 HYG-F HYG-R YG-F AGTCGACGACAACTACCATCG ATCTGACGACGCCTGGTTGCTA CGCCTGAATAAGTG TGACAGTTCTGGTTAGCCGTCA CCAGTGTAACGCTGGTGAAAG TAAAAGATGCTGAAGAT GTATTGACCGATTCCTTGCGGT CCGAA RP to amplify hph from pCB1004 FP for hph overlapping region, pair with HYG-R HY-R CGATGTAGGAGGGCGTGGATA TGTCC aFP = forward primer bRP = reverse primer RP for hph overlapping region, pair with HYG-F 93 Results and Discussion Identification of Terpene Synthase Genes and Homologous Gene Analysis Six of the 17 putative TPS F. graminearum genes had increased levels of gene expression on straw, when compared to gene expression during plant infection (Figure A-3). Orthologs of F. graminearum TPS genes were identified throughout the Fusarium genus, and functional information was available for four of the six TPS orthologues in Fusarium fujikuroi (Homann et al. 1996, Mende et al. 1997, Linnemannstöns et al. 2002, Zhao et al. 2010). Nucleotide alignments of the TPS genes and their orthologs indicated they are conserved across Fusarium spp. (Table A- 4). Phyre2 analysis of F. graminearum TPS protein sequences and the corresponding F. fujikuroi homologs indicated 73% of the residues modelled at >90% confidence, indicating secondary and tertiary structures of F. graminearum and F. fujikuroi TPS are similar. Based on Phyre2 analysis, we hypothesize the TPS homologs in F. graminearum and F. fujikuroi could have similar functions. Functional and chemical analysis of F. graminearum gene products is needed to confirm this hypothesis. Table A-4. Summary of Fusarium spp. sequence alignments. Gene ID FGSG_03066 FGSG_06784 FGSG_09381 FGSG_10933 FGSG_12186 FGSG_17725 Num. in MSA Length (bp) 619 907 463 457 526 497 9 6 9 10 9 9 Identical sites (bp) Pairwise % Identity 83.6 83.3 91.6 95.3 76.4 91.1 331 280 366 396 243 325 94 Confirmation of Mutant Phenotypes Pathogenicity of one mutant strain, ∆12186.3B, during head infection was significantly reduced when compared with wild-type (Figure A-4A). Perithecium development on straw was not significantly different between wild-type and knockout strains. Perithecia were formed through stomata down to the second node in all strains (Figure A-4B). Numbers of in vitro perithecium produced were not significantly different across the strains (Figure A-4C). Disease incidence, during seedling infection or after challenging against protective endophytes, did not differ across wildtype and mutant strains (Figure A-5). Conidia production of the wild-type strain was higher than the production of both mutant strains (Figure A-6). The rates of production of conidia by both mutant strains were significantly reduced from the wild-type when measured at 72 hours (Figure A-6). The function of the F. graminearum TPS gene, FGSC_12186, does not appear to be involved in pathogenicity or perithecium development. However, it may be involved in conidial germination or growth, as the mutant strains ∆12186.2B and ∆12186.3B had reduced conidial production compared to wild-type F. graminearum. The rates of conidial production differed between the two mutant strains, therefore generation and phenotypic analysis of additional ∆12186 strains will confirm the reduced conidia production phenotype. Complementation of the FGSC_12186 gene in the mutant strains is also required to verify the conidial production phenotype. Future work should focus on chemical or metabolite analysis of the wild-type and the ∆12186 strains. Additionally, comparison of F. graminearum TPS gene products to the known products of F. fujikuroi TPS genes would confirm if the sequence homology and subsequent similar structural predictions from Phyre2 analysis are accurate. 95 Figure A-3. Relative gene expression of F. graminearum terpene synthase genes. Expression levels during perithecia development on straw (A), and during plant infection (B), DW and BW indicate barley cultivars. Expression data from Hallen et al. 2007, Hallen and Trail 2008, Guenther et al. 2009, Hallen-Adams et al. 2011b. 96 Figure A-4. Disease incidence of wild-type (PH-1) and independent knockout strains of FGSC_12186 during wheat head infection (A), distribution of perithecia production on wheat straw (B), and in vitro perithecia production (C). Bars in panel A represents the average of ten or more inoculated plants, repeated three independent times. Bars in panel B represent three sets of inoculate stems, repeated three independent times. Bars in panel C represent the number of perithecia on two cores per plate, six independent plates. Errors bars indicate standard deviation of the mean. Analysis of variance and Tukey’s honest significant difference were used to test significance; asterisk indicates significance (p < 0.001). 97 Figure A-5. Disease incidence of wild-type (PH1) and knockout strains during seedling infection. Subset of seedlings were also inoculated with best protective endophytes. Each bar represents ten plants repeated two independent times. 2B = ∆12186.2B; 3B = ∆12186.3B; 37, 38, 40, 70 are fungal endophyte strains from Gdanetz and Trail, 2017. 