Biotic Filters in Fungal Endophyte Community Assembly

BIOTIC FILTERS IN FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY

Megan Saunders

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Ecology and Evolutionary Biology University of Toronto

BIOTIC FILTERS IN FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY Degree of Doctor of Philosophy Megan Saunders Graduate Department of Ecology and Evolutionary Biology University of Toronto 2010

ABSTRACT

My work focuses on the community ecology of symbioses, specifically of fungal endophytes and their hosts. This thesis describes how plant defense compounds and a seed endophyte influence community structure of maize fungal endophytes. Maize produces benzoxazinoids (BXs), compounds toxic to microbes and insects. I assessed the influence of three factors on endophyte communities: host BX production, host neighbor identity and presence of a BX-detoxifying endophyte, Fusarium verticillioides (FV). To determine the influence of BXs on communities, two BX- producing (BX+) and one BX-nonproducing (BX–) genotype were planted in Ridgetown and Harrow, Ontario (triculture). Fungi were isolated and tested for tolerance to 2-benzoxazolinone (BOA), a toxic BX byproduct. Species and functional diversity (community distribution of BOA tolerance levels) was calculated. In seedling roots and mature leaves, the community proportion with low BOA tolerance was greater in BX– than BX+ plants. Fusarium abundance was up to 35 times higher in mature leaves of BX+ than BX– plants. Next, to assess the effect of host neighbor identity on communities, BX– monocultures were planted, and communities from BX– plants in monoculture and triculture compared. Monoculture root communities had higher species diversity than those in triculture. In vitro experiments were conducted to evaluate the influence of BOA on endophyte species interactions. FV facilitated species with lower BOA tolerance in the presence of BOA. Finally, fields were planted with a BX+ and BX– genotype in Ontario, Canada and Georgia, USA. Seed was inoculated with FV (FV+) or sterilized (FV–). FV abundance was highest in BX+FV+ plants, and Fusarium abundance was greater in

BX+ than BX– plants in mature leaves. In Georgia, BX+FV+ communities in below ground tissue had lower abundance of BOA sensitive species than BX+FV–. Overall, results suggest that BXs are a habitat filter that increased colonization by horizontally transmitted and seed-born Fusarium species. This invokes the hypothesis that selective breeding for enhanced BX concentrations increased abundance of Fusarium species in maize. The in vitro study indicated that FV could facilitate other species. In contrast, field results suggest that FV interacts competitively with community members, a trait enhanced in the presence of BXs.

iii ACKNOWLEDGEMENTS

I would first like to thank my advisor, Linda Kohn, who has been there to support and guide me through my graduate school career. Her enthusiasm and dedication to science are an inspiration. I am also grateful to the members of my Ph.D. committee for providing advice on all aspects of becoming a scientist. I thank Jim Anderson for his encouragement and insight throughout this project. Peter Kotanen allowed me to take up residence in his lab for five years, has had unwavering enthusiasm for my work, and was always willing to provide feedback. Thank you to Tony Glenn for hosting me in his lab, and for being a constant source of advice and support over the years.

This project would not have been possible without the help of a handful of excellent field and lab assistants. Lisa Hopcroft, Arthur Snare and Bobby Diaz made significant contributions to the project by sequencing fungal DNA and assisting with culture maintenance. Albert Tenuta, Britton Ormiston, and Terry Anderson assisted greatly in the field, and Jeff Brotherton provided the maize seed used in this project. Thank you also to Keith Seifert and John Leslie for providing fungal isolates. I would also like to thank Lorraine Thompson for hosting me while I was in Georgia doing field work.

I have been fortunate to share space with labmates that are outstanding both as scientists and as people; I would like to thank members of the Kohn lab, Anderson lab and Kotanen lab for their support. Caroline Sirjusingh always made herself available for technical training. I shared countless scientific discussions over afternoon tea with Andrew MacDonald, Steve Hill and Marion Andrew, which never failed to ignite my excitement about ecology.

Finally, I could not have done any of this without the love and support of my family. I am extremely lucky to have loving parents, Lynn and George, who have always believed in my ability to accomplish whatever I put my mind to. Thanks to my sister, Allyson, for doing everything possible to help me be happy. I have been fortunate enough to inherit a Canadian family, the Caradonna family, all of whom have been supportive and caring. My husband Steve has been a constant source of love, encouragement, and comfort from the moment that we met. Thank you Steve, for everything.

TABLE OF CONTENTS

ABSTRACT ii

ACKNOWLEDGEMENTS iv

TABLE OF CONTENTS v

LIST OF TABLES ix

LIST OF FIGURES xi

CHAPTER 1. GENERAL INTRODUCTION 1 Thesis overview 2 Relationship between fungal endophytes and their plant hosts 3 Biotic filters in fungal endophyte community assembly 8 Host plant species 8 Host plant genotype 10 Biochemical plant defense mechanisms 12 Fungal endophyte infection can induce host defense reactions 14 Environmentally acquired plant defense mechanisms 16 Microbial species interactions 17 Objectives of thesis 21

CHAPTER 2. EVIDENCE FOR ALTERATION OF FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY BY HOST DEFENSE COMPOUNDS 24 Abstract 25 Introduction 26 Materials and Methods 28 Maize genotypes 28 Study site and collection times 29 Isolation of fungi from plant tissue 29 Identification of fungal isolates 30 DNA isolation, polymerase chain reaction amplification (PCR) and sequencing 31 v

Assignment of isolates to BOA tolerance threshold group 32 Statistical analyses 32 Abundance of Fusarium in 9-wk-old plants 32 Diversity and similarity of fungal endophyte communities 32 BOA tolerance of fungal endophyte communities 34 Results 34 Relative abundance and diversity of endophytes 34 BOA tolerance levels of endophytic fungi 38 Partitioning of endophyte communities by BOA tolerance level 41 Abundance of Fusarium in 9-wk-old plants 41 Discussion 45 Presence of BXs influences endophyte community structure 45 BX+ plants have a higher incidence of Fusarium than BX- plants 48 Commonalities between endophyte community ecology in agricultural and naturally 49 occurring plants Conclusions 50

CHAPTER 3. PRODUCTION OF DEFENSE COMPOUNDS BY PLANT 52 NEIGHBORS DECREASES FUNGAL ENDOPHYTE ABUNDANCE AND DIVERSITY IN MAIZE ROOTS Abstract 53 Introduction 54 Materials and Methods 56 Study site, maize varieties and tissue collection 56 Isolation of fungi from root and leaf tissue 57 Identification of fungal isolates 57 Statistical analyses 58 Results 61 Discussion 64

CHAPTER 4. HOST-SYNTHESIZED SECONDARY COMPOUNDS INFLUENCE 69 THE IN VITRO INTERACTIONS BETWEEN FUNGAL ENDOPHYTES OF MAIZE Abstract 70 Introduction 71 vi

Materials and Methods 74 Characterization of BOA tolerance in maize endophytes 74 Strains assessed for BOA tolerance 74 Isolation of fungal endophytes from maize 76 Sequential inoculation experiments 77 Assessment of presence of BOA in the medium 78 Statistical analyses 79 Results 79 Characterization of BOA tolerance 79 Sequential inoculation of species pairs 79 Assessment of biodegradation of BOA in the medium 85 Discussion 85

CHAPTER 5. HOST DEFENSE COMPOUNDS AND THE SEED ENDOPHYTE, 92 FUSARIUM VERTICILLIOIDES, AS FILTERS IN FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY Abstract 93 Introduction 94 Materials and Methods 97 Field experiment 97 Isolation and identification of fungi 98 Species diversity of fungal endophyte communities 100 Functional diversity of fungal endophyte communities 101 Abundance of Fusarium 102 BX concentration of maize plants 102 Results 103 Species diversity of fungal endophyte communities 103 Functional diversity of fungal endophyte communities 109 Abundance of Fusarium 117 BX concentration in maize plants 117 Discussion 120

CHAPTER 6. SUMMARY 125

vii

LITERATURE CITED 130

viii

LIST OF TABLES

Table 2.1. BOA tolerance thresholds of fungal endophyte species or 33 morphotypes isolated from maize (bxbx, B37, W22 genotypes).

Table 2.2. Diversity of fungal endophyte communities in maize. 39

Table 2.3. Results of χ2 tests for a difference between proportion of 42 isolates in BOA tolerance threshold groups 0.25 (0.25% BOA tolerance threshold) and 0.50 in endophyte communities from maize bxbx (BX– genotype), W22 and B37 (BX+ genotypes).

Table 2.4. Identity of a subset of isolates obtained from 9-wk-old maize 46 tissue plated on BOA medium.

Table 3.1. Species isolated from leaf and root tissue of BX- (bxbx) maize 60 grown in monoculture and in triculture.

Table 4.1. BOA tolerance threshold (highest concentration of BOA 75 supporting growth) of common maize endophytes.

Table 5.1. Abundance of species and morphotypes (with No. designations) 104 isolated from maize in Ontario and Georgia.

Table 5.2. BOA tolerance thresholds (highest concentration of BOA 110 supporting growth in mg per ml) of maize fungal endophytes isolated from plants grown in Georgia and Ontario.

Table 5.3. MANOVA of BOA tolerance threshold groups with maize 113 genotype (BX+ or BX-) and infection status (FV+ or FV-) as

factors.

Table 5.4. MANOVA of average community BOA tolerance level, and 116 average community BOA tolerance level with Fusarium verticillioides removed from the analysis with maize genotype (BX+ or BX-) and infection status (FV+ or FV-) as factors.

Table 5.5. Identity of isolates obtained on BOA medium from plants grown 119 in Ontario and Georgia. Isolates from 2 plants/block were identified.

LIST OF FIGURES

Figure 1.1. Movement of species from a regional species pool to a local 7 community.

Figure 2.1. Collection data and community bar graphs of isolates obtained 35-37 from root (a) and leaf (b) tissue of maize on PDA.

Figure 2.2. Similarity coefficients of fungal endophyte communities from 40 leaf and root tissue of maize grown in Harrow and Ridgetown, Ontario.

Figure 2.3. Distribution of fungal endophyte community members isolated 43-44 on potato dextrose agar (PDA) into BOA tolerance threshold groups.

Figure 2.4. Mean number of isolates per plant obtained on BOA-medium 47 from leaf (a) and root (b) tissue of 9-wk-old maize plants in Harrow and Ridgetown, Ontario, Canada.

Figure 3.1. Species diversity, abundance of fungal endophytes, and 62-63 abundance of Fusarium in BX- (bxbx) maize grown in monoculture and in triculture.

Figure 3.2. Distribution of fungal endophytes into BOA tolerance 65 threshold groups. Leaf and root tissue was collected from bxbx maize (BX-) in Harrow and Ridgetown, Ontario.

Figure 4.1. Benzoxazinoids and their degradation products. 72

Figure 4.2. Interaction set 1 (Fusarium verticillioides, F. subglutinans 80

and F. proliferatum) mean colony area values of solo control or secondary colonizers on the BOA-amended medium (a) and on the BOA-free medium (b).

Figure 4.3. Interaction set 2 (N. oryzae, A. zeae, and P. macrospinosa) 81 mean colony area values of solo control or secondary colonizers on the BOA-amended medium (a) and on the BOA-free medium (b).

Figure 4.4. Percent differences between mean colony area values in the 83 autospecific and heterospecific interactions for interaction set 1 (Fusarium verticillioides, F. subglutinans, and F. proliferatum).

Figure 4.5. Percent differences between mean colony areas in the 84 autospecific and heterospecific interactions for interaction set 2 (Nigrospora oryzae, Acremonium zeae, and P. macrospinosa).

Figure 4.6. Qualitative assessment of presence of BOA in the medium. 86

Figure 5.1. Species diversity of maize fungal endophtye communities as 106 measured by Fisher’s alpha.

Figure 5.2. Mean number of Fusarium verticillioides isolates per plant 107-108 obtained on potato dextrose agar medium.

Figure 5.3. Mean number of isolates in each tolerance threshold group in 111-112 below ground tissue (a) and leaves (b).

Figure 5.4. Average community BOA tolerance level in fungal enodophyte 114-115 communities of maize.

xii

Figure 5.5. Mean number of Fusarium isolates per plant obtained on 118 Fusarium selective medium (BOA medium).

Figure 5.6. Concentration of benzoxazinones in above and below ground 121 tissue of maize plants.

xiii

CHAPTER 1

GENERAL INTRODUCTION

CHAPTER 1. GENERAL INTRODUCTION 2

THESIS OVERVIEW

The quest for a general theory that connects species traits to environmental conditions has engaged community ecologists for decades. A central objective is the elucidation of assembly rules - conditions that prevent movement of a species from a regional pool to a specific community. Such conditions are defined by the inability of a species to overcome particular abiotic or biotic stressors, which thereby prevent establishment. McGill et al. (2006) have suggested that finding general rules that explain the assembly of speciose communities should start with quantification of functional trait variation across environmental gradients. I am interested in the functional and community ecology of plant-fungal symbioses. More specifically, I am intrigued by the community-level response of fungal endophytes to biotic stressors. Culturable microbes offer us the unique opportunity to collect community members and hold them in stasis until laboratory experiments on trait variation are performed. I have tried to capitalize on this by testing for a connection between the phenotypes of the host and its associated fungal endophytes.

What variables determine the composition of a fungal community associated with a given plant host? Is the ability of a fungal endophyte to colonize a plant influenced equally by host defense mechanisms and fungal competitors? How might these two factors interact to shape the endophyte communities? I sought to address these questions by quantifying how plant defense compounds influence community structure of fungal endophytes in maize (Zea mays), and by evaluating the potential of these compounds to interact with other biotic factors that affect endophyte colonization. I have investigated the influence of three factors on maize endophyte diversity: direct presence of plant-produced defense compounds in host plants, indirect presence of these defense compounds in the host neighborhood, and presence of a seed endophyte that can detoxify the defense compounds in host plants. A unique aspect of my work is the incorporation of data on species diversity and functional diversity of endophyte communities. My approach has been to characterize communities by synthesizing data on 1) fungal species identity, 2) tolerance of fungal species to plant defense compounds, 3) production of defense compounds by the plant, and 4) in vitro competition between fungal species. Production CHAPTER 1. GENERAL INTRODUCTION 3 of defense compounds is ubiquitous among plant species, some of which also harbor fungal endophytes in seed. The ability of fungi to detoxify host - produced compounds is evidenced in numerous species, suggesting that these toxins are a significant selective pressure on plant-associated fungi. By assessing this and other factors shared among many plant-fungal relationships, I hope to contribute to our understanding of fungal endophyte community ecology in many plant hosts.

RELATIONSHIP BETWEEN ENDOPHYTES AND THEIR PLANT HOSTS

Almost every plant species surveyed to date harbors non-mycorrhizal fungal endophytes, species that reside within host tissue without causing visible symptoms of disease (Petrini 1991, Stone et al. 2000; Arnold 2007). These are an extremely diverse and polyphyletic group of fungi, primarily belonging to the Ascomycota (Petrini 1986; Jumpponen and Trappe 1998, Arnold 2007). Fungal endophytes have been isolated from representatives of all major lineages of land plants, and from every plant organ that has been investigated (e.g. Carroll and Petrini 1983; Fisher 1996; Arnold and Herre 2003; Davis and Shaw 2008; Kauserud et al. 2008;). To date, the most well studied group of fungal endophytes are species in the family Clavicipitaceae that form associations with some cool and warm season grass species (C endophytes). Phylogenetic evidence indicates a history of co- evolution between C endophytes and their plant hosts. These species are vertically transmitted, abundant in host tissue, and can provide a wealth of fitness benefits to the host, for example, herbivore deterrence (e.g. Clay et al. 1985; Siegel 1990). The relationship between C endophytes and their hosts differs markedly from most plant- endophyte systems. The vast majority of plant species harbor a diverse assemblage of horizontally transmitted endophyte species, often at high density in tissue. Host specificity, as observed between C endophytes and their hosts, appears to be rare among endophytes of non-grass plant species. Non-C endophytes can provide fitness benefits to the host, although little evidence of co-evolution in these relationships has been documented. Several excellent reviews on the C endophytes are available (see Clay and Schardl 2002; Schardl et al. 2004; Rudgers and Clay 2007). Here, the focus is on non-C species, and the term ‘fungal endophyte’ is applied solely to non-mycorrhizal, non-C endophytes. CHAPTER 1. GENERAL INTRODUCTION 4

Fungal endophytes can have myriad effects on host plant fitness, with the outcome of interactions ranging beneficial to antagonistic. Benefits include protection against damage by pathogens, herbivores, and salt or water stress (Arnold et al. 2003; Obledo et al. 2003; Donoso et al. 2008; Miller et al. 2008). For example, Arnold et al. (2003) compared the ability of endophyte infected and uninfected leaves of Theobroma cacao to resist damage caused by a common foliar pathogen, and found that leaves infected with a community of endophytes had significantly less visible evidence of disease. Endophyte infection can also enhance host photosynthesis. When Agave victoria-reginae was grown with its fungal endophyte Fusarium oxysporum, total chlorophyll and sugar content increased, resulting in increased photosynthetic efficiency compared to endophyte-free plants (Obledo et al. 2003). Conversely, fungal endophytes have been noted to decrease photosynthetic efficiency and increase water loss from leaves under drought conditions (Pinto et al. 2000; Arnold and Englebrecht 2007). The status of a fungal species as an endophyte is not necessarily static; an asymptomatic endophyte may prove to have been a latent pathogen by shifting to pathogenesis. This often occurs when the host is physiologically weakened; e.g., Fusarium verticillioides typically resides in tissue as a fungal endophyte, but can become pathogenic under conditions of water stress (Dodd 1980).

Some fungal endophyte species provide their host with elevated tolerance to extreme environmental conditions. The grass species Dichanthelium lanuginosum is able to withstand high soil temperatures only when in association with the fungal endophyte Curvularia protuberata (Redman et al. 2002). The symbionts commonly co-occur in geothermal sites at Yellowstone National Park, where soil temperatures can reach 57o C. When inoculated with C. protuberata, D. lanuginosum could tolerate temperatures as high as 70o C, while in the absence of the endophyte the maximum temperature that supported growth was 40o C. Without its host C. protuberata could not withstand temperatures above 38o C. Both partners gain the benefit of thermotolerance when in the association, from which the hypothesis can be inferred that the association arose as an adaptation to heat stress (Rodriguez et al. 2004). A subsequent study revealed that the benefit provided by C. protuberata is dependent on the presence a fungal virus (Márquez CHAPTER 1. GENERAL INTRODUCTION 5 et al. 2007). Plants infected with a C. protuberata isolate carrying the virus were able to withstand prolonged exposure to elevated soil temperatures, while uninfected plants and those infected with axenic C. protuberata did not.

Given that fungal endophytes influence host plant health, it is likely that they will also indirectly affect plant community dynamics. Previous studies indicate that mycorrhizal fungi and C endophytes can have a significant impact on plant communities, and can instigate a cascade of events that ultimately affects ecosystem-level processes (Hartnett and Wilson 1999; Rudgers et al. 2004; Lemons et al. 2005; Rudgers et al. 2007; van der Heijden et al. 2008). Such fungi have been noted to suppress forest succession, depress growth of native trees and enhance the ability of host plants to invade diverse communities (Marler et al. 1999; Rudgers et al. 2007). Often symbiotic fungi indirectly alter plant communities by manipulating microbial or herbivore communities. The influence of non-C fungal endophytes on plant communities has rarely been investigated. Endophyte species are, however, often observed to modify the abundance of other microbial species. Biocontrol studies frequently report decreased abundance of pathogenic fungi in the presence of endophytes. Recent work indicates that endophytes may also protect against nematodes and insect damage (e.g. Fravel et al. 2003; Miller et al. 2008; Sikora et al. 2008). For example, the ‘bodyguard hypothesis’ predicts that host plants deploy entomopathogenic fungi as an anti-herbivore defense mechanism (Elliot et al. 2000). The entomopathogenic fungus Beauveria bassiana can reside as an endophyte in maize tissue, where it protects against damage by the herbivore Ostrinia nubilalis via feeding deterrence or antibiosis (Lewis and Bing 1991; Ownley et al. 2008). Recent studies indicate that a diversity of plant species harbor entomopathogens as endophytes, suggesting that such tri-trophic interactions are common (Vega 2008, Vega et al. 2008). Overall, mounting data suggest that the multi-trophic interactions described for endophytic fungi will indirectly influence the plant community.

Elucidation of factors that regulate fungal endophyte community dynamics is crucial to determining the role of these fungi in higher-level interactions. Ecological theory predicts that specific environmental variables can act as habitat filters by preventing CHAPTER 1. GENERAL INTRODUCTION 6 establishment of species that lack the phenotype required to survive (Fig. 1.1; Diamond 1975; Weiher and Keddy 1999). If a species is able to tolerate the habitat filters in a given environment, successful colonization will hinge on the ability to navigate interactions with other species in the community. By testing for correlations between environmental parameters and phenotypic variation of community members, we can identify specific traits that are important in community assembly. This was demonstrated in maize, where Saunders and Kohn (2009) showed that host genotypes that produce defense compounds have fungal endophyte communities dominated by species with relatively high tolerance to the compounds [Saunders and Kohn 2009 (Chapter 2); Chapter 5]. Endophyte communities from genotypes that differed in ability to produce defense compounds were not distinguishable using standard diversity indices based on species identity alone, indicating that phenotypic profiling of communities may be required to identify the traits that influence community composition.

My objective in this introduction is to review recent research that has identified biotic factors involved in assembly of fungal endophyte communities, and to highlight biotic factors that hold promise for explaining community level patterns. This includes studies that have elucidated mechanisms important to the colonization success of one or a few fungal species, and studies on broader scales that address the effect of specific factors on whole communities. My intent is to cast a broad net and discuss factors that can explain community dynamics in many endophyte systems, rather than to exhaustively review every endophyte study focused on biotic interactions. First, I will briefly discuss the impact of host species identity on endophyte infection density (isolation frequency), species diversity and abundance, and will touch on how the interaction between environment and host species can shape communities. Next, the impact of host genotype and response of the endophyte community to intraspecific variation in host traits will be presented. The host plant response to endophyte infection is then discussed, with an emphasis on how recent studies on host-pathogen interactions can inform future investigations of endophyte communities. I will close with a consideration of research on microbial species interactions.

CHAPTER 1. GENERAL INTRODUCTION 7

Figure 1.1. Movement of species from a regional species pool to the local community. As species move from a regional species pool to become established as part of a community, the assemblage of fungi within an individual host, they must have the ability to disperse to the habitat and pass through environmental stressors (habitat filters) that prevent some species from colonizing. Once in the habitat, the species may experience competitive or facilitative interactions with other species. The endophyte community assemblage may influence composition of the species pool; inoculum will reside within/on plant matter until it is wind, water or animal dispersed (dotted line). Historical events, physiological constraints of the taxon and stochasticity can influence every step of this cycle.

CHAPTER 1. GENERAL INTRODUCTION 8

BIOTIC FILTERS IN FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY

Host plant species The most well studied aspect of fungal endophyte ecology is the characterization of communities associated with different host species (Carroll 1995; Stone et al. 2000; Arnold 2007). Previous research has found that in general, endophyte communities differ significantly between host species, including hosts that are closely related (Arnold 2007; Hoffman and Arnold 2008). Preliminary evidence indicates that host phylogeny may predict composition of endophyte communities; hosts that are relatively closely related can have endophyte assemblages that are more similar to one another than those in distantly related hosts. Arnold et al. (2007) found that con-familial host plants tend to be dominated by con-class endophyte species. For example, hosts in the family Fagaceae had endophyte communities that were dominated by species in the Sordariomycetes, while plants in the Cupressaceae were dominated by species in the Dothideomycetes. Similarly, a study on foliar pathogens showed that relatedness of host species is positively correlated with disease severity (Gilbert and Webb 2007). Gilbert and Webb (2007) detected a phylogenetic signal in host range of foliar pathogens; the likelihood that a fungal species could infect two plant species decreased as phylogenetic distance between host species increased. In contrast, no phylogenetic signal in host range was detected for epiphytic fungi (Gilbert et al. 2007). Instead, environmental conditions were correlated with overall epiphyte abundance, and abundance of particular species. Specifically, canopy openness was associated with distribution of fungi in understory forest plants; plants in the dark understory had higher epiphyte abundance than those in relatively exposed areas.

Results from numerous studies indicate that environmental factors, in addition to species identity, can have a significant effect on fungal endophyte communities. Other environmental variables that can influence endophyte communities include fertilizer application, experimental warming, water stress and seasonal moisture regimes (Seghers et al. 2004; Gonthier et al. 2006; Fujimura et al., 2008; Suryanarayanan et al. 2002). The CHAPTER 1. GENERAL INTRODUCTION 9 interaction between host species and environment can also significantly impact endophyte communities. Hoffman and Arnold (2008) compared endophyte communities from species in the Cupressaceae family, and found that there was relatively low similarity between endophyte communities from different host species at the same locality and between communities from the same host species in different localities, but there was high similarity from individuals from the same host species in the same locality. This suggests that both host identity and environment shape endophyte assemblages. One of the most striking environmental patterns associated with endophyte community species composition is a correlation between latitude and foliar endophyte assemblages. Arnold and Lutzoni (2007) surveyed foliar endophytes from 28 plant species along a latitudinal gradient from central Panama to the Canadian arctic, and found that species diversity was negatively correlated with latitude. Communities associated with tropical trees were, however, dominated by fewer fungal classes than those from boreal species. Results from these previous studies raise the question of how plant- endophyte associations change across the host range. Evidence indicates that abundance of host generalists may decrease as the distance to the edge of the plant distribution increases. Fisher et al. (1993, 1994) found that species richness, infection density and abundance of host-specific endophytes was lower in trees grown outside of their native range than in those found within the native range.

Most fungal endophyte species are polyphagous, and are therefore not expected to faithfully track the range of one particular host species. In general, an individual plant harbors fungal species that span the range of specificity in their associations with the host species (Carroll 1995; Stone et al. 2000; Arnold 2007; Arnold and Lutzoni 2007). Some endophyte species are host specific, i.e., with a host range of one species. Other species exhibit host preference – they are frequently associated with a particular host species, but are occasionally isolated from alternate hosts. Many endophyte species are host generalists, documented to associate with tens or even hundreds of host species. Some host generalists may be ‘incidental opportunists’ – species that happened to land on a plant host by chance, but do not typically reside in living plant tissue, and are therefore unable to proliferate or interact significantly with the host plant (Carroll 1995; Stone et CHAPTER 1. GENERAL INTRODUCTION 10 al. 2000; Schulz and Boyle 2005; Arnold 2007; Arnold and Lutzoni 2007). The coexistence in plant tissue of a mosaic of species with different levels of host specificity raises some intriguing questions. How do host specialists and generalists interact differentially with a common host? Are there phenotypic traits that consistently differentiate specialists and generalists? If host generalists and specialists have different effects on host plant fitness, selection for mean endophyte abundance could differ among genotypes at different distances from the range edge.

Host plant genotype For a fungal endophyte to affect host plant evolution, three criteria must be met. First, presence or abundance of endophytes must differ among host genotypes. Second, this variation among host genotypes is heritable. Third, the endophyte must have a significant effect on host fitness. Compared to the many studies that evaluate differences among host species, few investigate the influence of variation in host genotype on endophyte assemblages. Several studies have reported that endophyte abundance is heritable, but none have tested for an effect of infection on host fitness. Elamo et al. (1999) documented endophyte occurrence in ten genotypes of Betula pubescens (mountain birch), and found that abundance of the two dominant endophyte genera, Fusicladium and Melanconium, could be predicted by host genotype. Not surprisingly, environment also had a significant effect on endophyte abundance. While abundance of Fusicladium was uniform across individuals of the same genotype grown in different environments, abundance of Melanconium varied both with host genotype and environment. Pan et al. (2008) also detected a difference in endophyte communities associated with host genotype. The interaction between maize genotype and plot location was also significant, even though the plots were located within the same agricultural field. Even in the relatively heterogeneous environment of a maize field, factors outside of host genotype greatly impacted the endophyte community, again highlighting the importance of environmental variation on endophyte community assembly.

Genotype-genotype interactions are thought to be important in defining the relationship between plants and their fungal pathogens, but have less often been documented for CHAPTER 1. GENERAL INTRODUCTION 11 fungal endophytes (Thompson and Burdon 1992). Ahlholm et al. (2002) demonstrated an association between mountain birch genotype, overall endophyte abundance, and genetic diversity of its endophyte, Venturia ditricha. Host genotypes supporting a high density of V. ditricha in tissue had genetically uniform V. ditricha populations. In contrast, host genotypes that had a low density of V. ditricha had relatively heterogeneous V. ditricha populations. These data suggest that genetic variation in the host may affect genetic diversity of the fungus.

Intraspecific variation in plant defenses can dramatically affect endophyte communities. For example, plant genotypes that vary in ability to produce toxic compounds can select for different endophyte communities [Saunders and Kohn, 2009 (Chapter 2); Chapter 5, Bailey et al. 2005]. Bailey et al. (2005) estimated fungal endophyte abundance in twigs of two hybridizing tree species, Populus fremontii and Populus angustifolia. Endophyte abundance was positively correlated with the introgression of P. fremontii genetic markers, and tannin concentration in bark was negatively correlated with introgression. Endophyte abundance was thus negatively correlated with tannin concentration, indicating that genetic regulation of tannin concentrations selected for endophyte abundance. I compared endophyte communities from maize genotypes that differed in ability to produce defense compounds, and found that host genotypes that do not produce defense compounds harbor a higher abundance of endophytes that are sensitive to the compounds than toxin-producing genotypes [Saunders and Kohn 2009 (Chapter 2)]. My related study indicates that proximity to plant neighbors that produce defense compounds can also influence endophyte assemblages (Chapter 3). Maize genotypes that do not produce defense compounds were grown in two types of fields: in a field with two defense compound-producing genotypes (triculture) and in monoculture. Monoculture plants had higher infection density and endophyte species diversity in root tissue compared to communities from plants grown in triculture. Such a trend is expected if defense compounds were leached into soil and functioned as habitat filters.

Genotype-genotype interactions between plants and fungal pathogens can be extremely important in determining the outcome of the pathogen’s ability to colonize and cause CHAPTER 1. GENERAL INTRODUCTION 12 disease, and it has also been noted that genotype-level interactions are likely important in plant-endophyte relationships (Petrini 1991; Arnold 2007). Research that assesses heritability of host-endophyte interactions and additionally determines how endophyte species impact host fitness will be an essential step in testing for plant adaptation to endophytic inhabitants. Hypotheses about co-evolution could be addressed by assessing fitness of both partners. An assessment of fitness in the context of different endophyte communities would provide a more realistic measure of the conditions under which co- evolution could occur.

