Reciprocal Interactions Between Non-Vector Community Members

Home , Entomophaga

RECIPROCAL INTERACTIONS BETWEEN NON-VECTOR COMMUNITY MEMBERS

AND AN INSECT-TRANSMITTED PLANT PATHOGEN

PAUL JOSEPH CHISHOLM

A dissertation submitted in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

WASHINGTON STATE UNIVERSITY Department of Entomology

MAY 2017

To the Faculty of Washington State University:

The members of the Committee appointed to examine the dissertation of PAUL JOSEPH CHISHOLM find it satisfactory and recommend that it be accepted.

______David W. Crowder, Ph.D., Chair

______William E. Snyder, Ph.D.

______Sanford D. Eigenbrode, Ph.D.

______Jesse E. Brunner, Ph.D.

ACKNOWLEDGEMENTS

When I first came to WSU in the winter of 2013, I was like a newborn baby—wrinkled, confused, and searching for purpose. I found that purpose in the form of an advisor, David

Crowder. Dr. Crowder grasped my tiny, post-fetal hand and led me through the winding corridors of the academic labyrinth, teaching me how to make my own way in the world of scientific research. Today, when I look in the mirror, I no longer see that screaming, red-faced infant, flailing mindlessly in a sea of amniotic fluid. I see a full-grown adult scientist.

Others guided me through the maze, as well. I sometimes refer to my dissertation committee as the “League of Extraordinary Gentlemen,” though Sean Connery would be lucky to possess half the pluck and intellectual acuity of any of these seasoned academic minds. When the path to robust experimental design became dark, Dr. William Snyder brought me a torch to light my way. When I became lost among the endlessly branching network of Palouse pea farmers,

Dr. Sanford Eigenbrode was there to hand me a map. And when the gods saw fit to torment me with unfathomable adversity (namely, R-code errors), Dr. Jesse Brunner stepped in to intercede on my behalf. To these gentlemen, I am tremendously grateful.

Today, as I near the end of my journey, I feel like the Jamaican bobsled team from “Cool

Runnings.” My technical abilities were admittedly raw when I began, and I took my share of crashes. But as I carry my broken sled to the finish line, I hear the beginnings of a syncopated slow clap. I look around, and I see the faces of my friends and family in the crowd, joined by my academic mentors and collaborators. The collective clap grows in frequency and amplitude. Is that a tear forming in the corner of my eye? I trudge onwards, dragging my rusty old toboggan, knowing the end is in sight.

Thanks to you, I’m going to finish this race.

iii

RECIPROCAL INTERACTIONS BETWEEN NON-VECTOR COMMUNITY MEMBERS

AND AN INSECT-TRANSMITTED PLANT PATHOGEN

Abstract

by Paul Joseph Chisholm, Ph.D. Washington State University May 2017

Chair: David W. Crowder

Plant viruses interact with arthropod vectors in complex ecosystems consisting of many different components. Non-vector insect herbivores and plant-mutualistic bacteria are often present in these systems, and may have impacts on pathogens and vectors. Conversely, plant pathogens and vectors may significantly affect herbivores and plant mutualists. Here, we show evidence of reciprocal interactions between a plant virus, a vector herbivore, a non-vector herbivore, and a species of soil-dwelling rhizobia bacteria.

In Chapter 1 I provide an overview of the different community components that may influence plant pathogens, and summarize the current literature on this topic. In Chapter 2, we show results from a study demonstrating reciprocal interactions between a non-vector herbivore

(Sitona lineatus L.) and a plant virus, Pea enation mosaic virus (PEMV). Virus-infected pea plants (Pisum sativum) are attractive to S. lineatus, and feeding by S. lineatus increases the titer of PEMV in infected plants. Moreover, feeding by S. lineatus on pea hosts influences the host selection behavior of the Acyrthosiphon pisum, the main vector of PEMV. In Chapter 3, we

iv demonstrate that S. lineatus can also influence inter-host spread of PEMV. We observed higher incidence of virus in experimental mesocosms occupied by S. lineatus, and discovered that A. pisum is competitively displaced by S. lineatus in a way that increases the inoculation efficiency of PEMV. In Chapter 4, we examine how the mutualistic relationship between P. sativum and a nitrogen-fixing bacterium, Rhizobium leguminosarum biovar. viciae, influences A. pisum and

PEMV. Inoculation with R. leguminosarum results in decreases in aphid population and a subsequent reduction in PEMV incidence, while soil sterilization increases aphid populations and reductions in yield. These results indicate that the traditional host-vector-pathogen model may be an oversimplification of complex pathosystems. Incorporating the effects of

other members of the community may provide more robust predictions of pathogen prevalence and spread.

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS……………………………………………………………....………iii

ABSTRACT……………………………………………………………………………………...iv

LIST OF TABLES………………………………………………………………………..…….viii

LIST OF FIGURES……………………………………………………………………..……...... ix

CHAPTER ONE: INTERACTIONS BETWEEN PLANT VIRUSES, ARTHROPOD VECTORS, AND ECOLOGICAL COMMUNITIES

Abstract……………………………………………………………………………………1

Introduction…………………………………………………………………………...…..2

Influence of Vectors and Hosts on Viral Spread…………………………………...……..5

Influence of Non-Vector Insects on Viral Spread………………………………………..10

Conclusion……………....…………………………...…………………………………..16

Literature Cited…………………………………………………………………………..18

CHAPTER TWO: RECIPROCAL PLANT-MEDIATED INTERACTIONS BETWEEN A PLANT VIRUS AND A NON-VECTOR HERBIVORE

Abstract………………………………………………………………………..…………33

Introduction………………………………………………………………………………34

Methods……………………………………………….………………………………….36

Results……………………………………………………………………...…………….41

Discussion………………………………………………………………………………..42

Acknowledgements………………………………………………………………………45

Literature Cited………………………………………………………………….……….46

CHAPTER THREE: NICHE PARTITIONING BETWEEN HERBIVORES PROMOTES THE SPREAD OF A VECTOR-BORNE PLANT VIRUS

Abstract………………………………………………………………………………..…56

Introduction………………………………………………………………………………57

Methods…………………………………………………………………………………..58

Results……………………………………………………………………………………63

Discussion……………………………………………………………………………..…64

Acknowledgements………………………………………………………………………67

Literature Cited………………………………………………………………………...... 68

CHAPTER FOUR: SOIL RHIZOBIA PROVIDE DIRECT AND INDIRECT BENEFITS TO PLANT YIELDS

Abstract……………………………………………………………………………….….80

Introduction………………………………………………………………………………81

Methods……………………………………………………………………..……………83

Results…………………………………………………………………………………....87

Discussion………………………………………………………………………………..89

Acknowledgements………………………………………………………………………92

Literature Cited…………………………………………………..………………………93

APPENDICES

Appendix S1 Supplemental figures…………………………………………………..………...107

Appendix S2 Supplemental tables…………………………………………………....………...122

Appendix S3 Supplemental methods………………………………………………....………...125

vii

LIST OF TABLES

Table S1. Number of healthy/infected plant pairs measured from each sampling location in the observational defoliation study from Palouse pea fields……………………………….………122

Table S2. Commonly co-occurring chewing, non-vector herbivores and insect-transmitted viruses in agricultural systems………………………………………………………….………123

Table S3. Predictor coding for each individual treatment from the soil mesocosm experiment………………………………………………………………………………………124

viii

LIST OF FIGURES

FIGURE 1.1. Generalized dispersal strategy employed by persistent and non-persistent plant viruses……………………………………………………………………………………...... 31

FIGURE 1.2. Visual conceptualization of interactions between hosts, predators, competitors, mutualists, and vectors in a plant pathosystem……………………………………………...…...32

FIGURE 2.1. Host preferences (infected vs. healthy) of S. lineatus and A. pisum from greenhouse assays and field observations…………………………………………………...... 53

FIGURE 2.2. PEMV titer as a response to weevil feeding………………………….………..…54

FIGURE 2.3. Response of plant phytohormones to PEMV infection and defoliation from S. lineatus…………………………………………………………………………………………...55

FIGURE 3.1. Virus incidence and the reproduction, survival, and movement of A. pisum to feeding in response to feeding by S. lineatus………………………………………………….…77

FIGURE 3.2. Observed feeding locations of S. lineatus and A. pisum on individual pea plants.78

FIGURE 3.3. Proportion of plants becoming successfully infected with PEMV following aphid inoculation at various locations on the plant……………………………………………………..79

FIGURE 4.1. Nodulation in soil treatments……………………………………………...…....102

FIGURE 4.2. Aphid population, infection incidence, and plant yield in response to soil treatment………………………………………………………..………………..……………..108

FIGURE 4.3. Effect of nodulation on aphid populations……………………………………...109

FIGURE 4.4. Correlations between aphids, virus, and yield………………………………..…110

FIGURE 4.5. Path diagram of variable interactions in the best-supported model explaining how soil treatment, aphids, and virus influence yield…………………………………………..……111

FIGURE S1. Photos of study system components……………………………………….……107

FIGURE S2. Experimental setup for greenhouse choice experiment with S. lineatus….……..108

FIGURE S3. Relative size and aphid populations of healthy and infected plants……………..109

FIGURE S4. Relative aphid abundances on healthy and infected plants from the greenhouse choice assay…………………………………………………………………………………….110

FIGURE S5. PEMV titers from plants damaged by S. lineatus prior to inoculation………….111

FIGURE S6. PEMV titer from plants defoliated by S. lineatus following inoculation………..112

FIGURE S7. Setup of experiment examining aphid movement and virus incidence in response to weevil feeding at Tukey Orchard…………………………………………………………….113

FIGURE S8. Graphical sequence of events for the field mesocosm experiment at Tukey Orchard…………………………………………………………………………………………114

FIGURE S9. Setup of experiment examining the effect of aphid feeding location on inoculation efficiency……………………………………………………………………………………….115

FIGURE S10. Effect of S. lineatus and vector infectivity on reproduction, survival, and movement of A. pisum………………………………………………………………………….116

FIGURE S11. Proportion of A. pisum feeding on different plant structures…………………..117

FIGURE S12. Density plot of residuals from the infections~aphids function………………...118

FIGURE S13. Path diagrams of three best-supported models…………………………………119

FIGURE S14. Setup of soil experiment at Spillman Farm…………………………………….120

FIGURE S15. Diagram demonstrating the distribution of plants within each replicate for the soil experiment………………………………………………………………………………………121

Dedication

This dissertation is dedicated to all of you. Really, thanks.

CHAPTER ONE

INTERACTIONS BETWEEN PLANT VIRUSES, ARTHROPOD VECTORS, AND

ECOLOGICAL COMMUNITIES

ABSTRACT

Vector-borne pathogens threaten plant and animal health across many natural and managed ecosystems. Pathogens such as viruses and bacteria often affect the fitness and behavior of vectors in ways that promote pathogen spread; for this reason many pathogens are considered to have a mutualistic relationship with vectors. While impacts of pathogens on vectors have been relatively well studied, vector species often forage in complex food webs, interacting with individuals of many species, including predators, competitors, and mutualists. These interactions have received scant attention in the literature despite the understanding that species interactions can alter the fitness and movement of vectors, and might thereby promote or inhibit pathogen spread. For example, predators reduce the number of vector herbivores, which might inhibit the spread of viruses; however, predators might also stimulate vector movement which could promote virus spread. Other groups of organisms, such as competitors and mutualists, might also indirectly impact the dynamics of pathogen spread through their impacts on vector species or host plants. In this chapter we review the literature on the interactions between pathogens, arthropod vectors, and the communities in which they are embedded. We conclude that researchers must move beyond simple models of pathogen-vector interactions to more broadly understand how species interactions within food webs impact the dynamics of pathogens.

KEYWORDS: Vector, plant virus, competition, predation, disease ecology, community ecology

INTRODUCTION

Ever since the discovery of microorganisms, scientists have appreciated the role of pathogens in shaping both micro- and macroscopic communities. Viruses, which infect all life, are among the most ecologically-important pathogens. More than 700 plant viruses have been identified (Strange and Scott 2005), and most rely upon insect vectors for transmission (Power

2000, Hogenhout et al. 2008). Due to the ubiquity and economic importance of vector-borne plant viruses, understanding the community-level factors that influence viral dynamics in plant populations is critical.

As most plant viruses are vector-borne, the dynamics of pathogen transmission are intrinsically linked to the life history of the vector species involved in transmission. The abundance, behavior, and physiology of insect vectors and their plant hosts strongly mediate virus dispersal through ecosystems (Power 1987, Power 1991, Long and Finke 2015). Moreover, the dynamics of vector-host relationships can be strongly affected by individuals of other species, including non-vector species. Herbivores, predators, and mutualists may indirectly influence the dispersal and performance of plant viruses by affecting the behavior, physiology, and abundance of hosts and vectors. This review outlines mechanisms by which five major community guilds (vector insects, plant hosts, competitors, predators, and mutualists) influence the spread of a virus through a plant population.

Transmission dynamics of vector-borne plant viruses

Around half of all arthropod-borne plant viruses are transmitted by aphids (Casteel and

Jander 2013), and most of the rest are transmitted by thrips, whiteflies, leafhoppers, mites, or planthoppers (Bragard et al. 2013). Vector-borne plant viruses can be categorized as persistent,

2 where the virus migrates through the hemolymph to the vector’s salivary glands, causing a vector to become infectious for the rest of their life, or non-persistent, where the virus is simply carried on the vector’s stylet and remains active for a relatively short duration (Gray and Banerjee 1999,

Nault 1997). Persistent viruses can be further classified as propagative (replicating within the vector) or non-propagative (not replicating within the vector) (Bragard et al. 2013).

Persistent viruses typically require longer periods of feeding for vectors to successfully acquire and transmit them than non-persistent viruses. Consequently, persistent viruses may be more likely to form mutualistic relationships with vectors (reviewed in Mauck et al. 2012). For instance, infection with cucumber mosaic virus (CMV), a non-persistent virus, reduces the host quality of cucumber plants for the CMV vectors Aphis gossypii and Myzus persicae. Vectors are initially attracted to CMV-infected plants, before quickly detecting the inferior host quality and dispersing to new hosts (Mauck et al. 2010). Since CMV is acquired relatively quickly, this behavior results in the successful spread of CMV to new hosts. Persistent viruses, on the other hand, are not acquired by a vector unless it feeds on an infected plant for an extended period of time, and vectors will not become infectious if they quickly disperse from an infected plant. As a result, many persistent viruses, such as tomato spotted wilt, increase the quality of infected plants for the vector (Belliure et al. 2005) to ensure that the vector feeds long enough to acquire the virus (Fig. 1.1). Thus, patterns of vector movement to and from infected and healthy hosts may have different outcomes for pathogen spread depending on the mode of pathogen transmission.

Ecological context of plant viruses

Plant viruses often have considerable impacts on community composition (Dobson and

Crawley 1994). Pathogens may prevent a plant species from establishing in a new region, while a

3 lack of pathogens may allow a species to become naturalized. For example, invasive plants are affected by 77% fewer pathogens on average in their introduced range than their native range

(Mitchell and Power 2003). In contrast, introduced plants that undergo strong selection by native viruses are not likely to establish in a novel environment, as evidenced by the history of cassava in eastern Africa. Originally native to South America, cassava was introduced as a food staple to

Africa in the 16th century (Legg 1999). In the 1990s, the emergence of a recombinant form of two native African viruses decimated cassava plants across the eastern half of the continent, completely eliminating the species from the agricultural landscape in many communities (Legg and Fauquet 2004). Despite intense counter-measures, cassava remains difficult to grow in regions of Africa where the virus is prevalent (Legg and Fauquet 2004). Thus, invading plants entering a pathogen-free environment tend to thrive, whereas those entering a pathogen-rich environment typically perish.

Additionally, plant viruses can also enhance insect host ranges by turning a normally unsuitable plant into a nutritious host. For example, several whitefly (Bemisia tabaci) biotypes survive on tobacco (Nicotiana tabacum) only if the host is infected with Tomato yellow leaf curl

China virus (Zhang et al. 2012). As a result, the virus facilitates the spread of these biotypes into areas that would otherwise be not be colonizable. These examples illustrate the capacity of plant viruses as key drivers of species distributions and community structure.

Plant viruses also have a substantial economic impact on global agricultural production.

An estimated ten percent of total global agricultural productivity is lost to disease (Strange and

Scott 2005), and climate change is contributing to the emergence of several new viral threats to agricultural staples such as rice, wheat, potatoes, and corn (Anderson et al. 2004, Rosenzweig et al. 2001). Addressing the challenges presented by emerging viruses to food security will likely

4 become one of the defining agricultural issues of the 21st century. Consequently, a firmer understanding of the ecological factors underlying viral spread would be invaluable from both a theoretical and a pragmatic standpoint.

INFLUENCE OF VECTORS AND HOSTS ON VIRAL SPREAD

The performance and behavior of insect vectors is crucial to the proliferation of plant viruses. Although many different factors affect virus dispersal, vector abundance (Landis and

Van der Werf 1997, Legg and Ogwal 1998) and vector movement (Cornwell 1960, Roitberg and

Myers 1978, Smyrnioudis et al. 2001) are two of the most influential. Consequently, evolution has resulted in an assortment of mechanisms by which plant viruses affect vector abundance and movement in ways that enhance viral spread. These mechanisms can act either directly, by manipulating the vector itself, or indirectly, by altering the physiology of the host plant.

Direct viral effects on vectors

Few instances have been detected of direct effects of plant viruses on vectors. Indeed, plant viruses were originally assumed to be biologically inert elements within the bodies of insect vectors because the ability to replicate within both insects and plants was thought to be evolutionarily unlikely (Carter 1962, Kennedy et al. 1962). However, examples of pathogen- mediated impacts on vector from animal pathosystems are relatively common. Mosquitoes carrying the dengue virus, for instance, take longer bloodmeals than non-infected mosquitoes, increasing the likelihood the mosquito will transmit the virus to the host (Platt et al. 1997).

Emerging molecular techniques are revealing that direct impacts of plant viruses on vectors may also be more common than previously suspected, For example, electron microscopy (Sylvester

1980) has revealed that several viruses, mostly from the family Rhabdoviridae, actively replicate within their insect vector (reviewed in Hogenhout et al. 2008). In these situations, it may be technically appropriate to define the insect as a secondary host rather than a vector. Some of these propagative viruses, such as tomato leaf curl virus in whiteflies, are so biologically active within the salivary glands of their hosts that they negatively influence the insect’s fitness by diverting resources from physiological processes to viral transcription (Sinisterra et al. 2005).

Moreover, recent research indicates that even non-propagative viruses may directly influence vector behavior. In host choice assays, the bird cherry oat aphid (Rhopalosiphum padi) initially selects hosts infected with BYDV at a higher rate than non-infected hosts; however, once it has acquired the virus from a feeding membrane, this trend reverses and R. padi preferentially selects non-infected hosts (Ingwell et al. 2012). Although the mechanisms driving this interaction are unclear, it indicates that propagative viruses are not the only class of virus that might directly manipulate vector behavior.

Indirect effects on vectors via infected plants

Viruses may also indirectly affect vector behavior through physiological changes in the infected host plant. When a plant is infected with a virus, it undergoes a series of physiological changes that affect the vector in a way that generally maximizes dispersal of the virus. Typically, vectors are more attracted to infected plants than non-infected plants (reviewed in Bosque-Perez and Eigenbrode 2011). This increased attractiveness has been attributed to enhanced plant volatile production (Mauck et al. 2010, Eigenbrode et al. 2002, Jimenez-Martinez 2004) and the yellowing, chlorotic appearance of diseased hosts (Ajayi and Dewar 1983, Fereres and Moreno

2009, Hodge and Powell 2010). The attractiveness of infected plants means that non-infected vectors are more likely to select an infected plant as a host and in turn acquire the virus.

In addition to being more odoriferous and visually appealing, many infected plants are also less toxic and more nutritious than non-infected plants (Belliure et al. 2005, Bosque-Perez and Eigenbrode 2011, Malmstrom et al. 2011). Vectors feeding on infected plants thus experience a fitness boost due to a variety of context-dependent mechanisms. Many viruses, for instance, halt the transcriptional pathways that produce defensive compounds through a gene- silencing mechanism (Lewsey et al. 2010, Zhang et al. 2012, Ziebell et al. 2011). Additionally, nutrients such as free amino acids may be more abundant in infected hosts (Casteel et al. 2014).

As a result, many vector-virus relationships may be categorized as mutualistic (Belliure et al.

2005). The enervated defenses and enhanced nutritional content of infected plants allow associated vector populations to grow faster on infected compared to non-infected plants, and this increase in vector abundance enhances the dispersal of the virus.

Some viruses, however, especially those that are non-persistently transmitted (Mauck et al. 2012), exert a negative fitness effect for their vector. Infection with non-persistent cucumber mosaic virus (CMV), for instance, reduces cucumber to a less nutritious host for its two primary vector species (Mauck et al. 2010). To mask the poor host quality of the infected plant, CMV stimulates the hyperproduction of insect-attracting volatiles in cucumber, such that potential vectors are attracted to plants despite their relatively low nutritional quality. When the aphids alight on infected host tissue, they soon discover the inferiority of the host and readily depart for another plant. Because CMV can be acquired in only a few minutes, this promotes the spread of

CMV to subsequent plants (Mauck et al. 2010). Consequently, insect-borne viruses can have negative consequences for vectors under certain situations.

Viral dispersal is also maximized by the way individual virions behave within a host.

Infected cells often produce different viral morphs, with some morphs specially equipped for long-distance dispersal through the plant (Blanc et al. 2011). Additionally, some viruses, such as cauliflower mosaic virus, stimulate the formation of specialized transmission bodies within plants (Drucker et al. 2002) that rearrange when the plant is attacked by a phloem feeder to increase the chance that a potential vector will pierce the body and acquire the virus (Martiniere et al. 2013). Both these mechanisms positively influence viral dispersal.

