UG ETD Template

Home , Contarinia, Contarinia nasturtii, Inostemma, Macroglenes, Synopeas

Distribution, Abundance, and Biological Control Potential of Synopeas myles, a Parasitoid of the Swede Midge

Charles-Étienne Ferland

A Thesis presented to The University of Guelph

In partial fulfilment of requirements for the degree of Master of Science in Environmental Sciences

Guelph, Ontario, Canada

ABSTRACT

DISTRIBUTION, ABUNDANCE, AND BIOLOGICAL CONTROL POTENTIAL

OF SYNOPEAS MYLES, A PARASITOID OF THE SWEDE MIDGE

Charles-Étienne Ferland Advisor:

University of Guelph, 2020 Professor R. H. Hallett

The swede midge, Contarinia nasturtii Kieffer (Diptera: Cecidomyiidae), is a major invasive pest of canola and other plants in the Brassicaceae family. Current pest management practices fail to ensure adequate crop protection. The parasitoid

Synopeas myles Walker (Hymenoptera: Platygastridae), an adventively introduced natural enemy of C. nasturtii, was discovered in Ontario in 2016. The overall goal of this project was to quantify the potential of S. myles as a biological control candidate to use in an integrated pest management approach for C. nasturtii in Ontario, as well as to assess possible biological control practices to promote this parasitoid. Synopeas myles was found to be widespread across canola growing regions of Ontario, with parasitism rates close to 6% on average. Floral supplements showed promising potential to increase parasitoid longevity, but not fecundity. Parasitism rates were positively associated with percent managed surroundings and presence of wild flowers near field edges, and negatively associated with soil silt content and C. nasturtii population density.

ACKNOWLEDGEMENTS

It was in the late-1990s that, as an insect-enthusiastic child, I met AAFC entomologist Dr. Charles Vincent, who lived across from my family on Oak Street, St-

Lambert (Vincent 2018). I am grateful for his patience in examining the specimens I would bring to him then (most of them isopods and myriapods I’m afraid), and for his generosity with all the books he lent me. Years later, I was introduced to the swede midge, Contarinia nasturtii Kieffer (Diptera: Cecidomyiidae), when working in an agronomic services company near Montreal with Julie Nichols, Nancy Palardy and

Gabriel Muñoz. How could such a small fly cause such devastating damage in cruciferous crops? Interested in the family Cecidomyiidae, I had some pleasant correspondences with Dr. Raymond Gagné and then sought further entomological experience at the Station Linné in Sweden in the winter of 2016 where I met, among others, Dave Karlsson, Annie Lander and Drs. Mathias and Catrin Jaschoff, all of whom encouraged me to pursue further studies. I am thankful to you all who accompanied me on these milestones of my entomological journey. And so, to the University of Guelph I came, to meet with the Canadian expert on C. nasturtii who would become my thesis advisor.

I wish to thank my advisor, Dr. Rebecca Hallett, for her support and guidance throughout our project. I also assure her that, with a great many efforts, I have left science-fiction out of this thesis. Should any reader encounter oversized anthropophagous wasps in these pages, please let me know; they are in the wrong

iii

book. I also wish to thank Dr. Paul Abram for his help and advice during this project, but also once more for our meeting together at the botanical garden of Montreal in 2015 before I started my master’s degree, for taking the time to discuss C. nasturtii and parasitoids, and answer questions I had about grad school.

Thank you to Jonathon Williams and Jamie Heal for welcoming me in the biological control laboratory in 2016,and helping me learn the standard operating protocols with respect to greenhouse maintenance and C. nasturtii rearing. I recall that

Jon and I met in 2013 or 2014 at one of the first annual meetings of the Entomological

Society of Canada I attended, and already discussed C. nasturtii back then! Many thanks to our teams of undergraduate students who worked very hard and persevered in the long days despite the heat in the midst of summer: Charlotte Mackay, Jamie

Bauman, Kate Lindsay, Elizabeth Porter, Jarrett Blair, Daniel Humphrey, Tasha Valente,

Kylie Barwise, Emily Sousa and Linley Sherin. Thank you to Spencer McGregor for conducting the 2016 survey, to our OMAFRA collaborator Meghan Moran, Carrie

James, Deb Campbell, and the Ontario Canola Growers Association. Thank you to Dr.

Boyd Mori, Jennifer Holowachuk, Dr. Peter Buhl and Dr. Owen Lonsdale for helping us identify our hymenopteran specimens. I would also like to thank canola growers John

Kidd, Ray McCabe, Leo and Jon Blydorp, Laverne Trimble, Brian Besley, Brian Wiley,

Holmes Agro and the Elora Research Farms for allowing us access to your fields and contributing to applied agricultural research. Thank you to friends and colleagues Dr.

Elisabeth Hodgdon, Jenny Liu, Cassie Russell, Matthew Muzzatti, Carol McLennan and

Dr. Angela Gradish. Thanks to Dr. Elisabeth Franklin for her help with assessing flower

presence within the vicinity of the canola fields, Dr. Michelle Edwards for her patience and support with my statistics experience, and Paula Martin from the UofG Data

Resource Centre for her help with geomatics tools. Thank you also to the team at the

Centre de recherche sur les grains inc. (CÉROM) for their help and support, and taking me on as a laboratory technician while I was in the final miles of my thesis, with particular thanks to Dr. Sébastien Boquel and Alexis Latraverse, and summer students

Sandrine Corriveau Tousignant, Frédérique Pilote, Cyril Le Maux, Étienne Minaud, and

Caroline Ferland. Special thanks also to Dr. Cynthia Scott-Dupree, Dr. Steve

Marshall, Dr. Steve Paiero and Dr. Morgan Jackson for broadening my understanding of entomology.

Carla Parodi, thank you for your presence at my side in the laboratory and in the fields, and for your everlasting support throughout this big project. You have always been there to cheer me up, and I could not have done it without you. Thank you to my parents who supported my interest in entomology from the very beginning, and to my family and friends who have supported me. Katie, Andy, Mickey, and Murphy thanks for the long walks, witty remarks, and insightful conversations.

I also wish to thank our funding sources for their financial support, the OMAFRA-

UofG Production Systems Research Program, Bunge Canada, Bayer Crop Sciences,

Ontario Canola Growers Association, and the Eastern Canada Oilseeds Development

Alliance (ECODA).

TABLE OF CONTENTS

Abstract ...... ii

Acknowledgements ...... iii

Table of Contents ...... vi

List of Tables ...... viii

List of Figures ...... ix

1 INTRODUCTION ...... 1

1.1 CONTARINIA NASTURTII ...... 2

1.1.1 Distribution and life history ...... 2

1.1.2 Pest status in Canada ...... 4

1.2 CONTARINIA NASTURTII MANAGEMENT ...... 6

1.2.1 Biological control of C. nasturtii ...... 6

1.3 SYNOPEAS MYLES ...... 10

1.3.1 Synopeas myles biology ...... 11

1.3.2 Synopeas myles as a potential biological control agent ...... 13

1.3.3 Maximum parasitism rates ...... 17

1.4 SUMMARY ...... 19

2 DISTRIBUTION AND ABUNDANCE OF AN ADVENTIVELY ESTABLISHED PARASITOID OF SWEDE MIDGE IN ONTARIO, CANADA ...... 20

2.1 INTRODUCTION ...... 20

2.2 MATERIALS & METHODS ...... 24

2.3 RESULTS ...... 31

2.4 DISCUSSION ...... 43 vi

3 ENVIRONMENTAL FACTORS AFFECTING PARASITISM ...... 54

3.1 INTRODUCTION ...... 54

3.2 MATERIALS & METHODS ...... 59

3.3 RESULTS ...... 68

3.4 DISCUSSION ...... 74

4 GENERAL DISCUSSION ...... 82

References Cited ...... 88

APPENDIX A ...... 104

vii

LIST OF TABLES

Table 2.1. Canola fields sampled for Synopeas myles in 2016, 2017, and 2018...... 26

Table 2.2. Sampling regimes of the central Ontario area collections and focus locations by year...... 27

Table 2.3. Mean percent parasitism rate of Contarinia nasturtii by Synopeas myles; total number of sampling events, S. myles-positive sampling events, and S. myles and individuals collected; and mean maximum S. myles parasitism rate (% ± SE) from canola fields sampled for S. myles in 2016, 2017, and 2018...... 34

Table 2.4. Weekly parasitism rate (%) of Contarinia nasturtii by Synopeas myles at central Ontario canola fields sampled weekly in 2016 (A), 2017 (B), and 2018 (C)...... 37

Table 2. 5. Mean number of days ± S.E. from collection of canola plant samples to adult emergence of Contarinia nasturtii and Synopeas myles at central Ontario and provincial survey sites...... 38

Table 2.6. Estimated dates of emergence of overwintered S. myles adults from the central Ontario collection, based on parasitoid development times from larval collection to adult emergence...... 42

Table A. 1. Sample sizes and production practice information provided by canola growers...... 107

viii

LIST OF FIGURES

Figure 2.1. Canola fields (n = 60) sampled for Synopeas myles in 2016, 2017, and 2018 Weekly sites (n = 16) are displayed as empty circles. Provincial sites (n = 44) are displayed as black circles. The districts of Kenora, Rainy River, and Thunder Bay are not shown as no sampling occurred in these locations...... 25

Figure 2.2. Canola fields (n = 60) sampled for Synopeas myles in 2016, 2017, and 2018. Black and empty circles indicate fields where S. myles was detected (n = 37) or not detected (n = 23), respectively. Districts of Kenora, Rainy River, and Thunder Bay are not shown as no sampling occurred in these locations...... 36

Figure 2.3. (A) Daily mean temperature (°C) and daily precipitation (mm) in Mono, ON in 2016 (Environment Canada 2019). (B) Seasonal abundance of total Synopeas myles (black) and Contarinia nasturtii (white) emergences from all canola sites from central Ontario in 2016. Data is displayed by insect emergence date, not plant collection...... 39

Figure 2.4. (A) Daily mean temperature (°C) and daily precipitation (mm) in Mono, ON in 2017 (Environment Canada 2019). (B) Seasonal abundance of total Synopeas myles (black) and Contarinia nasturtii (white) emergences from all canola sites from central Ontario in 2017. Data is displayed by insect emergence date, not plant collection...... 40

Figure 2.5. (A) Daily mean temperature (°C) and daily precipitation (mm) in Mono, ON in 2018 (Environment Canada 2019). (B) Seasonal abundance of total Synopeas myles (black) and Contarinia nasturtii (white) emergences from all canola sites from central Ontario in 2018. Data is displayed by insect emergence date, not plant collection...... 41

Figure 3.1. Significant associations of parasitism rate of Synopeas myles with environmental parameters in 2017 & 2018 (excluding vegetative cover): percent 2 2 managed surrounding (LRT; χ = 7.72, df = 1, P = 0.00547), percent silt in soil (LRT; χ = 12.50, df = 1, P = 0.000413), and cumulative number of male Contarinia nasturtii 2 captured in 4 pheromone traps per field (LRT; χ = 5.28, df = 1, P = 0.0215) from first to last capture (May 24 to October 30). Note that fits are not on partial residuals; the black line represents logistic model predictions based on all model factors and data points on each plot show raw data values...... 70

Figure 3.2. Association between parasitism rate of Synopeas myles and mean percent cover of flowers in field edges (LRT; χ2 = 14.10, df = 1, P = 0.000172) or percent silt in the field soil (LRT; χ2 = 31.50. df = 1, P = 0.0000000198) in 2018 (including vegetative cover). The black line represents the logistic curve. Note that fits are not on partial

residuals; the black line represents logistic model predictions based on all model factors and data points on each plot show raw data values...... 71

Figure 3.3. Total number of sampling events from which adult Contarinia nasturtii (black) and Synopeas myles (white) emerged at various distances (m) from the nearest field edge (n plants sampled = 8,156) in 2018 (whole season)...... 71

Figure 3. 4. Survival probability of male and female Synopeas myles according to treatment (control, C; sweet alyssum, SA; and sugar) and sex (total n = 210). Curves labeled with different letters have significantly different mean survival times, according to Tukey multiple comparisons on the cox model, using the multcomp package (P < 0.01)...... 72

Figure 3. 5. Mean number of oocytes per female S. myles (n = 79) according to sugar treatment and age of Synopeas myles (days). The bold horizontal line inside the boxplots represents the mean, the gray boxes represent quartiles 2 and 3 (from percentile 25 to 75), and the vertical line shows the distribution of 95% of the data. The circle represents an outlier...... 73

Figure 3. 6. The mean number of eggs laid by S. myles in C. nasturtii larvae by treatment (n = 39). The bold horizontal line inside the boxplots represents the mean, the gray boxes represent quartiles 2 and 3 (from percentile 25 to 75), and the vertical line shows the distribution of 95% of the data. The circles represent outliers...... 73

Figure A. 1. Canola growing practices questionnaire submitted to producers...... 104

Figure A. 2. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the date at which the crop was seeded (day of year) (±SE) (n sites = 11)...... 108

Figure A. 3. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the seed variety used (±SE) (n sites = 11)...... 108

Figure A. 4. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the seed treatment applied (±SE) (n sites = 11)...... 109

Figure A. 5. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of insecticide applications (±SE) (n sites = 11)...... 109

Figure A. 6. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of herbicide applications (±SE) (n sites = 11)...... 110

Figure A. 7. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of fungicide applications (±SE) (n sites = 11)...... 110

Figure A. 8. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of summers without canola in that field (±SE) (n sites = 11)...... 111

Figure A. 9. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the presence of adjacent canola in previous year (±SE) (n sites = 11)...... 111

1 INTRODUCTION

In Canada, the production of canola (Brassica napus L., B. rapa L., and B. juncea

L.) occupies 9 million hectares (Statistics Canada 2017) and contributes $26.7 billion to the economy annually (Canola Council of Canada 2017a). Since 2000, one of the most challenging pests Ontario Brassica growers have faced is the swede midge, Contarinia nasturtii (Kieffer 1888) (Diptera: Cecidomyiidae). Contarinia nasturtii was first reported in Ontario canola in 2003 (Hallett 2017), although potential symptoms of C. nasturtii injury had been reported as early as 1996 (Hallett and Heal 2001). This small fly lays its eggs on meristematic tissue and in developing flower buds. By extraoral digestion of the plant, C. nasturtii larvae trigger swelling and distortion of tissues, and consequently mild to severe economic damage (Hallett 2017). Current management practices for C. nasturtii rely heavily on insecticides and are not always effective. Therefore, there is a need to develop alternative management tactics for C. nasturtii (Hallett 2017).

The parasitoid Synopeas myles (Walker 1836) (Hymenoptera: Platygastridae) is a natural enemy of C. nasturtii (Abram et al. 2012a, b) that was discovered in Ontario in

2016 (Hallett and McGregor, unpublished data). The introduction of S. myles is assumed to be accidental. In Europe, parasitism rates of C. nasturtii by S. myles are usually low, close to 3% on average, and rarely as high as 28% (Abram et al. 2012a). A

2016 survey in Ontario canola agroecosystems found S. myles parasitism rates of 13% on average, with a maximum of 30% (Hallett and McGregor, unpublished data). The aim of this MSc research project was to determine the distribution, abundance, and

biological control potential of S. myles in Ontario, with the ultimate goal of determining whether S. myles could be a useful biological control agent of C. nasturtii in Ontario and how it could be integrated in a biological control program against C. nasturtii.