Biochemical plant defense mechanisms Variation in plant tissue chemistry can alter the ability of individual fungal species to colonize, and can influence the abundance, diversity and species composition of fungal communities. Numerous fungal taxa are tolerant to host-produced toxins, and it is likely that such compounds are among the most common stressors encountered by plant- associated fungi. The diversity of secondary compounds produced by plants is astounding. Most that have been tested demonstrate antifungal or antimicrobial properties. Some of these compounds, such as saponins and cyanogenic glycosides, are produced in a diverse group of polyphyletic plant species while others, such as the glucosinolates, are taxon specific (Hammerschmidt 1999, Osbourn 1999, VanEtten et al. 2001, Osbourn et al. 2003).

Mechanisms of fungal tolerance to toxic compounds include enzymatic detoxification, activation of membrane transporters that exude toxins from the cell, and structural alteration of the target of the toxin. The first two are known mechanisms of tolerance to plant defense compounds, with detoxification the most commonly documented mechanism. Several pathways for biotransformation of defense compounds have been elucidated, and in most cases the final byproduct is less toxic than the initial compound (VanEtten et al. 2001). The majority of research on fungal tolerance to plant toxins has focused on pathogenic species, for which tolerance can be positively correlated with quantitative virulence (Soby et al. 1996; Delserone et al. 1999). Endophytic species are a notable example of no observed correlation between tolerance and virulence (e.g. Glenn CHAPTER 1. GENERAL INTRODUCTION 13 et al. 2002).

The fungal endophyte, Fusarium verticillioides has a high level of tolerance to maize defense compounds, relative to other fungal species. Maize produces benzoxazinoids (BXs), compounds toxic to animals, plants and microbes (Barry and Darrah 1991; Niemeyer and Perez 1995; Hashimoto and Shudo 1996). These compounds are produced in the cell vacuole as biologically inactive beta-glucosides, and are converted to aglycones when the vacuole is disrupted. The aglycones are unstable, and degrade to benzoxazolinones, compounds that are also widely toxic and can persist in plant tissue and in soil (Krogh et al. 2006). F. verticillioides has a high level of tolerance to benzoxazolinones [e.g. 2-benzoxazolinone (BOA)], and converts them to less toxic products. Although F. verticillioides usually appears as a symptomless endophyte, it can become pathogenic when the plant is physiologically stressed, for example under drought stress (Dodd 1980). No correlation between BOA tolerance and virulence has been detected in F. verticillioides, (Glenn et al. 2002). However, BOA tolerance may provide an ecological advantage in colonization and establishment in maize by F. verticillioides [Saunders and Kohn 2009 (Chapter 2); Chapter 5]. In a comparison between BX producing and non-producing maize genotypes that were inoculated with F. verticillioides prior to planting, abundance of F. verticillioides was significantly greater in BX producing maize [Saunders and Kohn 2009 (Chapter 2)]. Presence of BXs also provides an advantage to Fusarium species as a whole, many of which exhibit high levels of tolerance to the compounds. In mature maize leaves, abundance of Fusarium species was up to 35 times higher in BX producers than in non-producers [Saunders and Kohn 2009 (Chapter 2)]. The correlation between production of defense compounds and dominance of endophyte species that can detoxify them has also been noted in endophyte communities associated with wheat. Carter et al. (1999) evaluated fungal communities in oat and wheat roots and found that the majority of species isolated from oat were tolerant to avenacin, a phytoanticipin that is exuded from oat roots, but that is not produced by wheat.

Compounds exuded from plant roots can have a strong impact on associated CHAPTER 1. GENERAL INTRODUCTION 14 microorganisms, for example by initiating cross talk between plants and beneficial microbes (Bais et al. 2004). Such interactions are important in the formation of relationships between leguminous hosts and nodulating bacteria and between plant roots and arbuscular mycorrhizae (Peters et al. 1986; Steinkellner et al. 2007). When compounds are secreted from roots, they exist in a matrix that also includes ions, water, enzymes and mucilage, all of which may interact to influence microbial communities. Root exudates have the potential to manipulate the pool of species available to colonize root tissue. Broeckling et al. (2008) executed a factorial design where soil microbe communities were exposed to four treatments: three generations of a plant species that naturally occurs in the soil (resident), three generations of a nonresident species, and the root exudates of each plant species in the absence of the living plant. Species richness of microbial communities changed significantly only in response to the nonresident plant and its exudates, and individual taxa were positively or negatively influenced by root exudates. This may indicate that some components of the exudates acted as defense compounds, while others behaved as signaling molecules that initiated an increase in abundance of specific fungal populations. However, non-linear interactions may also explain these results.

Fungal endophyte infection can induce host defense reactions Over the past five years, a widely accepted model for a general disease resistance response in plants has been developed (Bent and Mackey 2007). Central to this model is a sequence of four events: 1) plants have an immune system that registers a microbial invader by detecting a well-conserved trait (e.g. chitin), 2) host-adapted microbes evolve virulence factors, which suppress at least part of the general defense response, 3) the host evolves a specific R gene that indirectly detects the microbial virulence factor by perceiving their effect on host proteins, and 4) the microbe avoids detection by suppressing the defenses that have been induced in step 3. An important component of the first step is ‘priming’, when a microbe instigates induced resistance that is characterized by an increased capacity to activate a defense response to a later infection (Conrath et al. 2002).

CHAPTER 1. GENERAL INTRODUCTION 15

Much of this work seeks to describe microbe-associated molecular patterns (MAMPs), microbial traits that are recognized by the host and thereby initiate induced resistance (Boller and Felix 2009). An exciting avenue of this research is the comparison between the MAMPs and corresponding host responses induced by beneficial and pathogenic microorganisms (van Wees et al. 2008). Comparisons between pathogenic and beneficial Pseudomonas species associated with a variety of hosts indicate that beneficial microbes primarily induce systemic resistance by way of pathways known to be involved in priming (step 1, above) and are able to tolerate systemic resistance defenses, while infection by pathogens often results in immediate activation of defense (steps 1, 2 and possibly 3 and 4). Relatively few studies have evaluated the influence of beneficial fungi on such host responses. In one such study, Djonovic et al. (2007) found that a hydrophobin-like elicitor of the fungal endophyte Trichoderma virens can induce systemic resistance in maize. Other studies have described how gene expression in the plant changes with infection by different types of symbionts. For example, in a comparison between mycorrhizal and pathogenic fungi of rice, Guimil et al. (2005) found that there was a 40% overlap in the genes expressed in response to the two types of symbionts, indicating that the reaction to different partners is markedly different. As more comparisons are made, we will be able to identify genes that are regulated differentially in response to endophytes or pathogens, and distinguish plant traits that mediate fungal endophyte infection.

Several studies have compared the physiological response of plants to infection by endophytes or pathogens. In some cases, the host responded similarly. For example, peroxidase activity in the apoplastic washing fluid of bean and barley shoots was detected after inoculation with a pathogen and an endophyte. Fusarium verticillioides can induce enhanced lignin deposition and defense compound production in maize, responses that are typically associated with pathogen infection (Yates et al. 1997; Chapter 5). Other studies report a difference in plant response. Bishop (2002) found an increase in peroxidase activity in apoplastic washing fluid after infecting wheat with an endophyte, but did not see such a response to a pathogenic species. The ability of fungal endophytes to induce a host defense response may influence colonization by other endophyte species. CHAPTER 1. GENERAL INTRODUCTION 16

Kniskern et al. (2007) examined how variation in two defense signaling pathways affects the diversity and composition of bacterial epiphyte and endophyte communities. The salicylic acid (SA) and jasmonic acid (JA) signaling pathways mediate induced resistance against a broad range of microbial pathogens. To evaluate the influence of these pathways on natural bacterial communities, several Arabidopsis thaliana genotypes were compared – mutants deficient in the SA and the JA signaling defense pathways, controls, plants with artificially elevated levels of defense, and a wild type. The two signaling pathways had different effects on bacterial communities; plants lacking JA-mediated defense had an increase in epiphyte diversity, and hosts in which SA-mediated defenses were induced experienced a reduction in endophytic bacterial community diversity. This indicates that if an endophyte species predictably influences a specific defense signaling pathway, it can indirectly effect fellow community members.

The rapidly growing body of work on molecular mechanisms of plant disease resistance provides a rich background for development of theory and experiments addressing the influence of fungal endophytes on plant defense, and the effect of induced resistance on communities. Whether a generalized host response to endophyte infection exists is a central question in endophyte biology. As a whole, previous studies invoke the hypothesis that primary colonizers gain a colonization benefit by ‘priming’ the plant host. If this occurs, then microbial species tolerant to host defenses would also gain a colonization advantage, likely resulting in a dramatically different endophyte assemblage than if priming did not occur.

Environmentally acquired plant defense mechanisms Some plant species utilize environmental toxins to protect against pathogens and herbivores. The elemental defense hypothesis presents metal hyper-accumulation by plant species as an example. Recent work indicates that elemental defense is widespread among metal hyperaccumulating plant species (Boyd et al. 1994; Boyd 2007). The implications of this hypothesis are wide; such a defense system may operate directly by maintaining a sterile environment within the plant, by supplementing other defense mechanisms, or by inducing systemic resistance. The later was observed in a study on CHAPTER 1. GENERAL INTRODUCTION 17 exposure of wheat to cadmium, in which seedlings expressed induced resistance that in turn protected against colonization by the fungal pathogen Fusarium oxysporum (Mittra et al. 2004). Predictably, fungal endophyte species that commonly grow in metal contaminated areas can be highly tolerant to them, and may sequester metals in tissue (Zhang et al. 2008). This dynamic can influence host plant health. For example, ectomycorrhizal species have been noted to accumulate metals in hyphae, thereby protecting the host from metal toxicity (Leyval et al. 1997, Van Tichelen et al. 2001). To the best of my knowledge, studies that evaluate the concentration of metals in plant tissue have not considered the endophyte component. It is therefore possible that fungal endophytes mediate hyper-accumulation in plants; heavy metal accumulation could be accomplished solely by the plant, one resident microbe, or by several symbionts in concert.

Salt is another toxicant that can convey protection against microbial colonization. Excretion of salt in leaves of mangrove plant species may defend against fungal pathogens and alter endophyte abundance. Mangrove species differ in mechanisms for surviving salt stress; some species excrete salt though glands in leaves, while others filter salt from water as it passes through roots. The former strategy can result in a high concentration of salt on the leaf surface, providing a potential mechanism of resistance to foliar microbes. Gilbert et al. (2002) found that salt excreting species had lower disease expression in response to foliar pathogens, lower endophyte species richness, and lower hyphal density on the leaf upper surface than did a root filtering plant species. In many cases it is unlikely that the uptake of environmental toxins originally evolved as a defense mechanism. However, if the fitness benefit that is provided by uptake or excretion of toxins is significant, then the genotypes most efficient in toxin relocation may be selected for.

Microbial species interactions In addition to plant defense mechanisms, an endophyte must contend with other microbial species when becoming established in a community. Species interactions range from facilitative to competitive, with competition being the most frequently described CHAPTER 1. GENERAL INTRODUCTION 18 outcome of interactions between fungal species. Competition, “the negative effects which one organism has upon another by consuming, or controlling access to, a resource that is limited in availability” is often conceptualized in two categories (Keddy 1989). The first, exploitative competition, describes an interaction in which one species consumes a resource thereby reducing availability to another species. The second, interference competition, applies when one species inhibits the growth of another. Fungi obtain nutrients by colonizing an organic substrate. Interference and exploitative competition are, therefore, not clearly defined. Boddy (2000) proposed alternative concepts to describe competition between fungal species focused on timing of arrival. A species may practice primary resource capture by growing on previously uncolonized substrate, or may obtain nutrients as a secondary colonizer, a strategy that likely involves interaction with a living tenant. For the primary colonizing species, wide dispersal and rapid spore germination, mycelial growth and metabolism are important for success (Cooke and Rayner 1984). Competitive ability is crucial for survival of secondary colonizers, and is perhaps the most important type of interaction for many fungal species. Biocontrol studies provide ample examples of fungal endophytes that prevent colonization by other microbial species (Zabalgogeazcoa 2008). In some circumstances the mechanism of inhibition is consistent with that of a priming effect induced by a primary colonizing endophyte, indicating that early colonizers escape from direct competition and indirectly influence growth of secondary colonizers (Fravel et al. 2003). Endophytic biocontrol species can reduce abundance of unwanted root microbes by competing for nutrients in the soil, and as either primary or secondary colonizers can compete for infection sites, thereafter preventing other microbes from establishing on host tissue (Fravel et al. 2003).

That such competition is mediated by production of antifungal compounds has been proposed. An in vitro study found that approximately 80% of fungal endophyte species produce secondary compounds with herbicidal, antibacterial or antifungal activity (Schulz et al. 2002). Experimental data show that endophyte species can compete directly by producing antifungal secondary compounds that diffuse in substrate, and prevent other species from encroaching. Relatively few studies have documented the CHAPTER 1. GENERAL INTRODUCTION 19 presence of endophyte secondary compounds in planta, likely because the compounds in low concentration may evade detection. There are only a few circumstances where endophyte secondary compounds are found in appreciable quantities in the host plant, for example the accumulation of secondary compounds produced by Fusarium species in maize (Bacon et al. 2008). It is likely that microbial interactions mediated by endophyte secondary compounds operate on small spatial scales, making them challenging to detect. Some endophyte species produce volatile compounds with antifungal activity, resulting in direct competition at spatial scales expected to be larger than those for antibiosis within a substrate. The suite of low molecular mass, volatile, antifungal compounds produced by Muscodor albus is one example (Strobel et al. 2001; Stinson et al. 2003). Tolerance of fungal species to the volatiles is variable, and the influence of the compounds on fungal communities is unknown.

Fungal species can facilitate the growth of other species, although this type of interaction is less frequently documented than competition. Facilitation between species is important in community dynamics of plants and animals; it is possible that facilitation between plant-associated fungi is not often observed because identification of competitive species is the goal of many studies. One study on saprophytic fungi demonstrated that positive interactions are important in maintaining species diversity (Tiunov and Shaeu 2005). Fungal communities with species richness ranging from a monoculture to five species were grown on two substrate types: complex and single- resource. Rate of decomposition was positively correlated with species richness on both substrates, indicating that facilitative interactions were more important than resource partitioning in maintaining species diversity. Evidence suggestive of facilitation has also been noted for wood-decay fungi. For example, in an extensive fungal survey, Niemela et al. (1995) found that the occurrence of over twenty basidiomycete species is statistically correlated with specific species of fungi that proceeded them, indicating a dependence on primary colonizers.

Few studies have tested for facilitation between fungal endophyte species. One excellent study evaluated communities in several genotypes of maize, and used null models to test CHAPTER 1. GENERAL INTRODUCTION 20 for fungal species that were found together more or less often than is expected by chance, hypothesized to be a signal of facilitation and competition (Pan and May 2009). Fungal communities in shoot tissue were assessed using both a culture-dependent and culture- independent approach. Data from the culture-dependent approach suggest that intraspecific facilitation is common, while data from the culture-independent approach either did not detect any significant pairwise interactions between species, or detected competition, depending on the scale of analysis. In another study on culturable endophytes of maize, Saunders et al. (Chapter 5) hypothesized that a primary colonizer, the seed endophyte Fusarium verticillioides, would facilitate ingress of other species by detoxifying plant defense compounds. The influence of facilitation and competition on species diversity is addressed in the modified Intermediate Disturbance Hypothesis (IDH). The IDH predicts that when disturbance or environmental stress is low, competition will dominate interactions, resulting in low species diversity because one or few species will prevent other species from inhabiting the community (Paine 1966; Sousa 1979; Hacker and Gaines 1997). When stress is high the focal species ameliorates the environmental stress thereby increasing species diversity. Our results did not conform to the expectations of the IDH, but rather indicated that overall, F. verticillioides decreased species diversity in below ground tissue of maize. In some environments, F. verticillioides prevented species with lower tolerance to maize toxins from colonizing below ground tissue, a result consistent with what is expected if F. verticillioides interacted competitively with community members, rather than facilitatively as expected. Compared to below ground tissue, F. verticillioides did not colonize leaves in great abundance, and did not significantly influence leaf endophyte communities.

Do above and below ground tissues foster the same type of microbial species interactions? Fungal endophyte communities in root and shoot tissue can differ markedly in infection density and species composition. Leaf endophyte communities are often characterized by highly localized, spatially small (<0.2 mm) colonies (Herre et al. 2007), whereas root endophytes species are more likely to colonize systemically, often with no apparent niche within the root system (Rodriguez et al. 2009; Schulz and Boyle 2005). Fungi living in above ground tissues are exposed to specific stressors, including higher CHAPTER 1. GENERAL INTRODUCTION 21 temperature variation and UV radiation than root-dwelling fungi. Foliar endophyte species may direct their resources to coping with stress and are therefore less likely to engage in extensive combat, explaining the high density of small, localized infections in leaf tissue.

Most plants simultaneously harbor a diverse assemblage of both fungal and bacterial endophyte species, and it is likely that bacterial and fungal species also interact (Fisher et al. 1992; Arnold 2007; Ryan et al. 2008). Bacon et al. (2007) evaluated the outcome of in vitro interactions between maize endophytes F. verticillioides and the bacterium, Bacillus mojavensis in the presence and absence of the BX byproduct, BOA. Results show that B. mojavensis can impede the ability of F. verticillioides to metabolize BOA by preventing the transformation of a toxic intermediate product into the less toxic HPMA. This resulted in high accumulation of 2-amino-3H-phenoxazin-3-one in medium, which is highly toxic to F. verticillioides. In an in vitro setting, B. mojavensis competed with F. verticillioides by disrupting its toxin tolerance mechanism, thereby creating an environment that was toxic to the fungus. This study illustrates the complex nature of microbial species interactions, and reveals some important interactions that are missed when only a subset of the microbial community is considered. The inclusion of fungal and bacterial endophytes in future studies will enable a more holistic understanding of in planta microbial dynamics.

OBJECTIVES OF THESIS

Fungal endophytes are subjected to a wide variety of biotic stressors that may influence the ability of particular taxa to establish within a community. Every endophyte species is likely to co-occur in planta with both plant defense compounds and a fungal species that is well adapted to these defense compounds. Ecological theory predicts that a species well adapted to an environmental stress will interact with the stressor to influence species diversity. The goal of my thesis was to investigate the influence of these biotic factors, and the interaction between them, on the functional and species diversity of fungal endophyte communities. While pursuing this goal, I hoped to develop methods for CHAPTER 1. GENERAL INTRODUCTION 22 measuring the phenotypic diversity of fungal communities by combining field observations with in vitro categorization of fungal trait variation.

My first objective was to determine the influence of BX production on fungal endophyte community assembly. I conducted a field experiment comparing endophyte communities from root and leaf tissues of maize genotypes that differ in ability to produce BXs. This is described in Chapter 2, which has been published (Saunders and Kohn 2009).

My second objective was to determine the influence of BX production on fungal endophyte communities of neighboring plants that do not produce BXs. This was an unplanned comparison that proved interesting, and is described in Chapter 3.

My third objective was to evaluate the influence of BOA on interactions between maize endophyte species in vitro. The influence of substrate alteration by a primary colonizer on subsequent endophyte colonization was assessed, as described in Saunders and Kohn (2008) and Chapter 4.

Finally, I tested the influence of BX production by the host, presence of the BX- detoxifying endophyte, Fusarium verticillioides in seed, and the interaction between these two factors on endophyte community assembly. My objective was to test the hypothesis that the influence of F. verticillioides on endophyte communities is conditional on the presence of BXs. In addition, I was able to re-test the hypothesis that BXs influence fungal communities and to determine if seed-born F. verticillioides also gains a colonization advantage in the presence of BXs (Chapter 5).

Understanding the biotic factors that influence assembly of fungal endophyte communities will help to predict the downstream effects of these fungi on surrounding communities and ecosystems. This is the first study to evaluate the influence of host defense compounds, the influence of a seed endophyte well adapted to the toxins, and the interaction between these factors on fungal endophyte communities. Examination of the effect of defense compound production on endophyte communities of plant neighbors is CHAPTER 1. GENERAL INTRODUCTION 23 also novel. It is my hope that identification of traits that impact assembly of fungal endophyte communities will contribute to our understanding of endophyte biology, and will facilitate the development of general principals in fungal community ecology.

CHAPTER 2

EVIDENCE FOR ALTERATION OF FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY BY HOST DEFENSE COMPOUNDS

Previously published as: Saunders M and Kohn LM. Evidence for alteration of fungal endophyte community assembly by host defense compounds. New Phytologist. 2009 April; 182 (1): 229-238. This chapter is reproduced with permission granted from © New Phytologist.

CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 25 ENDOPHYTE COMMUNITIES ABSTRACT

Plant defense compounds are common stressors encountered by endophytes. Fungi readily evolve tolerance to these compounds, yet few studies have addressed the influence of intra-specific variation in defense compound production on endophyte colonization. We compared the influence of defense compound production on the composition of fungal endophyte communities in replicated field experiments. Maize (Zea mays) produces benzoxazinoids (BXs), compounds with antifungal byproducts persistent in the environment. To determine the influence of BX production on fungal endophyte communities, fungi were isolated from leaf and root tissue of two maize genotypes that produce BXs, and a natural mutant that does not. Isolates representing the species recovered were tested for tolerance to 2-benzoxazolinone (BOA), a toxic BX byproduct. Species diversity and functional diversity, defined as the distribution of BOA tolerance levels among communities, was assessed. In seedling roots and mature leaves, the community proportion with low BOA tolerance was significantly greater in BX nonproducers than producers. Mean isolation frequency of Fusarium species was up to 35 times higher in mature leaves of BX producers than nonproducers. Data suggest that fungal species with relatively high tolerance to BOA are more abundant in BX producing than BX nonproducing maize. Production of BXs may increase colonization by Fusarium species in maize, including agents of animal toxicosis and yield-reducing disease. Overall, results indicate that production of defense compounds can significantly alter endophyte community assembly.

CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 26 ENDOPHYTE COMMUNITIES INTRODUCTION

Communities of nonmycorrhizal fungal endophytes occupy most plants (Schulz and Boyle 2005; Arnold 2007). They can directly affect plant fitness and indirectly influence surrounding plant and arthropod communities (Arnold et al. 2003; Finkes et al. 2006; Rudgers et al. 2007). Fungal endophytes are transmitted in seed (vertically), or from plant to plant via fungal propagules (horizontally). An excellent example of obligate vertical transmission is the well-studied ascomycete, Neotyphodium, which forms associations with temperate grasses mediated by molecular signaling dependent on the compatibility of endophyte and host species (reviewed in Schardl et al. 2004). Phylogenetic and molecular genetic evidence indicate co-evolution of some Neotyphodium species with their plant hosts (for reviews see Clay and Schardl 2002; Schardl et al. 2004). By contrast, the majority of fungal endophytes are transmitted horizontally. These fungi can form diverse species assemblages within plants, often at high density (Arnold 2007). There are still many open questions about the mechanisms that influence colonization and community structure of horizontally transmitted endophytes, but certain abiotic environmental factors and plant defense compounds are known to be important.

Abiotic factors such as fertilizer application can influence the community structure of fungal endophytes in maize (Zea mays) (Seghers et al. 2004), and water activity and temperature can significantly influence the outcome of interactions between maize endophytes (Marin et al. 1998). Plant defense compounds can also influence interactions between the plant and its surrounding community of fungi, bacteria, insects, and plants (Niemeyer and Perez 1995). All plant species studied to date produce defense compounds (Hashimoto and Shudo 1996; Dixon 2001). Common tolerance strategies adopted by fungi include activation of membrane transporters that pump toxicants out of cells, and enzymatic detoxification (VanEtten et al. 2001). Detoxification of host compounds can be a virulence factor among pathogens (VanEtten et al. 2001). 230 Res Ability to detoxify may also increase competitiveness of root endophytes (Carter et al. CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 27 ENDOPHYTE COMMUNITIES 1999). Increase in endophyte competitiveness among tolerant species is likely the result of higher growth rates in the presence of host toxins than less well-adapted species [Arnold and Herre 2003; Nicol et al. 2003; Saunders and Kohn 2008, (Chapter 4)]. This fitness benefit would be expected to influence endophyte community assembly. In a comparison of fungal communities in Avena sativa (oat) and Triticum sp. (wheat) roots, Carter et al. (1999) found that the majority of oat-derived isolates tolerated the oat defense compound, avenacin A-1. The spatial scale of influence may increase when plant defense compounds are secreted in soil, affecting colonization by mycorrhizal and soil fungi (Stinson et al. 2006; Broeckling et al. 2008). Commercial maize has been selectively bred to produce high quantities of defense compounds, the benzoxazinoids (BXs). Several BX byproducts are toxic to microbes, insects and plants (Barry and Darrah 1991; Niemeyer and Perez 1995; Hashimoto and Shudo 1996). The primary BXs found in maize are 2,4-dihydroxy-7-methoxy-2H-1,4-benzoxazin-3-one (DIMBOA) and 2,4-dihydroxy-2H-1,4-benzoxazin-3-one (DIBOA). These compounds reside in the cell vacuole as biologically inactive beta-glucosides. They are enzymatically converted to toxic benzoxazinoids upon cell disruption (Hashimoto and Shudo 1996), ultimately degrading to the biologically active and stable benzoxazolinones, 6-methoxy-2- benzoxazolinone (MBOA) and 2-benzoxazolinone (BOA). These are formed systemically and secreted from root tissue (Park et al. 2004). They can be produced both constitutively and in response to tissue damage (Cambier et al. 2000; Oikawa et al. 2004).

All commercial maize genotypes produce BXs. Concentrations of DIMBOA have been recorded ranging from 2.86 to 12.90 mmol per kg fresh weight (Zuniga et al. 1983; Cambier et al. 2000). The concentrations of BXs and their toxic byproducts in plant tissue can vary with plant age, tissue, genotype and environment (Zuniga et al. 1983; Richardson and Bacon 1993; Cambier et al. 2000). Uptake of BX byproducts from soil by plants has been reported (Chiapusio et al. 2004). Accumulation of BX byproducts in soil is expected to depend on all of these factors, as well as the interactions between members of the microbial community (Bacon et al. 2007). It has been proposed that the ability to detoxify benzoxazolinones enhances colonization success in maize (Glenn et al. CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 28 ENDOPHYTE COMMUNITIES 2001). Some detoxifying species, particularly Fusarium verticillioides, Fusarium subglutinans, Fusarium proliferatum and Fusarium graminearum cause disease in corn, but are also common endophytes. As endophytes, they can lead to the asymptomatic contamination of grain, in some cases producing toxins that cause mycotoxicosis in animals, and are suspected risk factors for cancers and other human health problems (Ueno et al. 1997; Marasas 2001; Marasas et al. 2004). Contamination of maize grain by Fusarium is estimated to cause millions of dollars of economic loss annually in the USA (Wu 2007; Wu and Munkvold 2008). Breeding programs aimed at deterring infection of maize by Fusarium using native resistance mechanisms have been largely unsuccessful (Munkvold 2003).

Here, we investigated the influence of BX production on the assembly and composition of fungal endophyte communities in maize. Specifically, three hypotheses were tested: (1) BX production increases the proportion of fungi tolerant to the toxic BX byproduct, BOA, in endophyte communities; (2) BX production increases the incidence of Fusarium in maize; and (3) BX producing genotypes harbor endophyte communities that are less diverse than genotypes that do not produce BXs. To test these hypotheses, fungal endophyte communities from three maize genotypes were compared: one was a natural mutant deficient in BXs (BX–) and the other two were commercial genotypes that produce BXs (BX+). The objective was to observe communities in BX– versus BX+ maize. Common species were tested for BOA tolerance in vitro. Community structure was consistent with expectations for BX influence in seedling roots and mature leaves. BX production significantly increased the frequency of Fusarium in leaves of mature plants.

MATERIALS AND METHODS

Maize genotypes Each location was planted with three maize genotypes W22 and B37, two genotypes both producing BXs and commonly used to produce commercial hybrids, and bxbx, the only recorded natural mutant lacking the ability to produce BXs (Hamilton 1962). The three CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 29 ENDOPHYTE COMMUNITIES genotypes are Yellow Dent maize, characterized by a genetic background of flint and floury maize and common kernel phenotypes (Smith et al. 2004). Because of the nonlinear dynamics of BX concentration in plants and soil, we chose to assess the influence of BX production, rather than concentration, on endophyte communities.

Study site and collection times Maize was planted in two Ontario locations approx. 123 km apart: Ridgetown, with a history of soybean crops from 2001 to 2004, and Harrow, with a history of maize from 1999 to 2004. Within the plot, each genotype was planted in twelve rows, one genotype per row, as a row intercropping design. Assignment of rows within the field was random. The Ridgetown plot was surrounded by soybean, which does not produce BXs, and the Harrow plot by BX-producing maize. Planting was on June 12, 2005, with sampling carried out at two and nine wk subsequently. Whole plants were collected in paper bags and stored at 4°C.

Isolation of fungi from plant tissue Plants were rinsed with distilled water and surface dried at room temperature. From each plant, eight healthy 1.0 x 2.0 cm segments were taken 0.5 cm from the midrib and above the leaf collar of the second and third leaf blades (2-wk-old plants) or fourth and fifth blades (9-wk-old plants). Plants collected at two wk had three emergent leaves (V3 growth stage). Plants collected at 9 wk were entering the R3 stage of growth, midway through kernel development and approx. 3 wk before physiological maturity (Ritchie et al. 1993). Leaf senescence begins at physiological maturity. Our goal was to analyze healthy plant material; tissue was therefore collected before maturity. Healthy root segments 2.0 cm long and 0.2–0.3 cm diameter were taken from the radicle and lateral seminal roots. Tissue segments were surface-sterilized first in 70% ethanol (2 min), then in 0.53% NaOCl (2 min) and finally in sterile double-distilled water (2 min).

Tissue segments were incubated on two growth media. A general, neutral medium, potato dextrose agar (PDA; Difco, Detroit, MI, USA) was used to capture a relatively broad snapshot of the fungal community. A selective medium amended with 1.00 mg per CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 30 ENDOPHYTE COMMUNITIES ml BOA (Glenn et al. 2001) was used to determine the mean number of Fusarium colonies per plant (subsequently termed, abundance), isolated from 9-wk-old plants; this time-point was initially selected as an indicator of potential infestation of crop debris. Potato dextrose agar is expected to be more favorable than BOA medium to fungi that are BOA-sensitive.

Four leaf and four root segments were incubated on PDA amended with antibiotics (1.00 g per liter streptomycin sulfate and 0.25 g per liter neomycin sulfate), and four segments of each tissue were incubated on BOA medium. Segments from a total of 144 plants were plated on PDA (12 plants, 1 plant per row x 2 locations x 2 times: 2 wk and 9 wk post-planting x 3 maize genotypes); 9–12 plants per treatment were analyzed. A total of 144 plants were plated on BOA medium (24 plants, 2 plants per row x 2 locations x 1 time: 9 wk post planting x 3 maize genotypes); 16–24 plants (8–12 rows) per treatment were analyzed. All tissues were surface sterilized and plated within 96 h of collection. Plates were incubated at room temperature under a 12-h light–12-h dark regime.