Regulation of viral spread by host diversity

The diversity and spatial arrangement of plants within a community can affect viral spread under certain conditions. Much of our current knowledge base is derived from the foundational work by Leonard (1969), who modeled the dispersal rate of a fungal pathogen in relation to the proportion of susceptible hosts. Though Leonard’s models have considerable bearing on viral pathogens, modeling the spread of a virus is complicated by the added component of insect vectors.

Genetically diverse plant populations generally inhibit viral dispersal more than less diverse communities (Power 1991, Zhu et al. 2000), and the presence of non-susceptible plant species in communities with high interspecific diversity may provide a physical barrier that slows pathogen spread (Power 1987, Finckh et al. 2000, Mitchell et al. 2002). Additionally, the vectors that are necessary for viral dispersal can also be challenged and negatively affected by diverse habitats (Underwood and Rausher 2000, Carr and Eubanks 2002, Bukovinsky et al.

2008). Agroecosystems are an excellent setting in which to study the effects of host diversity on pathogen spread because plant diversity can easily be manipulated on large scales. For instance,

8 wheat fields containing only one cultivar have a significantly higher incidence of barley yellow dwarf virus (BYDV) than fields planted to multiple cultivars (Power 1991). Similarly, intercropping corn with beans suppressed vector populations and the incidence of viral disease in

Nicaragua (Power 1987). Additionally, diverse communities of grassland plants experience a lower incidence of foliar fungal diseases than simpler plant communities (Mitchell et al. 2002).

The role of plant diversity suppressing outbreaks of viral disease is highly variable.

Smithson and Lenne (1996) found that while most species and varietal mixtures provided some level of protection against pathogens in agricultural environments, pathogen incidence in diverse stands can range from lower to higher than the incidence observed in pure stands. This broad range of results may be due to functional ecological diversity, and not necessarily genetic diversity, being the primary driver behind disease reduction (Leonard 1969). To provide a buffer against disease spread, different plant genotypes must have functional differences that variably affect the specific pathogen or vector (Mundt 2002); similarly, adding a new plant species to the community will only make a difference if its susceptibility to the vector, pathogen, or both varies from the species already present (Leonard 1969). Additionally, spatial arrangement of plant species or genotypes can be as important as diversity in predicting plant virus epidemics (Garrett and Mundt 1999). Thus, a stand containing many functionally diverse species arranged in a spatially heterogeneous manner is likely to be a more challenging environment for plant viruses than a less diverse stand.

The pathogen spillover effect

Greater plant host diversity may increase viral prevalence in some scenarios involving a generalist pathogen, termed the pathogen spillover effect (sensu Daszak et al. 2000). In situations

9 where a virus is capable of infecting multiple plant species and possesses varying levels of interspecific pathogenicity, the addition of a new species may exacerbate infectious disease if it serves as a reservoir harboring long-term, relatively high titers of the virus. This phenomenon has been observed in BYDV, which infects multiple species of wild annual grasses. Power and

Mitchell (2004) found that when diversity in experimental plots was augmented with the addition of wild oats (Avena fatua), a highly susceptible reservoir of BYDV, disease prevalence was much higher than simpler, A. fatua-free plots. The role of A. fatua as a reservoir for BYDV may explain why invasive A. fatua rapidly displaces native grasses in California (Malmstrom et al.

2005). In these ways, the pathogen spillover effect has similar impacts as apparent competition between herbivores. Consequently, in managed environments such as agriculture and landscaping this phenomenon underscores the importance of identifying a particular species’ functional relationship to the pathogen before introducing it to augment biodiversity.

INFLUENCE OF NON-VECTOR INSECTS ON VIRAL SPREAD

Non-vector invertebrates, including predators and competing herbivores, have considerable impacts on the behavior and abundance of vector species (reviewed in Price et al.

1980, Denno et al. 1995) and the physiology of potential plant hosts (Karban and Baldwin 1997).

Non-vectors may consequently influence viral spread indirectly through their interactions with vectors and hosts (Fig. 1.2), even though they do not directly host or vector the virus. Exploring how non-vector guilds influence plant virus dispersal is important to achieving a holistic understanding of viral dynamics in any given community.

Role of herbivore competitors

Surprisingly little research has examined the role of non-vector herbivores in influencing viral dispersal, despite the fact that such herbivores compete for resources on the same plants as vectors. Research examining host and vector responses to herbivory, however, allows us to identify several potential mechanisms by which herbivory may influence viral dispersal.

Herbivory triggers a wide range of physiological changes in host plants, and these changes can influence several key components of viral spread such as vector fitness, vector host selection, and host susceptibility.

First, herbivory may reduce the nutritional content of a plant (Weibull 1987, de

Mazencourt et al. 1998). Although nutritional reduction of the host means lower vector fitness and thus fewer potential vectors (Awmack and Leather 2002), nutritionally deficient plants also stimulate enhanced herbivore movement as the insects search for more suitable hosts (Kareiva

1982). High vector mobility has been correlated with enhanced viral spread (Cornwell 1960,

Roitberg and Myers 1978, Thresh et al. 1983), so reductions in host quality from competing herbivore species could potentially enhance viral spread by increasing vector movement despite a potential reduction in host fitness. However, this hypothesis has yet to be directly tested.

Additionally, herbivory often induces the production of phytohormones that regulate the production of defensive compounds (Karban and Baldwin 1997, Glazebrook 2005). The hormone salicylic acid (SA) tends to produce compounds that defend against biotrophic pathogens such as viruses (Delaney et al. 1994, Verberne et al. 2000), whereas the jasmonic acid

(JA) pathway stimulates production of compounds with activity against necrotrophic pathogens and insect herbivores (Arimura et al. 2000, Ozawa et al. 2000). However, many piercing-sucking insects, such as aphids and whiteflies, stimulate the salicylic pathway (Leitner et al. 2005, Gosset et al. 2009, Ozawa et al. 2000, Zarate et al. 2007), and plants that are attacked by aphids often

11 show a transcriptional and phenotypic response similar to the response of plants attacked by biotrophic pathogens (Kaloshian and Walling 2005). Other piercing-sucking insects stimulate both pathways simultaneously (Walling 2008). The diversity of plant defense responses to piercing-sucking insects can make predicting these responses difficult.

Establishing the specific defense pathway elicited by an insect is important, because it may influence defense against pathogens. The JA-SA defense pathways have often been characterized as antagonistic, meaning that the induction of one pathway suppresses the expression of the other (Niki et al. 1998, Thaler et al. 2012). Consequently, while plants that are attacked by pests triggering a given pathway may become tolerant to subsequent attackers that trigger the same pathway, they may be more susceptible to pests that require another pathway for effective defense. For instance, tomato plants that are infected with Tobacco mosaic virus show elevated levels of SA, which suppresses production of JA and renders them susceptible to attack from defoliating caterpillars (Thaler et al. 2010). This cross-talk between phytohormones provides a theoretical framework for predicting plant performance in the context of attack from multiple antagonists.

Extension of the theoretical basis underlying phythormone crosstalk would result in a prediction that herbivores triggering the SA pathway would increase host plant tolerance to viruses and other biotrophic pathogens, while herbivores triggering the JA pathway would reduce tolerance. However, while there is ample evidence that biotrophic pathogens triggering the SA pathway render plants more susceptible to chewing herbivores (Spoel et al. 2007, Thaler et al. 2010), there is little evidence that JA increases susceptibility to biotrophic pathogens (Oka et al. 2013), and virtually no evidence that herbivores triggering the JA pathway increase pathogen susceptibility. The suppressive effects of JA on SA may be weaker than the

12 suppressive effects of SA on JA (Thaler et al. 2012), which may explain this relative dearth of evidence. However, several authors have noted that JA is often expressed in concert with SA when defending against biotrophic pathogens (Salzman et al. 2005, Zhu et al. 2014a, Yang et al.

2016), and elevated levels of JA have been shown to confer tolerance to viruses (Zhu et al.

2014a, Zhu et al. 2014b, Zhang et al. 2016). For instance, foliar application of methyl jasmonate, a JA derivative, increases rice plant tolerance to rice black streaked dwarf virus (He et al. 2016).

These observations seemingly defy the paradigm of SA-JA antagonism. Additionally, while increases in SA (either through exogenous application or genetic modification) seem to increase tolerance to biotrophic pathogens (Delaney et al. 1994), there is little evidence that piercing- sucking insects triggering the SA pathway will similarly increase tolerance to pathogens. Thus, while theory predicts that feeding by non-vector insect herbivores should influence host susceptibility to viruses, empirical evidence is scant, and inferring broad trends based on limited observations may not be prudent.

In addition to altering host susceptibility to pathogens, activation of anti-herbivore defenses by non-vectors could alter plant suitability for future herbivores, including potential vectors. By inducing phytohormones and defense compounds, non-vector herbivores could have either positive or negative effects on vector species, depending on which pathway is induced. For instance, population growth of the aphid Aphis nerii is reduced on milkweed plants damaged by monarch caterpillars (Ali and Agrawal, 2014), but population growth of the aphid Brevicoryne brassicae is higher on caterpillar-damaged Brassica oleracea plants (Soler et al. 2012). Shifts in plant host quality mediated by competing herbivores may also result in changes in vector behavior, specifically host-to-host movement, which could have implications for disease spread.

Although lower-quality hosts may result in suppression of vector populations, insect movement

13 often increases in situations of low host quality (Kareiva 1982). While this may promote disease spread under certain situations, viruses that require long feeding periods for acquisition, such as persistently transmitted barley yellow dwarf virus, are negatively influenced by high rates of vector movement (Power 1991, Long and Finke 2015). Thus, there are a wide range of potential host-mediated effects of non-vector competitors on vectors, and the resulting outcome for insect- borne pathogens is dependent upon the life history of the pathogen.

Spatial distributions of vector species could also be influenced by competing non-vector herbivores. Competition for a resource may result in niche partitioning and competitive displacement (Connell 1961), and while direct empirical evidence is scant, it is likely that competitive interactions between herbivores influence insect-borne pathogen spread.

Competitive exclusion of a vector species could negatively impact a pathogen, while elimination of a competitor could result in vector range expansion and increases in pathogen prevalence.

Interspecific competition is a significant constraint on the ranges of Aedes albopictus and A. aegypti (Juliano and Luonibos 2005), invasive mosquitos that transmit several animal pathogens, and these patterns of competitive exclusion are likely to have consequences for Aedes- transmitted pathogens such as yellow fever and dengue. Although the effects of non-vector competitors on plant pathosystems have received less attention, many of the same principles are likely to apply.

Role of predators

Predators can indirectly influence viral transmission by affecting vector behavior and abundance. Natural enemies slow the growth of vector populations (Fox et al. 2004), which may reduce viral transmission. In addition to directly ingesting vector species, predators also elicit

14 anti-predator behaviors in vector species (Tamaki et al. 1970) that come at a negative fitness cost to the vector (Lind and Cresswell 2005). In accordance with this trend, predators of the vector species M. persicae significantly suppress the spread of beet yellows virus in sugar beet fields

(Landis and Van der Werf 1997). However, high predator densities do not always effectively control vectors (Grylls 1972), and decreased vector abundances do not always correlate with lower incidence of insect-vectored disease (Bragard et al. 2013, Finke 2012). Consequently, some researchers question the disease-suppressing quality of natural enemies.

Natural enemies could potentially enhance the spread of a virus by increasing the mobility of infectious vectors. Predator foraging behavior increases aphid movement (Dixon

1958), and aphids regularly produce alarm pheromones in the presence of predators that cause them to drop from their host plant (Nault 1973). If these dislodged aphids are inoculative and re- establish on a new host, they could potentially spread the pathogen. Enclosure experiments have demonstrated that lady beetles regularly dislodge aphids and thus contribute to viral spread

(Roitberg and Myers 1978, Smyrnioudis et al. 2001), but cages with specialist parasitoids maintain the same rate of virus spread as predator-less controls (Smyrnioudis et al. 2001).

Additionally, Irwin (1999) noted that predators failed to provide effective suppression of soybean mosaic potyvirus in soybean crops. Predators of aphids transmitting barley yellow dwarf virus resulted in increased vector movement, resulting in diminished pathogen spread because high rates of movement interfered with pathogen acquisition (Long and Finke 2014). In sum, predators have the potential to both suppress and enhance the spread of a plant virus, depending on the environment.

Role of mutualists

Mutualistic interactions between vector and non-vector species could also influence pathogen spread, although this idea has rarely been tested. For example, ants often form mutualistic relationships with aphids, providing services such as protection in return for nutrition

(honeydew excrement) (Way 1963, Goggin 2007). The implications of these mutualisms for viral dispersal, however, are unclear. In the southern United States, invasive red imported fire ant

(Solenopsis invicta) colonies have begun tending broods of the cotton aphid, Aphis gossypii

(Rice and Eubanks 2013), a vector of several economically important plant viruses such as cucumber mosaic virus (Escriu et al. 2000). The potential role of S. invicta in virus spread found that aphid colonies tended by ants not only exhibited enhanced growth, but also tended to switch hosts more frequently (Cooper 2005). The influence of mutualistic ants on viral dispersal patterns, however, has never been directly measured, despite the potentially significant impact these studies suggest.

CONCLUSION

Insect-vectored plant viruses are often key contributors to community structure.

Consequently, we need to understand the ecological factors underlying their spread. A better understanding of plant virus outbreaks may allow more effective mitigation practices in agricultural landscapes, where viruses are economically damaging. Most research in this area has focused on the ecological role of virus vectors and plant hosts, especially ways in which viruses have manipulated these two guilds to their advantage. However, non-vector members of the insect community, including competitors, predators, and mutualists, have the potential to indirectly influence viral spread by altering the physiology, fitness, and behavior of vectors and hosts alike. Current understanding of how these functional assemblages may be shaping viral

16 disease dynamics in plants is incomplete. Consequently, more research is needed to deconstruct the various mechanisms at work in these multi-faceted pathosystems.

LITERATURE CITED

Ajayi, O., and A. M. Dewar. 1983. The effect of barley yellow dwarf virus on field populations

of the cereal aphids, Sitobion avenae and Metopolophium dirhodum. Annals of Applied

Biology 103: 1-11.

Ali, J. G. and A. A. Agrawal. 2014. Asymmetry of plant-mediated interactions between specialist

aphids and caterpillars on two milkweeds. Functional Ecology 28: 1404-1412.

Anderson, P.K., A. A. Cunningham, N. G. Patel, F. J. Morales, P. R. Epstein, and P. Daszak.

2004. Emerging infectious diseases of plants: pathogen pollution, climate change and

agrotechnology drivers. Trends in Ecology and Evolution 19: 535–544.

Arimura, G., R. Ozawa, T. Shimoda, T. Nishioka, W. Boland, and J. Takabayashi, J. 2000.

Herbivore-induced volatiles elicit defense genes in lima bean leaves. Nature 406: 512–515.

Awmack, C.S. and S. R. Leather. 2002. Host plant quality and fecundity in herbivorous insects.

Annual Review of Entomology 47: 817–844.

Belliure, B., A. Janssen, P. C. Maris, D. Peters, and M. W. Sabelis. 2005. Herbivore arthropods

benefit from vectoring plant viruses. Ecology Letters 8: 70–79.

Blanc, S., M. Uzest, and M. Drucker. 2011. New research horizons in vector-transmission of

plant viruses. Current Opinion in Microbiology 14: 483–491.

Bosque-Pérez, N.A., and S. D. Eigenbrode. 2011. The influence of virus-induced changes in

plants on aphid vectors: Insights from luteovirus pathosystems. Virus Research 159: 201–

205.

Bragard, C., P. Caciagli, O. Lemaire, J. J. Lopez-Moya, S. MacFarlane, D. Peters, P. Susi, and

L. Torrance. 2013. Status and prospects of plant virus control through interference with

vector transmission. Annual Review of Phytopathology 51: 177–201.

Bukovinszky, T., F. J. F. van Veen, Y. Jongema, and M. Dicke. 2008. Direct and indirect effects

of resource quality on food web structure. Science 319: 804–807.

Carr, D.E. and M. D. Eubanks. 2002. Inbreeding alters resistance to insect herbivory and host

plant quality in Mimulus guttatus (Scrophulariaceae). Evolution 56: 22–30.

Carter, W. 1962. Insects in relation to plant disease. Interscience (Wiley), New York.

Casteel, C. L. and G. Jander. 2013. New synthesis: investigating mutualisms in virus-vector

interactions. Journal of Chemical Ecology 39: 809.

Casteel, C. L., C. Yang, A. C. Nanduri, H. D. De Jong, S. A. Whitham, and G. Jander. 2014. The

Nia-Pro protein of Turnip mosaic virus improves growth and reproduction of the aphid

vector, Myzus persicae (green peach aphid). The Plant Journal 77: 653-663.

Connell, J. H. 1961. The influence of interspecific competition and other factors on the

distribution of the barnacle Chthamalus stellatus. Ecology 42: 710-723.

Cooper, L. 2005. The potential effects of red imported fire ants (Solenopsis invicta) on arthropod

abundance and Cucumber mosaic virus. Auburn University Graduate Thesis.

Cornwell, P.B. 1960. Movements of the vectors of virus diseases of cacao in Ghana—Wind

movements and aerial dispersal. Bulletin of Entomological Research 51: 175–201.

Daszak, P., A. A. Cunningham, and A. D. Hyatt. 2000. Emerging infectious diseases of wildlife--

threats to biodiversity and human health. Science 287: 443–449. de Mazancourt, C., M. Loreau, and L. Abbadie. 1998. Grazing optimization and nutrient cycling:

When do herbivores enhance plant production? Ecology 79: 2242–2252.

Delaney, T.P., S. Uknes, B. Vernooij, L. Friedrich, K. Weymann, D. Negrotto, T. Gaffney, M.

Gut-Rella, H. Kessmann, E. Ward, and J. Ryals. 1994. A central role of salicylic acid in plant

disease resistance. Science 266: 1247–1250.

Denno, R.F., M. S. McClure, and J. R. Ott. 1995. Interspecific interactions in phytophagous

insects: Competition reexamined and resurrected. Annual Review of Entomology 40: 297–

331.

Dixon, A.F.G. 1958. The escape responses shown by certain aphids to the presence of the

coccinellid Adalia decempunctata. Transactions of the Royal Entomological Society of

London 110: 319–334.

Dobson, A. and M. Crawley. 1994. Pathogens and the structure of plant communities. Trends in

Ecology and Evolution 9: 393–398.

Drucker, M., R. Froissart, E. Hébrard, M. Uzest, M. Ravallec, P. Espérandieu, J.-C. Mani, M.

Pugnière, F. Roquet, A. Fereres, and. S. Blanc. 2002. Intracellular distribution of viral gene

products regulates a complex mechanism of Cauliflower mosaic virus acquisition by its aphid

vector. PNAS 99: 2422–2427.

Eigenbrode, S.D., H. Ding, P. Shiel, and P. H. Berger. 2002. Volatiles from potato plants

infected with potato leafroll virus attract and arrest the virus vector, Myzus persicae

(Homoptera: Aphididae). Proc. R. Soc. Lond. B 269: 455–460.

Escriu, F., K. L. Perry, and F. García-Arenal. 2000. Transmissibility of cucumber mosaic virus

by Aphis gossypii correlates with viral accumulation and is affected by the presence of its

satellite RNA. Phytopathology 90: 1068–1072.

Fereres, A. and A. Moreno. 2009. Behavioural aspects influencing plant virus transmission by

homopteran insects. Virus Research 141: 158–168.

Finckh, M.R., E. S. Gacek, H. Goyeau, C. Lannou, U. Merz, C. C. Mundt, L. Munk, J. Nadziak,

A. C. Newton, C, de Vallavieille-Pope, and M. S. Wolfe. 2000. Cereal variety and species

mixtures in practice, with emphasis on disease resistance. Agronomie 20: 813–837.

Finke, D. L. 2012. Contrasting the consumptive and non-consumptive cascading effects of

natural enemies on vector-borne pathogens. Entomol. Exp. Appl. 144: 45-55.

Fox, T.B., D. A. Landis, F. F. Cardoso, and C. D. Difonzo. 2004. Predators suppress Aphis

glycines Matsumura population growth in soybean. Environmental Entomology 33: 608–618.

Garrett, K.A., and C. C. Mundt. 1999. Epidemiology in mixed host populations. Phytopathology

89: 984–990.

Glazebrook, J. 2005. Contrasting mechanisms of defense against biotrophic and necrotrophic

pathogens. Annual Review of Phytopathology 43: 205–227.

Goggin, F.L. 2007. Plant–aphid interactions: molecular and ecological perspectives. Current

Opinion in Plant Biology 10: 399–408.

Gosset, V., N. Harmel, C. Göbel, F. Francis, E. Haubruge, J.-P. Wathelet, P. du Jardin, I.

Feussner, and M.-L. Fauconnier. 2009. Attacks by a piercing-sucking insect (Myzus persicae

Sultzer) or a chewing insect (Leptinotarsa decemlineata Say) on potato plants (Solanum

tuberosum L.) induce differential changes in volatile compound release and oxylipin

synthesis. J. Exp. Bot. 60: 1231–1240.

Gray, S.M. and N. Banerjee. 1999. Mechanisms of arthropod transmission of plant and animal

viruses. Microbiol. Mol. Biol. Rev. 63: 128–148.

Grylls, N. 1972. Aphid infestation and virus infection of peas and beans on the Central

Tablelands of New South Wales. Aust. J. Exp. Agric. 12: 668–674.

He, Y., Zhang, H., Sun, Z., Li, J., Hong, G., Zhu, Q., Zhou, X., MacFarlane, S., Yan, F., and J.

Chen. 2016. Jasmonic acid-mediated defense suppresses brassinosteroid-mediated

susceptibility to Rice black streaked dwarf virus infection in rice. New Phytologist 214: 388–

399.

Hodge, S. and G. Powell. 2010. Conditional facilitation of an aphid vector, Acyrthosiphon pisum,

by the plant pathogen, pea enation mosaic virus. J. Insect Sci. 10: 155.