The purpose of this introductory literature review is to investigate the current state of knowledge on the parasitoid S. myles. An introduction to C. nasturtii, S. myles, and the role of biological control within integrated pest management programs are included.

1.1 CONTARINIA NASTURTII

1.1.1 Distribution and life history

Contarinia nasturtii a member of the Cecidomyiidae, the highly diverse family of gall midges. The family contains 6,203 described species in 736 genera (Gagné and

Jaschhof 2014). Contarinia nasturtii belongs to the subfamily Cecidomyiinae and is best known for its association with plants in the Brassicaceae family.

In the Palearctic region, C. nasturtii is present in southwestern Asia (Olfert et al.

2006), Austria, Belgium, Bosnia and Herzegovina, Bulgaria, Czech Republic, France,

Greece, Ireland, Italy, Kosovo, Latvia, Lithuania, Malta, Norway, Poland, Romania,

Serbia, Sicily, Slovakia, Slovenia, Sweden, Switzerland, Turkey, Ukraine, and throughout the Iberian Peninsula (Olfert et al. 2006). In the Nearctic, its invaded range,

C. nasturtii is found in Ontario, Manitoba, Québec, Prince Edward Island, Nova Scotia,

New York, Connecticut, Massachusetts, New Jersey, Vermont, Ohio, Michigan, and

Minnesota (Chen et al. 2011, Canola Council of Canada 2017b, Philips et al. 2017). It was believed that C. nasturtii had reached the Canadian Prairies but, in fact, another species has been found to be infesting prairie canola: Contarinia brassicola Sinclair

(Mori et al. 2019). Contarinia nasturtii was not even found in later surveys, casting doubt on whether it was even present there in the first place. According to bioclimatic models

(Olfert et al. 2006, Mika et al. 2008), C. nasturtii has the potential to establish in most of

Canada, with southwestern British Columbia, southern Ontario and Québec, New

Brunswick, Nova Scotia, and Prince Edward Island being the most suitable habitats.

There are concerns about the damage C. nasturtii is currently causing in its invaded range and also about its further spread in North America (Chen et al. 2011), and therefore there is interest in improving management practices.

In late-May and early-June, C. nasturtii females lay opaque, white, oval eggs (ca.

0.37 mm long) of unisexual broods (Barnes 1950) in batches of 2 – 50 on the developing tissue of the host plant (Readshaw 1961, 1966). Depending on temperature, eggs hatch within 1 – 10 days. Larvae feed gregariously by extraoral dissolution of the plant cuticle, liquefying cells beneath and digesting the fluid for 7 – 21 days (Readshaw

1961, 1966). The larvae are the only stage that causes host injury (Readshaw 1961).

Fourth instar C. nasturtii larvae jump off their host and enter the top 1 cm of soil, where they pupate and either spin a spherical cocoon (if entering diapause) or an oval cocoon (if not diapausing) (Readshaw 1961, 1966). The pupa exits the cocoon and wriggles towards the soil surface, where the imago frees itself of its pupal case. The

duration of C. nasturtii life cycle varies according to temperature and soil moisture

(Readshaw 1961); in Ontario, C. nasturtii can complete its entire cycle in 3 weeks

(Hallett et al. 2009a).

Contarinia nasturtii adults live 1 – 4 days in the wild (Barnes 1946; Readshaw

1961, 1966). They are non-feeding but sometimes drink from moistened surfaces

(Readshaw 1961). Females emerge from pupation with their full complement of eggs, ready to be fertilized within 8 – 12 hours (Readshaw 1961). Females can lay up to 95 eggs in a lifetime (Readshaw 1961), although 18 – 28 eggs have been more commonly observed in laboratory settings (Hallett 2017).

Contarinia nasturtii adults are 2 – 3 mm long and have hairy wings with reduced venation, folded horizontally over the pale yellow abdomen at rest (Barnes 1946,

Readshaw 1961). Females have a telescopic ovipositor (Readshaw 1961). Claspers in males are a key distinguishing feature in species-level identification (Harris 1966). The antennae are sexually dimorphic: 12 flagellomeres, bilobed in males (giving an impression of 24 beads) and cylindrical in females (Readshaw 1961).

In Ontario and Québec, there are 2 emergence phenotypes, each with 4 generations of C. nasturtii per growing season (Hallett et al. 2009a), with a possible fifth generation under future climate models (Hallett et al. 2009b). There is also evidence for a third phenotype emerging later in the summer (Hallett et al. 2009b).

1.1.2 Pest status in Canada

Contarinia nasturtii was first reported in Canada in 2000 (Hallett and Heal 2001).

It had likely been present since 1996 or before, but plant injury symptoms during that time were confused with molybdenum deficiencies, temperature stress, or mechanical injury (Hallett and Heal 2001).

Over 100 cruciferous (Brassicaceae) plants are suitable for C. nasturtii development (Stokes 1953a, b; Hallett 2007; Chen et al. 2009). Some plant hosts are of economic importance, including canola (B. napus L., B. juncea L., B. rapa L.), Brussels sprouts (B. oleracea L. var. gemmifera), cabbage (B. oleracea L. var. capitata, tuba, and sabauda), broccoli (B. oleracea L. var. botrytis), turnip (B. rapa L. var. rapa), and cauliflower (B. oleracea var. botrytis) (Barnes 1946; Stokes 1953a; Readshaw 1961;

Hallett 2007). Characteristic symptoms of C. nasturtii-infested plants are leaf crumpling, swelling of buds and petioles, and corky scars (Barnes 1946). In canola, severe C. nasturtii infestation can lead to death of meristematic tissue, preventing stem elongation and causing flower buds to rot due to secondary pathogen invasions (Hallett 2017).

Scouting for C. nasturtii is challenging (Kikkert et al. 2006) because symptoms of plant injury may be confused with mechanical injury, herbivory, herbicide phytotoxicity, genetic variation, or temperature stress (Barnes 1946); and injury may not appear until

5 – 10 days after larval feeding has begun (Hallett 2007). Contarinia nasturtii injury decreases produce marketability and crop yield, and in North America, commercial cruciferous vegetable crops have suffered losses as high as 85% (Hallett and Heal

2001).

1.2 CONTARINIA NASTURTII MANAGEMENT

Contarinia nasturtii owes its success partly to a plastic life history (Hallett 2017). Its cryptic larval lifestyle, overlapping multivoltinism, variable diapause patterns, wide host range, lack of natural enemies, and short lifespan must be taken into account in the development of an integrated pest management (IPM) program (Hallett 2017). IPM programs draw on a range of tactics, including biological, physical, cultural, mechanical, behavioural, genetic and chemical control techniques, to suppress pests effectively and economically while minimizing environmental impacts (Luckmann and Metcalf 1994). An

IPM approach is important for managing C. nasturtii as insecticides alone often provide insufficient C. nasturtii control. To minimize C. nasturtii injury, crop rotation, early planting, and population monitoring with pheromone traps to optimize insecticide application timing are recommended (Hallett 2017). Despite these techniques, C. nasturtii remains difficult to manage (Hallett 2017).

1.2.1 Biological control of C. nasturtii

Biological control is the practice of using live organisms, such as insects or pathogens, to manage pest populations (CABI 2017). There are 3 types of biological control approaches: classical, augmentative, and conservation (van Lenteren 2012).

Classical biological control is the introduction of an exotic natural enemy for permanent establishment to a new geographical area where it is not present to control a pest (van

den Bosch 1971, Kenis et al. 2017). Approximately 2,700 natural enemies have been introduced to 196 countries (Heimpel et al. 2017); however, it is now recognized that the practice needs to be used with caution as it may pose threats to non-target species

(Louda et al. 2003, Heimpel et al. 2017, Hajek et al. 2018). While the number of classical biological control introductions has declined in the last few decades, in part due to the increasing legal requirements for importing, exporting, and releasing natural enemies, augmentative and conservation biological control strategies have been on the rise (Heimpel et al. 2017, Hajek et al. 2018).

The following qualities are typically sought in a good biological control agent: high effective attack rates, high egg complement, host-parasitoid emergence synchrony with the adequate life stages, host-specificity and low handling time (Beddington et al. 1978).

Host specificity must be considered particularly carefully when screening potential biological control agents since specific parasitoids are less likely to prey upon non- target species (Bigler et al. 2006).

According to the enemy release hypothesis, exotic invaders demonstrate increased success in their range of invasion due to the absence of co-evolved natural enemies (Elton 1958, Keane and Crawley 2002, Heger and Jeschke 2018). Invasive species, escaping from regulation by its specialist natural enemies (parasites, parasitoids, pathogens, and predators) from their area of origin, tend to perform better and undergo rapid increases in distribution and abundance (Keane and Crawley 2002,

Colautti et al. 2004, Heger and Jeschke 2018). However, the enemy release hypothesis has received mixed support when its assumptions are tested (Keane and Crawley 2002,

Roy et al. 2011, Jeschke et al. 2012). Nonetheless, the enemy release hypothesis still forms the theoretical basis for the practice of classical biological control, that is to retrieve an adapted natural enemy from the invasive pest’s native habitat and to release them in the invaded area, in the hope that they become established and maintain pest populations low (van den Bosch 1971).

Augmentative biological control is the release of mass-reared supplementary individuals of a native natural enemy, without the intention of establishing permanent populations of that natural enemy (Hoy 2008, Heimpel et al. 2017, Hajek et al. 2018).

Introductions can be inoculative (single introduction at one point in time) or inundative

(multiple introductions throughout a period of time) (Sivinski 2013). Augmentative biological control has been used successfully in greenhouses (Heimpel et al. 2017,

Hajek et al. 2018), and in fruit orchards, maize, cotton, sugarcane, soybean, and vineyards (van Lenteren 2012). Conservation biological control is the implementation of practices, such as establishing and maintaining non-crop habitat (forest, flowered field margins, meadows) (Tscharntke et al. 2008), avoiding pesticide use (Heimpel et al.

2017), intercropping, cover-cropping, or use of polycultures (Heimpel et al. 2017), to sustain and promote the reproduction and survival of established natural enemies

(McCravy 2008, Begg et al. 2017). Conservation and augmentative strategies are not mutually exclusive, since the success of augmentative biological control is greatly influenced by the ability of the landscape to support the natural enemies (Alvarez et al.

2019).

In the context of biological control of C. nasturtii, a few organisms have been studied as potential biological control candidates. For example, two polyphagous predatory Coccinellidae, Harmonia axyridis Pallas and Coccinella septempunctata L., were evaluated as potential natural enemies of C. nasturtii; however, these generalist predators were incapable of locating larvae on a broccoli plant (Corlay et al. 2007).

Three species of entomopathogenic nematodes (Steinernema feltiae Filipjev,

Steinernema carpocapsae Weiser, and Heterorhabditis bacteriophora Poinar) were also evaluated (Corlay et al. 2007). At a concentration of 1000 infective juveniles per C. nasturtii larva, H. bacteriophora induced 90 – 100% mortality of larvae in a diversity of non-compacted soils (Corlay et al. 2007). While all 3 nematode species can reproduce using C. nasturtii as a host, H. bacteriophora was found to be the most effective species in laboratory studies (Evans et al. 2015). Foliar sprays of nematodes on infested broccoli buds caused no significant mortality of larval C. nasturtii (Evans et al. 2015).

Field trials with nematodes and M. brunneum in soil showed mixed results: Contarinia nasturtii adult emergence was reduced in one field, but not in the other (Evans et al.

2015). Since none of the organisms studied to date provide adequate control of C. nasturtii, there is a need to investigate other biological control candidates.

The Platygastrinae is a highly diverse subfamily of tiny wasps known to parasitize phytophagous gall midges (Diptera: Cecidomyiidae), but their biology and taxonomy has not been thoroughly investigated (Abram et al. 2012b, Buhl and Jałoszyński 2016). The

Platygastrinae includes 40 genera and 1,829 species worldwide (Buhl 2016), 375 of which belong in the genus Synopeas (Hymenoptera Online 2017). Since some

Platygastridae parasitoids (e.g., Platygaster demades Walker) have been used successfully to reduce pest populations of Cecidomyiidae (Dasineura mali Kieffer)

(Zhao He and Wang 2011, 2015), there has been interest in investigating parasitoids of

C. nasturtii. In Europe, 4 species have been found to parasitize C. nasturtii: Synopeas myles Walker (Hymenoptera: Platygastridae), Macroglenes chalybeus Haliday

(Hymenoptera: Pteromalidae), Inostemma opacum Thompson (Hymenoptera:

Platygastridae), and Synopeas osaces Walker (Hymenoptera: Platygastridae) (Abram et al. 2012a).

1.3 SYNOPEAS MYLES

Synopeas myles is known to be a widespread parasitoid of C. nasturtii in Europe

(Buhl and Notton 2009, Abram et al. 2012a). After investigating the biology of S. myles,

Abram et al. (2012a) suggested that this parasitoid wasp would not be a good candidate for introduction to North America for classical biological control of C. nasturtii because it was not sufficiently host-specific and only parasitized on average 2.9 ± 0.4% of the surveyed C. nasturtii populations, rarely parasitizing as much as 28.7%.

Accidental biological control introductions have occurred quite frequently in recent years (Mason et al. 2017, Weber et al. 2017). Such events are typically discovered quite some time after a host has established in an area (Mason et al. 2017).

Unexpectedly, S. myles was found in Ontario, Canada in 2016 (Hallett and McGregor, unpublished data), rougly 20 years after detection of C. nasturtii (Hallett and Heal 2001).

These North American parasitoid populations are assumed to be adventive. Because S. myles is the first parasitoid of C. nasturtii that has been detected in Noth America, there is now interest in understanding the biological control potential of S. myles in the invasive range of C. nasturtii.

1.3.1 Synopeas myles biology

In Europe, S. myles occurs in Switzerland, Germany, France, Belgium, Slovenia,

England, Sweden, and Denmark (Abram et al. 2012a).

Female S. myles lay their eggs next to the midgut of C. nasturtii larvae.

Synopeas myles prefers to parasitize second and third instars of C. nasturtii larvae

(Abram et al. 2012b), possibly because younger larvae tend to hide within the inner regions of the growing tip of plants, making them less accessible (Abram et al. 2012b).

This pattern of host stage use is rare among the Platygastridae, which usually prefer to parasitize eggs or earlier instars (Abram et al. 2012b). No egg parasitism was observed by Abram et al. (2012b).

The eggs of S. myles become easily visible approximately 2 days after C. nasturtii larvae have been parasitized, while the embryo is barely visible within the trophamnion (the nutritive sheath within the egg) (Abram et al. 2012b). Three to four days after parasitization, the C. nasturtii larva jumps off the plant (as in normal development) and spins its cocoon. The parasitoid larvae emerge from the trophamnion as first instars and feed within the body cavity of the host (Abram et al. 2012b).