Effectiveness of the surface sterilization procedure was tested on PDA by plating out 500 µl of the rinse water from the sterilization procedure, and independently by using the tissue imprint method described by Schulz and Boyle (2005). Approximately 25% of tissue samples processed were tested, and no surface contaminants were detected.

Identification of fungal isolates Fungi emerging from plant tissue were isolated and established in axenic culture. Sporulating cultures were identified morphologically. Each Fusarium isolate was established in axenic culture from a single spore and identified morphologically when diagnostic characters were evident, or using DNA sequence data when such characters were ambiguous or absent. Isolates of Fusarium were grown on Carnation Leaf Agar (Leslie et al. 2006) and PDA for morphological identification (Summerell et al. 2003; Leslie et al. 2006). For each Fusarium species that was identified morphologically, the identity of a subset of isolates was verified with DNA sequence data. Nonsporulating fungi were grouped into morphotypes. To maintain consistency, all morphotyping was CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 31 ENDOPHYTE COMMUNITIES done by one of the authors (M.S.). Morphotypes were then confirmed in blind tests by the other author (L.M.K.).

Isolates obtained on PDA were used to characterize 24 fungal endophyte communities (2 locations x 2 times x 2 tissue types x 3 maize genotypes). All isolates obtained from PDA were morphotyped, and isolates recovered more than three times were identified taxonomically. Of the isolates obtained on BOA medium, 57 isolates, 18 from leaf tissue and 39 from root tissue, were identified using molecular sequence data.

DNA isolation, polymerase chain reaction amplification (PCR) and sequencing Total genomic DNA was isolated using the DNeasy Plant Minikit (Qiagen, Mississauga, ON, Canada). For identification of morphotypes, the nuclear ribosomal internal transcribed spacer region (ITS) was amplified by PCR using primers ITS- 1F and NLB-3 (c. 700 bp) (Gardes and Bruns 1993; Martin and Rygiewicz 2005). For identification of Fusarium isolates, the translation elongation factor 1-alpha (TEF) gene was amplified using primers TEF-1 and TEF-2 (c. 700 bp) (Geiser et al. 2004).

The final volumes of the PCR mixture (50-µl volume) were 9.75 µl glass-distilled H2O, 5.00 µl 10x PCR buffer, 5.00 µl deoxynucleoside triphosphates, 0.50 µl of each primer (50 mm), 4.00 µl of magnesium chloride, 0.25 µl Amplitaq DNA polymerase (Perkin- Elmer, Norwalk, CT, USA), and 25.00 µl of a 50-fold dilution of genomic DNA. A Perkin- Elmer GeneAmp System 9600 or 9700 thermocycler was used to amplify PCR product. The PCR program used was: 1 cycle of 95°C for 8 min, 35 cycles of 95°C for 1 min, 55°C for 30 s and 72°C for 1 min, and 1 cycle of 72°C for 10 min. Sequencing was performed at the Genetic Analysis Core Facility (USDA-ARS-ERRC, Wyndmoor, PA, USA) using an ABI 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Contigs were assembled and edited in sequencher 4.6 (Gene Codes Corporation, Ann Arbor, MI, USA), and blast searches of the NCBI GenBank database were conducted for tentative identification. A 99% sequence match to a sequence of known origin in the CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 32 ENDOPHYTE COMMUNITIES database was counted as a correct species identification. DNA sequences have been deposited in GenBank as FJ496215-FJ496332.

Assignment of isolates to BOA tolerance threshold group Isolates were assigned to a tolerance threshold group corresponding to the highest concentration of BOA that supported growth (Table 2.1). The PDA was amended with BOA (stock solution of 100 mg per ml in anhydrous ethanol) in each of the following concentrations: 0.00, 0.25, 0.50, 0.75, 1.00, 1.10, 1.20 mg per ml. From a total of 24 species/morphotypes, 61 isolates were assigned to BOA tolerance threshold groups. Two isolates per species/morphotype were assigned when available. All Fusarium species and species/morphotypes recovered more than three times were included. Strains were incubated in quadrant Petri dishes for 14 d in the dark at 22°C, and scored for growth or no growth.

Statistical analyses For all statistical analyses, each row was considered the unit of replication. For the PDA assays, one plant per row was sampled. For the BOA assay, two plants per row were sampled. The average number of isolates per plant, averaged across each row, was considered a replicate.

Abundance of Fusarium in 9-wk-old plants One-way ANOVAs were followed by Tukey–Kramer Honestly Significant Difference (HSD) tests for pairwise comparisons between the BX– and BX+ genotypes. The threshold for statistical significance was P ≤ 0.05. Analyses were conducted using JMP in (version 5.1; SAS Institute Inc., Cary, NC, USA).

Diversity and similarity of fungal endophyte communities As described earlier in the section Identification of Fungal Isolates, 24 fungal endophyte communities were characterized from isolates obtained on PDA. Diversity was measured using the Simpson’s inverse diversity index (D) and Fisher’s alpha (α) (Magurran, 2004).

CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 33 ENDOPHYTE COMMUNITIES Table 2.1. BOA tolerance thresholds of fungal endophyte species/morphotypes isolated from maize (bxbx, B37, W22 genotypes).

BOA tolerance Species / Morphotype Order Number of Number of threshold group (mg/ml) isolates in group isolates tested

0.25 Alternaria alternata Pleosporales 4 4

m14 Sordariales 4 4

Periconia macrospinosa anamorphic ascomycete 3 3

0.50 Cladosporium sp. Capnodiales 2 2

Epicoccum nigrum anamorphic ascomycete 2 2

Fusarium acuminatum Hypocreales 1 1

Fusarium incarnatum-equiseti species complex Hypocreales 5 5

Fusarium oxysporum species complex Hypocreales 6 13

Fusarium proliferatum Hypocreales 1 1

Fusarium redolens Hypocreales 1 1

Fusarium sporotrichioides Hypocreales 1 1

Fusarium tricinctum Hypocreales 1 1

Penicillium sp. Eurotiales 3 3

Periconia circinaturm anamorphic ascomycete 3 3

0.75 Fusarium oxysporum species complex Hypocreales 7 13

Fusarium solani species complex Hypocreales 2 2

Fusarium verticillioides Hypocreales 1 1

m22 anamorphic ascomycete 2 2

m51 anamorphic ascomycete 1 1

Nigrospora oryzae Trichosphaeriales 2 4

Trichocladium sp. Pleosporales 2 4

Trichoderma sp. Hypocreales 2 2

1.00 Fusarium culmorum Hypocreales 4 4

Fusarium graminearum Hypocreales 1 1

Trichocladium sp. Pleosporales 2 4

Nigrospora oryzae Trichosphaeriales 2 4

1.10 Rhizopus sp. Mucorales 3 3

CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 34 ENDOPHYTE COMMUNITIES Similarity was determined using the Jaccard’s (Magurran 2004) and Morisita–Horn indices (Chao et al. 2005). All were calculated using estimates (Colwell 2000).

BOA tolerance of fungal endophyte communities The distribution of isolates in BOA tolerance threshold groups was determined for each community. To test for a difference between the distribution of isolates in BOA tolerance threshold groups in BX+ and BX– plants, a likelihood ratio test was conducted using the JMP statistical analysis software package. When necessary, results were then adjusted for small expected values using the formula (Gotelli and Ellison 2004): 2 2 X adjusted = X / qmin

where n is the number of rows, m is the number of columns, N is the total sample size, v is the degrees of freedom, and Yi,j is the frequency of observations in row i, column j.

When global likelihood ratio tests detected a significant difference between the distribution of isolates in BX+ and BX– plants, further tests for a difference between the proportion of fungi in the 0.25 and 0.50 BOA tolerance threshold groups were conducted. Follow-up likelihood ratio tests were conducted on the 0.25 and 0.50 tolerance threshold groups, because a mean of 83% of isolates in each community were in these tolerance threshold groups.

RESULTS

Relative abundance and diversity of endophytes Infection density, measured as the proportion of tissue segments yielding isolates, ranged CHAPTER 2. INFLUENCE OF BX PRODUCTION 35 ON ENDOPHYTE COMMUNITIES

Relative abundance of community members a b c d e f

2 weeks H bxbx 44 64 32

W22 48 55 29

B37 44 54 28

R bxbx 48 60 28

W22 48 54 31

B37 44 57 29

9 weeks H bxbx 40 85 53

W22 44 84 53

B37 48 69 42

R bxbx 40 76 45

W22 40 73 47

B37 36 83 49

Figure 2.1a 1 10 20 30 40 CHAPTER 2. INFLUENCE OF BX PRODUCTION 36 ON ENDOPHYTE COMMUNITIES

Relative abundance of community members a b c d e f

2 weeks H bxbx 48 71 33

W22 44 66 31

B37 48 63 31

R bxbx 48 44 21

W22 48 50 24

B37 48 65 32 c

9 weeks H bxbx 40 100 60

W22 36 100 64

B37 40 100 65

R bxbx 40 100 56

W22 40 100 69

B37 40 100 66

Figure 2.1b 1 10 20 30 40 CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 37 ENDOPHYTE COMMUNITIES Figure 2.1. Collection data and community bar graphs of isolates obtained from root (a) and leaf (b) tissue of maize on PDA. The community bar graph shows the isolation frequency of each species/morphotype in each treatment. Isolates to the left of the grey line dividing each bar graph were tested for BOA tolerance, isolates to the right of the grey line were isolated 3 times or less and were therefore not tested for BOA tolerance.

Columns: age of plant (a), location (b; H=Harrow and R=Ridgetown), maize genotype

(c), number of tissue segments analyzed (d), infection frequency (e), total number of isolates (f). Relative abundance of community members (left to right): BOA tolerance threshold group (TG) 0.25 (BOA tolerance threshold= 0.25% concentration): Alternaria alternata, Morphotype no. 14, Periconia macrospinosa; TG

0.50: Cladosporium sp., Epicoccum nigrum, Fusarium acuminatum, Fusarium incarnatum-equiseti species complex, Fusarium proliferatum, Fusarium redolens,

Fusarium sporotrichioides, Penicillium sp., Periconia circinatum; TG 0.50 and 0.75:

Fusarium oxysporum species complex; TG 0.75: Fusarium solani species complex,

Fusarium verticillioides, Trichoderma sp. Morphotype no. 22, Morphotype no. 51; TG

0.75 and 1.00: Nigrospora oryzae, Trichocladium sp., TG 1.00: Fusarium culmorum,

Fusarium graminearum; TG 1.10 V: Rhizopus sp. Isolates not tested for BOA tolerance: Chaetomium globosum, Colletotrichum sp., Diaporthe phaseolorum,

Geotrichum sp., Irpex lacteus, Scytalidium sp., Sphaeropsis sp., Stemphyllium sp.,

Morphotype no. 32, Morphotype no. 39, Morphotype no. 29, Morphotype no. 42,

Morphotype no. 44, Morphotype no. 46, Morphotype no. 47, Morphotype no. 48,

Morphotype no. 50, Morphotype no. 54, Morphotype no. 55.

CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 38 ENDOPHYTE COMMUNITIES from 54 to 100%, and was higher in August than in June in leaf and root tissue of all maize genotypes (Fig. 2.1). Of 1495 fungal isolates obtained, 1018 were isolated on PDA, and 477 on BOA. The genus Fusarium includes several species complexes. Consequently, the Fusarium oxysporum species complex, the Fusarium incarnatum- equiseti species complex, and the Fusarium solani species complex were each treated as single species in the analyses. A total of 43 species/morphotypes were identified. All of the 13 species BOA were also isolated on PDA. On BOA medium, roots yielded approximately twice the number of isolates as leaves.

Species richness and presence/absence differed between communities from leaf and root tissue on PDA. Leaf and root tissue had 31% (13/42) species/morphotypes in common, root communities had 60% (25/42) species/morphotypes present that were not isolated from leaves, and leaf communities had 9% (4/42) species/morphotypes that were not isolated from roots. The diversity of endophyte communities was assessed using isolates obtained on PDA. Values of Simpson’s D (inverse) and Fisher’s α were similar across communities from all three maize genotypes, with one exception: in communities isolated from seedling roots in Harrow, BX– plants had higher diversity than BX+ plants (Table 2.2). In all treatments (2 locations x 2 collection times x 2 tissue types), diversity of communities was lower in August than in June.

Based on the similarity indices, the expectation that endophyte communities from BX+ plants would be more similar to one another than either were to communities from BX– plants was observed in four of eight treatments (Fig. 2.2). Overall, diversity and similarity indices detected a difference between BX+ and BX− communities in Harrow more often than Ridgetown.

BOA tolerance levels of endophytic fungi The BOA tolerance threshold of isolates ranged from 0.25 to 1.10% concentration. No isolates were able to grow at 1.20%. Isolates were categorized in five BOA tolerance threshold groups (Table 2.1). Twenty-one of 24 species/morphotypes had no intraspecific variation in tolerance level. The remaining three species, Trichocladium sp., CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 39 ENDOPHYTE COMMUNITIES

Table 2.2. Diversity of fungal endophyte communities in maize. Alpha= Fisher’s alpha and Simpson= Simpson’s inverse index of diversity.

Maize Tissue Plant age Location genotype Alpha Simpson

Leaf 2-weeks Harrow bxbx 1.64 2.27

B37 2.48 2.49

W22 2.46 2.23

Ridgetown bxbx 2.72 4.36

B37 2.90 4.96

W22 2.26 4.29

9-weeks Harrow bxbx 0.96 1.61

B37 1.32 1.89

W22 1.18 1.92

Ridgetown bxbx 0.99 1.87

B37 1.04 2.04

W22 0.94 2.14

Root 2-weeks Harrow bxbx 7.49 11.23

B37 5.53 8.69

W22 5.66 8.31

Ridgetown bxbx 4.47 9.23

B37 5.99 9.11

W22 7.12 10.78

9-weeks Harrow bxbx 4.36 6.81

B37 4.41 8.09

W22 5.48 8.19

Ridgetown bxbx 2.40 6.02

B37 3.37 5.88

W22 4.44 6.11 CHAPTER 2. INFLUENCE OF BX PRODUCTION 40 ON ENDOPHYTE COMMUNITIES

Harrow Ridgetown

Leaf Communities of 2-week old plants

bxbx B37 W22 bxbx B37 W22 bxbx 0.85 0.924 bxbx 0.704 0.541 B37 0.125 0.939 B37 0.375 0.894 W22 0.286 0.333 W22 0.429 0.5

Leaf Communities of 9-week old plants

bxbx B37 W22 bxbx B37 W22 bxbx 0.925 0.952 bxbx 0.98 0.943 B37 0.333 0.997 B37 0.4 0.97 W22 0.4 0.75 W22 0.6 0.75

Root Communities of 2-week old plants

bxbx B37 W22 bxbx B37 W22 bxbx 0.527 0.573 bxbx 0.547 0.439 B37 0.429 0.864 B37 0.357 0.514 W22 0.375 0.455 W22 0.286 0.267

Root Communities of 9-week old plants

bxbx B37 W22 bxbx B37 W22 bxbx 0.863 0.913 bxbx 0.75 0.828 B37 0.643 0.845 B37 0.545 0.753 W22 0.5 0.471 W22 0.545 0.429

Figure 2.2. Similarity coefficients of fungal endophyte communities from leaf and root tissue of maize grown in Harrow and Ridgetown, Ontario. Jaccard’s Index is based on species presence/absence and is below the diagonal. The Morisita-Horn Index incorporates species frequency data and is above the diagonal. Values in bold indicate the communities that are most similar to one another. CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 41 ENDOPHYTE COMMUNITIES Nigrospora oryzae, and members of the Fusarium oxysporum species complex had isolates in two adjacent tolerance threshold groups.

Partitioning of endophyte communities by BOA tolerance level The distribution of isolates in BOA tolerance threshold groups was determined for each community characterized from PDA (Fig. 2.3). Isolates of the species/morphotypes that had intraspecific variation in tolerance level were equally distributed in the adjacent tolerance threshold groups for the analyses.

A significant difference in the tolerance threshold groups distribution was seen in four of the eight treatments, in which BX– plants had a significantly greater proportion of isolates in the 0.25 tolerance threshold groups than did BX+ plants (Table 2.3, Fig. 2.3). These four treatments were communities from seedling roots in Ridgetown (bxbx by W22: χ2 = 7.64, P = 0.0219; bxbx by B37: χ2 = 23.71, P = <0.00001) and in Harrow (bxbx by W22: χ2 = 13.64, P = 0.0041; bxbx by B37: χ2 = 11.83, P = 0.0027), and communities from 9-wk-old leaves in Ridgetown (bxbx by W22: χ2 = 10.95, P = 0.012; bxbx by B37: χ2 = 10.21, P = 0.0169) and in Harrow (bxbx by W22: χ2 = 12.96, P = 0.0047; bxbx by B37: χ2 = 5.94, P = 0.0148). In communities from seedling roots in Harrow, BX+ plants had a significantly higher proportion of isolates in the 0.50 tolerance threshold group compared with BX− plants (Table 2.3). In seedling roots, BX− plants had dominant species in the 0.25 tolerance threshold group (Alternaria alternata in Harrow and Periconia macrospinosa in Ridgetown), while BX+ plants had dominant species in the 0.50 and 0.75 tolerance threshold groups (members of the Fusarium equiseti-incarnatum species complex and the F. oxysporum species complex in both locations). In communities from 9-wk-old leaves, A. alternata was dominant in the three plant genotypes, but was isolated less frequently in BX– than in BX+ plants.

Abundance of Fusarium in 9-wk-old plants To compare the difference in relative abundance of Fusarium isolates per plant, the mean number of colonies isolated on BOA medium from the BX+ genotypes was compared with that from the BX– plants. A significant difference was detected in leaves (Harrow F CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 42 ENDOPHYTE COMMUNITIES Table 2.3. Results of χ2 tests for a difference between proportion of isolates in BOA tolerance threshold groups 0.25 (0.25 mg per ml BOA tolerance threshold) and 0.50 in endophyte communities from maize bxbx (BX– genotype), W22 and B37 (BX+ genotypes). Values in bold are statistically significant.

Tissue Plant age Location Tolerance Genotype Chi-square P-value threshold group comparison

Leaf 9-weeks Harrow 0.25 bxbx and W22 12.72 0.0004

bxbx and B37 11.42 0.0007

0.50 bxbx and W22 5.17 0.023

bxbx and B37 5.04 0.0248

Ridgetown 0.25 bxbx and W22 8.38 0.0038

bxbx and B37 23.43 <0.0001

0.50 bxbx and W22 0.43 0.513

bxbx and B37 10.78 0.001

Root 2-weeks Harrow 0.25 bxbx and W22 9.29 0.0023

bxbx and B37 5.24 0.022

0.50 bxbx and W22 6.64 0.01

bxbx and B37 0.84 0.3584

Ridgetown 0.25 bxbx and W22 5.25 0.022

bxbx and B37 8.52 0.0035

0.50 bxbx and W22 1.99 0.1586

bxbx and B37 6.71 0.0096

CHAPTER 2. INFLUENCE OF BX PRODUCTION 43 ON ENDOPHYTE COMMUNITIES

Harrow Ridgetown 2-week old plants 2-week old plants 1.0

0.8

0.6

0.4

0.2

0.0

1.0 9-week old plants 9-week old plants

0.8 * *

0.6

0.4

Proportion of isolates in leaf community leaf in isolates of Proportion 0.2

0.0

1.0 2-week old plants 2-week old plants 0.8 * 0.6 * *

0.4

0.2

0.0

1.0 9-week old plants 9-week old plants 0.8

0.6

0.4

Proportion of isolates in root community root in isolates of Proportion 0.2

0.0 0.25 0.50 0.75 1.00 1.10 0.25 0.50 0.75 1.10 1.10 Tolerance threshold (mg / ml) CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 44 ENDOPHYTE COMMUNITIES Figure 2.3. Distribution of fungal endophyte community members isolated on potato dextrose agar (PDA) into BOA tolerance threshold groups. The tolerance threshold is the highest concentration of 2-benzoxazolinone (BOA) that supported growth. Leaf and root endophyte communities from bxbx (BX–, circles), W22 (BX+, squares) and B37 (BX+, triangles) were characterized in 2-wk-old and 9-wk-old maize plants. Bold type indicates treatments indicated by global Chi-square tests to have communities distributed differently among tolerance threshold groups in BX+ and BX– plants. Asterisks show where follow-up Chi-square tests indicate that the proportion of isolates was significantly different between BX– and both BX+ genotypes. The dashed lines emphasize trends; they do not indicate a series of measurements over time.

CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 45 ENDOPHYTE COMMUNITIES

= 7.03, P = 0.0046; Ridgetown F = 9.12, P = 0.0014), but not in roots (Harrow F = 0.29, P = 0.7501; Ridgetown F = 0.08, P = 0.921). In leaf tissue, averaging values for the two BX+ genotypes, BOA-tolerant fungi were 14 times (Harrow) to 35 times (Ridgetown) more abundant in BX+ genotypes than in the BX− genotype (Fig. 2.4). From the isolates obtained on BOA medium that were identified, 96.5% were Fusarium species (Table 2.4). A total of nine Fusarium species, one isolate of Nigrospora oryzae and one isolate of Trichocladium sp. were identified. Of these species, six were isolated from leaf and root tissue, four from roots but not from leaves, and one only from leaf tissue. The most abundant species were members of the F. oxysporum species complex and the F. equiseti-incarnatum species complex.

DISCUSSION

Presence of BXs influences endophyte community structure These results suggest that BX production contributes to the structuring of endophyte communities in seedling roots and 9-wk-old leaves (Fig. 2.3). The differences over time in these tissues could arise from temporal changes in BX byproduct concentrations and allocation within tissue. Concentrations of constitutive DIMBOA and related compounds have been demonstrated to decrease with shoot and root age (Cambier et al. 2000). This would explain why there were more BOA tolerant fungi in BX+ than in BX− roots in the seedling, but not in 9-wk-old plants.

Fungal-mediated changes in MBOA concentration may be a factor in the relatively high abundance of BOA sensitive isolates in leaves of 9-wk-old BX− plants compared with BX+ plants. Oikawa et al. (2004) observed that inoculation of A. alternata in mature maize leaves induced production of HDMBOA-glucoside, a proposed precursor to MBOA. Tolerance of fungi to MBOA is positively correlated with tolerance to BOA (Glenn et al. 2001). In the present study, A. alternata (0.25 tolerance threshold group) was isolated more frequently from BX– leaves than BX+ leaves in 9-wk-old plants. Induction of HDMBOA-glucoside production by A. alternata could result in release and CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 46 ENDOPHYTE COMMUNITIES

Table 2.4. Identity of a subset of isolates obtained from 9-wk-old maize tissue plated on 2-benzoxazolinone (BOA) medium. Taxon Total Total Total identified from leaves from roots

Fusarium oxysporum species complex 27 5 22

Fusarium incarnatum- equiseti species complex 8 4 4

Fusarium subglutinans 4 3 1

Fusarium graminarum 4 1 3

Fusarium culmorum 3 0 3

Fusarium sporotrichoides 4 3 1

Fusarium proliferatum 2 0 2

Fusarium verticillioides 2 1 1

Fusarium solani 1 0 1

Nigrospora oryzae 1 1 0

Trichocladium sp. 1 0 1

CHAPTER 2. INFLUENCE OF BX PRODUCTION 47 ON ENDOPHYTE COMMUNITIES

3.0 a. 2.5 A A A 2.0 A 1.5

1.0

0.5 B B 0.0 3.0 C C b. C C 2.5 C C 2.0

1.5

Mean number of isolates per plant per isolates of number Mean 1.0

0.5

0.0 bxbx W22 B37 bxbx W22 B37 Harrow Ridgetown

Figure 2.4. Mean number of isolates per plant obtained on 2-benzoxazolinone (BOA)-medium from leaf (a) and root (b) tissue of 9-wk-old maize plants in Harrow and Ridgetown, Ontario, Canada.

Based on identification of a sample of these isolates, approx. 96.5% of isolates obtained on BOA medium were Fusarium species. Tukey–Kramer HSD tests were conducted to compare mean number of isolates per plant obtained from bxbx (BX–), W22 (BX+), and B37 (BX+) genotypes.

Analysis of leaf and root tissue was conducted separately. The same letter above two columns indicates no significant difference between means. Vertical bars, ± SE. CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 48 ENDOPHYTE COMMUNITIES accumulation of MBOA, which would be self-inhibitory, allowing colonization by other species with higher BOA/MBOA tolerance.

Genotypes of a host species can vary significantly in their influence on community organization (Whitham et al. 2006). Assessing the influence of Ustilago maydis resistance on endophyte community assembly in maize, Pan et al. (2008) found that community structure was not correlated with resistance, but rather with host genotype. Consistent with this, our results demonstrate differences in endophyte communities among all three genotypes, but with striking differences in tolerance of host toxins among community members between the BX− mutant and the BX+ genotypes.

McGill et al. (2006) propose that the most direct route to understanding the mechanisms underlying assembly of speciose communities is through study of functional trait variation across environmental gradients. Here, we have compared the extremes of the BX concentration gradient. Traditional species diversity measures did not detect a difference between endophyte communities from BX+ and BX− genotypes. When the distribution of functional traits in communities was analyzed, clear differences between the communities of BX+ and BX– genotypes were apparent. Such a shift in functional diversity, despite no change in species diversity, has frequently been noted in plant communities (for review of plant functional types in ecology see Duckworth et al. 2000).

BX+ plants have a higher incidence of Fusarium than BX− plants Fusarium was isolated 14 to 35 times more frequently from BX+ leaves than from BX− leaves in 9-wk-old plants. This suggests that Fusarium species are more competitive in the presence of BXs. Successive cropping of maize or other BX-producing plants in the same locality may increase the dominance of Fusarium in the endophyte community. Carryover across seasons would be expected through crop residue, which is considered to be the most important source of inoculum for endophyte infection (Sutton 1992). Our data suggest that crop residue of BX+ genotypes will have more Fusarium inoculum than the BX– genotype. If BX concentrations are maintained in crop residue and soil, our data suggest that Fusarium is likely to accumulate. Contamination of maize with Fusarium CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 49 ENDOPHYTE COMMUNITIES species can lead to human disease, yield loss and livestock toxicosis (Ueno et al. 1997; Marasas 2001; Marasas et al. 2004; Wu 2007). Reducing inoculum is the most direct approach to reducing fungal infection (Jouany 2007).

Commercial maize has been selectively bred to produce elevated levels of BXs. Results from our study suggest that presence of BXs can significantly enhance the competitive ability of Fusarium species. Given that BXs are general phytoprotectants, commercial cultivation of BX− maize is not realistic. Investigation of a relationship between BX concentration and abundance of Fusarium in maize tissue could inform crop management strategies. It is possible that there is a threshold concentration of BXs that is high enough to provide insecticidal benefits to the plant, but low enough to obviate any colonization benefit to Fusarium species.

Commonalities between endophyte community ecology in agricultural and naturally occurring plants There is an assumption that endophyte communities of agricultural plants and plants in their natural habitat are fundamentally different. Likely this stems from emphasis on individual fungal species as pathogens in agricultural crops, versus a more holistic approach to understanding endophyte communities of wild plants. However, we see four major commonalities between endophyte communities in cultivated and naturally occurring plants. First, the present study found that frequency of endophyte infection increases with plant age, with infection density of leaves reaching 100% in 9-wk-old leaves. Second, we found that in all treatments, diversity decreased over time (Table 2.2). Both of these trends have also been observed in endophyte communities of tropical plants (Herre et al. 2007). Third, results presented here indicate that above- and belowground tissue harbor distinct endophyte assemblages, a pattern observed in a diversity of plants (e.g. Kumar and Hyde 2004). Finally, plant residue is considered the most important source of fungal inoculum in the life histories of both agricultural and wild endophytes (Sutton 1992; Herre et al. 2007).

Another potential similarity between endophyte communities of plants in agricultural and CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 50 ENDOPHYTE COMMUNITIES natural environments is the role of interspecific interactions between fungi in mediating community structure. Arnold and Herre (2003) proposed that interspecific competition mediated by leaf chemistry is a common mechanism shaping endophyte communities. Our results are consistent with this hypothesis. Fusarium was significantly more abundant in BX+ plants compared with BX– plants, indicating that BXs provide a competitive advantage to Fusarium. Observational data from the present study indicate that interspecific facilitation may also influence endophyte communities. When isolating fungi from leaf tissue on BOA medium, colonies of A. alternaria and P. macrospinosa occasionally emerged from the colony center of BOA detoxifying Fusarium species. All of the A. alternaria and P. macrospinosa tested, including isolates obtained on BOA medium (1.00% concentration), could not grow above a 0.25% concentration of BOA. This may indicate that BOA detoxifying Fusarium species can facilitate the growth of less tolerant species. A previous in vitro study found that maize endophytes able to tolerate 1.00% and 1.10% concentrations of BOA facilitated growth of maize endophytes unable to grow above a 0.25% and 0.50% concentration (Saunders and Kohn 2008). These commonalities between agricultural and wild endophyte communities may represent general mechanisms in endophyte community assembly, and would therefore be constructive areas for future research.

CONCLUSIONS

We found that plant defense compounds are a significant factor in structuring fungal endophyte communities of maize. Glenn et al. (2001) proposed that BX production by maize enhances the ecological success of Fusarium species in maize; our results support this hypothesis. Further, we found that non-Fusarium species with intermediate BOA tolerance levels had a colonization advantage over BOA-sensitive fungi in BX-producing plants, indicating that a large proportion of the fungal community is influenced by defense compound production.

Our data point to the possibility that breeding for elevated concentrations of BXs in maize has unintentionally allowed for increased colonization by Fusarium and the CHAPTER 2. INFLUENCE OF BX PRODUCTION ON 51 ENDOPHYTE COMMUNITIES possibility of increased inoculum load in plant residue and in soil. These results are preliminary to the next logical step of breeding isogenic corn lines that only differ in quantitative production of BXs. The caveat to this approach will be that concentrations of BX are not static in planta, and are also likely to be dynamic in the surrounding soil and crop residue.

Future experiments on endophyte communities of other host species are needed to test the possibility that host defense compounds are a general mechanism in structuring communities. Studies incorporating measurements of multiple factors, such as competition and facilitation between microbes, species composition in the soil-borne ‘spore-bank’ and microbial niche overlap will help to identify key fungal traits in colonization. From this, we will be better able to form hypotheses about evolution of host colonization strategies within a community context.

