Enemy of My Enemy: Can the Rhizosphere Biota of Vincetoxicum Rossicum Act As Its “Ally” During Invasion?

Home , Astragalus canadensis, Desmodium canadense, Lespedeza hirta, Lupinus nootkatensis, Potentilla recta, Solidago pulchra

Enemy of my Enemy: Can the Rhizosphere Biota of Vincetoxicum rossicum Act as its “Ally” During Invasion?

Angela Dukes

A Thesis presented to The University of Guelph

In partial fulfillment of the requirements for the degree of Master of Science in Environmental Biology

Guelph, Ontario, Canada

Enemy of my Enemy: Can the Rhizosphere Biota of Vincetoxicum rossicum Act as its “Ally” During Invasion?

Angela Dukes Advisors: Dr. Pedro Antunes University of Guelph, 2017 Dr. Kari Dunfield

The ‘Enemy of my enemy’ (EE) is a major hypothesis in invasion ecology. It states that a non-native invader ‘accumulates generalist pathogens, which limit competition from indigenous competitors’. Few empirical studies have tested the EE hypothesis in plant invasions, especially on biotic rhizosphere interactions. Here, the EE hypothesis was tested by applying rhizosphere biota from the invasive plant Vincetoxicum rossicum

(VIRO) to five co-occurring native plant species, and four native legume species, respectively. Each of the native plant species, and VIRO were grown under controlled conditions for three months, either in presence or absence of soil biota from VIRO invaded and non-invaded soils. Rhizosphere biota from invaded areas had variable effects among native plants (including legumes). It was concluded that the accumulation of rhizosphere enemies that ‘spill’ onto native plants may not be a major factor in the invasive success of VIRO. The EE hypothesis was not supported.

iii

ACKNOWLEDGEMENTS

I deeply appreciated the patience of my supervisors: Dr. Kari Dunfield and Dr. Pedro Antunes. I worked in the Plant and Soil Ecology Lab at Algoma University with Dr. Pedro Antunes. He stood by me, affording me the time to take this journey. I had a steep learning curve, learning the world of soil biota, learning to write at a higher level, and to think more critically. But throughout I had his confidence – and that meant a lot. Pedro also allowed me time to take some tangents, working as a TA, and volunteering with several departmental workshops providing environmental education for children. He held me to a high level of achievement, but also encouraged me, guided me, was a confidant and friend – for this he was one of the best teachers I have ever had. Thank you for your time and confidence Pedro.

This work would not have been possible without the assistance of Dr. Akihiro Koyama, our Post Doc in the Plant and Soil Ecology Lab. He conducted the bootstrap analysis in R for this thesis. With these outputs I was able to analyse net biotic effect of Vincetoxicum rossicum on native plant species. Aki also conducted the phylogenetic analysis component of this thesis.

I am overwhelmingly grateful to my long-suffering family. They allowed me to defer my full responsibilities as a mom, they stood by me, and encouraged me. I am blown away at how fast my little boys have grown up! And honored to have a husband that so thoroughly nourished our children through my countless weekends away.

Many thanks to the Soil and Plant Ecology Lab at Algoma University: Undergraduate students (Kyle Snape, Alecia Finateri, Derissa Vincentini), and Lab coordinator Meaghan Sutherland. Your help in planting and harvest were much appreciated. I am grateful to the staff at the Ontario Forest Research Institute, particularly Darren Derbowka, and Keven Maloney for their suggestions and support in the use of the growth facilities. I deeply appreciate the assistance of Susan Meades, curator of the Northern Ontario Plant Database, for her help with plant identification. Finally, I owe a big thank-you to the staff of the Central Lake Ontario Conservation Authority (Jamie Davidson), Ganaraska Region Conservation Authority, (Ken Towle), Orono Crown Lands (John Standeven), Toronto Zoo (Elizabeth Heighington) and Ontario Ministry of Natural Resources (Bohdan Kowalyk), for your support and access to field sites. The work described in the following pages is the result of funding through Ontario Graduate Scholarship awarded to myself, the NSERC Discovery Grant awarded to Dr. Pedro Antunes, and bursaries from the University of Guelph.

TABLE OF CONTENTS

ABSTRACT ...... ii ACKNOWLEDGEMENTS ...... iii TABLE OF CONTENTS ...... iv LIST OF TABLES ...... vi LIST OF FIGURES ...... viii Chapter 1 Introduction 1.1 Invasion Biology ...... 1 1.2 Plant-Biota Interactions and Their Role in Plant Invasions ...... 3 1.3 Facilitation of Invasion by Pathogen Accumulation...... 7 1.4 Pathogen Accumulation in Vincetoxicum rossicum ...... 9 1.5 Aim and Objectives ...... 10 Chapter 2 Response of Native Plants to Vincetoxicum rossicum Rhizosphere Biota 2.1 Abstract ...... 11 2.2 Introduction ...... 13 2.3 Materials and Methods 2.3.1 Collection of Biotic Soil Communities ...... 17 2.3.2 Experimental Design and Growth Conditions ...... 18 2.3.3 Harvest and Response Variables ...... 21 2.3.4 Statistical Analysis...... 22 2.4 Results 2.4.1 Effect of Invasion on Native Plant Species Biomass ...... 26 2.4.2 Analysis of Phylogenetic Signal in Net Biotic Effect ...... 30 2.4.3 Effect of Invasion on Native Plant Reproduction ...... 32 2.4.4 Effect of Invasion on Nodule Formation in Desmodium canadense ...... 33 2.4.5 Effects of Invasion on Incidence of Root Lesions ...... 34 2.4.6 Effects of Invasion on Colonization by Arbuscular Mycorrhizal Fungi ...... 36 2.5 Discussion 2.5.1 Does ‘Enemy of my Enemy’ Influence Vincetoxicum rossicum Invasion? ...... 39 2.5.2 Does Phylogenetic Relatedness Influence Response to Invasion? ...... 44 2.5.3 Conclusions ...... 45

Chapter 3 Response of Native Legumes to Vincetoxicum rossicum Rhizosphere Biota 3.1 Abstract ...... 46 3.2 Introduction ...... 48 3.3 Materials and Methods 3.3.1 Collection of Biotic Soil Communities ...... 52 3.3.2 Experimental Design and Growth Conditions ...... 53 3.3.3 Harvest and Response Variables ...... 56 3.3.4 Statistical Analysis...... 56 3.4 Results 3.4.1 Effect of Invasion on Native Plant Species Biomass ...... 58 3.4.2 Effect of Invasion on Desmodium canadense Reproductive Potential ...... 63 3.4.3 Effect of Invasion on Nodule Formation ...... 65 3.5 Discussion 3.5.1 The Effect of Vincetoxicum rossicum Rhizosphere Biota on Native Legumes Productivity ...... 67 3.5.2 The Role of ‘Enemy of my Enemy’ in Vincetoxicum rossicum Invasion ...... 72 3.5.3 Conclusions ...... 73 CONCLUSION ...... 74 REFERENCE LIST ...... 77 APPENDICES Appendix A Coordinates for Soil Collection Sites...... 93 Appendix B Plant community of Collection Sites ...... 93 Appendix C Nutrient Analysis of Field Soil ...... 97 Appendix D Effect of sterilization on root lesion development ...... 100 Appendix E Effect of serilization on AMF Colonization ...... 103 Appendix F Germination response to V. rossicum rhizosphere biota ...... 105 Appendix G Enemy of my Enemy Literature Review ...... 108 Appendix H Seedling survival ...... 114 Appendix I Effects of Electrical Fire on 2nd (Fabaceae) experiment...... 115 Appendix J Relationship between NBE and Phylogenetic Distance ...... 119

LIST OF TABLES

Table 2.1 A list of speciers used in the study with taxonomic designation, common name and growth form...... 20 Table 2.2 Analysis of variance (ANOVA) statistics showing the effects ‘Species’ and ‘invasion’ on the biomass of native plants and Vincetoxicum rossicum produced in Live, Sterile treatments. ‘Species x Invasion’ effect of Net Biotic Feedback was calculated for native plants only; V. rossicum was excluded from net biotic effect,b ...... 26 Table 2.3 Independent t-tests statistics showig the effect of ‘invasion’ on the net biotic effect on plant biomass for each species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and Vincetoxicum rossicum). Analysis conducted on natrual log transformed data...... 29 Table 2.4 Phylogenetic signal measured for average ‘Net Biotic Effect’ on plant biomass in response to ‘invasion’...... 31 Table 2.5 Analysis of variance (ANOVA) statistics showing the effect of ‘Invasion’ and ‘Plant Species’ in live treatments on percentage occurance of decay and herviroy-type lesions in roots of native species...... 35 Table 2.6 Independent t-tests statistics assuming unequal variances showing the effect of Vincetoxicum rossicum invasion history (invaded vs. uninvaded) on lesion formation (by decay and herbivory) for each plant species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and V. rossicum)...... 35 Table 2.7 Analysis of variance (ANOVA) statistics showing the effect ‘Invasion’ (invaded vs. uninvaded) and ‘Plant Species’ (N=5) on the percent colonization by Arbuscular mycorrhizal fungi (arbuscules (AC), vesicles (VC), and hyphae (HC)) in roots of native species. Analysis was conducted on arcsine transformed percent colonization data from live treatments...... 36 Table 2.8 Independent t-tests statistics on arcsine transformed data assuming unequal variances showing the effect of Vincetoxicum rossicum invasion history (invaded vs. uninvaded) on vesicle and hyphal colonization for each plant species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and V. rossicum)...... 38 Table 3.1 A list of speciers used in the study with taxonomic designation, common name and abbreviation ...... 54 Table 3.2 Analysis of variance (ANOVA) showing the effect of ‘Species’ and ‘invasion’ on the plant biomass of native legumes and Vincetoxicum rossicum produced in Live and Sterile treatments. Species x Invasion effects from Net Biotic Effect (NBE) was calculated for native legumes only; V. rossicum was excluded from NBE...... 58

vii

Table 3.3 One sample t-test showing wether the Net Biotic Effect (NBE) of rhizosphere biota on total, shoot and root biomass deviated from zero. Positive t-values indicate biomass promotion by NBE, negative t-values indicate biomass inhbition by NBE, values nearest zero indicate nutral effects. P values <0.05 show significant differences from zero (i.e. rhizosphere biota did not significantly affect plant productivity). Analysis conducted on the combined NBE of rhizosphere biota from invaded and uninvaded soils (log transformed). .61 Table 3.4 Independent t-tests statistics showig the effect of ‘invasion’ on the net biotic effect on plant biomass for each species (legumes: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa), and the invasive vine: Vincetoxicum rossicum). Analysis conducted on natrual log transformed data...... 62 Table 3.5 Independent t-tests showing the effect of ‘Invasion’ (rhizosphere biota from Vincetoxicum rossicum invaded vs. uninvaded soils) on the reproductive potential of Desmodium canadense including bud, flower, seed, seed weight and all reproductive structures (combined bud, flower and seed)...... 64 Table 3.6 Analysis of variance (ANOVA) statistics showing the effect of ‘Invasion’ and ‘ Species’ on nodule formation in native legumes with rhizosphere biota (live) compared to the control (sterile/without biota). Analysis was conducted on square root transformed (0.1+count) data...... 65 Table 3.7 Independent t-test showing the effect of rhizosphere biota from soils invaded and uninvaded by Vincetoxicum rossicum (effect of ‘invasion’) on nodule formation in four legume speciers. Analysis was conduced on square root (+0.1) transformed data...... 66

viii

LIST OF FIGURES

Figure 2.1 Map of sampling sites used for collection of field soils. Collections sites are indicated by solid triangles. Site names and coordinates are from left to right: Toronto Zoo (43.815721, -79.187969), (Crows Pass Conservation Area; 44.037153 -79.042661), (Orono Crown Lands; 43.978888 -78.64607), and (Rice Lake Conservation Area; 44.083999 - 78.300137). Cities are given as landmarks (circles). Created with SimpleMappr, http://www.simplemappr.net...... 17 Figure 2.2 Phylogenetic tree showing evolutionary relationships between native plants used to test effects of rhizosphere biota from V. rossicum invaded and uninvaded soils. The horizontal lines are branches that represent genetic change; the longer the branch the greater the genetic distance to V. rossiucm. Native species are in labeled in black. Vincetoxicum rossicum is labled in red. Additional information of order and broad group is given on the right. Constructed using phyloT: Phylogenetic Tree Generator, 2016. http://phylot.biobyte.de/...... 20 Figure 2.3 A phylogenetic tree for the plant species tested in the present study based on Brownian motion of evolution. The horizonal lines are branches that represent evolutionary lineages changing through time. Every ~1cm of branch length is representative of 10 units of genetic change (see top left side of image: Tree Scale); the longer the branch, the larger the amount of change. Nodes are intersections of branches and represent genetic divergences from a common ancestor. The species are seporated into groups of increasing relatedness. Node tips are the last node along the branch terminating in the plant species name. The numbers above each branch are the numberical quantification of branch length based nucleotide substitutions per site which is the number of changes or 'substitutions' divided by the length of the genetic sequence. Thus, the numbers are representative of genetic variation. The shared branch length is equivelent to covariation between species. For example V. rossicum and A. gerardii do not share branch lengths (covariation = 0). Conversely, V. rossicum and R. hirta share the branch lengths 38 and 22.2 (covariation = (38+22.2)= 60.2). Thus, the closer the relationship between species, the greater the covariation value. Total distance from root to tip is the varience (in the tree above =161). This information on varience-covarience can then be used to calculate Pagel’s Lambda...... 24 Figure 2.4 The total biomass produced by five native herbs when grown with and without rhizosphere biota (Live vs. Sterile treatments) from soils invaded and uninvaded by Vincetoxicum rossicum. Bars represent mean total biomass (± SE). Species abbreviations are: Andropogon gerardii (AnGe), Desmodium canadense (DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) and Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRo)...... 26

Figure 2.5 The net biotic effect (NBE) of rhizosphere biota on plant total (a), shoot (b) and root (c) biomass from soils invaded (dark gray) and uninvaded (light gray) by V. orssicum. Bars represent mean (±SE) NBE; positive values indicate biomass promotion by rhizosphere biota, and negative values indicate biomass inhibition. Species abbreviations are: Andropogon gerardii (AnGe), Desmodium canadense( DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) and Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRo). Asterisks show significant (P<0.05) from independent t-test results for ‘Invasion” on natrual log transformed data: ***P<0.001, **P=0.01, * P=0.05...... 28 Figure 2.6 The quantity of reproductive structures produced by (a) Desmodium canadense (DeCa) and (b) Rudbeckia hirta (RuHi) when grown with and withouts soil biota (live vs. sterile) from soils invaded (dark gray) and uinvaded (light gray) by Vincetoxicum rossicum. Bars represent mean (± SE) number of reproductive structures (peduncles by D. canadense and flowers by R. hirta). Significance was determined using t-test analysis performed on square root transformed data (+0.1). Asterisks indicate significant differences between Live and sterile (long horrizontal lines) and invaded and uninvaded (short horozontal lines) at P<0.05 (*** P<0.001, ** P=0.01, * P=0.05)...... 32 Figure 2.7 The number of root nodules produced by Desmodium canadense in response to rhizosphere biota from soils invaded (dark gray) and uninvaded (light gray) by Vincetoxicum rossicum. Live treatments are compared to sterile controls (no nodules were produced in sterile-uninvaded treatments). Bars represent mean number of nodules (± SE) per plant. Differences between live and sterile treatments (long horizontal line) and invaded and uninvaded treatments (short horizontal line) were determined using independent t-tests on square root transformed data (+0.01). Asterisks indicate significant differences between invaded and uninvaded treatments at P<0.05 (***P<0.01, **P=0.01, * P=0.05)...... 33 Figure 2.8 The amount of root decay (a) and herbivory-type (b) lesions on native (compared to Vincetoxicum rossicum) plants grown with rhizosphere biota from soils invaded (dark gray) and uninvaded (light gray) treatment. Bars represent mean percent root lesions (± SE) per plant. Species abbreviations are: Andropogon gerardii (AnGe), Desmodium canadense (DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRo). Independent t-tests were used to detect differences based on the invasion history of rhizosphere biota (non were detected)...... 34 Figure 2.9 The percent arbuscular mycorhizal fungal colonization of five native plants and Vincetoxicum rossicum grown with rhizosphere communities from soils invaded (dark gray) and uninvaded (light gray) by V. rossicum. Bars represent mean percent root colonization (± SE) of hyphae (a), vesicles (b) and arbuscules (c) per plant. Species abbreviations are: Andropogon gerardii (AnGe), Desmodium canadense (DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRo). Asterisks show significant (P<0.05) from independent t-test results for ‘Invasion” on arcsine transformed data: ***P<0.001, **P=0.01, * P=0.05...... 37

Figure 3.1 Map of sampling sites used for collection of field soils. Collections sites are indicated by solid triangles. Site names and coordinates are from left to right: Toronto Zoo (43.815721, -79.187969), (Crows Pass Conservation Area; 44.037153 -79.042661), (Orono Crown Lands; 43.978888 -78.64607), and (Rice Lake Conservation Area; 44.083999 - 78.300137). Cities are given as landmarks (circles). Created with SimpleMappr, http://www.simplemappr.net...... 52 Figure 3.2 Phylogenetic tree showing evolutionary relationships between native legumes used to test effects of rhizosphere biota from V. rossicum invaded and uninvaded soils. The horizontal lines are branches that represent genetic change; the longer the branch the greater the genetic distance to V. rossiucm. Native species are in labeled in black. Vincetoxicum rossicum is labled in red. Additional information of order and broad group is given on the right. Constructed using phyloT: Phylogenetic Tree Generator, 2016. http://phylot.biobyte.de/...... 54 Figure 3.3 The total biomass produced by four native legumes when grown with and without rhizosphere biota (Live vs. Sterile treatments) from soils invaded and uninvaded by Vincetoxicum rossicum. Bars represent mean total biomass (± SE). Species abbreviations are: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa)...... 58 Figure 3.4 The net biotic effect (NBE) of rhizosphere biota on legume total (a), shoot (b) and root (c) biomass from soils invaded (dark gray) and uninvaded (light gray) by V. orssicum. Bars represent mean (±SE) NBE; positive values indicate biomass promotion by rhizosphere biota, and negative values indicate biomass inhibition. Species abbreviations are: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa). Asterisks show significant (P<0.05) from independent t-test results for ‘Invasion” on natrual log transformed data: ***P<0.001, **P=0.01, * P=0.05...... 60 Figure 3.5 The total number of reproductive structures produced by Desmodium canadense plants when growth with and without soil biota (live vs. sterile) from soils invaded (dark gray) and uinvaded (light gray) by V. rossicum. Bars represent mean (± SE) number of total reproductive structures (combined buds, flowers and seeds) per plant. Independent t-tests were used to detect significant differencences between treatments. Analysis was performed on square root transformed data (+0.1). Asterisks indicate significant differences between Live and sterile (long horrizontal lines) at P<0.05 (*** P<0.001, ** P=0.01, * P=0.05)...... 63 Figure 3.6 The number of buds, flowers and seeds produced by Desmodium canadense in response to rhizosphere biota (i.e. live treatments) from Vincetoxicum rossicum invaded and uninvaded soils. Bars represent mean (± SE) number of buds, flowers and seeds produced per plant...... 64 Figure 3.7 The weight of seeds produced by Desmodium canadense in response to rhizosphere biota (i.e. live treatments) from Vincetoxicum rossicum invaded and uninvaded soils. Bars represent mean seed weight (± SE) per plant...... 64

Figure 3.8 The number of root nodules produced by native legumes in response to rhizosphere biota from soils invaded (dark gray) and uninvaded (light gray) by V. rossicum. Species abbriviations are: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa). Bars represent mean number of nodules (± SE) per plant. Independent t-tests on square root transformed data (+0.01) were used to detect differences between invaded and uninvaded treatements with live rhizosphere biota. Asterisks indicate significant differences between invaded and uninvaded treatments at P<0.05 (***P<0.01, **P=0.01, * P=0.05)...... 66

Chapter 1

Introduction

1.1 Invasion Biology

Biologically speaking, the ultimate goal of any organism is reproduction; passing its genetic heritage from one generation to the next. Thus, mechanisms have evolved to optimize fitness. One such mechanism is dispersal, which allows a species to overcome obstacles such as enemies or resource competition. By chance or design, humans have assisted in this process, moving plants and animals and other organisms around the world since time immemorial (e.g., Gunn et al., 2011). Following the industrial revolution, the frequency and distance at which humans regularly travel (thus spreading non-native species), has increased exponentially. New trade partnerships continue to be formed, international travel continues to increase and the number of introduced species continues to rise (Bradley et al., 2012; Early et al., 2016; Ricciardi et al., 2017).

Invasive species are defined as plants, fungus, or animals that experience rapid population expansion with negative effects on the environment (CBD, 2000; IUCN, 2000) regardless of their environmental effects (Blackburn et al., 2011; Richardson et al., 2000; Richardson, Pyšek, & Carlton, 2011). While not every non-native species becomes invasive and not every native species is benign; non-native species are deemed to be of greater concern. Simberloff et al. (2012) calculated that exotic species are as much as 40% more likely to become invasive. Plants represent a large portion of invasive non-native species (INNS) and they have profound effects on ecosystems (reviewed by: Fridley 2011; Simberloff 2011). Most significantly, INNS threaten native species biodiversity and abundance (Vilà et al., 2011). Reductions in biodiversity can cause top-down or bottom up disruptions in ecosystem functioning and stability. For example, if a keystone species is removed from an ecosystem, a cascade of effects down the trophic chain can occur. When wolves were hunted to extirpation in Yellowstone National Park, WI, USA, it altered the course of rivers and disrupted fish populations (Ripple & Beschta, 2012).

Wolves keep the elk population in check and so without them, elk can become overabundant and are also able remain in favored but dangerous habitats for longer periods of time. For example, places like the waters edge that have abundant brush for elk to feed but which are open and easy for wolves to hunt). When wolves were reintroduced, the overabundant elk population were forced to spend less time near rivers. As a result, the abundance of trees along rivers increased. In turn, beavers had more plentiful resources to build dams. The increase in the number of dams altered the river flow and increased fish spawning habitat (Ripple & Beschta, 2012). The Yellowstone story demonstrates that the removal of just one species from an ecosystem can have profound effects with causes that may not be immediately evident. That is until interactions between organisms and the biotic and abiotic environment are considered.

Some invasions do not result in outright extinctions but rather reductions in species abundance (reviewed by Gurevitch & Padilla, 2004). A meta-population analysis by Gilbert & Levine (2013) showed INN plant species can reduce abundance of native counterparts and drive them to “refugia” (small islands of natives in a sea of INNS). Although this does not reduce biodiversity immediatelly, it creates extinction debts as the populations in these refugia do not have sufficient genetic diversity to make them viable. In addition, should the newly dominant plants have a different set of characteristics (e.g., stems more susceptible to fire), it could lead to changes in disturbance regime, hydrology and complementary among native species (Gilbert & Levine, 2013; Simberloff, 2005; Vitousek et al., 2011). In turn, ecosystem services could be affected. Ultimately, INNS endanger the global environment, the economy and human health (reviewed by Hooper et al., 2005; Isbell et al., 2017; Vitousek et al., 1997).

Currently, 1/6th of Earth’s land surface is under significant pressure by INNS, including 16% of the global biodiversity hotspots (Early et al., 2016). Many of these places are in poorly developed countries which are the least able to respond. As it stands, INNS are considered a global threat, comparable in scale to climate change or a natural disaster (Early et al., 2016; Ricciardi et al., 2011) and are the main contributors to “the sixth mass extinction in Earth’s history” (Isbell et al., 2017).

Understanding the mechanisms of biological invasions is therefore an important part of the larger movement towards the sustainability of human society. Researching these mechanisms will allow us to better predict the long-term effects of biological invasions and identify the most vulnerable ecosystems which would allow for timely and efficient mitigation (van der Putten et al., 2013).

There are many hypotheses on the mechanisms of invasion (summarized by Catford et al. 2009). However, the consensus is that biological invasions result from context-dependent interactions between three main factors: 1) propagule pressure, 2) the invader’s traits and those of the invaded community and 3) the abiotic characteristics of the invaded range (Catford et al., 2009).

This thesis focuses on the effects invasive plants have on rhizosphere biota (biota that live in and around plant roots), and how these interactions affect native plant species within the invaded community. These interactions are particularly difficult to study. Plants can alter the soil chemically (e.g., changing pH), physically (e.g., changing soil aggregation properties), and biologically - affecting species assemblages in soil communities (reviewed by van der Putten et al., 2013). Interactions between plants and specific groups of rhizosphere biota (including species of bacterial and fungal endophytes, archaea, oomycetes, amoebae and nematodes) are thought to have one of the most significant roles in plant community dynamics – particularly during plant invasions (reviewed by Bardgett et al., 2005; Bardgett & Wardle, 2010; Pieterse et al., 2016; Rasmann & Turlings, 2016; Reinhart & Callaway, 2006; Wagg et al., 2014).

1.2 Plant-Biota Interactions and Their Role in Plant Invasions

Interactions between plants and associated rhizosphere biota involve detection, communication and control of each other’s activity. Plants can detect the presence of soil organisms through various detection mechanisms. For example, sensing the presence of certain structures in cellular membranes (e.g., chitin) helps the plant identify the type of organism present (reviewed by Shinya et al., 2015). Once recognized, the plant can then deploy the appropriate response. For example, plants can prevent or minimize pathogen attack by producing anti-microbial metabolites, and by undergoing physiological changes such as thickening their cell walls (mechanisms reviewed by Yamazaki & Hayashi, 2015).

Plants release a variety of organic compounds into the surrounding soils including sugars, amino acids, plant hormones and other primary and secondary metabolites (reviewed by Antunes & Koyama, 2017; Bais et al., 2006; Pieterse et al., 2016; Rasmann & Turlings, 2016). Although many details remain unclear, variations within these compounds are thought to act as chemical signalling mechanisms, carrying information about plant status and condition. Through these signals, plants control organisms in their rhizosphere, encouraging or discouraging

interactions with specific taxa. Recently, Corral-Lugo et al. (2016) showed that rosmarinic acid, produced by plants mimic quorum-sensing compounds. These compounds are produced by bacteria to signal aggregation (formation of biofilm) or dispersal. By varying the concentration and types of metabolite excretions, plants can inhibit parasitism or pathogenesis and promote mutualisms (Corral-Lugo et al., 2016; Gao, Teplitski et al., 2003; Pérez-Montaño et al., 2013).

Rhizosphere biota can also sense, respond and communicate with plants (Buer et al. 2010; Shinya et al., 2015; Turrà et al., 2015; Yamazaki & Hayashi, 2015). For example, effector proteins that are released by arbuscular mycorrhizal fungi (AMF) can alter gene expression in host plants, resulting in inhibition of plant defence hormones and initiation of symbiosis (Plett & Martin, 2015). Similarly, fungal pathogens may use effector proteins to avoid chitin detection mechanisms in the plant – effectively hiding from the potential host (Shinya et al., 2015). Bacteria can directly synthesize plant hormones like auxins (reviewed by Ng et al., 2015) and cytokinin (e.g., Großkinsky et al., 2016), allowing them to manipulate pathways involved in plant development, nodule formation and root branching. As with plants, many of the communication mechanisms of rhizosphere biota have yet to be studied (Plett & Martin, 2015).

Recently, it has been recognized that microbe-microbe interactions are also important for plant productivity. Fungi, for example, have their own set of associated bacterial endophytes, the effects of which are just beginning to be studied. Recently, Hernández-Montiel et al. (2013) found that synergistic interactions between AMF and rhizosphere bacteria conferred pathogen protection to the host plants.

Generally, plant associations with rhizosphere biota are broadly divided in two groups: mutualistic and antagonistic (parasitic or pathogenic). On one hand, mutualistic relationships benefit both partners. In exchange for carbon, these organisms provide host plants with access to additional nutrient pools, as well as improving host drought tolerance, and increasing pathogen protection, ultimately bettering overall host fitness (reviewed by Bardgett et al., 2005). On the other hand, antagonistic relationships result when organisms extract phyto-carbon without reciprocal exchange (parasites), and in some cases cause disease (pathogens), diminishing host growth and/or fitness.

The organisms that comprise plant-associated biota are a phylogenetically diverse group, but most are contained within specific taxa. For example, 96% of plant colonizing bacteria fall within four phyla (Proteobacteria, Actinobacteria, Firmicutes and Bacteroidetes) (Beattie, 2006; Hardoim et al., 2015). Similarly, plant associated fungi belong mostly to the Glomeromycotina,

Ascomycota, and Basidiomycota (Hardoim et al., 2015). Within plant tissues, oomycetes, viruses and micro-invertebrates such as nematodes can also be found (Faillace et al., 2017; Jones et al., 2013). Each of these groups (except subphylum subphylum Glomeromycotina; Spatafora et al., 2016) contain both well-known plant pathogens (see top ten pathogens: Dean et al., 2012; Kamoun et al., 2015; Mansfield et al., 2012; Scholthof et al., 2011), and plant mutualists (e.g., the polyphyogenetic group called rhizobia which colonize root nodules and fix nitrogen for plant use). Only, Glomeromycotina is exclusively comprised of biota which form a common functional group, specifically AMF which are well known plant mutualists. Thus, the function of soil biota within their host plants cannot be easily predicted from their taxonomic identity alone; function is determined by host x biota genotype x environmental parameters interactions (Bardgett et al., 2005; Bardgett & Wardle, 2010; Johnson & Graham, 2012; Pieterse et al., 2016; van der Putten et al., 2013).

Mutualistic and antagonistic interactions occur simultaneously, and it is clear from what little we know, that interactions between plants and rhizosphere biota are incredibly intricate. As such, plant-rhizosphere biota interactions are best considered along a continuum from mutualistic to parasitic/pathogenic (Aguilar-Trigueros et al., 2014; Faillace et al., 2017; Johnson & Graham, 2012; Johnson et al., 1997; Jones & Smith, 2004; Kobayashi & Crouch, 2009; Veneault-Fourrey et al., 2013). Klironomos (2003) tested 64 plant species from an old-field community and found that plant growth responses to AMF varied from strongly positive to strongly negative, depending on host-AMF species combinations. Thus, although AMF are generally considered plant mutualists, they may act as parasites in some situations (e.g., certain AMF-plant species combinations).

Scale-of-observation is also important for determining the overall effect of the interaction between plant and soil biota (Carroll, 1988; Faillace et al., 2017; Kobayashi & Crouch, 2009). Most of our understanding of such interactions come from greenhouse studies and from a limited number of hosts and associated biota. How these interactions play out in nature can be quite different from the individual effects observed in the greenhouse. For example: a virus might be antagonistic to a specific host in the greenhouse, but this effect may be relatively small in comparison to effects of the virus among other members of the plant community (reviewed by Faillace et al., 2017).

These interactions between plants and rhizosphere biota lead to a continuous cycle of actions and reactions between plant and soil community (collectively termed “plant soil feedback”; PSF).

As plant-rhizosphere biota interactions affect plant growth and fitness, it results in changing the soil biotic community, which then affects the host and alters the host plant community. Since plants also affect soil nutrients, pH, and chemical composition (e.g., through the release of allelochemicals). Plant-Soil feedback is also used to describe these cyclical interactions between plant and soil abiotic environment. This thesis focuses specifically on “biological” PSF.

Plant soil feedback can be a positive or negative interaction. Positive feedback occurs when interactions between plants and rhizosphere biota promote growth of the host plant or plant species whereas negative feedback occurs when these interactions inhibit host growth (Bardgett & Wardle, 2010; Reinhart & Callaway, 2006; van der Putten et al., 2013).

Mutualistic and antagonistic interactions, resulting in positive and negative feedback, occur simultaneously; hence the net outcome of these interactions determine plant community composition. Positive feedback is driven by mutualistic associations (e.g., Bachelot et al., 2015; Wehner et al., 2011) and is generally thought to facilitate the maintenance of its host species in the community (Bardgett et al., 2005; Reinhart & Callaway, 2006). Thus, positive feedback has been suggested as one of the main mechanisms driving plant invasions (Klironomos, 2002). For example: in a recent study on rhizosphere fungi on native and introduced maple trees in north America, introduced maples were shown to associate with a greater diversity of mutualistic fungi than native maples (Toole et al., 2017). These fungi included nearly all the mutualist fungi found on native maples plus additional mutualist fungi. If an introduced plant is benefiting more from mutualist biota than the native plant species in the community, there is a good chance that it could gain the competitive advantage. This relationship could lead to the replacement of native species, which could lead to changes hydrological and disturbance regimes – a major concern associated with invasive species (Gilbert & Levine, 2013; Simberloff, 2005; Vitousek et al., 2011).

In contrast, negative feedback is driven by accumulation of parasitic/pathogenic organisms. It is the dominant interaction in native plant communities and is thought to limit the abundance of the host species and promote species diversity (Bever et al., 1997; Klironomos, 2002; Reinhart & Callaway, 2006). As such, it was thought that over time, invasive species would accumulate antagonistic rhizosphere biota (including viruses) as native biota adapted to the presence of the new host, ultimately bringing invasive species’ populations under control. The system would thus switch back to a negative-feedback system; assumed to be dominant among native plant communities (Flory & Clay, 2013; Lankau et al., 2009). However, despite common plant

pathogens being found on the roots of invasive species, this does not always result in invasive population declines (Day, Antunes, & Dunfield, 2015; Flory & Clay, 2013). Therefore, it is necessary to investigate the wide range of effects parasites and pathogens have on invasive plants.