Synopeas myles larvae feed for 9 – 11 days, followed by a pupal stage of 10 – 12 days, so that approximately 21 days after parasitization, the S. myles imago emerges (Abram et al. 2012b).

Sugar-fed females live up to 2 weeks on average in the lab, whereas sugar-fed males live just over one week on average (Abram et al. 2012b). If females are only given water, they tend to live for about 2 days and males for about 3 days (Abram et al.

2012b). It is possible that S. myles searches not only for hosts, but also for sources of carbohydrates such as nectar (Abram et al. 2012b), which would be consistent with other parasitoids (Gillespie et al. 2016).

Egg-to-adult developmental times are longer as temperature decreases (Abram et al. 2012b). At 22.5 °C, S. myles was able to complete development within approximately 25 days, whereas at 13.5 °C, approximately 75 days was needed. No parasitoids emerged from samples maintained at temperatures under 9.5 °C or over 30

°C within 100 days (Abram et al. 2012b).

If males are available, female S. myles can mate almost immediately after emergence (Abram et al. 2012b). Females lay 67 eggs on average if provided with a continuous supply of hosts. Under serial oviposition scenarios (where females are provided daily with new larvae to parasitize and host availability is not a limiting factor), females lived for only 3 days (in comparison to a maximum of 12 days when females are provided with sugar), suggesting a trade-off between longevity and offspring production (Abram et al. 2012b). Younger females tend to be most fecund, laying

approximately 33 eggs on their first day in comparison to approximately 13 on their fourth day (Abram et al. 2012b).

Superparasitism, a form of parasitism in which multiple conspecifics lay eggs inside a common host, has been observed in S. myles in the field and laboratory

(Abram et al. 2012b). In the laboratory, the rate of superparasitism by S. myles was close to 6%, whereas in the field the rate was approximately 19% (Abram et al. 2012b).

Superparasitism tends to increase parasitoid development time and the proportion of emerging females, but does not impact emerging parasitoid size or emergence probability (Abram et al. 2012b). The maximum number of S. myles eggs found within a

C. nasturtii larva was eight (Abram et al. 2012b). However, in all cases only one S. myles larva per superparasitized host survived, suggesting that S. myles is a solitary endoparasitoid like most Platygastrinae (Austin et al. 2005, Abram et al. 2012b).

1.3.2 Synopeas myles as a potential biological control agent

To create and use a biological program against C. nasturtii involving S. myles, there is a need to determine the parasitoid’s distribution, abundance, and parasitism rates in Ontario, as well as environmental factors that influence its survival, reproduction and population growth. Such factors may include soil particle size, vegetation and floral resources in field edges, management of field surroundings; canola growing practices also should be assessed. Soil particle size should be considered because soil texture and moisture are known to influence insects that have a developmental life stage

occurring in the soil (Macdonald and Ellis 1990, Pacchioli and Hower 2004, Shililu et al.

2004, Bagyaraj et al. 2016). Different types of soil (loam fine sand, fine sand, clay loam, muck, shale loam, and silt loam soil) were examined to determine if the nature of the soil affected C. nasturtii pupation and survival (Chen and Shelton 2007). While the soil type did not affect adult emergence, extreme dryness or moisture reduced the number of emerging adults (Chen and Shelton 2007). Because S. myles larvae likely remain inside C. nasturtii larvae when the latter jumps off its host plant to pupate in the soil, S. myles is assumed to spend part of its life cycle in the soil. It is currently unknown if soil particle size affects S. myles as adults when they exit the pupal sheath of their host and attempt to reach for the soil surface, and whether the parasitoid may be more successful at navigating through certain soils more than others to reach the surface.

Field edge cover and management are important factors to examine because structurally complex landscape favours higher rates of parasitism and can help reduce crop damage compared to simple landscapes with intensive agricultural use (Thies and

Tscharntke 1999). For instance, rape pollen beetles Meligethes aeneus (Coleoptera:

Nitidulidae) in fields in complex landscapes (uncultivated land, perennial habitats) were subject to higher rates of parasitism by larval ichneumonid parasitoids (Thies and

Tscharntke 1999). Complex landscape neighbouring annual crops enhances immigration of natural enemy populations (Thies and Tscharntke 1999). Habitat management is increasingly being recognized as a tool to boost biological control

(Bianchi and Wäckers 2008). For instance, providing nectar resources to satisfy parasitoid requirements has potential to increase their effectiveness as biological control

agents (Bianchi and Wäckers 2008). It is currently unknown if and how landscape structure influences S. myles.

Dispersal of S. myles within the must be examined because it can reveal preferential habitat and suggest dispersion capacity. For instance, Euplectrus platyhypenae Howard (Hymenoptera: Eulophidae), an ectoparasitoid of the fall armyworm Spodoptera frugiperda Smith (Lepidoptera: Noctuidae), was more abundant next to field edges. (Hay-Roe et al. 2016). In contrast, the fall armyworm parasitoids

Cotesia sp. and Meteorus sp. have higher dispersal capacities and are distributed throughout fields (Hay-Roe et al. 2016). By remaining in the field edges, E. platyhypenae benefits from an environment that provides protective cover, and nectar and pollen from wildflowers (Hay-Roe et al. 2016). It is currently unknown whether S. myles disperses within a field or if it remains in field edges. A good biological control agent should be able to match its target pest in terms of dispersion to exert control on the undesirable insect population (Hay-Roe et al. 2016). Canola growing practices include seeding date, seed variety, seed treatment, phytosanitary treatments, and crop rotation. Such field information is critical to determine whether certain actions undertaken by growers have favoured or hindered the presence and abundance of S. myles, and examining these factors could lead to issuing recommendations to favour establishment of the parasitoid.

Female parasitoid fecundity should be tuned to match the number of hosts the parasitoid is expected to encounter in its lifetime (Rosenheim 1996). High parasitoid fecundity can suggest potential for population growth, which is among the key

characteristics of a good potential biological control agent (Meyer 2003). Nectar feeding can increase parasitoid fecundity and reproductive lifespan, and accelerate egg maturation rates (Baggen and Gurr 1998, Schmale et al. 2001, Winkler et al. 2006). In the context of biological control of C. nasturtii, finding plants able to support natural enemy populations would be a first step in the process of sustaining parasitoids to reducing crop damage due to C. nasturtii. The effect of floral supplements on the fecundity in S. myles has not yet been investigated, and could provide insights on future biological practices in an IPM program against C. nasturtii. Abram et al. (2012b) found that adult S. myles females are capable of laying eggs immediately after emergence. In indirect comparisons, their lifetime fecundity is equal to their egg quantity at emergence, which indicates that this species is likely proovigenic (emerges with fully developed eggs and does not develop any during its adult life) (Jervis et al. 2001). While proovigeny is the current suggested mode of oogenesis in S. myles, it is currently unknown if S. myles could also exhibits some traits of synovigeny and whether its fecundity can benefit from floral plants such as sweet alyssum (Lobularia maritima

Desvaux) to synthesize additional eggs. Although uncommon, there are known

Platygastridae exhibiting synovigeny (Jervis et al. 2001), and the degree of synovigeny could potentially be modified by floral supplementation.

Floral supplements can also increase parasitoid longevity (Lee and Heimpel

2008). Parasitoid longevity is important in the context of biological control because it allows more time to mate (Benelli et al. 2017), to parasitize hosts (Sandanayaka and

Charles 2006), and for synovigenic species to increase ovigenesis (Tena et al. 2015).

Parasitoids provided with sweet alyssum survived longer than those that fed on a sugar solution or on water (Aparicio et al. 2018), and attractiveness to alyssum has also been observed in the Platygastridae (e.g., Trissolcus basalis Wollaston, Foti et al. 2017). It is currently unknown whether sweet alyssum influences longevity of S. myles and if the plant could be used to benefit the longevity and fecundity this natural enemy of C. nasturtii.

1.3.3 Maximum parasitism rates

Hawkins (1994) reviewed 64 parasitoids introduced in classical biological control programs and ranked them by maximum parasitism rate. Biological control attempts involving parasitoids with a parasitism rate lower than 40% tended to be unsuccessful.

A biological control attempt was deemed successful when an introduction resulted in reducing pest densities below economic injury levels and thus yield economic benefits

(Hawkins 1993). Hawkin’s (1994) model pooled partial, substantial, and complete successes as successes. There was a threshold for maximum rates below 32%. No control was achieved with parasitism rates below 32% and that only partial control was achieved at rates up to 40%.

Maximum parasitism rates are associated with the rate of success of biological control agents (Hawkins 1994) but should not be regarded as a predictive tool. Percent parasitism is a measure of how much of a host population has been parasitized (Van

Driesche 1983). It is a parameter used in studies that aim to present data on parasitoid

presence and abundance in particular locations, times, or hosts (Van Driesche 1983).

Percent parasitism can help determine how common a species is, and measure the impact of parasitoids on their host population (Van Driesche 1983). However, parasitism rate has some limitations: realized parasitism rates, as calculated from the number of adult parasitoids and adult midges that survived until adulthood, are influenced by the environmental conditions that they are exposed to (e.g., controlled environment). This influence is not represented in the parasitism rate. When measured using adult emergence, parasitism rates do not account for the number of parasitoid eggs laid and the number of host larvae (Van Driesche 1983). Mortality for both immature parasitoids and hosts (parasitized or not) that remained in their growing medium and did not emerge as adults may not be accounted for (Van Driesche 1983). Also, nonreproductive effects of parasitoids on their hosts (e.g. altered behaviour, reproduction, or development) (Abram et al. 2019) are not represented in the parasitism rate (Van

Driesche 1983). Diapausing and aestivating insects may also not be accounted for.

Therefore, soil sampling of overwintering midges (Chavalle et al. 2018) can be pursued as a complementary method to the collection of fresh infested plant matter.

While the maximum parasitism rate of C. nasturtii by S. myles in Europe (28%) was below Hawkins’s (1994) threshold for impacting pest populations, preliminary surveys conducted in Quebec revealed maximum parasitism rates of approximately

76% (Sébastien Boquel, personal communication). Further investigation is therefore necessary to better understand the potential of S. myles in the biological control of C. nasturtii.

1.4 SUMMARY

Contarinia nasturtii is a serious pest of Brassica crops in Ontario (Hallett 2017).

The parasitoid S. myles is a natural enemy of C. nasturtii and could qualify as a potential biological control agent. However, it has only recently been found in North

America and further investigation is required to evaluate whether S. myles could be integrated into a biological control program for C. nasturtii.

The goal of this project was to begin to evaluate the potential of S. myles as a biological control candidate to use in an integrated pest management approach for C. nasturtii in Ontario, as well as to assess possible biological control practices to promote the integration of this parasitoid. The specific objectives of this project were to determine:

• the identity of C. nasturtii parasitoids in Ontario,

• the abundance and spatial distribution of S. myles at provincial and field scales,

• the phenology of S. myles in relation to that of C. nasturtii,

• the environmental parameters promoting the presence and abundance of S.

myles, and

• the influence of floral supplementation on the longevity and fecundity of S. myles.

2 DISTRIBUTION AND ABUNDANCE OF AN ADVENTIVELY ESTABLISHED

PARASITOID OF SWEDE MIDGE IN ONTARIO, CANADA

2.1 INTRODUCTION

Adventive biological control introductions have rapidly grown in frequency in recent years (Mason et al. 2017, Weber et al. 2017). Unintentionally introduced natural enemies are believed to have increased due to rising international trade, inability to detect and eliminate minute organisms such as parasitoids, and an absence of screening for species that are not associated with live vegetative or animal material, among others (Weber et al. 2017). Such introductions are typically discovered much later after a host has established in an area (Mason et al. 2017). In this paper, another case of an invasive pest and its adventively established natural enemy is presented.

The swede midge (Contarinia nasturtii Kieffer) (Diptera: Cecidomyiidae), discovered in Ontario in 2000 (Hallett and Heal 2001), is an invasive pest of Brassica crops in Canada (Hallett and Heal 2001), and belongs in the large family of gall midges.

Host crop plants of C. nasturtii include canola (Brassica napus Linnaeus, B. rapa L. and

B. juncea), Brussels sprouts (B. oleracea L. var. gemmifera), cabbage (B. oleracea L. var. capitata, tuba, sabauda, and acephala), and broccoli (B. oleracea L. var. botrytis).

When C. nasturtii larvae feed on their host through extraoral digestion, their saliva causes distortion of leaves, fusing and swelling of buds, and corky scarring (Barnes

1946). Larval feeding on the meristem may lead to the death of the meristem and

reduced yields (Hallett et al. 2007). Contarinia nasturtii injury has caused up to 81% yield loss in canola in Ontario (Hallett 2017).

Contarinia nasturtii is a challenging pest to manage because it is multivoltine, the larvae feed cryptically within deformed buds, and it has variable patterns of diapause

(Hallett 2017). Contarinia nasturtii has a wide host range, and it lacks effective natural enemies in North America (Hallett 2017). Current management practices rely on insecticides, and the insecticides currently registered for C. nasturtii on canola in

Canada are not very efficacious (Hallett 2017). An integrated pest management approach is required to better protect Brassica crops like canola from C. nasturtii and to reduce reliance on insecticides (Hallett 2017).

A number of taxa are known natural enemies of gall midges; however, none of them appear to specialize in targeting C. nasturtii and none are used as biological control agents (Keller and Wilding 1985, Gagné 1989, Tschamtke 1992, Gagné 1994,

Holland and Thomas 2000, Mendonca and Romanowski 2002, Sampson et al. 2002,

Goodfellow 2005). There are no current conservation or augmentative biological control agents recommended to manage C. nasturtii. In 2005, four years after the presence of

C. nasturtii was documented in Ontario by Hallett and Heal (2001), no parasitoids were found in Quebec, suggesting that C. nasturtii was introduced without its associated natural enemies from its area of origin (Corlay et al. 2007). Abram et al. (2013) assessed Synopeas myles (Walker) (Hymenoptera: Platygastridae), a parasitoid of C. nasturtii in its native range, as a candidate for introduction to North America for classical biological control of C. nasturtii. Synopeas myles was not found to be a good candidate

for introduction because it did not appear to be sufficiently host-specific and did not parasitize a significant portion of the sampled C. nasturtii populations in its native range

(3% on average) (Abram et al. 2013). However, S. myles was found in Quebec in 2015

(Hallett 2017) and in Ontario in 2016 (Hallett, unpublished data). It is not known how or when S. myles arrived in North America, and no information exists regarding the parasitoid’s life history, distribution, abundance and behaviour in the North American invasive range of C. nasturtii. Information on the distribution, abundance, parasitism rates, and development times of S. myles in Ontario is crucial for determining whether

S. myles could contribute to biological control of C. nasturtii.

Platygastridae parasitoids have previously been used to reduce populations of

Cecidomyiidae successfully; for example, Platygaster demades Walker appeared to contribute to controlling Dasineura mali Kieffer with up to 87% parasitism rate (Zhao He and Wang 2011). Some Platygastridae, while not always integrated in a biological program, can parasitize high proportions of gall midge populations. For example,

Synopeas sp., Platygaster sp., and Inostemma sp. parasitized 30 – 40% of blueberry gall midge, Dasineura oxycoccana Johnson, populations (Sampson et al. 2006).