CHAPTER 3

PRODUCTION OF DEFENSE COMPOUNDS BY PLANT NEIGHBORS DECREASES FUNGAL ENDOPHYTE ABUNDANCE AND DIVERSITY IN MAIZE ROOTS

CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 53 ENDOPHYTE COMMUNITIES ABSTRACT

An individual plant can affect surrounding plants by manipulating the abundance of pathogenic or mutualistic fungi in the common habitat. Despite this, little is known about the influence of plant neighbors on fungal endophyte communities. Here, we assess the response of fungal endophyte communities to production of defense compounds by neighboring maize plants. Maize produces benzoxazinoids (BXs), compounds with byproducts toxic to insects, bacteria, fungi and plants. A maize genotype that does not produce BXs (BX-) was grown both in monoculture, and in fields with two BX- producing maize genotypes (BX+). Fungal community members were isolated from leaf and root tissue of the mature, BX- plants. Fungal species were assessed for tolerance to 2-benzoxazolinone (BOA), a toxic BX byproduct. Species and functional diversity were compared between fungal communities from the BX- monoculture and BX- plants grown in triculture. Functional diversity is here defined as the distribution of BOA tolerance levels within communities. Fungal communities from the roots of BX-, monoculture- grown plants had higher species diversity, overall endophyte abundance, and abundance of common maize inhabiting species in the genus Fusarium, than communities from BX- plants grown in triculture. In contrast, leaf endophyte communities from monoculture- cultivated hosts were not significantly different from triculture-cultivated hosts. Functional diversity was significantly different between monoculture and triculture only in leaf tissue in one location, indicating that functional diversity is not influenced by the genotype of host neighbors. Results suggest that presence of conspecific plant neighbors results in a decrease in endophytic colonization of BX- plant roots. This raises the hypothesis that exudation of BX compounds from roots influences the community of soil fungi.

CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 54 ENDOPHYTE COMMUNITIES INTRODUCTION

An individual plant can indirectly affect neighboring plants by altering the abundance of their microbial associates (Burdon et al. 2006; Power and Mitchell 2004). Variation in plant traits such as susceptibility to pathogen colonization and concentration of allelochemicals can influence neighboring plant genotypes and species. Plants can experience “associational susceptibility” or “associational resistance” when presence of neighboring plants cause a change in the amount of herbivore- or pathogen-induced damage that is sustained (Futuyma and Wasserman 1980; Power and Mitchell 2004; Burdon et al. 2006; Russell et al. 2007). A common example of associational susceptibility is when a “reservoir” host accumulates a high density of infective propagules that are in turn transmitted to neighboring plants (Power and Mitchell 2004; Burdon et al. 2006). Plants can also affect neighboring species or genotypes by reducing the abundance of mutualistic fungi. For example, garlic mustard (Alliaria petiolata), which is non-mycorrhizal, can dramatically reduce the activity of mycorrhizal fungi in soil, thereby suppressing the growth of native tree species (Stinson et al. 2006; Wolfe et al. 2008). Evidence strongly suggests that the mechanism of this suppression is the release of fungitoxic compounds into the soil by garlic mustard (Stinson et al. 2006).

Plants can acquire associational resistance when abundance of herbivores or pathogens is decreased in the vicinity of plant neighbors. In agriculture there is ample, though conflicting, evidence for reduced disease incidence with intraspecific or interspecific multicropping (Burdon et al. 2006). In disease management one approach is to include a genotype or species that produces defense compounds with the aim of decreasing soilborne inoculum and infection of subsequent crops (e.g. Khanh et al. 2005).

In the study reported here, we used maize (Zea mays) to investigate the influence of host genotypes that produce defense compounds on the fungal endophyte communities of neighboring genotypes that do not produce them. The main defense compounds produced by maize are the benzoxazinoids (BXs), and those that are typically produced in the highest concentration are 2,4-dihydroxy-7-methoxy-2H-1,4-benzoxazin-3-one and CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 55 ENDOPHYTE COMMUNITIES 2,4-dihydroxy-2H-1,4-benzoxazin-3-one. These compounds reside in the cell vacuole as biologically inactive beta-glucosides, and upon plant cell disruption are converted to the aglyconic forms. The aglycones are highly toxic but unstable, and rapidly degrade to the benzoxazolinones, 6-methoxy-2-benzoxazolinone (MBOA) and 2-benzoxazolinone (BOA), respectively (Hashimoto and Shudo 1996). Both MBOA and BOA are widely toxic to microbes, insects and plants (Barry and Darrah 1991; Hashimoto and Shudo 1996; Niemeyer and Perez 1995). BX byproducts are exuded from plants into soil; persistence of these products in soil can change with environmental conditions (Krogh et al. 2006).

Fungal species vary widely in their tolerance to benzoxazolinones [Friebe et al. 1998; Glenn et al. 2001; Saunders and Kohn 2009 (Chapter 2)]. Of those tested, among the most tolerant are species of Fusarium. The abundance of Fusarium in maize is of particular interest, because several species in this genus asymptomatically reside in maize while producing toxins that are a risk to human and animal health, causing severe economic loss worldwide (Ueno et al. 1997; Marasas 2001; Marasas et al. 2004; Wu and Munkvold 2008). Production of BXs has been shown to provide a colonization advantage for Fusarium in maize. In a comparison between leaf tissue of BX+ maize genotypes and a genotype that does not produce BXs, abundance of Fusarium was up to 35 times higher in BX+ plants [Saunders and Kohn 2009 (Chapter 2)].

In the present study, we compared fungal endophyte communities of BX- plants that were grown in monoculture with communities from BX- plants grown in fields with two BX+ genotypes (triculture). The comparison between endophyte communities from mono- and triculture was unplanned, but later caught our attention. With one field of each treatment in two locations, these data are pseudoreplicated (Hurlbert 1984). Distance between the mono- and triculture fields was approximately 4.5 meters in Harrow and Ridgetown. To the best of our knowledge there is no reason to assume that there is environmental variation that would contribute to a difference between fields in the same location.

CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 56 ENDOPHYTE COMMUNITIES We addressed several questions: 1) Do leaf and root endophyte communities respond differently to these monoculture and triculture host neighborhoods? Root tissue is typically colonized by soil fungi, and leaves by airborne propagules. It is therefore expected that if BXs are present in soil, they will only alter root endophyte communities. 2) Do hosts grown in the triculture have a lower endophyte infection density than those grown in the monoculture? If the BX+ genotypes suppress fungal growth in soil, then we expect that triculture plants will have a lower abundance of endophytes in tissue than monoculture grown plants. 3) Does the presence of BX+ neighbors influence the species and functional diversity in fungal endophyte communities of BX- monoculture plants? It is expected that if BXs are present in soil, they will reduce growth of fungi that are sensitive to BXs, thereby decreasing diversity in triculture plants as compared to monoculture plants. The functional diversity measure considered here is the distribution of BOA tolerance levels of fungi within communities. 4) Does production of BXs by host neighbors influence abundance of Fusarium in tissue? Fusarium species are expected to be more abundant in root tissue of triculture grown plants, because many species are tolerant to toxic BX byproducts.

MATERIALS AND METHODS

Study site, maize varieties and tissue collection Maize fields were planted on June 12, 2005 in Ridgetown and Harrow, Ontario. One triculture and one monoculture field was planted in each location. Three genotypes were planted in the triculture fields: W22 and B37, genotypes that produce BXs, and bxbx, the only recorded natural mutant lacking the ability to produce BXs (Hamilton 1962). The triculture and monoculture fields were of the same dimensions. In triculture each genotype was planted in 12 rows, with one genotype per row, for a total of 36 rows. Assignment of rows within the field was random. In the monoculture field, BX- plants were planted in 36 rows. The Ridgetown location had a previous history of soybean crops in 2001-2004, and was surrounded by soybean, which does not produce BXs. The Harrow location was surrounded by BX-producing maize, and has a history of maize cultivation in 1999-2004. At the time of collection, the 9-wk old plants were entering the CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 57 ENDOPHYTE COMMUNITIES R3 stage of growth, approximately 3 wk prior to physiological maturity (Ritchie et al. 1993). This study is part of a larger study that evaluated endophyte communities from all three genotypes in the triculture only [Saunders and Kohn 2009 (Chapter 2)].

Isolation of fungi from root and leaf tissue BX- plants were rinsed with distilled water and surface dried at room temperature. From each plant sampled, 4 healthy leaf and 4 healthy root segments were incubated on two growth media: potato dextrose agar (PDA) or BOA medium [15 g peptone (Difco), 1 g

KH2PO4, 0.5 g MgSO4 - 7H2O, 0.3 mg streptomycin, 20 g agar, 1 L H2O, 1 g BOA (Sigma)], (Glenn et al. 2001). PDA (Difco, Detroit, MI) amended with antibiotics (1.00 g per L streptomycin sulfate and 0.25 g per L neomycin sulfate, Sigma Chemical Co., St. Louis, MO) was used to isolate a wide diversity of fungi, and BOA medium is a Fusarium-selective medium that was used to quantify the mean number of Fusarium colonies per plant. PDA has been shown to favor growth of more fungi that are BOA- sensitive than BOA medium [Saunders and Kohn 2009 (Chapter 2)]. Leaf segments measuring 1.0 x 2.0 cm, and root segments 2.0 cm in length and 0.2-0.3 cm in diameter were cut, surface sterilized and incubated as previously described [Saunders and Kohn 2009 (Chapter 2)]. Segments from a total of 48 plants were plated on PDA, including one plant per row (9-12 plants per treatment were analyzed). A total of 96 plants were plated on BOA medium, including two plants per row (16-24 plants per treatment were analyzed). Effectiveness of the surface sterilization procedure was tested as described in Saunders and Kohn (2009) (Chapter 2).

Identification of fungal isolates Fungi emerging from plant tissue were isolated and established in pure culture and identified taxonomically or grouped into morphotypes. Each Fusarium isolate was established in pure culture from a single spore and identified morphologically, or molecularly when morphological characters were not sufficient for identification at the species level. To identify Fusarium morphologically, isolates were grown on Carnation leaf agar and PDA (Summerell et al. 2003; Leslie et al. 2006). For each Fusarium species the identity of a subset of isolates was verified with DNA sequence data. All CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 58 ENDOPHYTE COMMUNITIES isolates obtained from PDA were morphotyped, and isolates recovered more than 3 times were identified taxonomically. To determine the range of Fusarium species isolated on BOA medium a subset of 35 isolates were identified using DNA sequence data.

DNA isolation, PCR amplification and sequencing were accomplished as described previously [Saunders and Kohn 2009 (Chapter 2)]. Contigs were assembled and edited in Sequencher 4.6 (Gene Codes Corporation, Ann Arbor, Michigan, USA), and BLAST searches of the NCBI GenBank database were conducted for tentative identification. A sequence match of 99% similarity to a sequence of known origin in the database was considered a valid species identification. DNA sequences will be deposited in GenBank.

Statistical analyses For all statistical analyses, each row was considered the unit of replication, and the two locations were analyzed separately. The number of isolates obtained on PDA was used to determine the overall endophyte abundance in leaf and root tissue of plants in monoculture as compared to triculture. One-way ANOVAs were used to compare the number of isolates from plants grown in mono- and triculture. Residuals did not significantly deviate from a normal distribution (Shapiro-Wilks test), and variances were homogeneous (Levene’s test). The threshold for statistical significance was P ≤ 0.05. Analyses were conducted using JMP (version 5.1, SAS Institute Inc., Cary, NC).

Isolates obtained on PDA were used to characterize eight fungal endophyte communities from BX- plants. We define a fungal community as a unique combination of location, treatment and tissue type (2 locations x 2 treatments x 2 tissue types). Again, each row was considered a replicate. Diversity of fungal endophyte communities isolated on PDA was estimated using Fisher’s α, because this diversity measure is robust for comparisons between treatments with uneven samples sizes (Magurran 2004). Fisher’s α was calculated using EstimateS (Colwell 2000).

We define functional diversity as the distribution of BOA tolerance threshold levels across communities. Fungal isolates were tested for BOA tolerance in the previous study CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 59 ENDOPHYTE COMMUNITIES [Saunders and Kohn 2009 (Chapter 2)]. Each species/morphotype isolated on PDA more than three times was assigned to a BOA tolerance threshold group defined by the highest concentration of BOA supporting growth. For each community, the distribution of isolates into BOA tolerance threshold groups was determined. To test for a difference between the distribution of isolates into tolerance threshold groups in plants grown in the monoculture as compared to the triculture, a likelihood ratio test was conducted using JMP. When necessary, results were then adjusted for small expected values using the formula (Gotelli and Ellison 2004):

2 2 X adjusted = X / qmin

where n is the number of rows (treatments), m is the number of columns (tolerance threshold groups), N is the total sample size, v is the degrees of freedom, and Yi,j is the frequency of observations in row i, column j.

The number of isolates obtained on BOA medium was used to estimate the abundance of Fusarium species in tissue. Results from previous studies indicate that approximately 97% of the isolates obtained on BOA medium are Fusarium species [Glenn et al. 2001; Saunders and Kohn 2009 (Chapter 2)]. One-way ANOVAs were conducted to test for differences between mean number of isolates obtained from plants grown in monoculture and triculture. Variances were homogeneous and residuals did not deviate from a normal distribution.

CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 60 ENDOPHYTE COMMUNITIES

Table 3.1. Species isolated from leaf and root tissue of BX- (bxbx) maize grown in monoculture and in triculture. Tissue was collected from Harrow and Ridgetown, Ontario, and plated on potato dextrose agar. 8 8 0 0 2 0 0 1 0 1 0 0 0 0 0 0 7 16 18 10 63 Ridgetown 6 2 2 0 0 0 2 2 0 0 0 0 0 0 0 0 8 14 26 10 64 Harrow Monoculture 7 9 6 6 2 0 0 1 0 0 0 0 1 1 0 0 0 1 11 45 10 Ridgetown 4 4 4 5 4 0 0 0 0 0 0 0 0 1 1 1 0 10 19 53 10 Root Harrow Triculture 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 3 40 22 64 Ridgetown 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 34 12 46 Harrow Monoculture 2 0 0 0 1 0 0 0 0 2 0 0 0 0 0 0 0 5 38 13 56 Ridgetown 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 4 46 24 72 Leaf Harrow Triculture 0.25 0.50 0.25 0.50 0.50 1.00 1.00 1.00 not tested not tested not tested not tested not tested not tested not tested not tested not tested not tested 0.50 and 0.75 BOA tolerance threshold (mg/ml) Species / morphotype A. alternata Epicoccum nigrum No. 14 (Pleosporales) Fusarium oxysporum species complex Fusarium equiseti-incarnatum species complex Cladosporium sp. Phoma sp. Fusarium graminearum Nigrospora oryzae No. 16 No. 39 No. 55 No. 56 Fusarium acuminatum Penicillium sp. Rhizopus sp. Trichoderma sp. No. 51 No. 32 Total isolates per community type Total species per community type CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 61 ENDOPHYTE COMMUNITIES RESULTS

A total of 746 fungal isolates were obtained, 464 on PDA, and 282 on BOA. A total of 19 species/morphotypes were identified from PDA (Table 3.1). Among these were isolates of the Fusarium oxysporum and Fusarium incarnatum-equiseti species complexes. When multiple isolates within a species complex were detected, the species complex was treated as a single species in the analysis.

Species richness and presence/absence differed between communities from leaf and root tissue on PDA, with root communities on average containing 2.5 times more species than leaf communities (Table 3.1). Leaf and root tissue had 33% species/morphotypes in common and root communities had 67% species/morphotypes present that were not isolated from leaves. Leaf communities did not have any unique species/morphotypes.

In root communities, species diversity and infection density were consistently, although not always significantly, higher in monoculture communities than in triculture (Fig. 3.1). There was no significant difference in the species diversity among leaf endophyte communities. Infection density can be measured in two ways, as the proportion of tissue segments yielding isolates, or as the number of colonies isolated per plant. Using both measurements, infection density was higher in root tissue of monoculture plants than in triculture. The proportion of tissue segments yielding isolates ranged from 45-100%. In root tissue, an average of 49% of tissue segments yielded colonies in triculture plants (53% in Harrow and 45% in Ridgetown), and 100% in monoculture plants in both locations. In leaf tissue, 98-100% of tissue segments yielded isolates. The mean number of colonies per plant isolated on PDA was significantly greater in monoculture roots than in triculture (Harrow: DF= 1, F= 11.41, P=0.003; Ridgetown: DF= 1, F=48.44, P=<0.0001; Fig. 3.1). In leaf tissue, endophyte abundance was greater in monoculture in Ridgetown, but there was no significant difference in mean number of colonies isolated from leaf tissue in mono- and triculture from Harrow (Harrow: DF= 1, F=0.3007, P=0.59; Ridgetown: DF= 1, F=7.75, P=0.0118).

CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 62 ENDOPHYTE COMMUNITIES

7 a. 6 5 4 3

per plant 2

Mean Fisher’s alpha 1 0

12 b. 10 * *

8 *

per plant 4

Mean number of colonies 0

7 c. * 6 5 4 3 2

colonies per plant 1

Mean number of Fusarium 0 Harrow Ridgetown Harrow Ridgetown Leaves Roots CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 63 ENDOPHYTE COMMUNITIES Figure 3.1. Species diversity, abundance of fungal endophytes, and abundance of Fusarium in BX- (bxbx) maize grown in monoculture and in triculture. Leaf and root tissue was collected from plants grown in Ridgetown and Harrow, Ontario. The dark grey bars = triculture and light grey bars = monoculture. Diversity of communities isolated on potato dextrose agar (PDA) was compared using Fisher’s alpha (a). Endophyte abundance is the mean number of colonies isolated per plant on PDA (b). The abundance of Fusarium is the mean number of isolates obtained on BOA medium, a Fusarium specific medium (c). Asterisks designate significant differences between communities from the monoculture and triculture using one-way ANOVAs. Bars indicate standard error.

CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 64 ENDOPHYTE COMMUNITIES

To compare the difference in relative abundance of Fusarium isolates per plant, the mean number of colonies isolated on BOA medium was compared between plants grown in monoculture and triculture (Harrow- Leaf: DF= 1, F=0.26, P=0.61; Ridgetown- Leaf: DF= 1, F=0.00, P=1; Harrow- Root: DF= 1, F=1.86, P=0.18; Ridgetown- Root: DF= 1, F=57.90, P=<0.0001). Only one significant difference was detected (Fig. 3.1). In root communities from Ridgetown, plants grown in monoculture had a significantly higher abundance of Fusarium as compared to plants grown in triculture. Of the thirty five isolates from BOA medium that were identified, all were Fusarium species. Twenty five were identified as members of the F. oxysporum species complex, five as members of the F. incarnatum-equiseti species complex, two isolates as F. acuminatum, one isolate each as F. culmorum, F. subglutinans and F. verticillioides.

The proportion of community members in the BOA tolerance threshold groups was determined for each community characterized (Table 3.1, Fig. 3.2). Isolates of members of the F. oxysporum species complex and of Nigrospora oryzae have intraspecific variation in tolerance levels with equal distribution in the adjacent tolerance threshold groups [Saunders and Kohn 2009 (Chapter 2)]. Isolates of these species were equally distributed in the adjacent tolerance threshold groups for the analyses. A significant difference in the tolerance threshold group distribution was seen in 1 of the 4 tests (Harrow- Leaf: DF=2, X2=13.95, P=0.0009; Ridgetown- Leaf: DF= 1, X2=1.81, P=0.18; Harrow- Root: DF= 2, X2=0.18, P=0.92; Ridgetown- Root: DF= 3, X2=2.10, P=0.56). In leaf communities in Harrow, plants grown in triculture had a significantly greater proportion of isolates in the most sensitive tolerance threshold group (species with a BOA tolerance threshold of 0.25 mg per ml) than did plants grown in monoculture.

DISCUSSION

BX- plants grown in monoculture had a higher overall abundance of endophytes in root tissue than did triculture grown plants. In addition, in Ridgetown, roots of monoculture plants had a higher abundance of Fusarium species than did triculture grown plants. Data CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 65 ENDOPHYTE COMMUNITIES

Harrow Ridgetown 1.0 * 0.8

0.6

Leaf 0.4

0.2

0.0

1.0

0.8 Proportion of community of Proportion

Root 0.6

0.4

0.2

0.0 0.25 0.50 0.75 1.00 1.10 0.25 0.50 0.75 1.00 1.10 BOA tolerance threshold (mg/ml)

Figure 3.2. Distribution of fungal endophytes into BOA tolerance threshold groups.

Leaf and root tissue was collected from bxbx maize (BX-) in Harrow and Ridgetown,

Ontario. Plants were grown in monoculture, or in triculture with two maize genotypes that produce benzoxazolinones. The circles = triculture and squares = monoculture. The asterisk designates a significant difference in Chi-square tests of the distribution of isolates in tolerance threshold groups among communities of monoculture and triculture grown plants. The dashed lines emphasize trends; all points connected by a line sum to 1. CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 66 ENDOPHYTE COMMUNITIES from a related field study indicated that BXs are a habitat filter advantageous to colonization by Fusarium species [Saunders and Kohn 2009 (Chapter 2)], therefore, the expectation was that Fusarium species would be more abundant in the triculture grown plants. Instead, data suggest that endophyte abundance, including that of Fusarium species, was greater in monoculture roots.

In the present study we did not measure the growth of host plants, and therefore did not directly assess associational resistance or susceptibility. However, presence of BX+ genotypes increased the resistance of BX- plants to overall colonization by asymptomatic endophytes, and the increase in abundance of Fusarium species has the potential to influence host plant health. Although the Fusarium species isolated in this study appeared to be asymptomatic endophytes, several species (e.g. Fusarium graminearum) can become pathogenic when the plant is physiologically compromised (Dodd 1980).

Endophyte communities from roots of BX- plants grown in monoculture also had higher species diversity than communities from triculture plants. Such a trend would be expected if BX byproducts moved from plants into the soil, reducing the number of fungal taxa available to colonize root tissue. Amendment of soil with root exudates is known to drastically alter the community structure of soil fungi. Broeckling et al. (2008) evaluated soil microbe communities that were exposed to four treatments: three generations of a plant species that naturally occurs in the soil (resident), three generations of a nonresident species, and the root exudates of each plant species. Species richness of microbial communities changed in response to the nonresident plant and its exudates, but did not change in response to the resident plant. The nonresident plant had positive and negative affects on the abundance of specific fungal taxa, and the authors hypothesized that components of the root exudates acted as signaling molecules for some taxa, and as defense compounds for others. The matrix of root exudates includes secondary compounds, water, mucilage and enzymes, all of which have the potential to influence endophyte colonization. In the present study it is possible that the differences in CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 67 ENDOPHYTE COMMUNITIES endophyte communities in mono- and triculture grown plants were mediated by components of root exudates other than BXs.

Surprisingly, although differences in endophyte abundance and diversity were seen, no clear trends in the distribution of BOA tolerance levels were detected between communities of monoculture and triculture plants. In a related study, leaf communities of mature BX- plants had a significantly greater abundance of BOA sensitive isolates than did BX-producing genotypes [Saunders and Kohn 2009 (Chapter 2)]. From this, we expected that root communities of plants grown in triculture, where there is a potential for species to be in contact with BXs in soil prior to host colonization, would have a greater abundance of fungi with high BOA tolerance relative to monoculture grown plants. However, we detected only one difference in BOA tolerance distribution between communities in mono- and triculture plants, and the difference was not in expected direction; in leaf communities in Harrow, triculture grown plants harbored more BOA sensitive isolates than monoculture-grown plants. The absence of a connection between neighborhood treatment and BOA tolerance distribution of endophyte communities suggests that concentration of BXs in soil were not high enough to affect the BOA tolerance distribution profile of endophyte communities in BX- plants.

If the mechanism for the neighborhood effect observed in this study is accumulation of toxins in soil, it is likely that endophyte communities of leaves will remain unaffected relative to root communities, as was observed. Aerial tissue, predominantly colonized by airborne propagules, and belowground tissue, by soil-borne propagules, typically harbor distinct microbial communities (for review see Arnold 2007). As well, physiological barriers could impede microbial movement between below- and above-ground tissues in systemic colonization. Oren et al. (2003) used GFP-expressing transgenic isolates of the common maize endophyte F. verticillioides to visualize early stages of the maize-fungus interaction in planta, using soil-borne inoculum. After 21 days, infection density was highest in the mesocotyl. Colonization of aboveground tissue was sparse, suggesting a barrier to systemic colonization by root endophytes, and a potential mechanism for delineation of root and shoot endophyte communities. CHAPTER 3. AFFECT OF PLANT NEIGHBORS ON 68 ENDOPHYTE COMMUNITIES

The difference in genetic diversity of the two field types could also have influenced the endophyte species diversity in monoculture and triculture plants. Results from previous studies indicate that the genotypic composition of the plant neighborhood can influence the species diversity of associated microbes or herbivores (Hughes et al. 2008). Often, an increase in genetic diversity of the host community is positively correlated with an increase in parasite diversity (e.g. Johnson et al. 2006). While we did not specifically test the influence of host genetic diversity on fungal communities, we did compare two levels of host diversity. In contrast to the positive correlation between host and parasite diversity that has been noted in previous studies, we found that the less diverse host community supported the most diverse fungal communities.

We found that species diversity and overall abundance of root fungal endophytes was greater in a neighborhood of one genotype of BX-nonproducing plants than in a triculture of BX- and BX+ genotypes. Abundance of Fusarium species was higher in monoculture grown plants than in triculture grown plants. Factors influencing Fusarium colonization are of interest because several species, including some isolated in this study (e.g. F. verticillioides and F. subglutinans) can colonize grain and produce toxins that cause mycotoxicosis in animals, and are suspected risk factors for several human diseases (Ueno et al. 1997; Marasas 2001; Marasas et al. 2004). Circumstantial evidence suggests that defense compounds present in soil impacted the common soil microbial community, ultimately causing a decrease in the species diversity of root endophyte communities.

CHAPTER 4

HOST-SYNTHESIZED SECONDARY COMPOUNDS INFLUENCE THE IN VITRO INTERACTIONS BETWEEN FUNGAL ENDOPHYTES OF MAIZE

Previously published as: Saunders M and Kohn LM. Host-synthesized secondary compounds influence the in vitro interactions between fungal endophtyes of maize. Applied and Environmental Microbiology. 2008 January; 74: 136-142. This chapter is reproduced with permission granted from © Applied and Environmental Microbiology.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 70

ABSTRACT

Maize produces a suite of allelopathic secondary metabolites, the benzoxazinoids. 2,4- Dihydroxy-7-methoxy-2H-1,4-benzoxazin-3-one and 2,4-dihydroxy-2H-1,4-benzoxazin- 3-one reside as glucosides in plant tissue and spontaneously degrade to 6-methoxy-2- benzoxazolinone (MBOA) and 2-benzoxazolinone (BOA) upon plant cell disruption. Several maize-associated fungi in the genus Fusarium can metabolize MBOA and BOA. BOA tolerance levels in 10 species of Fusarium and in the maize endophytes Nigrospora oryzae, Acremonium zeae, and Periconia macrospinosa were characterized. BOA tolerance ranged from 0.25 to 1.10 mg per ml among species. The influence of substrate alteration by one species on the subsequent growth of another species was assessed in the presence and absence of BOA. The colony area of the secondary colonizer in heterospecific interactions was compared to that in autospecific interactions (one isolate follows itself). In the presence of BOA, four of six secondary colonizers had greater growth (facilitation) when primary colonizers had higher BOA tolerance than the secondary colonizer. When the primary colonizer had lower tolerance than the secondary, three of six secondary colonizers were inhibited (competition) and three not significantly affected. In BOA-free medium, the number of isolates that were facilitated or inhibited was the same regardless of the tolerance level of the primary colonizer. Two of six secondary colonizers were facilitated, two inhibited, and two not significantly affected. This study provides some support for facilitation in stressful conditions under the Menge-Sutherland model. The results are not consistent with the corresponding prediction of competition in the absence of stress. The hypothesis drawn from these data is that in the presence of a toxin, fungal species that detoxify their substrate can enhance the colonization rate of less tolerant fungi.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 71

INTRODUCTION

Maize harbors a diverse community of nonmycorrhizal fungal endophytes (Fisher et al. 1992). Several species in the fungal genus Fusarium, notably F. verticillioides, F. subglutinans, F. proliferatum, and F. graminearum, are commonly associated with maize as pathogens or asymptomatic endophytes. These species can produce secondary metabolites that are toxic to livestock, including moniliformin, fumonisin, and deoxynivanol (Marasas et al. 1984; Desjardins and Proctor 2001). F. verticillioides is frequently encountered as an asymptomatic maize endophyte, but as a pathogen it can cause seedling blight and stem, seed, root, and ear rot. This species is globally distributed and has a wide host range, including sorghum, millet, and sugar cane, but it is most frequently isolated from maize. In some localities, maize fields have rates of greater than 90% infection with F. verticillioides, which can be transmitted via seed, crop residue in soil, or airborne conidia (Munkvold et al. 1997). Other common preharvest endophytes include Acremonium zeae, Periconia macrospinosa, and Nigrospora oryzae (Wicklow et al. 1998). Maize has several mechanical and protein-based mechanisms of resistance to microbial invasion (Cordero et al. 1994; Warfield and Davis 1996; Bravo et al. 2003; Chen et al. 2006). In addition, maize contains a suite of toxic secondary metabolites, the benzoxazinoids (also known as cyclic hydroxamic acids). The primary benzoxazinoids found in maize are 2,4-dihydroxy-7-methoxy-2H-1,4-benzoxazin-3-one (DIMBOA) and its desmethoxy derivative, 2,4-dihydroxy-2H-1,4-benzoxazin-3-one (DIBOA). These compounds originate in the cell vacuole as biologically inactive β-glucosides, 2-[2,4- dihydroxy-7-methoxy- 2H-1,4-benzoxazin-3(4H)-one]-β-D-glucopyranoside and its desmethoxy derivative, and are enzymatically converted to the biologically active benzoxazinoids upon cell disruption (Hashimoto and Shudo 1996). The benzoxazinoids are unstable in aqueous solution and spontaneously degrade to the corresponding benzoxazolinones, 6-methoxy- 2-benzoxazolinone (MBOA) and 2-benzoxazolinone (BOA) (Fig. 4.1). The benzoxazolinones are biologically active and stable in aqueous solution. They can be allelopathic and protect against herbivory, fungal, and bacterial infection (Long et al. 1977; Corcuera et al. 1978; Bravo et al. 1997, Yan et al. 1999; Macias et al. 2006). Several species within the Poaceae, including CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 72

Figure 4.1. Benzoxazinoids and their degradation products. The biologically inactive compounds DIMBOA-Glu and DIBOA-Glu reside in plants and are enzymatically converted by the plant to the fungitoxic aglucones DIMBOA and DIBOA. These compounds spontaneously degrade to the biologically active compounds MBOA and BOA. Adapted from Morrissey and Osbourn (1999) with permission.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 73 wheat, rye, maize, teosinte, and some wild barley species, produce benzoxazinoids (Niemeyer 1988; Sicker et al. 2000). DIMBOA and DIBOA are synthesized upon germination and accumulate within the tissues of the growing seedling, thereby providing a constitutive defense against invasion (Klun and Robinson 1969; Cambier et al. 2000). The highest levels of DIMBOA and DIBOA are found in the young seedling, and the concentrations of these compounds decrease linearly with age. The concentration of DIMBOA has been shown to peak at between 1 to 10 mmol per kg fresh weight at seven d after germination. At approximately 15 d, the concentration of DIMBOA decreases to about 0.3 to 3 mmol per kg fresh weight (Argandona and Corcuera 1985). DIMBOA and DIBOA are formed systemically and are actively exuded from root tissue (Sicker et al. 2001; Park et al. 2004). F. verticillioides is known to tolerate MBOA and BOA. Tolerance is associated with metabolic conversion of MBOA and BOA to N (2-hydroxy- 4-methoxyphenyl) malonamic acid and N-(2-hydroxy-phenyl)malonamic acid, respectively (Richardson and Bacon 1995; Yue et al. 1998). Other important maize- associated fungi, including F. subglutinans, F. proliferatum and F. graminearum, are also known to detoxify MBOA and BOA (Glenn et al. 2001). Fusarium species are highly variable in their tolerance to these compounds, with F. verticillioides demonstrating the highest tolerance. This characteristic has been suggested to increase the ecological fitness of F. verticillioides in maize fields (Glenn et al. 2001). A fungal species in host tissue may have a facilitative, competitive, or neutral effect on subsequent fungal infection. For example, in a study of 135 saprophytic fungi, Clonostachys rosea was found to suppress conidiation of several Fusarium species by up to 80% on maize stalks. Some species suppressed conidiation only slightly or not at all (Luongo et al. 2005). Given that F. verticillioides is often in maize seed, and can begin to inhabit tissue upon seedling germination, this species has the potential to influence the colonization success of subsequent fungal invaders. The objective of this work was to test the influence of BOA biotransformation on the outcome of indirect interactions between common maize- associated fungi. This work is a complement to ongoing field studies aimed at determining the influence of benzoxazinoid production on fungal colonization in maize fields. The approach taken was threefold. First, BOA tolerance levels in ten species of Fusarium and the common maize endophytes Nigrospora oryzae, Acremonium zeae, and CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 74

Periconia macrospinosa were characterized. Second, sequential inoculations were performed in vitro to determine the influence of substrate alteration by one species on the growth of another in the presence or absence of BOA. The effect of F. verticillioides as a primary colonizer on subsequent fungal colonization and the influence of BOA on the outcome of species interactions were determined. Third, the isolates used in sequential inoculation experiments were evaluated for their ability to degrade BOA in the medium.