1.3 Facilitation of Invasion by Pathogen Accumulation

The role that rhizosphere enemies play in plant invasions is not well understood. When a plant is introduced, it is either introduced with or without associated soil biota. This exotic biota includes mutualists and/or antagonistic organisms (such as parasites and pathogens) (henceforth enemies). Upon arrival, introduced plants also encounter a new suite of mutualists and enemies. Whether the introduced plant becomes invasive depends largely on the interactions between the introduced plant and its own mutualists and enemies (if they survived introduction), native mutualists and enemies and the relative effects of mutualists and/or enemies (regardless of their biogeographic origin) on the native plant population (reviewed by (van der Putten et al., 2013)

One of the leading hypotheses is the enemy release hypothesis (ERH) which states that invasive species are competitively superior in their introduced range because they have left their co-evolved enemies behind (Keane & Crawley, 2002). Unlike native plants, which are still supressed by their enemies, invasive plants have escaped theirs and are able to redirect resources to growth and reproduction; ultimately gaining competitive advantage over native plants (Colautti et al., 2004; Keane & Crawley, 2002). Empirical support for the ERH is split however, particularly with respect to rhizosphere biota (Jeschke et al., 2012). A combination of other mechanisms of invasion from the perspective of belowground plant-soil feedback are likely (reviewed by (Colautti et al., 2004; Flory & Clay, 2013; Inderjit & van der Putten, 2010; Reinhart & Callaway, 2006)

One hypothesis that has received little attention is the “enemy of my enemy is my friend” hypothesis (Enemy of my Enemy or EEH, for short). It states that a ‘facilitation of invasions through the amplification of enemies in a host population that serves as a reservoir for further infection of other species in the community’ (Colautti et al., 2004). The enemies may be either co-introduced with the invader or originate from the native soil community. Introduced biota may ‘spillover’ on to native plants on which they have greater negative effects (i.e., cause negative

feedback). This increase in negative feedback is because the invader will have co-evolved with the introduced rhizosphere biota and have evolved sufficient defence mechanisms. In contrast the native plants will not have mechanisms to ward off the introduced biota. Alternatively, the invader may also be infected by generalist enemies, already present in the native community, that can thrive along with the invader’s population. They can function as commensals or even mutualists in the invader but can spillback into incoming native plants thereby promoting invasion through facilitation (apparent competition). In this respect, the EEH is similar, if not identical to other independently developed hypotheses, specifically the local pathogen accumulation hypothesis (Eppinga et al., 2006) and the Spillover (or Spillback) hypothesis (Mangla et al., 2008); ‘Spillover’ referring to the accumulation of non-native parasites/pathogens; most likely coevolved with the invader-host, and ‘Spillback’ referring to the specific accumulation of native parasites/pathogens on an invader-host (Flory & Clay, 2013; Strauss et al., 2012).

One of the main reasons why so many hypotheses have been developed around the concept of ‘facilitation of invasions through the amplification of enemies of native plants’, is likely because the geographic origin of the enemy is difficult to determine, or not even considered. Thus, some authors use the terms generally. For example: the term ‘Spillover’, which has been adapted from human and animal pathology (Kelly et al., 2009), specifically refers to enemies that have been co-introduced with the invader. Despite this definition, ‘Spillover’ has been used by some researchers to describe the entire phenomenon (e.g., Crocker et al., 2015; Power & Mitchell, 2004). To avoid confusion, the EEH is used as a general term for the phenomenon, which encompasses both introduced and native enemies.

Few studies have been published on EEH in non-native plant invasions (see Appendix G). Of those that have been published, most have found that 1) parasites/pathogens accumulate on invasive-species, 2) they have relatively little impact on the invader-host (and sometimes act as mutualists promoting the host’s growth), and 3) they inhibit native species fitness (Beckstead et al., 2010; Day et al., 2016; H. Li et al., 2014; Mangla et al., 2008; but see Crocker et al., 2015, 2016; Mordecai, 2013). However, most of these studies have involved examining the EEH between invasive and native grasses (Poaceae: see Appendix G). Only a few studies have tested the effects of invasion on native plants outside the Poaceae family. A literature search using Web of Science and Google Scholar yielded 46 papers that referenced one of four search terms in the context of plant biology (“pathogen spillover”, “pathogen spillback”, “accumulation of local pathogens” and “enemy of my enemy”; see Appendix G). Of the relevant papers 11

involved empirical research on invasive species and only 4/11 of these described experiments involving native plants that did not belong to the Poaceae family (Crocker et al., 2015, 2016; Day et al., 2016; Mangla et al., 2008). A phylogenetically restricted approach makes it difficult to make inferences about the role of EEH has in plant invasions.

1.4 Pathogen Accumulation in Vincetoxicum rossicum

Day et al. (2016) were the first to find evidence of EEH in relation to Vincetoxicum rossicum, a perennial herbaceous small twining vine that forms dense monocultures. Introduced to Eastern North America from the Ukraine between 1848 and 1889 (Sheeley & Raynal, 1996), it can grow in a wide range of habitats and is a target of research and control efforts (e.g. Ontario’s Invading Species Awareness Program, 2017). Vincetoxicum rossicum can easily be propagated from seed or root crown and grows relatively quickly. As such, V. rossicum shows promise as a model organism for the study of EEH from a belowground soil feedback perspective.

Vincetoxicum rossicum has been found to host many fungal plant pathogens (Bongard, Navaranjan et al., 2013; Day et al., 2016). Day et al. (2016) were able to isolate some of these pathogens and infect seedlings of two native species (Solidago canadensis and Asclepias syriaca) and V. rossicum. They found that combinations of two of the isolates inhibited the growth of one of the native species (S. canadensis), while at the same time, promoted V. rossicum. That this inhibitory effect was observed when only multiple pathogen isolates were applied, suggests synergistic interactions within soil microbial communities. Whole communities of rhizosphere biota need to be considered in investigations of EEH. Indeed, many of the previous studies on EEH have been conducted on single species of rhizosphere biota (e.g., Beckstead et al., 2010; H. Li et al., 2014; Mordecai, 2013; Power & Mitchell, 2004). Furthermore, because only one native plant species exhibited a response to potential pathogens isolated from within the roots of V. rossicum suggests EEH might not be a universal mechanism for V. rossicum invasions. Tests of the EEH in plants has been primarily tested between invasive and native plants that are closely related to grasses (i.e., generally in the order Poales; see Appendix G). Focusing on native species along a phylogenetic gradient including different orders would provide information to improve inferences of the ecological effects of EEH in plant invasions.

1.5 Aim and Objectives

The overarching aim of this thesis was to investigate whether the EEH is an important driver of plant invasions. Using V. rossicum as a model organism, EEH was tested by inoculating native plants at increasing phylogenetic distances from V. rossicum with soil biota from invaded populations. This thesis is divided into two main chapters. In the first chapter, a plant-soil feedback study was used to investigate whether rhizosphere biota associated with V. rossicum affected native plant seedling growth. In the second chapter, the EEH was tested using a group of native legumes. This chapter follows from results of the first which indicated strong positive feedback on the native legume Desmodium canadensis.

Chapter 2

Response of Native Plants to Vincetoxicum rossicum Rhizosphere Biota

2.1 Abstract

Interactions between plants and soil biota are gaining recognition as major drivers of plant invasions. ‘Enemy of my enemy’ (EEH) is a major hypothesis in invasion ecology, also known as ‘accumulation of local enemies’ hypothesis. It states that a non-native invader “accumulates generalist pathogens, which limit the invader’s abundance, but limit indigenous competitors more”. One hypothesized mechanism is that invasive plants amplify rhizosphere enemies (e.g., pathogens) by acting as a reservoir that can ‘spill’ onto native plant species, ultimately leading to relatively greater declines of native populations compared to non-native invasive ones. Few empirical studies have tested the EEH in plant invasions, especially on below-ground biotic rhizosphere interactions. Here we used Vincetoxicum rossicum (VIRO), one of the most invasive plant species in Eastern North America, and five, functionaly and phylogenetically diverse co-occuring native plant species, to test whether rhizoshepre biota accumulating in VIRO invaded soil contribute to VIRO invasion. Specifically, we predicted that native species would grow less in presence of soil biota from VIRO-invaded soils compared to that from non- invaded soils. In addition, we predicted that support for the ‘EE’ hypothesis would correlate with phylogenetic distance to VIRO (i.e., native plants closely related to VIRO would grow less in comparison with more distantly related native plants). Field soils were collected from VIRO- invaded and neighboring non-invaded areas. Each of the five native plant species (Andropogon gerardii, Desmodium canadense, Solidago canadensis, Rudbeckia hirta, and Asclepias syriaca) and VIRO were grown in a plant growth chamber for three months, either in presence or absence of soil biota from the invaded and non-invaded soils. We found that the growth of S. canadensis was smaller in response to soil biota from VIRO-invaded compared to non-invaded soils, as predicted by the EEH. However, opposite responses were observed in D. canadense and A. gerardii. The remaining two species (R. hirta and A. syriaca) were not significantly affected by invasion history of the soil biota. In addition, native plant response to soil biota was not reflective of the phylogenetic distances from VIRO. In conclusion, we did not find consistent

support for the ‘EEH’ between VIRO and the native co-occurring species, but rather negative soil feedback effects of VIRO on native plants were species-specific. We concluded that, the accumulation of rhizosphere enemies that can ‘spill’ onto native plants may not be a major factor in the invasive success of VIRO.

2.2 Introduction

Invasive Non-native species (INNS) can interrupt the functioning of the ecosystems they invade. This disruption can have significant negative effects on ecosystem services, including ecosystem stability, pollution filtration, and resource production (Fridley, 2011; Fridley, & Sax, 2014; Gilbert, & Levine, 2013; Simberloff, 2005; Vitousek et al., 2011). Currently it is estimated that 1/6th of Earth’s land surface is under significant pressure by INNS including 16% of the global biodiversity hotspots, many of which are in poorly developed countries whom are the least able to respond (Early et al., 2016).

The most significant threat posed specifically by invasive non-native plants are their negative effects on biodiversity and species abundance (Vilà et al., 2011). Loss of a key stone species or reductions of native species to refugia (islands of native plants surrounded by non-native plant species) can transform entire ecosystems (e.g. Gilbert, & Levine, 2013; Ripple, & Beschta, 2012). For example, should the invading species have a different set of characteristics like stems that are more susceptible to fire, they can significantly change an ecosystem’s hydrology and disturbance regime. As more plant species are reduced or replaced, the interactions between organisms within the invaded landscape change, leading to structural and functional shifts within the ecosystem and further degradation of the environment (Fridley, 2011; Fridley, & Sax, 2014; Gilbert, & Levine, 2013; Simberloff, 2005; Vitousek et al., 2011).

The introduction of non-native plant species has significantly altered the national flora of many; if not all countries. In Canada, for example, roughly one quarter of the national flora are non- native. Among the non-native flora, ~40% are considered weedy or invasive (Canadian Food Inspection Agency, 2008). However, it is likely the effects of invasions have been underestimated, especially if the subtle interactions between plants and rhizosphere biota are considered (Jaric & Cvijanovic, 2012; Simberloff et al., 2013). Recent advancements in molecular tools is allowing for more targeted investigations into the consequences of interactions between plants and their rhizosphere biota (henceforth rhizosphere biota). The relationship between plants and arbuscular mycorrhizal fungi (AMF) has received much attention. However, antagonistic rhizosphere biota (parasites and pathogens) in the context of belowground plant-soil interactions, are not often considered; despite the general assumption that they (along with AMF and mutualistic rhizosphere biota) are important drivers of plant community dynamics (reviewed by Bardgett et al., 2005; Bardgett, & van der Putten, 2014;

Bardgett, & Wardle, 2010; Pieterse et al., 2016; van der Putten et al., 2013; Rasmann, & Turlings, 2016; Reinhart, & Callaway, 2006; Wagg et al., 2014).

Plants and rhizosphere biota can alter their response to each other depending on plant and biota genotype and environmental parameters, (Bardgett et al., 2005; Bardgett & Wardle, 2010; Catford et al., 2009; Johnson & Graham, 2012; Pieterse et al., 2016; van der Putten et al., 2013). Individual species of rhizosphere biota interact with plants along a continuous scale of mutualistic to antagonistic interactions (Faillace et al., 2017; Johnson & Graham, 2012; Johnson et al., 1997; Smith & Smith, 2015; Vandegrift et al., 2015). Mutualistic and antagonistic interactions occur simultaneously as plants engage in a continuous cycle of action and reaction with the rhizobia community (a biological form of Plant-Soil Feedback or PSF) (Klironomos, 2002).

One of the leading hypotheses’ regarding the role antagonistic soil biota have in plant invasions is the ‘Enemy Release Hypothesis’ (ERH) which states that invasive species are competitively superior in their introduced range because they have left their co-evolved enemies behind. Unlike native plants, which are still supressed by their enemies, invasive plants have escaped their own enemies and are able to redirect resources to growth and reproduction; ultimately gaining competitive advantage over native plants (Colautti et al., 2004; Keane & Crawley, 2002; Richardson & Pyšek, 2006). Empirical support for the ERH is split however, particularly with respect to rhizosphere biota (Colautti et al., 2004). Native plant pathogens are known to colonize non-native plants (e.g., Bongard et al., 2013). Over time, it is hypothesised that invasive species should steadily accumulate antagonistic soil biota (including viruses), as local pathogens evolve to evade host defences, and contributing to the invasive species’ decline (Flory & Clay, 2013; Lankau et al., 2009). However, despite common plant pathogens being found among the roots of invasive species, accumulation of pathogens does not always lead to invasive decline (Day, Dunfield, & Antunes, 2015).

An alternative hypothesis to the ERH is the ‘Enemy of my Enemy Hypothesis’ (EEH) which states that ‘plant invasions are facilitated by amplification of enemies in a host population that serves as a reservoir for further infection of other species in the community’ (Colautti et al., 2004). Invasion is facilitated when the growth of the non-native invasive species is not affected by antagonistic soil biota while native species growth is reduced by the effects of negative relationships with soil biota. Thus, the invasive species dominates through a form of apparent

competition whereby the invasive plant benefits from the ill health of the native plant and not through direct competition itself (Power & Mitchell, 2004).

The EEH is also known as the ‘Local Pathogen Accumulation Hypothesis’ (Eppinga et al., 2006), and the ‘‘Spillover’ and Spillback’ hypotheses (Mangla et al., 2008; Power & Mitchell, 2004). ‘Spillover’ is specific to the accumulation of non-native parasites/pathogens; most likely co-evolved with the invader-host, and ‘Spillback’ refers to the specific accumulation of native parasites/pathogens on the invader-host (Flory & Clay, 2013; Strauss et al., 2012). It is not often possible to determine the geographic origin of the rhizosphere biota. Therefore, EEH is used as a collective term for the phenomenon of accumulation of pathogens to facilitate invasion regardless of origin.

Most studies on EEH from a belowground plant-soil feedback study have only considered invasive plants in the Poaceae (Beckstead et al., 2010; Day et al., 2016; H. Li et al., 2014; Mangla et al., 2008; but see Crocker et al. 2015, 2016; Mordecai 2013). Day et al. (2016) were one of the first to find evidence of EEH from an invasive species outside the Poaceae family, specifically Vincetoxicum rossicum (Kleopov) Barbaricz. Vincetoxicum rossicum is a perennial herbaceous small twining vine that forms dense monocultures in its invaded range of Eastern North America. It was introduced from South-Western Russia in the mid 1800s (Sheeley & Raynal, 1996), and can grow in a variety of habitats, from prairie to open forests (DiTommaso et al., 2005; Sanderson et al., 2015). It produces copious amounts of seeds, which are larger and more viable when formed in open areas (Cappuccino et al., 2002; St Denis & Cappuccino, 2004; and reviewed by DiTommaso et al., 2005). Vincetoxicum rossicum can be easily propagated from seed or root crowns. In addition, several studies have shown that V. rossicum may have significant ecological impacts (DiTommaso & Losey, 2003; Ernst & Cappuccino, 2005; Mattila & Otis, 2003). These combined characteristics make V. rossicum a promising as a model organism for the study of EEH.

Vincetoxicum rossicum is known to host many fungal plant pathogens, but their effects are unknown (Bongard et al., 2013; Day et al., 2016). Day et al. (2016) isolated some of these pathogens and infected seedlings of V. rossicum and of two native species (Solidago canadensis and Asclepias syriaca). They found that combinations of two fungal isolates inhibited the growth of one of the native species (S. canadensis) while promoting that of V. rossicum. Since this inhibitory effect was observed when only multiple pathogen isolates were applied it is likely that synergistic interactions between rhizosphere biota within whole soil

communities are needed to be considered in investigations of EEH. The objective of this study was to explore the effects of EEH using rhizosphere communities trained by V. rossicum on a phylogenetic range of native plant species. Evidence suggests that non-native invasive plants are more successful when invading more distantly related native plant populations than when they invade more closely related native populations (Feng & van Kleunen, 2016; Strauss et al., 2006). These observations hold true with the naturalization hypothesis (Darwin, 1859) which predicts that non-native species invading areas with closely related relatives should be less successful because they require the same resources (e.g. nutrients) and share common specialized enemies (e.g. pathogens).

The accumulation of soil dwelling plant pathogens has long been recognized in agroecosystems (i.e. thousands of years) and has some supporting evidence (e.g. Miller & Menalled, 2015). Some experiments on native systems show that soil biota plays a likely role in phylogenetic aligned soil-feedback (Anacker et al., 2014; Brandt et al., 2009; Callaway et al., 2013). Yet, the notion that closely related plants share common enemies because of common physiology is far from consistent across studies. Several high-profile studies looking at plant communities in general (Fitzpatrick et al., 2017; Mehrabi et al., 2015; Mehrabi & Tuck, 2015) and specifically at native and non-native invasive plants (Dostál & Palečková, 2011) show no phylogenetic signal.

The contradictions in the literature may exist because of differences in the scale of observation. Diez et al. (2008) found that whether related plants shared enemies depended on the abundance and number of the related species, and whether the plants were invasive or naturalized non-native plants. In addition, if positive phylogenetically linked plant-soil feedback due to a portion of the rhizosphere biota (i.e. AMF; see Dostál & Palečková, 2011; Reinhart et al., 2012) are not accounted for; a phylogenetic effect might be hidden (Fitzpatrick et al., 2017).

Native seedlings grown with V. rossicum invaded and uninvaded soil with and without rhizosphere biota was used to test two hypotheses: (1) Biota from V. rossicum invaded soil is more detrimental to native plants than that from uninvaded soil areas due to pathogen accumulation; (2) phylogenetic relatedness of native plants to V. rossicum predicts their responses to rhizosphere biota from V. rossicum invaded areas.

2.3 Materials and Methods

2.3.1 Collection of Biotic Soil Communities

Soil was collected from each of four Conservation Areas in Southern Ontario, Canada (Figure 2.1; See Appendix A – Table A1). Sites were first identified using the online mapping system, EDDMaps (http://www.eddmaps.org/ontario). Local conservation authorities were then contacted to confirm that specific V. rossicum invasions were present for at least five years.

Vincetoxicum rossicum shows a high degree of phenotypic plasticity and can invade a range of habitats, from closed-canopy forest to open-meadow/grasslands (DiTommaso et al., 2005). Nevertheless, seedling success and propagation vary across habitat types. Vincetoxicum rossicum populations were selected from grassland habitats where they are most successful (Cappuccino, 2004; Cappuccino et al., 2002; St Denis & Cappuccino, 2004).

Collection sites were all within the Mixed-Wood Plains Ecozone, Ecoregion 6E (Crins et al., 2009) on calcareous parent material (Middle-Upper Ordovician limestone/dolostone/shale)

(Ontario Geological Survey, 1991). All sites had similar plant communities (see Appendix B) and soil texture (i.e., sandy clay loam; except Rice Lake which was a loam) (see Appendix C).

In August 2015, soil was collected from the centre of dense V. rossicum populations to ensure broad representation of rhizosphere communities affected by this species (i.e. “Invaded” soil communities). Likewise, soil was collected from plant communities adjacent to the V. rossicum populations (i.e. “Uninvaded” soil communities). Selected invaded and uninvaded plant communities were located adjacent to one another to isolate the effects of V. rossicum on rhizosphere biota while controlling for potential differences in abiotic conditions. Indeed, soil fertility between invaded and uninvaded populations were similar across sites (see Appendix C - Table C.2 & Table C.3).

Invaded (>5m2) and adjacent uninvaded populations were identified as having 75-100%, and <5% V. rossicum abundance, respectively (see Appendix B for approximate community composition of invaded and uninvaded populations based on cover abundance). Soil collected from the uninvaded plant communities was at least 2-4 m from the edge of a V. rossicum population and at least 1m from any adult V. rossicum individual.

To obtain soil biotic communities specific to the rhizobiome of V. rossicum, soil from the root zone (upper 30 cm) of 5-8 randomly selected adult V. rossicum plants was collected from each of the four populations. In addition, soil cores of equivalent size (20 cm wide x 30 cm deep) were collected from adjacent uninvaded plant communities. Collection tools were sterilized with 10% bleach to avoid cross-contamination between invaded and uninvaded soils. Soils were placed separately into coolers for transportation to Algoma University (1-2 days). Soils were sieved (4-mm mesh) to remove roots, rocks and macroinvertebrates. Roots were collected from invaded and uninvaded soils and cut into 1cm sections for use in the experiment (see below). Soils and root material across sites were pooled according to invasion history and stored at 4oC for eight weeks prior to the onset of the experiment. Although pooling soil among sites may overestimate local effects of rhizosphere biota (Reinhart & Rinella, 2016), since the objective of this study was to test for a potential biotic effect of V. rossicum on native plants via changes in rhizosphere biota communities, having wide representation of the rhizosphere biota present in soils either invaded by V. rossicum or yet to be invaded was important. Therefore, pooling soil in this manner is appropriate (see Karst et al. 2015; Cahill et al. 2016; but also see Reinhart & Rinella 2016).

2.3.2 Experimental Design and Growth Conditions

The experiment consisted of a completely randomized fully crossed design with three main factors and six replicates per treatment (N=6). The three main factors were ‘Rhizosphere biota’, 'Invasion' and 'Plant Species'. ‘Rhizosphere biota’ effects were tested using live and sterile treatments, ‘Invasion’, invaded and uninvaded treatments, and ‘Plant Species’ by using five native species and V. rossicum, for a total of 144 experimental units.

Treatments consisted of a 2:1 mixture of Turface (a montmorillonite clay, Turface Athletics MVP, Profile Products LLC, Buffalo Grove, IL, USA) and non-calcareous granitic Sand (Hutcheson Sand and Mixes, Huntsville, ON), combined with 15% (by volume) of either ‘invaded’ or ‘uninvaded’ field soil (i.e., inoculum). Sterile Inoculum consisted of sub-samples of 'invaded' and 'uninvaded' field soils and root material that had been autoclaved twice (one-hour cycle at 121° C with), and stored at 4oC until the beginning of the experiment. Additionally, three grams of root material were added to each respective treatment as an added source of inoculum for rhizosphere biota not able to survive without a plant host (J. D. Bradley, 2013; Day, Dunfield, et al., 2015). Specifically, sterile-roots from uninvaded field soil were added to sterile-uninvaded treatments, live-uninvaded roots were added to live-uninvaded treatments, and like-wise for live and sterile invaded treatments.

Round Standard Pots (6”, 1.5L; Stuewe and Sons Inc., Corvallis, OR, USA) were filled with 1.5L of the respective experimental substrates. Pots were placed on individual saucers to minimize cross contamination and nutrient losses through leaching. Care was taken to sterilize all pots and utensils with 10% bleach, to prevent contamination between treatments.

The plant species used in this study included five native species commonly found co-occurring in Southern Ontario's grasslands (Desmodium canadense, Solodago canadensis, Rudbeckia hirta, Andropogon gerardii and Asclepias syriaca) and V. rossicum (Table 2.1). Native plant species were selected from a short-list created by cross-referencing the Southern Ontario Vascular Plants List (Bradley 2013) with lists of Ontario Tallgrass Prarie and Local Toronto Plants (Limestone Creek Restoration Nursary, 1996; Tall Grass Ontario, n.d.). Only perennial species that appeared on all three lists were selected. Preference was then given to plant species present in previous V. rossicum community surveys (Day 2015). Within the short-list, native species were selected along a gradient of relatedness to V. rossicum (Figure 2.2).

Table 2.1 A list of speciers used in the study with taxonomic designation, common name and growth form.

Order Family Species name Abbreviation Common name Poales Poaceae Andropogon gerardii Vitman AnGe Big bluestem Fabales Fabaceae Desmodium canadense (L.) de Candolle DeCa Showy tick-trefoil Astrales Asteraceae Rudbeckia hirta L. RuHi Black-eyed Susan Astrales Asteraceae Solidago Canadensis L. SoCa Canada goldenrod Gentianales Apocynaceae Asclepias syriaca L. AsSy Common milkweed Gentianales Apocynaceae Vincetoxicum rossicum (Klepow) Barbaricz ViRo Dog-strangling vine

Poales (Monocot)Fabales (Rosids) Astrales (Asterids I) Gentinales (Astrids II ) Figure 2.2 Phylogenetic tree showing evolutionary relationships between native plants used to test effects of rhizosphere biota from V. rossicum invaded and uninvaded soils. The horizontal lines are branches that represent genetic change; the longer the branch the greater the genetic distance to V. rossiucm. Native species are in labeled in black. Vincetoxicum rossicum is labled in red. Additional information of order and broad group is given on the right. Constructed using phyloT: Phylogenetic Tree Generator, 2016. http://phylot.biobyte.de/.

Seeds were acquired from commercial sources (Ontario Seed Co., Limited, Waterloo, Ontario) and were representative of US Midwest genetic stock (OSC pers. comm.). In addition, seeds of V. rossicum were field-collected on August 6th, 2015 from Crows Pass Conservation Area.

Seeds were surface sterilized in 10% bleach for 3-5min. and thoroughly rinsed in deionized water for 10 minutes. Smaller seeds (S. canadensis and R. hirta) were more sensitive to the sterilization process (data not shown) and were bleached for one minute. Seeds were cold-wet stratified for 1 week at 4oC to break dormancy. They were then germinated for 1-2 weeks at 23oC between moistened paper towels in sealed germination trays to maintain a high humidity. Four seedlings were planted into each pot and thinned to one individual after one week of growth. Seedling survival was maximized by replacing unsuccessful seedlings with excess seedlings, maintained in the original germination trays, in the first week of growth.

Plants were grown during the fall and early winter of 2015/16 in a walk-in plant growth-room at Ontario Forest Research Institute, Ontario, Canada. The chamber was set to 65% humidity,

21oC/18oC day/night temp with 15 hr day length, at 200-250 mol light intensity. Plants were watered to field capacity every 2-3 days (200-300 ml).

Mild signs of nutrient stress were observed at 3wks of growth. After which, all pots were fertilized with 50mL (0.48g/L) of Miracle-Gro 24:8:16 (The Scotts Company LLC, Mississauga, ON, Canada) every two weeks. By the end of the experimental period each pot received a total of 25mg N, 3.6mg P, and 13.8mg K. Studies have shown that as soil fertility increases feedback with mutualists while pathogens declines (see model outlined in Smith-Ramesh & Reynolds 2017). Because this experiment focused on plant-soil biotic interactions, only the minimum amount of fertilizer required to support plant growth was applied.

2.3.3 Harvest and Response Variables Plants were harvested after 13wk, as determined by the first species to enter senescence (R. hirta). Roots and shoots were separated at the soil surface. Roots were gently removed from the experimental substrate and washed in cool water to remove excess substrate. Root samples were taken for future molecular analysis of the soil microbial community (analysis not included). Roots were patted dry and weighed before and after samples were taken. Shoots and roots were placed in separate bags, dried for three days at 60°C and weighed to nearest 0.001g. Percent moisture of oven-dried root was used to calculate the total dry root weight. Reproductive structures (combined number of buds, flowers and seeds, or panicle branches in D. canadense) and root nodules (D. canadense only) were counted to provide additional information on the effects of invasion due to belowground interactions. Flowering phenology and reproductive fitness are not often assessed in response to soil biotic interactions, but have the potential to provide additional insights into the effects of plant-soil interactions not necessarily expressed in biomass production (Wagner et al., 2014).

Root herbivory, damage and decay (together referred as ‘lesions’) were used as a proxy for pathogen activity using a modification of the Magnified Intersections Method (McGonigle, Miller, Evans, Fairchild, & Swan, 1990). Dried root samples were rehydrated overnight in 50% ethanol, cleared in 10% KOH at 80oC in a water bath for 30 minutes. Roots were kept in 50% glycerol until use. Randomly selected root segments (0.5-2cm long; <0.1mm wide) were mounted on microscope slides in 10 rows along the width of the slide. To minimize bias when a branched root was selected, branches were removed before mounting; these root segments were added back into the pool of root segments. Roots were mounted in 50% glycerol, covered with

22x50mm cover slips (Fisher Scientific; NJ, USA), and the edges sealed with clear-coat nail polish (Rimmel London, SalonPRO; NY, USA).

Slides were mounted and observed at x200 magnification. However, instead of using the cross- hair intersection approach (McGonigle et al., 1990), the entire root segment in the field of view (FOV) was considered an intersection. Multiple roots in a FOV were treated as separate intersections. A minimum of 100 root intersections were assessed for each sample, to minimize observer bias (McGonigle et al., 1990).

Each intersection was assessed for the presence/absence of different types of root lesions, and enumerated using a manual counter labeled with four categories. The categories are described as follows: None (e.g. clear roots, maybe dark lines in cell walls), Decay (e.g. asymmetrical discoloration, usually in external walls), Herbivory (e.g. jagged discontinuity of root wall), and Decay+Herbivory (a combination of the two lesion types). Only one ‘lesion’ category was counted per intersection (e.g. if Decay and Herbivory lesion types were present, this was counted enumerated in the Decay+Herbivory category only). Total root lesion abundance was calculated using a modification of the percent AM fungal colonization (McGonigle et al., 1990), whereby:

푡표푡푎푙 # 표푓 푙푒푠푖표푛 푙푒푠푖표푛 푑푎푚푎푔푒 = 푖 i = lesion type (decay or herbivory) 푖 푡표푡푎푙 # 표푓 푟표표푡 푖푛푡푒푟푠푒푐푡푡푖표푛푠

To estimate AMF colonization, Roots were rehydrated and cleared as above. Root samples were then acidified with 1% HCl for 15 minutes to improve stain adherence to chitin in fungal structures, and stained with a 0.05% trypan blue solution (1:1:1 glycerol, lactic acid and deionized water - by volume) at 80oC for 15 minutes, before storage in 50% glycerol. Slides were mounted and observed at x200 magnification. Percent colonization by AMF structures (arbuscules, vesicles and hyphae) was quantified using the ‘Magnified Intersections Method’ by McGonigle et al. (1990).

2.3.4 Statistical Analysis To verify that biomass production was indeed the result of rhizosphere biota, effects of live and sterile treatments were analysed separately using factorial ANOVAs with ‘Plant Species’ (including V. rossicum, N=6) and 'Invasion’ (invaded and uninvaded) as fixed factors. Data were ln transformed to meet the assumptions of normality and stabilize the residual variance. Lack of a significant ‘Plant species’ x ‘Invasion’ interaction in sterile treatments would then indicate that significant effects in live soil were due to biotic effects.

Net Biotic Effect (NBE) was calculated to further isolate the biotic effects from artifact abiotic factors (e.g. Non-significant variations in soil nutrients). This variable was calculated using bootstrapped ratios (Live: Sterile) of total, shoot and root biomass respectively (999 iterations; R version 2.11.1: R CoreTeam 2011). Bootstrapped samples were randomly selected within 95% confidence intervals and ln transformed to meet the assumptions of normality and stabilize the residual variance. Thus, NBE represents an estimate of the proportion of plant biomass which is a result of the interaction with rhizosphere biota (two categories: invaded and uninvaded) of defined as the net effect of rhizosphere biota on plant biomass (hence forth “the effect of rhizosphere biota (or rhizosphere biota) on plant biomass”.

The effect of invasion on native plants (i.e. excluding V. rossicum) was analysed using a factorial ANOVA with ‘Plant Species’ (N=5) and ‘Invasion’ (i.e. NBE invaded, NBE uninvaded; N=2) as fixed factors. Following significant ‘Plant Species X Invasion’ interactions, separate 2- tailed t-tests were performed on each species to test how invasion influenced responses. Vincetoxicum rossicum was assessed separately using t-tests.