Platygaster diplosisae Risbec combined with Aprostocetus procerae Risbec

(Hymenoptera: Eulophidae) parasitized up to 77% of African rice gall midge Orseolia oryzivora Harris & Gagné populations (Ogah et al. 2010). Because the Platygastridae are an under-investigated taxonomic group, ecological studies are often constrained to the family level (Maia et al. 2009). Thus, the ecology of many Platygastridae-

Cecidomyiidae complexes remains to be investigated.

Recent surveys in Quebec have found rates of C. nasturtii parasitism by S. myles as high as 76% (Sébastien Boquel, CEROM, personal communication, 2019). These preliminary findings justify further investigation of S. myles parasitism rates in North

America, given that parasitism rates in Europe reached 28% only on rare occasions

(Abram et al. 2013). Although Abram et al. (2012) did not deem S. myles a good candidate for biological control, now that it is established in North America, it is worth re- assessing its potential impact on its host in its adventive range. For example, climatic differences may lead to parasitoids being better synchronized with hosts in an area of introduction than in their area of origin (Rand et al. 2016). The extent to which S. myles parasitizes C. nasturtii in Ontario is currently unknown, as well as its distribution, phenology, and environmental factors affecting its presence. This information is required to assess the potential of S. myles to contribute to integrated pest management programs against C. nasturtii.

The objectives of this study were (i) to determine the distribution and abundance of S. myles in Ontario, (ii) to identify other potential parasitoids of C. nasturtii throughout canola-growing counties in Ontario, and (iii) to compare parasitism between years in the context of weather data. The overall goal of the project was to determine whether S. myles would be a biological control candidate worthy of incorporation into integrated pest management programs against C. nasturtii.

2.2 MATERIALS & METHODS

Central Ontario parasitoid collection

Sixteen canola fields located near Shelburne, ON were sampled throughout June and July in 2016, 2017, and 2018 from Dufferin, Grey and Wellington counties (Figure

2.1, Table 2.1). The fields were sampled weekly beginning at the emergence of first true leaves (canola stage 2.1, Harper and Berkenkamp 1975). In 20161, plant parts from the edges (first 3 m from the perimeter, where C. nasturtii injury is typically most prominent) of each field were collected weekly (Table 2.2) and included the rosette, primary bud inflorescence, secondary buds, or tertiary buds, depending on the growth stage of the plants at time of collection. Whole plants were collected in June and July from field edges in 2017 and throughout the field in 2018 (Table 2.2). In 2017 and 2018, half of the sites were sampled on Mondays, the other half on Thursdays.

In 2018, tissue was collected from plants in 10 randomly selected quadrats in each field. Randomization was performed by dividing an aerial photograph of the field into 100 quadrats and selecting 10 quadrats using a random number generator.

In the early season collecting events, before C. nasturtii injury was visible

(typically weeks 1 and 2), plants were collected haphazardly: collaborators were instructed to walk through the field edges (2016 and 2017) or in their respective quadrat

(2018), stop and point at the ground, and select the nearest plant.

1 In 2016, all sampling and monitoring of C. nasturtii and S. myles emergence was conducted by Spencer McGregor, University of Guelph, Guelph, ON. 24

Table 2.1. Canola fields sampled for Synopeas myles in 2016, 2017, and 2018. No. of No. plant sampling parts Field Year County GPS coordinatesa events collected DU1-16 2016 Dufferin 44.02, -80.17 8 791 DU2-16 2016 Dufferin 43.91, -80.16 8 768 DU3-16 2016 Dufferin 44.04, -80.21 8 774 DU4-16 2016 Dufferin 44.13, -80.29 8 795 GR1-17 2017 Grey 44.48, -80.57 8 2,897 DU5-17 2017 Dufferin 44.06, -80.19 8 2,852 DU6-17 2017 Dufferin 44.02, -80.17 7 2,751 DU7-17 2017 Dufferin 44.09, -80.23 7 2,026 DU8-17 2017 Dufferin 44.00, -80.16 7 2,018 WE1-17 2017 Wellington 43.64, -80.40 7 1,665 WE2-17 2017 Wellington 43.91, -80.82 1 120 WE3-17 2017 Wellington 43.82, -80.57 1 136 WE4-17 2017 Wellington 43.91, -80.56 1 81 DU9-17 2017 Dufferin 44.20, -80.25 1 95 BR1-17 2017 Bruce 44.94, -81.27 1 63 BR2-17 2017 Bruce 44.29, -81.30 1 117 BR3-17 2017 Bruce 44.55, -81.05 2 341 RE1-17 2017 Renfrew 45.00, -76.81 1 -b RE2-17 2017 Renfrew 45.54, -76.92 1 - RE3-17 2017 Renfrew 45.43, -76.40 1 - PE1-17 2017 Peterborough 44.29, -78.13 1 - PE2-17 2017 Peterborough 44.28, -78.16 1 - PE3-17 2017 Peterborough 44.28, -78.64 1 163 CO1-17 2017 Cochrane 48.52, -81.40 1 78 CO2-17 2017 Cochrane 48.51, -81.38 1 79 TE1-17 2017 Temiskaming 47.73, -79.76 2 428 TE2-17 2017 Temiskaming 47.71, -79.92 2 281 TE3-17 2017 Temiskaming 47.69, -79.88 2 238 TE4-17 2017 Temiskaming 47.58, -79.83 2 226 TE5-17 2017 Temiskaming 47.57, -79.85 1 141 TE6-17 2017 Temiskaming 47.62, -79.54 1 96 WN1-17 2017 W. Nipissing 46.36, -80.17 2 325 WN2-17 2017 W. Nipissing 46.39, -80.03 2 346 WN3-17 2017 W. Nipissing 46.39, -80.00 2 293 DU10-18 2018 Dufferin 44.00, -80.16 8 1,930 DU11-18 2018 Dufferin 44.13, -80.29 8 1,278 DU12-18 2018 Dufferin 44.09, -80.37 8 2,209 DU13-18 2018 Dufferin 44.09, -80.37 6 1,218 DU14-18 2018 Dufferin 44.01, -80.31 5 2,261 WE5-18 2018 Wellington 43.64, -80.40 4 1,113 26

WE6-18 2018 Wellington 43.82, -80.56 2 235 WE7-18 2018 Wellington 43.90, -80.61 2 257 WE8-18 2018 Wellington 43.88, -80.81 1 226 DH1-18 2018 Durham 44.28, -78.70 2 277 DH2-18 2018 Durham 44.31, -78.77 2 312 DU3-18 2018 Durham 44.07, -78.63 1 83 BR4-18 2018 Bruce 44.93, -81.32 1 332 BR5-18 2018 Bruce 44.30, -81.30 1 275 BR6-18 2018 Bruce 44.94, -81.31 1 147 RE4-18 2018 Renfrew 45.54, -76.92 2 177 RE5-18 2018 Renfrew 45.81, -77.08 2 185 RE6-18 2018 Renfrew 45.74, -76.83 2 190 CO3-18 2018 Cochrane 48.51, -81.41 2 395 CO4-18 2018 Cochrane 48.39, -81.39 2 242 CO5-18 2018 Cochrane 48.52, -81.39 2 318 TE7-18 2018 Temiskaming 47.71, -79.88 2 502 TE8-18 2018 Temiskaming 47.58, -79.80 2 508 TE9-18 2018 Temiskaming 47.58, -79.82 2 275 WN4-18 2018 W. Nipissing 46.39, -80.04 2 358 WN5-18 2018 W. Nipissing 46.37, -80.03 2 472 aGPS coordinates have been reported to 2 decimal places to conserve anonymity of participating growers. bDashes represent data that has been lost and could not be reported.

Table 2.2. Sampling regimes of the central Ontario area collections and focus locations by year. Sampling regime Year June July Location 2016 100 plant parts 75 plant parts Field edgesa 2017 100 full plants 75 full plants Field edgesa 2018 80 full plants 40 full plants Throughout field (quadrats) a First 3 m from the perimeter

Once C. nasturtii injury symptoms on plants became obvious, only plant tissue with evidence of infestation was collected. After the first 4 weeks of plant collection, the number of plants collected per field was reduced, due to their increased size and

multiple growth points suitable for infestation (Table 2.2). Sampling continued until plants were fully podded (late July) and there was no longer adequate tissue for C. nasturtii larval development. In 2018, collection during week 8 took place strictly inside the field edges because the interior of the field had reached maturity. Growers practiced conventional agriculture, and pest management was left to their discretion. The study was designed to collect only individuals developing within the season, not overwintered individuals.

Provincial parasitoid collection

The edges from 44 additional fields (Figure 2.1), located across 9 canola growing counties, were sampled once or twice throughout the season (Table 2.1), by, or under the supervision of, Meghan Moran, Canola and Edible Beans Specialist, Ontario

Ministry of Agriculture, Food and Rural Affairs. Plant samples were brought to University of Guelph for processing within 1 – 3 days of collection.

Monitoring insect emergence

Growing tips and bud clusters were trimmed from each plant and placed inside

500-mL clear plastic deli containers (Pro-Kal PK16SC, Shortreed Paper Inc, Guelph,

ON) containing a 1-cm layer of potting mix (BX Mycorrhizae, Pro-Mix, Rivière-du-loup,

QC) mixed with water. A 7-cm diameter hole was cut in each container lid, and fine mesh was glued over the hole for ventilation. In 2016, the tissue from multiple plants

was placed together in containers (13.4 ± 0.3 plant parts per container), but in 2017 and

2018, the tissue collected from each plant was placed in a separate container.

Tissue samples were kept in a growth chamber at 25 ± 1 °C on a 16:8 h

(light:dark) photoperiod. The substrate in each container was moistened with approx. 15 mL of water as needed. Plant tissue was removed when mouldy. Containers were checked for emergence of parasitoids or midges 3 times per week for 7 weeks. Adult parasitoids and midges were removed from the containers using a mouth aspirator and counted.

Insect identification

All midges were assumed to be C. nasturtii based on their association with canola: C. nasturtii is the only known species of gall midge to feed on the canola parts collected.

Emerged parasitoids were kept in 95% ethanol in glass vials at approximately 4

°C until identification. Parasitoids were initially identified to the superfamily or family level using Goulet and Huber (1993). Platygastridae parasitoid subsamples were identified by morphological methods (2016: n = 21, 2017: n = 52, 2018: n = 117) by Dr.

Peter N. Buhl (Natural History Museum of Denmark, Copenhagen) and by molecular methods (2016: n = 10, 2017: n = 45, 2018: n = 87) by Dr. Boyd Mori (Agriculture and

Agri-Food Canada Saskatoon Research and Development Centre, Saskatoon,

Saskatchewan). Subsamples sent for identification consisted of both males and females from the first and final dates of parasitoid emergence, as well as one date in the middle

of the emergence period. Non-Platygastridae adults of similar size to C. nasturtii and S. myles were sent to Dr. Owen Lonsdale (Canadian National Collection, Ottawa) for taxonomic identification.

For molecular identification, the 700 bp fragment of the cytochrome c oxidase subunit I (COI) gene was sequenced using the universal primer pair LCO1490-F (5’-

GGTCAACAAATCATAAAGATATTGG-3’) and HCO2198-R (5’-

TAAACTTCAGGGTGACCAAAAAATCA-3’) (Folmer et al. 1994). The purified amplicons were Sanger sequenced at the National Research Council of Canada on the University of Saskatchewan campus (Jennifer Holowachuk, AAFC Saskatoon, personal communication). The accession numbers for the GenBank reference nucleotide sequences are: BankIt2272019 Sm_Cn_CHa_1COI MN560786 to BankIt2272019

Sm_Di_CHd_3COI MN560812. Voucher S. myles specimens (n = 10) were deposited in the Canadian National Collection.

Parasitism rate

Percent parasitism of C. nasturtii per sampling event was calculated as the total number of S. myles divided by the sum of the total number of S. myles and C. nasturtii

(since each parasitoid makes use of a single C. nasturtii larva for its development) and multiplied by 100 (Abram et al. 2012a). Mean parasitism rates were calculated only for sampling events when S. myles was present.

C. nasturtii pheromone traps

Four C. nasturtii pheromone traps (Solida Distributions Inc., Saint-Ferréol-les-

Neiges, QC) were installed each year at each of the 16 central Ontario sites. One trap was installed in the centre of each field edge, approx. 40 cm above the ground. Trap liners were changed twice weekly, and all C. nasturtii males were counted. Sex pheromone dispensers were changed monthly.

Weather data

The closest Environment Canada weather station to Shelburne, ON, where weekly parasitoid collection occurred, is 14.6 km away in Mono, ON. Temperature and precipitation data from Mono during the period of emergence of adult C. nasturtii and S. myles over 2016, 2017, and 2018 (day 170 through 250) were retrieved from the

Environment Canada archives website (Environment Canada 2019). There were, however, missing temperature observations in the 2017 dataset; the temperature for certain days was unavailable and could not be retrieved (Environment Canada 2019). It is unknown what factor (e.g., faulty sensor) caused the missing data.

2.3 RESULTS

Parasitoid identification

All individuals in the 2016 subsample of presumed S. myles were confirmed as S. myles by morphological (n = 21) or molecular methods (n = 10). In the 2017 subsample,

43 individuals were identified as S. myles by morphological methods. The other 9 individuals were identified as a Platygaster sp. similar to P. tuberosula Kieffer (Dr. Peter

Buhl, personal communication). Forty-three individuals from the 2017 subsample were identified as S. myles by molecular methods. The other 2 individuals, different from one another, were unidentified Playgastrids, and neither of their haplotypes matched that of

P. tuberosula (Dr. Boyd Mori, Agriculture and Agri-Food Canada Saskatoon, personal communication 2017). All individuals of the 2018 subsample of presumed S. myles were identified as S. myles by morphological (n = 117) and molecular (n = 87) methods.

Identification of non-Synopeas parasitoids Of the non-Synopeas Hymenoptera, 4 male specimens were identified as Lyrcus sp. (Pteromalidae: Pteromalinae), possibly Lyrcus nigroaeneus Ashmead, a known parasitoid of gall-forming cecidomyiids, though C. nasturtii is not a previously recorded host (Dr. Gary Gibson, personal communication 2017).

All of the remaining specimens belonged to groups known to be parasitoids of other taxa and not including the Cecidomyiidae, including one female Pteromalus sp.

(Pteromalidae: Pteromalinae), one male Diaeretiella rapae McIntosh (Braconindae:

Aphidiinae), a cosmopolitan species that parasitizes aphids; one female Microplitis sp.