Hogenhout, S.A., E.-D. Ammar, A. E. Whitfield, and M. G. Redinbaugh. 2008. Insect vector

interactions with persistently transmitted viruses. Annual Review of Phytopathology 46: 327–

359.

Ingwell, L.L., S. D. Eigenbrode, and N. A. Bosque-Pérez. 2012. Plant viruses alter insect

behavior to enhance their spread. Sci. Rep. 2: 578.

Irwin, M.E. 1999. Implications of movement in developing and deploying integrated pest

management strategies. Agricultural and Forest Meteorology 97: 235–248.

Jiménez-Martínez, E.S., N. A. Bosque-Pérez, P. H. Berger, R. S. Zemetra, H. Ding, and S, D,

Eigenbrode. 2004. Volatile cues influence the response of Rhopalosiphum padi (Homoptera:

Aphididae) to barley yellow dwarf virus–infected transgenic and untransformed wheat.

Environmental Entomology 33: 1207–1216.

Juliano, S. A. and L. P. Lounibos. 2005. Ecology of invasive mosquitoes: effects on resident

species and on human health. Ecology Letters 8: 558-574.

Kaloshian, I., and L. L. Walling. 2005. Hemipterans as plant pathogens. Annual Review of

Phytopathology 43: 491–521.

Karban, R., and I. Baldwin. 1997. Induced responses to herbivory. University of Chicago Press,

Chicago.

Kareiva, P. 1982. Experimental and mathematical analyses of herbivore movement: quantifying

the influence of plant spacing and quality on foraging discrimination. Ecological

Monographs 52: 261–282.

Kennedy, J.S., M. F. Day, and V. F. Eastop. 1962. A conspectus of aphids as vectors of plant

viruses. Eastern Press Ltd., London.

Landis, D.A. and V. W. der Werf. 1997. Early-season predation impacts the establishment of

aphids and spread of beet yellows virus in sugar beet. Entomophaga 42: 499–516.

Legg, J. 1999. Emergence, spread and strategies for controlling the pandemic of cassava mosaic

virus disease in east and central Africa. Crop Protection 18: 627–637.

Legg, J.P. and C. M. Fauquet. 2004. Cassava mosaic geminiviruses in Africa. Plant Mol Biol 56:

585–599.

Legg, J.P., and S. Ogwal. 1998. Changes in the incidence of African cassava mosaic virus

disease and the abundance of its whitefly vector along south-north transects in Uganda.

Journal of Applied Entomology 122: 169-178.

Leitner, M., W. Boland, and A. Mithöfer. 2005. Direct and indirect defenses induced by

piercing-sucking and chewing herbivores in Medicago truncatula. New Phytologist 167:

597–606.

Leonard, K.J. 1969. Factors affecting rates of stem rust increase in mixed plantings of

susceptible and resistant oat varieties. Phytopathology 59.

Lewsey, M.G., A. M. Murphy, D. MacLean, N. Dalchau, J. H. Westwood, K. Macaulay, M. H.

Bennett, M. Moulin, D. E. Hanke, G. Powell, A. G. Smith and J. P. Carr. 2010. Disruption of

two defensive signaling pathways by a viral RNA silencing suppressor. Molecular Plant-

Microbe Interactions 23: 835–845.

Lind, J. and W. Cresswell. 2005. Determining the fitness consequences of antipredation

behavior. Behavioral Ecology 16: 945–956.

Long, E. Y. and D. L. Finke. 2014. Contribution of predator identity to the suppression of

herbivores by a diverse predator assemblage. Environmental Entomology 43: 569-576.

Long, E. Y. and D. L. Finke. 2015. Predators indirectly reduce the prevalence of an insect-

vectored plant pathogen independent of predator diversity. Oecologia 177: 1067-1074.

Malmstrom, C.M., A. J. McCullough, H. A. Johnson, L. A. Newton, and E. T. Borer. 2005.

Invasive annual grasses indirectly increase virus incidence in California native perennial

bunchgrasses. Oecologia 145: 153–164.

Malmstrom, C.M., U. Melcher, and N. A. Bosque-Pérez. 2011. The expanding field of plant

virus ecology: Historical foundations, knowledge gaps, and research directions. Virus

Research 159: 84–94.

Martinière, A., A. Bak, J.-L. Macia, N. Lautredou, D. Gargani, J. Doumayrou, E. Garzo, A.

Moreno, A. Fereres, S. Blanc, and M. Drucker. 2013. A virus responds instantly to the

presence of the vector on the host and forms transmission morphs. eLife 2.

Mauck, K.E., C. M. D. Moraes, and M. C. Mescher. 2010. Deceptive chemical signals induced

by a plant virus attract insect vectors to inferior hosts. PNAS 107: 3600–3605.

Mauck, K., N. A. Bosque-Pérez, S, D, Eigenbrode, C. M. De Moraes, and M. C. Mescher, M.C.

2012. Transmission mechanisms shape pathogen effects on host–vector interactions:

evidence from plant viruses. Functional Ecology 26: 1162–1175.

McElhany, P., L. A. Real, and A. G. Power, A.G. 1995. Vector preference and disease dynamics:

A study of Barley yellow dwarf virus. Ecology 76: 444–457.

Mitchell, C.E. and A. G. Power. 2003. Release of invasive plants from fungal and viral

pathogens. Nature 421: 625–627.

Mitchell, C.E., D. Tilman, and J. V. Groth. 2002. Effects of grassland plant species diversity,

abundance, and composition on foliar fungal disease. Ecology 83: 1713–1726.

Mundt, C.C. 2002. Use of multiline cultivars and cultivar mixtures for disease management.

Annual Review of Phytopathology 40: 381–410.

Nault, L.R. 1973. Alarm pheromones help aphids escape predators. Ohio Rep. Res. Develop Agr.

Home Econ. Natur. Resour. 58.

Nault, L.R. 1997. Arthropod transmission of plant viruses: A new synthesis. Annals of the

Entomological Society of America 90: 521–541.

Niki, T., I. Mitsuhara, S. Seo, N. Ohtsubo, and Y. Ohashi. 1998. Antagonistic effect of salicylic

acid and jasmonic acid on the expression of pathogenesis-related (PR) protein genes in

wounded mature tobacco leaves. Plant Cell Physiology 39: 500–507.

Oka, K., M. Kobayashi, I. Mitsuhara, and S. Seo. 2013. Jasmonic acid vegatively regulates

resistance to Tobacco mosaic virus in tobacco. Plant & Cell Physiology 54: 1999-2010.

Ozawa, R., G. Arimura, J. Takabayashi, T. Shimoda, and T. Nishioka. 2000. Involvement of

jasmonate- and salicylate-related signaling pathways for the production of specific herbivore-

induced volatiles in plants. Plant Cell Physiology 41: 391–398.

Platt, K.B., K. J. Linthicum, K. S. Myint, B. L. Innis, K. Lerdthusnee, and D. W. Vaughn. 1997.

Impact of dengue virus infection on feeding behavior of Aedes aegypti. Am. J. Trop. Med.

Hyg. 57: 119–125.

Power, A.G. 1987. Plant community diversity, herbivore movement, and an insect-transmitted

disease of maize. Ecology 68: 1658–1669.

Power, A.G. 1991. Virus spread and vector dynamics in genetically diverse plant populations.

Ecology 72: 232.

Power, A.G. 2000. Insect transmission of plant viruses: a constraint on virus variability. Current

Opinion in Plant Biology 3: 336–340.

Power, A.G., and C. E. Mitchell. 2004. Pathogen spillover in disease epidemics. The American

Naturalist 164: S79–S89.

Price, P.W., C. E. Bouton, P. Gross, B. A. McPheron, J. N. Thompson, and A. E. Weis. 1980.

Interactions among three trophic levels: Influence of plants on interactions between insect

herbivores and natural enemies. Annual Review of Ecology and Systematics 11: 41–65.

Rice, K.B. and M. D. Eubanks. 2013. No enemies needed: Cotton aphids (Hemiptera: Aphididae)

directly benefit from red imported fire ant (Hymenoptera: Formicidae) tending. Florida

Entomologist 96: 929–932.

Roitberg, B.D. and J. H. Myers. 1978. Effect of adult coccinellidae on the spread of a plant virus

by an aphid. Journal of Applied Ecology 15: 775–779.

Rosenzweig, C., A. Iglesias, X. B. Yang, P. R. Epstein, and E. Chivian. 2001. Climate change

and extreme weather events: Implications for food production, plant diseases, and pests.

Global Change and Human Health 2: 90–104.

Salzman, R. A., J. A. Brady, S. A. Finlayson, C. D. Buchanan, E. J. Summer, F. Sun, P. E. Klein,

R. R. Klein, L. H. Pratt, M.-M. Cordonnier-Pratt, and J. E. Mullet. 2005. Transcriptional

profiling of sorghum induced by methyl jasmonate, salicylic acid and aminocyclopropane

carboxylic acid reveals cooperative regulation and novel gene responses. Plant Physiology

138: 352-368.

Sinisterra, X.H., C. L. McKenzie, W. B. Hunter, C. A. Powell, and R. G. Shatters, R.G. 2005.

Differential transcriptional activity of plant-pathogenic begomoviruses in their whitefly

vector (Bemisia tabaci, Gennadius: Hemiptera Aleyrodidae). J. Gen. Virol. 86: 1525–1532.

Smithson, J.B., and J. M. Lenné. 1996. Varietal mixtures: a viable strategy for sustainable

productivity in subsistence agriculture. Annals of Applied Biology 128: 127–158.

Smyrnioudis, I.N., R. Harrington, S. J. Clark, and S. J. Katis. 2001. The effect of natural enemies

on the spread of Barley yellow dwarf virus (BYDV) by Rhopalosiphum padi (Hemiptera:

Aphididae). Bull. Entomol. Res. 91: 301–306.

Soler, R., G. R. Badenes-Perez, C. Broekgaarden, S.-J. Zhen, A. David, W. Boland, and M.

Dicke. 2012. Plant-mediated facilitation between a leaf-feeding and a phloem-feeding insect

in the brassicaceous plant: from insect performance to gene transcription. Functional Ecology

26: 156-166.

Spoel, S. H., J. S. Johnson, X. Dong. 2007. Regulation of tradeoffs between plant defenses

against pathogens with different lifestyles. Proc Natl Acad Sci USA 104: 18842-18847.

Strange, R.N., and P. R. Scott. 2005. Plant disease: A threat to global food security. Annual

Review of Phytopathology 43: 83–116.

Sylvester, E.S. 1980. Circulative and propagative virus transmission by aphids. Annual Review of

Entomology 25: 257–286.

Tamaki, G., J. E. Halfhill, and D. O. Hathaway. 1970. Dispersal and reduction of colonies of pea

aphids by Aphidius smithi (Hymenoptera: Aphidiidae). Annals of the Entomological Society

of America 63: 973–980.

Thaler, J. S., A. A. Agrawal, and R. Halitschke. 2010. Salicylate-mediated interactions between

pathogens and herbivores. Ecology 91: 1075-1082.

Thaler, J.S., P. T. Humphrey, and N. K. Whiteman. 2012. Evolution of jasmonate and salicylate

signal crosstalk. Trends in Plant Science 17: 260–270.

Thresh, J.M., R. S. Scorer, R. Harrington, D. E. Pedgley, P. A. Nuttall, and R. F. Sellers. 1983.

The long-range dispersal of plant viruses by arthropod vectors. Phil. Trans. R. Soc. Lond. B

302: 497–528.

Underwood, N., and M. D. Rausher. 2000. The effects of host-plant genotype on herbivore

population dynamics. Ecology 81: 1565–1576.

Verberne, M.C., R. Verpoorte, J. F. Bol, J. Mercado-Blanco, and H. J. M. Linthorst. 2000.

Overproduction of salicylic acid in plants by bacterial transgenes enhances pathogen

resistance. Nat Biotech 18: 779–783.

Walling, L.L. 2008. Avoiding effective defenses: Strategies employed by phloem-feeding

insects. Plant Physiol. 146: 859–866.

Way, M.J. 1963. Mutualism between ants and honeydew-producing Homoptera. Annual Review

of Entomology 8: 307–344.

Weibull, J. 1987. Seasonal changes in the free amino acids of oat and barley phloem sap in

relation to plant growth stage and growth of Rhopalosiphum padi. Annals of Applied Biology

111: 729–737.

Yang, T., Meng, Y., Chen, L. J., Lin, H. H., and D. H. Xi. 2016. The roles of alpha-momorcharin

and jasmonic acid in modulating the response of Momordica charantia to Cucumber mosaic

virus. Front Microbiol 7.

Zarate, S.I., L. A. Kempema, and L. L. Walling. 2007. Silverleaf whitefly induces salicylic acid

defenses and suppresses effectual jasmonic acid defenses. Plant Physiol. 143: 866–875.

Zhang, C., Ding, Z., Wu, K., Yang, L., Li, Y., Yang, Z., Shi, S., Liu, X., Zhao, S., Yang, Z., et al.

2016. Suppression of jasmonic acid-mediated defense by viral-inducible MicroRNA319

facilitates virus infection in rice. Molecular Plant 9: 1302–1314.

Zhang, T., J.-B. Luan, J.-F. Qi, C.-J. Huang, M. Li, X.-P. Zhou, and S.-S. Liu. 2012.

Begomovirus–whitefly mutualism is achieved through repression of plant defenses by a virus

pathogenicity factor. Molecular Ecology 21: 1294–1304.

Zhu, Y., H. Chen, J. Fan, Y. Wang, J. Chen, J. Fan, S. Yang, L. Hu, H. Leung, T. W. Mew, P. A.

Teng, Z. Wang and C. C. Mundt. 2000. Genetic diversity and disease control in rice. Nature

406: 718-722.

Zhu, F., Xi, D. H., Deng, X. G., Peng, X. J., Tang, H., Chen, Y. J., Jian, W., Feng, H., and H. H.

Lin. 2014a. The Chilli veinal mottle virus regulates expression of the Tobacco mosaic virus

resistance gene N and jasmonic acid/ethylene signaling is essential for systemic resistance

against Chilli veinal mottle virus in tobacco. Plant Mol Biol Rep 32: 382–394.

Zhu, F., Xi, D. H., Yuan, S., Xu, F., Zhang, D. W., and H. H. Lin. 2014b. Salicylic acid and

jasmonic acid are essential for systemic resistance against Tobacco mosaic virus in Nicotiana

benthamiana. MPMI 27: 567–577.

Ziebell, H., A. M. Murphy, S. C. Groen, T. Tungadi, J. H. Westwood, M. G. Lewsey, M. Moulin,

A. Kleczkowski, A. G. Smith, M. Stevens, G. Powell, and J. P. Carr. 2011. Cucumber mosaic

virus and its 2b RNA silencing suppressor modify plant-aphid interactions in tobacco.

Science Reports 1: 187.

FIGURE LEGENDS

FIG. 1.1. Generalized dispersal strategy employed by persistent (A) and non-persistent (B) plant viruses. Persistent viruses generally increase host quality for the vector and require long periods of feeding for acquisition and inoculation. Vectors are attracted to an infected host and reproduce before dispersing. In contrast, non-persistent viruses generally decrease host quality and require relatively short periods of feeding for acquisition and inoculation. Vectors are attracted to an infected host, but quickly disperse because of poor host quality (Mauck et al. 2010).

FIG. 1.2. Non-vector insects have a variety of direct and indirect on pathogen vectors. Predators decrease vector populations and may increase vector movement. Herbivores directly impact vectors through physical displacement and competition, and indirectly through feeding-mediated shifts in plant quality and defenses. Mutualists, such as aphid-tending ants, can directly provision vectors and may also repel predators and competing herbivores.

FIG. 1.1.

FIG. 1.2.

CHAPTER TWO

RECIPROCAL PLANT-MEDIATED INTERACTIONS BETWEEN A PLANT VIRUS

AND A NON-VECTOR HERBIVORE

ABSTRACT

Plant viruses often alter physical and chemical traits of hosts, indirectly affecting the fitness and behavior of vectors in ways that promote virus transmission. However, whether viruses induce similar plant-mediated behavioral and fitness shifts in non-vector herbivores remains relatively unknown. Here we evaluated reciprocal interactions between pea enation mosaic virus (PEMV) and Sitona lineatus, a non-vector herbivore. In the field, PEMV-infected plants experienced greater defoliation by S. lineatus than uninfected plants; behavior assays similarly showed S. lineatus individuals preferred PEMV-infected plants. The phytohormone jasmonic acid, implicated in defense against S. lineatus, was suppressed in PEMV-infected plants, possibly explaining these preferences. In turn, inoculative pea aphid vectors preferred plants damaged by

S. lineatus, and herbivory from S. lineatus promoted increased titer of PEMV in plant hosts. Our results show strong reciprocal interactions between PEMV and S. lineatus, which are mediated by changes in weevil behavior in response to plant infection as well as changes in aphid behavior and plant susceptibility in response to S. lineatus. Importantly, these interactions appear to benefit the pathogen. As plant virus and vectors exist in diverse communities, our study highlights the importance of non-vector species in influencing plant pathogens, and vice versa, through host-mediated effects.

KEYWORDS: disease ecology, plant-insect interactions, plant-pathogen interactions, indirect effects, pathosystems, phytohormone cross-talk

INTRODUCTION

Plant viruses are a major driver of biodiversity loss in natural plant populations, and have devastating effects on global crop productivity (Strange and Scott 2005). Plant viruses, which rely almost exclusively on arthropod vectors for dispersal to new hosts (Power 2000; Hogenhout et al. 2008), often form mutualistic relationships with vectors (e.g., Fereres et al. 1989, Castle and Berger 1993, Belliure et al. 2005, Bosque-Perez and Eigenbrode 2011, Mauck et al. 2012).

In plants, viruses often suppress defensive responses targeting vectors (Zhang et al. 2012, Luan et al. 2013, Li et al. 2014) and increase nutrient availability for insect herbivores (Casteel et al.

2014). While such vector-virus mutualisms are well studied, vectors exist in diverse communities, where they interact with individuals of many species, including plants, competitors, predators, and mutualists. Yet, the role of these interactions on disease ecology has received relatively little attention (Selakovic et al. 2014, Seabloom et al. 2015).

As plant viruses affect plant defenses and nutrients, they might also impact non-vector herbivores that co-occur with vector species. Indeed, empirical studies have observed positive

(Hare and Dodds 1987, Belliure et al. 2010, Kersch-Becker and Thaler 2013, Thaler et al. 2010), negative (Mauck et al. 2010b, Molken et al. 2012, Pan et al. 2013), and neutral (Sadeghi et al.

2016) effects of plant viruses on the fitness of non-vector herbivores. However, most of these studies were conducted in controlled environments, and evidence of interactions between plant viruses and non-vector herbivores in the field remains scant (but see Mauck et al. 2010a).

Moreover, the potential effects of plant viruses on non-vector behavior have been poorly studied.

Conversely, because herbivory stimulates a wide range of physiological responses in plants (Karban and Baldwin 1997), non-vector herbivores might alter plant susceptibility to pathogens. Phytohormone crosstalk, where induction of plant defenses related to herbivory leads to a suppression of plant defenses targeting pathogens (Thaler et al. 2010, Abe et al. 2012), might mediate such interactions. For example, many insect herbivores stimulate the jasmonic acid pathway in plants (Bowles 1991), which can cause a down-regulation of salicylic acid through phytohormone crosstalk (Thaler et al. 2010, Abe et al. 2012, Thaler et al. 2012). As salicylic acid is commonly implicated in defense against biotropic pathogens (Thaler et al. 2010, Robert-

Seilaniantz et al. 2011), such interactions could have strong effects on pathogen dynamics.

Moreover, shifts in plant physiology induced by non-vector herbivores could affect vector fitness or host selection behavior. For example, Agrawal (1999) observed that aphid colonization of caterpillar-damaged plants was 30% less than undamaged controls, although this study did not examine the incidence of pathogens vectored by aphids. Consequently, while several studies have examined interactions between insect herbivores and potential vectors, the direct and indirect impacts of non-vector herbivores on plant pathogens transmitted by vectors have received comparatively little attention in the literature.

To address these knowledge gaps, we examined reciprocal interactions between a vector- borne plant virus, pea enation mosaic virus (PEMV), and a non-vector herbivore, the pea leaf weevil (Sitona lineatus). We hypothesized that S. lineatus feeding might impact plant defenses, which could affect plant susceptibility to PEMV and behavior of the primary PEMV vector, the pea aphid (Acrythosiphon pisum). Conversely, PEMV might affect the behavior of S. lineatus by affecting plant physiology. We combined a large-scale field study examining relationships between S. lineatus and PEMV with a series of greenhouse assays to examine these interactions.

Our study shows that reciprocal plant-mediated interactions between a virus and a non-vector herbivore exert complex effects on vectors and plants in ways that promote virus transmission.

METHODS

Natural history of study system

PEMV infects multiple crop (Clement 2006) and native (Chisholm, personal observation) plant species in the Palouse region of eastern Washington and northern Idaho. PEMV induces symptoms such as leaf-thickening, a mosaic-patterned chlorosis, and hyperplastic outgrowths

(enations) on undersides of leaves (Appendix S1: Fig. S1a). PEMV is obligately-transmitted by insects in a persistent, circulative manner, with pea aphids (A. pisum; Appendix S1: Fig. S1b) the primary vector (Clement et al. 2010). Once A. pisum individuals acquire PEMV from a plant, they can transmit it until their death. Sitona lineatus, the pea leaf weevil, is a common co- occuring herbivore that overwinters as an adult and infests new plants each spring. Adult S. lineatus (Appendix S1: Fig. S1c) feed on leaf margins and lay eggs in early summer; larvae hatch and feed on roots before pupating. Although S. lineatus cannot transmit PEMV, it co- occurs with aphids, PEMV, and plant hosts, and thus interactions between these species could influence PEMV transmission in the field.