To determine the production of biomass that could be explained by phylogenetic distance from V. rossicum (i.e. phylogenetic signal), the variation in biomass produced by native plants was compared to the variation expected based on the phylogenetic relatedness of native plants to V. rossicum. First, the evolutionary relatedness between each species was determined by the amount of genetic change over time estimated by Brownian motion in evolution (See phylogenetic tree Figure 2.3). Variance was measured by the total distance from phylogenetic root to tip (variance = 161 Figure 2.3). Covariation between each species based on genetic change was measured by adding shared branch lengths between each set of species. Closer evolutionary relationships therefore shared greater the amounts of covariation. To determine if responses of the six plant-species to the treatments were influenced by their phylogenetic relationship, we employed Pagel’s Lambda () (Pagel, 1999, Freckleton et al., 2002).  is an index to quantify correlation between phylogenetic distances among a set of species and their traits (Freckleton et al., 2002). A ’ estimate of “1” indicates that the variation in the response variable in a model (in this case the NBE on biomass) are directly in proportion to the variation expected based on their shared evolutionary history. A ’ of “0” indicates that the model is not related to the shared branch lengths and the response variable is independent of phylogeny.

The results of Pagel’s Lambda were confirmed using K-statistics (Blomberg et al., 2003). The K- statistic compares the strength of the observed phylogenetic signal to the strength of the

expected phylogenetic signal (Blomberg et al., 2003). When K = “1” the phylogenetic signal in response variables (in this case NBE on biomass) are equivalent what is expected Brownian motion evolution (Blomberg et al., 2003). When K > “1” this indicates that the phylogenetic signal is stronger than expected and when K < “1” it indicates that the phylogenetic signal in the observed data is less than expected.

The average NBE on shoot, root and total biomass of each species (N=6) was used to calculate Pagel’s ’ and K-statistics. All analysis was conducted in ape (Paradis et al., 2004) and phytools (Revell, 2012) packages in R 3.2.2 (R CoreTeam, 2015). Linear regression analysis was used to confirm Pagel’s ’ and K-statistics and the how much of the variability in the average NBE on total biomass could be explained by phylogenetic distance of native plants to V. rosicum.

Figure 2.3 A phylogenetic tree for the plant species tested in the present study based on Brownian motion of evolution. The horizonal lines are branches that represent evolutionary lineages changing through time. Every ~1cm of branch length is representative of 10 units of genetic change (see top left side of image: Tree Scale); the longer the branch, the larger the amount of change. Nodes are intersections of branches and represent genetic divergences from a common ancestor. The species are seporated into groups of increasing relatedness. Node tips are the last node along the branch terminating in the plant species name. The numbers above each branch are the numberical quantification of branch length based nucleotide substitutions per site which is the number of changes or 'substitutions' divided by the length of the genetic sequence. Thus, the numbers are representative of genetic variation. The shared branch length is equivelent to covariation between species. For example V. rossicum and A. gerardii do not share branch lengths (covariation = 0). Conversely, V. rossicum and R. hirta share the branch lengths 38 and 22.2 (covariation = (38+22.2)= 60.2). Thus, the closer the relationship between species, the greater the covariation value. Total distance from root to tip is the varience (in the tree above =161). This information on varience-covarience can then be used to calculate Pagel’s Lambda.

Counts of reproductive structures and root nodules (D. canadense only) were square root transformed (O ’Hara & Kotze, 2010) to meet the assumptions of normality and stabilize the residual variance. F-test two sample for variance tests were used to determine equality of variance. T-tests for independent samples were used to test for differences between plants, of a

given species, grown in invaded compared to uninvaded treatments. In all cases unequal variances were assumed.

Percentage colonization by AMF and Lesions were arcsine transformed to meet the assumptions of normality and stabilize the residual variance. As expected, sterilization significantly reduced colonization of AMF and occurrence of root lesions (see Appendix D& Appendix E for details), so the effects of invasion on native plant species were assessed for live treatments, alone.

The effects of invasion on root colonization by lesions and AMF, respectively, were determined using two-way ANOVAs with ‘Plant Species’ (N=5) and ‘Invasion’ (N=2) as factors. Analysis was conducted separately for arbuscular (AC), vesicle (VC) and hyphal (HC) colonization, decay and herbivory-type lesions. Following significant ‘Plant Species X Invasion interactions, separate t- tests were performed for each species to determine how invasion influenced responses.

Statistical tests were conduced in Statistica (Dell STATISTICA, Dell Inc. 2015. version 13. software.dell.com) and statistical results of t-tests were reported assuming equal variances - unless otherwise specified.

2.4 Results

2.4.1 Effect of Invasion on Native Plant Species Biomass

In sterilized soil treatments, biomass was unaffected by ‘Invasion’ (Figure 2.4; Table 2.2), whereas, in live-soil treatments, there were significant species-specific responses to ‘Invasion’ for shoot, root, and total biomass (Figure 2.4; Table 2.2).

14 Live (invaded) 12 Live (uninvaded) Sterile (invaded) 10 Sterile (uninvaded) 8 6

Biomass (g) Biomass 4 2 0 AnGe DeCa RuHi SoCa AsSy ViRo Figure 2.4 The total biomass produced by five native herbs when grown with and without rhizosphere biota (Live vs. Sterile treatments) from soils invaded and uninvaded by Vincetoxicum rossicum. Bars represent mean total biomass (± SE). Species abbreviations are: Andropogon gerardii (AnGe), Desmodium canadense (DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) and Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRO).

Table 2.2 Analysis of variance (ANOVA) statistics showing the effects ‘Species’ and ‘invasion’ on the biomass of native plants and Vincetoxicum rossicum produced in Live, Sterile treatments. ‘Species x Invasion’ effect of Net Biotic Feedback was calculated for native plants only; V. rossicum was excluded from net biotic effect,b Plant Species Invasion Species X Invasion Response F P F P F P Shoot Biomass Sterile 50.85 <0.001a 0.45 0.51b 1.03 0.410a Live 90.29 <0.001a 8.97 0.004b 7.35 <0.001a Net biotic effect 449.03 <0.001c 42.76 <0.001d 16.48 <0.001c Root Biomass Sterile 29.02 0.035a 0.88 0.353b 2.32 0.055a Live 38.66 <0.001a 11.23 0.001b 4.17 0.003a Net biotic effect 122.90 <0.001c 0.52 0.472d 11.52 <0.001c Total Biomass Sterile 42.55 <0.001a 0.03 0.857b 0.95 0.453a Live 84.95 <0.001a 16.89 <0.001b 8.70 <0.001a Net biotic effect 503.31 <0.001c 34.29 <0.001d 30.63 <0.001c a df=(5,59), b df=(1,59), c df=(5,49); d df=(1,49); Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

There was a significant ‘Plant Species’ x ‘Invasion’ for net biotic effect (Table 2.2). All native species, except D. canadense, responded negatively to rhizosphere biota from both invaded and uninvaded treatments (Table 2.2). The negative effect of rhizosphere biota on S. canadensis (total and shoot biomass) was significantly stronger in invaded compared to uninvaded treatments (Figure 2.5a & b; Table 2.3).

In contrast, total biomass of R. hirta and A. syriaca did not differ in response to ’Invasion (Figure 2.5a). The effect of rhizosphere biota on total biomass (NBE) in both species was different from the response to rhizosphere biota in shoot and root biomass. More shoot biomass (significant only for R. hirta) was produced in response to invaded relative to uninvaded rhizosphere biota (Figure 2.5b; Table 2.3). More R. hirta root biomass and marginally more A. syriaca biomass was produced in response to uninvaded relative to invaded rhizosphere biota (Figure 2.5c; Table 2.3). In retrospect, the opposite responses of shoot and root biomass to the factor of ‘invasion’ cancel out the effect of invasion recorded in total biomass.

The negative effects of rhizosphere biota on A. gerardii biomass were significantly greater in uninvaded compared to invaded treatments; more biomass was produced in response to rhizosphere biota without a history of V. rossicum invasion (Figure 2.5; Table 2.3).

Unlike the other native species, D. canadense responded positively to rhizosphere biota and this effect was greater in response to invaded verses uninvaded treatments; more biomass was produced in response to invaded relative to uninvaded rhizosphere biota (Figure 2.5; Table 2.3). The effect of rhizosphere biota was clear in live treatments where Shoot, root and total biomass produced 79%, 70% and 74% more biomass in invaded compared to uninvaded treatments.

The invasive plant V. rossicum also responded positively to rhizosphere biota overall. However, actual positive rhizosphere biota feedback (i.e., response to ‘Invasion’) was negligible (Figure 2.5a; Table 2.3).

1.5 (a) invaded *** uninvaded

1 Total Total 0.5 * *

0 Biomass -0.5

Net Biotic Effect on on Effect Biotic Net -1

-1.5 AnGe DeCa RuHi SoCa AsSy ViRo

1.5 (b) *** 1

0.5 * * 0 ***

Biomass -0.5

-1 Net Biotic Effect on Shoot Shoot on Effect Biotic Net -1.5 AnGe DeCa RuHi SoCa AsSy ViRo

1.5 (c) 1 ** 0.5 ** 0 **

Biomass -0.5

-1 Net Biotic Effect on Root Root on Effect Biotic Net -1.5 AnGe DeCa RuHi SoCa AsSy ViRo

Table 2.3 Independent t-tests statistics showig the effect of ‘invasion’ on the net biotic effect on plant biomass for each species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and Vincetoxicum rossicum). Analysis conducted on natrual log transformed data. Invaded Uninvaded Response M SD M SD df t-value P

Net Biotic effect on Shoot Biomass A. gerardii -0.54 0.16 -0.76 0.14 10 2.51 0.031 D. canadense 1.33 0.16 0.65 0.12 10 8.50 <0.001 R. hirta -0.77 0.06 -1.06 0.08 10 7.09 <0.001 S. canadensis -1.02 0.08 -0.85 0.15 10 -2.51 0.031 A. syriaca -0.28 0.11 -0.38 0.20 9 1.11 0.296 V. rossicum 0.71 0.15 0.62 0.32 9 0.57 0.583 Net Biotic effect on Root Biomass A. gerardii -0.23 0.15 -0.48 0.09 10 3.34 0.007 D. canadense 0.62 0.33 0.12 0.15 10 3.41 0.007 R. hirta -1.00 0.18 -0.63 0.09 10 -4.59 <0.001 S. canadensis -1.10 0.19 -1.05 0.19 10 -0.52 0.612 A. syriaca -0.29 0.15 -0.13 0.05 9 -2.21 0.054 V. rossicum 0.62 0.11 0.46 0.13 9 0.39 0.706 Net biotic effect on Total Biomass A. gerardii -0.38 0.20 -0.64 0.13 10 2.65 0.024 D. canadense 1.12 0.11 0.39 0.06 10 14.64 <0.001 R. hirta -0.89 0.08 -0.93 0.08 10 0.92 0.378 S. canadensis -1.05 0.06 -0.95 0.07 10 -2.62 0.026 A. syriaca -0.32 0.10 -0.21 0.12 9 -1.59 0.147 V. rossicum 0.62 0.11 0.46 0.13 9 2.05 0.070 M=mean; SD= standard deviation; df = group degrees of freedom; tdf = adjusted degrees of freedom for unequal variances; t value. = t statistic assuming equal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.7.

2.4.2 Analysis of Phylogenetic signal in Net Biotic Effect

Phylogenetic relatedness of the plant species did not explain significant variations in the NBE in plant biomass (Table 2.4). Both invaded and uninvaded NBE response variables (root, shoot and total) had had lambda (’) statistics near zero Table 2.4. Pagel’s Lambda (’) is a value which represents how close the covariation observed in the data set response variables of a set of species is to the covariation expected from these species based on background rates of genetic changes based on Brownian motion of Evolution, thus can be used to determine whether a phylogenetic signal can be seen in the data set (Freckleton et al., 2002; Pagel, 1999). If ’ approaches “1” then variation between species in the response variable (in this case the NBE on biomass) is directly in proportion to the variation expected based on their shared evolutionary history. If ’ approaches “0” then the variation observed between species is not related to the shared evolutionary history and indicates the differences in the data are independent of phylogeny. Therefore, a nearly consistent ’ < 5 x 10-5 for NBE on shoot, root and total biomass in response invaded and uninvaded rhizosphere biota indicates that these responses were independent of the phylogeny of the species observed. The one was NBE on total biomass in response to invaded rhizosphere biota ( =0.721) but this result was not significant (P=1.000; see Table 2.4).

K-statistics support this result. The K-statistic compares the strength of the observed phylogenetic signal based on the variation between species to the strength of the expected phylogenetic signal (Blomberg et al., 2003). When K = “1” the phylogenetic signal in response variables (in this case NBE on biomass) are equivalent to what is expected Brownian motion evolution (Blomberg et al., 2003). When K > “1” this indicates that the phylogenetic signal is stronger than expected and when K < “1” it indicates that the phylogenetic signal in the observed data is less than expected. The K-statistic for all response variables was less than one indicating that variation in the data did not fit the variation expected if the variation in the data were affected by phylogeny. Together these results show that there was no phylogenetic signal in the response of plant biomass to rhizosphere biota.

Linear regression analysis on the relationship between mean NBE and phylogenetic distance supported the results of both the K-statistic and Pagel’s Lambda. Less than 3% of the variation in mean total NBE could be predicted by phylogenetic distance (invaded R2=0.025; uninvaded R2=0.011). The correlation between these two variables was non-significant (R=0.159, P=0.799; and R=0.103, P=0.869 respectively) (Appendix J).

Table 2.4 Phylogenetic signal measured for average ‘Net Biotic Effect’ on plant biomass in response to ‘invasion’. P of Log Likelihood that observed (’) Response Variable ’ K P of K differs from expected () Invaded Net Biotic Effect Shoot NBE 7.44 x 10-5 1.000 0.609 0.275 Root NBE 6.24 x 10-5 1.000 0.610 0.277 Total NBE 0.721 0.850 0.697 0.226 Uninvaded (control) Net Biotic Effect Shoot NBE 7.44 x 10-5 1.000 0.588 0.306 Root NBE 7.11 x 10-5 1.000 0.538 0.368 Total NBE 7.44 x 10-5 1.000 0.629 0.306  and K= phylogenetic signal statistic. When =1 the response variable can be explained by the phylogenetic distance among the plant species tested (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and Vincetoxicum rossicum). When =0 there is no phylogenetic signal. When K<1 it implies that relatives resemble each other less than expected; when K>1 it implies that relatives resemble each other more than expected. P = significance of K statistic. P values are statistically significant at P<0.05.

2.4.3 Effect of Invasion on Native Plant Reproduction

Two of the five native species produced reproductive structures by the end the 13 week growth period (Desmodium canadense and Rudbeckia hirta). Desmodium canadense produced significantly more reproductive structures (panicle branches) in Live (Mean number of reproductive structures = 2.0 SD=1.06) compared to Sterile treatments and (mean=0.18 SD=0.26; df=12.34, t=5.8, P<0.001; Figure 2.6). In addition, D. canadense produced marginally more panicle branches in live- invaded (mean =2.60, SD=0.28) compared to uninvaded treatments (Mean =1.39, SD=1.22; df=5.5, t=2.37, P=0.059; Figure 2.6).

Rudbekia hirta had an opposite response comopared to D. canadense. Significantly more reproductive structures (flowers) were produced by R. hirta in sterile (mean=1.16, SD=0.79) compared to live treatments (mean=0.40, SD=0.45; df= 17.35, t=-2.91, P=0.010; Figure 2.6). There were no difference in the number of reproductive structures produced in Live-invaded (mean=0.25, SD=0.37) and uninvaded treatments (mean=0.55, SD=0.50; df= 9.25, t= -1.95, P=0.262). However, R. hirta produced significantly more flowers in Sterile-uninvaded treatments (mean=1.73, SD=0) compared to Sterile- invaded treatments (mean=0.59, SD=0.77; df=5, t=- 3.64, P=0.015; Figure 2.6).

9 Invaded 8 *** Uninvaded 7 6 ** 5 4 per Plant per ** 3 2 1

Mean Number of Reproductive Structures Structures Reproductive of Number Mean 0 Live Sterile Live Sterile DeCa RuHi

Figure A The quantity of reproductive structures produced by (a) Desmodium canadense (DeCa) and (b) Rudbeckia hirta (RuHi) when grown with and withouts soil biota (live vs. sterile) from soils invaded (dark gray) and uinvaded (light gray) by V. rossicum. Bars represent mean (± SE) number of reproductive structures (peduncles by D. canadense and flowers by R. hirta). Significance was determined using t-test analysis performed on square root transformed data (+0.1). Asterisks indicate significant differences between Live and sterile (long horrizontal lines) and invaded and uninvaded (short horozontal lines) at P<0.05 (*** P<0.001, ** P=0.01, * P=0.05).

2.4.4 Effect of Invasion on Nodule Formation in Desmodium canadense

Desmodium canadense produced significantly more nodules in live (mean=6.76, SD=3.95) compared to sterile treatments (mean=0.78, SD=1.17; df=12.91, t=5.04, P<0.001). Within Live treatments, Desmodium canadense produced significantly more nodules (df=10, t=10.77, P<0.001) in invaded (M=10.38, SD=1.10) compared to uninvaded treatments (mean=3.12, SD=1.24; Figure 2.7).

140 invaded *** uninvaded 120 ***

100

60 Mean Number of Nodules of Number Mean 40

0 live sterile Figure 2.7 The number of root nodules produced by Desmodium canadense in response to rhizosphere biota from soils invaded (dark gray) and uninvaded (light gray) by V. rossicum. Live treatments are compared to sterile controls (no nodules were produced in sterile-uninvaded treatments). Bars represent mean number of nodules (± SE) per plant. Differences between live and sterile treatments (long horizontal line) and invaded and uninvaded treatments (short horizontal line) were determined using independent t-tests on square root transformed data (+0.01). Asterisks indicate significant differences between invaded and uninvaded treatments at P<0.05 (***P<0.01, **P=0.01, * P=0.05).

2.4.5 Effects of Invasion on Incidence of Root Lesions

There was a marginally significant 60 invaded ‘Plant Species X Invasion’ interaction (a) uninvaded for decay type root lesions (Table 2.5). 50 Most species tested had more decay in 40

invaded treatments (A. gerardii, S.

type Lesions type - canadensis, A. syriaca and V. 30 rossicum) compared to uninvaded treatments. Desmodium canadense Decay of % 20 and R. hirta had more decay in 10 uninvaded compared to invaded treatments). However, differences in 0 the percentage decay between AnGe DeCa RuHi SoCa AsSy ViRo treatments were not significant (Table 2.5). Only A. gerardii had a marginally 8 (b) larger percentage of root decay when 7 in invaded compared to uninvaded soil 6

(Table 2.6; Figure 2.8) Lesions type - 5 The percentage root herbivory differed 4 significantly between species (Table 3

2.5) but the level of herbivory-type Herbivory of % 2 lesions was not significantly affected by 1 invasion in any of the native species 0 (Table 2.5; Table 2.6). AnGe DeCa RuHi SoCa AsSy ViRo V. rossicum was not significantly Figure B The amount of root decay (a) and herbivory-type (b) lesions on native (compared to Vincetoxicum rossicum) plants grown affected by ‘Invasion’ in either the with rhizosphere biota from soils invaded (dark gray) and uninvaded percentage of decay-type lesions or (light gray) treatment. Bars represent mean percent root lesions (± herbivory-type lesions (Table 2.6; SE) per plant. Species abbreviations are: Andropogon gerardii (ANGE), Desmodium canadense (DeCa), Rudbeckia hirta (RuHi), Figure 2.8). Solidago canadensis (SoCa) Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRO). Independent t-tests were used to detect differences based on the invasion history of rhizosphere biota (non were detected).

Table 2.5 Analysis of variance (ANOVA) statistics showing the effect of ‘Invasion’ and ‘Plant Species’ in live treatments on percentage occurance of decay and herviroy-type lesions in roots of native species. Plant Speciesa Invasionb Plant Species X Invasiona Lesion Type F P F P F P Decay 33.76 <0.001 2.13 0.151 2.39 0.063 Herbivory 46.60 <0.001 2.03 0.160 2.14 0.090 a df= (4,50), b df= (1,50). Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

Table 2.6 Independent t-tests statistics assuming unequal variances showing the effect of Vincetoxicum rossicum invasion history (invaded vs. uninvaded) on lesion formation (by decay and herbivory) for each plant species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and V. rossicum).

Species invaded Uninvaded M SD M SD Ndf tdf t separ. P

Decay A. gerardii 0.39 0.10 0.26 0.10 10 9.98 2.14 0.058 D. canadense 0.37 0.08 0.45 0.14 10 8.22 -1.19 0.268 R. hirta 0.03 0.01 0.04 0.03 10 6.94 -0.95 0.372 S. canadensis 0.17 0.07 0.10 0.05 10 8.78 2.04 0.073 A. syriaca 0.23 0.09 0.17 0.13 10 8.74 1.00 0.346 V. rossicum 0.30 0.10 0.21 0.08 9 8.99 1.72 0.119

Herbivory A. gerardii 0.06 0.02 0.05 0.02 10 9.45 0.80 0.444 D. canadense 0.00 0.01 0.01 0.01 10 9.99 -1.08 0.304 R. hirta 0.00 0.00 0.00 0.01 10 5.00 -1.48 0.198 S. canadensis 0.00 0.00 0.00 0.00 10 5.00 1.54 0.185 A. syriaca 0.02 0.02 0.00 0.00 10 5.16 2.01 0.098 V. rossicum 0.01 0.01 0.01 0.01 9 7.92 0.26 0.799 M=mean; SD= standard deviation; Ndf = group degrees of freedom; tdf = adjusted degrees of freedom for unequal variances; t sep. var. = t statistic assuming unequal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07

2.4.6 Effects of Invasion on Colonization by Arbuscular Mycorrhizal Fungi

The effects of invasion were detected in AMF hyphal colonization, as demonstrated by a significant ‘Plant Species X Invasion’ interaction (in live treatments) (F (4,49) =2.97, P=0.028; Table 2.7).

Among native plant species, only D. canadense had a significantly higher percentage of hyphae in invaded compared to uninvaded treatments (Figure 2.9a; Table 2.8). Two other species, S. canadensis and A. syriaca, showed a similar trend, with a marginally significant number of hyphal colonization in invaded compared to uninvaded treatments (Figure 2.9a; Table 2.8).

There was a marginal species-specific response shown for vesicular colonization of native plant roots (F (4,49) =2.53, P=0.052; Table 2.7). Only D. canadense responded significantly (t=3.32, P=0.045; Figure 2.9b), with more vesicles in invaded treatments compared to uninvaded treatments (Table 2.8).

In contrast to D. canadense, AMF colonization of V. rossicum roots in invaded (its own biota) treatments was not significantly different compared to uninvaded treatments (Table 2.8).

Table 2.7 Analysis of variance (ANOVA) statistics showing the effect ‘Invasion’ (invaded vs. uninvaded) and ‘Plant Species’ (N=5) on the percent colonization by Arbuscular mycorrhizal fungi (arbuscules (AC), vesicles (VC), and hyphae (HC)) in roots of native species. Analysis was conducted on arcsine transformed percent colonization data from live treatments.

Plant Species a Invasion b Plant Species X Invasion a Response F P F P F P AC 6.97 <0.001 1.08 0.304 1.45 0.232 VC 14.71 <0.001 1.36 0.248 2.53 0.052 HC 7.65 <0.001 4.34 0.042 2.97 0.028 a df= (4,49), b df= (1, 49). Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

invaded 80 (a) 70 uninvaded 60 50 40 * 30

20 % Hyphal Colonization Hyphal % 10 0 AnGe DeCa RuHi SoCa AsSy ViRo

16 (b) 14 12 10 8 6 *

4 % Vessicle Colonization Vessicle% 2 0 AnGe DeCa RuHi SoCa AsSy ViRo

18 (c) 16 14 12 10 8 6 4

% Arbuscular Colonization Arbuscular% 2 0 AnGe DeCa RuHi SoCa AsSy ViRo

Figure 2.9 The percent arbuscular mycorhizal fungal colonization of five native plants and Vincetoxicum rossicum grown with rhizosphere communities from soils invaded (dark gray) and uninvaded (light gray) by V. rossicum. Bars represent mean percent root colonization (± SE) of hyphae (a), vesicles (b) and arbuscules (c) per plant. Species abbreviations are: Andropogon gerardii (AnGe), Desmodium canadense (DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRO). Asterisks show significant (P<0.05) from independent t-test results for ‘Invasion” on arcsine transformed data: ***P<0.001, **P=0.01, * P=0.05.

Table A Independent t-tests statistics on arcsine transformed data assuming unequal variances showing the effect of Vincetoxicum rossicum invasion history (invaded vs. uninvaded) on vesicle and hyphal colonization for each plant species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta, Solidago canadensis, Asclepias syriaca, and V. rossicum).

Invaded Uninvaded M SD M SD ND tdf t sep var P

Hyphal colonization A. gerardii 0.37 0.06 0.40 0.13 10 7.25 -0.64 0.544 D. canadense 0.28 0.12 0.11 0.05 10 6.63 3.32 0.014 R. hirta 0.43 0.14 0.54 0.19 10 8.98 -1.10 0.299 S. canadensis 0.35 0.14 0.20 0.13 10 9.93 1.84 0.096 A. syriaca 0.45 0.17 0.26 0.16 9 8.35 1.91 0.092 V. rossicum 0.52 0.32 0.49 0.24 10 9.22 0.16 0.876

Vesicle colonization A. gerardii 0.07 0.03 0.11 0.07 10 6.85 -1.27 0.245 D. canadense 0.03 0.02 0.01 0.01 10 8.22 2.36 0.045 R. hirta 0.09 0.04 0.07 0.02 10 6.21 1.17 0.284 S. canadensis 0.06 0.05 0.02 0.01 10 5.56 2.18 0.076 A. syriaca 0.01 0.01 0.01 0.01 9 8.89 0.62 0.548 V. rossicum 0.03 0.03 0.03 0.03 10 9.95 -0.22 0.832 M=mean; SD= standard deviation; Ndf = group degrees of freedom; tdf = adjusted degrees of freedom for unequal variances; t sep. var. = t statistic assuming unequal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

2.5 Discussion

2.5.1 Does ‘Enemy of my Enemy’ Influence Vincetoxicum rossicum Invasion?

The results of this study show that rhizosphere communities from areas invaded by V. rossicum have differential effects on native plant biomass. Contrary to predictions, only one species (Solidago canadensis), experienced reduced growth in the presence of rhizosphere biota from V. rossicum invaded compared to uninvaded soil suggesting ‘Enemy of my Enemy’ as a mechanism (see Figure 2.5a). This specific result was consistent with the findings of Day et al. (2016), who demonstrated that communities of fungal pathogens isolated from V. rossicum inhibited the growth of S. canadensis. In the present study, abundance of root lesions was used as a proxy for pathogen activity. However, the quantity of root lesions on S. canadensis did not differ between invaded and uninvaded treatments and therefore relative reductions in S. canadensis biomass did not correspond with an increase in root lesions as expected (Figure 2.8). Root biomass of S. canadensis was not significantly affected by invasion, but shoot biomass was reduced in response to invaded compared to uninvaded rhizosphere biota (Figure 2.5c). As such, the inhibitory effects that V. rossicum rhizosphere biota had on shoot biomass, contributed the most to the overall reduction in growth of S. canadensis in invaded compared to uninvaded treatments (Figure 2.5c). The absence of invasion effects on root lesion and root biomass parameters, in addition to the presence of invasion effects on reduced shoot growth, suggests two possibilities: (1) mutualist rhizosphere biota in uninvaded treatments alleviated the overall negative effects of soil antagonists that affected S. canadensis biomass (Figure 2.5) (e.g. Hernández-Montiel et al. 2013; Mordecai, 2013); or (2) consistent with the EE hypothesis, soil antagonists were elevated in V. rossicum rhizosphere biota communities which reduced S. canadensis biomass in invaded relative to uninvaded treatments, but these antagonists did not always produce lesions.

Arbuscular mycorrhizal fungi are well known plant mutualists providing phosphorus (and a variety of other beneficial effects) in exchange for carbon, leading to improved host plant growth (Philippot et al., 2013). Mutualistic interactions with AMF are one of the most common beneficial interactions, occurring in more than 80% of vascular plants (Philippot et al., 2013). If S. canadense was benefiting from AMF mutualistic interactions in uninvaded but not invaded treatments, then it should be reflected in AMF root colonization. However, no significant differences in AMF colonization between treatments were observed (Figure 2.9). There are many other mutualists besides AMF that were not measured which could have benefited S.

canadensis; for example: plant growth promoting root endophytes (Hardoim et al., 2015). However, in this study, rhizosphere biota inhibited S. canadensis biomass production in both invaded and uninvaded soil biota, suggesting that antagonistic and not mutualistic interactions may be dominant in S. canadensis relationships with soil biota.

Counting lesions allow for the direct observations of the effects of pathogens such as root truncation which reduces access to nutrients (Coyne, Nicol, & Claudius-Cole, 2014). However, several authors have recognized that current methods of counting lesions may underestimate disease effects (Martínez-García, Pietrangelo, & Antunes, 2016; Mitchell, 2003; Schnitzer et al., 2011). The method used in this study was an attempt to improve precision by using a standard procedure to control for user bias.

There are other reasons why disease effects may be underestimated. These methods assume that pathogens will form visible lesions, but not all antagonistic rhizosphere biota form lesions at all stages of their life cycle. Increasingly, it has been observed that some plant pathogens are asymptomatic to at least some of their hosts (Stergiopoulos & Gordon, 2014). Whether cryptic pathogens impact their host is still an open question; recent research suggests they do (e.g. Aguilar-Trigueros & Rillig, 2016; Shaw et al., 2016). Relative reductions of S. canadensis biomass in response to invaded rhizosphere biota compared to uninvaded rhizosphere biota, could be due to cryptic pathogens with no obvious signs of disease (Faillace et al., 2017).

Other possibilities not explored here are multipartite interactions between multiple biota and their host. Often, multipartite interactions are considered in terms of beneficial synergies (Hardoim et al., 2015), but interactions between multiple species of microbes residing in a single host can also produce antagonistic effects on the host which would not occur otherwise (Kobayashi & Crouch, 2009; Lamichhane & Venturi, 2015; Syller, 2012). For example, tomato plants have been shown to exhibit more severe disease symptoms when co-infected by tomato chlorosis virus and tomato spotted wilt virus, than when infected by either virus alone (García- Cano et al. 2006). The two viruses can interact synergistically in the host, inducing disease symptoms that are greater than would be expected if the viruses acted in an additive manner (García-Cano et al. 2006). Synergistic interactions which produce antagonistic effects in a host can occur between taxa as well as within taxa including bacterial-fungal interactions such as the interaction between Colletorichum coccodes and Pyrenochaeta lycopersici in tomato brown- root-rot (Vrisman et al., 2017).

Given that soil antagonists do not always yield visual evidence of their presence, it is still possible that the reduced growth observed in S. canadensis in response to invaded treatments is due to the accumulation of enemies in V. rossicum soils. Accumulation of cryptic pathogens would still allow V. rossicum to gain dominance over S. canadensis through apparent competition, thus providing support for the EE hypothesis. However, this negative response of S. canadense to V. rossicum rhizosphere biota was not consistent with the other native species tested in this study (A. gerardii, D. canadense, R. hirta, and A. syriaca) (Figure 2.5). In fact, D. canadense and A. gerardii responded more positively to invaded, relative to uninvaded rhizosphere biota.

Other studies have shown that native plants are not consistently sensitive to changes in soil biotic communities driven by plant invasions (Del Fabbro & Prati, 2015; Jordan et al., 2011). Furthermore, pathogen symptoms can be alleviated by strong associations between native plants and mutualistic soil organisms such as when the mutualistic association between two beneficial organisms (e.g., AMF and Pseudomonas sp.) reduce disease severity from a pathogen (e.g., the wilting caused by Fusarium oxysporum on Carica papaya) (Hernández- Montiel et al., 2013). There is great diversity in the interactions between plants and organisms in their microbiome, largely driven by genetic differences and environmental circumstances (Pieterse, et al. 2016). Thus, different plant species will have different abilities to resist to generalist pathogens depending on different mutualisms they might form with microbiota or their innate susceptibility and tolerance to that pathogen as the associations that it might form (e.g., specific character traits) (Mordecai, 2013). It is plausible that rhizosphere communities that have developed in response to V. rossicum invasion (i.e. were ‘trained’) include generalist beneficial symbionts that can promote native plant growth, or at least alleviate some of the negative effects of rhizosphere antagonists (i.e., ‘the friend of my enemy is my enemy’). Indeed, previous studies have shown V. rossicum is heavily colonized by AMF, relative to native species and the association often drives positive feedback (Smith et al. 2008). However, Day, Antunes, & Dunfield, (2015) cautions that site variability may influence V. rossicum biomass results; of the two sites they investigated, significant positive soil feedback was only observed at one.

In this study, AMF colonization of V. rossicum did not differ between invaded and uninvaded treatments. This similarity between treatments was reflected by positive biomass responses in both treatments (Figure 2.5), indicating that rhizosphere biota was beneficial to V. rossicum which was present in both invaded and uninvaded sites. Yet the fact that V. rossicum associated rhizosphere biota affected the growth of native plants differently than rhizosphere

biota from adjacent uninvaded plant communities shows that V. rossicum invasion drives changes in rhizosphere communities.