(Braconindae: Microgastrinae), probably near Microplitis plutellae Haliday, but species in this subfamilyare exclusively parasitoids of lepidopteran larvae; other Braconidae specimens in the subfamily Alysiinae, which are exclusive parasitoids of cyclorraphous

Diptera (suborder Brachycera), and therefore unlikely to be parasitizing C. nasturtii; and several specimens in the genus Alloxysta (Figitidae: Charipinae), which are 32

hyperparasitoids of braconids in aphids (Drs. Matthew Buffington, José Fernández-

Triana, and Gary Gibson, Canadian National Collection, personal communications,

2017 – 2019).

Distribution and parasitism rates of S. myles

Synopeas myles was found at 37 of the 60 sites sampled over the 3 years of the study (Table 2.3). Synopeas myles was detected as far south as Wellington County, as far north and west as Temiskaming District, and as far east as Renfrew County (Figure

2.2). A total of 7,396 S. myles and 71,117 C. nasturtii were collected. The average parasitism rate was 6.36 ± 1.02% (based on parasitism rates of all sites and positive S. myles sampling events) (Table 2.3). The highest parasitism rates (up to 31%) were generally found in Dufferin, Temiskaming, Cochrane, West Nipissing, and Renfrew

Counties (Table 2.3).

Abundance of S. myles varied among fields and years (Table 2.3). For instance, in 2016, 6,040 S. myles were collected, and approximately 84% of those individuals were collected from a single field. In 2017, a total of 260 S. myles were collected, approximately 58% from one field, and in 2018, 1,077 parasitoids were collected, with approximately 44% from one field.

In 2016, the mean weekly parasitism rate across all sites was highest on July 7

(collection week 5), followed by July 21 (week 7) and July 28 (week 8) (Table 2.4).

In 2017, the mean weekly parasitism rate across all sites was highest on July 24,

2017 (collection week 8) (Table 2.4), but only 2 fields were still being sampled at that

point in time. The second highest rate was observed from samples collected on July 4 and 7, 2017 (collection week 5) (Table 2.4).

In 2018, the mean weekly parasitism rate across all sites was highest in samples collected on July 30 and August 2, 2018 (week 8) with a weekly parasitism rate of

5.68%. However, this result might be an artifact of sampling only plants from the edge of the field for this last sampling date. The second highest parasitism rate was found in samples collected on July 9 and 12, 2018 (collection week 5) (Table 2.4).

Sampling events Total number Mean Max. Sampling with of adults emerged parasitism parasitism Field events S. myles S. myles C. nasturtii rate (% ± SE) rate (%) DU1-16 8 6 5,060 21,319 22.9 ± 4.3 30.8 DU2-16 8 4 143 3,172 9.2 ± 3.9 19.9 DU3-16 8 6 136 2,261 6.7 ± 2.1 10.4 DU4-16 8 5 701 9,699 15.0 ± 5.9 27.8 GR1-17 8 3 150 3,453 5.0 ± 0.9 6.7 DU5-17 8 0 - 761 - -a DU6-17 7 5 36 3,004 5.5 ± 4.2 22.2 DU7-17 7 2 2 790 1.0 ± 0.9 1.9 DU8-17 7 2 8 1,307 0.9 ± 0.2 1.1 WE1-17 7 4 18 2,116 1.9 ± 1.2 5.4 WE2-17 1 0 - 1 - - WE3-17 1 0 - 55 - - WE4-17 1 0 - 10 - - DU9-17 1 1 12 447 2.6 ± 0.0 2.6 BR1-17 1 0 - 6 - - BR2-17 1 0 - 31 - - BR3-17 2 1 1 310 0.4 ± 0.0 0.4 RE1-17 1 0 - 109 - - RE2-17 1 0 - 150 - - RE3-17 1 0 - 12 - -

PE1-17 1 0 - 33 - - PE2-17 1 0 - 0 - - PE3-17 1 1 1 122 0.8 ± 0.0 0.8 CO1-17 1 0 - 10 - - CO2-17 1 0 - 1 - - TE1-17 2 0 - 578 - - TE2-17 2 1 3 274 4.11 ± 0.9 4.11 TE3-17 2 2 8 442 2.0 ± 0.9 2.9 TE4-17 2 0 - 294 - - TE5-17 1 0 - 2 - - TE6-17 1 0 - 228 - - WN1-17 2 2 6 532 4.1 ± 3.4 7.5 WN2-17 2 2 13 490 3.3 ± 0.8 4.2 WN3-17 2 1 2 256 0.8 ± 0.0 0.8 DU10-18 8 3 20 563 4.6 ± 1.4 6.6 DU11-18 8 2 4 190 12.5 ± 7.5 20.0 DU12-18 8 4 41 2,050 2.48 ± 1.1 5.7 DU13-18 6 0 - 35 - - DU14-18 5 2 19 585 6.9 ± 6.2 13.2 WE5-18 4 1 1 263 0.4 ± 0.0 0.4 WE6-18 2 0 - 277 - - WE7-18 2 0 - 68 - - WE8-18 1 0 - 0 - - DH1-18 2 1 1 241 0.5 ± 1.0 0.5 DH2-18 2 1 2 973 1.6 ± 0.0 1.6 DU3-18 1 0 - 1 - - BR4-18 1 1 2 97 2.0 ± 0.0 2.0 BR5-18 1 0 - 4 - - BR6-18 1 1 6 73 7.6 ± 0.0 7.6 RE4-18 2 2 23 135 13.8 ± 12.4 26.2 RE5-18 2 1 6 99 6.1 ± 0.0 6.1 RE6-18 2 1 2 72 3.6 ± 0.0 3.6 CO3-18 2 2 99 997 8.3 ± 3.4 11.6 CO4-18 2 2 98 958 9.2 ± 1.2 10.4 CO5-18 2 2 71 773 7.8 ± 1.7 9.5 TE7-18 2 2 497 3,004 11.5 ± 9.4 20.9 TE8-18 2 2 75 1,974 3.9 ± 0.5 4.4 TE9-18 2 2 67 779 17.3 ± 12.7 30.0 WN4-18 2 2 16 1,949 3.4 ± 2.7 6.1 WN5-18 2 2 1 1 25.1 ± 24.0 50.0 a A dash (-) represents the absence of S. myles.

Table 2.4. Weekly parasitism rate (%) of Contarinia nasturtii by Synopeas myles at central Ontario canola fields sampled weekly in 2016 (A), 2017 (B), and 2018 (C).

A. 2016 Parasitism Rate (%) by Field Week Date DU1-16 DU2-16 DU3-16 DU4-16 All 1a June 9 0 0 0 0 0 2 June 16 0 0 0 0 0 3 June 23 5.4 0 0 0.8 4.2 4 June 30 14.6 7.2 2.4 0.5 8.9 5 July 7 28.9 19.9 10.0 27.8 27.9 6 July 14 30.8 0 0 0 9.9 7 July 21 27.6 8.2 10.4 21.8 20.6 8 July 28 30.6 1.6 3.9 23.9 17.5 All All 23.0 9.2 6.7 15.0 14.2

B. 2017 Parasitism Rate (%) by Field Week Date GR1-17 DU5-17 DU6-17 DU7-17 DU8-17 WE1-17 All 1 June 5 – 8 0 0 0 0 0 0 0 2 June 12 – 15 0 0 0 0 0 0 0 3 June 19 – 22 0 0 0 0.3 0 0 0.2 4 June 26 – 29 0 0 0 0.3 0.7 5.4 0.3 5 July 4 – 7 3.6 1.9 0 2.1 0 1.3 2.1 6 July 10 – 13 4.6 0.2 0 2.4 1.1 0.3 1.6 7 July 17 – 21 0 0 0 22.2 0 0.8 1.3 8 July 24 6.7 -b 0 - - - 6.4 All All 4.2 1.0 0 1.2 0.6 0.8 1.8

C. 2018 Parasitism Rate (%) by Field Week Date DU10-18 DU11-18 DU12-18 DU13-18 DU14-18 WE5-18 All 1 June 11 – 14 0 0 0 0 0 0 0 2 June 18 – 21 0 0 0 0 0 0.4 0.2 3 June 25 – 28 0 0 0 0 0 0 0.1 4 July 2 – 5 0 0 0 0 0 0 0 5 July 9 – 12 0 0 0 1.4 13.2 - 4.7 6 July 16 – 19 0.9 0 0 1.4 - - 0.8 7 July 23 – 26 6.7 4.9 - 1.4 - - 1.9 8 July 30 – Aug 2 5.2 20.0 - 5.7 - - 5.7 All All 3.4 2.1 0 2.0 3.2 0.4 2.3 a Week refers to the specimens that emerged from that specific week. Here, parasitism rate per site is a percent value derived from the sum of midges and parasitoids from all containers from that site and week. b A dash (-) indicates absence of data when canola plants were mature and beyond sampling.

Seasonal abundance, phenology and synchrony of S. myles and C. nasturtii 37

The mean number of days from the date of collection of C. nasturtii-infested plants until emergence was 13.8 ± 1.0 days for C. nasturtii and 21.1 ± 0.6 days for S. myles (Table 2.5).

In 2016, 4 peaks of adult S. myles emergence are distinguishable on July 20

(Day 202), July 28 (Day 210), August 10 (Day 223), and August 16 (Day 229) (Figure

2.3)2. In 2017, 4 peaks of adult S. myles emergence occurred on July 26 (Day 207),

July 28 (Day 209), August 2 (Day 214), and August 14 (Day 226) (Figure 2.4). In 2018,

6 peaks of adult S. myles emergence appeared on July 23 (Day 204), August 1 (Day

213), August 13 (Day 225), August 20 (Day 232), August 24 (Day 236), and August 29

(Day 241) (Figure 2.5).

Table 2. 5. Mean number of days ± S.E. from collection of canola plant samples to adult emergence of Contarinia nasturtii and Synopeas myles at central Ontario and provincial survey sites. Mean (± S.E.) no. of days from larval collection to adult emergence Contarinia Sites and year nasturtii n Synopeas myles n Shelburne 2016 12.9 ± 0.2 36,460 20.8 ± 0.2 6,052 Shelburne 2017 12.7 ± 1.7 11,431 18.9 ± 1.7 214 Provincial 2017 17.3 ± 2.3 4,395 30.4 ± 4.0 46 Shelburne 2018 18.9 ± 1.1 4,893 29.4 ± 1.6 88 Provincial 2018 14.5 ± 1.9 13,938 22.1 ± 2.1 993 Overall meana 13.8 ± 1.0 71,117 21.1 ± 0.6 7,393 a The overall mean was calculated by averaging all insect’s development times from larval collection to adult emergence, not averaging means of sites and years. A.

2 Weather data was obtained from the Environment Canada station in Mono. Emergence of adult insects was illustrated using data from 4 sites in 2016, 6 sites in 2017, and 6 sites in 2018. 38

Mean Temperature (°C) Total Precipitation (mm)

30 80

C) ° 60 20 40 10 20

Mean Temperature ( Temperature Mean 0 0 170 190 210 230 250 Precipitation(mm) Total Day of Year

B. C. nasturtii S. myles 7000 1200 6000 1000 5000 800

4000 myles S. C. nasturtii C. 600 3000 2000 400 200

1000 ofNumber Number ofNumber 0 0 170 180 190 200 210 220 230 240 250 Emergence Day (Day of Year)

A. Mean Temperature (°C) Total Precipitation (mm)

30 80

C) ° 60 20 40 10 20

Mean Temperature ( Temperature Mean 0 0 170 190 210 230 250 Precipitation(mm) Total Day of Year

B. C. nasturtii S. myles 2000 100

1500 75 S. myles S. C. nasturtii C. 1000 50

500 25 Number ofNumber Number ofNumber 0 0 170 180 190 200 210 220 230 240 250 Emergence Day (Day of Year)

A. Mean Temperature (°C) Total Precipitation (mm)

30 80

C) ° 60 20 40 10 20

Mean Temperature ( Temperature Mean 0 0 170 190 210 230 250 (mm) Precipitation Total Day of Year

B. C. nasturtii S. myles 1600 20

1200 15

nasturtii S. myles S. 800 10

400 5

Number ofNumber Number of C.of Number 0 0 170 180 190 200 210 220 230 240 250 Emergence Day (Day of Year)

In Shelburne, ON, the earliest S. myles emergence was recorded on July 11,

2016; July 4, 2017 and; July 16, 2018. Based on approximate S. myles development times (20.81 ± 0.24 days in 2016, 18.94 ± 1.72 days in 2017, and 29.38 ± 1.63 days in

2018) (Table 2.6), oviposition likely occurred on or before June 20, 2016, on or before

June 15, 2017, and on or before June 17, 2018. Therefore, it appears that the overwintering generation of S. myles typically emerges in mid-June in central Ontario

(Table 2.6). As a reference for future surveys, C. nasturtii emerged approximately 14 days after plant collection, while S. myles emerged 7 days later, approximately 21 days after plant collection.

Table 2.6. Estimated dates of emergence of overwintered S. myles adults from the central Ontario collection, based on parasitoid development times from larval collection to adult emergence. Development Emergence of 1st emergence of time from overwintering Year 1st generation collection (days) Min. Max. generation 2016 July 11 20.81 ± 0.24 6 53 June 20 2017 July 4 18.94 ± 1.72 2 37 June 15 2018 July 16 29.38 ± 1.63 13 44 June 17

From Day 170 through to Day 250, mean daily temperature was 20.60 ± 0.34 °C,

16.61 ± 0.44 °C, and 20.21 ± 0.34 °C, in 2016, 2017, and 2018, respectively. Total precipitation from Day 170 to Day 250 was 141.0 mm, 287.8 mm, and 309.9 mm, in

2016, 2017, and 2018, respectively. Due to currently insufficient information, it is too

early to make any conclusions about the relationships between temperature, precipitation, and parasitism rates.

2.4 DISCUSSION

This study determined S. myles to be widespread in Ontario, indicating that the parasitoid has spread and established throughout the province, and further confirms previous reports (Buhl and Notton 2009, Abram et al. 2012a) that C. nasturtii is a host for this Platygastridae parasitoid. This study also demonstrates another case in which a candidate biological control agent has established on a new continent without human intervention (Mason et al. 2017, Weber et al. 2017), and now appears to be widespread and well established, with parasitism rates being similar to its presumed native range.

Parasitism rates

Parasitism rates in this study were similar to those observed in Europe. The average parasitism rate by S. myles was 2.9 ± 0.5% in Europe (Abram et al., 2012a), compared to 6.4 ± 1.0% in Ontario. The maximum parasitism rates of S. myles recorded in Europe and Ontario are not meaningfully different from each other (28.7% and

30.0%, respectively), despite sampling methodologies that were very different. Abram et al. (2012a) sampled many different crop plants, not just canola.

The rates alone observed in both studies do not suggest that S. myles is a strong candidate as a biological control agent. However, S. myles should not yet be dismissed

as a potential biological control agent as some aspects of the host-parasitoid complex remain to be investigated. For instance, the maximum parasitism rate of approximately

76% reported in Témiscamingue, QC suggests that S. myles may induce higher levels of C. nasturtii mortality under some, yet to be determined, conditions.