98 Figure A-6. Conidia growth curves of wild-type (PH-1) and knockout strains. Each point represents the average of three independent replicates, ± standard error of the mean. Analysis of variance and Tukey’s honest significant difference were used to test significance. PH-1 displayed significantly higher rate of conidia production than both knockout strains at 72 hours, and higher conidia production than ∆12186.2B at 96 hours (p < 0.05). Strain ∆12186.3B displayed significantly higher conidia production than PH-1 and ∆12186.2B at 120 hours (p < 0.05). Figure A-7. In vitro perithecium development. Squash mounts of F. graminearum perithecia at 120 hours post-induction of sexual development, showing mature asci of wild-type (A) and delayed development of knockout strain 12186.3B (B). 99 REFERENCES 100 REFERENCES Abanda-Nkpwatt D, Musch M, Tschiersch J, Boettner M, Schwab W. 2006. Molecular interaction between Methylobacterium extorquens and seedlings: growth promotion, methanol consumption, and localization of the methanol emission site. J Exp Bot. 57:4025–4032. Agler MT, Ruhe J, Kroll S, Morhenn C, Kim ST, Weigel D, Kemen EM. 2016. Microbial hub taxa link host and abiotic factors to plant microbiome variation. PLoS Biol.14:e1002352. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol. 215:403-10. Anonymous. Stop neglecting fungi. 2017. Stop neglecting fungi. Nat Microbiol. 2:17120. APS. Eds: Leach J, Beattie G, Eversole K. 2016. Phytobiomes: a roadmap for research and translation. 1–16. Arthur JC. 1891. Wheat scab. Indiana Agricultural Experiment Station Bulletin. 36:192-32 Bakker MG, Schlatter DC, Otto-Hanson L, Kinkel LL. 2014. Diffuse symbioses: roles of plant- plant, plant-microbe and microbe-microbe interactions in structuring the soil microbiome. Mol Ecol. 23:1571–1583. Baldwin TK, Gaffoor I, Antoniw J, Andries C, Guenther J, Urban M, Hallen-Adams HE, Pitkin J, Hammond-Kosack KE, Trai, F. 2010. A partial chromosomal deletion caused by random plasmid integration resulted in a reduced virulence phenotype in Fusarium graminearum. Mol Plant Microbe In. 23:1083–1096. Bálint M, Tiffin P, Hallström B, O’Hara RB, Olson MS, Fankhauser JD, Piepenbring M, Schmitt I. 2013. Host genotype shapes the foliar fungal microbiome of balsam poplar (Populus balsamifera). PLoS ONE. 8:e53987. Balint-Kurti P, Simmons SJ, Blum JE. 2010. Maize leaf epiphytic bacteria diversity patterns are genetically correlated with resistance to fungal pathogen infection. Mol Plant Microb In. 23:473– 484. Bazzicalupo AL, Bálint M, Schmitt I. 2013. Comparison of ITS1 and ITS2 rDNA in 454 sequencing of hyperdiverse fungal communities. Fungal Ecol. 6:102-9. Bennett JW, Klich M. 2003. Mycotoxins. Clin Microbiol Rev. 16:497–516. Berch SM, Massicotte HB, Tackaberry LE. 2004. Re-publication of a translation of “The vegetative organs of Monotropa hypopitys L.” published by F. Kamienski in 1882, with an update on Monotropa mycorrhizas. Mycorrhiza. 15:323–332. 101 Berg G, Rybakova D, Grube M, Köberl M. 2016. The plant microbiome explored: implications for experimental botany. J Exp Bot. 67:995–1002. Bernard E, Larkin RP, Tavantzis S, Erich MS, Alyokhin A, Sewell G, Lannan A, Gross SD. 2012. Compost, rapeseed rotation, and biocontrol agents significantly impact soil microbial communities in organic and conventional potato production systems. Appl Soil Ecol. 52:29–41. Berthrong ST, Buckley DH, Drinkwater LE. 2013. Agricultural management and labile carbon additions affect soil microbial community structure and interact with carbon and nitrogen cycling. Microb Ecol. 66:158–170. Biere A, Goverse A. 2016. Plant-mediated systemic interactions between pathogens, parasitic nematodes, and herbivores above- and belowground. Annu Rev Phytopathol. 54:499–527. Bhuvaneswari TV, Turgeon BG, Bauer WD. 1980. Early events in the infection of soybean (Glycine max L. Merr) by Rhizobium japonicum. Plant Pathol. 1027–1031. Blaalid R, Kumar S, Nilsson RH, Abarenkov K, Kirk PM, Kauserud H. 2013. ITS1 versus ITS2 as DNA metabarcodes for fungi. Mol Ecol Resour. 13:218-24. Bontio G, Reynolds H, Robeson MS, Nelson J, Hodkinson BP, Tuskan G, Schadt CS, Vilgalys R. 2014. Plant host and soil origin influence fungal and bacterial assemblages in the roots of woody plants. Mol Ecol. 23:3356-3370. Borrell AN, Shi Y, Gan Y, Bainard LD, Germida JJ, Hamel C. 2017. Fungal diversity associated with pulses and its influence on the subsequent wheat crop in the Canadian prairies. Plant Soil. 414:13–31. Bouffaud M-L, Poirier M-A, Muller D, Moënne-Loccoz Y. 2014. Root microbiome relates to plant host evolution in maize and other Poaceae. Environ Microbiol. 16:2804–2814. Brown DW, Proctor RH. 2013. Fusarium: Genomics, Molecular and Cellular Biology. Poole, UK: Horizon Scientific Press. Bruns TD, Gardes M. 1993. Molecular tools for the identification of ectomycorrhizal fungi — taxon-specific oligonucleotide probes for Suilloid fungi. Mol Ecol. 2:233–242. agro-ecosystems. Environ Microbiol. 