MATERIALS AND METHODS

Characterization of BOA tolerance in maize endophytes. A panel of ten Fusarium species (two isolates of each species), two isolates of A. zeae, two isolates of N. oryzae, and two isolates of P. macrospinosa were scored for growth or no growth on a range of concentrations of BOA (Table 4.1). Potato dextrose agar (PDA) (Difco, Detroit, MI) was amended with BOA (Sigma Chemical Co., St Louis, MO) in each of the following concentrations of BOA with respect to PDA: 0.00, 0.25, 0.50, 0.75, 1.00, 1.10, 1.20 mg per ml. Medium was aliquoted into quadrant Petri dishes containing 5.00 ml medium per quadrant. The treatment concentrations in PDA were made from a stock solution of BOA in anhydrous ethanol (EtOH) (100 mg per ml). The 0.00% BOA control was a 1.20% concentration of anhydrous EtOH in PDA, to match the highest concentration of BOA solution in the treatments (1.20 mg per ml). The addition of EtOH to the medium did not inhibit the growth of any of the strains used in this study. Strains were incubated for 14 d in the dark at 22°C and scored for growth or no growth at a given concentration. These experiments were repeated twice with four replicates per isolate.

Strains assessed for BOA tolerance The following strains were scored for BOA tolerance: Fusarium verticillioides JFL A- 00149 (= FGSC 7600) and JFL A-00999 (= FGSC 7603), Fusarium sacchari JFL B- 03582 (= FGSC 7610) and JFL B-03853 (= FGSC 7611), Fusarium fujikuroi JFL C- 01993 and C-01995, Fusarium proliferatum JFL D-04853 (= FGSC 7614) and JFL D-

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 75

Table 4.1. BOA tolerance threshold (highest concentration of BOA supporting growth) of common maize endophytes.

BOA tolerance Taxon threshold (mg/ml)

0.25 Periconia macrospinosa (MS 31) P. macrospinosa (MS 113)

0.50 Fusarium thapsinum (JFL F-4093) F. thapsinum (JFL F-4094) Fusarium sacchari (JFL B-03852) F. sacchari (JFL B-03853) Fusarium fujikuroi (JFL C-01993) F. fujikuroi (JFL C-01995) Fusarium proliferatum (JFL D-04853) F. proliferatum (JFL D-04854) Acremonium zeae (NRRL 6415) A. zeae (NRRL 13540)

0.75 Fusarium konzum (JFL I-11615) F. konzum (JFL I-11616) Gibberella zeae (JFL 3639) G. zeae (JFL 3634) Fusarium circinatum (JFL H-10847) F. circinatum (JFL H-10850) Fusarium nygamai (JFL G-5111) F. nygamai (JFL G-5112) Nigrospora oryzae (MS 581)

1.00 Fusarium subglutinans (JFL E-00990) F. subglutinans (JFL E-02192) Nigrospora oryzae (MS 81)

1.10 Fusarium verticillioides (JFL A-00149) F. verticillioides (JFL A-00999)

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 76

04854 (= FGSC 7615), Fusarium subglutinans JFL E-00990 (= FGSC 7616) and JFL E- 02192 (= FGSC 7617), Fusarium thapsinum JFL F-4093 (= FGSC 7056) and JFL F-4094 (= FGSC 7057), Fusarium nygamai JFL G-5111 and JFL G-5112, Fusarium circinatum JFL H-10847 and JFL H-10850, Fusarium konzum JFL I-11615 and JFL I-11616, Gibberella zeae JFL 3639 and JFL 3634, Acremonium zeae NRRL 6415 and NRRL 13540) Nigrospora oryzae MS 81 (= DAOM 239206) and MS 581 (= DAOM 239207), and Periconia macrospinosa MS 31 (= DAOM 239204) and MS 113 (= DAOM 239205). All isolates of Fusarium were obtained directly from J. F. Leslie (Kansas State University, Manhattan, KS), A. zeae was obtained from D. Wicklow (U.S. Department of Agriculture, Peoria, IL), and N. oryzae and P. macrospinosa were isolated from maize in a related field experiment in Ontario in 2005 (Saunders and Kohn 2009). Designations for sources of strains are as follows: DAOM, National Mycological Herbarium, Agriculture and Agri-Food Canada, Ottawa, Ontario; FGSC, Fungal Genetics Stock Center, Department of Microbiology, University of Kansas Medical Center, Kansas City, KS; JFL, John F. Leslie, Department of Plant Pathology, Kansas State University, Manhattan; KS; MS, Megan Saunders, Department of Ecology and Evolutionary Biology, University of Toronto, Toronto, Ontario; and NRRL, Northern Regional Research Laboratory, USDA, ARS, Peoria, IL.

Isolation of fungal endophytes from maize N. oryzae and P. macrospinosa were isolated from the roots of maize seedlings grown in Ridgetown, Ontario. Root segments approximately 2.00 to 3.00 mm in diameter and 2.00 cm in length were surface sterilized in sequential washes of 70.00% EtOH (2 min) and 0.53% NaOCl (2 min) and finally were rinsed in sterile distilled water (2 min). Root tissue was plated onto PDA amended with 1.00 g per liter streptomycin sulfate and 0.25 g per liter neomycin sulfate. The efficiency of the surface sterilization procedure was tested using the tissue imprint method described by Schulz and Boyle (2005), and the procedure was found to be effective.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 77

Sequential inoculation experiments From the initial screen for BOA tolerance, two groups of species (referred to as interaction set 1 and interaction set 2) were chosen for sequential inoculation. The members of each interaction set were chosen so that commonly co-occurring species that have different BOA tolerance profiles would comprise each set. Interaction set 1 included F. verticillioides (JFL 00999), F. subglutinans (JFL 02192), and F. proliferatum (JFL 04854), and interaction set 2 included Nigrospora oryzae (MS 81), Acremonium zeae (NRRL 13540), and Periconia macrospinosa (MS 113). Each experiment was repeated twice with a minimum of three replicates. To determine the effect of substrate alteration by one species on the growth of another, isolates were inoculated sequentially on the same medium, using a technique maintaining sterility for both inoculations. For interaction set 1, two types of media were prepared in standard petri dishes: PDA amended with BOA (0.50 mg per ml), and PDA with 0.5% EtOH. For interaction set 2, PDA amended with 0.25 mg per ml BOA and PDA amended with 0.25% EtOH were prepared in standard Petri dishes. These concentrations were chosen because they were the highest tolerated by the most sensitive species within each interaction set. Cellophane membranes (VWR International, Mississauga, Ontario) were cut into 9- by 9-cm squares and sterilized by autoclaving for 20 min at 121°C in double distilled water. The membranes were overlaid smoothly on the surface of each Petri dish. The center of the membrane was inoculated with a mycelial plug 5 mm in diameter (taken with a no. 1 cork borer). Plates were incubated at 22°C in the dark, and radial growth measurements were taken at 72 h by tracing the colony margin on the plate. Directly after the measurement was taken, the membrane supporting the fungal colony was removed and replaced with a new sterile membrane. The medium that had been biotransformed by the preceding colonizing isolate was then inoculated with the second colonizer. Radial growth measurements of the second colonizers were taken at 72 h. At the completion of the experiment, the colony area measurements were traced onto paper, scanned, and saved as electronic files. Areas of the colonies were measured using the ImageJ digital analysis program (version 1.32; W. S. Rasband, U.S. National Institutes of Health, Bethesda, MD [http://rsb.info.nih.gov/ij/]). Prior to beginning these experiments, it was determined that all isolates grew on the surface of the membrane without penetration. F. graminearum CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 78 could not be included because it rapidly degraded the cellophane membrane.

Assessment of presence of BOA in the medium The degradation of BOA was evaluated using thin-layer chromatography (TLC) as described by Glenn et al. (2001). Media and inocula of the focal strains were prepared as described in the previous section. After 72 h of growth, the cellophane membranes were removed, and a plug of medium was taken from the center of the area previously covered by the colony. Glass-backed silica gel TLC plates (10 by 20 cm) precoated with F254, a fluorescent indicator, were used (VWR International, Mississauga, ON). TLC plates were placed in a 50°C oven for 10 min prior to use to remove ambient moisture. Plugs of medium were placed on oven-dried plates, with the upper surface down, for 30 s. The compounds were then mobilized in toluene-ethyl acetate-formic acid (50:40:10), allowed to air dry, and subsequently viewed and photographed using a 254-nm UV lamp (Mineralight UVSII; Ultra-Violet Products Inc., San Gabriel, CA). TLC plates were run with a BOA standard spot (10 ng) on each plate. In addition, uninoculated, BOA- amended PDA was screened. One plug from the center of each colony was sampled. The intensity of a spot is a qualitative, relative measurement of the amount of BOA in the medium. A darker spot indicates a greater quantity of BOA. High-pressure liquid chromatography (HPLC) was used to verify that the compound viewed using TLC was truly BOA. A second set of TLC plates was created to purify the compound for HPLC, using a protocol similar to that described by Glenn et al. (2003). After the first colonizer had grown for 72 h, the colony-containing membrane was removed. A total of 19 plugs were taken from the center of each petri plate using a no. 5 cork borer and processed as described above. TLC plates were left to dry under a laminar flow hood overnight, after which the silica gel containing the putative BOA spots was scraped off of the plate, eluted in methanol, and filtered through a 0.45-nm nylon filter. HPLC was performed on an HP series 1100 variable-wavelength detector, using a Perkin-Elmer Brownlee C18 column (column dimensions, 250 by 4.6 mm; bead size, 5 µm). The mobile phase was 40% aqueous methanol for the entire 10-min run. The flow rate was 1.00 ml per minute. The elutant (15 µl) was injected onto the column and monitored at 280 nm at room temperature. CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 79

Statistical analyses In this study the objective of the sequential inoculations was to determine if a species would grow faster when it succeeded unlike species (heterospecific interaction) than when it followed itself (autospecific interaction). If the colony area of a species is greater in a heterospecific than in an autospecific interaction, the heterospecific sequence is interpreted as facilitative. If the area of a species is less in a heterospecific than in an autospecific interaction, the heterospecific sequence is competitive. Tukey-Kramer honestly significant difference (HSD) tests were conducted to compare colony area values for different treatments. The threshold for statistical significance was set at a P value of ≤0.05. The analyses were conducted using JMP (version 5.1; SAS Institute Inc., Cary, NC). Interaction set 1 and interaction set 2 were analyzed separately.

RESULTS

Characterization of BOA tolerance Five BOA tolerance threshold groups were recognized (Table 4.1), based on the highest concentration of BOA that supported growth (mg per ml): 1.10 (F. verticillioides); 1.00 (F. subglutinans and N. oryzae); 0.75 (F. konzum, F. graminearum, F. circinatum, F. nygamai, and N. oryzae); 0.50 (F. thapsinum, F. sacchari, F. fujikuroi, F. proliferatum, and A. zeae); and 0.25 (P. macrospinosa). No species were able to grow in a 1.20 mg per ml concentration. The two isolates of N. oryzae differed in tolerance. There was no variation in tolerance levels between the two isolates tested in any of the other species.

Sequential inoculation of species pairs Mean colony areas for the three treatments in BOA-amended and BOA-free media are shown in Fig. 4.2 and 4.3. The three treatments are as follows: (i) solo control (growth of a strain on previously uncolonized medium), (ii) autospecific interaction (growth of a strain following itself), and (iii) heterospecific interaction (growth of a strain following an unlike species). The statistical significance of the difference between the mean colony area values of a single strain for each treatment is indicated. ANOVAs were conducted CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 80

Figure 4.2. Interaction set 1 (Fusarium verticillioides, F. subglutinans and F. proliferatum) mean colony area values of solo control or secondary colonizers on the BOA-amended medium (a) and on the BOA-free medium (b). Solo control, growth of a species on previously uncolonized medium; primary col., growth of secondary colonizer following F. verticillioides, F. subglutinans, or F. proliferatum as indicated. Error bars indicate standard errors. Tukey-Kramer HSD tests were used to compare mean colony area values of the same species in different treatments. The same letter above two bars indicates that there is no significant difference between means.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 81

Figure 4.3. Interaction set 2 (N. oryzae, A. zeae, and P. macrospinosa) mean colony area values of solo control or secondary colonizers on the BOA-amended medium (a) and on the BOA-free medium (b). Note that data are plotted on a log scale. Solo control, growth of a species on previously uncolonized medium; primary col., growth of secondary colonizer following N. oryzae, A. zeae, or P. macrospinosa as indicated. Error bars indicate standard errors. Tukey-Kramer HSD tests were used to compare mean colony area values of the same species in different treatments. The same letter above two bars indicates that there is no significant difference between means.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 82

to test for a significant difference between experimental replicates, and no significant differences were detected (results not shown). Figs. 4.4 and 4.5 compare the outcomes of autospecific and heterospecific interactions for the secondary colonizer. Values listed are the percent differences between mean colony area values in the autospecific interaction compared to a heterospecific interaction. A positive value indicates a greater mean colony area in the heterospecific interaction than in the autospecific interaction (facilitation). A negative value indicates that the colony area of the heterospecific interaction was less than that in the autospecific interaction (competition). In the presence of BOA, four out of six secondary colonizers were facilitated when the primary colonizers had higher BOA tolerance than the secondary colonizer (Fig. 4.4 and 4.5, above the diagonal). For example, in interaction set 1, F. verticillioides (Table 4.1, tolerance threshold group 1.10) had the greatest impact on the colony area of the least BOA-tolerant strain, F. proliferatum (group 0.50), increasing its colony area by 173% (Fig. 4.4). In interaction set 1, the isolates with the highest BOA tolerance, F. verticillioides and F. subglutinans (groups 1.10 and 1.00), achieved a larger colony area in the autospecific interaction than in the solo control (Fig. 4.2a). When the primary colonizer had lower tolerance than the secondary, three out of six secondary colonizers were inhibited (competition), and the remaining three had interactions that were not significant (neutral) (Fig. 4.4 and 4.5, below the diagonal). In BOA-free medium, two out of six secondary colonizers were facilitated, two were inhibited, and two were not significantly affected (Fig. 4.4 and 4.5). The solo control of all strains in the BOA-free medium had a larger colony area than in the autospecific interaction (Fig. 4.2b and 4.3b). In some cases the direction of heterospecific interactions was different in the BOA- amended and BOA-free media. In four out of six examples where the primary colonizer was more tolerant than the secondary colonizer, the direction of interaction changed depending on the medium. For example, when F. proliferatum followed F. subglutinans, the mean colony area was greater in the presence of BOA but was not affected in the BOA-free medium. When P. macrospinosa (group 0.25) followed N. oryzae (group 1.00), the mean colony area of P. macrospinosa was greater in the BOA-amended medium and decreased in the BOA-free medium. Some interactions were the same in CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 83

Figure 4.4. Percent differences between mean colony area values in the autospecific and heterospecific interactions for interaction set 1 (Fusarium verticillioides, F. subglutinans, and F. proliferatum). Mean colony area was measured at 72 h. Percent differences in colony area for the BOA-amended environment are on the top line of each cell, and those for the BOA-free environment are on the bottom. A positive value indicates a facilitative interaction and a negative value a competitive one. Asterisks indicate values where the heterospecific and autospecific mean colony area values are significantly different.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 84

Figure 4.5. Percent differences between mean colony areas in the autospecific and heterospecific interactions for interaction set 2 (Nigrospora oryzae, Acremonium zeae, and P. macrospinosa). Mean colony area was measured at 72 h. Percent differences for the BOA-amended environment are on the top line of each cell, and those for the BOA- free environment are on the bottom. A positive value indicates a facilitative interaction and a negative value a competitive one. Asterisks indicate values where the heterospecific and autospecific mean colony area values are significantly different.

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 85 direction, if not in magnitude, in both BOA-amended and BOA-free media. The colony area of F. verticillioides was always greater in autospecific interactions than in heterospecific interactions (Fig. 4.2). When F. verticillioides was a primary colonizer, the colony area of heterospecific secondary colonizers was greater, with the exception of F. subglutinans in BOA-amended medium.

Assessment of biodegradation of BOA in the medium The degradation of BOA in the medium was evaluated using TLC and HPLC. In interaction set 1, the plugs taken from medium in which F. verticillioides and F. subglutinans were incubated produced a spot of BOA lighter than that produced by F. proliferatum and the uninoculated control plate (Fig. 4.6). In interaction set 2, plugs from both N. oryzae and A. zeae produced a spot of BOA lighter than that of P. macrospinosa and the control plate. The results show that BOA tolerance is in part associated with chemical alteration of BOA. We observed no variation in intensity of BOA spotting on the TLC plates between biological replicates. Results of the HPLC analysis confirm that the compound identified using TLC was BOA. The retention times of the samples, including the BOA standard, were nearly identical, all peaks were symmetrical, and only one peak per chromatograph was observed (results not shown). Results could not be quantified because the sequential inoculation experiments were tractable only on solid medium.

DISCUSSION

The patterns of BOA tolerance observed in this study generally conform to expectations under the assumption that tolerant fungi will be associated with benzoxazinoid-producing hosts. The species in BOA tolerance threshold groups 1.10 and 1.00, i.e., F. verticillioides, F. subglutinans, and N. oryzae, are most frequently isolated from maize. The isolates in groups 0.75, 0.50 and 0.25 are associated with a variety of hosts, not all of which are known to produce benzoxazinoids (Leslie 1995; Klaasen and Nelson 1998; Wicklow et al. 1998; Desjardins 2003; Zeller et al. 2003; Gordon 2006).

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 86

Figure 4.6. Qualitative assessment of presence of BOA in the medium. Plugs were taken from the center of the colony after 72 h of incubation, after removal of the cellophane membrane. BOA spots were identified using a BOA standard (10 ng).

CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 87

Nine out of the 13 species compared here were previously tested for BOA tolerance. Glenn et al. (2001) assessed the ability of 29 Fusarium species to grow in a 1.00 mg per ml concentration of BOA. Eleven of these species were tolerant. Tolerance of this concentration in F. verticillioides, F. subglutinans, F. nygamai, F. thapsinum, F. sacchari, and F. fujikuroi was also observed in the present study. Tolerance of this concentration in F. graminearum, F. circinatum, and F. proliferatum was not repeated in the present study, in which the highest concentration tolerated by F. graminearum and F. circinatum was 0.75 mg per ml and the highest concentration tolerated by F. proliferatum was 0.50 mg per ml. With five exceptions, different strains were used in the two studies. The five strains that were used in the previous and present studies had the same tolerance test results. Divergent tolerance results might be due to fixed differences between strains or to differences between environments in the two studies, influencing growth kinetics. For example, the previous study was conducted with standard Petri dishes at 23°C, while the present study used subdivided Petri dishes (5 ml medium per compartment) at 22°C; in both studies the focal tests were performed with PDA amended with BOA.

Tolerance to BOA in F. konzum, N. oryzae, A. zeae, and P. macrospinosa is reported in the present study for the first time. F. konzum is a recently described species that appears to be endemic to the Konza Prairie in Kansas, where it is associated with wild grasses (Zeller et al. 2003). Benzoxazinoid levels in hosts of F. konzum have not been ascertained. Several agriculturally important members of the Poaceae produce benzoxazinoids, but few wild grasses have been investigated. P. macrospinosa was isolated frequently from maize in a related field experiment (Saunders and Kohn 2009), and A. zeae is a common maize endophyte (Wicklow et al. 1998). A. zeae decreases BOA in the medium and therefore has a mechanism of tolerance to BOA, although it cannot grow in concentrations above 0.50 mg per ml (Fig. 4.6). Species with relatively low BOA tolerance that are frequently associated with benzoxazinoid-producing hosts may survive in planta by any of several mechanisms. Benzoxazinoids and their derivatives are highest in concentration in the vascular tissue of the seedling and could be avoided spatially or temporally. Another strategy for colonization without tolerance could be concurrent colonization with a species that is able to detoxify benzoxaxinoid CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 88 degradation products.

In the field of community ecology, researchers have often focused on the role of interspecific competition in community organization and function. Recent studies indicate the prevalence of interspecific facilitation in these processes as well (for a review, see Bruno et al. 2003). A general trend emerging from this work is that primary colonizers often dramatically affect the establishment of communities by ameliorating stress in the environment, consequently facilitating colonization by additional species (Stachowicz 2001; Callaway et al. 2002; Bruno et al. 2003).

The modified Menge-Sutherland model (MSM) is a conceptual model proposing a trade- off between interspecific competitive ability and stress tolerance (Menge and Sutherland 1976; Menge and Sutherland 1987; Bruno et al. 2003). The direction of species interactions should therefore change predictably along a gradient of environmental stress, such that in less stressful conditions, species interactions are competitive, leading to the ecological dominance of one or a few species (Stachowicz 2001). In relatively high levels of environmental stress, well-adapted species are predicted to facilitate the population growth of community members. For example, environmentally dependent facilitation can occur when abiotic stress is ameliorated. Callaway (1994) evaluated the outcome of interactions between salt marsh plants and found that winter annual plants can facilitate the growth of perennial shrubs by creating shade, which acts to increase water availability and reduce salinity. Other studies have shown that when the stress of salinity is removed, species interactions become competitive rather than facilitative (see Bertness and Yeh 1994). The outcome of this dynamic is predicted to increase with the size or density of the facilitative individuals (Callaway and Walker 1997). The goal of these studies is not only to quantify the outcome of species interactions under specific conditions but to determine the various factors that contribute to the outcome of these interactions. Data describing the primary variables that shape species interactions will enable us to better predict community change over time.

While our study is not an explicit test of the MSM, this conceptual model provided CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 89 expectations against which observed data could be compared, offering a useful ecological framework for interpreting our results. In this study we measured the outcome of species interactions in the presence or absence of an environmental stress while keeping the starting population size constant. The extent of support for the MSM can be seen in Fig. 4.4 and 4.5, above the diagonal. In the presence of BOA, in four cases out of six, a primary colonizer with higher BOA tolerance facilitated a secondary colonizer with lower tolerance. In the absence of BOA, only two of these six interactions evidenced the competition effect expected under the MSM. Evidence is stronger for facilitation under stress than for corresponding competition in the absence of stress. The MSM is not sufficient to describe this system. A hypothesis that merits general testing is that under environmental stress, fungal species that detoxify their substrate can enhance the colonization rate of less tolerant fungi.

It is possible that several factors other than BOA tolerance and metabolism may have contributed to the direction and strength of interactions between species. For example, drawing essential nutrients from the medium could be a mechanism of competition for the primary colonizer. The degree to which a species was detrimental to the growth of itself (autospecific interaction) in the absence of BOA could be an approximate, relative measure of essential nutrient use by that species. In both interaction sets, the species most inhibited in the autospecific interaction in BOA free medium was neutral or competitive with other species in the same medium (F. subglutinans and N. oryzae in interaction sets 1 and 2, respectively). This pattern would be expected if F. subglutinans and N. oryzae depleted a greater essential nutrient load than the other species. In contrast, in both interaction sets, the species least detrimental to its own growth in BOA- free medium was neutral or facilitative to heterospecifics (F. verticilloides and A. zeae).

Another factor that may have contributed to the outcome of these species interactions is the production of antifungal compounds. The Fusarium species tested here each produce a suite of toxigenic compounds (Desjardins and Proctor 2001). One of the most toxic and abundant of these compounds is fumonisin B-1 (FB-1), a known fungitoxin produced by all of the Fusarium species used in this study (Keyser et al. 1999). F. verticillioides and CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 90

F. subglutinans are not sensitive to FB-1, and F. proliferatum is sensitive only to concentrations of 40 mM and higher. The isolate of F. subglutinans used in the sequential competition experiment does not produce FB-1 (Leslie et al. 1992). Because the other focal strains are not sensitive to low concentrations of FB-1 and Fusarium species do not produce mycotoxins in high quantities on PDA, it is unlikely that production of this compound influenced our results.

Several Fusarium mycotoxins, such as fusaric acid, are known to have antibacterial properties that may influence interactions between Fusarium spp. and bacteria in the maize environment (Bacon et al. 2006). Little is known about the effects of Fusarium mycotoxins on filamentous fungi. It is possible that mycotoxins other than FB-1 also have antifungal properties. This would be an important factor to consider when implementing biocontrol strategies. A consideration of both facilitative and competitive interactions between microbes could be a meaningful way to evaluate potential biocontrol agents.

Species used in interaction set 2 are also known to produce antifungal compounds. An organic extract from N. oryzae inhibited the colony growth and spore germination of several Fusarium species (Szewczuk et al. 1991), and A. zeae produces pyrrocidines, compounds that are antagonistic to F. verticillioides and Aspergillus flavus (Wicklow et al. 2005). Consequently, we cannot distinguish competition for nutrients from a potential antifungal interaction.

Many experimental studies have shown that positive and negative interactions co-occur in stressful environments, and the beneficial component of the interaction often outweighs the negative component (see Bertness and Yeh 1994; Callaway and Walker 1997). In this study, the outcome of some heterospecific interactions changed with addition of BOA to the medium. This may be because the effect of BOA detoxification is stronger than the effect of unknown mechanisms of competition in the presence of BOA, while in the absence of BOA, mechanisms such as nutrient competition or production of antifungal compounds alter the direction of the interaction. Such a CHAPTER 4. INFLUENCE OF BOA ON ENDOPHYTE SPECIES 91 dynamic would be explained by the MSM. Further work is needed to address these factors directly.

This study was designed to evaluate indirect species interactions, simulating a successional series where heterospecific mycelia do not intermingle. In some cases these species may also reside concurrently, with heterospecific hyphae in close proximity to one another. In such circumstances, factors such as BOA metabolism or mycotoxin production may change and shift the outcome of species interactions. The patterns observed here set an experimental baseline that is being compared with dynamics in planta in a related field study assessing the influence of F. verticillioides and benzoxazinoid production on endophyte community structure and dynamics (Chapter 5). The investigation of mechanisms contributing to the outcome of species interactions, under controlled conditions, will supplement field data and increase our understanding of endophyte community assembly.

CHAPTER 5

HOST DEFENSE COMPOUNDS AND THE SEED ENDOPHYTE, FUSARIUM VERTICILLIOIDES, AS FILTERS IN FUNGAL ENDOPHYTE COMMUNITY ASSEMBLY

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 93 ON ENDOPHYTE COMMUNITIES

ABSTRACT

Ecological theory predicts that species well adapted to an environmental stressor will facilitate establishment of less tolerant species. Every fungal endophyte species is likely to co-occur in planta with both host-produced defense compounds and a fungal species that is tolerant to the compounds. Here, maize (Zea mays) was used to test two hypotheses: i. plant-produced toxins are habitat filters in fungal endophyte community assembly, and ii. endophyte species that detoxify plant toxins relax the filter by facilitating less tolerant fungi. Maize produces toxins, the benzoxazinoids (BXs), and harbors the BX detoxifying endophyte Fusarium verticillioides (FV). Fields in Ontario, Canada and Georgia, USA were planted with BX producing (BX+) and non-producing (BX-) genotypes, and seed was inoculated with FV (FV+) or sterilized (FV-). Fungal isolates were tested for tolerance to 2-benzoxazolinone (BOA), a toxic BX byproduct. Species and functional diversity (distribution of BOA tolerance levels in endophyte communities) was assessed. Results show that FV abundance was highest in BX+FV+ plants. In leaves, abundance of Fusarium was greater in BX+ than BX- plants. In below ground tissue in Georgia, BX+FV+ communities had lower abundance of BOA sensitive species than BX+FV-. Overall, communities from FV- plants had higher species diversity than FV+. Previous evidence that BXs are a habitat filter that can increase Fusarium colonization was confirmed and newly demonstrated for seed-born FV. In BX+ plants, FV can prevent less tolerant species from colonizing, rather than facilitating as predicted.

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 94 ON ENDOPHYTE COMMUNITIES

INTRODUCTION

Every species will encounter multiple factors that may prevent or allow establishment within a community. Community ecology theory proposes that “filters” or “rules” mediate community assembly through a series of nested processes that result in the coexistence of particular species at a site (Diamond 1975; Weiher and Keddy 1999). These processes can be roughly divided into two categories: habitat filtering and species interactions between community members. Habitat filtering refers to the ability of specific environmental variables to prevent a species from establishing. For plant- associated fungi, putative habitat filters include host defenses such as tissue chemistry or enhanced lignification, soil pH or annual temperature. Interactions between fungal species are also likely to influence community assembly. Species can compete directly via exudation of antifungal metabolites, or can interact indirectly by influencing host physiology or competing for nutrients (Morrissey and Osbourn 1999; Fravel et al. 2003; Tiunov and Scheu 2005; Bennett et al. 2009).