Parasitism rates, as measured here, may not give a fully accurate picture of the actual impact of the parasitoids on the host population (Van Driesche 1983). This present study reported on realized parasitism rates, calculated from the number of adult wasps and adult midges that survived until adulthood under the conditions that they were subjected to. The number of parasitoid eggs laid and the number of C. nasturtii larvae were not measured. Mortality for both immature parasitoids and midges

(parasitized or not) that remained in the soil and did not emerge as adults also was not accounted for; mortality of parasitoids would have underestimated the impact of parasitism, while mortality of the midges would have overestimated the impact of parasitism. Also, non-lethal effects of parasitoids on their hosts (e.g., altered behaviour, reproduction, or development) (Abram et al. 2019) were not assessed, thus potentially underestimating the impact of the parasitoids on host populations.

Diapausing and aestivating insects were not accounted for in this project; however, high variability in parasitism has previously been noted in Cecidomyiidae-

Platygastridae complexes (Murchie 1996, Zhao He and Wang 2007). The Brassica pod midge, Dasineura brassicae Winnertz, and its parasitoid, Platygaster subuliformis

Kieffer (Hymenoptera: Platygastridae), may provide insights to the C. nasturtii-S. myles interaction. Pre-diapause, only 0 – 18% of collected D. brassicae larvae were

parasitized and gave rise to parasitoids, whereas post-diapause parasitism rates of 4 –

74% were observed (Murchie 1996). Similarly, increases in parasitism rates from 41 –

73% were observed between P. demades emerging in a given summer from D. mali larvae and overwintered P. demades emerging the next summer (Zhao He and Wang

2007). Abram et al. (2012a) found that 12.5% of sampled European S. myles entered diapause and the post-diapause parasitism rates were quite low (2 – 6%). Future North

American studies should conserve samples to allow parasitoids in diapausing larvae and aestivating parasitoids to develop and emerge as adults in the next year. Should high parasitism rates by S. myles be observed in diapausing C. nasturtii or aestivating

S. myles similar to that of P. subuliformis or P. demades, then S. myles would likely be considered as a much better candidate as a biological control agent.

High parasitism rates are not consistently required for a candidate parasitoid to be evaluated as a potential biological control agent; the parasitoid Collyria catoptron

Wahl (Hymenoptera: Ichneumonidae) exerts 1.9 – 38.3% parasitism on the wheat stem sawfly Cephus cinctus Norton (Hymenoptera: Cephidae) and was studied as a potential candidate because it possessed excellent synchrony with its host (Rand et al. 2016).

Synchrony is a crucial factor in good biological control agents (Rand et al. 2016). It is critical that a natural enemy encounters the adequate life stage of its host when it emerges (Rand et al. 2016). It has been found that weather conditions made American populations of Co. catoptron better synchronized with their hosts than in Asia, their continent of origin (Rand et al. 2016). The maximum rate of Co. catoptron was not very far from the maximum rate observed in S. myles, and this natural enemy was still

evaluated as a promising biological control agent (though it was ultimately excluded for other reasons). Thus, additional studies should investigate S. myles development and parasitism rates under different environmental conditions to determine whether a maximum parasitism rate above a theoretical threshold for success can be found.

Parasitism rates of zero were found at multiple sampling locations both in Europe and Ontario (Abram et al. 2012a, Table 2.3). Similar to Abram et al. (2012a) and Van

Driesche (1983), only certain collection dates yielded high numbers of parasitoids. In both Abram’s (2012a) study and this present project, parasitism could be low or nonexistent even when C. nasturtii infestations were severe. It is possible that certain collection events took place between the emergence of a generation of S. myles. It is also possible that sampling of life stages not yet fully susceptible to S. myles took place.

For example, if plants were sampled with mostly eggs and early instar larvae, which are less acceptable to S. myles (Abram et al. 2012b), parasitism rates would be lower compared to samples with mostly late-instar larvae, even if the absolute presence of S. myles was the same (Van Driesche 1983). The number of S. myles generations remains unknown. This information could be useful to understand population dynamics better and identify activity peaks, to avoid using an incompatible IPM tool during high S. myles activity. To increase detection rates and increase accuracy of parasitism rate data, each site should have a minimum of 2 sampling events and plants with symptoms of C. nasturtii injury should be inspected for presence of larvae.

Adult S. myles were found emerging from C. nasturtii between June 30 and

September 11. Although the maximum parasitism rate was 50% at a West Nipissing site

in 2018, only a single S. myles and a single C. nasturtii emerged from the 40 plants collected at this site (Table 2.3). A more accurate estimate of the maximum parasitism rate is likely provided by the 2016 Dufferin site, which had a 30.8% parasitism rate arising from a sample size of 5,060 S. myles and 21,319 C. nasturtii.

It should be acknowledged that sampling methods can affect measures of insect abundance and calculated parasitism rates. For instance, in 2018, the mean weekly parasitism rate across all sites was highest in samples collected on July 30 and August

2, 2018 (week 8) with a weekly parasitism rate of 5.68%. However, only plants from the edge of the field were collected on this last sampling date, as the field interior contained mature plants with no tissue infested with C. nasturtii. Sampling throughout the field in

2018 was conducted mainly to evaluate the spatial distribution of S. myles in an area of infestation of its host, rather than placing the focus on generating parasitism rates in areas of the fields under more severe C. nasturtii infestation. Consequently, the 2018 collection may have favoured collection of C. nasturtii over S. myles, and therefore have underestimated parasitism rates compared to the 2017 method. The latter specifically targeted field edges, which are typically more subject to injury by C. nasturtii. Because of the spatial variability of S. myles in canola fields, future surveys should organize their methods according to their research question; if the purpose of the survey is to obtain a large sample size of parasitoids, to maximize odds of detection, or to estimate parasitism rate in areas most often infested by C. nasturtii, then field edges might be the optimal sampling location.

Abram (2012a) noted that it would be hard to justify introducing S. myles as a biological control agent to North America unless investigations could show that (i) the parasitoid is an important factor in supressing populations of C. nasturtii, and that (ii) the parasitoid is unlikely to attack non-target cecidomyiids in its area of introduction. The parasitism rates from this study did not show that the parasitoid was an important factor in supressing C. nasturtii populations when data was averaged over an 8-week collection period (excluding weeks of parasitoid inactivity, i.e., when no parasitoids emerged from the samples). However, weekly parasitism rates can identify periods during which S. myles is most active and more likely to kill higher proportions of C. nasturtii populations. That S. myles can parasitize up to a third of a sampled population in certain weeks in the summer (Table 2.5) is encouraging for future augmentative and conservation biological control bioassays. Further weekly collection should be undertaken to follow variations in parasitism rates between years; additional sampling in a broader spatial and temporal perspective is critical to stabilize host-parasitoid interactions in theoretical models and empirical studies (Hassel et al. 1991, Stireman and Singer 2002). It is often the case that high levels of variability in parasitism rates are due to variable population densities and microhabitat effects (Stireman and Singer

2002). A multi-scale spatial and temporal sampling programme (Stireman and Singer

2002) needs to be pursued to understand better the ecological interactions between the parasitoid and its host.

Distribution

Synopeas myles was widespread across Ontario (Figure 2.2), from Shelburne

(southernmost positive site) to Timmins (northernmost positive site), and from the Bruce

Peninsula (westernmost positive site) to Arnprior (easternmost positive site). It is possible that this study underestimates the distribution of S. myles in Ontario since only canola fields and canola-growing regions were sampled. Contarinia nasturtii is an important pest of other Brassicaceae crops (Hallett 2007), and thus S. myles could be present in other crops further south than Wellington County. Synopeas myles is known to be present in Quebec, but studies of parasitoids of C. nasturtii in the United States have not yet been conducted.

Of the 23 sites that were negative for S. myles, 17 were sampled only once

(Table 2.2). Future studies should aim to sample fields a minimum of 2 times during the season, at different canola growth stages, to maximize the likelihood of encountering S. myles and thus improve confidence in parasitoid presence and distribution data.

Seasonal abundance

The number of insects collected each year greatly varied (Table 2.6), as did collection methods. Plant sampling was most limited in 2016, but yielded more insects than 2017 and 2018 sampling efforts. It should be noted that the 2016 samples were monitored daily during the week, in comparison to 3 times per week in 2017 and 2018.

These differences in monitoring frequency may have favoured collection of more parasitoids when containers were checked daily, especially if some managed to escape

the emergence containers. However, the minimum longevity of S. myles is typically 2 days (Abram et al. 2012b), therefore it is unlikely that many parasitoids were unaccounted due to mortality.

The known temporal distribution of adult S. myles in Ontario is from June 23 –

September 11 (Day 174 – 254). Synopeas myles could be present later in areas, or under conditions, where canola may have matured later in the season. However, sampling was not pursued beyond canola maturity (flowering of all meristems finished, seed pods filling, no more buds available).

Co-occurring parasitoids

In this study, several specimens of Hymenoptera across multiple families were collected from the canola samples infested with C. nasturtii larvae. Abram et al. (2012a) reported that 4 parasitoids were found parasitizing C. nasturtii in Europe: Macroglenes chalybeus Haliday (Hymenoptera: Pteromalidae), S. myles, Inostemma opacum

Thomson (Hymenoptera: Platygastridae), and Synopeas osaces Walker (Hymenoptera:

Platygastridae). Of the 4 European parasitoids, only S. myles was found in Ontario.

Two unidentified Platygaster specimens were found in 2017 and have not yet been identified to species level, due to the lack of current taxonomic expertise in

Nearctic Platygastridae. Further investigation should pursue species description and identification. However, the data from this present study suggest that Platygaster spp. abundance was negligible, and thus it is unlikely that the Platygaster spp. will play any important role in controlling C. nasturtii populations.

Of the non-Platygastridae specimens, none were identified as common parasitoids of C. nasturtii. Therefore, S. myles appears to be the only widespread parasitoid of C. nasturtii in Ontario at this time.

Potential for biological control

In 2012, S. myles was evaluated for introduction in North America as a potential classical biological control agent (Abram et al. 2012a). However, introduction of S. myles was not recommended because of its low parasitism rate in its native range and seemingly broad host range (Abram et al. 2013).

While the Ontario mean parasitism rates were slightly higher than those of

Europe (2.9 ± 0.4%, Abram et al. 2012a), it does not appear that S. myles is a significant factor in reducing C. nasturtii populations since it was found to parasitize only

6.36% of the sampled C. nasturtii on average, and parasitized as much as 30% in only

11% of weekly collections. Experiments have not yet been conducted to evaluate the efficacy of S. myles in conservation or augmentation biological control approaches, compared to the parasitism rates it exhibits in its natural occurrence.

Assessment of the success of biological control agents has shown that biological control attempts involving parasitoids with a maximum parasitism rate of ≤32% did not provide effective pest control (Hawkins 1994). A biological control attempt was considered a success when an introduction reduced pest densities below economic injury levels and consequently yielded economic benefits (Hawkins 1993). Hawkin’s

(1994) model deemed partial, substantial, and complete successes as successes. A

32% threshold for maximum parasitism rates was identified as below this value, no control was achieved.

Parasitism rates superior to Hawkin’s (1994) threshold for parasitoid success would be expected to reduce a significant portion of C. nasturtii populations, but high rates have only been observed rarely in Ontario. However, it should be acknowledged that Hawkin’s (1994) threshold was empirically derived for a large and general dataset; it therefore is not predictive of success for any one system in particular, but should rather be seen as a rough indicator for potential success. Investigation of the field parameters associated with variations of abundance could help to identify which environmental factors promote the presence of S. myles (Chapter 3), and help to identify conservation biological control tactics that may promote S. myles presence to levels that reduce C. nasturtii populations.

In summary, average parasitism rates of S. myles on C. nasturtii were very low.

Synopeas myles surveys must be pursued further to evaluate how parasitism rates fluctuate as this species further establishes. Future studies should investigate whether

(i) higher parasitism rates occur and under what conditions, as well as whether (ii) S. myles is likely to parasitize non-target species in the invasive range of C. nasturtii.

Augmentative biological control could potentially be successful if the maximum parasitism rate can be achieved more regularly. The success of augmentative biological control depends on the parasitoid population, quality, mass-rearing optimization, framework or technology to implement releases, timing, frequency, date of release,

weather, crop, host, predation, pesticides and dispersal (Smith 1996, van Lenteren

2000). Moreover, post-diapause parasitism rates of D. brassicae from 4 – 74% were observed, in contrast to 0 – 18% pre-diapause (Murchie 1996). Post-diapause parasitism rates of S. myles should be investigated to assess fully the role of this potential biological control agent in various scenarios.

The highly variable parasitism rates found in this study offer an opportunity to investigate conservation biological control initiatives. Such initiatives typically involve the implementation of practices to sustain and promote the reproduction and survival of established natural enemies (McCravy 2008, Begg et al. 2017), such as establishing and maintaining non-crop habitat (e.g., forest, flowered field margins, meadows)

(Tscharntke et al. 2008); avoiding pesticide use (Heimpel et al. 2017); intercropping, cover-cropping, or use of polycultures (Heimpel et al. 2017). If S. myles were to be promoted in the context of conservation biological control, environmental parameters enhancing the success of the parasitoid and increase parasitism rate should be investigated.

3 ENVIRONMENTAL FACTORS AFFECTING PARASITISM

OF CONTARINIA NASTURTII BY SYNOPEAS MYLES

3.1 INTRODUCTION

Habitat management has increasingly been recognized as a strategy to boost biological control of insect pests (Bianchi and Wäckers 2008, Karp et al. 2018, Miall et al. 2019). Providing supplemental nectar to parasitoids has the potential to increase their effectiveness as biological control agents (Bianchi and Wäckers 2008). However, there is conflicting evidence that non-crop habitat surrounding farms benefits yields and improves pest management (Karp et al. 2018). Habitat management, while intended to benefit a targeted insect, is also likely to impact other members of the arthropod community, including pest species (Baggen and Gurr 1998, Williams and Hendrix

2008). There is a need to identify cases in which habitat conservation truly represents a win – win opportunity to conserve biodiversity and protect yields (Karp et al. 2018). To address this need, studies must investigate how landscape effects are affected by farm management and how they affect the biology of crop pests and their natural enemies

(Karp et al. 2018).

There are many environmental parameters that influence habitat suitability for pest insects and their natural enemies. Soil abiotic factors, such as texture and moisture, are known to influence the survival and emergence of insects that have a developmental life stage inside the soil (Macdonald and Ellis 1990, Pacchioli and Hower

2004, Shililu et al. 2004, Bagyaraj et al. 2016). Cover and management of field edges

are important factors to examine because structurally complex landscapes favour higher rates of parasitism and can help reduce crop damage compared to simple landscapes with intensive agricultural use (Thies and Tscharntke 1999). Dispersal of natural enemies must be examined because it can reveal preferential habitat (Hay-Roe et al.

2016), and also potentially determine the efficacy of conservation biological control

(Heimpel 2019). Impacts of agronomic practices and non-target effects of pesticides on natural enemies should also be considered to determine whether certain actions undertaken by growers have favoured or hindered the presence and abundance of parasitoids, and examining these factors could lead to recommendations for growing practices that favour proliferation of the parasitoid (Weaver et al. 2004, Abbes et al.