Andropogon gerardii, was negatively affected by rhizosphere biota overall, but was less negatively affected by rhizosphere biota in invaded relative to uninvaded treatments (Figure 2.9). This result suggests that either certain rhizosphere biota in V. rossicum soil were alleviating the negative effects of soil antagonists found in both invaded and uninvaded plant communities, or soil from uninvaded areas had more antagonistic biota. AMF colonization (AC, VC, or HC) did not significantly differ between invaded and uninvaded treatments. However, levels of decay-type lesions were marginally higher on A. gerardii roots grown in invaded compared to uninvaded treatments treatment (Figure 2.8). Taken together, these results suggest that organisms other than AMF may be benefiting this species. Indeed, studies have shown higher numbers of ammonia-oxidizing bacteria (AOB) in association with many invasive species of many different herbaceous forms, such as grass, shrubs and trees (McLeod et al., 2016; Rodrigues et al., 2015). Furthermore, invasive species have also been associated with greater nitrate and ammonium turnover (Rodrigues et al., 2015). In the present study, nitrate concentrations were not measured directly, however total nitrogen concentrations as well as other limiting nutrients did not differ between V. rossicum invaded and uninvaded areas (see Appendix C - Table C.2 & Table C.3). Consistent nutrients between invaded and uninvaded areas at all soil collection sites indicates any differences observed during the study in the productivity of native plants between invaded and uninvaded treatments were likely the result of different rhizosphere communities formed from of V. rossicum invasion (or lack there of; uninvaded areas) rather than differences in nutrients between invaded and uninvaded soils.

Desmodium canadense biomass was promoted by rhizosphere biota in both invaded and uninvaded treatments. The positive effect on biomass was stronger with rhizosphere communities ‘trained’ by V. rossicum (Figure 2.5). Reproductive structures were also significantly higher in (live) invaded compared to uninvaded treatments (Figure 2.6), indicating that V. rossicum rhizosphere biota improved the overall fitness of D. canadense.

The number of root lesions on D. canadensis did not differ between treatments (Figure 2.8). However, hyphal and vesicle colonization as well as the number of root nodules were significantly elevated on plants grown in live- invaded compared to uninvaded treatments (Figure 2.9). Root nodules and AMF colonization represent two different symbiotic relationships. Root nodules form during symbiotic interactions between legume plants and rhizobia, a

polyphylogenetic group of nitrogen-fixing bacteria, which provide the host ready access to plant available forms of nitrogen (nitrate and ammonium) (Philippot et al., 2013). AMF provide the host ready access to phosphorus which is often limited because it is generally immobile in the soil. By forming associations with both AMF and rhizobacteria, legumes can benefit from improved access to important nutrients (N and P). It is unclear why soil communities associated with V. rossicum would improve AMF colonization and nodule formation while not also benefiting V. rossicum.

Vincetoxicum rossicum, is a generalist host and has been recorded associating with a variety of opportunistic AMF (Bongard et al., 2013). It has also been shown to elevate AMF colonization of its local soils (Smith et al., 2008; Day, Antunes, & Dunfield, 2015). In the present study, rhizosphere biota promoted V. rossicum biomass in general but there were no significant differences in biomass production nor AMF colonization between treatments (Figure 2.5; Figure 2.9).

Fungal endophytes, including AMF, have been observed in colonizing roots and root nodules of legumes but generally the effects of this are unknown (Dudeja et al., 2012). It is possible that D. canadensis benefits from a tripartite association between legumes and nitrogen-fixing bacteria. Indeed, legumes have been shown yield up to 15 times more biomass when grown with both AMF and rhizobia compared to either symbiont alone (van der Heijden et al., 2015; but see Bauer et al., 2012; Larimer et al., 2010). The increased access to nutrients may allow the plant to divert additional carbon to its symbionts thereby promoting their growth in a positive feedback loop.

If V. rossicum amplified certain species of rhizosphere biota that then enhanced the benefit of root-nodule bacteria (either directly as helper bacteria or indirectly in a tripartite synergistic relationship; AMF + root-nodule bacteria + host plant), this might explain why D. canadense grew more in live- invaded compared to uninvaded treatments and why V. rossicum did not similarly benefit in invaded (its own biota) treatments.

Desmodium spp. form mutualistic associations with root-nodule bacteria, mostly in the Bradyrhizobium group (Parker et al., 2015) but associations with other rhizobia are likely; specifically, Rhizobium and Mesorhizobium (Gu et al., 2007; Tlusty et al., 2004). Vincetoxicum rossicum may associate with ammonium oxidizing bacteria (AOB) like other invasive species trees (McLeod et al., 2016; Rodrigues et al., 2015). However, it is not clear how associations with AOB would influence the rhizosphere biota community of D. canadensis or other legumes.

Is the growth promoting effects of V. rossicum rhizosphere biota consistent across all legumes species? Further research is needed to determine identity of the microbial species in the nodules of D. canadense, including rhizobia and other bacterial and fungal endophytes and whether these are species also supported by V. rossicum.

2.5.2 Does Phylogenetic Relatedness Influence Response to Invasion?

This study demonstrated that phylogenetic relatedness to V. rossicum could not explain the growth response of native plants to biota from soil invaded for at least five years by V. rossicum. It has been predicted that more closely related plant species should share common enemies because of similar physiological traits; once an enemy has evolved a mechanism to overcome the defences of one plant species it likely could overcome defences of similar makeup (Anacker et al., 2014; Brandt et al., 2009; Callaway et al., 2013). The phylogenetic signal (the degree and direction of the response native plants had to V. rossicum soils) varied between species independent of phylogenetic relationship to V. rossicum (Table 2.4). Although, S. canadensis (which is relatively closely related to V. rossicum) responded negatively to V. rossicum rhizobia and D. canadense (which is comparatively more distantly to V. rossicum), no pattern emerged across the species tested. There are several studies that have not found a phylogenetic signal for plant communities (Fitzpatrick et al., 2017; Mehrabi et al., 2015; Mehrabi & Tuck, 2015), and specifically native and non-native invasive comparisons (Dostál & Palečková, 2011). Diez et al. (2008) suggests the reason why the evidence for phylogenetic signals in plant responses to soil biota is spilt is because observations are often made at different scales, for example: recent introductions compared to historical introductions of non-native plant species. Once the invading species is established (i.e. naturalized) the phylogenetic signal may be reduced or even lost (Diez et al., 2008). In this study, rhizosphere biota was collected from areas that had been invaded by V. rossicum for at least 5 years. However, V. rossicum has been recorded in the region for over 100 years (DiTommaso et al., 2005; Sheeley & Raynal, 1996), if Diez et al. (2008) is correct then the phylogenetic signal may be too weak to detect using the current design and measurement scale of this study.

2.5.3 Conclusions

In conclusion, the evidence presented in this study suggests the EE hypothesis is not a global mechanism of V. rossicum invasion. While V. rossicum harbours pathogenic organisms with the

potential to inhibit growth of some native species it also harbors rhizosphere biota that are beneficial to the growth of other native species; even when these rhizosphere biotas do not appear to benefit V. rossicum itself. This was particularly evident in D. canadense, for which V. rossicum associated rhizosphere biota promoted vegetative and reproductive production. Something other than AMF is likely responsible for V. rossicum benefits to native species. The evidence presented here points towards nitrogen fixing bacteria however, more research is needed to determine which organisms confer benefits to native plants. With increased understanding of V. rossicum driven changes in soil biota, and how this affects native plants, particularly legumes (e.g. D. canadense), could be identified for effective use in remediation efforts for V. rossicum invasions.

Additionally, the results of this study add to the discussion of the influence of phylogenetic relatedness on the effects of invasion for native plants. More research on phylogenetic relatedness in plant-soil feedback under different conditions is needed – especially in relation to plant traits (Anacker et al., 2014; Feng & van Kleunen, 2016; Fitzpatrick et al., 2017). A greater understanding of how phylogenetic relatedness influences plant-soil feedback will allow us to better understand plant community dynamics and better predict the progression of plant invasions and their ecological consequences.

Chapter 3

Response of Native Legumes to Vincetoxicum rossicum Rhizosphere Biota

3.1 Abstract

Interactions between plants and their soil biota are among the primary driving factors in plant community dynamics – particularly during plant invasions. It is generally assumed that invasive species are governed by positive feedbacks with mutualistic soil biota while native species are governed primarily by negative feedbacks with antagonistic soil biota (parasites/pathogens). Although, occasionally, invasive species have been shown to facilitate co-occurring species regardless of their geographic origin, little is known about the mechanisms behind this phenomenon. Recently, it has been shown that non-legume plants can facilitate nodulation in legumes by exuding certain compounds through their roots. This suggests that plants ‘communicate’ with soil biota to influence the relationship they have with neighboring plants. Earlier in this thesis (Chapter 2) Desmodium canadense (a native legume) produced more biomass when grown with rhizosphere biota from soil invaded by Vincetoxicum rossicum compared to rhizosphere biota from uninvaded areas. Here, V. rossicum and four phylogenetically diverse native legumes were used to test whether rhizosphere biota from V. rossicum-invaded soils promote the growth of native legumes in general, or whether the positive effect previously observed is specific to D. canadense. In addition, the responses of the native legumes were used to further test the ‘Enemy of my Enemy’ (EE) hypothesis which states that pathogen accumulation on invasive species facilitates invasions by supressing the growth of surrounding plant species relative to that of the invasive host. Each of the four legumes (D. canadense, Lespedeza hirta, Lupinus perennis, and Astragalus canadensis) and V. rossicum, were grown for three months with and without rhizosphere biota from V. rossicum invaded and uninvaded soils. Biomass, reproductive potential and nodule formation were measured. The results showed that V. rossicum rhizosphere communities specifically promoted D. canadense biomass relative to rhizosphere communities from adjacent soils. The other legumes had variable responses. Only one other legume (L. perennis) benefited significantly more from

invaded relative to uninvaded rhizosphere biota; however, the response was smaller than that of D. canadense (0.6g vs. 3.9g). Astragalus canadensis and L. hirta were unaffected and inhibited by V. rossicum soil biota, respectively. The results of this study support the findings of the previous study (Chapter 2) in that rhizosphere biota associated with V. rossicum have variable effects on native plants, and the mechanisms of pathogen accumulation (as described by the EE hypothesis) is unlikely to be important in facilitating invasions of V. rossicum. In conclusion, V. rossicum rhizosphere biota appear to specifically promote D. canadense fitness.

3.2 Introduction

Non-native plant invasions are characterized by strong competition (or apparent competition) with native and naturalized plants, such that the invading species dominates the plant community. Consequently, plant species abundance and richness decline in these invaded communities. In turn, ecological processes are altered and/or interrupted which ultimately leads to ecological degradation (Early et al., 2016; Gilbert & Levine, 2013; Isbell et al., 2017; Simberloff, 2005; Vilà et al., 2011; Vitousek et al., 2011).

Interactions between plants and soil biota are thought to be the primary factors driving plant community dynamics – particularly during plant invasions (Bardgett et al. 2005; Bardgett and Wardle 2010; Pieterse et al. 2016; Rasmann & Turlings 2016; Reinhart & Callaway 2006; Wagg et al. 2014). It is generally thought that invasive species experience positive feedbacks driven by mutualistic soil biota (Klironomos, 2002). Conversely, native species tend to more readily experience negative feedbacks driven by antagonistic (parasites/pathogens) soil biota (Bever et al., 1997; Reinhart & Callaway, 2006).

Interactions between plants and soil biota are dependent upon the context of environmental and biotic (genetics, host status) parameters. Even interactions between the same host and biota pairs can have variable outcomes depending on the context of their interaction (Catford et al., 2009; Johnson & Graham, 2012; Johnson et al., 1997; Veneault-Fourrey et al., 2013). In addition, the diversity and function of biota in association with their plant hosts and other biota is poorly understood. Therefore, the consequences that arise from many of these interactions remain unknown (e.g. see root-nodule endophytes in Velázquez et al., 2017; Hardoim et al., 2015).

Among the interactions that facilitate plant invasions, non-native plants (i.e. invasive) can associate with soil biota, either introduced or indigenous, with which they may establish mutualistic relationships. However, what is a mutualism on one host may instead be antagonistic on another, particularly in native plants (see: enemy-of-my-enemy, local pathogen accumulation, and/or spillover hypothesis in Colautti et al., 2004; Eppinga et al., 2006; Mangla et al., 2008). At the same time, the enemy release hypothesis (ERH) suggests invasive plants may experience enemy release from soil antagonists in their native range, allowing them to divert resources away from costly defence mechanisms toward more vigorous growth (Colautti et al., 2004; Keane & Crawley, 2002; Richardson & Pyšek, 2006). Novel interactions between

non-native invasive plants and generalist mutualists may further enhance the benefits of enemy release described as the “evolution of increased competitive ability hypothesis” (EICA) (Blossey & Notzold, 1995; Jeschke et al., 2012).

Empirical support for the ERH and EICA is split, particularly with respect to rhizosphere biota (Colautti et al., 2004; Jeschke et al., 2012). Native plant pathogens are known to colonize non- native plants (e.g., Bongard et al., 2013). Over time, it is hypothesised that invasive species should steadily accumulate antagonistic soil biota (including viruses) because local pathogens evolve to evade host defences, contributing to the invasive species’ decline (Flory & Clay, 2013; Lankau et al., 2009). However, despite common plant pathogens being found among the roots of invasive species, accumulation of pathogens does not always lead to invasive declines (Day, Dunfield, & Antunes, 2015). It is not also clear how often invasive plant mutualists, commensals and/or weak antagonists become pathogens on native plants. In the experiment described in the previous chapter (Chapter 2), the effects of V. rossicum rhizosphere associated biota were tested on six phylogenetically diverse native plant species. Rhizosphere communities were collected from field soils of V. rossicum-invaded and uninvaded areas. Half of the invaded and half of the uninvaded soils were sterilized to remove rhizosphere biota. Plants were grown in the presence or absence of rhizosphere biota from each of the invaded and uninvaded soils. The rhizosphere biota from V. rossicum soils had a variable effect on native plants that ranged from negative to neutral and, in the case of Desmodium canadensis, V. rossicum rhizosphere biota had a positive effect.

It is not uncommon for studies to report neutral effects from invasive species on the native plant population. For instance, Jordan et al. (2008) found that three plant invaders of the Great Plains region of North America, caused shifts in soil biota which negatively impacted growth of some but not all native species. Some native species were relatively insensitive to the changes in soil biota caused by the invasive plants tested. Rarely however, are invasive plants reported to promote native plants. Del Fabbro & Prati (2015) compared invasive and weedy plants and the effects of their ‘trained’ soils (a combined effect of allelopathic compounds and soil biota) on target native plants and found that invasive plants were no more inhibitory than weedy native plants (Del Fabbro & Prati, 2015). In fact, some invasive species were found to promote target native species.

One reason why some invasive plants have been observed promoting native plants is because they contribute to enhanced soil fertility. Increases in soil nitrogen and nitrification rates due to

ammonia oxidizing bacteria have been shown to occur after plant invasions by non-legumes (McLeod et al., 2016; Xiao, Schaefer, & Yang, 2017). This increase in nitrogen could potentially benefit any plants in the vicinity including native plants. In the previous study of this thesis (Chapter 2) no differences in total nitrogen concentrations were observed in Vincetoxicum rossicum (an invasive vine introduced from Ukraine and Southern Russia) invaded areas compared to adjacent uninvaded areas at southern Ontario field sites. However, effects of rhizosphere biota associated with V. rossicum were detected on the growth of native plants.

In the previous study (Chapter 2), Desmodium canadense was the only legume and the only native species to be promoted (increased biomass and the amount of reproductive structures and nodules) by rhizosphere biota from soil invaded by V. rossicum. Fabaceae (the legume family) form mutualistic associations with a polyphylogenetic clade of N2 fixing root-nodule bacteria consisting of Rhizobiales (e.g., Bradyrhizobium) the Phyllobacteriaceae (e.g., Mesorhizobium), and the Rhizobiaceae (e.g., Rhizobium) (Oldroyd, Murray, Poole, & Downie, 2011). Collectively they are referred to as rhizobia and act as pseudo-organelles called bacteroids, fixing nitrogen (N2) in the plant roots (Oldroyd et al., 2011).

Generally, legumes improve the nitrogen status of the surrounding soil, which makes them an important crop in agriculture (Jackson et al., 2008). As such, the symbiosis with N2 fixing rhizobia have been extensively studied (Oldroyd, 2013; Oldroyd et al., 2011; Oldroyd, & Downie, 2008, 2004). Rhizobia can be generalist or specialists when it comes to the selection of their hosts, although host-specific interactions are a common phenomenon among the Fabaceae (Ehinger et al., 2014; Parker et al., 2015). Recently, it has been shown that non-rhizobia plants can enhance nodulation of legumes (reviewed by Coskun et al., 2017). One study showed that corn facilitates nodule development in two agricultural legumes through its root exudates (Li, B. et al., 2016), but other mechanisms might also have been involved. In a study on the root nodule bacteria of Lotus corniculatus and Anthyllis vulneraria, Ampomah & Huss-Danell (2011) found genetic sequences (nod factors) in non-rhizobia (actinobacteria) that are typically associated with Mesorhizobium, the dominant rhizobia genus in the examined nodules. Further tests revealed that these non-rhizobia taxa had the ability to nodulate their legume hosts. This was the first study to show that actinobacteria were capeable of triggering nodulation in legumes, and adds to the growing list of non-rhizobia bacteria that posess functional nod-factors generally thought to be obtained through lateral gene transfer from rhizobia (Martínez-Hidalgo & Hirsch, 2017).

It is not well understood how prevalent cross-species nodule promotion from a non-legume to legume is for non-agricultural crops. The previous study in this thesis (Chapter 2) provided evidence that at least one legume species (D. canadense) was promoted by V. rossicum and this was accompanied by an increase in nodule formation. But questions remain. Does V. rossicum promote the growth of D. canadensis by promoting the formation of root nodules, or was this an artifact of the experiment? Is growth promotion of legumes a common effect of V. rossicum rhizosphere biota? If true, it would suggest that V. rossicum changes the rhizosphere community in ways that promote the relationship between legumes and their N2 fixing bacteria.

In the present study, I tested the hypothesis that V. rossicum rhizosphere biota promotes native legumes (likely by promoting beneficial associations between legumes and N2-fixing bacteria). If generalist rhizosphere biotas are involved, then all legumes should benefit from the interaction with V. rossicum rhizosphere biota. However, if specific soil biotas are involved it would be expected that only some legumes benefit from this interaction. The present experiment also provides the opportunity to test the EE hypothesis; specifically, that V. rossicum rhizosphere communities contain antagonistic biota which inhibit native plant growth (in this case tested for legumes) thereby facilitating V. rossicum.

3.3 Materials and Methods

3.3.1 Collection of Biotic Soil Communities

Methods were adapted from Chapter 2. Soil was collected from the same four Conservation Areas in Southern Ontario, Canada (Figure 3.1; See Appendix A) where V. rossicum had been present for a minimum of five years. Soils from grassland habitats were selected to acquire rhizosphere biota (henceforth rhizosphere biota) from locations where V. rossicum is the most productive (Cappuccino, 2004; Cappuccino et al., 2002; St Denis & Cappuccino, 2004).

Figure C Map of sampling sites used for collection of field soils. Collections sites are indicated by solid triangles. Site names and coordinates are from left to right: Toronto Zoo (43.815721, -79.187969), (Crows Pass Conservation Area; 44.037153 -79.042661), (Orono Crown Lands; 43.978888 -78.64607), and (Rice Lake Conservation Area; 44.083999 -78.300137). Cities are given as landmarks (circles). Created with SimpleMappr, http://www.simplemappr.net..

Collection sites were all within the Mixed-wood Plains Ecozone, Ecoregion 6E (Crins et al., 2009) on calcareous parent material (Middle-Upper Ordovician limestone/dolostone/shale) (Ontario Geological Survey, 1991). All sites had similar plant communities consisting of a mixture of native and non-native species. Plants that were dominant in uninvaded communities also tended be present in invaded communities, albeit at lower abundances (see Appendix B). Field soils at all collection sites were sandy clay-loam to loam in texture and of similar fertility (see Appendix C).

In May 2016, soil was collected at each of the four sites from dense V. rossicum populations (75-100% cover) and adjacent uninvaded plant communities, to ensure broad representation of V. rossicum-invaded and uninvaded rhizosphere biota. Specifically, the upper 5cm of soil was collected to acquire rhizosphere biota representative of rhizosphere communities that would be initially encountered by incoming seedlings. Soil was collected from 8-10 randomly selected adult V. rossicum plants in invaded areas, and 8-10 randomly located soil plugs in uninvaded areas (20 cm wide x 5 cm deep). Soil collected from the uninvaded plant communities was at least 2-4 m from the edge of a V. rossicum population and at least 1m from any adult V. rossicum. Care was taken to sterilize tools with 10% bleach to avoid contamination between invaded and uninvaded soil samples. Soils were placed separately into coolers for transportation back to Algoma University (1-2 days) where they were sieved (4-mm) to remove roots, rocks and macroinvertebrates.

Soils across sites were pooled according to invasion history and stored at 10oC for six weeks prior to the onset of the experiment. Although pooling soil among sites may overestimate local effects of rhizosphere biota (Reinhart & Rinella, 2016), the objective of this study was to test for a potential biotic effect of V. rossicum on native legumes via changes in rhizosphere communities. This objective therefore required wide representation of the rhizosphere biota present in soils either invaded by V. rossicum, or yet to be invaded (i.e. uninvaded). Pooling soil in this manner is appropriate to answer the questions in this study (see Karst et al. 2015; Cahill et al. 2016; but also see Reinhart & Rinella 2016).

3.3.2 Experimental Design and Growth Conditions

The experiment consisted of a completely randomized fully crossed design with three main factors and eight replicates per treatment (N=8). The three main factors were ‘Rhizosphere biota’, 'Invasion' and ‘Legume Species'. ‘Rhizosphere biota’ tested the overall effects of rhizosphere biota using live and sterile treatments. ‘Invasion’ tested the relative effects of V. rossicum invaded soils and consisted of invaded and uninvaded treatments. ‘Legume Species’ tested the effects of ‘Rhizosphere biota’ and ‘Invasion’ on four native legume species (Desmodium canadense, Lespedeza hirta, Lupinus perennis and Astragalus canadensis; Table 3.1) and V. rossicum, for a total of 160 experimental units.

Treatments consisted of a 2:1 mixture of Turface (a montmorillonite clay, Turface Athletics MVP, Profile Products LLC, Buffalo Grove, IL, USA) and non-calcareous granitic Sand (Hutcheson

Sand and Mixes, Huntsville, ON), combined with 15% (by volume) ‘invaded’ or ‘uninvaded’ field soil that either contained rhizosphere biota (‘live’ treatments) or did not (‘sterile’ treatments; field soil twice autoclaved at 121° C for one-hour each cycle).

Clean (10% bleach) Round Standard Pots (6”, 1.5L; Stuewe and Sons Inc., Corvallis, OR, USA) were filled with the experimental substrates (1L; 1200g). Pots were placed on individual saucers to minimize cross contamination and nutrient losses through leaching. The four legumes and V. rossicum were grown in experimental substrates of each treatment combination (live-invaded, live-uninvaded, sterile-invaded, sterile-uninvaded). Twelve of the 160 plants died before harvest. As a result, a total of 148 experimental units were used for analysis. Corrections for lost units were not needed because these were evenly distributed across treatments (see Appendix H).

Table 3.1 A list of speciers used in the study with taxonomic designation, common name and abbreviation Species name Abbreviation Common name Desmodium canadense (L.) de Candolle DeCa Showy tick-trefoil Lespedeza hirta (L.) Hornemann LeHi Hairy bush-clover Lupinus perennis L. LuPe Perennial Lupine Astragalus canadensis L. AsCa Canada milk-vetch Vincetoxicum rossicum (Klepow) Barbaricz ViRo Dog-strangling vine

To test whether the effects of V. rossicum-invasion on D. canadense were unique to this species or common among the Fabaceae, a phylogenetic range of native legume species were selected (Figure 3.2). Preference was given to perennial native species listed on three Ontario plant lists: Southern Ontario Vascular Plants List (Bradley 2013), Ontario Tallgrass Prarie and Local Toronto Plants (Limestone Creek Restoration Nursary, 1996; Tall Grass Ontario, n.d.).

Astrales (Asterids I)

Fabales (Rosids)

Figure 3.2 Phylogenetic tree showing evolutionary relationships between native legumes used to test effects of rhizosphere biota from V. rossicum invaded and uninvaded soils. The horizontal lines are branches that represent genetic change; the longer the branch the greater the genetic distance to V. rossiucm. Native species are in labeled in black. Vincetoxicum rossicum is labled in red. Additional information of order and broad group is given on the right. Constructed using phyloT: Phylogenetic Tree Generator, 2016. http://phylot.biobyte.de/.

Seeds of native legumes were acquired from commercial sources (Ontario Seed Co., Limited, Waterloo, Ontario) and were representative of US Midwest genetic stock. Seeds of V. rossicum were field-collected in October 2015 from Crows Pass Conservation area stored in dry conditions at 4oC until use (1 year). All seeds were surface sterilized in 10% bleach for 5 min and thoroughly rinsed in deionized water for 10 minutes. Seeds were cold-wet stratified in germination trays on sterile vermiculite for 1 week at 4oC to break dormancy. Trays were sealed to maintain a high humidity (80-90%). Seeds germinated over the course of 1-2 weeks at 23oC in the germination trays and seedlings were then transplanted (1-2 days after germination), into the experimental substrate of each respective treatment. Four seedlings were planted into each pot and thinned to one individual after one week of growth. Some seedlings died soon after transplantation, possibly because of damage suffered during transplantation. Seedlings were replaced with excess seedlings (maintained in germination trays) in the first week of the experiment.

Plants were grown for 13wks under controlled conditions (June to September 2016), in a walk-in growth chamber at the Ontario Forest Research Institute in Sault Ste. Marie (Ontario Ministry of Natural Resources and Forestry). The growth chamber was set to 65% humidity, 21oC/18oC day/night temp with 15 hr day length, at 200-250 mol light intensity. These light levels would be considered high/low relative to shaded forest understory (Darren Derbowka, Growth Facilities Coordinator. OFRI, Sault Ste. Marie. per. communication), and thus experimental conditions were representative of seedling growth in treefall gaps or forest edges, typical of V. rossicum populations (Cappuccino, 2004; Cappuccino et al., 2002; St Denis & Cappuccino, 2004).

Plants were watered to field capacity every 2-3 days (200-300 ml) or as necessary to prevent soil drying. All pots were fertilized with 50mL (0.48g/L) of Miracle-Gro 24:8:16 (The Scotts Company LLC, Mississauga, ON, Canada) every two weeks to a total of 25mg N, 3.6mg P, and 13.8mg K per pot (for details see Chapter 2, section 2.3.2: Experimental Design and Growth Conditions).

An electrical fire outside the growth chamber occurred half way through the experiment (wk 7). This caused the chamber to fill with smoke. Plants were removed from the chamber as soon as possible. Plants were maintained under reduced light (50-90umol) conditions for 2-3wk before original growth conditions could be restored. Height was measured immediately after the fire and at the end of the experiment. The overall trends of the treatments (observed in height) did

not differ between 7wk and 13wk (Appendix I); final measurements (@13wk) were conserved for analysis.

3.3.3 Harvest and Response Variables

Fresh leaf samples (2-4 leaves) from each plant were removed at the end of the experiment. Leaves were scanned using WinFlora Pro (2016) scanning software (Regent Instruments Canada Inc.) system with the Epson Expression 10000 XL scanner, and specific leaf area (mm2/mg) was calculated. Reproductive output (seeds) could only be measured in Desmodium canadense for which seed pods and seeds were counted. Roots and shoots were separated at the soil surface. Roots were gently removed from the experimental substrate and washed in cool water to remove excess media. Roots were stored in individual sterile pots containing DI water and stored at 4oC until they could be processed. Root nodules were counted on all species and assessed based on size (small=1-5mm, medium=5-10mm, and large 10mm+), location (within 2cm of stem, or not) and colour (pink or brown) (Greenwood, & Pankhurst, 1977; Shelby et al. 2016). Pink nodules or nodules with pink centres are indicative of N fixation activity (Greenwood, & Pankhurst, 1977; Shelby et al. 2016) and served as a proxy of mutualistic relationship between plant and root-nodule bacteria. To assess root architecture roots and count nodules were scanned with the Epson Expression 10000 XL scanner. Nodules were counted using the Cell Count plug-in in Fiji, an open source platform for biological-image analysis (Schindelin et al., 2012). Root architecture was analysed using WinRhizo Pro (2016) (Regent Instruments Canada Inc.).

Random root samples were taken, weighed and stored at -80oC for future molecular analysis of soil microbial community (analysis not included). Plant material was dried for three days at 60°C and shoot, root and total biomass were measured. Percent moisture of oven-dried root was used to calculate weights of root samples removed for molecular analysis, and total dry root weight.

3.3.4 Statistical Analysis

To verify that biomass production was indeed the result of rhizosphere biota, effects of live and sterile treatments were analysed separately using factorial ANOVAs (with ‘Legume Species’

(including V. rossicum, N=5) and 'Invasion’ (invaded and uninvaded) as fixed factors. Data were ln transformed to meet the assumptions of normality and stabilize the residual variance.

Once it was confirmed that soil sterilization reduced the effect of invasion, ‘Rhizosphere biota’ and ‘Invasion’ factors were combined to calculate the Net Biotic Effect (NBE) of invasion. This allowed the specific effects of rhizosphere biota to be observed, separate from abiotic factors (e.g. Non-significant variations in soil nutrients). Values for NBE were randomly selected (95% confidence intervals) from bootstrapped ratios (Live: Sterile) of total, shoot and root biomass respectively (999 iterations; R version 2.11.1: R CoreTeam 2011). All NBE values were natural log transformed to meet the assumptions of normality and stabilize the residual variance.

The effect of invasion on native legumes (i.e. excluding V. rossicum) was analysed using a factorial ANOVA with ‘Legume Species’ (N=4) and ‘Invasion’ (i.e. NBE invaded, NBE uninvaded; N=2) as fixed factors. Following significant ‘Legume Species X Invasion’ interactions, separate 2-tailed t-tests were performed on each species to test how invasion influenced responses. Vincetoxicum rossicum was assessed separately using t-tests.

Root nodules, and seeds counts (D. canadense only), were square root transformed (count + 0.01) (O ’Hara & Kotze, 2010) to meet the assumptions of normality and stabilize the residual variance. Equality of variance was tested using “F-test Two Sample for Variances” tests in the Excel ‘Data Analysis’ package (Microsoft Office 365 Excel, version 1708, Microsoft Corp.). T- tests for independent samples were used to test for differences between plants, of a given species, grown in invaded compared to uninvaded treatments.

Statistical tests were conducted in Statistica (Dell STATISTICA, Dell Inc. 2015. version 13. software.dell.com) and statistical results of t-tests were reported assuming equal variances - unless otherwise specified.

3.4 Results

3.4.1 Effect of Invasion on Native Legume Biomass

In sterilized soil treatments biomass was unaffected by ‘Invasion’. However, in live soil treatments, there were significant species-specific responses to invasion for shoot, root and total biomass (see species-invasion interaction (Table 3.2; Figure 3.3).

12 Live (invaded) Live (uninvaded)

10 Sterile (invaded) Sterile (uninvaded) 8

6 Biomass (g) Biomass 4

DeCa LeHi LuPe AsCa ViRo Figure 3.3 The total biomass produced by four native legumes when grown with and without rhizosphere biota (Live vs. Sterile treatments) from soils invaded and uninvaded by Vincetoxicum rossicum. Bars represent mean total biomass (± SE). Species abbreviations are: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa).

Table 3.2 Analysis of variance (ANOVA) showing the effect of ‘Species’ and ‘invasion’ on the plant biomass of native legumes and Vincetoxicum rossicum produced in Live and Sterile treatments. Species x Invasion effects from Net Biotic Effect (NBE) was calculated for native legumes only; V. rossicum was excluded from NBE. Legume Species1 Invasion2 Species X Invasion1 Response F P F P F P Total Biomass Sterilea 10.46 <0.001 1.01 0.320 0.54 0.709 Liveb 36.24 <0.001 0.49 0.486 3.79 0.008 Net Biotic Effectc 98.96 <0.001 0.05 0.832 22.87 <0.001 Shoot Biomass Sterilea 5.46a 0.001 1.63 0.206aa 0.56 0.692 Liveb 38.78b <0.001 0.08 0.772bb 3.86 0.007 Net Biotic Effectc 95.50c <0.001 2.21 0.143cc 31.21 <0.001 Root Biomass Sterilea 21.40 <0.001 0.66 0.418 0.59 0.670 Liveb 40.14 <0.001 1.45 0.233 3.57 0.001 Net Biotic Effectc 76.02 <0.001 6.58 0.013 25.08 <0.001 (Degrees of freedom, Error): a1 =(4,63); a2 =(1;63); b1=(4,65); b2 =(1,65); c1=(3,49); c2=(1, 49); Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

Rhizosphere biota generally had a significant effect on the total biomass of 3/4 native legume species tested (D. canadense, A. canadensis and L. hirta), as well as a significant effect on the root biomass of all the native legumes tested (Table 3.3). Whether legumes were promoted, inhibited or unaffected by rhizosphere biota depended on plant species identity and the invasion status of the rhizosphere biota that plants were grown with. Accordingly, there was a significant interaction between ‘Legume Species’ and ‘Invasion’ for Net Biotic Effect (NBE) in all biomass measurements (Total, Shoot and Root; Table 3.2; Figure 3.4).