5:441–452. Buckley DH, Schmidt TM. 2001. The structure of microbial communities in soil and the lasting impact of cultivation. Microb Ecol. 42:11–21. Buckley DH, Schmidt TM. 2003. Diversity and dynamics of microbial communities in soils from Burgess LW, Backhouse D, Summerell BA, Swan LJ. 2001. Crown rot of wheat. In: Summerell B, Leslie J, Backhouse D, Bryden W, Burgess L, eds. Fusarium: Paul E. Nelson Memorial Symposium. St. Paul, Minnesota: APS Press. p. 271–294. 102 Busby PE, Soman C, Wagner MR, Friesen ML, Kremer J, Bennett A, Morsy M, Eisen JA, Leach JE, Dangl JL. 2017. Research priorities for harnessing plant microbiomes in sustainable agriculture. PLoS Biol. 15:e2001793–14. Cappellini RA, Peterson JL. 1965. Macroconidium formation in submerged cultures by a non- sporulating strain of Gibberella zeae. Mycologia. 57:962–966. Carroll AM, Sweigard JA, Valent B. 1994. Improved vectors for selecting resistance to hygromycin. Fung Genet Newsl. 21:22. Cavinder B, Sikhakolli U, Fellows KM, Trail F. 2012. Sexual development and ascospore discharge in Fusarium graminearum. JoVE. e3895. Cavinder B, Trail F. 2012. Role of Fig1, a component of the low-affinity calcium uptake system, in growth and sexual development of filamentous fungi. Eukaryot Cell. 11:978–988. Chaparro JM, Sheflin AM, Manter DK, Vivanco JM. 2012. Manipulating the soil microbiome to increase soil health and plant fertility. Biol Fert Soils. 48:489–499. Chávez-Romero Y, Navarro-Noya YE, Reynoso-Martínez SC, Sarria-Guzmán Y, Govaerts B, Verhulst N, Dendooven L, Luna-Guido M. 2016. 16S metagenomics reveals changes in the soil bacterial community driven by soil organic C, N-fertilizer and tillage-crop residue management. Soil Till Res. 159:1–8. Chen W, Turkington TK, Lévesque CA, Bamforth JM, Patrick SK, Lewis CT, Chapados JT, Gaba D, Tittlemier SA, MacLeod A, Gräfenhan T. 2016. Geography and agronomical practices drive diversification of the epiphytic mycoflora associated with barley and its malt end product in western Canada. Agr Ecosyst Environ. 226:43–55. Cole JR, Wang Q, Fish JA, Chai B, McGarrell D M, Sun Y, Brown CT, Porras-Alfaro A, Kuske CR, Tiedje JM. 2013. Ribosomal Database Project: data and tools for high throughput rRNA analysis. Nucl Acids Res. 42:D633–D642. Coleman JJ, Ghosh S, Okoli I, Mylonakis E. 2011. Antifungal activity of microbial secondary metabolites. PLoS ONE. 6:e25321. Coleman-Derr D, Desgarennes D, Fonseca-Garcia C, Gross S, Clingenpeel S, Woyke T, North G, Visel A, Partida-Martinex LP, Tringe SG. 2015. Plant compartment and biogeography affect microbiome composition in cultivated and native Agave species. New Phytol. 209:798–811. Cook RJ. 1981. Fusarium diseases of wheat and other small grains in North America. In: Nelson PE, Tousson TA, Cook RJ, eds. Fusarium Diseases, Biology and Control. State College, PA: Pennsylvania State Univ. Press. p. 39—52. Copeland JK, Yuan L, Layeghifard M, Wang PW, Guttman DS. 2015. Seasonal community succession of the phyllosphere microbiome. Mol Plant Microbe In. 28:274–285. 103 Cuomo CA, Guldener U, Xu JR, Trail F, Turgeon BG, Di Pietro A, Walton JD, Ma LJ, Baker SE, Rep M, Adam G, Antoniw J, Baldwin T, Calvo S, Chang YL, DeCaprio D, Gale LR, Gnerre S, Goswami RS, Hammond-Kosack K, Harris LJ, Hilburn K, Kennell JC, Kroken S, Magnuson JK, Mannhaupt G, Mauceli E, Mewes HW, Mitterbauer R, et al. 2007. The Fusarium graminearum genome reveals a link between localized polymorphism and pathogen specialization. Science. 317:1400–1402. Delmotte N, Knief C, Chaffron S, Innerebner G, Roschitzki B, Schlapbach R, Mering von C, Vorholt JA. 2009. Community proteogenomics reveals insights into the physiology of phyllosphere bacteria. Proc Nat Acad Sci. 106:16428–16433. Dentinger BT, Didukh MY, Moncalvo JM. 2011. Comparing COI and ITS as DNA barcode markers for mushrooms and allies (Agaricomycotina). PLoS ONE. 6:e25081. Deshpande V, Wang Q, Greenfield P, Charleston M, Porras-Alfaro A, Kuske CR, Cole JR, Midgley DJ, Tran-Dinh N. 2016. Fungal identification using a Bayesian classifier and the Warcup training set of internal transcribed spacer sequences. Mycologia. 108:1-5. Desjardins AE, Proctor RH. 2007. Molecular biology of Fusarium mycotoxins. Int J Food Microbiol. 119:47–50. de Souza RSC, Okura VK, Armanhi JSL, Jorrín B, Lozano N, da Silva MJ, González-Guerrero M, de Araújo LM, Verza NC, Bagheri HC, Imperial J, Arruda,P. 2016. Unlocking the bacterial and fungal communities assemblages of sugarcane microbiome. Sci Rep. 6:1–15. Donn S, Kirkegaard JA, Perera G, Richardson AE, Watt M. 2014. Evolution of bacterial communities in the wheat crop rhizosphere. Environ Microbiol. 17:610–621. Dorner JW, Lamb MC. 2006. Development and commercial use of Afla-Guard®, an aflatoxin biocontrol agent. Mycotoxin Res. 22:33–38. Duncan DS, Jewell KA, Suen G, Jackson RD. 2016. Detection of short-term cropping system- induced changes to soil bacterial communities differs among four molecular characterization methods. Soil Biol Biochem. 96:160–168. Durrant WE, Dong X. 2004. Systemic acquired resistance. Annu Rev Phytopathol. 42:185–209. Dvorska JE, Surai PF, Speake BK, Sparks NHC. 2001. Effect of the mycotoxin aurofusarin on the antioxidant compositon and fatty acid profile of quail eggs. Brit Poultry Sci. 42:643–649. Edgar RC. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 26:2460-1. Edgar RC, Haas BJ, Clemente JC, Quince C, Knight, R. 2011. UCHIME improves sensitivity and speed of chimera detection. Bioinformatics. 27:2194–2200. 104 Edgar RC, Flyvbjerg H. 2015. Error filtering, pair assembly and error correction for next- generation sequencing reads. Bioinformatics. 31:3476–3482. Ellis JG. 2017. Can plant microbiome studies lead to effective biocontrol of plant diseases? Mol Plant Microbe In. 30:190–193. Essarioui A, LeBlanc N, Kistler HC, Kinkel LL. 2017. Plant community richness mediates inhibitory interactions and resource competition between Streptomyces and Fusarium populations in the rhizosphere. Microb Ecol. 74:157–167. Evans S, Martiny JB, Allison SD. 2016. Effects of dispersal and selection on stochastic assembly in microbial communities. ISME J. 1–10. Finkel OM, Burch AY, Lindow SE, Post AF, Belkin S. 2011. Geographical location determines the population structure in phyllosphere microbial communities of a salt-excreting desert tree. Appl Environ Microb. 77:7647–7655. Foster JM, Wheeler RM, Pamphile R. 2014. Host-microbe interactions in microgravity: assessment and implications. Life. 4:250–266. Fravel DR. 2005. Commercialization and implementation of biocontrol. Annu Rev Phytopathol. 43:337–359. Gdanetz K, Benucci GMN, Vande Pol N, Bonito G. 2017. CONSTAX: a tool for improved taxonomic resolution of environmental fungal ITS sequences. BMC Bioinformatics. 18:538. Gdanetz K, Trail F. 2017. The wheat microbiome under four management strategies, and potential for endophytes in disease protection. Phytobiomes. 1:158–168. Gosling P, Mead A, Proctor M, Hammond JP, Bending GD. 2013. Contrasting arbuscular mycorrhizal communities colonizing different host plants show a similar response to a soil phosphorus concentration gradient. New Phytol. 198:546–556. Granzow S, Kaiser K, Wemheuer B, Pfeiffer B, Daniel R, Vidal S, Wemheuer F. 2017. The effects of cropping regimes on fungal and bacterial communities of wheat and faba bean in a greenhouse pot experiment differ between plant species and compartment. Front Microbiol. 8:305–22. Guenther JC, Trail F. 2005. The development and differentiation of Gibberella zeae (anamorph: Fusarium graminearum) during colonization of wheat. Mycologia. 97:229–237. Guenther JC, Hallen-Adams HE, Bücking H, Shachar-Hill Y, Trail F. 2009. Triacylglyceride metabolism by Fusarium graminearum during colonization and sexual development on wheat. Mol Plant Microbe In. 22:1492–1503. 105 Hallen HE, Huebner M, Shiu S-H, Güldener U, Trail F. 2007. Gene expression shifts during perithecium development in Gibberella zeae (anamorph Fusarium graminearum), with particular emphasis on ion transport proteins. Fungal Genet Biol. 44:1146–1156. Hallen HE, Trail F. 2008. The L-type calcium ion channel cch1 affects ascospore discharge and mycelial growth in the filamentous fungus Gibberella zeae (anamorph Fusarium graminearum). Eukaryot Cell. 7:415–424. Hallen-Adams HE, Cavinder BL, Trail F. 2011a. Fusarium graminearum from expression analysis to functional assays. In: Xu JR, Bluhm BH, eds. Fungal genomics: methods and protocols. Totowa, NJ: Humana Press. pp. 79–101. Hallen-Adams HE, Wenner N, Kuldau GA, Trail F. 2011b. Deoxynivalenol biosynthesis-related gene expression during wheat kernel colonization by Fusarium graminearum. Phytopathology. 101:1091–1096. Hamid MI, Hussain M, Wu Y, Zhang X, Xiang M, Liu X. 2017. Successive soybean- monoculture cropping assembles rhizosphere microbial communities for the soil suppression of soybean cyst nematode. FEMS Microbiol Ecol. 93:1-10. Hartmann M, Frey B, Mayer J, der PMA, Widmer F. 2014. Distinct soil microbial diversity under long-term organic and conventional farming. ISME J. 9:1177–1194. Hibbett D, Abarenkov K, Kõljalg U, Öpik M, Chai B, Cole J, Wang Q, Crous P, Robert V, Helgason T, Herr JR. 2016. Sequence-based classification and identification of Fungi. Mycologia. 108:1049-68. Homann V, Mende K, Arntz C, Ilardi V, Macino G, Morelli G, Böse G, Tudzynski B. 1996. The isoprenoid pathway: cloning and characterization of fungal FPPS genes. Curr Genet. 30:232– 239. Huang, Y., Kuang, Z., Wang, W., and Cao, L. 2016. Exploring potential bacterial and fungal biocontrol agents transmitted from seeds to sprouts of wheat. Biol Control. 98:27–33. Hubbard M, Germida J, Vujanovic V. 2012. Fungal endophytes improve wheat seed germination under heat and drought stress. Botany. 90:137–149. Ioos R, Belhadj A, Menez M, Faure A. 2005. The effects of fungicides on Fusarium spp. and Microdochium nivale and their associated trichothecene mycotoxins in French naturally-infected cereal grains. Crop Prot. 24:894 –902. Jangid K, Williams MA, Franzluebbers AJ, Schmidt TM, Coleman DC, Whitman WB. 2011. Land-use history has a stronger impact on soil microbial community composition than aboveground vegetation and soil properties. Soil Biol Biochem. 43:2184–2193. 106 Johnston-Monje D, Lundberg DS, Lazarovits G, Reis VM, Raizada MN. 2016. Bacterial populations in juvenile maize rhizospheres originate from both seed and soil. Plant Soil. 405:337-355. Karlsson I, Friberg H, Kolseth A-K, Steinberg C, Persson P. 2017. Organic farming increases richness of fungal taxa in the wheat phyllosphere. Mol Ecol. 26:3424–3436. Katoh K, Rozewicki J, Yamada KD. 2017. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Brief Bioinform. 1–7. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJE. 2015. The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc. 10:845–858. Kelly BJ, Gross R, Bittinger K, Sherrill-Mix S, Lewis JD, Collman RG, Bushman FD, Li H. 2015. Power and sample-size estimation for microbiome studies using pairwise distances and PERMANOVA. Bioinformatics. 31:2461-2468. King R, Urban M, Hammond-Kosack MCU, Hassani-Pak K, Hammond-Kosack KE. 2015. The completed genome sequence of the pathogenic ascomycete fungus Fusarium graminearum. BMC Genomics. 1–21. Kinkel LL, Bakker MG, Schlatter DC. 2011. A coevolutionary framework for managing disease- suppressive soils. Annu Rev Phytopathol. 49:47–67. Kinkel LL, Schlatter DC, Bakker MG, Arenz BE. 2012. Streptomyces competition and co- evolution in relation to plant disease suppression. Research in Microbiologoy. 163:490–499. Klittich C, Leslie JF. 1988. Nitrate reduction mutants of Fusarium moniliforme (Gibberella fujikuroi). Genetics. 118:417–423. Koehler B. 1942. Natural mode of entrance of fungi into corn ears and some symptoms that indicate infection. J Agricultural Res. 64:421–442. Kõljalg U, Nilsson RH, Abarenkov K, Tedersoo L, Taylor AFS, Bahram M, Bates ST, Bruns TD, Bengtsson-Palme J, Callaghan TM, Douglas B, Drenkhan T, Eberhardt U, Duenas M, Grebenc T, Griffith GW, Hartmann M, Kirk PM, Kohout P, Larsson E, Lindahl BD, Luecking R, Martin MP, Matheny PB, Nguyen NH, Niskanen T, Oja J, Peay KG, Peintner U, Peterson M, Põldmaa K, Saag L, Saar I, Schüßler A, Scott JA, Senés C, Smith ME, Suija A, Taylor DL, Telleria MT, Weiss M, Larsson KH. 2013. Towards a unified paradigm for sequence-based identification of fungi. Mol Ecol. 22:5271–5277. Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. 2013. Development of a dual- index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Appl Environ Microb. 79:5112–5120. 107 Kroken S, Glass NL, Taylor JW, Yoder OC, Turgeon BG. 2003. Phylogenomic analysis of type I polyketide genes in pathogenic and saprobic ascomycetes. Proc Natl Acad Sci. 100:15670– 15675. Kurtz ZD, Müller CL, Miraldi ER, Littman DR, Blaser MJ, Bonneau RA. 2015. Sparse and compositionally robust inference of microbial ecological networks. PLoS Comp Biol. 11:e1004226–25. Kwaśna H, Bateman GL, Ward E. 2010. Microbiota in wheat roots evaluated by cloning of ITS1/2 rDNA and sequencing. J Phytopathol. 158:278–287. Lauber CL, Ramirez KS, Aanderud Z, Lennon J, Fierer N. 2013. Temporal variability in soil microbial communities across land-use types. ISME J. 7:1641–1650. Lazcano C, Gómez-Brandón M, Revilla P, Domínguez J. 2013. Short-term effects of organic and inorganic fertilizers on soil microbial community structure and function. Biol Fert Soils. 49:723– 733. Leach JE, Triplett LR, Argueso CT, Trivedi P. 2017. Communication in the phytobiome. Cell. 169:587–596. Lenc L, Kwaśna H, Sadowski C, Grabowski A. 2014. Microbiota in wheat roots, rhizosphere and soil in crops grown in organic and other production systems. J Phytopathol. 163:245–263. Li X, Deng Y, Li Q, Lu C, Wang J, Zhang H, Zhu J, Zhou J, He Z. 2012. Shifts of functional gene representation in wheat rhizosphere microbial communities under elevated ozone. ISME J. 7:660–671. Lindow SE, Brandl MT. 2003. Microbiology of the phyllosphere. Appl Environ Microbiol. 69:1875–1883. Linnemannstöns P, Prado M, Fernández-Martín R, Tudzynski B, Avalos J. 2002. A carotenoid biosynthesis gene cluster in Fusarium fujikuroi: the genes carB and carRA. Mol Genet Genomics. 267:593–602. Lira MA, Nascimento LRS, Fracetto GGM. 2015. Legume-rhizobia signal exchange: promiscuity and environmental effects. Front Microbiol. 6:945. Liu B, Shen W, Wei H, Smith H, Louws FJ, Steadman JR, Correll JC. 2015. Rhizoctonia communities in soybean fields and their relation with other microbes and nematode communities. Eur J Plant Pathol. 144:671–686. Liu H, Carvalhais LC, Schenk PM, Dennis PG. 2017. Effects of jasmonic acid signaling on the wheat microbiome differ between body sites. Sci Rep. 1–8. 108 Lu G-H, Zhu Y-L, Kong L-R, Cheng J, Tang C-Y, Hua X-M, Meng F-F, Pang Y-J, Yang R-W, Qi J-L, Yang Y-H. 2017. Impact of a glyphosate-tolerant soybean line on the rhizobacteria, revealed by Illumina MiSeq. J Microbiol Biotechn. 27:561–572. Lupatini M, Korthals GW, de Hollander M, Janssens TKS, Kuramae EE. 2017. Soil microbiome is more heterogeneous in organic than in conventional farming system. Front Microbiol. 7:683– 13. Mahoney A K, Yin C, Hulbert SH. 2017. Community structure, species variation, and potential functions of rhizosphere-associated bacteria of different winter wheat (Triticum aestivum) cultivars. Front Plant Sci. 8:2276–14. Mark Welch JL, Rossetti BJ, Rieken CW, Dewhirst FE, Borisy GG. 2016. Biogeography of a human oral microbiome at the micron scale. Proc Natl Acad Sci. 113:E791–E800. Marx DH, Rowan SJ. 1981. Fungicides influence growth and development of specific ectomycorrhizae on loblolly pine seedlings. Forest Sci. 27:167–176. Mashiane RA, Ezeokoli OT, Adeleke RA, Bezuidenhout CC. 2017. Metagenomic analyses of bacterial endophytes associated with the phyllosphere of a Bt maize cultivar and its isogenic parental line from South Africa. World J Microbiol Biotechnol. 