Host defense compounds are among the most common habitat filters that fungal endophytes are likely to encounter. Equally prevalent are the fungal species that are able to detoxify the compounds of their host (VanEtten et al. 2001). Dramatic responses of microbial communities to defense compounds have been particularly well documented in the rhizosphere (Bais et al. 2006). Exudation of defense compounds from roots can result in dominance of highly tolerant fungal species (e.g. Carter et al. 1999). Tolerance is often achieved by detoxification of the compounds, indicating that defense compounds are a strong selective pressure on plant-associated fungi (Rettenmaier and Lingens 1985; Morrissey and Osbourn 1999; Dixon 2001; VanEtten et al. 2001). Despite the ubiquity of host defense compounds and the fungi that can detoxify them, the potential for these factors to interact and influence endophyte communities has rarely been investigated.

Another factor that is likely to influence fungal community assembly is the presence of a primary colonizer. Many plant species harbor seed endophytes that will serve as the primary colonizers of the plant environment. Community assembly of plants, animals CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 95 ON ENDOPHYTE COMMUNITIES and microbes can be altered by species in early succession, with the net outcome of interactions in the range from facilitative to competitive (for discussion see Walker and Chapin 1987). In the fungi this is perhaps best documented in biocontrol studies where the aim is to identify fungi that out-compete community members. Genotypes that are useful in biocontrol are often effective at competing for soil nutrients or infection sites, and may induce systemic resistance in the host (Fravel et al. 2003). Much of the work on interspecific interactions between fungi has documented evidence of competition (e.g. Boddy 2000) although there is evidence for facilitation in the functioning of decomposer communities (Tiunov and Scheu 2005).

The influence of competition on communities is highlighted in the Intermediate Disturbance Hypothesis (IDH), which predicts that dominance of a community by one or few species will prevent other species from inhabiting the community when disturbance (environmental stress) is low. Species diversity is predicted to be greatest at intermediate stress levels; at high stress levels few species can survive (Paine 1966; Sousa 1979). Hacker and Gaines (1997) modified the IDH to include the effect of positive interactions. Central to the modified IDH is the concept of an “ecosystem engineer”, a species that ameliorates the growth-limiting environment (Lawton 1994). Under this model, an ecosystem engineer increases diversity when stress is high by providing a modified habitat that can support growth. When stress is low, competition is expected to dominate interactions. Although well documented in some systems (Callaway et al. 2002), support for the model is mixed (Maestre et al. 2009). To the best of our knowledge, the modified IDH has not previously been invoked to describe microbial community dynamics. We hypothesize that fungal species able to detoxify plant defense compounds engineer a less hostile environment in planta, thereby facilitating colonization by species that are sensitive to the toxins. The modified IDH suggests that in the presence of defense compounds, microbial species diversity should be greater when a detoxifier is present than when it is absent.

Here, we evaluate the influence of host defense compounds and a seed endophyte able to detoxify them on the fungal endophyte communities of maize (Zea mays). Maize CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 96 ON ENDOPHYTE COMMUNITIES produces benzoxazinoids (BXs), compounds toxic to microbes, insects and plants (Barry and Darrah 1991; Niemeyer and Perez 1995; Hashimoto and Shudo 1996). The primary BXs produced by maize are 2,4-dihydroxy-7-methoxy-2H-1,4-benzoxazin-3-one (DIMBOA) and 2,4-dihydroxy-2H-1,4-benzoxazin-3-one (DIBOA). DIMBOA and DIBOA reside in the cell vacuole as biologically inactive beta-glucosides. When the vacuole is disrupted, they are converted to the aglyconic forms, which rapidly degrade to the benzoxazolinones, 6-methoxy-2-benzoxazolinone (MBOA) and 2-benzoxazolinone (BOA) (Hashimoto and Shudo 1996). The benzoxazolinones are widely toxic and can be exuded from roots into soil (Krogh et al. 2006).

Fusarium verticillioides is a benzoxazolinone-detoxifying species that commonly inhabits maize seed and can colonize the plant systemically (Glenn et al. 2002; Glenn et al. 2003). Maize is frequently colonized by endophytic F. verticillioides, with fields often having infection rates >90% (Bacon and White 2000; Munkvold et al. 1997). F. verticillioides is among the most tolerant of the fungal species that have been tested with benzoxazolinones [Friebe et al. 1998; Glenn et al. 2001; Saunders and Kohn 2009 (Chapter 2)]. Several other Fusarium species, including Fusarium subglutinans and Fusarium graminearum, are also highly tolerant. BOA can influence the direction of interactions between these maize endophyte species. In an in vitro study, Saunders and Kohn (2008) found that BOA detoxifying species facilitated the growth of less tolerant species in the presence of BOA, suggesting that the detoxifiers act as ecosystem engineers (Chapter 4). Notably, F. verticillioides facilitated other Fusarium species in both the presence and absence of BOA.

Field data indicate that BXs are a habitat filter in fungal endophyte community assembly. In mature maize leaves, abundance of BOA tolerant Fusarium species was up to 35 times greater in BX-producing than BX non-producing plants, indicating that BXs are a filter advantageous for Fusarium species [Saunders and Kohn 2009 (Chapter 2)]. Results from a related study suggest that BX production has a spatial scale of influence that extends to neighbors of the BX-producing plant. When maize genotypes deficient in BXs were grown with BX-producing genotypes, species diversity and infection density of root CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 97 ON ENDOPHYTE COMMUNITIES endophytes was less than when BX-deficient genotypes were grown alone (Chapter 3). These data are consistent with what is expected if BXs are exuded into soil and then influence species frequencies in the microbial community available for root colonization of neighboring plants.

To determine if F. verticillioides influences fungal community assembly, and if the presence of BXs affects this dynamic, maize was planted in a fully factorial design including BX-producing (BX+) and non-producing (BX-) genotypes that were either inoculated with F. verticillioides (FV+) or purged of it (FV-) prior to planting. We consider the BX+ genotype to be a high stress, and the BX- to be relatively low stress environment for endophytes. Previous work documented that BXs are a habitat filter in endophyte community assembly. Here, we tested the hypothesis that F. verticillioides can relax that filter by facilitating species with relatively low tolerance to BXs. We had two main predictions. First, if BXs are a habitat filter permitting colonization by fungi with high tolerance, then abundance of Fusarium should be highest in BX+ plants, and abundance of FV should be highest in BX+FV+ plants. Second, if FV facilitates colonization by species with lower tolerance to BXs, then BX+FV+ communities should have higher species diversity and a lower mean BOA tolerance level than BX+FV- communities. We did not anticipate that FV would influence communities in BX- plants, as the abundance of FV would be expected to be low in the absence of BXs.

MATERIALS AND METHODS

Field experiment Fields were planted in a fully factorial design in Ridgetown, ON and Watkinsville, GA. Each location was planted with two maize genotypes: W22, which produces BXs, and bxbx the only recorded natural mutant lacking the ability to produce BXs (Hamilton 1962). Both genotypes are Yellow Dent maize, characterized by a genetic background of flint and floury maize and common kernel phenotypes (Smith et al., 2004). The seed of each genotype was either sterilized or inoculated with F. verticillioides (FV) prior to planting. Maize in ON was inoculated with FV strain DAOM 216410 (National CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 98 ON ENDOPHYTE COMMUNITIES

Mycological Herbarium, Agriculture and Agri-Food Canada, Ottawa, Ontario, Canada), isolated from maize in ON. In GA, maize seed was inoculated with RRC 408 (Richard B. Russell Agricultural Research Center, United States Department of Agriculture, Athens, Georgia, USA), isolated from maize in GA.

All operations with seed were conducted in sterile, glass-distilled H2O, referred to as H2O below. Seed was first externally sterilized in 100% commercial bleach (3-6% NaOCl) while shaking for 10 min, rinsed with H2O, then moved to fresh H2O and allowed to imbibe for 4 hr. Seeds were then incubated in a 60-61°C water bath for 5 min with frequent shaking, and subsequently cooled in fresh H2O for 5 min. Batches of 100 drained seeds were incubated in a FV spore suspension of 106 conidia per ml for 18 hr at

25°C. The uninoculated control seed was incubated in H2O. After the 18 hr incubation period, seeds were planted immediately in 4 blocks per treatment. Blocks were ten feet apart. Each block contained one treatment, with 64 plants in 4 rows, plants one foot apart. Plant material was collected at 2, 4 and 9 wks post planting (further details on sampling described in the following sections).

Isolation and identification of fungi Plants were rinsed with distilled water and surface dried. Healthy segments of leaf and below ground tissue were collected from each plant. Eight leaf segments 0.5 x 1.0 cm were taken 0.5 cm from the midrib and above the leaf collar of the second and third leaf blades. The below ground tissue included the one segment of the mesocotyl and seven root segments, all of which were 2.0 cm long and 0.2–0.3 cm diameter.

Tissue segments were surface-sterilized first in 70% ethanol (1 min), then in 0.53%

NaOCl (1 min) and rinsed for 2 min in H2O. Tissue was incubated on two growth media. A general medium, potato dextrose agar (PDA, Difco, Detroit, MI, USA) amended with 1.00 g per liter streptomycin sulfate and 0.25 g per liter neomycin sulfate was used to capture as many species in the fungal community as possible. BOA medium, a selective medium amended with 1.00 mg ml−1 BOA (Glenn et al. 2001) was used to determine the mean number of Fusarium colonies per plant. Four segments of each tissue type were CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 99 ON ENDOPHYTE COMMUNITIES placed onto both media for incubation within 2 d of sampling from the field. Effectiveness of the surface sterilization procedure was tested on PDA using the imprint method described by Schulz and Boyle (2005). Approximately 25% of tissue samples were tested, and 1 contaminant was detected, a Trichoderma from a root sample.

For the PDA assay, fungi emerging from tissue were established in axenic culture on PDA and 2% malt extract agar (MEA). Cultures were identified morphologically when diagnostic characters were evident, or by using DNA sequence data when such characters were ambiguous or absent. To identify Fusarium, each isolate was established in culture from a single spore, and subsequently grown on Carnation Leaf Agar (Leslie et al. 2006) and PDA for morphological identification (Summerell et al. 2003; Leslie et al. 2006). For each Fusarium species that was identified morphologically, the identity of a subset of isolates was verified with DNA sequence data. Isolates that did not present sufficient characters for species identification were grouped into morphotypes. All morphotypes that were isolated more than 4 times during the course of the study were identified with DNA sequence data.

Fusarium species in soil were also identified. On the day of planting, soil cores were taken from 4 randomly selected blocks in both GA and ON. Soil dilutions of 10-1, 10-2, 10-3, 10-4, and 10-5 were made, and 100 µl from each was plated onto 2-3 plates of BOA medium (0.8 mg/ml). Emerging colonies were then isolated in pure culture. A subset was identified as described above.

Total genomic DNA was isolated using the DNeasy Plant Minikit (Qiagen, Mississauga, ON, Canada). Morphotypes were tentatively identified using the nuclear ribosomal internal transcribed spacer region (ITS), amplified by PCR using primers ITS-1F and ITS-4 (c. 600 bp) (Gardes and Bruns 1993; White et al. 1990). F. verticillioides was identified using the intergenic spacer region of rDNA, amplified using species specific primers VERT-1 and VERT-2 (Patino et al. 2004). When PCR resulted in successful amplification this was considered a positive identification. PCR amplification using these primers was conducted on isolates morphologically identified as members of the CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 100 ON ENDOPHYTE COMMUNITIES

Gibberella fujikuroi species complex (of which FV is a member). All other Fusarium isolates, including those that did not yield a product with the FV specific primers, were identified using the translation elongation factor 1-alpha (TEF) gene, amplified using primers TEF-1 and TEF-2 (c. 700 bp) (Geiser et al. 2004). To verify that the likelihood of obtaining false positives using the VERT-1/2 primers was low, 20 isolates that were positively identified as FV using the VERT -1/2 primers were sequenced using the TEF- 1/2 primers. All of the isolates that did not yield a product of amplification were also sequenced using TEF-1/2, and thus provided a test for false negatives.

The PCR conditions were as described in Saunders and Kohn (2009) (Chapter 2). Sequencing was performed at the Genetic Analysis Core Facility (USDA-ARS-ERRC, Wyndmoor, PA, USA) using an ABI 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Contigs were assembled and edited in Sequencher 4.6 (Gene Codes Corporation, Ann Arbor, MI, USA), and blast searches of the NCBI GenBank database were conducted for tentative identification. A 99% sequence match to a sequence of known origin was considered a correct identification. Sequences will be deposited in GenBank.

Species diversity of fungal endophyte communities Isolates obtained on PDA were used to ascertain the species diversity of fungal endophyte communities. For the PDA assay, below ground tissue was collected at two and four-wks, and leaf tissue at four-wks. Four plants were collected from each block, for a total of 256 plants. Each unique combination of location, treatment, tissue type and time was considered a distinct fungal community. Diversity was measured using the Simpson’s inverse diversity index (D) and Fisher’s alpha (α) (Magurran 2004). All were calculated using EstimateS (Colwell 2000).

ANOVAs were used to analyze the abundance of FV isolated from PDA. Because the repeated measurements are not independent of one another, a Bonferroni correction was applied (α = 0.05/2) to give an adjusted significance level of P < 0.025 for below ground tissue data. When significant differences were detected, follow-up Tukey-Kramer CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 101 ON ENDOPHYTE COMMUNITIES honestly significant difference tests (TK) were conducted. Residuals did not significantly deviate from a normal distribution (Shapiro-Wilks test), and variances were homogeneous (Levene’s test). All tests were conducted in JMP (version 5.1; SAS Institute Inc., Cary, NC, USA).

Functional diversity of fungal endophyte communities We define functional diversity in two ways i) the distribution of BOA tolerance threshold levels across communities, and ii) the mean tolerance levels of communities. The BOA tolerance threshold of species (highest concentration of BOA supporting growth) from each location was determined. Each species/morphotype that was isolated more than four times during the experiment was tested for tolerance to BOA. In circumstances where members of a species complex were not distinguishable, for example in the Fusarium oxysporum species complex, eight isolates per species from each location were tested. For the remaining species, four isolates per species from each location were tested. Both of the isolates used to inoculate seed were also tested. BOA tolerance was assessed in PDA amended with BOA (stock solution of 100 mg per ml in anhydrous ethanol) in each of the following concentrations: 0.00, 0.25, 0.50, 0.75, 1.00, 1.10, 1.20 mg per ml. Strains were incubated in quadrant Petri dishes for 14 d in the dark at room temperature (approx. 20°C), and scored for growth or no growth. Isolates were then assigned to a tolerance threshold group corresponding to the highest concentration of BOA that supported growth. The entire experiment was repeated twice, with 2 replicates per isolate in each experiment.

MANOVAs were used to analyze the distribution of isolates in the 6 tolerance threshold groups (BOA tolerance thresholds of 0.25, 0.50, 0.75, 1.00, 1.10 and 1.20 mg per ml). When significant differences were detected, TK tests were used to test for differences between treatments at each of the tolerance threshold groups. Residuals did not significantly deviate from a normal distribution, and variances were homogeneous.

The mean tolerance of each community was calculated as the average tolerance value of species in the community, weighted by the relative abundance of the species carrying CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 102 ON ENDOPHYTE COMMUNITIES each value. To determine if FV caused shifts in mean tolerance of other species in the communities, the mean tolerance of each community was also calculated with FV removed from the data set. Each of these values was analyzed using MANOVAs that included the independent effects of BX production, FV in seed and the interaction between these factors. To account for the repeated measures a Bonferroni correction was applied to below ground tissue data (α = 0.05/2). When significant differences were detected follow-up TK tests were conducted.

To determine if FV significantly changed the mean tolerance of each community, one- way ANOVAs were conducted to test for a difference between the mean tolerance of the whole community and the mean tolerance of the community with FV removed.

Abundance of Fusarium Isolates obtained on BOA medium were used to assay Fusarium abundance, defined as the number of isolates per plant. Previous studies have shown that up to 97% of the isolates obtained on BOA medium are Fusarium species [Glenn et al. 2001; Saunders and Kohn 2009 (Chapter 2)]. In both ON and GA, above and below ground tissue were collected at 2 and 4 wks. In ON, 9-wk old leaf tissue was also collected. Seven plants per block were collected for a total of 560 plants. One-way ANOVAs were followed by TK tests to evaluate differences between mean numbers of colonies. Variances were homogeneous and residuals did not deviate from a normal distribution. In order to determine the species of Fusarium present, isolates from two plants per block were identified using DNA sequence data.

BX concentration of maize plants The concentration of BXs in 2- and 4-wk old plants grown in ON and GA was determined. Three plants per block were collected and stored at -80C. Lyophilized maize plants were separated into above and below ground tissue. Samples were placed in scintillation vials with 15 ml methanol (MeOH). BOA (1.2 mg per ml) was added as an internal standard. After grinding the samples with a Polytron Grinder (Brinkmann Instruments, Inc., Westbury, NY), the solutions were filtered and aliquots analyzed by CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 103 ON ENDOPHYTE COMMUNITIES reversed phase HPLC, using a H2O/MeOH linear gradient from 10% to 100% MeOH in 35 min, a flow rate of 1 ml/min, and detection at 280 nm. Each solvent contained 0.1%

H3PO4. Analyses were performed with a Beckman Ultrasphere C18, 5 micron column (4.6 X 250 mm, Beckman Instruments, Norcross, GA) using a Hewlett-Packard 1090 diode array HPLC. Quantification was performed using BOA's response factor. Benzoxazinoids and benzoxazolinones were identified by mass spectrometry using a Finnegan LCQ Duo (ESI). The system was equipped with a SpectraSystem P2000 HPLC pump, a SpectraSystem AS3000 auto injector, and a SpectraSystem UV6000LP UV/Vis detector (recording at 280 nm) for tandem UV-MS analyses. With the instrument in positive ion mode, the mass spectrometer parameters were tuned on the 136 (M+H) m/z ion obtained by direct injection of the BOA standard (Aldrich, 98%). The solvent system used was methanol-water, with 1% formic acid added pumped at 0.2 ml per min. A linear gradient was used, beginning at 10% methanol and increasing to 100% over 120 min. The solvent was then held at 100% methanol for another 10 min. Xcalibur, the system control software, was run in data dependent mode, isolating and fragmenting the most intense ion (by collisional ion dissociation, CID, 35%) found in each scan. The resulting chromatograms were then searched for the m/z’s of known benzoxazinoids and benzoxazolinones (Cambier et al. 1999) and matched to the tandem UV chromatograms. ANOVAs were used to test for a difference in BX concentration in BX+ and BX- maize, and between different treatments of the same maize genotype.

RESULTS

Species diversity of fungal endophyte communities A total of 74 species or morphotypes were identified from 1433 isolates obtained on PDA (Table 5.1). Isolates from several species complexes were detected: F. oxysporum, Fusarium incarnatum-equiseti and Fusarium solani. When multiple isolates within a species complex were detected, the species complex was treated as a single species in the analysis. Plants in ON yielded approximately twice the number of isolates of those in GA (ON: 959, GA: 474). Our hypothesis was that if FV facilitates the growth of species with lower BOA tolerance in BX+ plants, then species diversity should be higher in CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 104 ON ENDOPHYTE COMMUNITIES

Table 5.1. Abundance of species and morphotypes (with No. designations) isolated from maize in Ontario and Georgia.

Location Taxon ON GA Alternaria alternata 192 53 Acremonium zeae 1 15 Cladosporium cladosporiades 1 0 Cladosporium herbanum 5 0 Cladosporium spinaerosperanum 2 0 Cladosporium sp. 3 0 2 Cochliobolus lunatus 0 1 Curvularia intermedia 0 1 Curvularia protuberata 0 4 Diaporthe phaseolorum 1 5 Dreschlera bisepta 2 4 Dreschlera sp. 2 6 3 Epicoccum nigrum 93 62 Exserohilum rostratum 0 1 Fusarium equiseti 37 19 Fusarium fujikuroi 3 18 Fusarium graminearum 4 1 Fusarium oxysporum species complex 263 103 Fusarium proliferatum 5 16 Fusarium subglutinans 8 0 Fusarium solani species complex 2 0 Fusarium sporotrichioides 0 1 Fusarium tricinctum 0 1 Fusarium verticillioides 99 47 Gaeumannomyces graminis 22 4 Hypocrea lixii 34 11 Microdochium sp. 1 0 Mucoraceae sp. 1 0 1 Nigrospora oryzae 5 1 Penicillium sp. 4 8 Periconia circinata 0 1 Periconia macrospinosa 27 2 Prosthecium acerophilum 0 1 Rhizopus sp. 0 1 Scytalidium sp.2 0 1 Stenocarpella maydis 0 3 Thielaviopsis sp. 0 4 Ulocladium sp. 1 0 Volutella sp. 4 0 No. 4 2 0 No. 14 85 56 No. 22 5 0 No. 57 3 0 No. 58 4 0 No. 59 11 0 No. 60 4 0 No. 61 1 0 No. 62 2 2 No. 63 0 1 No. 66 0 1 No. 68 1 0 No. 69 3 0 No. 70 1 0 No. 71 4 0 No. 72 0 1 No. 73 1 0 No. 74 1 0 No. 76 0 1 No. 77 0 2 No. 78 0 1 No. 79 0 1 No. 82 0 1 No. 83 0 1 No. 84 3 0 No. 85 0 2 No. 86 0 1 No. 87 1 0 No. 88 1 0 No. 89 0 1 No. 93 0 1 No. 94 3 4 No. 95 0 1 No. 97 1 0 No. 100 0 1 CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 105 ON ENDOPHYTE COMMUNITIES

BX+FV+ than in BX+FV- plants. We did not observe this trend. In below ground tissue species diversity was higher in FV- treatments than in FV+ treatments, with the exception of 4-wk old BX+ roots in GA (Fig. 5.1). No trends in species diversity of leaf communities were detected. Results from Simpson’s Inverse D and Fisher’s alpha indicate the same trends. The data suggest that FV often decreases species diversity of communities in below ground tissue.

Results indicate that seed-born FV colonizes roots of BX+ plants more readily than BX- plants (Fig. 5.2). In below ground tissue, FV was most abundant in BX+FV+ maize in 2- wk old plants in ON and 4-wk old plants in ON and GA (ON, 2-wk: DF=3, F=42.70, P=<0.0001; GA, 2-wk: DF=3, F=5.47, P=0.013; ON, 4-wk: DF=3, F=15.84, P=0.0002; GA, 4-wk: DF=3, F=14.79, P=0.0002). In below ground tissue of 2-wk old plants in GA, when seed of BX+ and BX- plants was inoculated with FV, plants had more FV than the controls. Only one isolate of FV was obtained from 4-wk old leaf tissue, suggesting that the seed endophyte generally does not colonize above ground tissue in the early stages of plant growth. Identification using the VERT-1/2 primers did not yield any false positives or false negatives.

The abundance of soil isolates recovered on BOA medium from ON ranged from 1-2 x10-4 isolates per grams of soil, and from 1-3 x10-4 isolates in GA. Of the Fusarium species identified from ON, 12 isolates belonged to the F. oxysporum species complex, 3 to the F. incarnatum-equiseti species complex, 1 to the F. solani species complex and 1 to F. subglutinans. Of those identified from GA, there were 9 F. solani species complex, 4 to the F. oxysporum species complex, 2 Fusarium fujikuroi, 1 Fusarium pallidoroseum and 1 Fusarium proliferatum. The purpose of this assay was to determine the detection rate of FV in soil. Because no isolates of FV were obtained from soil, and very few from the FV- treatments (Fig. 5.2), it is likely that most of the FV isolated in this study originated from seed inoculum.

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 106 ON ENDOPHYTE COMMUNITIES

14 Ontario Georgia

Leaf tissue 6

0 4-wk 4-wk 14

Fisher’s alphaFisher’s

Below ground tissue 2

0 2-wk 4-wk 2-wk 4-wk Plant age

Figure 5.1. Species diversity of maize fungal endophtye communities as

measured by Fisher’s alpha. Solid black = BX+FV+, dotted/striped black

= BX+FV-, solid grey= BX-FV+, dotted/striped grey= BX-FV-. In root

tissue, diversity of FV- is higher than FV+ in all cases but communities

from 4-wk old BX+ plants in GA. CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 107 ON ENDOPHYTE COMMUNITIES

2.5 Ontario Georgia

1.5

0.5 Leaf tissue (4-wk old) (4-wk tissue Leaf

2.5 a

2 b 1.5

1 a a 0.5 c Below ground tissue (2-wk old) (2-wk tissue ground Below b b

isolates / plant / isolates verticillioides Fusarium Mean 2.5

2 a a 1.5

b 1 b

b 0.5 b

Below ground tissue (4-wk old) (4-wk tissue ground Below b b 0 FV: + - + - + - + - BX: - + - + CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 108 ON ENDOPHYTE COMMUNITIES

Figure 5.2. Mean number of Fusarium verticillioides isolates per plant obtained on potato dextrose agar medium. BX: benzoxazinoid (+ is producing maize genotype, - is non-producing genotype). FV: F. verticillioides (+ indicates maize that was inoculated, - indicates that it was not inoculated). Each unique combination of tissue type and time was analyzed separately. The same letter above two columns indicates no significant difference between means. Vertical bars, ± SE.

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 109 ON ENDOPHYTE COMMUNITIES

Functional diversity of fungal endophyte communities Two measures of functional diversity were used: i. the distribution of BOA tolerance thresholds, and ii. the mean tolerance levels of communities. The BOA tolerance thresholds of 12 species (68 isolates) from GA, and 16 species (80 isolates) from ON were determined (Table 5.2). In total, species were classified in 6 BOA tolerance threshold groups ranging from 0.25-1.20 mg per ml (Fig. 5.3). The two isolates used to inoculate seed had a BOA tolerance threshold of 1.20 mg per ml, as did all of the FV isolates tested. More differences between treatments were seen in communities from below ground tissue than aerial tissue. In below ground tissue, infection status (FV+ or FV-) was usually the best predictor of tolerance threshold group distributions (Table 5.3). In both locations, functional diversity of below ground communities was altered; BX+FV+ plants had more isolates in the 1.20 tolerance threshold group than the other treatments. In ON, the 1.20 tolerance threshold group included FV as the sole species, while in GA it was comprised of FV and members of the F. oxysporum species complex. The abundance of isolates in the 0.75 tolerance threshold group differed between treatments in GA. In below ground tissue of 4-wk old plants, the abundance of isolates in the 0.75 tolerance group was lowest in BX+FV+ plants and highest in BX-FV- plants. In ON, there was no influence of treatment on isolates in any tolerance threshold group other than 1.20. Production of BXs increased the abundance of isolates in the highest tolerance threshold group in both locations. In contrast, the presence of FV had an effect specific to the GA location, where it decreased the abundance of less tolerant species.

The second measure of functional diversity, mean community tolerance, varied between treatments in below ground tissue, but not in leaves (Fig. 5.4: mean community tolerance is the whole bar; Table 5.4). The mean tolerance of BX+FV+ communities in below ground tissue of 4-wk old plants in both locations was significantly greater than in BX+FV-. In BX- plants, there was no significant difference in mean tolerance of FV+ and FV- communities. To determine the influence of FV on this metric, mean community tolerance was calculated with FV removed from the data set (dark grey portion of bars). We found that in below ground tissues, FV significantly increased the mean tolerance of BX-FV+ communities (asterisks- comparison between dark grey and CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 110 ON ENDOPHYTE COMMUNITIES

Table 5.2. BOA tolerance thresholds (highest concentration of BOA supporting growth in mg per ml) of maize fungal endophytes isolated from plants grown in Georgia and Ontario. Each fraction represents the number of isolates in that tolerance threshold group out of the total number of isolates tested of the species.

BOA tolerance threshold Location Taxon 0.25 0.5 0.75 1 1.1 1.2 Georgia Alternaria alternata 1/4 3/4 Acremonium zeae 2/4 2/4 Diaporthe phaseolorum 4/4 Epicoccum nigrum 3/4 1/4 Fusarium incarnatum-equiseti species complex 4/4 Fusarium fujikuroi 4/4 Fusarium oxysporum species complex 1/8 3/8 3/8 1/8 Fusarium proliferatum 4/4 Fusarium verticillioides 8/8 Hypocrea lixii 8/8 Morphotype no. 14 4/4 Penicillium sp. 2/8 4/8 1/8 1/8 Ontario Alternaria alternata 4/4 Cladosporium herbanum 4/4 Dreschlera sp. 2 2/4 2/4 Epicoccum nigrum 2/4 2/4 Fusarium incarnatum-equiseti species complex 4/4 Fusarium oxysporum species complex 4/4 Fusarium proliferatum 2/4 1/4 1/4 Fusarium subglutinans 4/4 Fusarium verticillioides 8/8 Gaeumannomyces graminis 4/4 Hypocrea lixii 8/8 Morphotype no. 14 2/4 2/4 Morphotype no. 22 4/4 Morphotype no. 59 4/4 Nigrospora oryzae 4/4 Periconia macrospinosa 4/4

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 111 ON ENDOPHYTE COMMUNITIES

18 (a) 15 a 12 a a a 9 Ontario a b 6 a a a a a 3 a a a a c c 0 18 15 12

9 a Georgia 6 a ab ab a a a a b ab 3 a a a b b a a a a Mean isolates in below ground tissue (2-wk) tissue ground below in isolates Mean 0 a 18 a a 15 a a 12 9 a Ontario a a a 6 a a a a b a b 3 b 0 18 15 12 a 9 Georgia ab a ab 6 b a a ab 3 a a b a a a a b Mean isolates in below ground tissue (4-wk) tissue ground below in isolates Mean 0 0.25 0.50 0.75 1.00 1.10 1.20 Tolerance threshold (BOA mg/ml) CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 112 ON ENDOPHYTE COMMUNITIES

18 (b) 15 12 9

Ontario 6 3 0 18 15 12 a

Mean isolates in leaves (2-wk) leaves in isolates Mean 9 Georgia b 6 b a a b ab ab 3 b b ab b 0 0.25 0.50 0.75 1.00 1.10 1.20 Tolerance threshold (BOA mg/ml)

Figure 5.3. Mean number of isolates in each tolerance threshold group in below ground tissue (a) and leaves (b). Light grey bars= BX-FV+, light grey striped= BX-FV-, dark grey = BX+FV+, dark grey striped= BX+FV-. Each unique combination of tissue type and time was analyzed separately. The same letter above two columns indicates no significant difference between means. Vertical bars, ± SE. CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 113 ON ENDOPHYTE COMMUNITIES

Table 5.3. MANOVA of BOA tolerance threshold groups with maize genotype (BX+ or BX-) and infection status (FV+ or FV-) as factors. Significant values are bolded.