Total biomass of D. canadense and L. perennis was positively affected by invaded-rhizosphere biota but unaffected by uninvaded-rhizosphere biota, as such there was a significant difference between invaded and uninvaded treatments in total biomass produced by these two species (see NBE Table 3.4; Figure 3.4a). Indeed, D. canadense produced an average of 3.9g (~70%) more total biomass when grown in live-invaded treatments compared to live-uninvaded treatments and L. perennis produced an average of 0.6g (~43%) more total biomass in live- invaded compared to uninvaded treatments (Figure 3.3). Shoot and root biomass of D. canadense showed similar patterns of significant growth promotion due to invaded-rhizosphere biota (and no effect from uninvaded-rhizosphere biota). Conversely, only L. perennis root (and not shoot) biomass was affected by the ‘invasion’ status of rhizosphere biota, whereby root biomass was promoted by invaded-rhizosphere biota and inhibited by uninvaded-rhizosphere biota resulting in a significant difference between the two treatments. (Table 3.4; Figure 3.4b&c).

In contrast to D. canadense and L. perennis, L. hirta biomass was significantly inhibited by invaded-rhizosphere biota but was not affected by uninvaded-rhizosphere biota. Thus, L. hirta biomass production was more strongly (and negatively) affected by rhizosphere biota from V. rossicum soils than rhizosphere biota from uninvaded soils (Table 3.4; Figure 3.4).

Shoot, root and total biomass of Astragalus canadensis responded positively to rhizosphere biota overall (Table 3.3), but total biomass production did not differ between invaded and uninvaded treatments (Table 3.4; Figure 3.4a). Shoot and root biomass responses to rhizosphere biota were inconsistent with total biomass response. Both shoot and root biomass production were significantly greater in response to rhizosphere biota in uninvaded compared to invaded treatments (Table 3.4; Figure 3.4b&c).

V. rossicum shoot, root and total biomass responded negatively to rhizosphere biota overall. The inhibition of biomass production was greater in response to its own (invaded) rhizosphere biota compared to the rhizosphere biota from uninvaded soils (Table 3.4; Figure 3.4).

1.5 (a) invaded uninvaded 1 *** 0.5 ** ** ***

0 Biomass -0.5

Net Biotic Effect on Total Total on Effect Biotic Net -1

-1.5 DeCa LeHi LuPe AsCa ViRo

2 (b) * 1.5 *** 1 0.5 ** **

0 Biomass -0.5

Net Biotic Effect on Shoot Shoot on Effect Biotic Net -1

-1.5 DeCa LeHi LuPe AsCa ViRo

2 (c) ** 1.5 1 *** 0.5 ** *** ***

0 Biomass -0.5 -1

Root on Effect Biotic Nebt -1.5 -2 DeCa LeHi LuPe AsCa ViRo Figure 3.4 The net biotic effect (NBE) of rhizosphere biota on legume total (a), shoot (b) and root (c) biomass from soils invaded (dark gray) and uninvaded (light gray) by V. orssicum. Bars represent mean (±SE) NBE; positive values indicate biomass promotion by rhizosphere biota, and negative values indicate biomass inhibition. Species abbreviations are: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa). Asterisks show significant (P<0.05) from independent t-test results for ‘Invasion” on natrual log transformed data: ***P<0.001, **P=0.01, * P=0.05.

Table 3.3 One sample t-test showing wether the Net Biotic Effect (NBE) of rhizosphere biota on total, shoot and root biomass deviated from zero. Positive t-values indicate biomass promotion by NBE, negative t-values indicate biomass inhbition by NBE, values nearest zero indicate nutral effects. P values <0.05 show significant differences from zero (i.e. rhizosphere biota did not significantly affect plant productivity). Analysis conducted on the combined NBE of rhizosphere biota from invaded and uninvaded soils (log transformed). Response Variable N M SE t-value df P Difference from zero NBE of Total Biomass D. canadense 16 0.434 0.090 4.800 15 0.000 L. hirta 11 -0.558 0.191 -2.928 10 0.015 L. perennis 15 0.081 0.051 1.574 14 0.138 A. canadensis 15 1.110 0.059 18.909 14 0.000 V. rossicum 15 -0.336 0.033 -10.087 14 0.000 NBE of Shoot Biomass D. canadense 16 0.558 0.108 5.175 15 0.000 L. hirta 11 -0.426 0.200 -2.126 10 0.059 L. perennis 15 0.128 0.053 2.416 14 0.030 A. canadensis 15 1.237 0.086 14.327 14 0.000 V. rossicum 15 -0.486 0.044 -11.154 14 0.000 NBE of Root Biomass D. canadense 16 0.406 0.066 6.109 15 0.000 L. hirta 11 -0.728 0.259 -2.814 10 0.018 L. perennis 15 0.031 0.086 0.362 14 0.723 A. canadensis 15 1.099 0.097 11.362 14 0.000 V. rossicum 15 -0.423 0.042 -10.042 14 0.000 M=mean; SD= standard deviation; df = group degrees of freedom; t-value = t statistic assuming equal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07

Table 3.4 Independent t-tests statistics showig the effect of ‘invasion’ on the net biotic effect on plant biomass for each species (legumes: Desmodium canadense (DeCa), Lespedeza hirta (LeHi), Lupinus perennis (LuPe), Astragalus canadensis (AsCa), and the invasive vine: Vincetoxicum rossicum). Analysis conducted on natrual log transformed data. Invaded uninvaded Net Biotic Effect of M SD M SD df t-value P

Total Biomass D. canadense 0.76 0.08 0.11 0.16 14 10.27 <0.0001 L. hirta -0.97 0.40 -0.06 0.49 9 -3.41 0.008 L. perennis 0.20 0.14 -0.06 0.17 13 3.22 0.007 A. canadensis 1.08 0.14 1.14 0.31 13 -0.52 0.611 V. rossicum -0.44 0.07 -0.22 0.07 13 -6.20 <0.0001

Shoot Biomass D. canadense 0.96 0.08 0.15 0.14 14 14.11 <0.0001 L. hirta -0.90 0.48 0.14 0.27 9 -4.33 0.002 L. perennis 0.21 0.18 0.03 0.20 13 1.77 0.100 A. canadensis 1.08 0.22 1.42 0.36 13 -2.26 0.041 V. rossicum -0.59 0.07 -0.37 0.17 13 -3.41 0.005

Root Biomass D. canadense 0.62 0.08 0.19 0.19 14 5.86 <0.0001 L. hirta -1.32 0.53 -0.01 0.57 9 -3.95 0.003 L. perennis 0.27 0.19 -0.24 0.23 13 4.81 <0.0001 A. canadensis 0.87 0.20 1.36 0.37 13 -3.29 0.006 V. rossicum -0.56 0.07 -0.26 0.03 13 -9.96 <0.0001 M=mean; SD= standard deviation; Ndf = group degrees of freedom; tdf = adjusted degrees of freedom for unequal variances; t- value = t statistic assuming equal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07

3.4.2 Effect of Invasion on Desmodium canadense Reproductive Potential

There were significantly more reproductive structures on D. canadense in live (mean=7.55, SD=6.42) compared to sterile soil (mean=1.55, SD=2.87; P=0.037) (Figure 3.5). However, there were no significant differences between live-invaded and live-uninvaded treatments for any reproductive structure measured nor for seed weight (Figure 3.6; Figure 3.7; Table 3.5).

160 Invaded Uninvaded 140 ***

120

100

Number of Structures of Number 60

0 All Live SterileAll

Figure 3.5 The total number of reproductive structures (combined buds, flowers and seeds) produced by Desmodium canadense plants when growth with and without soil biota (live vs. sterile) from soils invaded (dark gray) and uinvaded (light gray) by V. rossicum. Bars represent mean (± SE) number of total reproductive structures per plant. Independent t-tests were used to detect significant differencences between treatments. Analysis was performed on square root transformed data (+0.1). Asterisks indicate significant differences between Live and sterile (long horrizontal lines) at P<0.05 (*** P<0.001, ** P=0.01, * P=0.05).

Invaded Invaded 160 Uninvaded 1800 Uninvaded 140 1600 120 1400 1200 100 1000 80 800 60 (g) Weight 600

Number of Structures of Number 40 400 20 200 0 0 Bud Flower Seed SeedsSeed Figure 3.6 The number of buds, flowers and seeds Figure 3.7 The weight of seeds produced by produced by Desmodium canadense in response to Desmodium canadense in response to rhizosphere biota rhizosphere biota (i.e. live treatments) from Vincetoxicum (i.e. live treatments) from Vincetoxicum rossicum rossicum invaded and uninvaded soils. Bars represent mean invaded and uninvaded soils. Bars represent mean seed (± SE) number of buds, flowers and seeds produced per weight (± SE) per plant. plant.

Table 3.5 Independent t-tests showing the effect of ‘Invasion’ (rhizosphere biota from Vincetoxicum rossicum invaded vs. uninvaded soils) on the reproductive potential of Desmodium canadense including bud, flower, seed, seed weight and all reproductive structures (combined bud, flower and seed).

Invaded uninvaded Variable M SD M SD df t-value p Bud 2.27 2.41 2.19 2.18 14 .068 .947 Flower 1.52 1.32 1.03 1.02 14 .836 .417 Seed 5.50 7.48 3.96 4.51 14 .498 .626 Seed Weight 1.11 4.74 1.32 3.94 14 -.098 .924 All 7.55 6.42 5.60 3.83 14 .738 .473

3.4.3 Effect of Invasion on Nodule Formation

Few nodules occurred on plants grown in sterilized treatments (overall Mean <1). There was one exception, one D. canadense plant grown in sterile uninvaded treatment (128 nodules). However, this did not influence the results; univariate tests showed the effect of ‘Invasion’ and the interaction between ‘Species and invasion” were statistically significant only in live treatments (Table 3.6).

Desmodium canadense and L. hirta had significantly more nodules when grown with live rhizosphere biota from invaded versus uninvaded treatments (Table 3.7; Figure 3.8). Conversely, A. canadensis had a marginally significant greater number of nodules in uninvaded than in invaded treatments. Nodule formation in L perennis did not differ significantly between treatments (Table 3.7; Figure 3.8).

Nodule morphology varied depending on plant species. Most were active (i.e., pink-red inside), and ranged from 1mm to 10mm in size. D. canadense was the only species containing inactive nodules (i.e., brown inside). The number of brown nodules formed in live invaded treatments (N=8, transformed M=2, SD=1.78) was not statistically different from that in live uninvaded treatments (N=8, transformed M=1, SD=1.07; t=1.696, df=14, P=0.112).

Table 3.6 Analysis of variance (ANOVA) statistics showing the effect of ‘Invasion’ and ‘ Species’ on nodule formation in native legumes with rhizosphere biota (live) compared to the control (sterile/without biota). Analysis was conducted on square root transformed (0.1+count) data. Sterile Live Response F P F P Legume Speciesa 5.456 0.003 44.417 <0.001 Invasionb 1.376 0.246 13.269 0.001 Sp.X Invasiona 0.645 0.589 16.248 <0.001 error 50 51 Degrees of freedom: a=3, b=1. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

Invaded 300 Uninvaded *** 250

200

150

100 Number of Nodules of Number

50 **

0 AsCa DeCa LeHi LuPe

Table 3.7 Independent t-test showing the effect of rhizosphere biota from soils invaded and uninvaded by Vincetoxicum rossicum (effect of ‘invasion’) on nodule formation in four legume speciers. Analysis was conduced on square root (+0.1) transformed data. Invaded Uninvaded Species M SD M SD t- value df sig (2 tail) D. canadense 14.22 3.11 5.37 1.65 7.11 14 <0.001 L. hirta 2.58 1.39 0.32 0.00 3.98 10 0.003 L. perennis 2.66 3.82 5.13 5.22 -1.06 13 0.310 A. canadensis 8.34 4.66 12.28 3.27 -1.96 14 0.070 M=mean; SD= standard deviation; df = group degrees of freedom; t-value = t statistic assuming equal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07

3.5 Discussion

3.5.1 The Effect of Vincetoxicum rossicum Rhizosphere Biota on Native Legumes Productivity

The results of this study show that rhizosphere communities from areas invaded by V. rossicum have differential effects on legume plant biomass. Contrary to predictions, legume biomass was not generally promoted by rhizosphere biota from V. rossicum invaded soil (Figure 3.4). There was a strong species-specific response to invasion, but only D. canadense and L. perennis responded as hypothesized; producing more total biomass in response to invaded-rhizosphere biota compared to uninvaded-rhizosphere biota (Figure 3.4). The remaining legume species were unaffected by the invasion status of rhizosphere biota (A. canadensis) or were inhibited by invaded-rhizosphere biota (L. hirta).

The response of D. canadense to biota from invaded soil was consistent with the results of the previous study (Chapter 2), where D. canadense was the only plant species whose biomass was significantly promoted (P<0.001) by invaded-rhizosphere biota (see NBE Figure 2.5a; Table 2.3). The effects of invasion were noticeable even when live treatments were considered on their own (i.e. the invasion status of the inoculum had little effect on native plant biomass when soils were sterilized but ‘invasion’ was significant on native plant biomass in live soils; Table 2.3); on average D. canadense produced 74% (5.7g) more total biomass in live-invaded versus in live-uninvaded treatments (Figure 2.4). In the present study, D. canadense also produced greater amounts of total biomass (on average 3.9g; 70%) in live-invaded than in live-uninvaded treatments. Significant differences in Net Biotic Effect (NBE) between invaded and uninvaded treatments confirm that the greater biomass associated with invaded treatments was largely due to rhizosphere biota that was specific to the invaded treatment (Figure 3.4). The only other legume to benefit significantly more from invaded-rhizosphere biota than uninvaded-rhizosphere biota (L. perennis; see NBE Figure 3.4) did not exhibit as noticeable a difference in live treatments as D. canadense; i.e. total biomass of L. perennis in live-invaded treatments weighed on average 0.6g (43%) more than live-uninvaded treatments. The similarities between the response of D. canadense to rhizobia in the previous and present studies (Chapter 2 and 3) occurred despite differences in experimental design. In the previous study (Chapter 2), rhizosphere soils were collected in August, while in the present study, soils were collected at the same location in early May of the following year.

Significant changes in microbial abundance and diversity have been shown to accompany distinct seasonal shifts between warm and cold weather (Lipson & Schmidt, 2004; Schadt, 2003). For example, bacterial biomass is thought to peak in spring to summer, as the bacteria feed on plant exudates produced when the plants are actively growing, while fungal biomass is thought to peak in summer-fall when complex plant residues are most abundant (Bardgett et al., 2005). However, seasonal dynamics are still not fully understood. Davison et al. (2012) looked at AMF colonization in forest soils throughout the growing season (in Northern Hemisphere: May-September), but no temporal differences were observed. Improved DNA-based identification methods are expanding the capacity to research temporal shifts in soil microbial communities and improve our understanding of how soil communities respond to changes in climate and plant productivity within and between seasons (Cross et al., 2011; Wingfield et al., 2011).

The persistent positive response that D. canadense had to rhizosphere biota despite seasonal differences in inoculum collection between studies (Chapter 2 vs. 3; May vs. August; spring vs. late-summer) is indicative of a strong and persistent relationship between D. canadense and certain rhizosphere biota found in V. rossicum soils. Interestingly, invaded rhizosphere communities do not appear to benefit V. rossicum. In the present study, V. rossicum biomass was generally inhibited by rhizosphere biota. The negative net biotic effect of biota on V. rossicum biomass was strongest in response to its own (invaded) biota (Figure 3.4), which contributed to greater biomass productivity in live-uninvaded compared to invaded (Figure 3.3). In the previous study V. rossicum was generally promoted by rhizosphere biota, but did not benefit more from its own (invaded) rhizosphere community relative to the uninvaded rhizosphere community. This response by V. rossicum was reflected by the absence of significant differences in biomass production between invaded and uninvaded treatments (Chapter 2; Figure 2.5). These results from the previous study correspond with the findings of Day, et al. (2015), currently the only other study to investigate feedback from invaded and uninvaded soil on V. rossicum biomass production. In their study, Day,et al.( 2015) found that V. rossicum produced more biomass in live compared to sterile controls, but there were no significant differences in biomass, root to shoot ratios and seed production between invaded and uninvaded treatments. Collectively, the research presented here and that of (Day et al., 2016; Day, Dunfield, Antunes, 2015) indicate that V. rossicum does not receive additional benefit from growing in its own soil community.

The species identity of mutualistic and antagonistic rhizosphere biota was not investigated in the present study. However, there were significantly more nodules on D. canadense plants grown in invaded treatments than in uninvaded treatments containing live inoculum (P<0.001; Figure 3.8). Elevated numbers of nodules indicate that nitrogen-fixing bacteria likely played a role in the positive response D. canadense had to V. rossicum rhizosphere biota. Furthermore, over 90% of nodules on D. canadense in live treatments were active (Greenwood, & Pankhurst, 1977; Shelby et al. 2016). Two of the three other legumes tested in the present study (L. hirta, L. perennis – but not A. canadensis) also had higher numbers of nodules in invaded versus uninvaded treatments (significant for L. hirta) (Figure 3.8). These numbers were small in comparison to D. canadense, which on average had 211 nodules in invaded treatments whereas L. hirta and L. perennis had 8 and 21 nodules, respectively. In addition, D. canadense was the only species that had both a higher number of nodules and biomass in invaded relative to uninvaded treatments. Lupinus perennis grew more in live- invaded compared to uninvaded treatments, but the number of nodules did not differ between these treatments. Lespedeza hirta had more nodules when grown in live- invaded verses uninvaded treatments, but this corresponded with an inhibition of biomass rather than a promotion (Figure 3.3). Nodules on D. canadense roots in the previous study (Chapter 2) were also higher on plants grown in invaded treatments than uninvaded treatments containing live inoculum. Collectively, these results support the notion that D. canadense is better able to form symbiotic relationships with root- nodule bacteria in invaded treatments than other legumes.

Associations with specialist root-nodule endophytes could help explain why D. canadense was positively affected by V. rossicum’s rhizosphere biota, while the other legumes did not generally benefit from the same interaction. Desmodium spp. form mutualistic associations with root- nodule bacteria mostly in the Bradyrhizobium group (Parker et al., 2015). Associations with other clades of nitrogen-fixing bacteria within the polyphylogenetic group Rhizobia (specifically: Rhizobium and Mesorhizobium) are also likely (Gu et al., 2007; Tlusty et al., 2004). Species- specific interactions between root-nodule bacteria and their hosts is a common phenomenon among the Fabaceae family (Ehinger et al., 2014; Parker et al., 2015), although this is highly dependent on which phylogenetic clade individual species belong to within Fabaceae. The species in the sub-family Papilionoideae, including the legumes used in the present study, are generally more promiscuous than other sub families among the Fabaceae, particularly those in the Desmodieae tribe (Andrews & Andrews, 2017). Two species in the present study belong to the Desmodieae tribe (D. canadense and L. hirta). Yet the responses of these two species to

rhizosphere biota were in opposition to one another; suggesting that the biota or biotic interactions which were benefiting D. canadense are host-specific.

It is unclear why V. rossicum would increase the abundance of rhizosphere biota specialist on D. canadense and/or amplify the relationship between D. canadense and its specific rhizosphere biota. Some studies have shown higher numbers of ammonia-oxidizing bacteria (AOB) in association with many invasive species of different herbaceous forms, such as grass, shrubs and trees, which can result in an increase in the nitrogen availability in the soils these invasive species (McLeod et al., 2016; Rodrigues et al., 2015). Vincetoxicum rossicum may associate with AOB like other invasive species trees (e.g. McLeod et al., 2016; Rodrigues et al., 2015), but investigations into V. rossicum interactions with bacteria have been restricted to the antibiotic properties of V. rossicum (Mogg et al., 2008). In the present study, only total nitrogen concentrations were measured, and no differences were found between V. rossicum-invaded and -uninvaded areas (see Appendix C - Table C.2 & Table C.3). In addition to rhizobia, legume nodules have been found to be colonized by a variety of other bacterial endophytes e.g. actinobacteria and firmicutes (Palaniappan, et al., 2010) and fungal endophytes (e.g., AMF) (Dudeja et al., 2012). Unlike the legume-rhizobia symbiosis, which is well studied (Oldroyd et al., 2011; Oldroyd, & Downie, 2008, 2004), very little is known about other organisms housed in root nodules, i.e. root-nodule endophytes (reviewed by Velázquez et al., 2017). Nodule endophytes may directly benefit their host, or may act as helper bacteria to enhance the effects of rhizobia (Dudeja et al., 2012). Nodule endophytes of D. canadense and multipartite interactions between V. rossicum and rhizosphere biota have not been considered in past research, thus future molecular analysis of the samples collected in this study will represent the first steps in this direction.

Fungal endophytes including AMF have also been observed in root nodules as well as the roots themselves, although it is unclear if their role in the nodule differs from their role in the rest of the root (Dudeja et al., 2012). Arbuscular MF are well known to impart positive effects on their hosts including increasing nutrient availability, particularly phosphorus, pathogen protection and other benefits that promote host fitness (Philippot et al., 2013). Recently, the community of fungal root colonists associated with V. rossicum has received renewed attention (Bongard et al., 2013; Day, et al., 2015; Day et al., 2016; Day, et al., 2015; Sanderson et al., 2015). These studies show that V. rossicum is highly mycorrhizal, associating with a wide diversity of AMF (Bongard et al., 2013; Day, Dunfield, Antunes, 2015; Day et al., 2016; Smith et al., 2008).

In the previous study (Chapter 2), D. canadense had greater hyphal colonization in live- invaded compared to uninvaded treatments which corresponded to a relative increase in biomass (Figure 2.9). Conversely, AMF colonization of V. rossicum did not differ between invaded and uninvaded treatments and this was reflected by non-significant differences in biomass (Figure 2.9). In the present study, although mycorrhizal colonization was not measured, it should be noted that V. rossicum experienced negative feedback. This contrasts with the work by Day, et al. (2015) which showed that differences in AMF community composition between invaded and uninvaded soil corresponded to positive feedback. Although, the rhizosphere fungal community of V. rossicum has been identified (Bongard et al., 2013; Day, Antunes, Dunfield, 2015; Day et al., 2016; Day, Dunfield, Antunes., 2015), the bacterial community has received less attention. Moreover, the microbial community D. canadense specifically has not been studied. The members of the tribe to which D. canadense belongs (Desmodieae) are generally promiscuous in their associations with rhizobia (Andrews & Andrews, 2017). However, the unique and consistent positive effect that to V. rossicum rhizosphere biota had on D. canadense biomass production suggests D. canadense may form specific interactions rhizosphere biota associated with V. rossicum. Multipartite symbioses and/or interactions are not well understood but may nevertheless be important. Indeed, Day, et al. (2016) found that certain V. rossicum root fungi had antagonistic effects on the biomass of the native plant Solidago canadensis but only when the fungal antagonists were applied together.

Legumes have been shown to yield up to 15 times more biomass when grown with both AMF and rhizobia compared to either symbiont alone (van der Heijden et al., 2015). However, it is unknown how consistently this phenomenon occurs in nature (Bauer et al., 2012; Larimer et al., 2010). Root-nodule bacteria convert dinitrogen into ammonium (a plant available form of nitrogen) and the fine hyphae of AMF improve host access to phosphorous pools in the bulk soil which are inaccessible to the larger plant roots. Since both phosphorous and nitrogen are limiting elements for the host plant, the plant benefits when symbiotic relationships are formed with both types of symbionts. Furthermore, improved host performance feeds back to benefit both symbionts, thus the presence of one symbiont can enhance the performance of the other (van der Heijden et al., 2015)

If V. rossicum amplified certain species of rhizosphere biota that enhanced the root-nodule bacteria of D. canadense either directly as helper bacteria or indirectly in tripartite synergistic relationships (AMF + root-nodule bacteria + host plant) this might explain why D. canadense grew more in live- invaded compared to -uninvaded treatments, and why V. rossicum and the

other legumes tested in this study did not similarly benefit from invaded rhizosphere biota. Further research is needed to determine the mechanisms behind the relationship between V. rossicum and D. canadense.

The consistent response of D. canadense to V. rossicum rhizosphere biota in the present and previous (Chapter 2) studies indicate that D. canadense could be an important species for use in restoration and reclamation programs following removal of V. rossicum. If D. canadense is promoted by V. rossicum rhizosphere biota under controlled conditions, then it might also be promoted by V. rossicum in the field, which could prevent the re-establishment of V. rossicum and improve the establishment of native species. Recent empirical evidence has found that some legumes may be able to improve establishment of native species by increasing nutrient availability and improving colonization by beneficial symbionts such as AMF (Mummey & Ramsey, 2017). Thus far the ability of legumes to supress negative soil feedback effects from invasive species and improve native species establishment have received little attention (Eviner & Hawkes, 2008). Indeed, the use of legumes to combat invasive species needs be approached with caution because some native legumes may promote invasive species (specifically legume species). Desmodium canadense was shown to promote nodulation in Cytisus scoparius (an invasive legume) during a common garden experiment, likely because the root-nodule bacteria Bradyrhizobium spp. passed from D. canadense to C. scopariusa (Parker et al., 2006). Thus, it is important to consider effects of D. canadense on V. rossicum, as well as other species (native and non-native) in the area intended for remediation, to fully assess the potential of D. canadense to mitigate V. rossicum invasions.

3.5.2 The Role of ‘Enemy of my Enemy’ in Vincetoxicum rossicum Invasion

The species-specific response of native legume biomass to rhizosphere biota in the present study was consistent with the results of the previous study (Chapter 2) in which a phylogenetically diverse set of native plants also showed a species based response (Chapter 2 Figure 2.5). Only one species in the present study (L. hirta) showed significant inhibition by rhizosphere biota in invaded compared to uninvaded treatments. In combination, the results of the present and previous study suggest that there is great variability in the role rhizosphere biota plays in V. rossicum invasions, even within groups of phylogenetically related native plants (e.g. Fabaceae).

It is not uncommon to find reports of species-specific responses to invasive plants (Day et al., 2016; Del Fabbro & Prati, 2015; Hanslin & Kollmann, 2016; Jordan et al., 2008; Leicht-Young, Bois, & Silander, 2015; O’connor et al., 2015; Perkins & Nowak, 2012). Indeed, there is growing evidence supporting that plant-community dynamics are driven by the interaction between plant and soil-biota genotypes, in the context of environmental conditions (Catford et al., 2009; Johnson & Graham, 2012; Oldroyd, 2013; Pan, Baumgarten, & May, 2008; Pieterse et al., 2016; Veneault-Fourrey et al., 2013). The results of this study provide further evidence showing that the accumulation of rhizosphere enemies does not ‘spill’ onto native plants (‘Enemy of my Enemy’ hypothesis; EEH) and may not be a major factor in the invasive success of V. rossicum.

3.5.3 Conclusions

In conclusion, no consistent support was found for the promotion or inhibition (EEH) of native legumes by V. rossicum rhizosphere biota, but rather the feedback effects of V. rossicum on native legumes was species-specific. Of the four legumes tested, only D. canadense significantly benefited from growing with rhizosphere biota from V. rossicum invaded soil with increases in biomass and nodulation in invaded verses uninvaded treatments. There are several explanations as to why V. rossicum would promote D. canadense biomass but not itself including symbiotic interactions with root-nodule bacteria (specifically rhizobia), root nodule endophytes, elevated colonization by AMF, and synergistic interactions between rhizobia and AMF or nodule endophytes (van der Heijden et al., 2015; Velázquez et al., 2017). Molecular analysis was not done in this study, so the identity of the root-nodule bacteria remains unknown. However, elevated nodule formation in the present study and elevated AMF and nodule colonization in the previous study (Chapter 2) suggest AMF and rhizobia and/or other fungal and bacterial nodule endophytes may be involved in positive effects V. rossicum rhizosphere biota had on D. canadense. Future research should focus on the unexplored area of root-nodule endophytes, as well as, multipartite synergies between root nodule bacteria and root dwelling microbiota and their role in legume productivity.

In addition, the positive effects of V. rossicum rhizosphere communities on D. canadense productivity may make D. canadense a useful tool for the mitigation of V. rossicum invasion. Abundant growth of D. canadense in V. rossicum soils could make it an effective competitor against V. rossicum seedlings germinating from seedbank or arriving through wind driven dispersal. The use of legumes to improve establishment of native plant species after removal of invasive plants has recently been suggested as an effective restoration strategy (Mummey &

Ramsey, 2017). Further research is needed to determine whether Desmodium canadense can be used for restoration purposes following V. rossicum invasions. Such research efforts should focus on how well D. canadense promotes the growth and establishment of native plant species, and what effects it might have on V. rossicum.

CONCLUSION

In this thesis I started by testing (Chapter 2) the enemy of my enemy hypothesis using V. rossicum as a model invasive species. Previous research had suggested that V. rossicum soil biota may inhibit the growth of native species. Six native plants were tested to see how widespread this phenomenon might be. However, there were no consistent reductions in biomass across species, nor were there evidence for elevated levels of pathogens when plants were exposed to soil biota from sites invaded by V. rossicum. In addition, there was no evidence for a phylogenetic signal in the native species’ responses. Thus, it is unlikely that the amplification of pathogens on V. rossicum rhizosphere are an important factor in V. rossicum invasion. In fact, soil biota from V. rossicum rhizosphere can promote the growth of some native species, specifically D. canadense, which showed higher numbers of both nodules and AMF colonization in invaded compared to uninvaded treatments. This lead to the question, was the positive response specific to D. canadense, or could it be a common response by native legumes in general?

To address this question, I set up a second study (Chapter 3) to investigate whether V. rossicum consistently promotes the growth of native legumes (i.e., D. canadense and three other native legumes). The results of this second study do not support the hypothesis that V. rossicum promotes native legumes in general, but rather D. canadense. This was consistent with the species-specific responses observed in the previous study and further supports the conclusion that antagonistic soil biota does not play an important role in V. rossicum invasions (even phylogenetically related groups). That D. canadense responded to V. rossicum soil biota in a consistently positive manner across both the first and second study (Chapter 2 and 3), even when V. rossicum did not, indicates that whichever soil biota are involved form a consistently mutualistic relationship with D. canadense. Desmodium canadense had significantly higher numbers of root-nodules in invaded compared to uninvaded treatments. To gain further understanding of how V. rossicum soil biota can affect native plant species, future research should investigate the identity of these soil biota, as well as, root-nodule endophytes, and their

functional interactions with their host plant. Also, the consistent positive response of D. canadense to V. rossicum soil biota means that D. canadense may be useful as a bridge- species – a species to plant in reclamation projects following V. rossicum removal. Desmodium canadense clearly benefits from V. rossicum soil biota. For D. canadense to be used in remediation programs, further research would be needed to determine if D. rossicum can promote the growth of successive native plants, as well as its effect on V. rossicum.

REFERENCE LIST

Aguilar-Trigueros, C. A., Powell, J. R., Anderson, I. C., Antonovics, J., & Rillig, M. C. (2014). Ecological understanding of root-infecting fungi using trait-based approaches. Trends in Plant Science, 19(7), 432–8. https://doi.org/10.1016/j.tplants.2014.02.006

Aguilar-Trigueros, C. A., & Rillig, M. C. (2016). Effect of different root endophytic fungi on plant community structure in experimental microcosms. Ecology and Evolution, 6(22), 8149– 8158. https://doi.org/10.1002/ece3.2416

Anacker, B. L., Klironomos, J. N., Maherali, H., Reinhart, K. O., & Strauss, S. Y. (2014). Phylogenetic conservatism in plant-soil feedback and its implications for plant abundance. Ecology Letters, 17(12), 1613–1621. https://doi.org/10.1111/ele.12378

Andrews, M., & Andrews, M. E. (2017). Specificity in legume-rhizobia symbioses. International Journal of Molecular Sciences, 18(4), 705. https://doi.org/10.3390/ijms18040705

Antunes, P. M., & Koyama, A. (2017). Mycorrhizas as Nutrient and Energy Pumps of Soil Food Webs: Multitrophic Interactions and Feedbacks. In J. Johnshon, N. C.; Gehring, C.; Jansa (Ed.), Mycorrhizal Mediation of Soil (pp. 149–173). Elsevier. https://doi.org/10.1016/B978- 0-12-804312-7.00009-7

Bachelot, B., Uriarte, M., & McGuire, K. (2015). Interactions among mutualism, competition, and predation foster species coexistence in diverse communities. Theoretical Ecology, 8(3), 297–312. https://doi.org/10.1007/s12080-015-0251-2

Bais, H. P., Weir, T. L., Perry, L. G., Gilroy, S., & Vivanco, J. M. (2006). The role of root exudates in rhizosphere interactions with plants and other organisms. Annual Review of Plant Biology, 57, 233–266. https://doi.org/10.1146/annurev.arplant.57.032905.105159

Bardgett, R. D., Bowman, W. D., Kaufmann, R., & Schmidt, S. K. (2005). A temporal approach to linking aboveground and belowground ecology. Trends in Ecology & Evolution, 20(11), 634–41. https://doi.org/10.1016/j.tree.2005.08.005

Bardgett, R. D., & van der Putten, W. H. (2014). Belowground biodiversity and ecosystem functioning. Nature, 515(7528), 505–511. https://doi.org/10.1038/nature13855

Bardgett, R. D., & Wardle, D. A. (2010). Aboveground-Belowground Linkages: Biotic Interactions, Ecosystem Processes and Global Change. Oxford University Press.