33:80. McMullen M, Jones R, Gallenberg D. 1997. Scab of wheat and barley: a re-emerging disease of devastating impact. Plant Dis. 81:1340–1348. McMullen M, Bergstrom G, De Wolf E, Dill-Macky R, Hershman D, Shaner G, Van Sanford D. 2012. A unified effort to fight an enemy of wheat and barley: Fusarium head blight. Plant Dis. 96:1712–1728. McMurdie PJ, Holmes S. 2013. phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data ed. Michael Watson. PLoS ONE. 8:e61217–11. McMurdie PJ, Holmes S. 2014. Waste not, want not: why rarefying microbiome data is inadmissible. PLoS Comp Biol. 10:e1003531–12. Mende K, Homann V, Tudzynski B. 1997. The geranylgeranyl diphosphate synthase gene of Gibberella fujikuroi: isolation and expression. Mol Gen Genet. 255:96–105. Mendentsev AG, Kotik AN, Trufanova VA, Akimenko VK. 1993. Identification of an aurofusarin from Fusarium graminearum that causes egg quality deterioration in hens. Prikladnaia Biokhimiia i Mikrobiologiia. 29:406-409. Mendes LW, Kuramae EE, Navarrete AACA, van Veen JA, Tsai SM. 2014. Taxonomical and functional microbial community selection in soybean rhizosphere. ISME J. 8:1577–1587. 109 Miksicek RJ. 1994. Interaction of naturally occurring nonsteroidal estrogens with expressed recombinant human estrogen receptor. J Steroid Biochem. 49:153–160. Mitter B, Pfaffenbichler N, Flavell R, Compant S, Antonielli L, Petric A, Berninger T, Naveed M, Sheibani-Tezerji R, Maltzahn von G, Sessitsch A. 2017. A new approach to modify plant microbiomes and traits by introducing beneficial bacteria at flowering into progeny seeds. Front Microbiol. 8:1–10. Mousa WK, Shearer C, Limay-Rios V, Ettinger CL. 2016. Root-hair endophyte stacking in finger millet creates a physicochemical barrier to trap the fungal pathogen Fusarium graminearum. Nature. 1:16167. Mueller UG, Sachs JL. 2015. Engineering microbiomes to improve plant and animal health. Trends Microbiol. 23:606–617. Müller MEH, Urban K, Köppen R, Siegel D, Korn U, Koch M. 2014. Mycotoxins as antagonistic or supporting agents in the interaction between phytopathogenic Fusarium and Alternaria fungi. World Mycotoxin J. 1:1–11. Naylor D, DeGraaf S, Purdom E, Coleman-Derr D. 2017. Drought and host selection influence bacterial community dynamics in the grass root microbiome. ISME J.11:2691–2704. Nettles R, Watkins J, Ricks K, Boyer M, Licht M, Atwood LW, Peoples M, Smith RG, Mortensen DA, Koide RT. 2016. Influence of pesticide seed treatments on rhizosphere fungal and bacterial communities and leaf fungal endophyte communities in maize and soybean. Appl Soil Ecol. 102:61–69. Niu B, Paulson JN, Zheng X, Kolter R. 2017. Simplified and representative bacterial community of maize roots. Proc Natl Acad Sci. 114:E2450–E2459. Ofek M, Voronov-Goldman M, Hadar Y, Minz D. 2013. Host signature effect on plant root- associated microbiomes revealed through analyses of resident vs. active communities. Environ Microbiol. 16:2157–2167. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, Minchin PR, OHara RB, Simpson GL, Solymos P, Stevens MHH, Szoecs E, Wagner H. 2016. Vegan: community ecology package. URL: https://CRAN.R-project.org/package=vegan Ottesen AR, Pena AG, White JR, Pettengill JB, Li C, Allard S, Rideout S, Allard M, Hill T, Evans P, Strain E, Musser S, Knight R, Brown E. 2013. Baseline survey of the anatomical microbial ecology of an important food plant: Solanum lycopersicum (tomato). BMC Microbiol. 13:114. Patron NJ, Waller RF, Cozijnsen AJ, Straney DC, Gardiner DM, Nierman WC, Howlett BJ. 2007. Origin and distribution of epipolythiodioxopiperazine (ETP) gene clusters in filamentous ascomycetes. BMC Evol Biol. 7:174. 110 Peiffer JA, Spor A, Koren O, Jin Z. 2013. Diversity and heritability of the maize rhizosphere microbiome under field conditions. Proc Natl Acad Sci. 110:6548–6553. Pineda M, Pérez-Bueno ML, Paredes V, Barón M. 2017. Use of multicolour fluorescence imaging for diagnosis of bacterial and fungal infection on zucchini by implementing machine learning. Functional Plant Biol. 44:563–572. Plante CJ. 2017. Defining disturbance for microbial ecology. Microb Ecol. 74:259–263. Poudel R, Jumpponen A, Schlatter DC, Paulitz TC, Gardener BBM, Kinkel LL, Garrett KA. 2016. Microbiome networks: a systems framework for identifying candidate microbial assemblages for disease management. Phytopathology. 106:1083–1096. R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL: https://www.R-project.org/ Rascovan N, Carbonetto B, Perrig D, Díaz M, Canciani W, Abalo M, Alloati J, González-Anta G, Vazquez MP. 2016. Integrated analysis of root microbiomes of soybean and wheat from agricultural fields. Sci Rep. 6:1–12. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W and Smyth GK. 2015. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43:e47. Robertson PG. 2015. In: Robertson PG, Hamilton SK, Doll JE, eds. The ecology of agricultural landscapes: Long-term research on the path to sustainability. Oxford, UK: Oxford University Press. Rocha O, Ansari K, Doohan FM. 2005. Effects of trichothecene mycotoxins on eukaryotic cells: a review. Food Addit Contam. 22, 369–378. Rodriguez RJ, Henson J, Van Volkenburgh E, Hoy M, Wright L, Beckwith F, Kim YO, Redman RS. 2008. Stress tolerance in plants via habitat-adapted symbiosis. ISME J. 2:404–416. Rodriguez RJ, White JF, Arnold AE, Redman RS. 2009. Fungal endophytes: diversity and functional roles. New Phytologist. 182:314–330. Rojas JA, Jacobs JL, Napieralski S, Karaj B, Bradley CA, Chase T, Esker PD, Giesler LJ, Jardine DJ, Malvick DK, Markell SG, Nelson BD, Robertson AE, Rupe JC, Smith DL, Sweets LE, Tenuta AU, Wise KA, Chilvers MI. 2017. Oomycete species associated with soybean seedlings in North America—Part II: diversity and ecology in relation to environmental and edaphic factors. Phytopathology. 107:293-304. Ryffel F, Helfrich EJ, Kiefer P, Peyriga L, Portais J-C, Piel JOR, Vorholt JA. 2015. Metabolic footprint of epiphytic bacteria on Arabidopsis thaliana leaves. ISME J. 10:632–643. 111 Sapkota R, Knorr K, Jørgensen LN, O'Hanlon KA, Nicolaisen M. 2015. Host genotype is an important determinant of the cereal phyllosphere mycobiome. New Phytol. 207:1134–1144. Sapkota R, Jørgensen LN, Nicolaisen M. 2017. Spatiotemporal variation and networks in the mycobiome of the wheat canopy. Front Plant Sci. 8:281–10. Saravanakumar K, Li Y, Yu C, Wang Q-Q, Wang M, Sun J, Gao J-X, Chen J. 2017. Effect of Trichoderma harzianum on maize rhizosphere microbiome and biocontrol of Fusarium stalk rot. Sci Rep. 1–13. Schlatter D, Kinkel L, Thomashow L, Weller D, Paulitz T. 2017. Disease suppressive soils: new insights from the soil microbiome. Phytopathology. 107:1284–1297. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, Sahl JW. 2009. Introducing mothur: open-source, platform- independent, community-supported software for describing and comparing microbial communities. Appl Environ Microb. 75:7537-41. Schroeder HW, Christensen JJ. 1963. Factors affecting resistance of wheat to scab caused by Gibberella zeae. Phytopathology. 53:831–838. Schwarz PB. 2017. Fusarium head blight and deoxynivalenol in malting and brewing: successes and future challenges. Trop Plant Pathol. 42:153–164. Shade A, Handelsman J. 2012. Beyond the venn diagram: the hunt for a core microbiome. Environ Microbiol. 14:4–12. Shade A, McManus PS, Handelsman J. 2013. Unexpected diversity during community succession in the apple flower microbiome. mBio. 4:e00602–e00612. Sieber T, Riesen TK, Müller E, Fried PM. 1988. Endophytic fungi in four winter wheat cultivars (Triticum aestivum L.) differing in resistance against Stagonospora nodorum (Berk.) Cast. & Germ.= Septoria nodorum (Berk.) Berk. J Phytopathology. 122:289–306. Sieber C, Lee W, Wong P, Münsterkötter M. 2014. The Fusarium graminearum genome reveals more secondary metabolite gene clusters and hints of horizontal gene transfer. PLoS ONE. 9:e110311. Siou D, Gélisse S, Laval V, Elbelt S, Repinçay C, Bourdat-Deschamps M, Suffert F, Lannou C. 2015. Interactions between head blight pathogens: consequences for disease development and toxin production in wheat spikes. Appl Environ Microbiol. 81:957–965. Smit E, Leeflang P, Glandorf B, van Elsas JD, Wernars K. 1999. Analysis of fungal diversity in the wheat rhizosphere by sequencing of cloned PCR-amplified genes encoding 18S rRNA and temperature gradient gel electrophoresis. Appl Environ Microbiol. 65:2614–2621. 112 Smith DP, Peay KG. 2014. Sequence depth, not PCR replication, improves ecological inference from next generation DNA sequencing. PLoS ONE 9:e90234. Smith SE, Read DJ. 2008. Mycorrhizal Symbiosis. San Diego, CA: Academic Press. Sobrova P, Adam V, Vasatkova A, Beklova M, Zeman L, Kizek R. 2010. Deoxynivalenol and its toxicity. Interdisciplinary Toxicology. 3:94–99. Souza RC, Cantão ME, Vasconcelos ATR, Nogueira MA, Hungria M. 2013. Soil metagenomics reveals differences under conventional and no-tillage with crop rotation or succession. Appl Soil Ecol. 72:49–61. Souza RC, Hungria M, Cantão ME, Vasconcelos ATR, Nogueira MA, Vicente VA. 2015. Metagenomic analysis reveals microbial functional redundancies and specificities in a soil under different tillage and crop-management regimes. Appl Soil Ecol 86:106–112. Suay I, Arenal F, Asensio FJ, Basilio A, Cabello MA, Díez MT, García JB, del Val AG, Gorrochategui J, Hernández P, Peláez F, Vincente MF. 2000. Screening of Basidiomycetes for antimicrobial activities. A Van Leeuw J Microb. 78:129–140. Sun W, Qian X, Gu J, Wang X-J, Li Y, Duan M-L. 2017. Effects of inoculation with organic- phosphorus-mineralizing bacteria on soybean (Glycine max) growth and indigenous bacterial community diversity. Can J Microbiol. 63:392–401. Sutton JC. 1982. Epidemiology of wheat head blight and maize ear rot caused by Fusarium graminearum. Can J Plant Pathol. 