Source of variation Plant age Tissue Location df Maize genotype Infection status Genotype x Infection 4-wk leaves ON 1 F=2.64, P=0.1303 F=0, P=1 F=1.35, P=0.2685 GA 1 F=6.96, P=0.0082 F=0.33, P=0.8059 F=4.85, P=0.0246 2-wk below ground ON 1 F=2.69, P=0.1027 F=19.74, P=0.0003 F=1.99, P=0.1844 GA 1 F=2.35, P=0.132 F=3.27, P=0.0648 F=0.77, P=0.5697 4-wk below ground ON 1 F=1.98, P=0.1853 F=7.53, P=0.0068 F=0.88, P=0.5331 GA 1 F=9.43, P=0.0028 F=33.68, P=<.0001 F=15.45, P=0.0008

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 114 ON ENDOPHYTE COMMUNITIES

1.0 Ontario Georgia

0.8

a 0.6 a a a a a a a

0.4 Leaves (4-wk old) (4-wk Leaves 0.2

0 1.0 *a * ab *a 0.8 ab ab b 0.6 b b

0.4

0.2 Below ground tissue (2-wk old) (2-wk tissue ground Below

0 Ave tolerance and Ave tolerance minus FV / block / FV minus tolerance Ave and tolerance Ave * 1.0 * a ab * a 0.8 ab ab b b b 0.6

0.4

0.2 Below ground tissue (4-wk old) (4-wk tissue ground Below 0 + - + - FV: + - + - + - + - BX: - + - + CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 115 ON ENDOPHYTE COMMUNITIES

Figure 5.4. Average community BOA tolerance level in fungal endophyte communities of maize (whole bar). The black fraction of each bar represents the mean community tolerance with Fusarium verticillioides (FV) removed from the analysis, and the grey fraction the amount that FV increased the mean tolerance of each community. Each unique combination of tissue type and time was analyzed separately. The same letter next to multiple points indicates no significant difference between means in tolerance of the whole community (entire bars: total of grey and black fractions). Asterisks indicate where FV significantly increased the mean tolerance for a given community, a comparison between the mean of the entire bar to the mean of the black fraction of each bar.

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 116 ON ENDOPHYTE COMMUNITIES

Table 5.4. MANOVA of average community BOA tolerance level, and average community BOA tolerance level with Fusarium verticillioides removed from the analysis with maize genotype (BX+ or BX-) and infection status (FV+ or FV-) as factors. Significant values are bolded.

Source of variation Plant age Tissue Location df Maize genotype Infection status Genotype x Infection 4-wk leaves ON 1 F=1, P=0.337 F=1, P=0.337 F=1, P=0.337 GA 1 F=3.57, P=0.0831 F=0.80, P=0.3881 F=2.87, P=0.0897 2-wk below ground ON 1 F=0.76, P=0.4016 F=20.32, P=0.0007 F=0.38, P=0.5469 GA 1 F=1.05, P=0.3257 F=21.88, P=0.0005 F=0.05, P=0.821 4-wk below ground ON 1 F=5.83, P=0.0327 F=14.94, P=0.0022 F=1.25, P=0.2848 GA 1 F=0.65, P=0.4345 F=14.52, P=0.0025 F=1.57, P=0.2344

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 117 ON ENDOPHYTE COMMUNITIES light grey portion of each bar: ON, 2-wk, BX-FV+: F=19.87, P=0.0043; BX-FV-: F=1.09, P=0.3357; BX+FV+: F=17.21, P=0.006; BX+FV-: F=3.35, P=0.1169. GA, 2- wk, BX-FV+: F=34.76, P=0.0011; BX-FV-: F=0.49, P=0.5061; BX+FV+: F=3.60, P=0.1065; BX+FV-: F=0, P=1. ON, 4-wk, BX-FV+: F=13.30, P=0.0107; BX-FV-: F=0.13, P=0.735; BX+FV+: F=5.78, P=0.0529; BX+FV-: F=0.78, P=0.4107. GA, 4- wk, BX-FV+: F=7.52, P=0.0336; BX-FV-: F=1.23, P=0.3106; BX+FV+: F=9.27, P=0.0226; BX+FV-: F=0, P=1; All tests have df=1). FV increased the mean tolerance of BX+ communities in below ground tissue of 2-wk old plants in ON and 4-wk old roots in GA. The presence of FV did not explain the increase in mean tolerance in below ground tissue of 2-wk old plants in GA and 4-wk old plants in ON, indicating that other species contributed to an increase in mean tolerance of these communities. BX+FV+ communities from below ground tissue had a higher mean tolerance than BX+FV- communities, with the exception of 2-wk old plants in GA.

Abundance of Fusarium A total of 640 isolates were obtained on BOA medium. Plants in ON yielded 3.4 times more isolates than those in GA (ON: 430, GA: 125). In 9-wk old leaves in ON, BX+ plants had approximately 4 times the abundance of Fusarium as BX- plants (Fig. 5.5; ON, leaves, 4-wk: F=8.79, P=0.0029; 9-wk: F=9.80, P=0.0019. GA, leaves, 4-wk: F=1.169, P=0.3621, all tests have df=3). In ON, BX+FV+ maize had the most isolates in 4-wk old leaves and below ground tissue of 2-wk old plants, (ON, below ground, 2-wk: F=6.14, P=0.009; 4-wk: F=0.08, P=0.9679. GA, below ground, 4-wk: F=1.07, P=0.3983, all tests have df=3). No significant difference was detected between treatments in GA. Of the isolates from BOA medium that were identified, ~90% were Fusarium species, on average. Members of the F. oxysporum species complex were dominant in both ON and GA (Table 5.5). The most important factor in determining abundance of Fusarium was production of BXs, as seen in 9-wk old leaves in ON.

BX concentration in maize plants BX concentrations in BX- maize were significantly lower than BX+ maize in all comparisons (all comparisons have a P-value of 0.0006 or less, data not shown). In 7/8 CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 118 ON ENDOPHYTE COMMUNITIES

Ontario Georgia 2.5 2.0 a 1.5 a a ab 1.0

Leaf tissue 0.5 ab b b b 0 a,a,a,a 2.5 2.0 a,a,a,a

1.5 a a a 1.0 b b,b a,a 0.5

Below ground tissue

isolates per plant per isolates Fusarium of number Mean 0 2 wk 4 wk 9 wk 2 wk 4 wk Plant age

Figure 5.5. Mean number of Fusarium isolates per plant obtained on Fusarium selective medium (BOA medium). Solid black line= BX+FV+, dotted black=

BX+FV-, solid grey= BX-FV+, dotted grey= BX-FV-. Each unique combination of tissue type and time was analyzed separately. The same letter next to multiple points indicates no significant difference between means. CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 119 ON ENDOPHYTE COMMUNITIES

Table 5.5. Identity of isolates obtained on BOA medium from plants grown in Ontario and Georgia. Isolates from 2 plants/block were identified.

Taxon Ontario Georgia Alternaria alternata 1 2 Acremonium zeae 0 1 Cladosporium sp. 3 0 1 Diaporthe phaseolorum 1 0 Epicoccum nigrum 1 0 Fusarium incarnatum-equiseti species complex 3 0 Fusarium fujikuroi 0 5 Fusarium graminearum 2 0 Fusarium oxysporum species complex 74 23 Fusarium proliferatum 2 3 Fusarium subglutinans 9 0 Fusarium verticillioides 27 7 Hypocrea lixii 2 0 Mucoraceae sp. 1 1 0 Morphotype no. 88 3 0 Nigrospora oryzae 6 0 Percentage Fusarium 88.6% 90.5%

CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 120 ON ENDOPHYTE COMMUNITIES collections, BXs, primarily in the aglyconic form, were detected in BX- maize. There was no significant difference in BX concentrations between BX-FV+ and BX-FV- plants. In below ground tissue of 2-wk old plants in GA, BX+FV+ maize had a lower mean BX concentration than BX+FV- plants (Fig. 5.6). In contrast, the concentration of BXs in 4- wk old leaves was significantly higher in BX+FV+ than in BX+FV- (Fig. S2). There was no significant difference between BX+ treatments in other tissues (ON, leaf, wk-4: F=0.04, P=0.8417. GA, leaf, wk-4: F=19.85, P=0.0043. ON, below ground, wk-2: F=0.05, P=0.8202; 4-wk: F=2.01, P=0.2060. GA, below ground, wk-2: F=8.42, P=0.0273; 4-wk: F=1.00, P=0.7618, all tests have df=3). Results indicate that although BXs are usually considered a constitutive defense, concentrations can increase or decrease with the presence of FV in seed.

DISCUSSION

In a previous study we established that BXs are a habitat filter that increases abundance of Fusarium species [Saunders and Kohn 2009 (Chapter 2)]. This was confirmed here, and results indicate that production of BXs can enhance colonization not only by horizontally transmitted Fusarium species, but also by seed-born F. verticillioides. In the present study we hypothesized that FV is an ecosystem engineer that facilitates colonization of less tolerant species in the presence of BXs. However, data suggest that FV does not act as an ecosystem engineer. Instead, species and functional diversity data are consistent with what is expected if FV prevents other species from colonizing. Presence of FV was generally associated with decreased species diversity in below ground communities, suggesting that it acts competitively with community members in both the high stress BX+ and the lower stress BX- environments. Data on functional diversity indicate that FV prevents less tolerant species from colonizing below ground tissues in GA. This is consistent with the IDH competition model, but not with the modified IDH that predicts facilitation by an ecosystem engineer. FV did not influence community members with lower BOA tolerance in ON. It is possible that the difference in the trends observed in ON and GA can be attributed to differences in the environments where the plants were grown. In the first month of growth, maize seedlings in GA were CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 121 ON ENDOPHYTE COMMUNITIES

Ontario Georgia 2.5 * 2.0 1.5

1.0 0.5

Leaves (4-wk old) 0 BX-FV+ BX-FV- BX+FV+ BX+FV-

4 *

Below ground (2-wk old) (2-wk ground Below BX+FV+ BX+FV-

Quantity of BXs (mg/g dry weight)

0 Below ground (4-wk old) (4-wk ground Below BX-FV+ BX-FV- BX+FV+ BX+FV- BX-FV+ BX-FV- BX+FV+ BX+FV-

Figure 5.6. Concentration of benzoxazinones in above and below ground tissue of maize plants. The dark grey fraction indicates the proportion made up by DIM2BOA-glucoside and HMBOA-glucoside, and the light grey fraction that from HMBOA, DIMBOA and TRIBOA. Each unique combination of tissue type and time was analyzed separately. Asterisks indicate where a significant difference between treatments was detected. CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 122 ON ENDOPHYTE COMMUNITIES subjected to a maximum temperature of 35oC and little to no precipitation. In ON, seedlings experienced a maximum temperature of 29oC and moderate precipitation. The elevated level of stress experienced by plants and fungi in GA may have enhanced the negative influence of FV on colonization of community members in the presence of BXs. The mechanism for this competition is unclear. If species inhabit the same niche space in the plant, direct physical interaction may occur, however, an indirect mechanism such as nutrient competition could also explain the patterns observed here.

Data on the BOA tolerance of communities also indicate that FV does not facilitate community members with lower tolerance. Below ground communities from BX+FV+ plants had a higher mean BOA tolerance level than in BX+FV- plants. This difference can be accounted for by FV alone; when FV was removed from the analysis the tolerance levels of all below ground communities were equivalent. The addition of FV to BX- plants also significantly increased the mean tolerance of communities, but not enough to differentiate the mean tolerance of BX-FV+ and BX-FV- communities. If detoxification of BXs by FV was occurring in planta, rates of detoxification were not sufficient to detect a change in BX concentration in plant tissue, with the exception of below ground tissue of 2-wk old plants in GA.

Our data indicate that Fusarium dominated a specific niche in the plant: the mesocotyl. In ON, 100% of mesocotyl segments were colonized, and 86% of the colonies isolated were Fusarium species. Only 76% of the root segments yielded isolates, of which 49% were Fusarium species, and 6% of leaf segments yielded Fusarium. A similar trend was noted in GA (100% of mesocotyl segments yielded isolates, 82% of which were Fusarium species; 82% of root segments yielded isolates; 36% of these were Fusarium, 3% of leaf segments contained Fusarium). The same species of Fusarium were isolated from mesocotyl and root tissue. This suggests that Fusarium dominates the mesocotyl and does not readily move to aerial organs. Our data are consistent with results from a greenhouse study in which GFP-expressing isolates of FV were used to visualize its movement in planta (Oren et al. 2003). Soil was inoculated with FV, and after three weeks FV was more abundant in the mesocotyl than in roots or leaves. Colonization of CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 123 ON ENDOPHYTE COMMUNITIES aboveground tissue was relatively low, suggesting that there was a barrier to colonization of aerial tissue by root endophytes; our data provide field evidence for this.

We chose to focus specifically on colonization by Fusarium because several species, including FV, F. subglutinans and F. proliferatum, pose an economic and health risk. These species can reside in grain as endophytes, often producing toxins that cause mycotoxicosis in domesticated animals and are suspected risk factors for cancers and other human health problems (Ueno et al. 1997; Marasas 2001; Marasas et al. 2004). Thus, contamination of maize by Fusarium causes millions of dollars of loss annually in the US alone (Wu 2007; Wu and Munkvold 2008). Interestingly, concentration of BXs is a highly variable trait, and maize has undergone selective breeding aimed at increasing BX concentrations to protect against insect damage. Results from the present and previous work raises the hypothesis that selective breeding has inadvertently increased colonization by Fusarium species [Saunders and Kohn 2009 (Chapter 2)].

One study has documented that a fungal pathogen can induce enhanced BX production in maize leaves (Oikawa et al. 2004). BXs are, however, generally considered to be a constitutive defense. In GA, BX concentration in 4-wk old BX+FV+ leaves was higher than in BX+FV-, supporting the hypothesis that BXs are inducible in leaf tissue. This could be a mechanism by which FV increases the strength of the habitat filter that provides it with a colonization advantage. Interestingly, we detected BXs in BX- plants, primarily as aglycones. Two likely explanations for this are that BX- plants produce small quantities of BXs that originate from metabolites not assayed for here, or that BX- plants take up the compounds from soil. The BX- genotype is usually not documented to produce BXs, with one exception reporting detection of small quantities (Zuniga et al. 1983). Uptake and translocation of BX byproducts has been documented in radish, but has not yet been tested for in maize (Chiapusio et al. 2004). It is well known that allelochemicals exuded into soil can detrimentally affect neighboring plants that take them up. To the best of our knowledge the ability of root exudates to act as defense compounds in the tissue of plants that have acquired them from plant neighbors has not CHAPTER 5. INFLUENCE OF BXS AND FUSARIUM VERTICILLIOIDES 124 ON ENDOPHYTE COMMUNITIES yet been investigated. Such a dynamic would extend the scale of influence of defense compounds and could change the outcome of plant-plant and plant-microbe interactions.

This study highlights the importance of habitat filtering and species interactions in fungal endophyte community assembly. We provide additional evidence that BXs are a habitat filter that allows for increased colonization by horizontally transmitted Fusarium species, and extend previous research results to show that this filter also acts on seed-born Fusarium. We found that FV influenced community assembly by decreasing the abundance of species less well adapted to host compounds, but only in a host with relatively high concentrations of the toxins. These results demonstrate the insight that can be gained by evaluating the effect of multiple, rather than only single, biotic factors on community assembly.

CHAPTER 6

SUMMARY

125

CHAPTER 6. SUMMARY 126

SUMMARY

The motivation for this project was my interest in the functional and species diversity of fungal endophyte communities, specifically the influence of species interactions on endophyte community dynamics. Maize and its associated endophytes appeared to be an ideal system with which to study biotic factors likely to influence assembly of fungal endophyte communities. Maize was known to commonly harbor BOA/MBOA tolerant Fusarium species, many of which produce unwanted mycotoxins. Despite this, the influence of BX production on colonization success of Fusarium species had never been tested. My field studies indicate that production of BXs by the host plant can select for enhanced colonization by Fusarium species in leaf tissue. This raises the hypothesis that selective breeding for enhanced BX concentrations in maize has unintentionally increased abundance of mycotoxin-producing Fusarium species. In future research, inbred maize lines created in the breeding program that selected for greater BX concentrations could be used to determine if selective breeding did in fact provide an ecological advantage to Fusarium in maize. Creation of isogenic maize lines that differ in ability to produce BXs would allow for a more direct test of BX production on endophyte colonization.

Ample evidence suggests that fungal endophytes have adapted to their host species. One of my long-term goals is to evaluate the influence of host plant evolution on endophyte communities and on interactions between endophyte species. Domesticated plants provide an ideal opportunity to address such questions. I hope to investigate the influence of maize domestication on endophyte communities, by comparing communities from maize genotypes that represent the progenitor of commercial maize (teosinte) to commonly used commercial hybrids, including maize lines selected for high BX concentrations. Because detailed records for maize breeding lines are available, the relationship of specific host traits on the functional and species diversity of fungal endophyte communities could be teased apart. For example, if increase in stalk size creates additional niche space for fungal endophytes, this could increase endophyte species diversity. Such an approach would be a novel test of the effect of crop CHAPTER 6. SUMMARY 127 domestication on microbial colonization.

Results from my initial field experiment comparing communities from BX-producing and non-producing maize genotypes showed that endophyte communities could be distinguished by BOA tolerance phenotype profiling, but not solely by the criterion of species diversity. This suggests that BXs present a barrier or filter through which potential endophytes must pass to successfully colonize the plant. The environmental filtering hypothesis utilizes the “assembly rules” concept to predict the formation of the community. Under the hypothesis, communities are constrained by species traits; the overall species abundance within a community is driven to a trait average optimal for the specific environment. This hypothesis has found success in predicting community structure of plants. In the future, I hope to test this hypothesis using fungal communities. This will involve the development of additional in vitro methods to quantify fungal phenotypes that are putatively important in plant colonization. Determination of thresholds and optima for temperature and moisture, production of compounds toxic to other fungi, and profiles of nutrient use could be used to categorize communities. Field observations and laboratory testing should expose traits important to endophyte success with a particular host species or genotype. This approach should help us to formulate evolutionary hypotheses about host-endophyte associations.

Assignment of trait values assessed in vitro should be undertaken with care, however, because many laboratory conditions have the potential to interact with the phenotype being tested. Variables such as starting population size, colony age, and the temperature, light and moisture conditions under which tests are conducted may influence fungal growth. For example, prior to BOA tolerance testing, I evaluated the influence of colony size, colony age, and nutrient quantity on the BOA tolerance of three Fusarium species (F. verticillioides, F. subglutinans, F. proliferatum). All three factors interacted with presence of BOA in the medium to influence growth rate, but none influenced the BOA tolerance threshold of isolates, suggesting that ability to grow on a specific substrate is less plastic than growth rate. Although community phenotype profiling may be a CHAPTER 6. SUMMARY 128 valuable way to characterize and compare communities, extreme care must be taken when designing and interpreting in vitro phenotype data.

The method used to describe a fungal community can influence the ability to detect microbial species interactions. In a recent study using maize, Pan and May (2009) described fungal endophyte communities using an experimental design in which they compared a culture-dependent method with a culture-independent method based on sequence data obtained directly from plant tissue. A null model was used to develop hypotheses about interactions between endophyte species, and results differed depending on the method that was used to define communities. Results from the culture dependent data indicated that positive species interactions were important, while those from the culture-independent approach detected negative interactions between endophyte species. The authors put forth the hypothesis that because culturing is likely to recover species that have similar resource requirements, culture-based community assessments are best suited for describing community assembly of fungi with similar niche requirements. A culture-independent approach may detect species from a wider diversity of niches, but can also bias estimates, for example by failing to detect whole fungal clades (Arnold et al. 2007). Currently, the most appropriate solution to this problem is the use of both culture-dependent and culture-independent methods in describing communities, an approach that I plan to employ in future studies.

I found preliminary evidence that BX-producing plants can influence the root endophyte communities of neighboring plants that are BX-deficient. An interesting avenue for future research is the influence that spatial proximity to defense compound producing plants has on the fungal communities of neighboring plant hosts. On what spatial scales can intraspecific community members manipulate that species pool of fungi available to colonize plant neighbors? This preliminary data also points to a mechanism through which neighborhood genetic diversity of the host species could affect fungal communities. The influence of genotype-level diversity on pathogen abundance and disease severity is well documented in agricultural systems, where varietal mixtures are sometimes used to reduce incidence of disease (Burdon et al. 2006). CHAPTER 6. SUMMARY 129

The phylogenetic community structure of interspecific plants hosts is known to have a significant impact on fungal communities. Phylogenetic distances among plant species can be correlated with plant community structure and pathogen host range, raising the hypothesis that there should be a phylogenetic signal in composition of fungal assemblages – plant species that are relatively phylogenetically similar should harbor endophyte communities that are more similar than distantly related hosts. Related to this is the hypothesis that plant genotypes that are more closely related should host more similar fungal assemblages. Characterization of endophyte communities associated with host genotypes of closely related species may reveal fungal traits that are fundamental in implementing a shift in host preference. This could be addressed by testing for a correlation between phylogenetic distance among host species and the mean value of fungal phenotypes that are of putative ecological importance. During the course of my career, I hope to contribute to our understanding of fungal ecology by incorporating studies on species diversity, functional traits, and phylogenetic structure of fungal communities and their hosts. The overarching goal is that such studies will help to develop general principals in fungal endophyte community ecology and will enhance our understanding of fungal evolution within a community context.

LITERATURE CITED

130

LITERATURE CITED 131

Ahlholm JU, Helander M, Henriksson J, Metzler M, Saikkonen K. 2002. Environmental conditions and host genotype direct genetic diversity of Venturia ditricha, a fungal endophyte of birch trees. Evolution 56: 1566-1573.

Argandona VH, Corcuera LJ. 1985. Distribution of hydroxamic acids in Zea mays tissues. Phytochemistry 24:177–178.

Arnold AE. 2007. Understanding the diversity of foliar fungal endophytes: progress, challenges and frontiers. Fungal Biology Reviews 21: 51–66.

Arnold AE, Engelbrecht BMJ. 2007. Fungal endophytes double minimum leaf conductance in seedlings of a tropical tree. Journal of Tropical Ecology 23: 369–372.

Arnold AE, Henk DA, Eells RL, Lutzoni F, Vilgalys R. 2007. Diversity and phylogenetic affinities of foliar fungal endophytes in loblolly pine inferred by culturing and environmental PCR. Mycologia 99: 185-206.

Arnold AE, Herre EA. 2003. Canopy cover and leaf age affect colonization by tropical fungal endophtyes: Ecological pattern and process in Theobroma cacao (Malvaceae). Mycologia 95: 388-398.

Arnold AE, Lutzoni F. 2007. Diversity and host range of foliar fungal endophytes: Are tropical leaves biodiversity hotspots? Ecology 88: 541-549.

Arnold AE, Mejia LC, Kyllo D, Rojas EI, Maynard Z, Robbins N, Herre EA. 2003. Fungal endophytes limit pathogen damage in a tropical tree. Proceedings of the National Academy of Sciences, USA 100: 15649–15654.

Bacon CW, Glenn AE, Yates IE. 2008. Fusarium verticillioides: Managing the endophytic association with maize for reduced fumonisins accumulation. Toxin Reviews 27: 411-446.

LITERATURE CITED 132

Bacon CW, Hinton DM, Glenn AE, Macias FA, Marin D. 2007. Interactions of Bacillus mojavensis and Fusarium verticillioides with a benzoxazolinone (BOA) and its transformation product, APO. Journal of Chemical Ecology 33: 1885–1897.

Bacon CW, Hinton DM, Hinton A. 2006. Growth-inhibiting effects of concentrations of fusaric acid on the growth of Bacillus mojavensis and other biocontrol Bacillus species. Journal of Applied Microbiology 100:185–194.

Bacon CW, White JF. (eds.) 2000. Microbial endophytes. New York, NY: Marcel Dekker, Inc.

Bailey JK, Deckert R, Schweitzer JA, Rehill BJ, Lindroth RL, Gehring C, Whitham TG. 2005. Host plant genetics affect hidden ecological players: links among Populus, condensed tannins, and fungal endophyte infection. Canadian Journal of Botany 83: 356- 361.

Bais HP, Park SW, Weir TL, Callaway RM, Vivanco JM. 2004. How plants communicate using the underground information superhighway. Trends in Plant Science 9: 26–32.

Bais HP, Weir TL, Perry LG, Gilroy S, Vivanco JM. 2006. The role of root exudates in rhizosphere interations with plants and other organisms. Annual Review of Plant Biology 57: 233-266.

Barry D, Darrah LL. 1991. Effect of research on commercial hybrid maize resistance to European corn-borer (Lepidoptera, Pyralidae). Journal of Economic Entomology 84: 1053–1059.

LITERATURE CITED 133

Bennett AE, Bever JD, Bowers MD. 2009. Arbuscular mycorrhizal fungal species suppress inducible plant responses and alter defensive strategies following herbivory. Oecologia 160: 771-779.

Bent AF, Mackey D. 2007. Elicitors, effectors, and R genes: The new paradigm and a lifetime supply of questions. Annual Review of Phytopathology 45: 399-436.

Bertness MD, Yeh SM. 1994. Cooperative and competitive interactions in the recruitment of marsh elders. Ecology 75:2416–2429.

Bishop DL. 2002. Gene expression of a vacuolar peroxidase with stress-induced pathogenesis in wheat sheaths. Physiological and Molecular Plant Pathology 61: 65–71.

Boddy L. 2000. Interspecific combative interactions between wood-decaying basidiomycetes. Fems Microbiology Ecology 31: 185-194.

Boller T, Felix G. 2009. A renaissance of elicitors: Perception of microbe-associated molecular patterns and danger signals by pattern-recognition receptors. Annual Review of Plant Biology 60: 379-406.

Boyd RS. 2007. The defense hypothesis of elemental hyperaccumulation: status, challenges and new directions. Plant and Soil 293: 153-176.

Boyd RS, Martens SN. 1994. Nickel Hyperaccumulated by Thlaspi montanum Var montanum is acutely toxic to an insect herbivore. Oikos 70: 21-25.

Bravo JM, Campo S, Murillo I, Coca M, Segundo BS. 2003. Fungus- and wound- induced accumulation of mRNA containing a class II chitinase of the pathogenesis- related protein 4 (PR-4) family of maize. Plant and Molecular Biology 52:745–759.

Bravo HR, Copaja SV, Lazo W. 1997. Antimicrobial activity of natural 2-

LITERATURE CITED 134 benzoxazolinones and related derivatives. Journal of Agriculture and Food Chemistry 45:3255–3257.

Broeckling CD, Broz AK, Bergelson J, Manter DK, Vivanco JM. 2008. Root exudates regulate soil fungal community composition and diversity. Applied and Environmental Microbiology 74: 738–744.

Bruno JF, Stachowicz JJ, Bertness MD. 2003. Inclusion of facilitation into ecological theory. Trends in Ecology and Evolution 18:119–125.

Burdon JJ, Thrall PH, Ericson L. 2006. The current and future dynamics of disease in plant communities. Annual Review of Phytopathology 44: 19-39.

Callaway RM. 1994. Facilitative and interfering effects of Arthrocnemum subterminale on winter annuals. Ecology 75:681–686.

Callaway RM, Brooker RW, Choler P, Kikvidze Z, Lortie CJ, Michalet R, Paolini L, Pugnaire FI, Newingham B, Aschehoug ET, Armas C, Kikodze D, Cook BJ. 2002. Positive interactions among alpine plants increase with stress. Nature 417:844–848.

Callaway RM, Walker LR. 1997. Competition and facilitation: a synthetic approach to interactions in plant communities. Ecology 78:1958–1965.

Cambier V, Hance T, de Hoffmann E. 1999. Non-injured Maize Contains Several 1,4- Benzoxazin-3-one Related Compounds but only as Glucoconjugates. Phytochemical Analysis. 10: 119-126.

Cambier V, Hance T, de Hoffmann E. 2000. Variation of DIMBOA and related compounds content in relation to the age and plant organ in maize. Phytochemistry 53: 223–229.

LITERATURE CITED 135

Carroll GC. 1995. Forest endophytes: pattern and process. Canadian Journal of Botany 73 (S1): 1316–1324.

Carroll GC, Petrini O. 1983. Patterns of substrate utilization by some fungal endophytes from coniferous foliage. Mycologia 75: 53-63.

Carter JP, Spink J, Cannon PF, Daniels MJ, Osbourn AE. 1999. Isolation, characterization, and avenacin sensitivity of a diverse collection of cereal-root-colonizing fungi. Applied and Environmental Microbiology 65: 3364–3372.

Chao A, Chazdon RL, Colwell RK, Shen T-J. 2005. A new statistical approach for assessing compositional similarity based on incidence and abundance data. Ecology Letters 8: 148–159.

Chen ZY, Brown RL, Rajasekaran K, Damann KE, Cleveland TE. 2006. Identification of a maize kernel pathogenesis-related protein and evidence for its involvement in resistance to Aspergillus flavus infection and aflatoxin production. Phytopathology 96:87–95.

Chiapusio G, Pellissier F, Gallet C. 2004. Uptake and translocation of phytochemical 2- benzoxazolinone (BOA) in radish seeds and seedlings. Journal of Experimental Botany 55: 1587–1592.

Clay K, Hardy TN, Hammond AM. 1985. Fungal endohytes of grasses and their effects on an insect herbivore. Oecologia 66: 1-5.

Clay K, Schardl C. 2002. Evolutionary origins and ecological consequences of endophyte symbiosis with grasses. American Naturalist 160: S99–S127.

Colwell RK. 2000. Estimates: statistical estimation of species richness. http://viceroy.eeb.uconn.edu/estimates

LITERATURE CITED 136

Conrath U, Pieterse CMJ, Mauch-Mani B. 2002. Priming in plant-pathogen interactions. Trends in Plant Science 7: 210-216.

Cooke RC, Rayner ADM. (eds.) 1984. The Ecology of Saprotrophic Fungi. London, UK: Longman.

Corcuera LJ, Woodward MD, Helgeson JP, Kelman A, Upper CD. 1978. 2,4- Dihydroxy-7-methoxy-2H-1,4-benzoxazin-3(4h)-one, an inhibitor from Zea mays with differential activity against soft rotting Erwinia species. Plant Physiology 61:791–795.

Cordero MJ, Raventos D, Sansegundo B. 1994. Differential expression and induction of chitinases and beta-1,3-glucanases in response to fungal infection during germination of maize seeds. Molecular Plant-Microbe Interactactions 7:23–31.

Davis EC, Shaw AJ. 2008. Biogeographic and phylogenetic patterns in diversity of liverwort-associated endophytes. American Journal of Botany 95: 914-924.

Desjardins AE. 2003. Gibberella from A (venaceae) to Z (eae). Annual Review Phytopathology 41:177–198.

Desjardins AE, Proctor RH. 2001. Biochemistry and genetics of Fusarium toxins. In B. A. Summerell, J. F. Leslie, D. Backhouse, W. L. Bryden, and L. W. Burgess (eds.), Fusarium: Paul E. Nelson memorial symposium. St. Paul, MN, USA: American Phytopathology Society Press., 50-69.