Bauer, J. T., Kleczewski, N. M., Bever, J. D., Clay, K., & Reynolds, H. L. (2012). Nitrogen-fixing bacteria, arbuscular mycorrhizal fungi, and the productivity and structure of prairie grassland communities. Oecologia, 170(4), 1089–1098. https://doi.org/10.1007/s00442- 012-2363-3

Beattie, G. A. (2006). Plant-associated bacteria: survey, molecular phylogeny, genomics and recent advances. In Plant-Associated Bacteria (pp. 1–56). Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-1-4020-4538-7_1

Beckstead, J., Meyer, S. E., Connolly, B. M., Huck, M. B., & Street, L. E. (2010). Cheatgrass facilitates spillover of a seed bank pathogen onto native grass species. Journal of Ecology,

98(1), 168–177. https://doi.org/10.1111/j.1365-2745.2009.01599.x

Bever, J. D., Westover, K. M., & Antonovics, J. (1997). Incorporating the Soil Community into Plant Population Dynamics: The Utility of the Feedback Approach. The Journal of Ecology, 85(5), 561. https://doi.org/10.2307/2960528

Blackburn, T. M., Pyšek, P., Bacher, S., Carlton, J. T., Duncan, R. P., Jarošík, V., … Richardson, D. M. (2011). A proposed unified framework for biological invasions. Trends in Ecology & Evolution, 26(7), 333–9. https://doi.org/10.1016/j.tree.2011.03.023

Blomberg, S. P., Garland, T., & Ives, A. R. (2003). Testing for Phylogenetic Signal in Comparative Data: Behavioral Traits are More Labile. Evolution, 57(4), 717. https://doi.org/10.1554/0014-3820(2003)057[0717:TFPSIC]2.0.CO;2

Bloom, S. A. (1981). Similarity indices in community studies: potential pitfalls. Marine Ecology Progress Series, 5(2), 125–128. https://doi.org/10.3354/meps005125.

Blossey, B., & Notzold, R. (1995). Evolution of Increased Competitive Ability in Invasive Nonindigenous Plants: A Hypothesis. The Journal of Ecology, 83(5), 887. https://doi.org/10.2307/2261425

Bongard, C. L., Navaranjan, G., Yan, W., & Fulthorpe, R. R. (2013). Fungal colonization of the invasive vine Vincetoxicum rossicum and native plants. Plant Ecology and Evolution, 146(1), 45–52. https://doi.org/10.5091/plecevo.2013.739

Bradley, B. A., Blumenthal, D. M., Early, R., Grosholz, E. D., Lawler, J. J., Miller, L. P., … Olden, J. D. (2012). Global change, global trade, and the next wave of plant invasions. Frontiers in Ecology and the Environment, 10(1), 20–28. https://doi.org/10.1890/110145

Bradley, J. D. (2013). Southern Ontario Vascular Plant Species List. Retrieved from http://www.forestry.utoronto.ca/Settledlandscapes/VSP/documents/SouthOntPlantList_Scie ntific.pdf

Brandt, A. J., Seabloom, E. W., & Hosseini, P. R. (2009). Phylogeny and provenance affect plant-soil feedbacks in invaded California grasslands. Ecology, 90(4), 1063–1072. https://doi.org/10.1890/08-0054.1

Bray, J. R., & Curtis, J. T. (1957). An Ordination of the Upland Forest Communities of Southern Wisconsin. Ecological Monographs, 27(4), 325–349. https://doi.org/10.2307/1942268

Buer, C. S., Imin, N., & Djordjevic, M. A. (2010). Flavonoids: New Roles for Old Molecules. Journal of Integrative Plant Biology, 52(1), 98–111. https://doi.org/10.1111/j.1744- 7909.2010.00905.x

Cahill, J. F., Cale, J. A., Karst, J., Bao, T., Pec, G. J., & Erbilgin, N. (2016). No silver bullet: Different soil handling techniques are useful for different research questions, exhibit differential type I and II error rates, and are sensitive to sampling intensity. New Phytologist, 216(1), 11–14. https://doi.org/10.1111/nph.14141

Callaway, R. M., Montesinos, D., Williams, K., & Maron, J. L. (2013). Native congeners provide biotic resistance to invasive Potentilla through soil biota. Ecology, 94(6), 1223–1229. https://doi.org/10.1890/12-1875.1

Callaway, R. M., Thelen, G. C., Barth, S., Ramsey, P. W., & Gannon, J. E. (2004). Soil fungi

alter interactions between the invader Centaurea maculosa and North American natives. Ecology, 85(4), 1062–1071. https://doi.org/10.1890/02-0775

Canadian Food Inspection Agency. (2008). Invasive Alien Plants in Canada: summary report [online]. Ottawa, ON: CFIA. Retrieved from http://www.agrireseau.qc.ca/argeneral/documents/SIPC Report - Summary Report - English Printed Version.pdf

Cappuccino, N. (2004). Allee effect in an invasive alien plant, pale swallow-wort Vincetoxicum rossicum (Asclepiadaceae). Oikos, 106(1), 3–8. https://doi.org/10.1111/j.0030- 1299.2004.12863.x

Cappuccino, N., Mackay, R., & Eisner, C. (2002). Spread of the Invasive Alien Vine Vincetoxicum rossicum: Tradeoffs Between Seed Dispersability and Seed Quality. The American Midland Naturalist, 148(2), 263. https://doi.org/10.1674/0003- 0031(2002)148[0263:SOTIAV]2.0.CO;2

Carroll, G. (1988). Fungal Endophytes in Stems and Leaves: From Latent Pathogen to Mutualistic Symbiont. Ecology, 69(1), 2–9. https://doi.org/10.2307/1943154

Catford, J. A., Jansson, R., & Nilsson, C. (2009). Reducing redundancy in invasion ecology by integrating hypotheses into a single theoretical framework. Diversity and Distributions, 15(1), 22–40. https://doi.org/10.1111/j.1472-4642.2008.00521.x

CBD. (2000). Alien species that threaten ecosystems, habitats or species. Nairobi, Kenya: Secretariat of the Convention on Biological Diversity.

Colautti, R. I., Ricciardi, A., Grigorovich, I. A., & MacIsaac, H. J. (2004). Is invasion success explained by the enemy release hypothesis? Ecology Letters, 7(8), 721–733. https://doi.org/10.1111/j.1461-0248.2004.00616.x

Corral-Lugo, A., Daddaoua, A., Ortega, A., Espinosa-Urgel, M., & Krell, T. (2016). Rosmarinic acid is a homoserine lactone mimic produced by plants that activates a bacterial quorum- sensing regulator. Science Signaling, 9(409), 1–10. https://doi.org/10.1126/scisignal.aaa8271

Coskun, D., Britto, D. T., Shi, W., & Kronzucker, H. J. (2017, August). How Plant Root Exudates Shape the Nitrogen Cycle. Trends in Plant Science. https://doi.org/10.1016/j.tplants.2017.05.004

Coyne, D. L., Nicol, J. M., & Claudius-Cole, B. (2014). Practical plant nematology: a field and laboratory guide (2nd ed.). Cotonou, Benin: SP-IPM Secretariat, International Institute of Tropical Agriculutre. Retrieved from https://cgspace.cgiar.org/bitstream/handle/10568/81172/U14BkCoynePracticalNothomDev. pdf-d663ec356760331c1acd9a16e3848f16.pdf?sequence=1

Crins, W. J., Gray, P. A., Uhlig, P. W. C., & Wester, M. C. (2009). The ecosystems of Ontario, Part 1: Ecozones and ecoregions. Technical Report SIB TER IMA TR-01. Peterborough, Ontario.

Crocker, E. V., Karp, M. A., & Nelson, E. B. (2015). Virulence of oomycete pathogens from Phragmites australis-invaded and noninvaded soils to seedlings of wetland plant species. Ecology and Evolution, 5(11), 2127–2139. https://doi.org/10.1002/ece3.1468

Crocker, E. V., Lanzafane, J. J., Karp, M. A., & Nelson, E. B. (2016). Overwintering seeds as reservoirs for seedling pathogens of wetland plant species. Ecosphere, 7(3), e01281. https://doi.org/10.1002/ecs2.1281

Cross, H. B., Lowe, A. J., & Gurge, C. F. D. (2011). DNA barcoding of invasive species. In D. M. Richardson (Ed.), Fifty years of invasion ecology. The legacy of Charles Elton (First, pp. 289–300). Oxford: Wiley-Blackwell.

Darwin, C. (1859). On the origin of species by means of natural selection. Murray, London.

Davison, J., Öpik, M., Zobel, M., Vasar, M., Metsis, M., & Moora, M. (2012). Communities of Arbuscular Mycorrhizal Fungi Detected in Forest Soil Are Spatially Heterogeneous but Do Not Vary throughout the Growing Season. PLoS ONE, 7(8), e41938. https://doi.org/10.1371/journal.pone.0041938

Day, N. J., Antunes, P. M., & Dunfield, K. E. (2015). Changes in arbuscular mycorrhizal fungal communities during invasion by an exotic invasive plant. Acta Oecologica, 67, 66–74. https://doi.org/10.1016/j.actao.2015.06.004

Day, N. J., Dunfield, K. E., & Antunes, P. M. (2015). Temporal dynamics of plant-soil feedback and root-associated fungal communities over 100years of invasion by a non-native plant. Journal of Ecology, 103(6), 1557–1569. https://doi.org/10.1111/1365-2745.12459

Day, N. J., Dunfield, K. E., & Antunes, P. M. (2016). Fungi from a non-native invasive plant increase its growth but have different growth effects on native plants. Biological Invasions, 18(1), 231–243. https://doi.org/10.1007/s10530-015-1004-2

Dean, R., Van Kan, J. A. L., Pretorius, Z. A., Hammond-Kosack, K. E., Di Pietro, A., Spanu, P. D., … Foster, G. D. (2012). The Top 10 fungal pathogens in molecular plant pathology. Molecular Plant Pathology, 13(4), 414–30. https://doi.org/10.1111/j.1364- 3703.2011.00783.x

Del Fabbro, C., & Prati, D. (2015). Invasive plant species do not create more negative soil conditions for other plants than natives. Perspectives in Plant Ecology, Evolution and Systematics, 17(2), 87–95. https://doi.org/10.1016/j.ppees.2015.02.002

Dengler, J., Chytrý, M., & Ewald, J. (2008). Phytosociology. Encyclopedia of Ecology, 4, 2767– 2779. Retrieved from http://www.elsevier.com/locate/permissionusematerial

Deyn, G. B. De. (2017). Plant life history and above-belowground interactions: missing links. Oikos, 126(4), 497–507. https://doi.org/10.1111/oik.03967

Diez, J. M., Sullivan, J. J., Hulme, P. E., Edwards, G., & Duncan, R. P. (2008). Darwin’s naturalization conundrum: Dissecting taxonomic patterns of species invasions. Ecology Letters, 11(7), 674–681. https://doi.org/10.1111/j.1461-0248.2008.01178.x

DiTommaso, A., Lawlor, F. M., & Darbyshire, S. J. (2005). The Biology of Invasive Alien Plants in Canada. 2. Cynanchum rossicum (Kleopow) Borhidi [= Vincetoxicum rossicum (Kleopow) Barbar.] and Cynanchum louiseae (L.) Kartesz & Gandhi [= Vincetoxicum nigrum (L.) Moench]. Canadian Journal of Plant Science, 85(1), 243–263. https://doi.org/10.4141/P03-056

DiTommaso, A., & Losey, J. E. (2003). Oviposition preference and larval performance of

monarch butterflies (Danaus plexippus) on two invasive swallow-wort species. Entomologia Experimentalis et Applicata, 108(3), 205–209. https://doi.org/10.1046/j.1570- 7458.2003.00089.x

Dostál, P., & Palečková, M. (2011). Does relatedness of natives used for soil conditioning influence plant-soil feedback of exotics? Biological Invasions, 13(2), 331–340. https://doi.org/10.1007/s10530-010-9824-6

Dudeja, S. S., Giri, R., Saini, R., Suneja-Madan, P., & Kothe, E. (2012). Interaction of endophytic microbes with legumes. Journal of Basic Microbiology, 52(3), 248–60. https://doi.org/10.1002/jobm.201100063

Early, R., Bradley, B. A., Dukes, J. S., Lawler, J. J., Olden, J. D., Blumenthal, D. M., … Tatem, A. J. (2016). Global threats from invasive alien species in the twenty-first century and national response capacities. Nature Communications, 7. https://doi.org/10.1038/ncomms12485

Ehinger, M., Mohr, T. J., Starcevich, J. B., Sachs, J. L., Porter, S. S., & Simms, E. L. (2014). Specialization-generalization trade-off in a Bradyrhizobium symbiosis with wild legume hosts. BMC Ecology, 14(1), 8. https://doi.org/10.1186/1472-6785-14-8

Elton, C. S. (1958). The Ecology of Invasions by Animals and Plants. Methuen and Co Ltd., London. The ecology of invasions by animals and plants. London: Methuen & Co. Ltd. Retrieved from https://www.cabdirect.org/cabdirect/abstract/19592202723

Eppinga, M. B., Rietkerk, M., Dekker, S. C., De Ruiter, P. C., & Van Der Putten, W. H. (2006). Accumulation of local pathogens: A new hypothesis to explain exotic plant invasions. Oikos, 114(1), 168–176. https://doi.org/10.1111/j.2006.0030-1299.14625.x

Ernst, C. M., & Cappuccino, N. (2005). The effect of an invasive alien vine, Vincetoxicum rossicum (Asclepiadaceae), on arthropod populations in Ontario old fields. Biological Invasions, 7(3), 417–425. https://doi.org/10.1007/s10530-004-4062-4

Eviner, V. T., & Hawkes, C. V. (2008). Embracing variability in the application of plant-soil interactions to the restoration of communities and ecosystems. Restoration Ecology, 16(4), 713–729. https://doi.org/10.1111/j.1526-100X.2008.00482.x

Faillace, C. A., Lorusso, N. S., & Duffy, S. (2017). Overlooking the smallest matter: viruses impact biological invasions. Ecology Letters, 20(4), 524–538. https://doi.org/10.1111/ele.12742

Feng, Y., & van Kleunen, M. (2016). Phylogenetic and functional mechanisms of direct and indirect interactions among alien and native plants. Journal of Ecology, 104(4), 1136–1148. https://doi.org/10.1111/1365-2745.12577

Fitzpatrick, C. R., Gehant, L., Kotanen, P. M., & Johnson, M. T. J. (2017). Phylogenetic relatedness, phenotypic similarity and plant-soil feedbacks. Journal of Ecology, 105(3), 786–800. https://doi.org/10.1111/1365-2745.12709

Flory, S. L., & Clay, K. (2013). Pathogen accumulation and long-term dynamics of plant invasions. Journal of Ecology, 101(3), 607–613. https://doi.org/10.1111/1365-2745.12078

Freckleton, R. P., Harvey, P. H., & Pagel, M. (2002). Phylogenetic analysis and comparative

data: a test and review of evidence. The American Naturalist, 160(6), 712–26. https://doi.org/10.1086/343873

Fridley, J. D. (2011). Biodiversity as a Bulwark Against Invasion: Conceptual Threads Since Elton. In D. M. Richardson (Ed.), Fifty Years of Invasion Ecology: The Legacy of Charles Elton (first, pp. 121–130). Oxford, UK: Wiley-Blackwell. https://doi.org/10.1002/9781444329988.ch10

Fridley, J. D., & Sax, D. F. (2014). The imbalance of nature: revisiting a Darwinian framework for invasion biology. Global Ecology and Biogeography, 23(11), 1–10. https://doi.org/10.1111/geb.12221

Gao, M., Teplitski, M., Robinson, J. B., & Bauer, W. D. (2003). Production of substances by Medicago truncatula that affect bacterial quorum sensing. Molecular Plant-Microbe Interactions, 16(9), 827–834. https://doi.org/10.1094/MPMI.2003.16.9.827

García-Cano, E., Resende, R. O., Fernández-Muñoz, R., & Moriones, E. (2006). Synergistic Interaction Between Tomato chlorosis virus and Tomato spotted wilt virus Results in Breakdown of Resistance in Tomato. Phytopathology, 96(11), 1263–1269. https://doi.org/10.1094/PHYTO-96-1263

Gilbert, B., & Levine, J. M. (2013). Plant invasions and extinction debts. Proceedings of the National Academy of Sciences of the United States of America, 110(5), 1744–9. https://doi.org/10.1073/pnas.1212375110

Greenwood, R. M., & Pankhurst, C. E. (1977). The rhizobium component of the nitrogen-fixing symbiosis. Proceedings NZ Grassland Association, 38, 167–174.

Großkinsky, D. K., Tafner, R., Moreno, M. V., Stenglein, S. A., García de Salamone, I. E., Nelson, L. M., … Roitsch, T. (2016). Cytokinin production by Pseudomonas fluorescens G20-18 determines biocontrol activity against Pseudomonas syringae in Arabidopsis. Scientific Reports, 6, 23310. https://doi.org/10.1038/srep23310

Gu, J., Wang, E. T., & Chen, W. X. (2007). Genetic diversity of rhizobia associated with Desmodium species grown in China. Letters in Applied Microbiology, 44(3), 286–292. https://doi.org/10.1111/j.1472-765X.2006.02071.x

Gunn, B. F., Baudouin, L., & Olsen, K. M. (2011). Independent origins of cultivated coconut (Cocos nucifera L.) in the old world tropics. PLoS ONE, 6(6), e21143. https://doi.org/10.1371/journal.pone.0021143

Gurevitch, J., & Padilla, D. K. (2004). Are invasive species a major cause of extinctions? Trends in Ecology & Evolution, 19(9), 470–4. https://doi.org/10.1016/j.tree.2004.07.005

Hanslin, H. M., & Kollmann, J. (2016). Positive responses of coastal dune plants to soil conditioning by the invasive Lupinus nootkatensis. Acta Oecologica, 77, 1–9. https://doi.org/10.1016/j.actao.2016.08.007

Hardoim, P. R., van Overbeek, L. S., Berg, G., Pirttilä, A. M., Compant, S., Campisano, A., … Sessitsch, A. (2015). The Hidden World within Plants: Ecological and Evolutionary Considerations for Defining Functioning of Microbial Endophytes. Microbiology and Molecular Biology Reviews, 79(3), 293–320. https://doi.org/10.1128/MMBR.00050-14

Hernández-Montiel, L. G., Rueda-Puente, E. O., Cordoba-Matson, M. V., Holguín-Peña, J. R., & Zulueta-Rodríguez, R. (2013). Mutualistic interaction of rhizobacteria with arbuscular mycorrhizal fungi and its antagonistic effect on Fusarium oxysporum in Carica papaya seedlings. Crop Protection, 47, 61–66. https://doi.org/10.1016/j.cropro.2013.01.008

Hooper, D. U., Chapin III, F. S., Ewel, J. J., Hector, A., Inchausti, P., Lavorel, S., … Wardle, D. A. (2005). Effects of biodiversity on ecosystem functioning: A consensus of current knowledge. Ecological Monographs, 75(1), 3–35. Retrieved from http://ecology- climate.org/files/Hooper_et_al__ESA2005.pdf

Inderjit, & van der Putten, W. H. (2010). Impacts of soil microbial communities on exotic plant invasions. Trends in Ecology & Evolution, 25(9), 512–9. https://doi.org/10.1016/j.tree.2010.06.006

Isbell, F., Gonzalez, A., Loreau, M., Cowles, J. A., Díaz, S., Hector, A., … Larigauderie, A. (2017). Linking the influence and dependence of people on biodiversity across scales. Nature, 546, 65–72. https://doi.org/10.1038/nature22899

IUCN. (2000). Guidelines for the prevention of biodiversity loss caused by alien invasive species. Gland. Retrieved from https://www.iucn.org/knowledge/focus/ipbes_focus/invasive_species_database/

Jackson, L. E., Burger, M., & Cavagnaro, T. R. (2008). Roots, Nitrogen Transformations, and Ecosystem Services. Annual Review of Plant Biology, 59, 341–363. https://doi.org/10.1146/annurev.arplant.59.032607.092932

Jeschke, J. M., Gómez Aparicio, L., Haider, S., Heger, T., Lortie, C., Pyšek, P., & Strayer, D. (2012). Support for major hypotheses in invasion biology is uneven and declining. NeoBiota, 14, 1–20. https://doi.org/10.3897/neobiota.14.3435

Johnson, N. C., & Graham, J. H. (2012). The continuum concept remains a useful framework for studying mycorrhizal functioning. Plant and Soil, 363(1–2), 411–419. https://doi.org/10.1007/s11104-012-1406-1

Johnson, N. C., Graham, J. H., & Smith, F. A. (1997). Functioning of mycorrhizal associations along the mutualism-parasitism continuum. New Phytologist, 135(4), 575–585. https://doi.org/10.1046/j.1469-8137.1997.00729.x

Jones, J. T., Haegeman, A., Danchin, E. G. J., Gaur, H. S., Helder, J., Jones, M. G. K., … Perry, R. N. (2013). Top 10 plant-parasitic nematodes in molecular plant pathology. Molecular Plant Pathology, 14(9), 946–961. https://doi.org/10.1111/mpp.12057

Jones, M. D., & Smith, S. E. (2004). Exploring functional definitions of mycorrhizas: Are mycorrhizas always mutualisms? Canadian Journal of Botany, 82(8), 1089–1109. https://doi.org/10.1139/b04-110

Jordan, N. R., Larson, D. L., & Huerd, S. C. (2008). Soil modification by invasive plants: Effects on native and invasive species of mixed-grass prairies. Biological Invasions, 10(2), 177– 190. https://doi.org/10.1007/s10530-007-9121-1

Jordan, N. R., Larson, D. L., & Huerd, S. C. (2011). Evidence of Qualitative Differences between Soil-Occupancy Effects of Invasive vs. Native Grassland Plant Species. Invasive Plant Science and Management, 4(1), 11–21. https://doi.org/10.1614/IPSM-D-10-00004.1

Kalra, Y. P., & Maynard, D. G. (1991). Particle-size Analysis. Methods maula for forest soil and plant analysis. Edmonton.

Kamoun, S., Furzer, O., Jones, J. D. G., Judelson, H. S., Ali, G. S., Dalio, R. J. D., … Govers, F. (2015). The Top 10 oomycete pathogens in molecular plant pathology. Molecular Plant Pathology, 16(4), 413–434. https://doi.org/10.1111/mpp.12190

Karst, J., Erbilgin, N., Pec, G. J., Cigan, P. W., Najar, A., Simard, S. W., & Cahill, J. F. (2015). Ectomycorrhizal fungi mediate indirect effects of a bark beetle outbreak on secondary chemistry and establishment of pine seedlings. New Phytologist, 208(3), 904–914. https://doi.org/10.1111/nph.13492

Keane, R. M., & Crawley, M. J. (2002). Exotic plant invasions and the enemy release hypothesis. Trends in Ecology and Evolution, 17(4), 164–170. https://doi.org/10.1016/S0169-5347(02)02499-0

Kelly, D. W., Paterson, R. A., Townsend, C. R., Poulin, R., & Tompkins, D. M. (2009). Parasite spillback: A neglected concept in invasion ecology? Ecology, 90(8), 2047–2056. https://doi.org/10.1890/08-1085.1

Klironomos, J. N. (2002). Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature, 417(6884), 67–70. https://doi.org/10.1038/417067a

Klironomos, J. N. (2003). Variation in plant response to native and exotic arbuscular mycorrhizal fungi. Ecology, 84(9), 2292–2301. https://doi.org/10.1890/02-0413

Kobayashi, D. Y., & Crouch, J. A. (2009). Bacterial/Fungal interactions: from pathogens to mutualistic endosymbionts. Annual Review of Phytopathology, 47, 63–82. https://doi.org/10.1146/annurev-phyto-080508-081729

Lamichhane, J. R., & Venturi, V. (2015). Synergisms between microbial pathogens in plant disease complexes: a growing trend. Frontiers in Plant Science, 6, 385. https://doi.org/10.3389/fpls.2015.00385

Lankau, R. A., Nuzzo, V., Spyreas, G., & Davis, A. S. (2009). Evolutionary limits ameliorate the negative impact of an invasive plant. Proceedings of the National Academy of Sciences of the United States of America, 106(36), 15362–7. https://doi.org/10.1073/pnas.0905446106

Larimer, A. L., Bever, J. D., & Clay, K. (2010). The interactive effects of plant microbial symbionts: A review and meta-analysis. Symbiosis, 51(2), 139–148. https://doi.org/10.1007/s13199-010-0083-1

Leicht-Young, S. A., Bois, S. T., & Silander, J. A. (2015). Impacts of Celastrus-primed soil on common native and invasive woodland species. Plant Ecology, 216(4), 503–516. https://doi.org/10.1007/s11258-015-0451-2

Li, B., Li, Y.-Y., Wu, H.-M., Zhang, F.-F., Li, C.-J., Li, X.-X., … Li, L. (2016). Root exudates drive interspecific facilitation by enhancing nodulation and N 2 fixation. Proceedings of the National Academy of Sciences, 113(23), 6496–6501. https://doi.org/10.1073/pnas.1523580113

Li, H., Zhang, X., Zheng, R., Li, X., Elmer, W. H., Wolfe, L. M., & Li, B. (2014). Indirect effects of non-native Spartina alterniflora and its fungal pathogen (Fusarium palustre) on native

saltmarsh plants in China. Journal of Ecology, 102, 1112–1119. https://doi.org/10.1111/1365-2745.12285

Limestone Creek Restoration Nursary. (1996). Restoring the Native Landscape. Toronto: Limestone Creek Restoration Nursary.

Lipson, D. A., & Schmidt, S. K. (2004). Seasonal changes in an alpine soil bacterial community in the colorado rocky mountains. Applied and Environmental Microbiology, 70(5), 2867–79. https://doi.org/10.1128/AEM.70.5.2867-2879.2004

Malmstrom, C. M., Hughes, C. C., Newton, L. A., & Stoner, C. J. (2005). Virus infection in remnant native bunchgrasses from invaded California grasslands. New Phytologist, 168(1), 217–230. https://doi.org/10.1111/j.1469-8137.2005.01479.x

Malmstrom, C. M., McCullough, A. J., Johnson, H. A., Newton, L. A., & Borer, E. T. (2005). Invasive annual grasses indirectly increase virus incidence in California native perennial bunchgrasses. Oecologia, 145(1), 153–164. https://doi.org/10.1007/s00442-005-0099-z

Malmstrom, C. M., Shu, R., Linton, E. W., Newton, L. A., & Cook, M. A. (2007). Barley yellow dwarf viruses (BYDVs) preserved in herbarium specimens illuminate historical disease ecology of invasive and native grasses. Journal of Ecology, 95(6), 1153–1166. https://doi.org/10.1111/j.1365-2745.2007.01307.x

Mangla, S., Inderjit, & Callaway, R. M. (2008). Exotic invasive plant accumulates native soil pathogens which inhibit native plants. Journal of Ecology, 96, 58–67. https://doi.org/10.1111/j.1365-2745.2007.01312.x

Mansfield, J., Genin, S., Magori, S., Citovsky, V., Sriariyanum, M., Ronald, P., … Foster, G. D. (2012). Top 10 plant pathogenic bacteria in molecular plant pathology. Molecular Plant Pathology, 13(6), 614–629. https://doi.org/10.1111/j.1364-3703.2012.00804.x

Marler, M. J., Zabinski, C. A., & Callaway, R. M. (1999). Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology, 80(4), 1180–1186. https://doi.org/10.1890/0012-9658(1999)080[1180:MIECEO]2.0.CO;2

Martínez-García, L. B., Pietrangelo, O., & Antunes, P. M. (2016). Parent tree distance- dependent recruitment limitation of native and exotic invasive seedlings in urban forests. Urban Ecosystems, 19(2), 969–981. https://doi.org/10.1007/s11252-015-0507-z

Martínez-Hidalgo, P., & Hirsch, A. M. (2017). The Nodule Microbiome: N 2 -Fixing Rhizobia Do Not Live Alone. Phytobiomes, 1(2), 70–82. https://doi.org/10.1094/PBIOMES-12-16-0019- RVW

Mattila, H. R., & Otis, G. W. (2003). A comparison of the host preference of monarch butterflies (Danaus plexippus) for milkweed (Asclepias syriaca) over dog-strangler vine (Vincetoxicum rossicum). Entomologia Experimentalis et Applicata, 107(3), 193–199. https://doi.org/10.1046/j.1570-7458.2003.00049.x

McGonigle, T. P., Miller, M. H., Evans, D. G., Fairchild, G. L., & Swan, J. A. (1990). A new method which gives an objective measure of colonization of roots by vesicular-arbuscular mycorrhizal fungi. New Phytologist, 115(3), 495–501. https://doi.org/10.1111/j.1469- 8137.1990.tb00476.x

McLeod, M. L., Cleveland, C. C., Lekberg, Y., Maron, J. L., Philippot, L., Bru, D., … Aerts, R. (2016). Exotic invasive plants increase productivity, abundance of ammonia-oxidizing bacteria and nitrogen availability in intermountain grasslands. Journal of Ecology, 104(4), 994–1002. https://doi.org/10.1111/1365-2745.12584

Mehrabi, Z., Bell, T., & Lewis, O. T. (2015). Plant-soil feedbacks from 30-year family-specific soil cultures: phylogeny, soil chemistry and plant life stage. Ecology and Evolution, 5(12), 2333–9. https://doi.org/10.1002/ece3.1487

Mehrabi, Z., & Tuck, S. L. (2015). Relatedness is a poor predictor of negative plant-soil feedbacks. New Phytologist, 205(3), 1071–1075. https://doi.org/10.1111/nph.13238

Miller, Z. J., & Menalled, F. D. (2015). Impact of species identity and phylogenetic relatedness on biologically-mediated plant-soil feedbacks in a low and a high intensity agroecosystem. Plant and Soil, 389(1–2), 171–183. https://doi.org/10.1007/s11104-014-2336-x

Mitchell, C. E. (2003). Trophic control of grassland production and biomass by pathogens. Ecology Letters, 6(2), 147–155. https://doi.org/10.1046/j.1461-0248.2003.00408.x

Mogg, C., Petit, P., Cappuccino, N., Durst, T., McKague, C., Foster, M., … Smith, M. L. (2008). Tests of the antibiotic properties of the invasive vine Vincetoxicum rossicum against bacteria, fungi and insects. Biochemical Systematics and Ecology, 36(5–6), 383–391. https://doi.org/10.1016/j.bse.2008.01.001

Mordecai, E. A. (2013). Despite spillover, a shared pathogen promotes native plant persistence in a cheatgrass-invaded grassland. Ecology, 94(12), 2744–2753. https://doi.org/10.1890/13-0086.1

Mummey, D. L., & Ramsey, P. W. (2017). Can sainfoin improve conditions for establishment of native forbs in crested wheatgrass stands? Ecological Restoration, 35(2), 127–137. https://doi.org/10.3368/er.35.2.127

Ng, J., Perrine-Walker, F., Wasson, A., & Mathesius, U. (2015). The Control of Auxin Transport in Parasitic and Symbiotic Root–Microbe Interactions. Plants, 4(3), 606–643. https://doi.org/10.3390/plants4030606

O’connor, J. J., Prevey, J. S., & Seastedt, T. R. (2015). Effects of Short-term soil conditioning by cheatgrass and western wheatgrass. The Prairie Naturalist, 47, 65–72. Retrieved from https://www.researchgate.net/profile/Janet_Prevey/publication/304215870_Effects_of_Shor t- term_Soil_Conditioning_by_Cheatgrass_and_Western_Wheatgrass/links/5769aae308ae2d 7145ba857f.pdf

O ’Hara, R. B., & Kotze, D. J. (2010). Do not log-transform count data. Methods in Ecology and Evolution, 1(2), 118–122. https://doi.org/10.1111/j.2041-210X.2010.00021.x

Oldroyd, G. E. D. (2013). Speak, friend, and enter: signalling systems that promote beneficial symbiotic associations in plants. Nature Reviews Microbiology, 11(4), 252–263. https://doi.org/10.1038/nrmicro2990

Oldroyd, G. E. D., & Downie, J. A. (2004). Calcium, kinases and nodulation signalling in legumes. Nature Reviews Molecular Cell Biology, 5(7), 566–576. https://doi.org/10.1038/nrm1424

Oldroyd, G. E. D., & Downie, J. A. (2008). Coordinating Nodule Morphogenesis with Rhizobial Infection in Legumes. Annual Review of Plant Biology, 59(1), 519–546. https://doi.org/10.1146/annurev.arplant.59.032607.092839

Oldroyd, G. E. D., Murray, J. D., Poole, P. S., & Downie, J. A. (2011). The Rules of Engagement in the Legume-Rhizobial Symbiosis. Annual Review of Genetics, 45(1), 119–144. https://doi.org/10.1146/annurev-genet-110410-132549

Ontario’s Invading Species Awareness Program. (2017). Dog-Strangling Vine. Retrieved August 10, 2017, from http://www.invadingspecies.com/dog-strangling-vine/

Ontario Geological Survey. (1991). Bedrock geology of Ontario, southern sheet; Ontario Geological Survey, Map2544, scale 1:1 000 000.