Field observations of the relationship between S. lineatus and PEMV

We first explored the relationship between PEMV and S. lineatus by assessing defoliation on infected and uninfected pea plants in June 2014 on 12 farms throughout the Palouse region. In each field, we visually identified 3 to 10 PEMV-infected plants; sample size varied based on

PEMV prevalence (Appendix S2, Table S1). For each infected plant, an adjacent (< 10 cm away)

36 healthy plant of similar size was identified, creating a paired design. For each plant, we estimated defoliation by counting the number of weevil feeding notches (Bardner et al. 1983). The growth stage (number of vegetative nodes) and aphid population size for each plant was also recorded visually. The infection status of plants was later confirmed by clipping a leaf from the highest fully-emerged node and analyzing it with a commercial ELISA kit(AC Diagnostics, Fayetteville,

AR), using the provided protocol (Appendix S3).

Behavioral choice assays

A relationship between S. lineatus and PEMV might result from preferences of S. lineatus for infected host plants, or preferences of inoculative pea aphids for plants damaged by

S. lineatus. We explored this with two greenhouse (Appendix S3) assays that tested (1) if S. lineatus adults preferred infected compared to uninfected plants and (2) whether pea aphid vectors preferred plants damaged by S. lineatus compared to undamaged plants.

The S. lineatus assay was conducted by randomly exposing 18, 22-d-old pea plants to one of two treatments: (1) sham-inoculated or (2) PEMV-inoculated. Sham-inoculated plants were fed on by 8, 6-d-old uninfected pea aphids for 48h; PEMV-inoculated plants were fed on by 8, 6- d-old inoculative pea aphids for 48h. After 48h, aphids were physically removed, and one sham- and one PEMV-inoculated plant were potted 10 cm apart in a 10×25×10 cm plastic bin (9 experimental units total). Fifteen d post-inoculation, when PEMV symptoms manifest, two adult

S. lineatus were released in each experimental unit from a vial equidistant from the two plants and allowed to feed for 6d (Appendix S1, Fig. S2A). After 6d, leaves were excised, photographed, and analyzed for defoliation (surface area of removed leaf area) with ImageJ

(NIH) following Glozer (2008). Excised leaves were then processed with the rest of the plant material and analyzed with ELISA to confirm infectivity.

The aphid assay was conducted in a greenhouse by randomly exposing 32 18-d old plants to one of two treatments: (1) undamaged – no feeding by S. lineatus and (2) damaged – feeding by 2 adult S. lineatus for 48h (16 plants received each treatment). The S. lineatus were removed and the 16 pairs of plants (one damaged and one undamaged) were added to 8 separate cages. An undamaged leaf from each plant was inserted into opposite ends of a 20cm clear vinyl tube

(Appendix S1, Fig. S2B). Twenty-five adult PEMV-inoculative or non-inoculative aphids were placed in a vial and inserted into an opening at the center of the tube. Each cage contained 2 pairs, one pair for each aphid type (inoculative or non-inoculative) (Appendix S1: Fig. S2C).

After 3 h, individual settlement on both plants was recorded. The experiment included 8 replicates with both inoculative and non-inoculative aphids and was conducted in a greenhouse.

Aphids that did not make a choice were excluded from analyses.

Effects of S. lineatus on susceptibility of host plants to PEMV

Feeding by S. lineatus might also affect host susceptibility to PEMV. We assessed this with greenhouse assays that explored (1) if feeding by S. lineatus prior to PEMV-inoculation affected PEMV titer and (2) if feeding by S. lineatus after PEMV-inoculation affected PEMV titer. Both were conducted in bucket cages. For the first experiment, we randomly exposed 40 plants to three treatments: (1) undamaged – no S. lineatus (20 replicates); (2) low damage – feeding by 1 S. lineatus adult for 72h (14 replicates); and (3) high damage – feeding by 2 S. lineatus adults for 72h (6 replicates). After removing the S. lineatus adults, we added 8, 3d old, inoculative aphids to each plant; these aphids were allowed to feed for 6d. For the second

38 experiment, individual 14d old pea plants were inoculated with PEMV by allowing 8, 3d old, inoculative pea aphids to feed for 48h in a clip cage. After 48h, clip cages and aphids were removed and either 0 (3 replicates), 1 (5 replicates) or 3 (8 replicates) field-collected adult pea leaf weevils were introduced and allowed to feed for 6d.

At the end of both expeirments, the entire above-ground biomass of each pea plant was harvested and analyzed with ELISA and a Pierce BCA protein assay (ThermoFisher Scientific) to determine relative PEMV titer. Relative titers were calculated based on the relative reactivity of the sample normalized against the sample’s protein content using the following formula:

Sample absorbance ⁄ Negative control absorbance

Sample protein content

Effect of PEMV and S. lineatus on the defense response of pea

Indirect interactions between herbivores and pathogens can be mediated by plant defenses. Accordingly, we conducted a study to determine how plant phytohormone production

(jasmonic acid and salicylic acid) responds to attack from S. lineatus compared to PEMV. The first assay randomly exposed 12, 14-d old, pea plants to two treatments: (1) sham-inoculated or

(2) PEMV-inoculated (six replicates each). The inoculations were performed as previously described, and plant tissue was collected 2wk after treatments. In the second assay, we randomly exposed 14, 28-d old pea plants to two treatments: (1) undamaged – no feeding by S. lineatus (6 replicates) and (2) damaged – feeding by 5 adult S. lineatus for 24h (8 replicates). After 24h, leaves with beetle damage were harvested. Plant tissue was harvested and tested for jasmonic acid and salicylic acid as in Casteel et al. (2015) (Appendix S3).

Data analysis

From the field data we tested the association between PEMV-infection (uninfected vs. infected) and three variables: (1) S. lineatus defoliation (number of feeding notches), (2) aphid density (aphids per plant), and (3) plant size (number of nodes) using paired t-tests. In these analyses, pairs of uninfected and infected plants across the 12 farms served as replicates.

From the behavioral assays we used paired t-tests to determine if S. lineatus preferred to feed on uninfected vs. infected plants; we conducted separate tests with total leaf area and proportion leaf area removed as response variables. Paired t-tests were also used to determine if uninfected and inoculative aphids preferred plants damaged by S. lineatus compared to undamaged plants.

To determine if prior S. lineatus feeding on uninfected pea plants affected PEMV titer, we conducted two analyses. First, we conducted a t-test assessing the effects of herbivore presence (pooled across two S. lineatus densities) vs. absence on PEMV titer. Second, we conducted a linear regression assessing whether S. lineatus density (0, 1, or 2) affected PEMV titer. In each of these analyses, 3 plants that did not become infected during the experiment were excluded. To determine if weevil feeding post-inoculation influenced PEMV titer, we first conducted an ANOVA assessing whether PEMV titer differed based on weevil density (0 vs. 1 vs. 3). Second, we analyzed PEMV titer as a function of total defoliation (cm2 leaf removed), or as a function of the proportion of leaf area defoliated, using separate linear regressions.

Finally, we used two-sample t-tests to determine if levels of jasmonic acid and salicylic acid differed for sham-inoculated vs. PEMV-inoculated plants, and for undamaged vs. S. lineatus-damaged plants. We excluded some experimental replicates when chromatograms were too noisy to assess phytohormones (Appendix S3). All statistical analyses were conducted in R

Studio (ver. 0.99.903).

RESULTS

Relationship between PEMV and S. lineatus in the field

PEMV-infected plants had significantly more S. lineatus feeding notches than adjacent uninfected plants (t89 = 3.99, P = 0.0001; Fig. 2.1A), despite infected plants being significantly smaller (t89 = -2.02, P = 0.046; Appendix S1: Fig. S3A). Aphid populations did not vary between infected and adjacent uninfected plants (t89 = 0.31, P = 0.76, Appendix S1: Fig. S3B).

Behavioral preferences of S. lineatus and A. pisum

Adult S. lineatus preferentially fed on PEMV-infected compared to uninfected host plants, as measured by total area defoliated (t8 = 2.85, P = 0.022, Fig. 2.1B) and the proportion of leaf area defoliated (t8 = 3.03, P = 0.016). Inoculative pea aphids also preferentially fed on plants damaged by S. lineatus compared to uninfected plants (Fig. 2.1C), although the effect was only marginally significant (t7 = -1.90, P = 0.099). Uninfected pea aphids did not show preferences for damaged vs. undamaged plants (t7 = 0.53, P = 0.62, Appendix S1: Fig. S4).

Effects of S. lineatus on PEMV accumulation in plants

Prior herbivory by S. lineatus on uninfected plants that were then inoculated with PEMV did not impact PEMV titer (t35 = 0.36, P = 0.72, Appendix S1: Fig. S5A), regardless of the number of S. lineatus (t1 = -0.082, P = 0.94, Appendix S1: Fig. S5B). However, greater numbers of S. lineatus increased the titer of PEMV on plants that were already inoculated (F2,13 = 5.51, P

= 0.019, Fig. 2.2A). There was also a positive relationship between the amount of leaf area removed by S. lineatus and PEMV titer (t1 = 2.04, P = 0.061, Fig. 2.2B); a similar relationship

41 was found between the proportion of leaf area removed and PEMV titer (t1 = 2.05, P = 0.060,

Appendix S1: Fig. S6).

Effects of PEMV and S. lineatus on phytohormone production

Plants inoculated with PEMV had a lower concentration of salicylic acid (t7 = 3.97, P =

0.0054, Fig. 2.3A), and a lower concentration of jasmonic acid (t9 = -3.12, P = 0.012, Fig. 2.3B), than sham-inoculated plants. Compared to undamaged plants, plants damaged by S. lineatus had similar levels of salicylic acid (t11 = 0.68, P = 0.51, Fig. 2.3C), but higher levels of jasmonic acid

(t9 = 2.45, P = 0.037, Fig. 2.3D).

DISCUSSION

Herbivores that transmit plant viruses are components of diverse food webs, where they interact with heterospecific individuals as they forage, seek mates, and avoid predation. While the study of food webs shows that species interactions can generate complex dynamics for focal populations (Polis and Strong 1996, Berlow et al. 2004) and ecosystems (Baiser et al. 2011), the impacts of species interactions for pathogens remains poorly understood (Johnson et al. 2015).

Here we demonstrate strong reciprocal interactions between a plant virus and a non-vector herbivore. While some studies have documented impacts of plant viruses on non-vector herbivores (Vega et al. 1995, Thaler et al. 2010, Mauck et al. 2010a), our study is among the first to show that a non-vector species can also affect the dynamics of a plant virus.

Interactions between S. lineatus and PEMV were mutualistic. We hypothesized that if S. lineatus arrives before aphids into fields, their feeding might make plants more susceptible to

PEMV or more attractive to aphids. Our results support the latter hypothesis. While chewing

42 herbivores often cause a decline in aphid fitness (i.e., Agrawal 1998; Thaler et al. 2001, Ali and

Agrawal 2014), such effects are strongly context-dependent and benefits of chewing herbivores for later arriving species have also been observed (Zarate et al. 2007, Thaler et al. 2010).

Interestingly, only inoculative pea aphids preferred damaged plants. While virus-infection can affect the preferences of vectors for host plants that are infected or not (reviewed by Mauck et al.

2012), our study demonstrates similar effects of virus-infection on vector preferences for plants damaged by S. lineatus. Models have shown that vector preferences can mediate rates of disease spread (i.e., McElheny et al. 1995; Roosein et al. 2013), suggesting that such behavioral outcomes of species interactions should be more broadly incorporated into studies of disease.

Feeding by S. lineatus prior to PEMV infection did not affect PEMV titers in infected hosts. However, feeding by S. lineatus increased titer in plants that were already infected, despite the fact that feeding by S. lineatus did not affect concentrations of salicylic acid, the primary defensive pathway induced by PEMV. While we do not know similar studies that have evaluated the context-dependency of herbivory on a plant pathogen, the order of herbivore arrival on plants strongly impacts plant-mediated effects on herbivore communities (Thaler et al. 2001,

Viswanathan et al. 2005). Similarly understanding the timing of interactions between non-vector herbivore and pathogens could aid in elucidating the impacts of these interactions.

Correlations between S. lineatus and PEMV might also be explained by preferences of S. lineatus for infected compared to healthy plants. Chewing herbivores have been observed to have increased performance on virus-infected plants in other systems (Hare and Dodds 1987, Musser et al. 2003, Thaler et al. 2010). Such benefits are often attributed to elevated levels of salicylic acid, and concurrent suppression of jasmonic acid, in virus-infected plants (Thaler et al. 2010,

Ali and Agrawal 2014). Our data similarly shows that plants upregulated production of jasmonic

43 acid in response to S. lineatus, indicating this pathway is important in anti-herbivore defense.

However, plants infected with PEMV had increased levels of salicylic acid and suppressed levels of jasmonic acid. Although our assays were too short to allow us to measure fitness metrics for S. lineatus, given these suppressed defenses and the preferences of S. lineatus, we may surmise that

S. lineatus benefitted from plants infected with PEMV. Consequently, viral outbreaks could contribute to increased abundance of S. lineatus through plant-mediated mechanisms.

Our study highlights the importance of studying reciprocal interactions between plant viruses and non-vector species, and suggests these interactions could have widespread implications for virus spread and evolution. If chewing herbivores allow viruses to attain higher titers within their hosts, and create feeding damage that benefits vector species, then natural selection may favor viruses that render plants attractive to chewers. Several authors have shown that viruses affect insect vectors, both through direct and plant-mediated effects, in ways that enhance viral spread (Fereres et al. 1989, Castle and Berger 1993, Belliure et al. 2005, Bosque-

Perez and Eigenbrode 2011, Mauck et al. 2010a). Our results suggest the intriguing possibility that such effects may extend to non-vector species as well. In turn, natural selection might favor non-vector herbivore species that provide a more conductive host environment for pathogens whenever plant infection provides reciprocal benefits for the herbivore.

Even in highly disturbed agroecosystems, plant pathosystems are embedded within complex ecological communities where interactions with non-vector species across multiple trophic levels are common. Our research shows that seemingly unrelated components of communities, such as a pathogen and a non-vector herbivore, can interact through strong plant- mediated effects. Continuing to move beyond simple studies of vector-virus dynamics will aid in

44 elucidating how such community-wide interactions affect disease ecology in natural and managed ecosystems

ACKNOWLEDGEMENTS

We thank Janelle Badger, Akaisha Charlton, Laura Rafferty, Isabel Brofsky, and Lee

Mendez for helping collect and process data; Sanford Eigenbrode, Ying Wu, and Brad Stokes for providing access to their sampling network involving growers Brian Busch, Greg Stout, Alan

Druffel, Bernie Schultheis, Jeff Anderson, and Jerry Mraz; and Steve Clement for providing essential background information on the system. This research was supported by a USDA-NIFA

Predoctoral Fellowship 2016-67011-24693.

LITERATURE CITED

Abe, H., Y. Tomitaka, T. Shimoda, S. Seo, T. Sakurai, S. Kugimiya, S. Tsuda, and M.

Kobayashi. 2012. Antagonistic plant defense system regulated by phytohormones assists

interactions among vector insects, thrips and a tospovirus. Plant Cell Physiol. 53: 204-212.

Agrawal, A. A. 1998. Induced responses to herbivory and increased plant performance. Science

279: 1201-1202.

Agrawal, A.A. 1999. Induced responses to herbivory in wild radish: effects on several herbivores

and plant fitness. Ecology 80: 1713–1723.

Ali, J. G. and A. A. Agrawal. 2014. Asymmetry of plant-mediated interactions between specialist

aphids and caterpillars on two milkweeds. Functional Ecology 28: 1404-1412.

Baiser, B., R. S. Ardeshiri, and A. M. Ellison. 2011. Species richness and trophic diversity

increase decomposition in a co-evolved food web. PLOS ONE 6: e20672.

Bardner, R., K. E. Fletcher and D. C. Griffiths. 1983. Chemical control of the pea and bean

weevil, Sitona lineatus, and subsequent effects on yield of field beans, Vicia faba L. The

Journal of Agricultural Science 101: 71-80.

Belliure, B., Janssen, A., Maris, P.C., Peters, D., and M.W. Sabelis. 2005. Herbivore arthropods

benefit from vectoring plant viruses. Ecology Letters 8: 70–79.

Belliure, B., M. W. Sabelis, and A. Janssen. 2010. Vector and virus induce plant responses that

benefit a non-vector herbivore. Basic and Applied Ecology 11: 162-169.

Berlow, E. L., A.-M. Neutel, J. E. Cohen, P. C. De Ruiter, B. Ebenman, M. Emmerson, J. W.

Fox, V. A. A. Jansen, J. I. Jones, G. D. Kokkoris, D. O Logofet, A. J. McKane, J. M.

Montoya, and O. Petchey. 2004. Interaction strengths in food webs: issues and opportunities.

Animal Ecology 73: 585-598.

Bosque-Pérez, N. A. and S. D. Eigenbrode. 2011. The influence of virus-induced changes in

plants on aphid vectors: insights from luteovirus pathosystems. Virus Res. 159: 201-205.

Bowles, D. 1991. The wound response of plants. Current Biology 1: 165-167.

Casteel CL, De Alwis M, Bak A, Dong H, Steven A, G. Jander. 2015. Disruption of ethylene

responses by Turnip mosaic virus mediates suppression of plant defense against the aphid

vector, Myzus persicae. Plant Physiology 169: 209-218.

Casteel, C. L., C. Yang, A. C. Nanduri, H. N. De Jong, S. A. Whitham, and G. Jander. 2014. The

Nia-Pro protein of Turnip mosaic virus improves growth and reproduction of the aphid

vector, Myzus persicae (green peach aphid). Plant J. 77: 653-663.

Castle, S. J. and P. H. Berger. 1993. Rates of growth and increase of Myzus persicae on virus-

infected potatoes according to type of virus-vector relationship. Entomologia Experimentalis

et Applicata 69: 51-60.

Clement, S. L. 2006. Pea aphid outbreaks and virus epidemics on peas in the US Pacific

Northwest: histories, mysteries and challenges. Online. Plant Health Progress.

doi:10.1094/PHP-2006-1018-01-RV

Clement, S.L., Husebye, D.S., and S.D. Eigenbrode. 2010. Ecological factors influencing pea

aphid outbreaks in the US Pacific Northwest. Pages 107-128 in P. Kindlmann, A.F.G. Dixon,

and J.P. Michaud, editors. Aphid Biodiversity under Environmental Change. Springer

Publishing, Berlin, Germany.

Fereres, A., R. M. Lister, J. E. Araya, and J. E. Foster. 1989. Development and reproduction of

the English grain aphid (Homoptera: Aphididae) on wheat cultivars infected with Barley

yellow dwarf virus. Environmental Entomology 18: 388-393.

Glozer, K. 2008. Protocol for leaf image analysis - surface area. University of California, Davis.

Hare, J. D. and J. A. Dodds. 1987. Survival of the Colorado potato beetle on virus-infected

tomato in relation to plant nitrogen and alkaloid content. Entomologia Experimentalis et

Applicata 44: 31-35.

Hogenhout, S. A., K. Oshima, E. D. Ammar, S. Kakizawa, H. N. Kingdom, and S. Namba. 2008.

Phytoplasmas: bacteria that manipulate plants and insects. Mol. Plant Pathol. 9: 403-423.

Johnson, P. T. J., De Roode, J. C. and A. Fenton. 2015. Why infectious disease biology needs

community ecology. Science 349: 1259504.

Karban, R. and I. T. Baldwin. 1997. Induced Responses to Herbivory. University of Chicago

Press, Chicago, Illinois, USA.

Kersch-Becker, M. F. and J. S. Thaler. 2013. Virus strains differentially induce plant

susceptibility to aphid vectors and chewing herbivores. Oecologia 174: 883-892.

Li, R., B. T. Weldegergis, J. Li, C. K. Jung, J. Qu, Y. W. Sun, H. M. Qian, C. S. Tee, J. J. A. van

Loon, M. Dicke, N. H. Chua, S. S. Liu, and J. Ye. 2014. Virulence factors of geminivirus

interact with MYC2 to subvert plant resistance and promote vector performance. Plant Cell

26: 4991-5008.

Luan J.B., D. M. Yao, T. Zhang, L. L. Walling, M. Yang, Y. J. Wang, and S. S. Liu. 2013.

Suppression of terpenoid synthesis in plants by a virus promotes mutualism with vectors.

Ecol Lett. 16: 390-398.

Mauck, K. E., C. M. D. Maraes, and M. C. Mescher. 2010a. Deceptive chemical signals induced

by a plant virus attract insect vectors to inferior hosts. PNAS 107: 3600-3605.

Mauck, K. E., Moraes, C. M. D., and M. C. Mescher. 2010b. Effects of Cucumber mosaic virus

infection on vector and non-vector herbivores of squash. Communicative and Integrative

Biology 3: 579–582.

Mauck, K., N. A. Bosque-Pérez, S, D, Eigenbrode, C. M. De Moraes, and M. C. Mescher, M.C.

2012. Transmission mechanisms shape pathogen effects on host–vector interactions:

evidence from plant viruses. Functional Ecology 26: 1162–1175.

McElhany, P., L. A. Real, and A. G. Power, A.G. 1995. Vector preference and disease dynamics:

A study of barley yellow dwarf virus. Ecology 76: 444–457.

Musser, R., S. M. Hum-Musser, G. W. Felton, and R. Gergerich. 2003. Increased larval growth

and preference for virus-infected leaves by the Mexican bean beetle, Eplilachna varivestis

Mulsant, a plant virus vector. Journal of Insect Behavior 16: 247-256

Pan, H. G. Chen, F. Li, Q. Wu, S. Wang, W. Xie, B. Liu, B. Xu and Y. Zhang. 2013. Tomato

spotted wilt virus infection reduces the fitness of a nonvector herbivore on pepper. Journal of

Economic Entomology 106: 924-928.