Delserone LM, McCluskey K, Matthews DE, Vanetten HD. 1999. Pisatin demethylation by fungal pathogens and nonpathogens of pea: association with pisatin tolerance and virulence. Physiological and Molecular Plant Pathology 55: 317-326.

LITERATURE CITED 137

Diamond JM. 1975. Assembly of species communities. In: Cody ML and Diamond JM, eds. Ecology and evolution of communities. Cambridge, MA, USA: Harvard University Press, 342–444.

Dixon RA. 2001. Natural products and plant disease resistance. Nature 411: 843–847.

Djonovic S, Vargas WA, Kolomiets MV, Horndeski M, Wiest A, Kenerley CM. 2007. A proteinaceous elicitor Sm1 from the beneficial fungus Trichoderma virens is required for induced systemic resistance in maize. Plant Physiology 145: 875-889.

Dodd J. 1980. The role of plant stresses in development of corn stalk rots. Plant Disease 64:533–537.

Donoso EP, Bustamante RO, Caru M, Niemeyer HM. 2008. Water deficit as a driver of the mutualistic relationship between the fungus Trichoderma harzianum and two wheat genotypes. Applied and Environmental Microbiology 74: 1412-1417.

Duckworth JC, Kent M, Ramsay PN. 2000. Plant functional types: an alternative to taxonomic plant community description in biogeography? Progress in Physical Geography 24: 515–542.

Elamo P, Helander ML, Saloniemi I, Neuvonen S. 1999. Birch family and environmental conditions affect endophytic fungi in leaves. Oecologia 118: 151-156.

Elliot SL, Sabelis MW, Janssen A, van der Geest LPS, Beerling EAM, Fransen J. 2000. Can plants use entomopathogens as bodyguards? Ecology Letters 3: 228-235.

Finkes LK, Cady AB, Mulroy JC, Clay K, Rudgers JA. 2006. Plant–fungus mutualism affects spider composition in successional fields. Ecology Letters 9: 344–353.

Fisher PJ. 1996. Survival and spread of the endophyte Stagonospora pteridiicola in

LITERATURE CITED 138

Pteridium aquilinum, other ferns and some flowering plants. New Phytologist 132: 119- 122.

Fisher PJ, Petrini O, Petrini LE, Sutton BC, 1994. Fungal endophytes from the leaves and twigs of Quercus ilex L. from England, Majorca, and Switzerland. New Phytologist 127: 133–137.

Fisher PJ, Petrini O, Scott HML. 1992. The distribution of some fungal and bacterial endophytes in maize (Zea mays L). New Phytologist 122: 299–305.

Fisher PJ, Petrini O, Sutton BC, 1993. A comparative study of fungal endophytes in xylem and bark of Eucalyptus nitens in Australia and England. Sydowia 45: 1–14.

Fravel D, Olivain C, Alabouvette C. 2003. Fusarium oxysporum and its biocontrol. New Phytologist 157: 493-502.

Friebe A, Vilich V, Hennig L, Kluge M, Sicker, D. 1998. Detoxification of benzoxazolinone allelochemicals from wheat by Gaeumannomyces graminis var. tritici, G-graminis var. graminis, G-graminis var. avenae, and Fusarium culmorum. Applied and Environmental Microbiology 64: 2386-2391.

Fujimura KE, Egger KN, Henry GH. 2008. The effect of experimental warming on the root-associated fungal community of Salix arctica. Isme Journal 2: 105-114.

Futuyma DJ, Wasserman SS. 1980. Resource Concentration and Herbivory in Oak Forests. Science 210: 920-922.

Gardes M, Bruns TD. 1993. ITS primers with enhanced specificity for basidiomycetes − application to the identification of mycorrhizae and rusts. Molecular Ecology 2: 113–118.

Geiser DM, Jimenez-Gasco MD, Kang SC, Makalowska I, Veeraraghavan

LITERATURE CITED 139

N, Ward TJ, Zhang N, Kuldau GA, O’Donnell K. 2004. FUSARIUMID v. 1.0: A DNA sequence database for identifying Fusarium. European Journal of Plant Pathology 110: 473–479.

Gilbert GS, Mejia-Chang N, Rojas E. 2002. Fungal diversity and plant disease in mangrove forests: salt excretion as a possible defense mechanism. Oecologia 132: 278- 285.

Gilbert GS, Reynolds DR, Bethancourt A. 2007. The patchiness of epifoliar fungi in tropical forests: Host range, host abundance, and environment. Ecology 88: 575-581. Gilbert GS, Webb CO. 2007. Phylogenetic signal in plant pathogen – host range. Proceedings of the National Academy of Sciences 104: 4979-4983.

Glenn AE, Gold SE and Bacon CW. 2002. Fdb1 and Fdb2, Fusarium verticillioides loci necessary for detoxification of preformed antimicrobials from corn. Molecular Plant- Microbe Interactions 15: 91-101.

Glenn AE, Hinton DM, Yates IE, Bacon CW. 2001. Detoxification of corn antimicrobial compounds as the basis for isolating Fusarium verticillioides and some other Fusarium species from corn. Applied and Environmental Microbiology 67: 2973– 2981.

Glenn AE, Meredith FI, Morrison WH, Bacon CW. 2003. Identification of intermediate and branch metabolites resulting from biotransformation of 2- benzoxazolinone by Fusarium verticillioides. Applied and Environmental Microbiology 69: 3165–3169.

Gonthier P, Gennaro M, Nicolotti G. 2006. Effects of water stress on the endophytic mycota of Quercus robur. Fungal Diversity 21: 69-80.

Gordon TR. 2006. Pitch canker disease of pines. Phytopathology 96:657–659.

LITERATURE CITED 140

Gotelli NJ, Ellison AM. 2004. The analysis of categorical data. In: Gotelli NJ, Ellison AM, eds. A primer of ecological statistics. Sunderland, MA, USA: Sinauer Associates, Inc., 349–382.

Guimil S, Chang HS, Zhu T, Sesma A, Osbourn A, Roux C, Ionnidis V, Oakeley EJ, Docquier M, Descombes P, Briggs SP, Paszkowski U. 2005. Comparative transcriptomics of rice reveals an ancient pattern of response to microbial colonization. Proceedings of the National Academy of Sciences of the United States of America 102: 8066-8070.

Hacker SD, Gaines SD. 1997. Some implications of direct positive interactions for community species diversity. Ecology 78: 1990-2003.

Hamilton RH. 1962. Segregation for a cyclic hydroxamate in maize seedlings. Maize Genetics Cooperation Newsletter 36: 71–72.

Hammerschmidt R. 1999. Phytoalexins: what have we learned after 60 years? Annual Review of Phytopathology 37: 285-306.

Hartnett DC, Wilson GWT. 1999. Mycorrhizae influence plant community structure and diversity in tallgrass prairie. Ecology 80: 1187-1195.

Hashimoto Y, Shudo K. 1996. Chemistry of biologically active benzoxazinoids. Phytochemistry 43:551–559.

Herre EA, Mejia LC, Kyllo DA, Rojas E, Maynard Z, Butler A, van Bael SA. 2007. Ecological implications of anti-pathogen effects of tropical fungal endophytes and mycorrhizae. Ecology 88: 550–558.

LITERATURE CITED 141

Hoffman MT, Arnold AE. 2008. Geographic locality and host identity shape fungal endophyte communities in cupressaceous trees. Mycological Research 112: 331-344.

Hughes AR, Inouye BD, Johnson MTJ, Underwood N, Vellend M. 2008. Ecological consequences of genetic diversity. Ecology Letters 11: 609-623.

Hurlbert SH. 1984. Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54: 187-211.

Johnson TMJ, Lajeunesse MJ, Agrawal AA. 2006. Additive and interactive effects of plant genotypic diversity on arthropod communities and plant fitness. Ecology Letters 9: 24-34.

Jouany JP. 2007. Methods for preventing, decontaminating and minimizing the toxicity of mycotoxins in feeds. Animal Feed Science and Technology 137: 342–362.

Jumpponen A, Trappe JM. 1998. Dark septate endophytes: a review of facultative biotrophic root-colonizing fungi. New Phytologist 140: 295-310.

Keyser Z, Vismer HF, Klaasen JA, Snijman PW, Marasas WFO. 1999. The antifungal effect of fumonisin B-1 on Fusarium and other fungal species. South African Journal of Science 95: 455–458.

Kauserud H, Mathiesen C, Ohlson M. 2008. High diversity of fungi associated with living parts of boreal forest bryophytes. Botany- Botanique 86: 1326-1333.

Keddy P. 1989. Competition. New York, NY, USA: Chapman and Hall.

Khanh TD, Chung MI, Xuan TD, Tawata S. 2005. The exploitation of crop allelopathy in sustainable agricultural production. Journal of Agronomy and Crop Science 191: 172- 184.

LITERATURE CITED 142

Klaasen JA, Nelson PE. 1998. Identity of Fusarium nygamai isolates with long and short microconidial chains from millet, sorghum and soil in Africa. Mycopathologia 140:171–176.

Klun JA, Robinson JF. 1969. Concentration of 2 1,4-benzoxazinones in dent corn at various stages of development of plant and its relation to resistance of host plant to European corn borer. Journal of Economic Entomology 62:214–220.

Kniskern JM, Traw MB, Bergelson J. 2007. Salicylic acid and jasmonic acid signaling defense pathways reduce natural bacterial diversity on Arabidopsis thaliana. Molecular Plant-Microbe Interactions 20: 1512-1522.

Krogh SS, Mensz SJM, Nielsen ST, Mortensen AG, Christophersen C, Fomsgaard IS. 2006. Fate of benzoxazinone allelochemicals in soil after incorporation of wheat and rye sprouts. Journal of Agriculture and Food Chemistry 54: 1064-1074.

Kumar DSS, Hyde KD. 2004. Biodiversity and tissue-recurrence of endophytic fungi in Tripterygium wilfordii. Fungal Diversity 17: 69–90.

Lawton JH. 1994. What Do Species Do in Ecosystems. Oikos 71: 367-374.

Lemons A, Clay K, Rudgers JA. 2005. Connecting plant-microbial interactions above and belowground: a fungal endophyte affects decomposition. Oecologia 145: 595-604.

Leslie JF. 1995. Gibberella fujikuroi: available populations and variable traits. Canadian Journal of Botany 73:S282–S291.

Leslie JF, Plattner RD, Desjardins AE, Klittich CJR. 1992. Fumonisin B1 production by strains from different mating populations of Gibberella fujikuroi (Fusarium section Liseola). Phytopathology 82:341–345.

LITERATURE CITED 143

Leslie JF, Summerell BA, Bullock S. 2006. The Fusarium laboratory manual. Oxford, UK: Blackwell Science.

Lewis LC, Bing LA. 1991. Bacillus thuringiensis Berliner and Beauveria bassiana (Balsamo) Vuillimen for European corn borer control – Program for immediate and season long suppression. Canadian Entomologist 123: 387-393.

Leyval C, Turnau K, Haselwandter K. 1997. Effect of heavy metal pollution on mycorrhizal colonization and function: physiological, ecological and applied aspects. Mycorrhiza 7: 139-153.

Long BJ, Dunn GM, Bowman JS, Routley DG. 1977. Relationship of hydroxamic acid content in corn and resistance to corn leaf aphid. Crop Science 17:55–58.

Luongo L, Galli M, Corazza L, Meekes E, De Haas L, Van der Plas CL, Kohl J. 2005. Potential of fungal antagonists for biocontrol of Fusarium spp. in wheat and maize through competition in crop debris. Biocontrol Science and Technology 15:229–242.

Macias FA, Marin D, Oliveros-Bastidas A, Castellano D, Simonet AM, Molinillo JMG. 2006. Structure-activity relationship (SAR) studies of benzoxazinones, their degradation products, and analogues. Phytotoxicity on problematic weeds Avena fatua L. and Lolium rigidum Gaud. Journal of Agriculture and Food Chemistry 54:1040–1048.

Maestre FT, Callaway RM, Valladares F, Lortie CJ. 2009. Refining the stress- gradient hypothesis for competition and facilitation in plant communities. Journal of Ecology 97: 199-205.

Magurran AE. 2004. Measuring biological diversity. Oxford, UK: Blackwell Science.

LITERATURE CITED 144

Marasas WFO. 2001. Discovery and occurrence of the fumonisins: a historical perspective. Environmental Health Perspectives 109: 239–243.

Marasas WFO, Nelson PE, Toussoun TA. 1984. Toxigenic Fusarium species: identity and mycotoxicology. University Park, PA, USA: Pennsylvania State University Press.

Marasas WFO, Riley RT, Hendricks KA, Stevens VL, Sadler TW, Gelineau-van Waes J, Missmer SA, Cabrera J, Torres O, Gelderblom WCA et al. 2004. Fumonisins disrupt sphingolipid metabolism, folate transport, and neural tube development in embryo culture in vivo: a potential risk factor for human neural tube defects among populations consuming fumonisin-contaminated maize. Journal of Nutrition 134: 711–716.

Marin S, Companys E, Sanchis V, Ramos AJ, Magan N. 1998. Effect of water activity and temperature on competing abilities of common maize fungi. Mycological Research 102: 959–964.

Marler MJ, Zabinski CA, Callaway RM. 1999. Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology 80: 1180-1186.

Marquez LM, Redman RS, Rodriguez RJ, Roossinck MJ. 2007. A virus in a fungus in a plant: Three-way symbiosis required for thermal tolerance. Science 315: 513-515.

Martin KJ, Rygiewicz PT. 2005. Fungal-specific PCR primers developed for analysis of the ITS region of environmental DNA extracts. BMC Microbiology 5: 1–10.

McGill BJ, Enquist BJ, Weiher E, Westoby M. 2006. Rebuilding community ecology from functional traits. Trends in Ecology and Evolution 21: 178–185.

Menge BA, Sutherland JP. 1976. Species-diversity gradients: synthesis of roles of predation, competition, and temporal heterogeneity. American Naturalist 110:351–369.

LITERATURE CITED 145

Menge BA, Sutherland JP. 1987. Community regulation: variation in disturbance, competition, and predation in relation to environmental stress and recruitment. American Naturalist 130:730–757.

Miller JD, Sumarah MW, Adams GW. 2008. Effect of a rugulosin-producing endophyte in Picea glauca on Choristoneura fumiferana. Journal of Chemical Ecology 34: 362-368.

Mittra B, Ghosh P, Henry SL, Mishra J, Das TK, Ghosh S, Babu CR, Mohanty P. 2004. Novel mode of resistance to Fusarium infection by a mild dose pre-exposure of cadmium in wheat. Plant Physiology and Biochemistry 42: 781-787.

Morrissey JP, Osbourn AE. 1999. Fungal resistance to plant antibiotics as a mechanism of pathogenesis. Microbial and Molecular Biology Review 63:708– 724.

Munkvold GP. 2003. Cultural and genetic approaches to managing mycotoxins in maize. Annual Review of Phytopathology 41: 99–116.

Munkvold GP, McGee DC, Carlton WM. 1997. Importance of different pathways for maize kernel infection by Fusarium moniliforme. Phytopathology 87:209 217.

Nicol RW, Yousef L, Traquair JA, Bernards MA. 2003. Ginsenosides stimulate the growth of soilborne pathogens of American ginseng. Phytochemistry 64: 257–264.

Niemela T, Renvall P, Pentilla R. 1995. Interactions of fungi at late stages of wood decomposition. Annales Botanici Fennici 32: 141-152.

Niemeyer HM. 1988. Hydroxamic acids (4-hydroxy-1,4-benzoxazin-3-ones), defense chemicals in the Gramineae. Phytochemistry 27:3349–3358.

LITERATURE CITED 146

Niemeyer HM, Perez FJ. 1995. Potential of hydroxamic acids in the control of cereal pests, diseases and weeds. In: Inderjit, Daksini KMM, Einhellig FA, eds. Allelopathy: organisms, processes and applications. Washington DC, WA, USA: ACS Symposium Series 582, American Chemical Society, 260-270.

Obledo EN, Barragan-Barragan LB, Gutierrez-Gonzalez P, Ramirez-Hernandez BC, Ramirez JJ, Rodriguez-Garay B. 2003. Increased photosyntethic efficiency generated by fungal symbiosis in Agave victoria-reginae. Plant Cell Tissue and Organ Culture 74: 237-241.

Oikawa A, Ishihara A, Tanaka C, Mori N, Tsuda M, Iwamura H. 2004. Accumulation of HDMBOA-Glc is induced by biotic stresses prior to the release of MBOA in maize leaves. Phytochemistry 65: 2995–3001.

Oren L, Ezrati S, Cohen D, Sharon A. 2003. Early events in the Fusarium verticillioides-maize interaction characterized by using a green fluorescent protein- expressing transgenic isolate. Applied and Environmental Microbiology 69: 1695-1701.

Osbourn AE. 1999. Antimicrobial phytoprotectants and fungal pathogens: a commentary. Fungal Genetics and Biology 26: 163-168.

Osbourn AE, Qi X, Townsend B, Qin B. 2003. Dissecting plant secondary metabolism constitutive chemical defences in cereals. New Phytologist 159: 101-108.

Ownley BH, Griffin MR, Klingeman WE, Gwinn KD, Moulton JK, Pereira RM. 2008. Beauveria bassiana: Endophytic colonization and plant disease control. Journal of Invertebrate Pathology 98: 267-370.

Paine RT. 1966. Food Web Complexity and Species Diversity. American Naturalist 100: 65-74.

LITERATURE CITED 147

Pan JJ, Baumgarten A, May G. 2008. Effects of host plant environment and Ustilago maydis infection on the fungal endophyte community of maize (Zea mays). New Phytologist 178: 147–156.

Pan JJ, May G. 2009. Fungal-fungal associations affect the assembly of endophyte communities in maize (Zea mays). Microbial Ecology DOI 10.1007/s00248-009-9543-7.

Park WJ, Hochholdinger F, Gierl M. 2004. Release of the benzoxazinoids defense molecules during lateral and crown root emergence in Zea mays. Journal of Plant Physiology 161: 981–985.

Patino B, Mirete S, Gonzalez-Jaen MT, Mule G, Rodriguez MT, Vazquez C. 2004. PCR detection assay of fumonisin-producing Fusarium verticillioides strains. Journal of Food Protection 67: 1278-1283.

Peters NK, Frost JW, Long SR. 1986. A plant flavone, luteolin, induces expression of Rhizobium meliloti nodulation genes. Science 233:977–80.

Petrini O. 1986. Taxonomy of endophytic fungi in aerial plant tissues. In: Fokkema NJ and van den Heuvel J, eds. Microbiology of the Phyllosphere. Cambridge, UK: Cambridge University Press, 175-187.

Petrini O. 1991. Fungal endophytes of tree leaves. In: Andrews J and Hirano S, eds. Microbial Ecology of Leaves. New York, NY, USA: Springer Verlag, 179-197.

Pinto LSRC, Azevedo JL, Pereira JO, Vierira MLC, Labate CA. 2000. Symptomless infection of banana and maize by endophytic fungi impairs photosynthetic efficiency. New Phytologist 147: 609–615.

Power AG, Mitchell CE. 2004. Pathogen spillover in disease epidemics. American Naturalist 164: S79-S89.

LITERATURE CITED 148

Redman RS, Sheehan KB, Stout RG, Rodriguez RJ, Henson JM. 2002. Thermotolerance generated by plant/fungal symbiosis. Science 298: 1581-1581.

Rettenmaier H, Lingens F. 1985. Purification and Some Properties of 2 Isofunctional Juglone Hydroxylases from Pseudomonas-Putida-J1. Biological Chemistry Hoppe-Seyler 366: 637-646.

Richardson MD, Bacon CW. 1993. Cyclic hydroxamic acid accumulation in corn seedlings exposed to reduced water potentials before, during, and after germination. Journal of Chemical Ecology 19: 1613–1624.

Richardson MD, Bacon CW. 1995. Catabolism of 6-methoxy benzoxazolinone and 2- belazoxazolinone by Fusarium moniliforme. Mycologia 87:510–517.

Ritchie SW, Hanway JJ, Benson GO, Herman JC, Lupkes ST. 1993. How a corn plant develops, Special Report No. 48. Ames, IA, USA: Iowa State University of Science and Technology, Cooperative Extensive Service.

Rodriguez RJ, Redman RS, Henson JM. 2004. The role of fungal symbioses in the adaptation of plants to high stress environments. Mitigation and Adaptation Strategies for Global Change 9: 261–272.

Rodriguez RJ, White JF, Arnold AE, Redman RS. 2009. Fungal endophytes: diversity and functional roles. New Phytologist 182: 314-330.

Rudgers JA, Clay K. 2007. Endophyte symbiosis with tall fescue: how strong are the impacts on communities and ecosystems? Fungal Biology Reviews. 21: 107-124.

Rudgers JA, Holah J, Orr SP, Clay K. 2007. Forest succession suppressed by an introduced plant–fungal symbiosis. Ecology 88: 18–25.

LITERATURE CITED 149

Rudgers JA, Koslow JM, Clay K. 2004. Endophytic fungi alter relationships between diversity and ecosystem properties. Ecology Letters 7: 42-51.

Russell FL, Louda SM, Rand TA, Kachman SD. 2007. Variation in herbivore- mediated indirect effects of an invasive plant on a native plant. Ecology 88: 413-423.

Ryan RP, Germaine K, Franks A, Ryan DJ, Dowling DN. 2008. Bacterial endophytes: Recent developments and applications. Fems Microbiology Letters 278: 1-9.

Saunders M, Kohn LM. 2008. Host-synthesized secondary compounds influence the in vitro interactions between fungal endophytes of maize. Applied and Environmental Microbiology 74: 136–142.

Saunders M, Kohn LM. 2009. Evidence for alteration of fungal endophyte community assembly by host defense compounds. New Phytologist 182: 229-238.

Schardl CL, Leuchtmann A, Spiering MJ. 2004. Symbioses of grasses with seedborne fungal endophytes. Annual Review of Plant Biology 55: 315–340.

Schulz B, Boyle C. 2005. The endophytic continuum. Mycological Research 109: 661– 686.

Schulz B, Boyle C, Draeger S, Rommert A-K, Krohn K. 2002. Endophytic fungi: a source of biologically active secondary metabolites. Mycological Research 106: 996– 1004.

Seghers D, Wittebolle L, Top EM, Verstraete W, Siciliano SD. 2004. Impact of agricultural practices on the Zea mays L. endophytic community. Applied and Environmental Microbiology 70: 1475–1482.

LITERATURE CITED 150

Sicker D, Frey M, Schulz M, Gierl A. 2000. Role of natural benzoxazinones in the survival strategy of plants. International Review of Cytology 198:319–346.

Sicker D, Schneider B, Hennig L, Knop M, Schulz M. 2001. Glycoside carbamates from benzoxazolin-2(3H)-one detoxification in extracts and exudates of corn roots. Phytochemistry 58:819–825.

Siegel MR. 1990. Fungal endophyte infected grasses – alkaloid accumulation and aphid response. Journal of Chemical Ecology 16: 3301-3315.

Sikora RA, Pocasangre L, zum Felde A, Niere B, Vu TT, Dababat AA. 2008. Mutualistic endophytic fungi and in planta suppressiveness to plant parasitic nematodes. Biological Control 46: 15-23.

Smith CW, Betran J, Runge ECA. 2004. Corn: origin, history, technology and production. Hoboken, NJ, USA: John Wiley and Sons, Inc.

Soby S, Caldera S, Bates R, VanEtten H. 1996. Detoxification of the phytoalexins maackiain and medicarpin by fungal pathogens of alfalfa. Phytochemistry 41: 759-765.

Sousa WP. 1979. Disturbance in Marine Inter-Tidal Boulder Fields - the Non- Equilibrium Maintenance of Species-Diversity. Ecology 60: 1225-1239.

Stachowicz JJ. 2001. Mutualism, facilitation, and the structure of ecological communities. Bioscience 51:235–246.

Steinkellner S, Lendzemo V, Langer I, Schweiger P, Khaosaad T, Toussaint JP, Vierheilig H. 2007. Flavonoids and strigolactones in root exudates as signals in symbiotic and pathogenic plant-fungus interactions. Molecules 12: 1290-1306.

LITERATURE CITED 151

Stinson KA, Campbell SA, Powell JR, Wolfe BE, Callaway RM, Thelen GC, Hallett SG, Prati D, Klironomos JN. 2006. Invasive plant suppresses the growth of native tree seedlings by disrupting belowground mutualisms. Plos Biology 4: 727-731.

Stinson AM, Zidack NK, Strobel GA, Jacobsen BJ. 2003. Mycofumigation with Muscodor albus and Muscodor roseus for control of seedling diseases of sugar beet and Verticillium wilt of eggplant. Plant Disease 87: 1349-1354.

Strobel GA, Dirkse E, Sears J, Markworth C. 2001. Volatile antimicrobials from Muscodor albus, a novel endophytic fungus. Microbiology 147: 2943-2950.

Stone JK, Bacon CW, White JF. 2000. An overview of endophytic microbes: endophytism defined. In: Bacon CW and White JF, eds. Microbial Endophytes. New York, NY, USA: Marcel Dekker, 3-30.

Summerell BA, Salleh B, Leslie JF. 2003. A utilitarian approach to Fusarium identification. Plant Disease 87: 117–128.

Suryanarayanan TS, Murali TS, Venkatesan G. 2002. Occurrence and distribution of fungal endophytes in tropical forests across a rainfall gradient. Canadian Journal of Botany 80: 818-826.

Sutton JC. 1992. Epidemiology of wheat head blight and maize ear rot caused by Fusarium graminearum. Canadian Journal of Plant Pathology 4: 195–209.

Szewczuk V, Kita W, Jarosz B, Truszkowska W, Siewinski A. 1991. Metabolites of Nigrospora: growth-inhibition of some phytopathogenic fungi by organic extracts from Nigrospora oryzae (Berkeley and Broome) Petch. Journal of Basic Microbiology 31:69– 73.

LITERATURE CITED 152

Thompson JN, Burdon JJ. 1992. Gene-for-Gene Coevolution between Plants and Parasites. Nature 360: 121-125.

Tiunov AV, Scheu S. 2005. Facilitative interactions rather than resource partitioning drive diversity-functioning relationships in laboratory fungal communities. Ecology Letters 8: 618-625.

Ueno Y, Iijima K, Wang SD, Sugiura Y, Sekijima M, Tanaka T, Chen C, Yu SZ. 1997. Fumonisins as a possible contributory risk factor for primary liver cancer: a 3-year study of corn harvested in Haimen, China, by HPLC and ELISA. Food Chemistry and Toxicology 35: 1143–1150. van der Heijden MGA, Bardgett RD, van Straalen NM. 2008. The unseen majority: soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecology Letters 11: 296-310.

Van Tichelen KK, Colpaert JV, Vangronsveld J. 2001. Ectomycorrhizal protection of Pinus sylvestris against copper toxicity. New Phytologist 150: 203-213.

Van Wees SCM, Van der Ent S, Pieterse CMJ. 2008. Plant immune responses triggered by beneficial microbes. Current Opinion in Plant Biology 11: 443-448.

VanEtten H, Temporini E, Wasmann C. 2001. Phytoalexin (and phytoanticipin) tolerance as a virulence trait: why is it not required by all pathogens? Physiological and Molecular Plant Pathology 59: 83–93.

Vega FE. 2008. Insect pathology and fungal endophytes. Journal of Invertebrate Pathology 98: 277-279.

Vega FE, Posada F, Aime MC, Pava-Ripoll M, Infante F, Rehner SA. 2008. Entomopathogenic fungal endophytes. Biological Control 46: 72-82.

LITERATURE CITED 153

Walker LR, Chapin FS. 1987. Interactions among Processes Controlling Successional Change. Oikos 50: 131-135.

Warfield CY, Davis RM. 1996. Importance of the husk covering on the susceptibility of corn hybrids to Fusarium ear rot. Plant Disease 80:208–210.

Weiher E and Keddy PA. 1999. Ecological assembly rules: perspectives, advances, retreats. Cambridge, UK: Cambridge University Press.

White TJ, Bruns TD, Lee SB, 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. San Diego, CA, USA, Academic Press, 315–322.

Whitham TG, Bailey JK, Schweitzer JA, Shuster SM, Bangert RK, Leroy CJ, Lonsdorf EV, Allan GJ, DiFazio SP, Potts BM et al. 2006. A framework for community and ecosystem genetics: from genes to ecosystems. Nature Reviews Genetics 7: 510–523.

Wicklow DT, Roth S, Deyrup ST, Gloer JB. 2005. A protective endophyte of maize: Acremonium zeae antibiotics inhibitory to Aspergillus flavus and Fusarium verticillioides. Mycological Research 109:610–618.

Wicklow DT, Weaver DK, Throne JE. 1998. Fungal colonists of maize grain conditioned at constant temperatures and humidities. Journal of Stored Product Research 34:355–361.

Wolfe BE, Rodgers VL, Stinson KA, Pringle A. 2008. The invasive plant Alliaria petiolata (garlic mustard) inhibits ectomycorrhizal fungi in its introduced range. Journal of Ecology 96: 777-783.

LITERATURE CITED 154

Wu F. 2007. Measuring the economic impacts of Fusarium toxins in animal feeds. Animal Feed Science and Technology 137: 363–374.

Wu F, Munkvold GP. 2008. Mycotoxins in ethanol co-products: modeling economic impacts on the livestock industry and management strategies. Journal of Agricultural and Food Chemistry 56: 3900–3911.

Yan F, Liang X, Zhu X. 1999. The role of DIMBOA on the feeding of Asian corn borer, Ostrinia furnacalis (Guenee) (Lep., Pyralidae). Journal of Applied Entomology 123:49– 53.

Yates IE, Bacon CW, Hinton DM. 1997. Effects of endophytic infection by Fusarium moniliforme on corn growth and cellular morphology. Plant Disease 81: 723-728.

Yue Q, Bacon CW, Richardson MD. 1998. Biotransformation of 2-benzoxazolinone and 6-methoxy-benoxazolinone by Fusarium moniliforme. Phytochemistry 48:451–454.

Zabalgogeazcoa I. 2008. Fungal endophytes and their interaction with plant pathogens. Spanish Journal of Agricultural Research 6: 138-146.

Zeller KA, Summerell BA, Bullock S, Leslie JF. 2003. Gibberella konza (Fusarium konzum) sp nov from prairie grasses, a new species in the Gibberella fujikuroi species complex. Mycologia 95:943–954.

Zhang Y, Zhang Y, Liu M, Xiaodong S, Zhiwei Z. 2008. Dark septate endophyte (DSE) fungi isolated from metal polluted soils: Their taxonomic position, tolerance, and accumulation of heavy metals in vitro. The Journal of Microbiology 46: 624-632.

Zuniga GE, Argandona VH, Niemeyer HM, Corcuera LJ. 1983. Hydroxamic acid

LITERATURE CITED 155 content in wild and cultivated Gramineae. Phytochemistry 22: 2665–2668.