Pagel, M. (1999). Inferring the historical patterns of biological evolution. Nature, 401(6756), 877–884. https://doi.org/10.1038/44766

Palaniappan, P., Chauhan, P. S., Saravanan, V. S., Anandham, R., & Sa, T. (2010). Isolation and characterization of plant growth promoting endophytic bacterial isolates from root nodule of Lespedeza sp. Biology and Fertility of Soils, 46(8), 807–816. https://doi.org/10.1007/s00374-010-0485-5

Pan, J. J., Baumgarten, A. M., & May, G. (2008). Effects of host plant environment and Ustilago maydis infection on the fungal endophyte community of maize (Zea mays). The New Phytologist, 178(1), 147–56. https://doi.org/10.1111/j.1469-8137.2007.02350.x

Paradis, E., Claude, J., & Strimmer, K. (2004). APE: Analyses of phylogenetics and evolution in R language. Bioinformatics, 20(2), 289–290. https://doi.org/10.1093/bioinformatics/btg412

Parker, M. A., Jankowiak, J. G., & Landrigan, G. K. (2015). Diversifying selection by Desmodiinae legume species on Bradyrhizobium symbionts. FEMS Microbiology Ecology, 91(7), fiv075. https://doi.org/10.1093/femsec/fiv075

Parker, M. A., Malek, W., & Parker, I. M. (2006). Growth of an invasive legume is symbiont limited in newly occupied habitats. Diversity and Distributions, 12(5), 563–571. https://doi.org/10.1111/j.1366-9516.2006.00255.x

Pérez-Montaño, F., Jiménez-Guerrero, I., Contreras Sánchez-Matamoros, R., López-Baena, F. J., Ollero, F. J., Rodríguez-Carvajal, M. A., … Espuny, M. R. (2013). Rice and bean AHL- mimic quorum-sensing signals specifically interfere with the capacity to form biofilms by plant-associated bacteria. Research in Microbiology, 164(7), 749–760. https://doi.org/10.1016/j.resmic.2013.04.001

Perkins, L. B., & Nowak, R. S. (2012). Soil conditioning and plant–soil feedbacks affect competitive relationships between native and invasive grasses. Plant Ecology, 213(8), 1337–1344. https://doi.org/10.1007/s11258-012-0092-7

Philippot, L., Raaijmakers, J. M., Lemanceau, P., & van der Putten, W. H. (2013). Going back to the roots: the microbial ecology of the rhizosphere. Nature Reviews. Microbiology, 11(11), 789–99. https://doi.org/10.1038/nrmicro3109 phyloT : Phylogenetic Tree Generator. (2016). Retrieved April 14, 2017, from http://phylot.biobyte.de/

Pieterse, C. M. J., de Jonge, R., & Berendsen, R. L. (2016). The Soil-Borne Supremacy. Trends in Plant Science, 21(3), 171–173. https://doi.org/10.1016/j.tplants.2016.01.018

Plett, J. M., & Martin, F. (2015). Reconsidering mutualistic plant-fungal interactions through the lens of effector biology. Current Opinion in Plant Biology, 26, 45–50. https://doi.org/10.1016/j.pbi.2015.06.001

Power, A. G., & Mitchell, C. E. (2004). Pathogen spillover in disease epidemics. The American Naturalist, 164 Suppl, S79-89. https://doi.org/10.1086/424610

R CoreTeam. (2011). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.r-project.org/

R CoreTeam. (2015). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.r-project.org/

Rasmann, S., & Turlings, T. C. J. (2016). Root signals that mediate mutualistic interactions in the rhizosphere. Current Opinion in Plant Biology, 32, 62–68. https://doi.org/10.1016/j.pbi.2016.06.017

Reinhart, K. O., & Callaway, R. M. (2006). Soil biota and invasive plants. The New Phytologist, 170(3), 445–57. https://doi.org/10.1111/j.1469-8137.2006.01715.x

Reinhart, K. O., & Rinella, M. J. (2016, May). A common soil handling technique can generate incorrect estimates of soil biota effects on plants. New Phytologist. https://doi.org/10.1111/nph.13822

Reinhart, K. O., Wilson, G. W. T., & Rinella, M. J. (2012). Predicting plant responses to mycorrhizae: Integrating evolutionary history and plant traits. Ecology Letters, 15(7), 689– 695. https://doi.org/10.1111/j.1461-0248.2012.01786.x

Revell, L. J. (2012). Phytools: An R package for phylogenetic comparative biology (and other things). Methods in Ecology and Evolution, 3(2), 217–223. https://doi.org/10.1111/j.2041- 210X.2011.00169.x

Ricciardi, A., Blackburn, T. M., Carlton, J. T., Dick, J. T. A., Hulme, P. E., Iacarella, J. C., … Aldridge, D. C. (2017). Invasion Science: A Horizon Scan of Emerging Challenges and Opportunities. Trends in Ecology & Evolution, 32(6), 464–474. https://doi.org/10.1016/j.tree.2017.03.007

Ricciardi, A., Palmer, M. E., & Yan, N. D. (2011). Should Biological Invasions Be Managed as Natural Disasters? BioScience, 61(4), 312–317. https://doi.org/10.1525/bio.2011.61.4.11

Richardson, D. M., & Pyšek, P. (2006). Plant invasions: merging the concepts of species invasiveness and community invasibility. Progress in Physical Geography, 30(3), 409–431. https://doi.org/10.1191/0309133306pp490pr

Richardson, D. M., Pyšek, P., & Carlton, J. T. (2011). A compendium of essential concepts and terminology in biological invasions. In D. M. Richardson (Ed.), Fifty years of invasion ecology: the legacy of Charles Elton (First, pp. 409–420). Oxford: Blackwell Publishing Ltd.

Richardson, D. M., Pyšek, P., Rejmanek, M., Barbour, M. G., Panetta, F. D., & West, C. J. (2000). Naturalization and invasion of alien plants: concepts and definitions. Diversity Distributions, 6(2), 93–107.

https://doi.org/10.1046/j.1472-4642.2000.00083.x

Ripple, W. J., & Beschta, R. L. (2012). Trophic cascades in Yellowstone: The first 15years after wolf reintroduction. Biological Conservation, 145(1), 205–213. https://doi.org/10.1016/j.biocon.2011.11.005

Rodrigues, R. R., Pineda, R. P., Barney, J. N., Nilsen, E. T., Barrett, J. E., & Williams, M. A. (2015). Plant invasions associated with change in root-zone microbial community structure and diversity. PLoS ONE, 10(10), e0141424. https://doi.org/10.1371/journal.pone.0141424

Sanderson, L. A., Day, N. J., & Antunes, P. M. (2015). Edaphic factors and feedback do not limit range expansion of an exotic invasive plant. Plant Ecology, 216(1), 133–141. https://doi.org/10.1007/s11258-014-0422-z

Schadt, C. W. (2003). Seasonal Dynamics of Previously Unknown Fungal Lineages in Tundra Soils. Science, 301(5638), 1359–1361. https://doi.org/10.1126/science.1086940

Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., … Cardona, A. (2012). Fiji: an open-source platform for biological-image analysis. Nature Methods, 9(7), 676–682. https://doi.org/10.1038/nmeth.2019

Schnitzer, S. A., Klironomos, J. N., HilleRisLambers, J., Kinkel, L. L., Reich, P. B., Xiao, K., … Scheffer, M. (2011). Soil microbes drive the classic plant diversity-productivity pattern. Ecology, 92(2), 296–303. https://doi.org/10.1890/10-0773.1

Scholthof, K.-B. G., Adkins, S., Czosnek, H., Palukaitis, P., Jacquot, E., Hohn, T., … Foster, G. D. (2011). Top 10 plant viruses in molecular plant pathology. Molecular Plant Pathology, 12(9), 938–954. https://doi.org/10.1111/j.1364-3703.2011.00752.x

Shaw, M. W., Emmanuel, C. J., Emilda, D., Terhem, R. B., Shafia, A., Tsamaidi, D., … van Kan, J. A. L. (2016). Analysis of Cryptic, Systemic Botrytis Infections in Symptomless Hosts. Frontiers in Plant Science, 7(May), 625. https://doi.org/10.3389/fpls.2016.00625

Sheeley, S. E., & Raynal, D. J. (1996). The distribution and status of species of Vincetoxicum in eastern North America. Bulletin of the Torrey Botanical Club, 123, 148–156. Retrieved from http://www.jstor.org/stable/2996072?seq=4#page_scan_tab_contents

Shelby, N., Duncan, R. P., van der Putten, W. H., McGinn, K. J., Weser, C., & Hulme, P. E. (2016). Plant mutualisms with rhizosphere microbiota in introduced versus native ranges. Journal of Ecology, 104(5), 1259–1270. https://doi.org/10.1111/1365-2745.12609

Shinya, T., Nakagawa, T., Kaku, H., & Shibuya, N. (2015). Chitin-mediated plant–fungal interactions: catching, hiding and handshaking. Current Opinion in Plant Biology, 26, 64– 71. https://doi.org/10.1016/j.pbi.2015.05.032

Simberloff, D. (2005). Non-native Species DO Threaten the Natural Environment! Journal of Agricultural and Environmental Ethics, 18(6), 595–607. https://doi.org/10.1007/s10806-005- 2851-0

Simberloff, D. (2011). Charles Elton: Neither Founder Nor Siren, but Prophet. In D. M. Richardson (Ed.), Fifty Years of Invasion Ecology: The Legacy of Charles Elton (first, pp. 11–24). Oxford, UK: Wiley-Blackwell. https://doi.org/10.1002/9781444329988.ch2

Simberloff, D., Souza, L., Nũnez, M. A., Barrios-Garcia, M. N., & Bunn, W. (2012). The natives

are restless, but not often and mostly when disturbed. Ecology, 93(3), 598–607. https://doi.org/10.1890/11-1232.1

Smith-Ramesh, L. M., & Reynolds, H. L. (2017, May). The next frontier of plant-soil feedback research: Unraveling context dependence across biotic and abiotic gradients. (R. Michalet, Ed.), Journal of Vegetation Science, pp. 484–494. https://doi.org/10.1111/jvs.12519

Smith, F. A., & Smith, S. E. (2015). How harmonious are arbuscular mycorrhizal symbioses? Inconsistent concepts reflect different mindsets as well as results. New Phytologist, 205(4), 1381–1384. https://doi.org/10.1111/nph.13202

Smith, L. L., DiTommaso, A., Lehmann, J., & Greipsson, S. (2008, April 30). Effects of Arbuscular Mycorrhizal Fungi on The Exotic Invasive Vine Pale Swallow-Wort (Vincetoxicum rossicum). Invasive Plant Science and Management. Weed Science Society of America 810 East 10th Street, Lawrence, KS 66044-8897. https://doi.org/10.1614/IPSM- 07-010.1

Spatafora, J. W., Chang, Y., Benny, G. L., Lazarus, K., Smith, M. E., Berbee, M. L., … Stajich, J. E. (2016). A phylum-level phylogenetic classification of zygomycete fungi based on genome-scale data. Mycologia, 108(5), 1028–1046. https://doi.org/10.3852/16-042

St Denis, M., & Cappuccino, N. (2004). Reproductive Biology of Vincetoxicum rossicum ( Kleo .) Barb . ( Asclepiadaceae ), an Invasive Alien in Ontario Author ( s ): Melissa St . Denis and Naomi Cappuccino Source : The Journal of the Torrey Botanical Society , Vol . 131 , No . 1 ( Jan . - Mar . Journal of the Torrey Botanical Society, 131(1), 8–15.

Stergiopoulos, I., & Gordon, T. R. (2014). Cryptic fungal infections: the hidden agenda of plant pathogens. Frontiers in Plant Science, 5, 506. https://doi.org/10.3389/fpls.2014.00506

Strauss, A., White, A., & Boots, M. (2012). Invading with biological weapons: The importance of disease-mediated invasions. Functional Ecology, 26(6), 1249–1261. https://doi.org/10.1111/1365-2435.12011

Strauss, S. Y., Webb, C. O., & Salamin, N. (2006). Exotic taxa less related to native species are more invasive. Proceedings of the National Academy of Sciences of the United States of America, 103(15), 5841–5. https://doi.org/10.1073/pnas.0508073103

Syller, J. (2012, February 1). Facilitative and antagonistic interactions between plant viruses in mixed infections. Molecular Plant Pathology. Blackwell Publishing Ltd. https://doi.org/10.1111/j.1364-3703.2011.00734.x

Tall Grass Ontario. (n.d.). Summer Flowers. Retrieved June 1, 2015, from http://www.tallgrassontario.org/summer_flowers.html

Tlusty, B., Grossman, J. M., & Graham, P. H. (2004). Selection of rhizobia for prairie legumes used in restoration and reconstruction programs in Minnesota. Canadian Journal of Microbiology, 50(11), 977–983. https://doi.org/10.1139/w04-084

Toole, D. R., Cannon, G. H., Brislawn, C. J., Graves, J. M., Lamendella, R., Muth, T. R., & Muth, N. Z. (2017). Differences in soil fungal assemblages associated with native and non-native tree species of varying weediness. Biological Invasions, 1–14. https://doi.org/10.1007/s10530-017-1580-4

Turrà, D., El Ghalid, M., Rossi, F., & Di Pietro, A. (2015). Fungal pathogen uses sex pheromone receptor for chemotropic sensing of host plant signals. Nature, 527(7579), 521–524. https://doi.org/10.1038/nature15516 van der Heijden, M. G. A., de Bruin, S., Luckerhoff, L., van Logtestijn, R. S. P., & Schlaeppi, K. (2015). A widespread plant-fungal-bacterial symbiosis promotes plant biodiversity, plant nutrition and seedling recruitment. The ISME Journal, 10(2), 1–11. https://doi.org/10.1038/ismej.2015.120 van der Putten, W. H., Bardgett, R. D., Bever, J. D., Bezemer, T. M., Casper, B. B., Fukami, T., … Wardle, D. A. (2013). Plant-soil feedbacks: the past, the present and future challenges. Journal of Ecology, 101(2), 265–276. https://doi.org/10.1111/1365-2745.12054

Vandegrift, R., Roy, B. A., Pfeifer-Meister, L., Johnson, B. R., & Bridgham, S. D. (2015). The herbaceous landlord: integrating the effects of symbiont consortia within a single host. PeerJ, 3, e1379. https://doi.org/10.7717/peerj.1379

Velázquez, E., Carro, L., Flores-Félix, J. D., Martínez-Hidalgo, P., Menéndez, E., Ramírez- Bahena, M.-H., … Peix, A. (2017). The Legume Nodule Microbiome: A Source of Plant Growth-Promoting Bacteria. In Probiotics and Plant Health (pp. 41–70). Singapore: Springer Singapore. https://doi.org/10.1007/978-981-10-3473-2_3

Veneault-Fourrey, C., Plett, J. M., & Martin, F. (2013). Who is Controlling whom within the Ectomycorrhizal Symbiosis: Insights from Genomic and Functional Analyses. In F. J. de Bruijin (Ed.), Molecular Microbial Ecology of the Rhizosphere (First, pp. 501–512). Hoboken, NJ, USA: John Wiley & Sons, Inc. https://doi.org/10.1002/9781118297674.ch47

Vilà, M., Espinar, J. L., Hejda, M., Hulme, P. E., Jarošík, V., Maron, J. L., … Pyšek, P. (2011). Ecological impacts of invasive alien plants: a meta-analysis of their effects on species, communities and ecosystems. Ecology Letters, 14(7), 702–8. https://doi.org/10.1111/j.1461-0248.2011.01628.x

Vitousek, P. M., D’Antonio, C. M., & Asner, G. P. (2011). Invasions and Ecosystems: vulnerabilities and the contribution of new technologies. In D. M. Richardson (Ed.), Fifty Years of Invasion Ecology: The Legacy of Charles Elton (pp. 227–286). Oxford, UK: Wiley- Blackwell.

Vitousek, P. M., Mooney, H. A., Lubchenco, J., & Melillo, J. M. (1997). Human Domination of Earth’s Ecosystems. Science, 277(5325), 494–499. https://doi.org/10.1126/science.277.5325.494

Vrisman, C. M., Testen, A. L., Elahi, F., & Miller, S. A. (2017). First Report of Tomato Brown Root Rot Complex Caused by Colletotrichum coccodes and Pyrenochaeta lycopersici in Ohio. Plant Disease, 101(1), 247–247. https://doi.org/10.1094/PDIS-05-16-0623-PDN

Wagg, C., Bender, S. F., Widmer, F., & van der Heijden, M. G. A. (2014). Soil biodiversity and soil community composition determine ecosystem multifunctionality. Proceedings of the National Academy of Sciences of the United States of America, 111(14), 5266–70. https://doi.org/10.1073/pnas.1320054111

Wagner, M. R., Lundberg, D. S., Coleman-Derr, D., Tringe, S. G., Dangl, J. L., & Mitchell-Olds, T. (2014). Natural soil microbes alter flowering phenology and the intensity of selection on flowering time in a wild Arabidopsis relative. Ecology Letters, 17(6), 717–26.

https://doi.org/10.1111/ele.12276

Wehner, J., Antunes, P. M., Powell, J. R., Caruso, T., & Rillig, M. C. (2011). Indigenous arbuscular mycorrhizal fungal assemblages protect grassland host plants from pathogens. PloS One, 6(11), e27381. https://doi.org/10.1371/journal.pone.0027381

Wilson, H. E., Carroll, G. C., Roy, B. A., & Blaisdell, G. K. (2014). Tall fescue is a potential spillover reservoir host for Alternaria species. Mycologia, 106(1), 22–31. https://doi.org/10.3852/12-330

Wingfield, M. J., Slippers, B., Jolanda Roux, E., & Wingfi, B. D. (2011). Fifty Years of Tree Pest and Pathogen Invasions, Increasingly Threatening World Forests. In D. M. Richardson (Ed.), Fifty years of Invasion Ecology: The Legacy of Charles Elton (First, pp. 89–99). Oxford: Wiley-Blackwell.

Xiao, H., Schaefer, D. A., & Yang, X. (2017). pH drives ammonia oxidizing bacteria rather than archaea thereby stimulate nitrification under Ageratina adenophora colonization. Soil Biology and Biochemistry, 114, 12–19. https://doi.org/10.1016/j.soilbio.2017.06.024

Yamazaki, A., & Hayashi, M. (2015). Building the interaction interfaces: host responses upon infection with microorganisms. Current Opinion in Plant Biology, 23, 132–9. https://doi.org/10.1016/j.pbi.2014.12.003

APPENDICES

Appendix A Coordinates for Soil Collection Sites

Table A.1 Coodinates of field soil collection sites used in experiments of net biotic effect from V. rossicum on native plant species (Chapter 2) and native legumeous species (Chapter 3) Site Management Latitude Longitude Toronto Zoo Toronto Zoo 43.815721 -79.187969 Crows Pass Conservation Area Central Lake Ontario Conservation Authority 44.037153 -79.042661 Orono Crown Lands Village of Orono 43.978888 -78.646070 Rice Lake Conservation Area Ganaraska Conservation Authority 44.083999 -78.300137

Appendix B Plant community of Collection Sites

Plant surveys were conducted in August 2015. Two 1 m2 survey plots were used at each of the four field soil collection sites to survey each V. rossicum-invaded and uninvaded plant community. Plots were randomly located and cover abundance was estimated for all species observed using a modification of the he Braun-Blanquet scale (Dengler, Chytrý, & Ewald, 2008). This scale uses 5 categories of cover abundance percentage (75-100, 50-75, 25-50, 5-25, <5) with the last category (<5%) split into three subcategories: <5% (many shoots), <5x (few shoots) and <5r (solitary shoots) (Table B3)

The similarity between V. rossicum-invaded and uninvaded communities was estimated using Czekanowski’s Quantitative Index recommended by (Bloom, 1981) for use in ecology studies (Equation B1).

Equation B.1: Czekanowski’s Quantitative Index (Bray & Curtis, 1957)

nc= the common species between sites; n1=the species of site 1, n2= the of species at site 2

Each set of two survey plots was treated as one sample, with duplicate species removed and the total number of species assumed to be representative of its respective plant community. The number of species found in both invaded and uninvaded samples at each site were calculated and similarity indices between invaded and uninvaded plots for each site were calculated. Invaded and uninvaded communities were found to be moderately similar with an average similarity index of 0.49 (measured on a scale of 0-1; 0 being completely dissimilar, and 1 being completely similar) (Table B1).

Table B.1 Czekanowski’s Quantitative Index for similarities between plant community composition in V. rossicum- invaded and uninvaded plant populations Invasion Status of Field Site Field Collection Site Invaded Uninvaded Common Species (nc) Similarity Index

(n1) (n2) Toronto Zoo 6 6 2 0.333 Crows Pass Conservation 7 7 3 0.429 Area Orono Crown Lands 6 10 3 0.375 Rice Lake Conservation 3 4 3 0.857 Area Sum 22 27 11 0.449 Average 0.499

To obtain an estimate of between-site similarities, similarity indices were also calculated between uninvaded plant populations at each of four sites (Table B2). Most sites were similar or moderately-similar to each other (similarity indices ranging from 0.53 to 0.89) (Table B2). The field collection site located at the Toronto Zoo, Toronto, Ontario was the least like the other sites (0.15-0.25).

Table B.2 Similarity Index Matrix of plant popoulations in areas uninvaded by V. rossicum compared between four field collection sites.

Zoo Crow’s Pass Orono Rice Lake Toronto Zoo 1 Crows Pass Conservation Area 0.153846 1 Orono Crown Lands 0.25 0.533333 1 Rice Lake Conservation Area 0.2 0.888889 0.571429 1

Plant species at the field collection sites consisted of a mix of native and non-native plant species, with non-native species being the most abundant plant type. Overall there were more non-native than native plants in all plots. All plots were dominated by non-native plants. Poa pratensis and V. rossicum were found at all four collection sites in similar quantities. Vincetoxicum rossicum invaded sites had 75-100% cover abundance and it was often codominant with Poa pratensis or Bromus inermis (50-100%). “Uninvaded” populations were dominated by Poa pratensis and/or Bromus inermis (50-100% cover-abundance). Vincetoxicum rossicum was present in uninvaded plant communities but at very low densities (<5% cover abundance consisting of a few seedlings – which were avoided when collecting soil for inoculum) (Table B.4).

Table B.3 The Mean (± SE) number of species and native species per invaded and uninvaded plot (N=2 respecitvely) at each field soil collection site (N=4). Mean Number of Mean number of native Map Species/plot species/plot Site Label Invaded Uninvaded Invaded uninvaded Toronto Zoo 1 5.0 ±0 4.0 ±2 0 ±0 0 ±0 Crows Pass Conservation Area 2 4.5 ±0.5 5.0 ±0 1.0 ±0 4.0 ±1 Orono Crown Lands 3 3.5 ±0.5 7.5 ±0.5 0.5 ±0.5 2.0 ±1 Rice Lake Conservation Area 4 2.5 ±0.5 3.5 ±0.5 0 ±0 0.5 ±0.5 Average species per plot over all 5 1.625 ±1.3 sites 3.875 ±0.8 ±1.4 0.375 ±0.4

Table B.4 Braun-Blanquet percent cover plant species found in each survey plot (Invaded and Uninvaded x2) at each of the field collection sites (N=4). Additional information on including common name and native species status are listed. Invaded Uninvaded Site Plot Species Common Name Status % Cover Species Common Name Status % Cover Toronto Zoo 1 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Poa pratensis Kentucky bluegrass Non-native 75-100

Poa pratensis Kentucky bluegrass Non-native 5-25 Lolium sp. Ryegrass Non-native 5-25

Linaria vulgaris Butter and Eggs Non-native <5 Vincetoxicum rossicum Dog-strangling Vine Non-native <5

Cirsium arvense Canada thistle Non-native <5 Plantago lanceolata English plantain Non-native <5

Poaceae sp. unknown grass unknown <5 Cichorium intybus Common chicory Non-native <5x

Daucus carota Queen Ann's Lace Non-native <5x 2 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Lolium sp. Ryegrass Non-native 75-100

Poa pratensis Kentucky bluegrass Non-native 75-100 Vincetoxicum rossicum Dog-strangling Vine Non-native <5

Linaria vulgaris Butter and Eggs Non-native <5x

Vicia cracca Cow vetch Non-native <5x Cirsium arvense Canada thistle Non-native <5r

Crows Pass 1 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Poa pratensis Kentucky bluegrass Non-native 75-100

Poaceae sp. unknown grass Non-native <5r Bromus inermis Smooth brome Non-native 5-25 Conservation Area Hypericum perforatum Common St. John's Wort Non-native <5r Vicia cracca Cow vetch Non-native <5

Equisetum hyemale rough horsetail Native <5r Equisetum hyemale rough horsetail Native <5x

Asclepias syriaca Common milkweed Native <5x 2 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Bromus inermis Smooth brome Non-native 75-100

Poa pratensis Kentucky bluegrass Non-native 25-50 Equisetum hyemale rough horsetail Native 5-25

Equisetum hyemale rough horsetail Native <5x Hypericum perforatum Common St. John's Wort Non-native 5-25

Vicia cracca Cow vetch Non-native <5r Asclepias syriaca Common milkweed Native <5x

Melilotus sp. Sweet Clover sp. Non-native <5 Anemone virginiana Tall thimble weed Native <5r Orono Crown 1 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Poa pratensis Kentucky bluegrass Non-native 50-75

Bromus inermis Smooth brome Non-native 75-100 Bromus inermis Smooth brome Non-native 5-25 Lands Lolium perenne perennial ryegrass Non-native 5-25 Pilosella sp. Hawkweed sp. Non-native 5-25

Fragaria virginiana Wild strawberries Native <5 Vincetoxicum rossicum Dog-strangling Vine Non-native <5x

Asclepias syriaca Common milkweed Native <5x

Potentilla recta Sulphur cinquefoil Non-native <5x

Vicia cracca Cow vetch Non-native <5x 2 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Bromus inermis Smooth brome Non-native 50-75

Poa pratensis Kentucky bluegrass Non-native 50-75 Vincetoxicum rossicum Dog-strangling Vine Non-native 5-25

Phleum pratense Timothy-grass Non-native 5-25 Pilosella sp. Hawkweed sp. Non-native 5-25

Asclepias syriaca Common milkweed Native <5

Erigeron canadensis Horseweed Native <5

Erigeron strigosus Prairie fleabane Native <5x

Potentilla recta Sulphur cinquefoil Non-native <5x

Verbascum thapsus Common mullein Non-native <5x Rice Lake 1 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Poa pratensis Kentucky bluegrass Non-native 75-100

Bromus inermis Smooth brome Non-native 5-25 Bromus inermis Smooth brome Non-native 50-75 Conservation Area Poa pratensis Kentucky bluegrass Non-native <5 Vincetoxicum rossicum Dog-strangling Vine Non-native <5r 2 Vincetoxicum rossicum Dog-strangling Vine Non-native 75-100 Bromus inermis Kentucky bluegrass Non-native 75-100

Bromus inermis Smooth brome Non-native 5-25 Poa pratensis Smooth brome Non-native 5-25

Asclepias syriaca Common milkweed Native 5-25

Vincetoxicum rossicum Dog-strangling Vine Non-native <5r

Appendix C Nutrient Analysis of Field Soil

In August 2015, soil particle size was analysed using the Bouyoucos Hydrometer Method (Kalra & Maynard, 1991). Two samples from each of the four field soil collection sites were collected and the average %Sand, Silt and Clay was calculated (Table C.1). Differences between the two samples at each site were minimal (+/- 0.63%). Differences in soil particle size (i.e. soil texture) between sites were negligible.

Table C.1 Texture of soils from each field collection sites. Map lables correspond to coordinates in Appendix A. Average Soil Particle Size Distribution (%) +/- 0.63% Site Map Texture %Clay %Silt %Sand Label Sandy Clay Toronto Zoo 1 26.46 27.50 46.05 Loam* Crows Pass Conservation Sandy Clay 2 Area 20.30 21.92 57.79 Loam Sandy Clay Orono Crown Lands 3 20.48 11.07 68.46 Loam Rice Lake Conservation 4 Area 26.01 39.60 34.40 Loam

In August 2015, soil samples from under V. rossicum invaded and uninvaded plant communities were collected from each of the four field soil collection sites and sent to A&L Canada Laboratories for quantification. Paired t-tests (N=4) were conducted on each a variety of soil properties to compare soil fertility in invaded plant communities and uninvaded plant communities (Table C.2). Generally, V. rossicum invasion had little effect on abiotic properties. There were no significant differences between soil fertility properties in invaded areas compared to adjacent uninvaded areas. Although, there were significantly elevated Mg (ppm) in invaded soils compared to adjacent uninvaded soil (Table C.2), the differences do not appear to be ecologically relevant and were not found when the test was repeated a year later (Table C.3).

In May 2016 soil samples were again taken. This time soil from adjacent invaded and uninvaded plant communities were sampled from six sites. Specifically, the previous four field soil collection size and two additional collection sites, one east and one of Toronto (see Appendix A for coordinates). Soil samples were sent to A&L Canada Laboratories for quantification. Paired t-tests (N=6) were conducted on each a variety of soil properties to compare soil fertility in

invaded plant communities and uninvaded plant communities (Table C.3). V. rossicum invasion had no significant effect on soil fertility properties.

Table C.2 Paired t-test statistics and P-values showing the effect of V. rossicum invasion (invaded vs. uninvaded) on soil fertility of field soil collcted in August 2015 across four field collection sites (N=4, df=3). P-values <0.05 indicate significant differences between sites. Invaded Uninvaded CI for Mean Difference Soil Property M SD M SD -95% +95% t P Organic Matter 3.58 0.35 3.68 1.09 -1.47497 1.274974 -0.23 0.832 % TKN 0.35 0.06 0.35 0.06 n/a n/a n/a n/a P (Bicarb) (ppm) 42.50 22.81 36.75 25.84 -8.56364 20.06364 1.28 0.291 K (ppm) 110.00 59.82 64.25 19.72 -28.3900 119.8900 1.96 0.144 Mg (ppm) 92.50 14.43 81.25 14.93 7.271942 15.22806 9.00 0.003 Ca (ppm) 2605.00 1612.15 2770.00 1984.78 -2251.28 1921.283 -0.25 0.818 Na (ppm) 9.75 2.06 8.50 1.00 -0.752245 3.252245 1.99 0.141 Al (ppm) 718.50 331.59 696.25 410.11 -162.726 207.2264 0.38 0.727 pH 6.83 0.62 6.90 0.88 -0.814248 0.664248 -0.32 0.768 Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

Table C.3 Paired t-test statistics and P-values showing the effect of V. rossicum invasion (invaded vs. uninvaded) on soil fertility of field soil collcted in May 2016 across four field collection sites (N=4, df=3). P-values <0.05 indicate significant differences between sites. Invaded Uninvaded CI for Mean Difference Soil Property M. SD M SD -95% +95% t P Organic Matter 4.65 1.09 4.22 1.62 -0.921482 1.788149 0.82 0.448 % TKN 0.30 0.09 0.28 0.10 -0.026176 0.059510 1.00 0.363 P (Bicarb) (ppm) 37.67 18.70 38.50 27.83 -13.0972 11.43055 -0.17 0.868 K (ppm) 90.17 21.65 80.83 30.42 -10.5746 29.24129 1.21 0.282 Mg (ppm) 86.67 42.27 80.00 39.12 -7.28201 20.61535 1.23 0.274 Ca (ppm) 2990.00 2293.38 2826.67 1857.04 -1075.31 1401.982 0.34 0.748 Na (ppm) 10.17 2.40 10.50 2.66 -2.87519 2.208526 -0.34 0.750 Al (ppm) 582.67 466.66 544.00 431.73 -129.729 207.0619 0.59 0.581 pH 7.02 0.53 7.12 0.75 -0.535231 0.335231 -0.59 0.580 Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.07.

Samples of inoculum (pooled field soil from the four field-collection – see Appendix A for coordinates) that were used in the net biotic effect experiments which tested the effects of V. rossicum invasion on native plants (see Chapter 2) and native legumes (see Chapter 3) were tested to see if sterilization influenced soil fertility properties. Soil samples were sent to A&L Canada Laboratories for quantification. The difference between live and sterile field inoculum was minimal. In field soil collected in August 2015 and May 2016, pH was slightly elevated in sterile soils Table C.4, Table C.5). However, since inoculum was only added at 15% volume to the experimental substrate small differences like these would not impact the experiment.