Polis, G. A. and D. R. Strong. 1996. Food web complexicity and community dynamics. The

American Naturalist 147: 813-846.

Power, A. G. 2000. Insect transmission of plant viruses: a constraint on virus variability. Current

Opinion in Plant Biology 3: 336-340.

Robert-Seilaniantz, A. M. Grant, and J. D. Jones. 2011. Hormone crosstalk in plant disease and

defense: more than just jasmonate-salicylate antagonism. Annual Review of Phytopathology

49: 317-343.

Roosein, B. K., R. Gomulkiewicz, L. L. Ingwell, N. A. Bosque-Pérez, D. Rajabaskar, and S.

D. Eigenbrode. 2013. Conditional vector preference aids the spread of plant pathogens:

results from a model. Environmental Entomology 42: 1299-1308.

Sadeghi, S. E., J. Bjur, L. Ingwell, L. Unger, N. A. Bosque-Pérez and S. D. Eigenbrode. 2016.

Interactions between Metopolophium festucae cerealium (Hemiptera: Aphididae) and Barley

yellow dwarf vius (BYDV-PAV). Journal of Insect Science 16: 21.

Seabloom, E. W. et al. 2015. Plant species’ origin predicts dominance and response to nutrient

enrichment and herbivores in global grasslands. Nat. Commun. 6: 7710.

Selakovic, S., P. C. de Ruitter, and H. Heesterbeek. 2014. Infectious disease agents mediate

interaction in food webs and ecosystems. Proc. Biol. Sci. 281: 20132709.

Strange, R. N. and P. R. Scott. 2005. Plant disease: a threat to global food security. Annu. Rev.

Phytopathol. 43: 83-116.

Thaler, J. S., A. A. Agrawal, and R. Halitschke. 2010. Salicylate-mediated interactions between

pathogens and herbivores. Ecology 91: 1075-1082.

Thaler, J. S., Humphrey, P. T. and Whiteman, N. K. 2012. Evolution of jasmonate and salicylate

signal crosstalk. Trends in Plant Science 17: 260–270.

Thaler, J. S., M. J. Stout, R. Karban, and S. S. Duffey. 2001. Jasmonate-mediated induced plant

resistance affects a community of herbivores. Ecological Entomology 26: 312-324.

Van Molken, T., H. de Caluwe, C. A. Hordijk, A. Leon-Reyes., T. A. L. Snoeren, N. M. van

Dam, and J. F. Stuefer. 2012. Virus infection decreases the attractiveness of white clover

plants for a non-vectoring herbivore. Oecologia 170: 433-444

Vega, F. E., P. F. Dowd, and R. J. Bartelt. 1995. Dissemination of microbial agents using and

autoinoculating device and several insect species as vectors. Biological Control 5: 545-552.

Viswanathan, D. V., A. J. T. Narwani, and J. S. Thaler. 2005. Specificity in induced plant

responses shapes patterns of herbivore occurrence on Solanum dulcamara. Ecology 86: 886-

896.

Zarate, S.I., L. A. Kempema, and L. L. Walling. 2007. Silverleaf whitefly induces salicylic acid

defenses and suppresses effectual jasmonic acid defenses. Plant Physiol. 143: 866–875.

Zhang, T., J.-B. Luan, J.-F. Qi, C. –J. Huang, M. Li, X.-P. Zhou, and S. –S. Liu. 2012.

Begomovirus–whitefly mutualism is achieved through repression of plant defenses by a virus

pathogenicity factor. Molecular Ecology 21: 1294–1304.

FIGURE LEGENDS

FIG. 2.1. (A) The relative defoliation by S. lineatus (number of feeding notches) on paired uninfected and PEMV-infected plants in commercial pea fields. (B) The relative defoliation

(total leaf area) by S. lineatus on paired uninfected and PEMV-infected plants in greenhouse choice assays. (C) The relative abundance of inoculative aphids on paired undamaged and S. lineatus-damaged plants in greenhouse choice assays. All points represent means ± SE. + - significant difference at α = 0.10; * - significant difference at α = 0.05.

FIG. 2.2. (A) The relative PEMV titer (absorbance/total protein) after 6d of S. lineatus feeding for pea plants that were exposed to either zero, one, or three S. lineatus after being inoculated with PEMV. All points represent means ± SE. Different letter represent significant differences at

α = 0.05. (B) The relative PEMV titer (absorbance/total protein) for pea plants that experienced varying defoliation from S. lineatus (total leaf area removed in cm2) after being inoculated with

PEMV; the dotted line shows the best-fit linear regression line.

Fig. 2.3. (A) Salicylic acid and (B) jasmonic acid levels in pea plants following sham-inoculation

(uninfected) or PEMV-inoculation (infected) treatments. (C) Salicylic acid and (D) jasmonic acid levels in pea plants that were never exposed to S. lineatus (undamaged) or exposed to S. lineatus feeding (damaged). All phytohormone levels shown were measured in nanograms per g of plant fresh weight. All points represent means ± SE. * - significant difference at α = 0.05

FIG. 2.1

FIG. 2.2.

FIG. 2.3.

CHAPTER THREE

NICHE PARTITIONING BETWEEN HERBIVORES PROMOTES THE SPREAD OF A

VECTOR-BORNE PLANT VIRUS

ABSTRACT

Herbivores that transmit plant pathogens often compete with non-vector species for shared plant resources. Such competition affects many traits of vectors including their fitness and behavior.

However, few studies have examined how competition might indirectly affect the spread of a plant pathogen. Here we conducted field and greenhouse experiments to show how a defoliating herbivore (Sitona lineatus) indirectly promoted the spread of a pea aphid-transmitted plant pathogen, pea enation mosaic virus (PEMV), through its effects on aphid behavior. In field mesocosms, we observed higher rates of viral spread when inoculative pea aphids shared plant resources with S. lineatus. This occurred despite S. lineatus appearing to have no effect on aphid fitness or plant-to-plant movement. Greenhouse experiments, however, showed that weevils and aphids partitioned the plant niche from bottom to top. Weevils tended to feed close to the soil surface, and when weevils were present aphids moved higher on plants. Importantly, plants were significantly more PEMV-susceptible when aphids fed higher on plants. Our results suggest that niche partitioning between S. lineatus and aphids on individual plants drives increased pathogen spread at larger scales. Given the ubiquity of competition between vectors and non-vector species, competition may shape pathogen dynamics in many natural and managed ecosystems.

KEYWORDS: niche partitioning, species interactions, competition, plant-herbivore interactions, disease ecology, vector-virus interactions

INTRODUCTION

Interspecific competition can shape the composition of herbivore communities on shared plant resources (Denno et al. 1995, Polis and Strong 1996, Reitz and Trumble 2002). While early studies suggested that resource competition in insect herbivore communities was generally weak

(Hairston et al. 1960, Shorrocks et al. 1984), recent reviews have shown that competition among arthropods regularly results in species exclusion (Reitz and Trumble 2002). Competition can also affect niche partitioning between herbivores, with weak competitors forced to avoid plant structures that are occupied by a dominant competitor (McEvoy 1986).

Plant-mediated competition between herbivores also occurs when feeding by one herbivore activates host plant defenses (Karban and Baldwin 1997) that impact the fitness (Ali and Agrawal 2014) and settlement behavior (Agrawal 1998, 1999) of other herbivores.

Herbivory also often alters the nutritional content of plants (Weibull 1987; de Mazencourt et al.

1998), potentially affecting the fitness of competing herbivores (Awmack and Leather 2002).

However, nutritionally-deficient plants may stimulate the production of mobile herbivore phenotypes (De Barro 1992) and stimulate herbivore movement as individuals search for more suitable hosts (Kareiva 1982). Induced variation in plant traits from competition among herbivores might in turn have strong impacts on mediating the structure of entire herbivore communities (Thaler et al. 2001, Visvanathan et al. 2005).

Many arthropod herbivores vector plant pathogens. Studies involving vector species show that competition can affect vector movement (Wratten et al. 1988, Senger et al. 2007,

Mazzi and Dorn 2012) and host selection behavior (Resetarits and Wilbur 1989, Rosenzweig

1991). Moreover, interacting species such as competitors can alter vector feeding behavior, such

57 as feeding duration (Prado and Tjallingii 1994, Fereres and Moreno 2009, Long and Finke 2015).

These factors might affect dynamics of pathogens, in addition to potential effects of competition on vector abundance (Muller and Godfray 1997, Pascual and Callejas 2004, Ali and Agrawal

2014). However, while several studies have shown how competition affects fitness and behavioral traits of species that function as vectors, few studies have examined resulting impacts on a pathogen.

Here we conducted a series of field and greenhouse experiments to test the indirect effects of competition between herbivores on an aphid-transmitted plant virus. Our model system included the defoliating non-vector herbivore, Sitona lineatus, and a vector, the pea aphid

Acyrthosiphon pisum. We hypothesized that herbivory by S. lineatus might impact the spread of a pathogen vectored by A. pisum by influencing fitness, plant-to-plant movement, or within-plant movement. Our results indicate that niche partitioning between S. lineatus and A. pisum increased the risk of viral infection for individual pea plants. In turn, competition between S. lineatus and A. pisum on individual plants altered rates of pathogen spread at broader scales.

METHODS

Study system

The Palouse region of eastern Washington and northern Idaho supports a diverse community of native and cultivated legumes including vetch, clover, pea, and lentil (Black et al,

). Two herbivore species that commonly co-occur on these hosts include the pea leaf weevil (S. lineatus), a defoliating insect, and the pea aphid (Acyrthosiphon pisum), a piercing-sucking insect (Clement 2006). Pea aphids vector a plant pathogen, pea enation mosaic virus (PEMV) to multiple different legume species (Cockbain and Gibbs, 1973). Given the common co-

58 occurrence of these organisms, we hypothesized that competition between S. lineatus and A. pisum might strongly impact the dynamics of PEMV transmission.

Pea aphids used in the experiments were obtained from greenhouse colonies (Appendix

S3). Adult S. lineatus used in experiments were collected from alfalfa plants growing in eastern

Washington and kept in an incubator at 4⁰ C until use. Peas were grown in a greenhouse

(Appendix S3) on the Washington State University (WSU) campus in 3.5” square pots.

Effects of S. lineatus on A. pisum and PEMV

We conducted a field experiment in 2015 to determine (1) whether S. lineatus and PEMV affected the behavior and fitness of A. pisum and (2) whether S. lineatus affected the spread of

PEMV. The experiment was conducted in 60cm3 mesh cages at the WSU Tukey Horticultural

Orchard; cages were arranged in a grid with 1m spacing (Appendix S1: Fig. S7A). The experimental design was a 2×2 factorial, with aphid infection status (inoculative or not) crossed with weevils (present or not); 6 replicates were conducted for each treatment. Treatments were applied in single-plant cages. The aphid and weevil treatments consisted of releasing 25 4d-old aphids (inoculative or not) and adult S. lineatus (2 or none), respectively, on single plants

(Appendix S1: Fig. S7B). Twenty-four hours later, the number of successfully established aphids was recorded and plants were moved, along with the insects, to the middle of an 8-plant ring enclosed by the mesh cage in the field (Appendix S3: Fig. S8).

Six days later, the number and stage (nymph or adult) of aphids infesting each of the 9 plants from each cage was recorded. Since pea aphids reach full size and reproductive maturity at

7-10d (Dixon 1974), the released aphids were clearly distinguishable from aphids that were born during the experiment. Terminal leaflets were clipped from each plant and tested with a

59 commercial double antibody sandwich (DAS) ELISA (AC Diagnostics, Fayetteville, AR) for the presence of PEMV using the provided protocol (Appendix S3). Samples were run in duplicate, and plants were diagnosed as infected if their absorbance value was greater than or equal to 2× the value of the negative control (Vemulapati et al. 2014).

Niche partitioning between S. lineatus and A. pisum

To determine if competition between S. lineatus and A. pisum influenced within-plant aphid distribution, we conducted a greenhouse (Appendix S3) experiment where individually caged 2wk-old pea plants were exposed to one of two treatments: (1) S. lineatus damage – two adults released on the plants or (2) undamaged – no adults were released; eight replicates were conducted for each treatment. After 2d of feeding, we released 15 4d-old A. pisum onto the plants. After 24 h, we recorded the location of each aphid on each plant according to structure

(stem, leaf, or tendril) and height (node). The number of S. lineatus feeding notches on each node’s leaf pair was also recorded for plants containing adult weevils.

Feeding location and inoculation success

To determine if the feeding location of A. pisum influenced inoculation efficiency of

PEMV, we conducted a 3×5 factorial greenhouse (Appendix S3) experiment manipulating A. pisum feeding location (top, middle, or bottom of plants) and density (2, 4, 6, 8, or 10) 6d-old inoculative aphids. Three replicates were conducted for each treatment. Cages at the top contained the top 2 vegetative nodes, and were constructed from a bag of lightweight mesh attached to a support. Bottom cages contained the lowest 2 vegetative nodes and were constructed from inverted plastic medicine cups with a hole to accommodate the plant stem; a

60 slit was cut in the side of the cup to allow it to be slid over the plant. Following the addition of aphids, the slit was taped and cotton balls were packed in the hole to ensure containment. Top and bottom cages were applied to each plant regardless of treatment to control for any cage effects on plant physiology (Appendix S1: Fig. S9). Aphids applied to the middle were not caged, but were restricted from feeding at the top and bottom by cages.

After an inoculation access period of 2d, aphids were killed with an application of 60 ml

0.07% imidacloprid to the soil of each plant; cages were subsequently removed. After aphids were killed, we waited 6d to allow viral titer to build to detectable levels, after which the aboveground portions of the plant was harvested and analyzed with ELISA (Appendix S3) to determine infection status and intensity. To compare relative viral titers, we calculated the relative reactivity of the sample normalized against the sample’s protein content as follows:

Sample absorbance ⁄ Negative control absorbance

Sample protein content

Data analysis

From the field experiment we tested effects of S. lineatus and PEMV on A. pisum movement, survival, and reproduction. Movement (number of A. pisum located off of the original host) and survival (number of established A. pisum adults that survived experiment) were analyzed using logistic regression. Aphids born during the experiment were clearly distinguishable due to their small size, and were not included in the movement analysis.

Reproduction (nymphs produced per adult) was analyzed with ANOVA. All models included weevil presence and PEMV-infection as main effects; interactions between these variables were never significant (α = 0.10) and were removed from final models. We also tested whether weevil presence influenced the infection status of plants using logistic regression. To account for

61 variation in the abundance of established A. pisum adults between cages, the number of established aphids at the start of the experiment was included as a covariate.

Analyses of the niche partitioning experiment tested whether weevil feeding affected the within-plant location of aphid feeding. We first analyzed weevil feeding patterns by testing the number of feeding notches as a function of leaf height (vegetative node) using a logistic model.

A positive coefficient for “node” would indicate a preference for higher leaves, while a negative coefficient would indicate a preference for lower leaves. Data from node 1 was excluded since leaves were too small to reliably measure herbivory. We next analyzed aphid preferences for (1) plant structures and (2) node height using logistic regression, where each individual aphid on each structure or node served as binary count data. We determined if S. lineatus feeding affected

A. pisum by using logistic regression to analyze aphid occurrence on plant nodes as a function of node height and the number of S. lineatus feeding notches occurring at that node. A positive coefficient for “notches” would indicate that aphids seek heavily notched leaves, while a negative coefficient would indicate aphids avoided heavily notched leaves. Moreover, we examined whether aphid preferences for particular locations on the plant (height or structure) was affected by weevils by adding a weevil*height or weevil*structure interaction term to the logistic regression. Finally, we used logistic regression to determine if aphids were more likely to settle in the top 4 nodes (roughly the top half of the plant) when S. lineatus were present compared to absent.

For the inoculation success experiment, we used logistic regression to model plant infection status as a function of aphid location (top, middle, bottom) and abundance. We used multiple regression to model the relative titer of each successfully inoculated plant as a function of aphid number and aphid location. Since there was no significant interaction (abundance ×

62 location) effect in either analysis (α = 0.10), final models did not include an interaction term. All statistical analyses were conducted in R Studio (v. 0.99.903). Parameter significance was based on the Z distribution for all generalized models.

RESULTS

Effect of S. lineatus on A. pisum and PEMV

Infection rates in individual field cages ranged from 11 to 66%. Mesocosms containing weevils exhibiting significantly higher rates of viral infection (Z = 2.42, P = 0.016, Fig. 3.1A).

Aphids had greater survival rates when PEMV was present in the system (Z = 2.07, P = 0.038;

Appendix S1: Fig. S10). However, S. lineatus did not affect aphid reproduction (F1,21 = 2.28, P =

0.15, Fig. 3.1B), survival (Z = -0.45, P = 0.65, Fig. 3.1C), or plant-to-plant movement (Z = 1.18,

P = 0.13, Fig. 3.1D).

Niche partitioning between S. lineatus and A. pisum

Adult S. lineatus showed decreased feeding rates (notch counts) as node height increased, indicating they preferred to feed close to the soil (Fig. 3.2A, Z = -4.79, P < 0.0001). Aphids also fed lower on plants (Z = -7.17, P < 0.0001) and preferred stems over tendrils (Z = -2.51, P =

0.0122, Appendix S1: Fig. S11,) and leaves (Z = -2.29, P = 0.022). However, while S. lineatus did not affect aphid preferences for particular plant structures (leaves: Z = 0.160, P = 0.873, tendrils: Z = 0.813, P = 0.416, Appendix S1: Fig. S11), aphid occurrence was negatively influenced by the number of S. lineatus feeding notches (Z = -3.542, P = 0.0004, Fig. 3.2B). The preference of aphids for lower portions of the plant was significantly influenced by weevils (Z =

1.871, P = 0.060), and aphids were more evenly distributed across the plant when weevils were

63 present (Fig. 3.2C). Consequently, aphids were more likely to be on the top half of plants when

S. lineatus was present (Z = 1.96, P = 0.050, Fig. 3.2D).

Feeding location and inoculation success

Aphid feeding location affected PEMV transmission, as inoculation success significantly increased as aphids fed higher on the plant (Z = 2.679, P = 0.0074, Fig. 3), regardless of the abundance of inoculating aphids (Z = 0.86, P = 0.39). When aphids were restricted to the top 2 nodes of the plant, inoculation success was 100%, compared to 53% when aphids were restricted to the bottom 2. However, aphid feeding location (t32 = 0.63, P = 0.53) and abundance (t32 =

0.650, P = 0.52) did not influence the severity (PEMV titer) of successful infections.

DISCUSSION

Our study shows that niche partitioning between a vector and non-vector herbivore can strongly affect a vector-borne pathogen. While niche partitioning in consumer communities has often been shown to influence rates of resource consumption (Schmitz 2008, Finke and Snyder

2010, Northfield et al. 2010), our study demonstrates it can also mediate pathogen transmission by altering vector feeding behavior. Ecological studies have similarly shown that host plant diversity (Power 1991) and predators (Long and Finke 2015) can affect vector feeding behavior, with implications for pathogen spread. Our study highlights the importance of species interactions in creating cascading effects for vectors and pathogens they transmit.

Our field mesocosm study showed that rates of disease spread increased when S. lineatus was present with aphids compared to when it was absent. We hypothesized that these patterns might be explained by effects of S. lineatus on aphid fitness or plant-to-plant movement, as these

64 factors are important factors shaping disease dynamics (Power 1991, McElheny et al. 1995,

Perring et al. 1999, McCallum et al. 2001, Finke 2012, Mauck et al. 2012). Surprisingly, however, none of these metrics were influenced by S. lineatus. As S. lineatus marginally impacted aphid population dynamics, we next explored whether S. lineatus indirectly affected the susceptibility of plants to PEMV.

We observed that aphid avoided feeding sites of S. lineatus, which resulted in aphids moving higher on plants when S. lineatus was present. Competition with S. lineatus therefore caused a shift in the habitat domain (Schmitz 2008) of aphids. In turn, the ability of aphids to inoculate plants with PEMV increased for aphids located higher on plants. Pathogen tolerance in plants generally increases with age (Leisner et al. 1993, Kus et al. 2002, Panter and Jones 2002,

Develey-Riviere and Galiana 2007). On individual plants, older leaves are also often more tolerant to pathogens than immature leaves (Roumen et al. 1992, Abedon and Tracy 1996). Our observation that viral inoculation success increased with leaf height is consistent with these findings.

While our results show that competition with S. lineatus affected the distribution of aphids on plants, other factors that affect vector micro-habitat use might also affect disease spread. For example, increasing temperatures often forces herbivores and other species to move lower on plants to avoid heat (Bale et al. 2002, Barton et al. 2009). The vertical distribution of vectors such as aphids can also vary based on population age structure, plant physiology, and time of the year (Jansson and Smilowitz 1986, Bale et al. 2002, McCornack et al. 2008).

Moreover, while younger plant tissue often has higher levels of defensive compounds such as glucosinolates (Wallace and Eigenbrode 2002), iridoid glycosides (Bowers and Stamp 1993), and protease inhibitors (Van Dam et al. 2001), herbivores often show a preference for younger

65 tissue (Fenner et al. 2001). Our study highlights that factors such as these that affect vector feeding behavior could have strong implications for pathogen spread.

Although many studies have examined effects of predators on vector-borne plant pathogens (Roitberg and Myers 1978, Smyrnioudis et al. 2001, Hodge et al. 2011, Long and

Finke 2015), ours is among the first to document effects of a competing non-vector herbivore on pathogen dynamics. In Chapter 2, we showed that S. lineatus influences virus-host interactions through defoliation, which increases viral titer in infected plants (Chisholm and Crowder, in review). Here, we demonstrate that weevils can also indirectly facilitate the host-to-host movement of the virus by promoting inoculation success on individual plants. In agriculture, pest management is typically based on direct damage inflicted, without considering interactions with other pests or diseases. While the damage caused by S. lineatus is typically minor (Carcamo and

Vankosky 2011), yield losses due to PEMV are a serious problem for pulse producers (Clement

2006). Given the observation that S. lineatus might promote viral spread, our results suggest this species should be given greater consideration for its indirect impacts as a pest.