2015).

The swede midge, Contarinia nasturtii Kieffer (Diptera: Cecidomyiidae), is a common invasive pest of Brassica crops in Ontario (Hallett and Heal 2001). Contarinia nasturtii can infest canola (B. napus L., B. juncea L., B. rapa L.), Brussels sprouts (B. oleracea L. var. gemmifera), cabbage (B. oleracea L. var. capitata, tuba, sabauda, and acephala), and broccoli (B. oleracea L. var. botrytis). Contarinia nasturtii larval feeding causes distortion of leaves, fusing and swelling of buds, and cork-like scars (Barnes

1946), which can lead to the death of the meristem and yield loss of host plants (Hallett et al. 2007).

Current control methods for C. nasturtii fail to provide adequate crop protection

(Hallett 2017). Management tactics must be improved because C. nasturtii can cause severe damage in its larval stage (Hallett 2017) and has the potential to expand its

distribution throughout most of North America (Olfert et al. 2006). Several taxa are known natural enemies of the Cecidomyiidae (Keller and Wilding 1985, Gagné 1989,

Tschamtke 1992, Gagné 1994, Holland and Thomas 2000, Mendonca and Romanowski

2002, Sampson et al. 2002, Goodfellow 2005); however, none of them appear to substantially mitigate C. nasturtii populations in North America.

There is currently no biological control program against C. nasturtii in North

America. Members of the Platygastridae (e.g., Platygaster demades Walker) have been used successfully to reduce populations of gall midges (e.g., Dasineura mali Kieffer)

(Zhao He and Wang 2011), and therefore, Synopeas myles Walker (Hymenoptera:

Platygastridae), an adventively introduced parasitoid of C. nasturtii, has been under investigation as a potential candidate for augmentative or conservation biological control of C. nasturtii in Ontario, Canada. This parasitoid is native to Europe (Abram et al.

2012a) and has only recently been found to be adventively established in Ontario and

Quebec, Canada (Hallett 2017). In laboratory conditions, sugar-fed females have lived on average 12 days and, unlike most Platygastridae, S. myles appears to prefer late- instar larvae (Abram et al. 2012a). Aside from C. nasturtii, S. myles is known to parasitize at least 9 other species of Cecidomyiidae across 2 different genera (Abram et al. 2012b, Chavalle et al. 2018). However, much remains to be known about S. myles, specifically regarding biotic and abiotic factors that may promote the presence and abundance of this natural enemy.

In pest-natural enemy systems such as with C. nasturtii and S. myles, habitat management holds potential to increase the effect of ecosystem services provided by

parasitoids (Fiedler et al. 2008). For instance, the choice in plant species with which to supplement a given crop is a key consideration in conservation biological control to attract beneficial insects and to improve efficacy of natural enemies to maximize pest suppression (Landis et al. 2000, Fiedler et al. 2008). Nectar feeding can extend the lifespan of parasitoids, increase their fecundity and reproductive lifespan, and accelerate egg maturation rates (Baggen and Gurr 1998, Schmale et al. 2001, Winkler et al. 2006). Providing additional nectar to parasitoids through habitat management could also increase parasitoid longevity, leading to increased parasitism rates and target host mortality (Miall et al. 2019). Increasing longevity grants more time to mate

(Benelli et al. 2017), more time to lay eggs and parasitize hosts (Sandanayaka and

Charles 2006), and more time for synovigenic species to increase ovigenesis (Tena et al. 2015). Egg production in parasitic wasps may be proovigenic or synovigenic, though there is a continuum between the 2 traits and species can exhibit both (Gordh and

Beardsley 1999, Jervis et al. 2001). Proovigenic species produce eggs prior to emergence, while synovigenic species generate eggs upon emergence (Gordh and

Beardsley 1999, Jervis et al. 2001). While proovigeny is the current suggested mode of oogenesis in S. myles, Abram et al. (2012a) used indirect comparisons to support this conclusion. Therefore, it is unknown whether S. myles also exhibits at least some traits of synovigeny and whether nectar feeding could indirectly benefit the realized fecundity of S. myles by enabling the parasitoid to live longer, or by preventing resources that could otherwise be allocated to egg production from being used for somatic maintenance.

Sweet alyssum, Lobularia maritima Desvaux (Brassicales: Brassicaceae), is among the most commonly used plants for attracting parasitoids (Legrand 2010). For example, parasitoids in the family Braconidae exposed to sweet alyssum survived longer than those that fed on a sugar solution or on water (Aparicio et al. 2018) and attractiveness to sweet alyssum has also been observed in the Platygastridae (e.g.,

Trissolcus basalis Wollaston) (Foti et al. 2017). Native to the Mediterranean area (Atlas of Living Australia 2019), sweet alyssum has been naturalized in Ontario (United States

Plants Germplasm System, 2019). It is a popular ornamental plant (Huang et al. 2015) that is highly attractive to multiple natural enemies and parasitoids (Rohrig et al. 2008,

Sivinski et al. 2011, Foti et al. 2017, Aparicio et al. 2018). Intercropping with sweet alyssum has been practiced (Brennan 2013, 2016), and, even though sweet alyssum is a member of the Brassicaceae family, intercropping with sweet alyssum in broccoli crops did not increase oviposition by C. nasturtii in one recent study (Brion 2015). It is currently unknown whether sweet alyssum may influence the longevity and fecundity of

S. myles or whether this plant could be used to augment the effectiveness of this natural enemy of C. nasturtii.

The goal of this study was to identify environmental factors that influence the occurrence and parasitism rates of S. myles to determine whether conservation biological control initiatives could help promote the presence of this parasitoid in agricultural habitats. To identify key parameters promoting the presence, abundance, longevity, and fecundity of S. myles to increase its biological control potential, the following environmental factors were examined: soil particle size, vegetative cover,

management of field surroundings, and S. myles dispersion. To examine the potential of sweet alyssum as a nectar-provisioning plant for S. myles, the effect of sugar feeding on longevity and fecundity was also examined.

3.2 MATERIALS & METHODS

Parasitoid collection

Sixty canola (Brassica napus L.) fields in Ontario, Canada were surveyed throughout June and July 20163, 2017, and 2018. Plants infested with C. nasturtii were collected and monitored for the emergence of C. nasturtii and parasitoids4. Meristems were cut from each plant and placed inside 500-mL clear plastic deli containers (Pro-Kal

PK16SC, Shortreed Paper Inc., Guelph, ON) containing a 1-cm layer of potting mix (BX

Mycorrhizae, Pro-Mix, Rivière-du-loup, QC) mixed with water. Each container lid included a circular 7-cm-diameter hole covered with fine mesh for ventilation. Samples were kept at 25 ± 1 °C on a 16:8 h (light:dark) photoperiod and moistened with approximately 15 mL of water as needed. Emergence of parasitoids and midges was checked 3 times per week for 7 weeks. Adult parasitoids and midges were retrieved using a mouth aspirator and counted.

3 In 2016, all sampling and monitoring of C. nasturtii and S. myles emergence was conducted by Spencer McGregor, University of Guelph, Guelph, ON. 4 Refer to Chapter 2, section 2.2 for additional details on how parasitoids were collected and how parasitism rates were calculated. 59

Soil characteristics

In 2017 and 2018, 3 cylindrical soil samples (radius 5.5 cm, depth 15 cm, volume

1425.5 cm3) were collected from a total of 12 canola fields post-harvest using a golf cup cutter (Par Aide Products Co., Minnesota, USA). The samples from each site were pooled and kept at 4 °C until analysis by the Agriculture and Food Laboratory, Laboratory

Services, University of Guelph for organic matter (Walkley and Black 1934) and soil texture by particle size: sand (2.0 to 0.05 mm in diameter), silt (0.05 to 0.002 mm) and clay <0.002 mm) (Skaggs et al. 2001).

Monitoring of C. nasturtii

Four C. nasturtii pheromone traps (Solida Distributions Inc., Saint-Ferréol-les-

Neiges, Quebec) were installed at each of the 16 fields that were visited weekly in 2016

(n=4), 2017 (n=6), and 2018 (n=6). Trapping began in the last week of April, before canola seeding and emergence of C. nasturtii, and ended in October when traps had zero midge captures for 2 consecutive weeks. Traps were installed approx. 40 cm from the ground equidistant to each other on each field edge. Trap liners were changed twice weekly, and the number of male C. nasturtii on each liner was counted. Pheromone dispensers were changed monthly.

Landscape surrounding canola fields

To determine whether parasitism by S. myles was affected by adjacent land use, the perimeter of each canola field was measured using Google MapsTM on satellite imagery. Using these measurements, similar to Elliott et al. (1999), the proportion of the 60

perimeter that was next to managed land (including other crops and residential properties with vegetative cover, and excluding forested area and non-cultivated meadows) was determined. The percent of the total perimeter land that was managed was then calculated for each field by dividing the managed perimeter length by the sum of both the managed and unmanaged perimeter length, and then multiplying by 100.

Percent vegetative cover of field edges

To determine whether floral resources around fields affected parasitism rates of

S. myles, total percent vegetative cover evaluations were conducted in 6 fields near

Shelburne, ON in 2018. Each of the 4 field edges were divided into 10 quadrats. Three quadrats per each field edge were selected at random weekly: a 1m x 1m PVC tube square was haphazardly dropped on the field edge of a randomly selected quadrat and photographed. A total of 12 pictures per field per week were taken. Percent bare soil, flower, and vegetation were estimated by overlaying a 5 x 5 grid (20 cm x 20 cm) to the photographs and the percent cover of each variable (bare soil, flower, and non-flowering vegetation) was estimated visually by assigning a categorical value between 0 and 4 percent to each of the 25 squares from the grid, and summing values from all squares to get the total percent. Due to time constraints, pictures were taken only in weeks 1

(June 11, 2018), 2 (June 18, 2018), 3 (June 25, 2018), and 6 (July 17, 2018). The mean percent value for each cover variable was calculated per field, averaged across all sampling dates.

Distance from field edge

To evaluate the dispersal capacity of S. myles into canola fields and differences in parasitism levels at different distances from field edges, C. nasturtii-infested plants were sampled from 10 randomly selected quadrats in 6 fields for 6 weeks (360 collection events) during the 2018 canola collection. The distance between the nearest field edge and each quadrat was measured using Google MapsTM, which indicated that samples were collected between 1.5 and 100 m into the field. Inside each quadrat, infested plants were sampled haphazardly. Samples were monitored for C. nasturtii and parasitoid emergence5. Total numbers of adult C. nasturtii and S. myles were then visualized with a frequency chart to show the distance from field edges where C. nasturtii infestation and parasitization are more likely to occur. The total number of sampling events at various distances (m) from the nearest field edge from which adult

C. nasturtii and S. myles emerged was calculated.

Collection of S. myles for experimental purposes

To source parasitoids for use in the longevity and fecundity experiments described below, 20 infested canola plants were sampled 4 times from 10 canola fields throughout the province of Quebec. Samples were monitored daily for the emergence of

S. myles. Emerged Platygastridae parasitoids were assumed to be S. myles because this species is currently the sole known parasitoid of C. nasturtii in Ontario and Quebec.

5 Refer to Chapter 2, section 2.2 for details on how samples were monitored for emergence of adult C. nasturtii and S. myles. 62

Voucher specimens of S. myles (n = 10) were deposited in the Canadian National

Collection, Ottawa.

Sweet alyssum as an experimental plant to provide nectar

Before use in experiments, sweet alyssum (cv. Carpet of Snow) plants (Les

Serres Dauphinais, Sainte-Julie, Quebec) were maintained in a greenhouse at 25 °C, and watered daily (130 mL). Plants were automatically fertilized in rotation: one week with 130 mL of 1 g/L Fafard All-purpose water-soluble fertilizer 20-20-20 (Teris, Laval,

Quebec) and the next with 130 mL of 1.5 g/L Teris T+ Fertilizer for flowering 15-30-15

(Teris, Laval, Quebec). Flower cuttings were trimmed periodically and provided to the parasitoids as a source of nectar.

Longevity of S. myles

To determine whether sugar resources from sweet alyssum influence the longevity of S. myles, 210 newly emerged (less than 24 h old) parasitoids (male:female ratio 1:1) were collected from emergence containers with a mouth aspirator. Parasitoids were divided evenly among three treatment groups. A single parasitoid was placed in each 500-mL clear plastic deli container (Pro-Kal PK16SC, Shortreed Paper Inc.,

Guelph, ON). A 7-cm-diameter hole was cut in each container lid, and fine mesh was glued over the hole. Each container received one of three treatments: a 10-cm cutting of sweet alyssum with flowers artificially removed (control), a 10-cm cutting of sweet alyssum with flowers intact, or a 3.81 x 0.95 cm cotton roll (#2, medium, Defend®, NY)

soaked with 10% sucrose solution and then placed in a dry 50 x 9 mm plastic Petri dish.

Cuttings were kept in a 5-mL Eppendorf vial with water and maintained upright with putty (Blue-tack, Bostik, USA). Assays were conducted at approximately 23 °C, 40% humidity, during a natural 15:9 h (light:dark) photoperiod. All containers included a 30- cm2 piece of filter paper (WhatmanTM #6, Maidstone, UK). Food and plant material were renewed every 3 days, and the filter paper in each container was moistened when dry.

Mortality was assessed every 12 h. Parasitoids that were not moving when probed were considered dead, and their time of death was assigned to the midpoint between the current and last evaluation. The experiment ran until all parasitoids were dead.

Potential fecundity of S. myles

To determine if S. myles synthesizes additional eggs after emergence, 81 females were distributed across a control (n = 27), sweet alyssum (n = 27), or a 10% sucrose treatment (n = 27), in 9 replicates. One female from each treatment and replicate was then removed at 0, 2, or 4 days after they were placed in the treatments, with the expectation that the number of ovarioles should increase if this species is synovigenic to some degree. The experimental set-up was the same as in the longevity experiment described above. Assays were conducted at approximately 23 °C, 40% humidity, on a 15:9 h (light:dark) natural photoperiod. Removed wasps were frozen at -

20 °C for at least 24 h before being dissected. Parasitoids were dissected using 2 pairs of very fine forceps (Dumont# 5, Fine Science Tools, Switzerland) to remove the abdomen from the body. The abdomen was placed at the edge of a small drop of water

on a microscope slide and ovaries were ejected into the drop of water by applying pressure on the abdomen. The ovaries were teased apart with the forceps to count the oocytes inside both mature and immature ovarioles.

Realized fecundity of S. myles

To determine whether the realized fecundity of adult S. myles is influenced by sugar provisioning, 39 males and 39 females were paired for 1 h in 5-mL Eppendorf vials, and mating was confirmed visually. The 39 mated females were placed in a control (n = 13), sweet alyssum (n = 13), or a 10% sucrose treatment (n = 13). A single female S. myles was held per treatment container. For 3 h every day, each female S. myles was placed in a closed Petri dish with a new group of approximately 60 (range 5

– 900) 4- to 8-day-old C. nasturtii larvae for oviposition, and then put back into its treatment container. The presumed parasitized C. nasturtii larvae were left undisturbed for 48 h to allow the S. myles eggs to absorb the hemolymph of their host and increase in size (Abram et al. 2012a). This oviposition sequence was repeated daily for 5 days.