Table C.4 Soil fertility properties of live and sterile inoculum used in the first study of this thesis (Chapter 2) which tesetd the effects of V. rossicum invasion on native plants. Organic TKN Bicarb K Mg Ca Na Al Inoculum Sterilization pH Matter (%) P (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) Invaded live 7.6 3.2 0.40 47 48 75 3830 8 784 Uninvaded live 7.7 3.2 0.30 47 52 85 3500 9 691 Invaded sterile 8.1 3 0.40 39 54 70 3430 11 693 Uninvaded sterile 8.3 2.9 0.30 49 54 60 3120 9 660

Table C.5 Soil fertility properties of live and sterile inoculum used in the second study of this thesis (Chapter 3) testing the effects of V. rossicum invasion on native legumes. Organic TKN Bicarb K Mg Ca Na Al Inoculum Sterilization pH Matter (%) P (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) Invaded live 7.2 3.9 0.30 35 89 70 2930 19 657 Uninvaded live 7.4 3 0.30 43 60 70 3260 12 652 Invaded sterile 7.7 3.4 0.30 50 94 65 3110 13 469 Uninvaded sterile 7.9 2.7 0.20 59 75 65 3140 22 587

100

Appendix D Effect of sterilization on root lesion development

The effect of sterilization was dependent the Plant species identity (Plant Species X

Sterilization’ Decay: F(5,130)=7.73, P<0.001; Herbivory F(5,130)=6.23, P<0.001). Half of the plant species had significantly more lesions in live treatments compared to sterile treatments, while the other half did not differ between treatments. One exception was D. canadense which had significantly more herbivory in sterile treatments (t= -2.31, P=0.038,

Lesions were detected in both live and sterile treatments. On average, plants grown in live treatments had 46% more decay-type lesions and 43% more herbivory-type lesions (Figure D. 1).

In sterilized treatments there were species specific responses to invasion (Plant Species X Invasion interaction). As such, Since, overall, there were significantly less lesions in sterile

101

compared to live soil treatments the methodology could be used to determine whether a greater abundance of enemies is present in soil invaded by V. rossicum compared to uninvaded soil.

30 live * 25

sterile 20

15 oot Lesions oot

10 % R % * 5 0 Decay Herbivory

Figure D. 1 Percent of root decay and herbivory-type lesions found on plants grown in live compared to sterile treatments in the first study of this thesis (Chapter 2) which tesetd the effects of V. rossicum invasion on native plants. Bars represent mean % root leasons (±SE) of the combined values of plants grown invaded and uninvaded treatments. Asterisks indicate significant differences between invaded and uninvaded treatments at P<0.05 (***P<0.01, **P=0.01, * P=0.05).

102

60 (a) Live (invaded) Live (uninvaded) 50 Sterile (invaded) 40 Sterile (invaded)

% Decay type Lesions type Decay % 10

0 AnGe DeCa RuHi SoCa AsSy DSV

7 (b) 6

4 3 2

% Herbivory type lesions type Herbivory % 1

0 AnGe DeCa RuHi SoCa AsSy DSV

Figure D. 2 Percent decay (a) and herbivory (b) -type lesions on native plant roots in all treatments from the first experiemtn of this thesis (chapter 2). Bars represent mean % root leasons (±SE) of each live and sterile, invaded and uninvaded treatment. Species abbreviations are: Andropogon gerardii (ANGE), Desmodium canadense( DeCa), Rudbeckia hirta (RuHi), Solidago canadensis (SoCa) and Asclepias syriaca (AsSy) and Vincetoxicum rossicum (ViRO).

103

Table D. 1 Statistical results of independent t-test assuming unequal variances assessing the effect of sterilization (live vs. sterile) for each species for decay and herbivory type lesions from the first experiment of this thesis (Chapter 2).

Live Sterile M SD M SD Ndf tdf t separ. P Decay A. gerardii 0.33 0.12 0.16 0.05 22 15.64 4.46 0.0004 D. canadense 0.41 0.12 0.45 0.18 22 18.73 -0.67 0.512 R. hirta 0.04 0.02 0.06 0.10 22 11.82 -0.77 0.459 S. canadensis 0.14 0.07 0.16 0.09 22 19.91 -0.58 0.571 A. syriaca 0.20 0.11 0.03 0.02 21 11.58 5.54 0.0001 V. rossicum 0.26 0.10 0.07 0.05 21 14.92 5.41 0.0001 Herbivory A. gerardii 0.05 0.02 0.03 0.02 22 21.86 2.66 0.014 D. canadense 0.01 0.01 0.02 0.02 22 13.06 -2.31 0.038 R. hirta 0.00 0.00 0.00 0.00 22 11.00 1.41 0.187 S. canadensis 0.00 0.00 0.00 0.00 22 11.00 1.45 0.175 A. syriaca 0.01 0.02 0.00 0.00 21 11.22 2.16 0.053 V. rossicum 0.01 0.01 0.00 0.00 21 11.21 2.95 0.013 M=mean; SD= standard deviation; Ndf = group degrees of freedom; tdf = adjusted degrees of freedom for unequal variances; t sep. var. = t statistic assuming unequal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05. Underlined and italicized values are of marginal significance P<0.1.

104

Appendix E Effect of sterilization on AMF Colonization

Sterilized treatments of all plants had little or no root colonisation by AMF (<0.8% combined arbuscular, vesicle and hyphal root colonization compared to 24% in live treatments).

Percentage root colonization by arbuscular mycorrhizal fungi (AMF) was effectively reduced by sterilization. Plants in live soil treatments had an average AMF colonization (AC, VC, HC combined) of 24%, while plants in sterilized treatments had an average colonization of 0.8% (Figure E. 1).

40 * live 35 sterile 30 25

15 10 * * 5 0 AC VC HC Other

Figure E. 1 Percentage root colonization by Arbuscular mycorrhizal (arbuscules (AC), vesicles (VC), and hyphae (HC)), and structures of “Other” fungi in native plants grown in live compared to sterile treatments from the first experiment of this thesis (Chapter 2) testing the effect of V. rossicum invasion on native

plants. Bars represent mean (±SE) percent colonization of the combind data from six native species (Andropogon gerardii, Desmodium canadense, Rudbeckia hirta), Solidago canadensis and Asclepias syriaca grown with invaded and uninvaded inoculum. Asterisks indicate significant differences (P<0.05) between live and sterile treatments based on a two-way ANOVA with species and sterilization as factors

of arcsine transformed data. (***P<0.01, **P=0.01, * P=0.05).

While plant species differed significantly in their response to colonization, as shown by a significant species x sterilization interaction for arbuscular (AC), vesicle (VC) and hyphal (HC) colonization (F (5,127) = 5.35, P=0.0002, F (5,127 ) = 10.56, P<0.001, F (5,157 ) = 5.29, P=0.0002 respectively); AC, HC, and VC were equally reduced due to sterilization in all species (Figure E. 1).

105

Percentage root colonization by “other fungi” differed significantly between species as well

(F(5,127 ) = 464.08, P<0.001). Solidago canadensis roots were heavily colonized (75-77%) compared to other species (1-5%). However, sterilization had no significant effect on colonization (F (1,127) = 1.98, P=0.16), nor on the species-sterilization interaction (F (5,127) = 0.17, P=0.97).

The effects of invasion on root colonization by AMF was determined using a factorial ANOVA with ‘Plant Species’ and ‘Invasion’ as cross factors, for live-treatments only (Dell STATISTICA, Dell Inc. 2015. version 13. software.dell.com). Analysis was conducted for arbuscular (AC), vesicle (VC) and hyphal (HC) colonization, and total colonization (combined AC, VC, HC). Following significant species X invasion interactions, separate t-test (assuming unequal variance) were performed for each species to test how invasion influenced responses. A few (3 random) replicates did not produce enough root material for analysis hence analysis was conducted assuming unequal variance.

Table E. 1 Independent t-test statistical results assuming unequal variances showing the effect of sterilization (live vs. sterile) for each species for arbuscular, vesicle and hyphal colonization in the first study of this thesis (Chapter 2). Response Species Live Sterile t sep 2-sided M SD M SD N df t df var. P Arbuscular colonization A. gerardii 0.07 0.04 0.00 0.00 21 11.00 6.68 0.00003 D. canadense 0.02 0.02 0.00 0.00 22 11.00 2.74 0.0191 R. hirta 0.11 0.06 0.00 0.00 22 11.00 6.18 0.00007 S. canadensis 0.10 0.08 0.00 0.00 22 11.03 4.33 0.001 A. syriaca 0.05 0.04 0.00 0.00 20 10.05 4.03 0.002 V. rossicum 0.06 0.05 0.00 0.00 20 11.00 3.99 0.002 Vesicle colonization A. gerardii 0.09 0.06 0.00 0.02 21 11.64 5.60 0.0001 D. canadense 0.02 0.02 0.00 0.01 22 11.15 3.53 0.0046 R. hirta 0.08 0.03 0.00 0.00 22 11.04 8.39 0.00004 S. canadensis 0.04 0.04 0.00 0.02 22 11.07 3.22 0.0081 A. syriaca 0.01 0.01 0.00 0.01 20 10.00 2.50 0.0315 V. rossicum 0.03 0.03 0.00 0.01 20 11.00 3.38 0.0062 Hyphal colonization A. gerardii 0.39 0.10 0.01 0.01 21 11.70 13.23 <0.001 D. canadense 0.20 0.12 0.00 0.00 22 11.06 5.36 0.0002 R. hirta 0.48 0.17 0.00 0.00 22 11.01 9.82 0.00001 S. canadensis 0.28 0.15 0.02 0.00 22 11.42 5.75 0.0001 A. syriaca 0.34 0.19 0.01 0.00 20 10.06 5.98 0.0001 V. rossicum 0.50 0.27 0.01 0.00 20 11.05 6.34 0.00005 M=mean; SD= standard deviation; Ndf = group degrees of freedom; tdf = adjusted degrees of freedom for unequal variances; t sep. var. = t statistic assuming unequal variance; 2-sided P = significance of t statistic. Bold P values are statistically significant at P<0.05.

106

Appendix F Germination response to V. rossicum rhizosphere biota

In May 2016, we collected field Lake Huron Lake Simcoe soil from six sites in southern Ontario; a combination of the four 4 5 2 3 Conservation Areas in Southern Lake Ontario Ontario, Canada from our earlier 1 experiment and two additional

sites from the same region 6 (Figure F. 1). As with our previous experiment field soil was collected from the centre of dense V. rossicum stands and Figure F. 1 Map of sampling sites used for collection of inocula (triangles): Toronto Zoo (1), Crows Pass Conservation Area (2) Orono Crown Lands adjacent plant communities to (3), Rice Lake Conservation Area (4), Long Sault Conservation Area (5), acquire soil communities that are and Kissing Bridge Trail-way (6). Cities are given as landmarks (circles). representative of “invaded” and “uninvaded” invasion history (see Chapter 2 section 2.3.1: Inoculum Collection).

Table F. 1 Coordinates of field soil collection sites used in experiment on the effects V. rossicum-associated rhizosphere biota on germination in native plants

Site Map Management Label Latitude Longitude Toronto Zoo 1 Toronto Zoo 43.815721 -79.187969 Crows Pass CA 2 Central Lake Ontario Cons. Auth. 44.037153 -79.042661 Orono Crown Lands 3 Village of Orono 43.978888 -78.646070 Rice Lake CA 4 Ganaraska Cons. Auth. 44.083999 -78.300137 Long Sault CA 5 Central Lake Ontario Cons. Auth. 44.043107 -78.740194 Kissing Bridge Trail Way 6 Woolwich Township 43.592942 -80.467150

Soil from the root zone (upper 5 cm) of 5-8 randomly selected adult V. rossicum plants was collected from each of the four conservation areas. The upper 5cm of soil was used to simulate soil communities which heterospecific seeds would initially encounter. Soil plugs of equivalent size (20 cm wide x 5 cm deep) were collected from adjacent uninvaded plant communities.

107

Collection tools were sterilized with 10% bleach to avoid contamination between invaded and uninvaded soils. Soils were placed separately into coolers for transportation back to Algoma University (1-2 days) where they were sieved (4-mm mesh) to remove roots and rocks and stored at 4oC until use. Soil samples from each collection site were sent to A&L Canada Laboratories for analysis of soil properties.

We established a germination experiment with ‘Species’ (N=5) and ‘Invasion’ (N=2) as cross factors. Germination was tested in invaded and uninvaded soils from six sites for the same five native plant species as in the previous experiment (Chapter 2).

A total of 12 Super heavy duty PermaNest Plant trays with clear domes (Growers Solutions LLC, USA; hereafter germination trays) were filled, separately, with 1L (1cm deep) of either Invaded or uninvaded field soil from each collection. Each site was treated as a separate replicate (n=6; [1 invaded tray+ 1 uninvaded tray] x 6 sites=12 trays). An additional germination tray was used for a sterile control and was filled with a growing medium, 1cm deep, composed of a 0.5cm layer of vermiculite covered by 0.5cm of non-calcareous granitic Sand (Hutcheson Sand and Mixes, Huntsville, ON). Sand and vermiculate were used as a growth medium because they provided similar soil texture, and a way to media moisture, respectively.

Seeds of the five-native species were acquired from commercial sources (Ontario Seed Co., Limited, Waterloo, Ontario) and were representative of US Midwest genetic stock (OSC pers. comm.). Vincetoxicum rossicum seeds were field-collected on August 6th, 2015 from Orono Crown Lands (Figure F. 1)) and stored in dry conditions at 4oC.

Each germination tray was sub-divided into 6 zones. Each species was randomly assigned a zone in each germination tray. Seeds of the five native-species and V. rossicum were surface sterilized seeds using the same methods as in the previous experiment (see Chapter 2). 72 seeds were planted in evenly spaced grids within their assigned zone in each germination tray such that individual seeds did not touch. A sprinkling of soil was added to the top of the seeds and the seeds were pressed into the soil to maximize contact with the soil or growth medium. The soil/medium surface was moistened by spritzing with DI water. Germination trays were covered, and seeds were cold-wet stratified (in the dark) at 4oC for 1 week (to break dormancy) before being moved to a walk-in growth chamber set at 23oC (70-90umol). Seeds were then germinated (in their respective germination trays) under controlled growth conditions in a walk-in growth chamber at Algoma University, in Sault Ste. Marie, ON., until germination stopped (8 wks.; defined as 3 consecutive days without germination in any tray). Soil moisture was

108

maintained by using covers (6” clear domes) on each tray and spritzing the soil surface of each tray every other day with DI water. Germination trays were randomly rotated around the growth chamber every other day to limit the effects of variation in lighting. Germination was counted daily.

Care was taken to sterilize all pots and utensils with 10% bleach, to prevent microbial contamination between inoculum types. Equipment was then thoroughly rinsed and dried to remove bleach that could harm seedlings and soil biota.

We tested how soil invasion history affected soil properties of the field soil used for inoculum. We used paired t-tests to detect differences between invaded or uninvaded areas from four inoculum collection sites (Dell STATISTICA, Dell Inc. 2015. version 13. software.dell.com). Separate analyses were conducted for each soil property Organic Matter (%), pH, TKN (%), P (ppm), K (ppm), Mg (ppm), Ca (ppm), Na (ppm), Al (ppm), % K, %Mg, % Ca, % Na, Saturation % P, Saturation % Al, and K/Mg Ratio.

109

Appendix G Enemy of my Enemy Literature Review

Search Criteria (latest search July 2017):

1) Random reading throughout my thesis 2) References in (Flory & Clay, 2013) 3) Web of Science search: Those that cited (Flory & Clay, 2013) = found 2/42 relevant (+3 already acquired) 4) Web of Science general search: (2013-2015*) within (plant) within (‘pathogen spillback’ or ‘pathogen spillover’) = found /213 (+2 already acquired). 5) General Google search: with at least one of the words (plant) within (2016-2017) within (‘pathogen spillback’ or ‘pathogen spillover’) =3/ 68 (+1 already acquired) 6) Web of Science search: (Enemy of my Enemy) = 4/36 (+ 1 already acquired)

In total 46 papers were searched for references to Spillover, spillback, accumulation of local pathogens or “Enemy of my Enemy” within the context of invasion and ecology. Below is a synthesis of the relevant papers both useful reviews and a comprehensive list empirical evidence (8 specifically testing spillover/spillback (Table H1)

110

Table G 1 Top Reviews and Models of Spillover/Spillback - using the terms spillover/spillback or ‘citing the Accumulation of Local Pathogens Hypothesis’ Reference Relevance to Spillover/Spillback hypothesis (Power & Mitchell, 2004) Review and experimental exploration of wild plants as reservoirs of pathogens that can ‘spillover’ to agricultural crops

(Colautti, Ricciardi, Grigorovich, & MacIsaac, 2004) Reviews Enemy Release Hypothesis and the alternate hypothesis of “Enemy of my Enemy”

(Eppinga, Rietkerk, Dekker, De Ruiter, & Van Der Putten, 2006) Development and Model of Accumulation of Local Pathogens Hypothesis – synonymous with spillover/spillback hypothesis

(Kelly, Paterson, Townsend, Poulin, & Tompkins, 2009) Reviews current state of Spillover hypothesis in invasive species in general

(Strauss, White, & Boots, 2012) Reviews current state of Spillover/Spillback hypothesis in invasive species in general

(Flory & Clay, 2013) Reviews current state of Spillover/Spillback hypothesis in invasive plants

(Faillace, Lorusso, & Duffy, 2017) Reviews in part Spillover/Spillback hypothesis in the context of plant viruses

111

Table G 2 Empirical evidence suporting Spillover of non-native parasite/pathogens from non-native host plant Spillover Native host type and origin Non-native Host and Pathogen/parasite type and Effect Reference origin origin

(Beckstead, North American grasses: Eurasia/Mediterranean SEED FUNGI: ●Both Native and non-native seeds are Meyer, Connolly, (Achnatherum hymeno ides, grass: Pyrenophora seminiperda ‘black infected. Huck, & Street, Elymus elymoides, Hesperostipa (Bromus tectorum) fingers of death’. Likely ●more native seeds had pathogen in invaded 2010) comata, Poa secunda and introduced from Eurasia than uninvaded communities – could be Pseudoroegneria spicata) spillover

(Mordecai, 2013) North American grass: Eurasia/Mediterranean SEED FUNGI: ●both native/exotic are infected (Elymus elymoides) grass: Pyrenophora seminiperda ‘black ●there is a cost to cheatgrass (Bromus tectorum) fingers of death’. Likely ●native benefited from pathogen @ community introduced from Eurasia level (Li et al., 2014) North American grass: (Spartina Eurasian grass: North American ROOT FUNGI: ●both native/exotic are infected alterniflora) (Phragmites australis – (Fusarium palustre) – study ●in field pathogen is present in mixed cultures, eurasian haplotype) provides indirect evidence lower in spartina monoculutres and absent in introduced – from north America) phragmites monocultures, and present in Spartina in home range – suggests spillover

112

Table G 3 Empirical evidence supporting Spillback of native parasite/pathogens from a non-native host plant Spillback Native host type and origin Non-native Host and Pathogen/parasite type and Effect Reference origin origin

(Mangla, Inderjit, Asian Shrub: Central/South American Asian SOIL FUNGI: ●invader-soil inhibits natives, has more & Callaway, (Bambusa arundinaceal; Poales - (Shrub) (Chromolaena (fusarium – generalist plant pathogen spores than surrounding native 2008).* Bamboo ) odorata) (Asterales) pathogens – possible sp. Id. F. veg(20maway) Origin unknown: semitectum) ●sterilization removed effects of pathogens (so (Amaranthus spinosus; its pathogen not nutrient) Caryophyllales) – South America?? (Power & North American grasses: European/Eurasian grass: VIRUS: Barley yellow dwarf ● Virus can infect many ‘wild’ plants (non- Mitchell, 2004) (Digitaria sanguinalis, (Avena fatua, Bromus Virus – origin unknown but domesticated plants) Panicum capillare) tectorum, Echinochloa others note the aphid vector is ● Study clearly establishes invasive plants as crus-galli, and Lolium from North America… reservoirs of this virus (although did not multiflorum) specifically address invasion ecology) African grass: ●Did not experimentally link observed virus (Setaria lutescens) levels to host effects

113

Table G 4 Empirical evidence supporting spillover/spillback of pathogens from a non-native host plant, but where the origin of the pathogen is not specified Spillover/Spillback Native host type and origin Non-native Host and Pathogen/parasite type and Effect Reference origin origin

(Wilson, Carroll, North America grasses: European grass: LEAF FUNGI: ●pathogen affected all 6 grass sp. Roy, & Blaisdell, (Agrostis exarata, Danthonia (Festuca arundinacea) (Alternaria sp.), not enough ●pathogen negatively affected 2 grass sp. 2014) californica and Deschampsia Effects on other information for origin ●tall fescue invader had high pathogen load, cespitosa) European/Eurasion and was affected more by the pathogen than grasses tested: (Holcus other species… some design problems are lanatus, addressed (Anthoxanthum odoratum) (Crocker, Karp, & North America grasses and Eurasian grass SOIL OOMYCETE`… mostly ●seeds are not affected Nelson, 2015) Forbs: (Phragmites australis - – phythium but there are others, ●seedlings are affected (Grass) Phragmites australis eurasian haplotype) isolated oomycetes for tests ●oomycete isolates from phragmities americanus, Muhlenbergia Effects on invaded and uninvaded are equally virulent glomerata, Phalaris arundinacea Eurasian/North African on native exotic species. - no difference b/t (Forb) Epilobium glandulosum , Forb tested: (Lythrum pathogen effects on natives compared to Euthamia graminifolia salicari) exotics (Crocker, North America grasses and Eurasian grass SOIL FUNGI + OOMYCETES ●seeds not affected (although seed Lanzafane, Karp, & Forbs: (Grass) Phragmites (Phragmites australis - – including phythium – unknown pathogens present) Nelson, 2016) australis, Calamagrostis eurasian haplotype) origin ●seedlings are affected by common fungal canadensis, Muhlenbergia Effects on pathogens glomerata (Sedge) Eurasian/North African ●No difference b/t pathogen (invader-soil) Carex comosa, Carex frankii Forb tested: (Lythrum effects on natives compared to exotic (some (Forb) salicari) but not all species were negatively impacted) Asclepias incarnata, Epilobium glandulosum

114

Table G 5 Emprical evidence with no use of the terminology of spillover/spillback but examine the concept in general and could be used for guiding spillover/spillback studies Not specified Reference Native host type and Non-native Host and Pathogen/parasite type and Effect origin origin origin (C.M. Malmstrom, Hughes, North American grasses: Eurasian grasses: VIRUS: Barley yellow dwarf ● Aphids are more abundant among invasive Newton, & Stoner, 2005; (Elymus glacus, Elymus (Avena fatua, Brumus virus – origin unknown plants. Carolyn M. Malmstrom, multisetus and Nassella sp.) Transmitted via a North ● virus plus other mechanisms may be McCullough, Johnson, pulchra) American aphid species. important for invasion Newton, & Borer, 2005; ● did not specifically link higher levels of Carolyn M. Malmstrom, Shu, infection to negative effects in native grass Linton, Newton, & Cook, populations – e.g., costs of different densities of 2007) in (Strauss et al., aphids 2012) (Callaway, Thelen, Barth, North American grasses: European Forb: FUNGI: North American: ● invasion enhanced in presence of native soil Ramsey, & Gannon, 2004; fungi (AMF) by reducing native plant cover (Festuca idahoensis, and (Centaurea maculosa) Soil Biota including AMF Marler, Zabinski, & relative to invading plants others) Callaway, 1999) in (Strauss ●Soil fungi plus other mechanisms may be et al., 2012) important for invasion ● This stretches the definition of spillback in this thesis, and the term was not used in this source, but it is worth mentioning because of arguments over the continuum of mutualist to parasite/pathogen

(Day et al., 2016) North American Forb: Eurasian Forb: FUNGI: origin not specified ●V. rossicum contains fungi that are listed as (Vincetoxicum rossicum) plant pathogens in central database (Asclepias syriaca and Solidago canadensis) ● Isolated soil fungal pathogens promote V. rossicum growth but inhibit S. canadensis growth but only when at least two fungal isolates are combined

115

Appendix H Seedling survival

Table H 1 Seeding survival rates of legumes seedlings grown during the second experiment of this thesis (Chapter 3) . Species Invasion Sterilization Survived of 8 %Survival DeCa Invaded Live 8 100 LeHi Invaded Live 6 75 LuPe Invaded Live 8 100 AsCa Invaded Live 8 100 ViRo Invaded Live 8 100 DeCa Invaded Sterile 8 100 LeHi Invaded Sterile 7 87.5 LuPe Invaded Sterile 8 100 AsCa Invaded Sterile 8 100 ViRo Invaded Sterile 8 100 DeCa Uninvaded Live 8 100 LeHi Uninvaded Live 6 75 LuPe Uninvaded Live 7 87.5 AsCa Uninvaded Live 8 100 ViRo Uninvaded Live 8 100 DeCa Uninvaded Sterile 8 100 LeHi Uninvaded Sterile 5 62.5 LuPe Uninvaded Sterile 7 87.5 AsCa Uninvaded Sterile 7 87.5 ViRo Uninvaded Sterile 7 87.5

116

Appendix I Effects of Electrical Fire on 2nd (Fabaceae) experiment

Table I 1 Native legume height (cm) from live treatments (only) at wk 7 - emediately after the fire experienced during the second experiment of this thesis (Chatper 3). Species Invasion Mean SD N

AsCa Invaded 12.56 4.995 8 Uninvaded 13.91 2.050 8 Total 13.24 3.754 16 DeCa Invaded 22.35 4.228 8 Table I 2 Analysis of Varience (ANOVA) statistics showing the effect of V. Uninvaded 16.36 2.915 8 rossicum invasionin live treatments on native legumes grown during the Total 19.36 4.677 16 second experiment of this thesis (Chapter 3) at 7 weeks growth (i.e. ViRo Invaded 12.79 4.500 8 emmediately after the fire). Uninvaded 15.33 2.767 8 Type III Sum Mean Total 14.06 3.839 16 Source of Squares df Square F Sig. LeHi Invaded 5.30 3.370 6 Species 2.937 4 .734 9.753 .000 Uninvaded 8.02 2.309 6 Invasion 1.715 1 1.715 22.781 .000 Total 6.66 3.099 12 Sp*Invasion .816 4 .204 2.709 .038 LuPe Invaded 6.80 1.969 8 Error 4.893 65 .075 Uninvaded 8.01 1.649 7 Total 607.640 75 Total 7.37 1.870 15

117

Table I 3 Native legume height (cm) from live treatments (only) at wk 13 - emediately after the fire experienced during the second experiment of this thesis (Chatper 3). Species Invasion Mean SD N

AsCa Invaded 27.75 10.905 8 Uninvaded 32.19 3.229 8 Total 29.97 8.100 16 DeCa Invaded 38.00 8.763 8 Uninvaded 32.06 5.641 8 Total 35.03 7.751 16 ViRo Invaded 26.00 5.782 8 Table I 4 Analysis of Varience (ANOVA) statistics showing the effect of V. Uninvaded 30.69 7.653 8 rossicum invasion in live treatments on native legumes grown during the Total 28.34 6.985 16 second experiment of this thesis (Chapter 3) at 13 weeks growth (i.e. LeHi Invaded 15.08 11.893 6 emmediately after the fire). Uninvaded 24.83 5.981 6 Type III Sum Mean Total 19.96 10.319 12 Source of Squares df Square F Sig. LuPe Invaded 9.69 3.139 8 Species 1.282 4 .321 3.324 .015 Uninvaded 9.50 3.175 7 Invasion .742 1 .742 7.697 .007 Total 9.60 3.043 15 Sp*Invasion 1.108 4 .277 2.871 .030 Total Invaded 23.74 12.951 38 Error 6.268 65 .096 Uninvaded 26.35 10.042 37 Total 873.732 75 Total 25.03 11.604 75

118

Table I 5 Native legume height (cm) from sterile treatments (only) at wk 7 - emediately after the fire experienced during the second experiment of this thesis (Chatper 3). Species Invasion Mean SD N

AsCa Invaded 14.01 2.558 8

Uninvaded 16.24 3.473 7

Total 15.05 3.125 15

DeCa Invaded 19.29 3.847 8

Uninvaded 24.98 5.934 8

Total 22.13 5.654 16

ViRo Invaded 19.45 5.632 8 Table I 6 Analysis of Varience (ANOVA) statistics showing the effect of V. Uninvaded 21.19 7.412 7 rossicum invasion in sterile tretments on native legumes grown during the Total 20.26 6.341 15 second experiment of this thesis (Chapter 3) at 7 weeks growth (i.e. LeHi Invaded 16.59 8.785 7 emmediately after the fire). Uninvaded 11.98 6.805 5 Total 14.67 8.035 12 Type III Sum Mean LuPe Invaded 8.28 1.594 8 Source of Squares df Square F Sig. Uninvaded 8.20 2.045 7 Species 5.183 4 1.296 4.823 .002 Total 8.24 1.751 15 Invasion .381 1 .381 1.418 .238 Total Invaded 15.49 6.357 39 Sp*Invasion .585 4 .146 .545 .704 Uninvaded 17.03 8.095 34 Error 16.924 63 .269 Total 16.21 7.208 73 Total 385.849 73

119

Table I 7 Native legume height (cm) from sterile treatments (only) at wk 13 - emediately after the fire experienced during the second experiment of this thesis (Chatper 3). Species Invasion Mean SD N

AsCa Invaded 23.44 5.017 8

Uninvaded 26.93 8.541 7

Total 25.07 6.863 15

DeCa Invaded 27.06 10.325 8

Uninvaded 32.50 15.829 8

Total 29.78 13.212 16 Table I 8 Analysis of Varience (ANOVA) statistics showing the effect of V. ViRo Invaded 43.56 10.048 8 rossicum invasion in sterile tretments on native legumes grown during the Uninvaded 37.36 12.671 7 second experiment of this thesis (Chapter 3) at 13 weeks growth (i.e. Total 40.67 11.382 15 emmediately after the fire). LeHi Invaded 33.64 17.329 7 Uninvaded 27.60 15.856 5 Type III Sum Mean Total 31.13 16.276 12 Source of Squares df Square F Sig. LuPe Invaded 11.88 3.410 8 Species 14.105 4 3.526 12.120 .000 Uninvaded 10.21 1.410 7 Invasion .116 1 .116 .400 .529 Total 11.10 2.720 15 Sp*Invasion .273 4 .068 .235 .918 Total Invaded 27.77 14.441 39 Error 18.330 63 .291 Uninvaded 27.04 14.709 34 Total 613.821 73 Total 27.43 14.469 73

120

Appendix J Relationship between NBE and phylogenetic distance

Figure J1 Linear regression of the natural log of NBE in total biomass of native plant species from rhizosphere biota originating in soil invaded by Vincetoxicum rossicum (y axis See Chapter 2) and the corresponding phylogenetic distance (x axis) of native plants to V. rossicum. Values for phylogenetic distance are the total genetic variability of all species tested (based on nucleotide substitutions per site) less the covarience (i.e. the number of nucleotide subsitutions shared by a given plant and V. rossicum; ie shared branch length). Values for NBE have no units and are based on the natrual log of the bootstrapped ratios between live and sterile biomass. Mean NBE records (gray circles) above the the reference line at the y-intersect of 0 indicate a positive effect on plant biomass. Mean NBE below zero indicate a negative effect on plant biomass. R2 is calculated from the line of best fit (black diagonal line).

121

Figure J2 Linear regression of the natural log of NBE in total biomass of native plant species from rhizosphere biota originating in soil uninvaded by Vincetoxicum rossicum (y axis See Chapter 2) and the corresponding phylogenetic distance (x axis) of native plants to V. rossicum. Values for phylogenetic distance are the total genetic variability of all species tested (based on nucleotide substitutions per site) less the covarience (i.e. the number of nucleotide subsitutions shared by a given plant and V. rossicum; ie shared branch length). Values for NBE have no units and are based on the natrual log of the bootstrapped ratios between live and sterile biomass. Mean NBE records (gray circles) above the the reference line at the y-intersect of 0 indicate a positive effect on plant biomass. Mean NBE below zero indicate a negative effect on plant biomass. R2 is calculated from the line of best fit (black diagonal line).

122

Table J1 Linear regression satistics on the correlation between the able J2 ANOVA statistics showing the significance of the predicitve power phylogenetic distance of native plants and its predictive capacity (R) to of phylogenetic distance (R statistic) to determine net biotic effect of determine net biotic effect (lnNBE_total) of rhizosphere biota from invaded rhizosphere biota from Vincetoxicum rossicum invaded soils on native plant soil on native plant biomass. biomass

R R Square Adjusted R Std. Error of Square the Estimate Sum of Mean Model Squares df Square F Sig. 0.159 0.025 -0.300 0.976 Regression 0.074 1 0.074 0.078 0.799 Residual 2.855 3 0.952

Total 2.929 4

rhizosphere biota from Vincetoxicum rossicum uninvaded soils on native plant biomass Table J3 Linear regression satistics on the correlation between the phylogenetic distance of native plants and its predictive capacity (R) to Sum of Mean determine net biotic effect (lnNBE_total) of rhizosphere biota from uninvaded soil on native plant biomass. Model Squares df Square F Sig. Regression 0.014 1 0.014 0.032 0.869 R R Square Adjusted R Std. Error of Square the Estimate Residual 1.259 3 0.420

0.103 0.011 -0.319 0.648 Total 1.272 4

Table J4 ANOVA statistics showing the significance of the predicitve power of phylogenetic distance (R statistic) to determine net biotic effect of