Our study highlights the role of interspecific competition in mediating the spread of a vector-borne plant pathogen. Competition is ubiquitous in natural and managed ecosystems, and complex herbivore assemblages often compete for the same plant resources (Denno et al. 1995,

Reitz and Trumble 2002). For example, among the most important agricultural crops in the world, non-vector defoliators commonly co-occur with vectors and viral pathogens they transmit

(Appendix S2: Table S2). Moreover, competition can also affect the ranges of species at larger spatial scales (Connell 1961, Schoener 1983). Understanding the interplay between competing vector and non-vector species is thus critical for understanding basic disease ecology and for developing effective disease mitigation strategies.

ACKNOWLEDGEMENTS

This research was supported by a USDA-NIFA Predoctoral Fellowship 2016-67011-

24693.

LITERATURE CITED

Abedon, B. G. and W. F. Tracy. 1996. Corngrass 1 of maize (Zea mays L.) delays development

of adult plant resistance to common rust (Puccinia sorghi Schw.) and European corn borer

(Ostrinia nubilalis Hubner). Journal of Heredity 87: 219-223.

Agrawal, A. A. 1998. Induced responses to herbivory and increased plant performance. Science

279: 1201-1202.

Agrawal, A. A. 1999. Induced responses to herbivory in wild radish: effects on several

herbivores and plant fitness. Ecology 80L 1713-1723.

Ali, J. G. and A. A. Agrawal. 2014. Asymmetry of plant-mediated interactions between specialist

aphids and caterpillars on two milkweeds. Functional Ecology 28: 1404-1412.

Anderson, P. K., A. A. Cunningham, N. G. Patel, F. J. Morales, P. R. Epstein, and P. Daszak.

2004. Emerging infectious diseases of plants: pathogen pollution, climate change, and

agrotechnology drivers. Trends in Ecology and Evolution 19: 535-544.

Awmack, C.S. and S. R. Leather. 2002. Host Plant Quality and Fecundity in Herbivorous Insects.

Annual Review of Entomology 47: 817–844.

Bailey, S. M., M. E. Irwin, G. E. Kampmeier, C. E. Eastman, and A. D. Hewings. 1995. Physical

and biological perturbations: their effect on the movement of apterous Rhopalosiphum padi

(Homoptera: Aphididae) and localized spread of barley yellow dwarf virus. Environ.

Entomology 24: 24-33.

Bale, J.S., Masters, G.J., Hodkinson, I.D., Awmack, C., Bezemer, T.M., Brown, V.K.,

Butterfield, J., Buse, A., Coulson, J.C., Farrar, J., et al. 2002. Herbivory in global climate

change research: direct effects of rising temperature on insect herbivores. Global Change

Biology 8: 1–16.

Barton, B. T., Beckerman, A. P., and O. J. Schmitz. 2009. Climate warming strengthens indirect

interactions in an old-field food web. Ecology 90: 2346–2351.

Black, A. E., E. Strand, P. Morgan, J. M. Scott, R. G. Wright, and C. Watson. 1998. Biodiversity

and landuse history of the Palouse bioregion: pre-European to present. Chapter 10 in T. D.

Sisk, editor. Perspectives on land-use survey. Biological Resources Division, Biological

Science Report USGS/BRD/BSR.

Bowers, M. D. and N. E. Stamp. 1993. Effects of plant-age, genotype, and herbivory on

Plantago performance in chemistry. Ecology 74: 1778-1791.

Burdon, J. J. and G. A. Chilvers. 1982. Host density as a factor in plant disease ecology. Annual

Rev. Phytopathol. 20: 143-166.

Canto, T., M. A. Aranda, and A. Fereres. 2009. Climate change effects on physiology and

population processes of host and vectors that influence the spread of hemipteran-borne plant

viruses. Global Change Biology 15: 1884-1894.

Carcamo, H. and M. Vankosky. 2011. Managing the pea leaf weevil in field peas. Prairie Soils

and Crops 4: 77-85.

Cardubale, B. J., J. P. Wright, M. W. Cadotte, I. T. Carroll, A. Hector, D. S. Srivastava, M.

Loreau, and J. J. Weis. 2007. Impacts of plant diversity on biomass production increase

through time because of species complementarity. PNAS 104: 18123-18128.

Clement, S. L. 2006. Pea aphid outbreaks and virus epidemics on peas in the US Pacific

Northwest: histories, mysteries and challenges. Online. Plant Health Progress.

doi:10.1094/PHP-2006-1018-01-RV

Cockbain, A. J. and A. J. Gibbs. 1973. Host range and overwintering sources of bean leaf roll

and pea enation mosaic viruses in England. Annals of Applied Biology 73: 177-187.

Connell, J. H. 1961. The influence of interspecific competition and other factors on distribution

of the barnacle chthamalus stellatus. Ecology 42: 710-723.

De Barro, P. J. 1992. The role of temperature, photoperiod, crowding and plant quality on the

production of alate viviparous females of the bird cherry-oat aphid, Rhopalosiphum padi.

Entomologia Experimentalis et Applicata 65: 205-214. de Mazancourt, C., M. Loreau, and L. Abbadie. 1998. Grazing optimization and nutrient cycling:

When do herbivores enhance plant production? Ecology 79: 2242–2252.

Denno, R. F., M. S. McClure, and J. R. Ott. 1995. Interspecific interactions in phytophagous

insects: competition revisited and resurrected. Annual Review of Entomology 40:297–331

Develey-Riviere, M. P. and E. Galiana. 2007. Resistance to pathogens and host developmental

stage: a multifaceted relationship within the plant kingdom. New Phytologist 175: 405-416.

Dixon, A. F. G. 1974. Biology of aphids. Physiological Entomology 48: 156.

Fenner, M., M. E. Hanley, and R. Lawrence. 2001. Comparison of seedling and adult palatability

in annual and perennial plants. Functional Ecology 13: 546-551.

Fereres, A. and A. Moreno. 2009. Behavioural aspects influencing plant virus transmission by

homopteran insects. Virus Research 141: 158-168.

Finke, D. L. 2012. Contrasting the consumptive and non-consumptive cascading effects of

natural enemies on vector-borne pathogens. Entomol. Exp. Appl. 144: 45-55.

Finke, D. L., and W. E. Snyder. 2010. Conserving the benefits of predator diversity. Biol.

Conserv. 143: 2260-2269.

Hairston, N. G., F. E. Smith, and L. B. Slododkin. 1960. Community structure, population

control and competition. The American Naturalist 94: 421-425.

Hodge, S. J. Hardie, and G. Powell. 2011. Parasitoids aid dispersal of a nonpersistenly

transmitted plant virus by disturbing the aphid vector. Agricultural and Forest Entomology

13: 83-88.

Hodgson, C. 1991. Dispersal of apterous aphids (Homoptera: Aphididae) from their host plant

and its significance. Bulletin of Entomological Research 81: 417-427.

Irwin M. E., G. E. Kampmeier and W. W. Weisser. 2007. Aphids as crop pests. Oxford

University Press, United Kingdom.

Jansson, R.K., and Z. Smilowitz. 1986. Influence of nitrogen on population parameters of potato

insects: abundance, population growth, and within-plant distribution of the green peach

aphid, Myzus persicae (Homoptera: Aphididae). Environ Entomol 15: 49–55.

Karban, R. and I. T. Baldwin. 1997. Induced Responses to Herbivory. University of Chicago

Press, Chicago, Illinois, USA.

Kareiva, P. 1982. Experimental and mathematical analyses of herbivore movement: quantifying

the influence of plant spacing and quality on foraging discrimination. Ecological

Monographs 52: 261–282.

Kennedy, G. G. 1976. Host plant resistance and the spread of plant viruses. Environ Entomol 5:

827–832.

Kus, J. V., K. Zaton, R. Sarkar, and R. K Cameron. 2002. Age-related resistance in Arabidopsis

is a developmentally regulated defense response to Pseudomonas syringae. The Plant Cell

14: 479-490.

Leisner, S., R. Turgeon, and S. H. Howell. 1993. Effect of host plant development and genetic

determinants on the long-distance movement of cauliflower mosaic virus in Arabidopsis. The

Plant Cell 5: 191-202.

Long, E. Y. and D. L Finke. 2015. Predators indirectly reduce the prevalence of an insect-

vectored plant pathogen independent of predator diversity. Oecologia 177: 1067-1074.

Mazzi, D. and S. Dorn. 2012. Movement of insect pests in agricultural landscapes. Annals of

Applied Biology 160: 97-113.

McCallum, H., N. Barlow, and J. Hone. 2001. How should pathogen transmission be modelled?

Trends in Ecology & Evolution 16: 295-300.

McCornack, B.P., Costamagna, A.C., and D. W. Ragsdale. 2008. Within-plant distribution of

soybean aphid (Hemiptera: Aphididae) and development of node-based sample units for

estimating whole-plant densities in soybean. Journal of Economic Entomology 101: 1488–

1500.

McElhany, P., L. A. Real, and A. G. Power, A.G. 1995. Vector preference and disease dynamics:

A study of barley yellow dwarf virus. Ecology 76: 444–457.

McEvoy, P. B. 1986. Niche partitioning in spittlebugs (Homoptera: Cercopidae) sharing shelters

on host plants. Ecology 67: 465-478.

Muller, R. and C. B. Godfray. 1997. Apparent competition between two aphid species. Journal

of Animal Ecology 66: 57-64.

Nault, L. R. and G. G. Gyrisco. 1966. Relation of the feeding process of the pea pahid to the

inoculation of pea enation mosaic virus. Annals of the Entomological Society of America 59:

1185-1197.

Northfield, T. D., G. B. Snyder, A. R. Ives and W. E. Snyder. 2010. Niche saturation reveals

resource partitioning among consumers. Ecology Letters 13: 338-348.

Panter, S. N. and D. A. Jones. 2002. Age-related resistance to plant pathogens. Advances in

Botanical Research 38: 251-280.

Pascual, S. and C. Callejas. 2004. Intra- and interspecific competition between biotypes B and Q

of Bemisia tabaci (Hemiptera: Aleyrodidae) from Spain. Bulletin of Entomological Research

94: 369-375

Perring, T. M., N. M. Gruenhagen, and C. A. Farrar. 1999. Management of plant viral diseases

through chemical control of insect vectors. Annual Review of Entomology 44: 457-481.

Polis, G. A. and D. R. Strong. 1996. Food web complexicity and community dynamics. The

American Naturalist 147: 813-846.

Power, A. G. 1991. Virus spread and vector dynamics in genetically diverse plant populations.

Ecology 72: 232-241.

Prado, E. and W. F. Tjallingii. 1994. Aphid activities during sieve element punctures.

Entomologia Experimentalis et Applicata 72: 157-165.

Reitz, S. R. and J. T. Trumble. 2002. Competitive displacement among insects and arachnids.

Annual Review of Entomology 47: 435-465.

Resetarits, Jr., W. and H. M. Wilbur. 1989 Choice of oviposition site by Hyla chrysoscelis: role

of predators and competitors. Ecology 70: 220-228.

Roitberg, B. D. and J. H. Myers. 1978. Effect of adult Coccinellidae on the spread of a plant

virus by an aphid. Journal of Applied Ecology 15: 775-779.

Rosenzweig, M. L. 1991. Habitat selection and population interactions: the search for

mechanism. Am. Nat. 137: S5-S28.

Roument, E. C., J. M. Bonman, and J. E. Parlevliet. 1992. Leaf age related partial resistance to

Pyricularia oryzae in tropical lowland rice cultivars as measured by number of sporulating

lesions. Phytopathology 82: 1414-1417.

Schmitz, O. J. 2008. Effects of predator hunting mode on grassland ecosystem function. Science

319: 952-954.

Schmitz, O. J. 2008. Effects of predator hunting mode on grassland ecosystem function. Science

319: 952-954.

Schoener, T. W. 1983. Field experiments on interspecific competition. The American Naturalist

122: 240-285.

Senger, S.E., Roitberg, B.D., and H. M. A. Thistlewood. 2007. Relative flight responses of

Rhagoletis indifferens as influenced by crowding, sex, and resources. Entomologia

Experimentalis et Applicata 123: 91–100.

Shorrocks, B., J. Rosewell, K. Edwards, and W. Atkinson. 1984. Interspecific competition is not

a major organizing force in many insect communities. Nature 310: 310-312.

Smyrnioudis, I., R. Harrington, S. J. Clark, and S. J. Katis. 2001. The effect of natural enemies

on the spread of barley yellow dwarf virus (BYDV) by Rhopalosiphum padi (Hemiptera:

Aphididae). Bulletin of Entomological Research 91: 301-306.

Strong, D. R., J. H. Lawton, and T. R. E. Southwood. 1984. Insects on Plants. Harvard

University Press, Cambridge, Massachusetts, USA.

Thaler, J. S., M. J. Stout, R. Karban, and S. S. Duffey. 2001. Jasmonate-mediated induced plant

resistance affects a community of herbivores. Ecological Entomology 26: 312-324.

Van Dam, N. M., M. Horn, M. Mares, and I. T. Baldwin. 2001. Ontogeny constrains systemic

protease inhibitor response in Nicotiana attenuata. Journal of Chemical Ecology 27: 547-

568.

Vemulapati, B. M., K. L. Druffel, D. Husebye, S. D. Eigenbrode, and H. R. Pappu. 2014.

Development and application of ELISA assays for the detection of two members of the

family Luteoviridae infecting legumes: Pea enation mosaic virus (genus Enamorivus) and

Bean leafroll virus (genus Luteovirus). Annals of Applied Biology 165: 130-136.

Viswanathan, D. V., A. J. T. Narwani, and J. S. Thaler. 2005. Specificity in induced plant

responses shapes patterns of herbivore occurrence on Solanum dulcamara. Ecology 86: 886-

896.

Wallace, S. K. and S. D. Eigenbrode. 2002. Changes in the glucosinolate-myrosinase defense

system in Brassica juncea cotyledons during seedling development. Journal of Chemical

Ecology 28: 243-256.

Weibull, J. 1987. Seasonal changes in the free amino acids of oat and barley phloem sap in

relation to plant growth stage and growth of Rhopalosiphum padi. Annals of Applied Biology

111: 729–737.

Wratten, S. D., P. J. Edwards, and L. Winder. 1988. Insect herbivory in relation to dynamic

changes in host plant quality. Bioogical Journal of the Linnean Society 35: 339-350.

FIGURE LEGENDS

FIG. 3.1. Effects of feeding by S. lineatus on (A) incidence of PEMV in individual field enclosures, (B) reproduction (nymphs per establish adult) of A. pisum vectors, (C) survival of established A. pisum, and (D) movement of A. pisum (number of aphids off-center release plant).

* - indicates a significant difference at α = 0.10.

FIG. 3.2. (A) Distribution of S. lineatus feeding notches on plants and (B) distribution of A. pisum in relation to S. lineatus feeding notches. (C) As A. pisum avoids S. lineatus feeding sites,

A. pisum adults skew more towards higher vegetative nodes when S. lineatus is present compared to absent, resulting in (D) more A. pisum on the upper-half of plants.

FIG. 3.3. Influence of aphid feeding location on the probability of successful inoculation of pea plants with PEMV. Values were pooled across treatments with variation in A. pisum density.

FIG. 3.1

FIG. 3.2

FIG. 3.3

0.9

0.8 0.7 0.6 0.5 0.4 0.3

0.2 Proportion plants infected plants Proportion 0.1 0 Bottom Middle Top Aphid location

CHAPTER FOUR

SOIL RHIZOBIA PROVIDE DIRECT AND INDIRECT BENEFITS TO PLANT YIELDS

ABSTRACT

Soil microbes can influence aboveground interactions between plants, herbivores, and pathogens.

Nitrogen-fixing root-colonizing bacteria such as rhizobia, which are mutualists with legumes, are highly abundant soil microbes. In addition to provisioning nitrogen, rhizobia might also affect plant tolerance to herbivores and pathogens, although this has rarely been tested. To explore how rhizobia influenced a vector-borne pathosystem, we conducted a mesocosm experiment whereby we excluded certain components of the soil microbial community and measured the responses of aphid vectors, viral pathogens, and pea plants. Our results were then analyzed using structural equation models to determine how soil variables influenced components of the pathosystem. Our results show that rhizobia indirectly increased plant yields by providing increased tolerance to aphids. In contrast, soil sterilization had both direct and aphid-mediated negative impacts on yield. The benefits of rhizobia did not appear to be simply due to nitrogen provisioning, as the addition of nitrogen fertilizer alone did not affect vectors, pathogens, or plants. Our results indicate that soil microbes exert considerable impact on aboveground pathosystems, and that these impacts are often mediated through reductions in pests and disease.

KEYWORDS: disease ecology, soil ecology, Rhizobia leguminosarum, symbiosis, plant- herbivore interactions, agriculture, pea enation mosaic virus

INTRODUCTION

Soil communities contain diverse networks of microbes (Roesch et al. 2007, Smith and

Reed 2008, Buee et al. 2009), arthropods, annelids, nematodes, and other invertebrates (Bardgett

2005). These communities can mediate aboveground community assembly (Marcel et al. 1998) and ecosystem processes (Wall et al. 2010; A’Bear et al. 2014). Soil microbes can affect plant tolerance to viral (Elbadry et al. 2006), bacterial (Huang et al. 2007), and fungal (Rabie 1998, de

Jensen et al. 2002) pathogens, herbivores (Raps and Vidal 1998, Jallow et al. 2004, Jaber and

Vidal 2009, Zehnder et al. 1997, Oosten et al. 2008, Valenzuela-Soto et al. 2010, Santos et al.

2014), and nematodes (Reitz et al. 2000, Hallmann et al. 2001). Many soil mutualists improve plant vigor even in the absence of pests (Avis et al. 2008, Azcon-Aguilar and Barea 1997), and may reduce impacts of abiotic stressors such as drought (Bae et al. 2009), nutrient deficiency

(Yang et al. 2009), salt (Evelin et al. 2009), and heavy metals (Weyens et al. 2009). However, benefits of mutualistic microbes for plant health and nutrition can also promote greater herbivore abundance (Pineda et al. 2010, Gehring and Bennett 2009, Shadler and Ballhorn 2016).

Rhizobial bacteria, which colonize roots and fix atmospheric nitrogen into mineralizable forms accessible to plants, are ecologically and economically important soil microbes. Plants respond to rhizobial infection by developing root nodules, specialized belowground structures that harbor rhizobia colonies. Rhizobial bacteria fix 40 to 60 million metric tons of nitrogen each year in agricultural systems, and an additional 3 to 5 million tons in natural ecosystems (Smil

1999). Although rhizobia exist naturally, many crop producers use commercial inoculants to ensure sufficient root colonization (Graham and Vance 2003). In addition to promoting plant health, nitrogen supplied by rhizobia may contribute to the synthesis of secondary metabolites

(Johnson et al. 1987, Ballhorn et al. 2013, Irmer et al. 2015, Dakora et al. 1993) with activity

81 against pathogens (Reitz et al. 2000, Hallmann et al. 2001, Omar and Abd-Alla 1998, Mabrouk et al 2007, Estevez de Jensen et al. 2002, Elbadry et al. 2006) and insect herbivores (Johnson et al. 1987, Thamer et al. 2011, Brunner et al. 2015). As with other mutualists, however, the superior nutritional content of plants fertilized by rhizobia can increase pest fitness (Wilson and

Stinner 1984, Kempel et al 2009, Whitaker et al. 2014).

While the nitrogen-fixing capacity of rhizobia benefits plants, rhizobia provide benefits to plants beyond what can be achieved through synthetic fertilization. Plants obtaining nitrogen from rhizobia, when compared to synthetic fertilizers, sometimes exhibit reduced pest abundance

(Dean et al. 2009, but see Dean et al. 2014). Additionally, higher levels of nodulation have been correlated with lower pest abundances (Brunner et al. 2015). Rhizobia produce anti-microbial compounds (Breil et al. 1996, Arora et al. 2001) that allow them to outcompete soil-borne plant pathogens (Avis et al. 2008), reducing the incidence of root diseases. Chemical by-products of rhizobia metabolism may also induce resistance to pests and pathogens in legumes (Reitz et al.

2000, 2002). Positive effects of rhizobial inoculation have even been observed on non-legume plants that do not form nodules (Antoun et al. 1998). Finally, provisioning plants with nitrogen via the rhizobial symbiosis avoids many negative consequences of synthetic fertilizers, such as ground water contamination and eutrophication of aquatic ecosystems (Matson et al. 1997).

Although several studies have examined how patterns of plant resistance shift in response to rhizobia, most have been conducted in controlled environments over short periods of time, or were confounded naturally-existing rhizobia. While such studies show rhizobia can influence plant performance via nitrogen provisioning, most have ignored secondary mechanisms by which rhizobia might impact plants, such as increased plant resistance to pest or pathogens. Quantifying the direct and indirect contributions of rhizobia to yield would improve our understanding of the

82 mechanisms underlying legume-rhizobia symbiosis. To address this gap, we conducted a large- scale field experiment to examine the impact of rhizobia exclusion and nitrogen fertilization on aphid pests, an aphid-borne viral pathogen, and resulting legume plant yields.

METHODS

Study system

Our experiment was conducted in a system consisting of pea (Pisum sativum) plants, pea aphid (Acyrthosiphon pisum) herbivores, and pea enation mosaic virus (PEMV), a devastating pathogen vectored by the pea aphid (Clement et al. 2010). These organisms commonly co-occur in the Palouse region of eastern Washington and northern Idaho, where peas are commonly grown commercially in rotation with wheat. Rhizobia leguminosarum biovar. viciae colonizes pea roots, stimulating nodule formation. Our field experiment explored how these rhizobial bacteria influenced plant productivity through direct and indirect mechanisms.