4:195–209. Tackenberg M, Volkmar C, Dammer K-H. 2016. Sensor-based variable-rate fungicide application in winter wheat. Pest Manag Sci. 72:1888–1896. Toju H, Tanabe AS, Yamamoto S, Sato H. 2012. High-coverage ITS primers for the DNA-based identification of Ascomycetes and Basidiomycetes in environmental samples. PLoS ONE. 7:e40863. Trail F, Common R. 2000. Perithecial development by Gibberella zeae: a light microscopy study. Mycologia. 92:130–138. Trail F. 2009. For Blighted Waves of Grain: Fusarium graminearum in the postgenomics era. Plant Pathol. 149:103–110. Tudzynski B. 2005. Gibberellin biosynthesis in fungi: genes, enzymes, evolution, and impact on biotechnology. Appl Microbiol Biotechnol. 66:597–611. USDA. 2016. Crop Production 2015 Summary 01/12/2016. eds. Robert Johansson and James M Harris. 113 Utermark J, Karlovsky P. 2007. Role of zearalenone lactonase in protection of Gliocladium roseum from fungitoxic effects of the mycotoxin zearalenone. Appl Environ Microbiol. 73:637– 642. van der Heijden MGA, Klironomos JN, Ursic M, Moutoglis P, Streitwolf-Engel R, Boller T, Wiemken A, Sanders IR. 1998. Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. Nature. 396:69–72. Vogel C, Bodenhausen N, Gruissem W, Vorholt JA. 2016. The Arabidopsis leaf transcriptome reveals distinct but also overlapping responses to colonization by phyllosphere commensals and pathogen infection with impact on plant health. New Phytol. 1–16. Vorholt JA. 2012. Microbial life in the phyllosphere. Nat Rev Microbiol. 10:828–840. Wang Q, Garrity GM, Tiedje JM, Cole JR. 2007. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microb. 73:5261–5267. Wang Q, Vera Buxa S, Furch A, Friedt W, Gottwald S. 2015. Insights into Triticum aestivum seedling root rot caused by Fusarium graminearum. Mol Plant Microbe In. 28:1288–1303. Warnes GR, Bolker B, Bonebakker L, Gentleman R, Huber W, Liaw A, Lumley T, Maechler M, Magnusson A, Moeller S, Schwartz S, Bill Venables. 2016. gplots: various R programming tools for plotting data. URL: https://CRAN.R-project.org/package=gplots Wegulo SN, Baenziger PS, Nopsa JH, Bockus WW, Hallen-Adams HE. 2015. Management of Fusarium Head Blight of wheat and barley. Crop Prot. 73:100–107. White TJ, Bruns T, Lee S, Taylor JW. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR Protocols a Guide to Methods and Applications. New York, NY: Academic Press Inc. p. 315–322. Wickham H. 2009. ggplot2: elegant graphics for data analysis. New York, NY: Springer. URL: http://R-project.org/package=ggplot2 Wiese MV. 1987. Compendium of Wheat Diseases, 2nd edn. St. Paul, Minnesota: APS Press. Wu K, Yuan S, Xun G, Shi W, Pan B, Guan H, Shen B, Shen Q. 2015. Root exudates from two tobacco cultivars affect colonization of Ralstonia solanacearum and the disease index. Eur J Plant Pathol. 141:667–677. Xiao X, Chen W, Zong L, Yang J, Jiao S, Lin Y, Wang E, Wei G. 2017. Two cultivated legume plants reveal the enrichment process of the microbiome in the rhizocompartments. Mol Ecol. 26:1641–1651. 114 Xu XM, Parry DW, Nicholson P, Thomsett MA, Simpson D, Edwards SG, Cooke BM, Doohan FM, Brennan JM, Moretti A, Tocco G, Mule G, Hornok L, Giczey G, Tatnell J. 2005. Predominance and association of pathogenic fungi causing Fusarium Ear Blight in wheat in four European countries. Eur J Plant Pathol 112:143–154. Xu X, Nicholson P. 2009. Community ecology of fungal pathogens causing wheat head blight. Annu Rev Phytopathol. 47:83–103. Xu XM, Jeffries P, Pautasso M, Jeger MJ. 2011. Combined use of biocontrol agents to manage plant diseases in theory and practice. Phytopathology. 101:1024–1031. Xue K, Wu L, Deng Y, He Z, Van Nostrand J, Robertson PG, Schmidt TM, Zhou J. 2013. Functional gene differences in soil microbial communities from conventional, low-input, and organic farmlands. Appl Environ Microb. 79:1284–1292. Yin C, Mueth N, Hulbert S, Schlatter D, Paulitz TC, Schroeder K, Prescott A, Dhingra A. 2017. Bacterial communities on wheat grown under long-term conventional tillage and no-till in the Pacific Northwest of the United States. Phytobiomes. 1:83-90. Young JPW, Haukka KE. 1996. Diversity and phylogeny of rhizobia. New Phytol. 133:87–94. Zadoks JC, Chang TT, Konzak CF. 1974. A decimal code for the growth stages of cereals. Weed Res. 14:415–421. Zarraonaindia I, Owens SM, Weisenhorn P, West K, Hampton-Marcell J, Lax S, Bokulich NA, Mills DA, Martin G, Taghavi S, van der Lelie D, Gilbert JA. 2015. The soil microbiome influences grapevine-associated microbiota. mBio. 6:e02527–14. Zhao R-Y, Xiao W, Cheng H-L, Zhu P, Cheng K-D. 2010. Cloning and characterization of squalene synthase gene from Fusarium fujikuroi (Saw.) Wr. J Ind Microbiol Biot. 37:1171–1182. Zhao J, Ni T, Li J, Lu Q, Fang Z, Huang Q, Zhang R, Li R, Shen B, Shen Q. 2016. Effects of organic–inorganic compound fertilizer with reduced chemical fertilizer application on crop yields, soil biological activity and bacterial community structure in a rice-wheat cropping system. Appl Soil Ecol. 99:1–12. Zhu S, Vivanco JM, Manter DK. 2016. Nitrogen fertilizer rate affects root exudation, the rhizosphere microbiome and nitrogen-use-efficiency of maize. Appl Soil Ecol. 107:324–333. 115