Field experiment

The experiment was conducted during the summer of 2016. Thirty-two holes were dug at the Washington State University Spillman Agronomy Farm in a 4×6 grid with 1m spacing. A single plastic tote (60L) was inserted into each hole prior to the initiation of the experiment to ensure that soil treatments were self-contained (Appendix S1: Figure S14). Peas have a shallow root structure (Weaver and Bruner 1927) and perform well in such conditions.

Our experiment consisted of a randomized block design with 8 blocks and 4 unique soil treatments; one replicate of each treatment was incorporated into each block. The four treatments were: (1) sterilized soil, (2) sterilized soil with rhizobia added, (3) non-sterilized soil with N-

83 fertilizer added, and (4) a non-sterilized control. High levels of nitrate fertilizer suppress legume- rhizobia symbiosis (Hopkins et al. 1932), so the N-fertilized treatment produces plants that obtain nitrogen from mineralized sources in the soil rather than rhizobia. Fertilized treatments were treated with ammonium nitrate at a rate of 90 kg/hectare. Soil was sterilized by bagging it in 61×91cm bags placed in a steam autoclave at 7psi/111⁰C for 8h. Pea seeds treated with R. leguminosarum biovar viciae were inoculated with N-Dure, a commercially available peet-based inoculant (Verdesian Life Sciences, Cary, NC) via the slurry method (Deaker et al. 2004).

Treated soil was added back to totes 48h later. Floorless, 60cm3 mesh cages (MegaView,

Taichung City, Taiwan) were erected over each tote to create fully-enclosed mesocosms. Peas

(Pisum sativum var. Banner) were initially planted on April 26 in a greenhouse (Appendix S3) on the Washington State University (WSU) campus. Plants were watered regularly and transplanted to the field on May 10, corresponding to a typical commercial planting date. Twenty-five plants were transplanted into each tub in a 5×5 grid (Appendix S1: Fig. S15), which is in line with typical densities within commercial pea fields. All plants were watered after transplanting to aid establishment, but were fed only by natural rainfall the remainder of the experiment.

After plants had a month to establish, 25 PEMV-inoculative A. pisum individuals were released in each cage at the lower-right corner on June 10. Colonies of PEMV-inoculative A. pisum were maintained on pea in a greenhouse on the WSU campus since 2013 (Appendix S3).

On July 6, the terminal leaflet was clipped from 9 randomly selected plants in each cage

(Appendix S1: Fig. S15). The tissue was analyzed with DAS-ELISA for the presence of PEMV using a commercial ELISA kit (A.C. Diagnostics, Fayetteville, AR) following the included procedure to determine infection prevalence in each cage.

Aphids were sampled on July 12 using a DVAC suction sampler (Rincon-Vitova,

Ventura CA). Cages were temporarily lifted off each tub, aphids were collected, and the cages were replaced. The number of aphids from each sample was then counted. Peas were harvested by hand on August 2, hulled, and dried in an oven for 5d at 37⁰C before weighing.

Soil nodulation assay

Although our goal was to measure the degree of nodulation occurring in plants from each tote in the field, the roots from different plants within a single tote became indistinguishable over the course of the experiment, making such estimates impossible. Consequently, we conducted a follow-up bioassay to confirm the presence/absence of rhizobia in each replicate at the conclusion of the experiment. Following harvest, a post-hole digger was used to take soil samples at 4 uniform locations in a square pattern from each tub. The soil sample was homogenized, mixed at a 1:1 ratio with sand, and potted in 2.6L pots. Four replicates were used from each tub. Two pea seeds were planted in each pot, and pots in which both seeds germinated were immediately thinned to one plant following emergence. Plants were allowed to grow for 47 to 50 d. The entire plant was gently removed from the pot, and the soil was washed off, exposing the roots. The total nodules occurring on each plant were counted, excised with a razor blade, dried for 5d at 37⁰C in a drying oven, and weighed.

Data analysis

All analyses were conducted in R Studio (ver. 0.99.903). For the field experiment, analyses determined the effects of sterilization, rhizobia, and nitrogen on aphids, PEMV infection rate, and yield (Appendix S2: Table S3). Yield (dry mass) and aphid abundance were

85 modelled using linear mixed-effects models that included a random block effect (lme4 package,

Bates et al. 2013), with significance tested using Satterthwaite’s approximations (lmerTest package, Kuznetsova et al. 2013). Infection rate (number of positive plants out of 9 sampled) was analyzed with mixed-effects logistic regression using the same predictors.

For the soil nodulation assay, we averaged the dry nodule biomass for the 4 plants in each tote. Pots in which no plants germinated, and two replicates from the autoclaved treatment that showed nodulation, were excluded. We then used ANOVA and Tukey’s HSD test to determine how treatment influenced average dry nodule biomass. We used linear mixed-effects models to determine effects of soil nodulation on aphids, PEMV, and yield. Since a reliable range of nodulation was only detected in the two non-sterilized treatments, sterilized treatments were excluded. These analyses were performed by assessing each response in the field experiment

(aphids, infection, and yield) as a function of dry nodule biomass (from the soil bioassay) and the presence/absence of nitrogen. This allowed us to parse apart the effect of nodulation from that of nitrogen addition. Aphid population and yield was analyzed using a linear mixed-effects model, while infection incidence was analyzed in a mixed-effects logistic regression.

To assess relationships between aphids, infection, and yield, linear mixed-effects models were used to analyze yield in response to infection rate and aphid population; logistic regression analyzed infection rate in response to aphids. However, since significant correlations were detected between yield, aphid population, and infection rate, we were unable to determine if trends in yield were directly related to soil treatments, or mediated through changes in aphid populations and infection rates. Consequently, we used a structural equation model to identify linkages between these variables. We used the piecewiseSEM() package (Lefcheck 2015) for model assembly and selection. The initial model was indicative of complete mediation, in which

86 the soil predictors influenced aphids, aphids influenced infection rate, and infection rate influenced yield. Due to our relatively small sample size (n = 30), all models were of a simple linear form. Although infection rate follows a binomial distribution, residuals did not depart from normality (Shapiro-Wilk test, W = 0.94, P > 0.05; Appendix S1: Figure S12). We used the sem.fit() function, which conducts a d-separation test (Shipley 2009) to identify missing linkages in the data and add all significant (P < 0.05) linkages. We then used the sem.coef() function, which calculates the individual significance of each linkage, to parse out non-significant (P >

0.05) links. Resulting models (n = 8) were evaluated using sample-size corrected AICc. Fisher’s

C statistic was used to test if discrepancies existed between models and data (Shipley 2009).

For models with significant support, standardized coefficients were calculated that approximated the relative effect size of each linkage. Coefficients of numeric predictors (aphids and infection) were standardized using the “scale” method, where the values reflect the change in the response variable (measured in SD) as a result of a one standard deviation change in the predictor. Coefficients of categorical predictors (rhizobia, sterilization, and nitrogen) were standardized using the “range” method, where the reported values reflect the change in the response variable (in SD) produced by the range of the predictor (0 to 1).

RESULTS

Field experiment and soil nodulation

Soil treatment influenced nodule biomass (F3,26 = 10.12, P = 0.00014), with autoclaved

(P = 0.001), autoclaved + rhizobia (P = 0.0004), and nitrogen (P = 0.001) treatments exhibiting less nodulation than controls (Fig. 4.1A). Aphids had the greatest abundance in initially- sterilized totes (P = 0.017), although this effect was reduced when rhizobia were added (P =

0.036). Nitrogen alone did not affect aphids (P = 0.48; Fig. 4.2A). PEMV increased due to sterilization (P = 0.0006) and decreased due to rhizobia (P = 0.031) and nitrogen (P = 0.021; Fig.

4.2B). Yield was reduced by sterilization (P = 0.0019) but unaffected by rhizobia (P = 0.19) or nitrogen (P = 0.88).

Within the non-sterilized treatments, aphid populations were positively affected by nodulation, as measured by nodule biomass (P = 0.038, Fig. 4.3). However, PEMV incidence (P

= 0.15) and yield (P = 0.15) were unaffected by nodulation. In these analyses, the addition of nitrogen fertilizer did not influence aphids (P = 0.84), PEMV (P = 0.43), or yield (P = 0.44).

Structural equation models

Significant correlations existed between response variables. Yield was negatively correlated with aphids (P < 0.0001; Fig. 4.4A) and PEMV (P < 0.0001; Fig. 4.4B), while PEMV was highly influenced by aphid populations (P < 0.0001; Fig. 4.4C). When analyzing the effects of aphid population (P = 0.035) and infection incidence (P = 0.002) on yield, results showedthat both factors negatively influenced yield, although these effects were independent.

Eight different models were tested using AICc. Although data from the best model is presented here (Fig. 4.5), two others were moderately supported (Appendix S1: Fig. S13). The model fit the data reasonably well, as there were no significant discrepancies between the model and the data (C = 7.62, P = 0.27). All three supported models excluded nitrogen fertilization as a predictor. In the best-supported model, sterilization negatively influenced yield directly (P =

0.06) and indirectly by promoting aphid abundance (P = 0.003). Rhizobia had an indirect positive effect on yield that was mediated by a reduction in aphids (P = 0.03). Aphids negatively

88 influenced yield directly (P = 0.05), and indirectly through an increase in PEMV (P = 0.0004).

Yield was most negatively influenced by PEMV (P = 0.01).

DISCUSSION

Our results show that soil microbes exert considerable impacts on aboveground communities. Inoculation with soil-borne rhizobia resulted in a significant reduction in the incidence of PEMV. Although other studies have observed rhizobia-associated decreases in viral susceptibility in individual plants challenged uniformly (Elbadry et al. 2006), few have examined the effect of rhizobia on viral incidence in a field setting where vector population and behavior plays a roll. Structural equation models reveal that this reduction in PEMV was mediated through reductions in aphid populations, which is consistent with previous observations that rhizobia inoculation reduced populations of phloem-feeding herbivores (Brunner et al. 2015,

Dean et al. 2009). Although this study examined more community complexity than comparable studies (Elbadry et al. 2006, Brunner et al. 2015, Kempel et al. 2009), other community guilds, such as insect predators, mycorrhizal fungi, and other pathogens, may also be influenced by soil community composition and could in turn be mediators of yield (Avis et al. 2008, Schadler and

Ballhorn 2016).

The structural equation approach we employed allowed us to quantify the individual contribution of each factor to each response variable. The best-supported model indicated that yield was directly, and negatively, influenced by PEMV incidence, aphid abundance, and soil sterilization. By comparing the standardized coefficients of these parameters, we can infer that

PEMV incidence was 40% more influential than aphids on yield, and 300% more influential than soil sterilization. However, this result does not include the indirect effect of soil sterilization on yield mediated by aphids and virus. Similarly, although sterilization resulted in higher aphid

89 abundance, the model indicates that this sterilization-mediated boost in aphid populations can be reduced by 69% simply through the addition of rhizobia back into the soil. This indicates that a significant portion of the soil community functionality that is lost during sterilization can be restored through the addition of a single influential species such as R. leguminosarum.

Results from the follow-up nodulation bioassay to estimate the degree of nodulation that occurred during the field experiment were used to confirm sterility for the autoclaved treatments and determine how nodulation may have influenced pest and disease loads in the experiment.

Unfortunately, reliable results were not obtained for the autoclaved + rhizobia treatment, as very few plants potted in that soil developed nodules, despite the fact that rhizobia-inoculated plants had been growing in that soil following sterilization. Although this result could indicate that the rhizobia-inoculation wasn’t successful, strong variation in performance between sterilized treatments with/without inoculation (Fig. 4.2) indicate successful inoculation. An alternative hypothesis is that plants in the experiment did nodulate, but that the proliferation of rhizobial colonies was localized in small areas near the plant roots resulting in relatively few bacteria being sampled (Wadisirisuk et al. 1989).

Interestingly, higher levels of nodulation were correlated with higher aphid abundances, though infection and yield were affected. This result is the opposite of previous studies (Brunner et al. 2015) examining the relationship between nodulation and herbivore abundances. Previous authors have speculated that higher rates of nodulation would mean more nitrogen for the plant and consequently higher levels of N-derived defenses (Bryant et al. 1983, Johnson et al. 1987,

Irmer et al. 2015). However, our study differed in that it included an aphid-transmitted pathogen,

PEMV. Given the demonstrated ability of plant viruses to down-regulate vector defenses in host plants (Zhang et al. 2012, Luan et al. 2013), it is possible that the enhanced defenses of highly-

90 nodulated plants were negated by the defense-suppressive effects of viral infection. If this were true, then the higher levels of plant-N in highly nodulated plants would be available to vectors without the corresponding negative impact of higher defense levels (Babiker et al. 1994).

Alternatively, herbivore specialization may determine the response of a specific pest to a rhizobia-mediated boost in nitrogen. Since specialists often possess host-specific adaptations for overcoming plant defenses, they may not be as hindered by high levels of defensive compounds as generalists (Van Oosten et al. 2008). Because A. pisum is highly specialized on P. sativum

(Peccoud et al. 2009), it is likely that A. pisum individuals were not negatively influenced by higher levels of defensive compounds associated with heavily nodulated host plants.

It is important to note the context in which each factor was analyzed, however. Our experiment was designed to compare three factors: (1) sterilization (autoclaved vs. control), (2) rhizobia (autoclaved+rhizobia vs. autoclaved), and (3) nitrogen fertilization (nitrogen vs. control). Consequently, the effects of rhizobia were examined in the context of fully sterilized soil, while the effects of nitrogen were examined in the context of healthy, non-sterilized soil.

Although our path analysis did not detect any significant effect of supplemental nitrogen on vector populations, pathogen prevalence, or host fitness, it is entirely likely that results would be different if supplemental nitrogen were examined in the context of sterilized soil.

The factors influencing plant yield are complex and inter-dependent. Insect herbivores, soil nutrition, plant pathogens, and soil-borne microbes may impact yield directly, but also by influencing each other. Consequently, the net influence of soil composition on plant yield may be highly dependent on ecological context, such as the relative prevalence of pests and pathogens in the environment. The relative economic importance of different threats (pests, pathogens, etc) varies between cropping systems, and threats may be artificially reduced through mitigation

91 practices. Consequently, discerning the relationships between these different factors is critical to understanding the response of plants to varying levels of soil health.

ACKNOWLEDGEMENTS

We would like to thank undergraduate technicians Akaisha Charlton, Lee Mendez, and

Lucy Eggleston who provided assistance for this study. We would also like to thank Amanda

Linskey and Dan Dreesman, who generously allowed us to use their autoclaving facilities. This research was supported by a USDA-NIFA Predoctoral Fellowship 2016-67011-24693.

LITERATURE CITED

A’Bear, A. D., Johnson, S. N., and T. H. Jones. 2014. Putting the “upstairs–downstairs” into

ecosystem service: What can aboveground–belowground ecology tell us? Biological Control

75: 97–107.

Antoun, H., C. J. Beauchamp, N. Goussard, R. Chabot, and Lalande, R. 1998. Potential of

Rhizobium and Bradyrhizobium species as plant growth promoting rhizobacteria on non-

legumes: Effect on radishes (Raphanus sativus L.). Plant and Soil 204: 57–67.

Arora, N. K., S. C. Kang, and D. K. Maheswari. 2001. Isolation of siderophore-producing strains

of Rhizobium meliloti and their biocontrol potential against Macrophomina phaseolina that

causes charcoal rot of groundnut. Current Science 81: 673-677.

Avis, T. J., V. Gravel, H. Antoun, and R. J. Tweddell. 2008. Multifaceted beneficial effects of

rhizosphere microorganisms on plant health and productivity. Soil Biology and Biochemistry

40: 1733–1740.

Azcón-Aguilar, C. and J. M. Barea. 1997. Arbuscular mycorrhizas and biological control of soil-

borne plant pathogens – an overview of the mechanisms involved. Mycorrhiza 6: 457–464.

Babiker, E. E., E. A. E. El Sheik, A. J. Osman, and A. H. El Tinay. 1994. Effect of nitrogen

fixation, nitrogen fertilization and viral infection on yield, tannin and protein contents and in

vitro protein digestibility of faba bean. Plant Foods for Human Nutrition 47: 257-263.

Bae, H., R. C. Sicher, M. S. Kim, S-H. Kim, M. D. Strem, R. L. Melnick and B. A. Bailey. 2009.

The beneficial endophyte Trichoderma hamatum isolate DIS 219b promotes growth and

delays the onset of the drought response in Theobroma cacao. J Exp Bot 60: 3279–3295.

Ballhorn, D. J., S. Kautz, and M. Schädler. 2013. Induced plant defense via volatile production is

dependent on rhizobial symbiosis. Oecologia 172: 833–846.

Bardgett, R. 2005. The Biology of Soil: A Community and Ecosystem Approach. First Edition.

Oxford University Press Inc., New York.

Breil, B., J. Borneman, and E. W. Triplett. 1996. A newly discovered gene, tfuA, involved in the

production of the ribosomally synthesized peptide antibiotic trifolitoxin. J. Bacteriol. 178:

4150–4156.

Brunner, S. M., R. J. Gos, S. J. Swenson, S. P. Foster, B. G. Schatz, Y. E. Lawley, and D. A,

Prishcmann-Voldseth. 2015. Impact of nitrogen fixing and plant growth-promoting bacteria

on a phloem-feeding soybean herbivore. Applied Soil Ecology 86: 71–81.

Bryant, J. P., F. S. Chapin III, and D. R. Klein. 1983. Coarbon/Nutrient balance of boreal plants

in realtion to vertebrate herbivory. Oikos 40: 657-368.

Buee, M., W. De Boer, F. Martin, L. van OVerbeek, and E. Jurkevitch. 2009. The rhizosphere

zoo: an overview of plant-associated communities of microorganisms, including phages,

bacteria, archaea, and fungi and of some of their structuring factors. Plant and Soil 231: 189-

212

Clement, S. L., Husebye, D. S., and S. D. Eigenbrode. 2010. Ecological factors influencing pea

aphid outbreaks in the US Pacific Northwest. Pages 107-128 in P. Kindlmann, A.F.G. Dixon,

and J.P. Michaud, editors. Aphid Biodiversity under Environmental Change. Springer

Publishing, Berlin, Germany.

Dakora, F. D., C. M. Joseph, and D. A. Phillips. 1993. A. Alfalfa (Medicago sativa L.) root

exudates contain isoflavonoids in the presence of Rhizobium meliloti. Plant Physiol 101:

819–824.

Deaker, R., R. J. Roughley, and I. R. Kennedy. 2004. Legume seed inoculation technology—a

review. Soil Biology and Biochemistry 36: 1275–1288.

Dean, J. M., M. C. Mescher, and C. M. D. Moraes. 2009. Plant–rhizobia mutualism influences

aphid abundance on soybean. Plant Soil 323: 187–196.

Dean, J. M., Mescher, M. C. and De Moraes, C. M. Plant Dependence on Rhizobia for Nitrogen

Influences Induced Plant Defenses and Herbivore Performance. International Journal of

Molecular Sciences 15, 1466–1480 (2014).

Elbadry, M., R. M. Taha, K. A. Eldougdoug, and H. Gamal-Eldin. 2006. Induction of systemic

resistance in faba bean (Vicia faba L.) to bean yellow mosaic potyvirus (BYMV) via seed

bacterization with plant growth promoting rhizobacteria. J Plant Dis Prot 113: 247–251.

Estevez de Jensen, C., J. A. Percich, and P. H. Graham. 2002. Integrated management strategies

of bean root rot with Bacillus subtilis and Rhizobium in Minnesota. Field Crops Research

74: 107–115.

Evelin, H., R. Kapoor, and B. Giri. 2009. Arbuscular mycorrhizal fungi in alleviation of salt

stress: a review. Annals of Botany 104: 1263-1280.

Gehring, C. and A. Bennett. 2009. Mycorrhizal fungal-plant-insect interactions: the importance

of a community approach. Environmental Entomology 38: 93-102

Graham, P. H. and C. P. Vance. 2003. Legumes: Importance and Constraints to Greater Use.

Plant Physiol. 131: 872–877.

Hallmann, J., A. Quadt-Hallmann, W. G. Miller, R. A. Sikora, and S. E. Lindow. 2001.

Endophytic Colonization of Plants by the Biocontrol Agent Rhizobium etli G12 in Relation to

Meloidogyne incognita Infection. Phytopathology 91: 415–422.

Hopkins, E. W., P. W. Wilson, and W. H. Peterson. 1932. Influence of potassium nitrate on

nodule formation and nitrogen fixation by clover. Plant Physiol 7: 597–611.

Huang, H. C., R. S. Erickson, and T. F. Hsieh. 2007. Control of bacterial wilt of bean

(Curtobacterium flaccumfaciens pv. flaccumfaciens) by seed treatment with Rhizobium

leguminosarum. Crop Protection 26: 1055–1061.

Irmer, S., N. Podzun, D. Langel, F. Heidemann, E. Kaltenegger, B. Schemmerling, C. M.

Geilfus, C. Zörb, and D. Ober. 2015. New aspect of plant–rhizobia interaction: Alkaloid

biosynthesis in Crotalaria depends on nodulation. PNAS 112: 4164–4169.

Jaber, L. R. and S. Vidal. 2009. Interactions between an endophytic fungus, aphids and

extrafloral nectaries: do endophytes induce extrafloral-mediated defences in Vicia faba?

Functional Ecology 23: 707–714.

Jallow, M. F. A., D. Dugassa-Gobena, and S. Vidal. 2004. Indirect interaction between an

unspecialized endophytic fungus and a polyphagous math. Basic and Applied Ecology 5:

183-191.

Johnson, N. D., B. Liu, and B. L. Bentley. 1987. The effects of nitrogen fixation, soil nitrate, and

defoliation on the growth, alkaloids, and nitrogen levels of Lupinus succulentus (Fabaceae).

Oecologia 74: 425–431.

Kempel, A., R. Brandl, and M. Schädler. 2009. Symbiotic soil microorganisms as players in

aboveground plant–herbivore interactions – the role of rhizobia. Oikos 118: 634–640.

Kuznetsova, A., Per B. Brockhoff and R. H. B. Christensen 2013. lmerTest: Tests for random

and fixed effects for linear mixed effect models (lmer objects of lme4 package).. R package

version 2.0-0. http://CRAN.R-project.org/package=lmerTest

Lefcheck, Jonathan S. 2015. piecewiseSEM: Piecewise structural equation modeling in R for

ecology, evolution, and systematics. Methods in Ecology and Evolution 7: 573-579.

Luan, J. B., D. M. Yao, T. Zhang, L. L. Walling, M. Yang, Y. J. Wang, and S. S. Liu. 2013.

Suppression of terpenoid synthesis in plants by a virus promotes its mutualism with vectors.

Ecology Letters 16: 390-398.

Mabrouk, Y., L. Zourgui, B. Sifi, P. Delavault, P. Simier and O Belhadj. 2007. Some compatible

Rhizobium leguminosarum strains in peas decrease infections when parasitised by Orobanche

crenata. Weed Research 47: 44–53.

Marcel, G. A., A. van der Heijden, J. N. Klironomos, M. Ursic, P. Moutoglis, R. Streitwolf-

Engel, T. Boller, A. Wiemken, and I. R. Sanders. 1998. Mycorrhizal fungal diversity

determines plan biodiversity, ecosystem variability and productivity. Nature 396: 69-72.

Matson, P. A., W. J. Parton, A. G. Power, and M. J. Swift. 1997. Agricultural intensification and

ecosystem properties. Science 277: 504-509.

Omar, S. A. and M. H. Abd-Alla. 1998. Biocontrol of fungal root rot diseases of crop plants by

the use of rhizobia and bradyrhizobia. Folia Microbiol 43: 431–437.

Peccoud, J., A. Ollivier, M. Plantegenets, and J.-C. Simon. 2009. A continuum of genetic

divergence from symatric host races to species in the pea aphid complex. PNAS 106: 7495-

7500.

Pineda, A., S.-J. Zheng, J. J. A. van Loon, C. M. J. Pieterse, and M. Dicke. 2010. Helping plants

to deal with insects: the role of beneficial soil-borne microbes. Trends in Plant Science 15:

507–514.

Rabie, G. H. 1998. Induction of fungal disease resistance in Vicia faba by dual inoculation with

Rhizobium leguminosarum and vesicular-arbuscular mycorrhizal fungi. Mycopathologia 141:

159–166.

Raps, A. and S. Vidal. 1998. Indirect effects of an unspecialized endophytic fungus on

specialized plant – herbivorous insect interactions. Oecologia 114: 541–547.

Reitz, M. et al. Importance of the O-antigen, core-region and lipid A of rhizobial

lipopolysaccharides for the induction of systemic resistance in potato to Globodera pallida.

Nematology 4, (2002).

Reitz, M., P. M. Oger, A. Meyer, K. Niehaus, S. K. Farrand, J. Hallmann, R, A, Sikora. 2000.

Lipopolysaccharides of Rhizobium etli Strain G12 Act in Potato Roots as an Inducing Agent

of Systemic Resistance to Infection by the Cyst Nematode Globodera pallida. Appl Environ

Microbiol 66: 3515–3518.

Roesch, L. F. W., R. R. Fulthorpe, A. Riva, G. Casella, A. K. Hadwin, A. D. Kent, S. H. Daroub,

F. A. Camargo, W. G. Farmerie, and E. W. Triplett. 2007. Pyrosequencing enumerates and

contrasts soil microbial diversity. ISME J 1: 283–290.

Santos, F., M. Fernanda, G. V. Peñaflor, P. W. Paré, P. A. Sanches, A. C. Kamiya, M. Tonelli, C.

Nardi, and J. M. S. Bento. 2014. A Novel Interaction between Plant-Beneficial Rhizobacteria

and Roots: Colonization Induces Corn Resistance against the Root Herbivore Diabrotica

speciosa. PLOS ONE 9: e113280.

Schädler, M. and D. J. Ballhorn. 2016. Beneficial soil microbiota as media of the plant defensive

phenotype and aboveground plant-herbivore interactions. Pages 305-343 in F. M. Cánovas,

F. M. Lüttge, and R. Matyssek, editors. Progress in Botany Vol. 78 Springer International

Publishing, New York, New York, USA.

Shipley, B. 2009. Confirmatory path analysis in a generalized multilevel context. Ecology 90:

363-368.

Smil, V. 1999. Nitrogen in crop production: An account of global flows. Global Biogeochem.

Cycles 13: 647–662.

Smith, S. E. and D. J. Read. 2008. Mycorrhizal symbiosis. Third edition. Academic press.

Thamer, S., M. Shädler, D. Bonte, and D. J. Ballhorn. 2011. Dual benefit from belowground

symbiosis: nitrogen fixing rhizobia promote growth and defense against a specialist

herbivore in a cyanogenic plant. Plant Soil 341: 209-219.

Valenzuela-Soto, J. H., M. G. Estrada-Hernández, E. Ibarra-Laclette, and J. P. Délano-Frier.

2010.Inoculation of tomato plants (Solanum lycopersicum) with growth-promoting Bacillus

subtilis retards whitefly Bemisia tabaci development. Planta 231: 397–410 (2010).

Van Oosten, V. R., N. Bodenhausen, P. Reymond, J. A. Van Pelt, L. C. Van Loon, M. Dicke and

C. M. J. Pieterse. 2008. Differential effectiveness of microbial induced resistance against

herbivorous insects in Arabidopsis. Molecular Plant-Microbe Interaction 21: 919-930.

Wadisirisuk, P., S. K. A. Danso, G. Hardarson, and G. D. Bowen. 1989. Influence of

Bradyrhizobium japonicum location and movement on nodulation and nitrogen fixation in

soybeans. Applied and Environmental Microbiology 55: 1711-1716.

Wall, D. H., R. D. Bardgett, and E. Kelly. 2010. Biodiversity in the dark. Nature Geosci. 3: 297–

298.

Weaver, J. E. and W. E. Bruner. 1927. Root development of vegetable crops. First Edition.

McGraw-Hill Book Company, New York City, New York, USA.

Weyens, N., D. Lelie, S. van der Taghavi, L. Newman, and J. Vangronsveld. 2009. Exploiting

plant–microbe partnerships to improve biomass production and remediation. Trends in

Biotechnology 27:591–298.

Whitaker, M. R. L., N. Katayama, and T. Ohgushi. 2014. Plant–rhizobia interactions alter aphid

honeydew composition. Arthropod-Plant Interactions 8: 213–220.

Wilson, K. G. and R. E. Stinner. 1984. A potential influence of rhizobium activity on the

availability of nitrogen to legume herbivores. Oecologia 61: 337–341.

Yang, J., J. W. Kloepper, and C.-M. Ryu. 2009. Rhizosphere bacteria help plants tolerate abiotic

stress. Trends Plant Sci. 14: 1–4.

Zehnder, G., J. Kloepper, C. Yao, and G. Wei. 1997. Induction of systemic resistance in

cucumber against cucumber beetles (Coleoptera: Chrysomelidae) by plant growth-promoting

Rhizobacteria. J Econ Entomol 90: 391–396.

Zhang, X., X. Zhang, J. Singh, D. Li and F. Qu. 2012. Temoerature-dependent survival of turnip-

crinckle virus-infected Arabidopsis plants relies on an RNA silencing-based defense that

requires DCL2, AGO2 and HEN1. Journal of Virology 88: 6847-6854.

100

FIGURE LEGENDS

FIG 4.1. Results from the soil nodulation assay. The (A) Average number of nodules/plant and

(B) dry nodule biomass (mg/plant) from various soil treatments (means ± SE are shown). + - significant difference at α = 0.10; * - significant difference at α = 0.05.

FIG. 4.2. (A) Aphid population, (B) number of infected plants (out of 9 tested), and (C) plant yield (g) for each soil treatment (means ± SE are shown)

FIG. 4.3. Effect of nodule biomass on observed aphid populations in the field experiment. Aphid populations increased with higher rates of nodulation (P = 0.038) and were not significantly affected by the addition of nitrogen fertilizer (P = 0.84).

FIG 4.4. Observed correlations between response variables. Yield was negatively associated with (A) PEMV infection (P < 0.0001) and (B) aphid abundance (P < 0.0001). (C) PEMV infection was positively associated with aphid population (P < 0.0001).

FIG 4.5. Path diagram (C = 7.62, P = 0.27) of variable interactions in the best-supported model.

P-values and standardized coefficients for each model link are reported. Coefficients of numeric predictors (aphids and infection) were standardized using the “scale” method, where the values reflect the change in the response variable (measured in number of standard deviations) as a result of a one standard deviation change in the predictor. Coefficients of categorical predictors

(rhizobia, sterilization, and nitrogen) were standardized using the “range” method, where the reported values reflect the change in the response variable (in standard deviations) produced by the range of the predictor (0 to 1).

101

FIG. 4.1

* * *

102

FIG. 4.2.

103

FIG. 4.3.

104

FIG. 4.4

105

FIG. 4.5.

106

Appendix S1 Supplemental figures

a) b) c)

Figure S1. Study system. a) A pea plant (Pisum sativum var. Banner) infected with PEMV. b) Pea aphids, Acyrthosiphon pisum (Photo courtesy of Shipher Wu under a Creative Commons Attribution 4.0 International Public License, https://creativecommons.org/licenses/by/ 4.0/legalcode) c) Adult pea leaf weevils, Sitona lineatus (Photo courtesy of Gail Hampshire under a Creative Commons Attribution 4.0 International Public License, https://creativecommons.org/licenses/by/4.0/legalcode)

107

a) b) c)

Figure S2. Experimental setup for greenhouse choice assays. a) The choice assay for weevils feeding on infected (left) and uninfected (right) pea plants. Two adult weevils were placed in a vial equidistant from each plant, and total defoliation (cm2) for each plant was measured after 6 days. b) The choice assay for either PEMV-inoculative or non-inoculative aphids choosing between weevil-damaged (right) and undamaged (left) plants. c) In order to control for the effect of lighting position, microclimate, and other variability within the greenhouse, each cage contained a plant pair with inoculative aphids and one pair with non-inoculative aphids.

108

Figure S3. The a) relative size of uninfected and PEMV-infected plants and b) relative number of aphids on uninfected and PEMV-infected plants from the field observational study. A value of 0.5 indicates that uninfected plants were of equal size, or had equal number of aphids, compared with paired PEMV-infected plants. All points represent means ± SE. * - significant difference at α = 0.05.

109

Figure S4. The relative abundance of uninfected aphids on paired undamaged and S. lineatus- damaged plants in greenhouse choice assays.

110

Figure S5. a) The relative PEMV titer (sample absorbance/negative control absorbance/total protein) after 6d of feeding by inoculative aphids for pea plants that had never been previously exposed to S. lineatus prior to aphid feeding (undamaged) or were previously exposed to feeding by S. lineatus (damaged) prior to aphid feeding (damaged) PEMV. All points represent means ± SE. b) The relative PEMV titer after 6d of feeding by inoculative aphids for pea plants that were exposed to varying numbers of S. lineatus adults prior to aphid feeding.

111

Figure S6. The relative PEMV titer (sample absorbance/negative control absorbance/total protein) after 6d of S. lineatus feeding for pea plants that experienced varying defoliation from S. lineatus (proportion leaf area removed) after being inoculated with PEMV. The dotted line shows the best-fit linear regression line.

112

a) b)

Figure S7. Experimental setup at Tukey Orchard. a) Cages were erected in a grid pattern with 1m spacing. b) Plants were first treated with 25 aphids (inoculative or not) and adult pea leaf weevils (2 or none) 24 hours before being placed in the middle of a 8-plant ring.

113

Figure S8. Graphical sequence of events for the field mesocosm experiment at Tukey Orchard. a) Plants were treated with aphids and weevils. b) Plants were placed in the center of an 8-plant grid. c) Aphids reproduced, and d) dispersed to surrounding plants. Aphid population on each of the 9 plants was recorded.

114

Figure S9. Cages were applied to the top and bottom of pea plants to restrict where vectors fed. Inoculation success and viral titer were measured in response.

115 a) b) c)

Figure S10: Effect of S. lineatus and aphid inoculation status (PEMV-inoculative or non- inoculative) on a) aphid reproduction (nymphs per established adult), b) aphid survival, and c) aphid movement (proportion off-center plant). * - indicates significant difference at α = 0.05.

116

Figure S11. Average proportion of aphids settling on each plant structure with or without S. lineatus present. * - indicates a significant difference at α = 0.05.

117

Figure S12. Density plot of residuals from the infections~aphids function fitted to data from the soil mesocosm experiment.

118

Figure S13. The three best-supported path models explaining data from the soil mesocosm experiment. Models are statistically indistinguishable from each other (AICc < 2).

119

Figure S14. Photo showing the process of putting tubs into holes at Spillman Farm and filling with treated soil.

120

20 adult PEMV-inoculative aphids released

Figure S15. Diagram demonstrating the distribution of plants within each replicate. Twenty-five peas were planted in each tub in a 5x5 grid. Dots denote a plant. Red dots denote a plant that was sampled and analyzed with ELISA for infection with PEMV.

121

Appendix S2. Supplemental tables.

Field # pairs measured

Mraz 8

Rimrock1 3

Rimrock2 5

Busch1 8

Busch2 8

Klemgard 8

Union Flat 8

Schultheis 8

Spillman 10

Kambitsch 8

Tukey 8

Tombstone 8

Table S1. Number of healthy/infected plant pairs measured from each sampling location in the observational study from pea fields.

122

Crop Virus Non-vector pest

Corn Maize Dwarf Mosaic European corn borer, corn rootworm

Rice Rice Grassy Stunt Rice water weevil

Peas Pea Enation Mosaic Pea leaf weevil

Soybean Soybean Mosaic Bean leaf beetle

Potato Potato Leafroll Colorado potato beetle

Wheat Barley Yellow Dwarf Hessian fly, cereal leaf beetle

Table S2: Commonly co-occurring chewing, non-vector herbivores and insect-transmitted viruses in agricultural systems.

123

Treatment Sterilized Rhizobia Nitrogen Control 0 0 0 Autoclaved 1 0 0 Autoclaved+Rhizobia 1 1 0 Nitrogen 0 0 1

Table S3: Predictor coding for each individual treatment from the soil mesocosm experiment.

124

Appendix S3. Supplemental methods.

DAS-ELISA Standard Operating Protocol, AC Diagnostics

1) Add coating antibody

a) Figure out how much coating antibody solution you will need. Each well will receive 100μl (0.1 ml) of solution. So, for one plate (96 wells) you should prepare 10ml of coating antibody solution.

b) Dilute your concentrated coating antibody buffer down to 1x. The buffers come concentrated at 10x. So, if you’re preparing 10ml of coating antibody buffer, you’ll add 1ml of the concentrated buffer to 9ml of distilled water.

c) Dilute your coating antibody in the coating antibody buffer you just prepared. The antibody comes concentrated at 200x. So, if you’re preparing 10ml of coating antibody solution, pipette 50 μl (0.05ml) of the antibody into your 10ml of buffer.

d) Pipette 100μl of your prepared coating antibody solution into each well.

e) Wrap in aluminum foil and incubate for 4 hours at 25 degrees C (room temperature), or at 4 degrees C (refrigerator) overnight.

2) Add sample

a) Figure out how much sample buffer you will need. The sample should be diluted down to about 1ml buffer/1g sample. A leaf is approximately 1g. So, if each sample is an individual leaf, you’ll need about 1ml of buffer for every sample you have. Thus, if you have 20 samples, you should prepare ~25ml of sample buffer.

b) Prepare your sample buffer. For each 1 ml of desired buffer, add 0.01g of Tween-20 and 0.0232g of sample buffer powder. So, if you’re preparing 25ml of sample buffer, add 0.25g (~250 μl) of Tween-20 and 0.58 g sample buffer powder to 25ml water.

c) Add desired amount of sample buffer (1ml of sample buffer per sample is a good place to start) to your samples in bags. Grind your samples.

d) If you are preparing controls for the first time, add 1.8ml sample buffer to each control vial and mix thoroughly. Pipette 220μl of this control solution into labeled tubes (this way, each tube will give you enough control for 2 wells). You probably won’t use all of this control solution today—save the controls you’re using today and put them in the fridge. The rest you can freeze until they are needed in the future.

d) Remove the plate from the aluminum foil and wash the plate 4x, allowing it to sit 2-3 minutes between washing.

125

e) Pipette your samples and controls into your plate, noting their location on a plate map.

f) Wrap in aluminum foil and incubate for 4 hours at 25 degrees C (room temperature), or at 4 degrees C (refrigerator) overnight. This is the best step to incubate overnight.

3) Add enzyme conjugate

a) Figure out how much enzyme conjugate solution you will need (it should be the same amount as the coating antibody solution). Each well will receive 100μl (0.1 ml) of solution. So, for one plate (96 wells) you should prepare 10ml of enzyme conjugate solution.

b) Dilute your concentrated enzyme conjugate buffer down to 1x. The buffers come concentrated at 10x. So, if you’re preparing 10ml of enzyme conjugate buffer, you’ll add 1ml of the concentrated buffer to 9ml of distilled water.

c) Dilute your enzyme conjugate in the enzyme conjugate buffer you just prepared. The conjugate comes concentrated at 200x. So, if you’re preparing 10ml of enzyme conjugate solution, pipette 50 μl (0.05ml) of the conjugate into your 10ml of buffer.

d) Remove the plate from the aluminum foil and wash the plate 4x, allowing it to sit 2-3 minutes between washing.

e) Pipette 100μl of your prepared enzyme conjugate solution into each well.

f) Wrap in aluminum foil and incubate for 2.5 hours at 25 degrees C (room temperature), or at 4 degrees C (refrigerator) overnight.

4) Add PNP color substrate

a) Figure out how much PNP solution you will need (it should be the same amount as the coating antibody solution and the enzyme conjugate solution). Each well will receive 100μl (0.1 ml) of solution. So, for one plate (96 wells) you should prepare 10ml of PNP solution.

b) PNP reacts with light. To minimize background reaction due to light, wrap your beaker in aluminum foil before using it to prepare your PNP solution.

c) Dilute your concentrated PNP buffer down to 1x. The buffer comes concentrated at 5x. So, if you’re preparing 10ml of PNP buffer, you’ll add 2ml of the concentrated buffer to 8ml of distilled water.

d) Add 1mg (1 PNP tablet) of PNP to every 5ml of solution. So, if you’re preparing 10ml of PNP solution, add 2 tablets to the 10ml of 1x PNP buffer you just prepared. Stir well until the tablet(s) is completely dissolved.

e) Remove the plate from the aluminum foil and wash the plate 4x, allowing it to sit 2-3 minutes between washing.

126 f) Wrap in aluminum foil and incubate for 30-60 minutes at 25 degrees C (room temperature). g) Read plate in a plate reader at a wavelength of 405nm. h) You can slow down the color reaction by putting the plate in a refrigerator. You can stop it completely by pipetting 50μl of 3M NaOH solution into each well.

127

Greenhouse settings

Greenhouse experiments were conducted in the entomology greenhouse on the campus of

Washington State University.

The greenhouse was programmed for a 16:8h photoperiod (light:dark) and temperatures that varied between 21 and 24°C during the light cycles, and between 16 and 18°C during the dark cycles.

128

Plants

All greenhouse experiments were conducted on pea plants grown from seed (variety

Banner) in Sunshine Mix LC1 potting media (Sun Gro Horticulture, Agawam, MA).

129

Aphids

All pea aphids used in the experiments were reared in greenhouse colonies (16:8h light:dark; 22°:17°C light:dark) on pea plants (variety Banner). We reared both uninfected and inoculative aphid colonies; both were started in 2012 from a field-collected population of inoculative aphids. To create the uninfected colony, we transferred 50 inoculative aphid adults to petri-dishes with filter paper and allowed them to reproduce for 3d. All aphid nymphs produced were uninfected, as PEMV is not maternally transmitted and the aphids never interacted with infected plants. Thereafter, these aphids were reared on uninfected pea plants; the inoculative aphid colony was reared on PEMV-infected plants. Genetic similarity between the colonies was maintained by transferring 100 adults from the uninfected to the infected colony every 2-3 months; at the same time we removed 100 adults from the infected colony and allow them to reproduce on filter paper for 3 d. Offspring of these aphids were uninfected and were transferred into the uninfected colony. This reciprocal transfer of aphids ensured that the genetic background was the same for the uninfected and infected colonies. Every two months we also test 20 aphids from each colony infection; our inoculative colony maintained 95-100% infection while our uninfected colony was 0% infected.

130

Phytohormone extraction and measurement

After the two assays we completed, all aboveground pea plant tissue was harvested.

Following harvest, tissue was frozen in liquid nitrogen and homogenized in a paint shaker for 20 s (Harbil 5G). Then, 5 µL of each extract was analyzed using a Thermo Electron LTQ-Orbitrap

XL Hybrid MS. Samples were separated on a Zorbax Extend-C18 HPLC column (Phenomenex

Kinetex, 5 µm, 100 × 2.10 mm) using 0.1% formic acid in water (Solvent A) and 0.1% (v/v) formic acid in acetonitrile (Solvent B) at a flow rate of 300 µL min–1. The gradient used was 0–

0.5 min, 20% B; 0.5–13.5 min, linear gradient to 100% B; 13.6–17 min, linear gradient to 30%

B; 17.1–18 min, linear gradient to 20% B.

For the first assay we excluded one sham-inoculated plant from the analyses of both jasmonic and salicylic acid because chromatograms were too noisy for analysis, and two PEMV- inoculated plants were excluded from the salicylic acid analysis for the same reason. For the second assay we excluded one undamaged plant from the analyses of both jasmonic and salicylic acid because chromatograms were too noisy for analysis; two S. lineatus-damaged plants were excluded from the jasmonic acid analysis for the same reason.

131