All C. nasturtii larvae were retrieved from the Petri dish 48 h after S. myles oviposition and stored in 70% ethanol until the eggs were counted. The eggs that were laid inside each C. nasturtii larva by S. myles were counted visually using a Keyence VHX-5000 digital microscope (Keyence, Mississauga, Canada). Assays were conducted at approximately 23 °C.

Statistical analyses

All statistical analyses were conducted in R version 3.6.1 (R Core Team 2019) with α = 0.05.

All samples from each site were pooled to calculate a single percent parasitism value per site. The parasitism rates were analyzed using generalized linear models with binomial error distribution. When overdispersion was observed, the data were analyzed using a generalized linear mixed model with observation-level random effect (function glmer in lme4 package) (Bates et al. 2019). Prior to analyses, collinearity between explanatory variables was verified using pair plots and variation inflation factor (VIF), and collinear variables were not included in the models. For the 2017 – 2018 model, the variables managed surroundings, silt, and C. nasturtii traps (as a proxy for population size) were retained. For the 2018 model, the variables managed surroundings, silt, and flower mean were retained. Interaction effects among the variables were not tested because there were too many explanatory variables (6) for too few data points (11).

Numerical explanatory variables were standardized to a mean of zero and standard deviation of one prior to performing analyses. The fixed effects were tested with likelihood ratio tests (function Anova in the car package (Fox et al. 2019)). The least significant variable was dropped, and the model was refitted. This procedure was repeated until all the remaining variables were significant. Coefficients of determination

(pseudo-R2) were calculated using the method described in Nakagawa et al. (2017), and the marginal coefficient was reported. The model assumptions were verified visually

(residuals against fitted values, and a quantile-quantile plot of residuals).

To analyze the longevity data, the survival package (Therneau 2019) was used to run a Cox proportional hazards model test with survival time as the dependent variable; and treatment, sex, and their interaction as categorical explanatory factors.

The significance of each factor was tested using likelihood ratio tests using the car package (Fox et al. 2019). Multiple comparisons among treatment levels were performed using Tukey tests with the multcomp package (Hothorn et al. 2019). The proportional hazards assumption was verified using the survival package.

The potential fecundity data were analyzed using a generalized linear model with number of eggs as the dependent variable; and parasitoid age, treatment, replicate, and the interaction between parasitoid age and treatment as independent variables. Prior to analysis, exploratory data analysis was conducted. The effects were tested using likelihood ratio tests (function drop1) and the most non-significant variables were dropped until all the variables were significant. In case of a significant effect, the means were compared using the least square means method (function emmeans) with Holm-

Bonferroni correction for multiple hypotheses tests. The model assumptions

(homogeneity of variance and normality of the residuals) were verified visually using residual and quantile-quantile plots. To analyze the realized fecundity data, the total number of eggs per female per day produced were pooled by treatment, and a one-way

ANOVA was computed using the total number of eggs laid as the response variable and treatment as a 3-level categorical factor. The model assumptions were verified visually using residual and quantile-quantile plots.

3.3 RESULTS

Parasitoid collection

In total, 71,117 C. nasturtii and 7,393 S. myles were collected throughout Ontario between 2016 and 2018. The average parasitism rate per field (n = 37) was 6.36 ±

1.02%, and the maximum rate was 30%6.

Two-year model (2017 – 2018): environmental parameters (excluding vegetation cover)

After model refinement and exclusion of non-significant explanatory variables, the remaining explanatory variables with respect to parasitism rates were percent

2 2 managed surrounding (LRT; χ = 7.72, df = 1, P = 0.00547), percent soil silt (LRT; χ

= 12.50, df = 1, P = 0.000413), and abundance of C. nasturtii as represented by

2 2 pheromone trap captures (LRT; χ = 5.28, df = 1, P = 0.0215) (Figure 3.1) (Pseudo-R =

0.27). Parasitism rates increased with increases in the percent managed surroundings and decreased with percent silt in the field soil and number of C. nasturtii captured in pheromone traps.

2018 model: environmental parameters (including vegetation cover)

After model refinement and exclusion of non-significant explanatory variables, the remaining explanatory variables with respect to parasitism rates in 2018 were percent soil silt (LRT; X(df=1) = 31.50; P = 0.0000000198), and mean percent cover of

6 Full survey details are available in Chapter 2. 68

2 flowers in field edges (LRT; X(df=1) = 14.10; P = 0.000172) (Figure 3.2) (Pseudo-R =

0.40). Parasitism rates increased with increases in the mean percent cover of flowers in field edges and decreased with percent silt in the field soil.

Distance from field edge

Synopeas myles (n = 32) were found 1.5 m into the field at one sampling event and 10 m at 4 sampling events (Figure 3.3). Synopeas myles were not found in any samples collected beyond 10 m into the field (Figure 3.3). Contarinia nasturtii (n =

2,528) were present in samples collected from 1.5 m into the field up to 100 m, though most C. nasturtii were found within the first 10 m of the field (Figure 3.3).

Influence of sugar treatment on the longevity of S. myles

Provision of sugar had a significant impact on survival time (X2 = 15.19, df = 2, P

= 0.005), but sex did not (X2 = 0.18, df = 1, P = 0.6675); the interaction of treatment and sex was not significant (X2 = 1.32, df = 2, P = 0.5163). Both sweet alyssum (P =

0.00463) and sugar (P < 0.001) significantly increased S. myles longevity compared to the control treatment (Figure 3.4). The longevity of S. myles given sugar was not different compared to those given sweet alyssum (P = 0.78034) (Figure 3.4). Both females and males had a lower mean longevity in the control group (approximately 3 days), compared to those in the sweet alyssum and the 10% sucrose treatments

(approximately 4 – 5 days). Across all treatments, S. myles lived from 0.69 – 12.56 days.

25 20 15 10 S. myles 5 C. nasturtii emergence 0

events with insectwith events 1.5 3 4.5 5 6 10 10.5 15 20 25 50 75 100 Frequency of sampling sampling of Frequency Distance into field from edge (m)

Influence of sugar treatment on the potential fecundity of S. myles

Whether Synopeas myles age (χ2 = 5.80, df = 2, P = 0.055) had a positive effect on egg load was unclear, though treatment (χ2 =1.03, df=2, P = 0.599) had no significant effect (Figure 3.5). The mean number of eggs per female was 73.33 ± 5.32.

Influence of sugar treatment on the realized fecundity of S. myles

Treatment did not significantly affect the number of eggs laid per adult female S. myles (F = 0.36, dfN = 2, dfD = 76; P = 0.7) (Figure 3.6).

3.4 DISCUSSION

The purpose of this study was to identify environmental factors associated with parasitism rates of C. nasturtii by S. myles. Analyses of environmental factors found weak but significant associations between parasitism rates and managed surroundings, presence of flowers adjacent to field edge, soil percent silt, and C. nasturtii population size. Moreover, sweet alyssum showed the potential to improve the longevity of S. myles, but not its fecundity.

Both the management of field surroundings and the mean percent cover of flowers were significant explanatory variables that increased parasitism rates. This finding suggests that S. myles benefits from the presence of managed surroundings and flowers, and that biological control initiatives should pursue evaluation of these factors when assessing field parasitism rates to determine whether and how to manipulate these factors. Parasitism rate increased with the proportion of managed field surroundings, where managed habitat consisted of cultivated land, roads, and private properties with maintained lawns. Complex landscapes are inconsistently found to be beneficial to parasitism rates (Menalled et al. 1999, Thies and Tscharntke 1999, Karp et al. 2018). However, this present study did not examine the nature of the management of the field surroundings (i.e., another crop or a concrete road and turfgrass). As adjacent cultivated fields, although flowering at times, were simply recorded as managed surroundings, and because more often than not the surroundings were fields rather than urban maintained properties, it is likely that the categorizations used were not

sufficiently reflective of the relevant variables. In this study, most of the managed surroundings were actually crop fields that flowered at some point during the season.

Future studies could examine how S. myles presence and parasitism rates are affected by the presence of other crop types, and whether or not they represent nectar sources for this parasitoid.

No C. nasturtii larvae were parasitized by S. myles further than 10 m inside the fields, which suggests that the area of the field just inside of the perimeter is a preferential habitat for the parasitoid, or that the parasitoids may only disperse into the field until they find hosts and not need, or have the ability, to go any further. Field edges are known to benefit parasitoids by provisioning refuge and nectar, which in turn benefits parasitoid fitness (Tillman 2016). Some parasitoids exhibit preferences for habitats near wooded areas, such as the ectoparasitoid Euplectrus platyhypenae

Howard (Hay-Roe et al. 2016). In the field perimeter, E. platyhypenae benefits from a refuge including availability of nectar and pollen from wildflowers (Hay-Roe et al. 2016), which could also be the case with S. myles. Another reason as to why S. myles may be found close to field edges is that the leaf litter may act as an overwintering refuge, if they overwinter as adults similar to Telenomus podisi Ashmead (Hymenoptera:

Platygastridae) (Lahiri et al. 2017). However, it has not yet been verified whether or not

S. myles hibernates in the leaf litter around the field, or if the parasitoid remains inside the overwintering pupae of C. nasturtii directly in the field post-harvest.

The preferred habitat and dispersal capacity of S. myles can be investigated by sampling from the edge into the centre of the field. For instance, Euplectrus

platyhypenae Howard (Hymenoptera: Eulophidae), an ectoparasitoid of the fall armyworm Spodoptera frugiperda Smith (Lepidoptera: Noctuidae), was found in higher abundance within the vicinity of field edges, in contrast to Cotesia sp. and Meteorus sp., which are more abundant further into the field (Hay-Roe et al. 2016). Cotesia sp. and

Meteorus sp., being fast fliers, have high dispersal capacities and can disperse to avoid competition (Hay-Roe et al. 2016). By remaining in the field edges, E. platyhypenae benefits from a refuge rich in wildflowers, nectar, and pollen (Hay-Roe et al. 2016). It was previously unknown whether S. myles is found to disperse within a field or if it remains near field edges. An effective biological control agent should be able to disperse to areas with high densities of its target pest, and S. myles does seem to be found in those areas of the field where C. nasturtii infestations are most severe. This observation suggests that conservation biological control initiatives such as floral supplementation practiced in the adjacent field margins would be accessible by S. myles.

Synopeas myles parasitism rates may be affected by soil type since survival and emergence could directly be impacted by soil particles, and/or indirectly by effects of soil on abundance of its host C. nasturtii. Silt was a significant explanatory variable negatively influencing parasitism rates of C. nasturtii by S. myles. While soil type does not hinder emergence of C. nasturtii (Chen and Shelton 2007), soil texture (particle size) can influence arthropods that spend part of their life cycle in the soil (Pan et al. 2018).

For instance, the influence of soil texture on ground-dwelling arthropods was observed in the field with pitfall traps (Li et al. 2013); coarsely textured soil (sand) was associated

with little vegetation and low species diversity. Soils with a fine texture (clay) are usually more favorable for the growth and survival of soil arthropods because they have relatively high water and nutrient availability (Pan et al. 2015). These findings are congruent with the evidence presented in this present study; while silt was negatively associated with parasitism rates, clay was inversely proportional to silt and therefore was likely positively associated with parasitism rates while sand remained relatively constant. However, the soil variables were highly colinear, limited to a range of 0 – 100

% which means that the regression coefficients were not uniquely determined. In this case, colinearity of the soil texture variables limited the interpretability of the model because they influence each other.

Soil type (loam fine sand, fine sand, clay loam, muck, Chenango shale loam, and silt loam soil) has been found not to affect C. nasturtii adult emergence; however, extremes of soil moisture (very dry or very wet) reduced the number of emerging adults

(Chen and Shelton 2007). Because S. myles larvae remain inside C. nasturtii larvae when the latter leaves its host plant to pupate in the soil, S. myles spends part of its life cycle in the soil. This study found evidence suggesting that the soil silt content negatively affected S. myles parasitism rates. The parasitoid may be less successful at navigating through silty soils, though it remains to be investigated whether this relationship is due to reduced success of emergence when adults exit the pupal sheath of their host and attempt to move toward the soil surface. Assays with controlled concentrations of soil particles could be pursued to better understand the influence of soil type on survival of C. nasturtii and S. myles, along with parasitism rate.

In the present study, parasitism rates were higher in fields with relatively low populations of C. nasturtii, as suggested by the binomial models. However, there were few fields with high C. nasturtii numbers included in this study. Future studies should target a spectrum of fields likely to host a gradient from low to high densities of C. nasturtii to validate the presence of a density-dependent relationship. Assuming that trap catches are representative of the actual densities of C. nasturtii in the field, this observation could be evidence of negative density-dependent parasitism (Chavalle et al.

2018), suggesting that S. myles may not be an effective parasitoid at high densities of

C. nasturtii. Because of the high reproductive success of the Cecidomyiidae, populations of Platygastridae parasitoids may be unable to saturate their host midge population (Chavalle et al. 2018). Functional response studies (Holling 1959) investigate the rise in prey numbers consumed per predator as prey density increases. Such studies should be conducted in controlled conditions to understand the capacity of S. myles to limit C. nasturtii populations. Functional response studies would allow one to characterize the prey-predator interaction and to predict predation to some extent

(Holling 1959).

Because percent flower cover was found to be associated with higher parasitism rates, sweet alyssum was expected to potentially increase the fecundity and longevity of

S. myles in laboratory conditions. In this study, sweet alyssum and sugar were found to increase the longevity of S. myles (Figure 3.11), but did not benefit potential and realized fecundity. The potential fecundity experiment (Figure 3.5) did not demonstrate significant differences in egg load, however there was a non-significant trend for an

increase in egg load over time in the control and sweet alyssum treatments. This trend could be investigated further to determine whether or not S. myles is strictly proovigenic.

While the literature suggests proovigeny in this species (Jervis et al. 2001, Abram

2012a), there has been some evidence suggesting synovigeny in Platygastridae (e.g.,

Platygaster diplosiae Risbec) (Jervis et al. 2001, Nacro and Nénon 2008). Ovigeny should be thought of as a continuum rather than as a dichotomous scenario of strict proovigeny or synovigeny (Jervis et al. 2001, Ellers and Jervis 2004). The realized fecundity data suggest that the schedule of egg laying was the same for all treatments, and sugar supplementation increased the time frame throughout which S. myles could oviposit but did not increase the number of eggs laid. Because of the low numbers of eggs laid, it is possible that the parasitoids did not have enough time to lay eggs, or were too stressed to do so, and future studies could investigate optimized oviposition scenarios with various lengths of time during which the undisturbed parasitoid is allowed to lay eggs. Experiments ended at 4 days because, most parasitoids lived from 3 – 5 days under the experimental conditions. It is possible that the full picture of potential fecundity was not captured, since the experiment was concluded before all wasps died

(a minority of wasps lived beyond 6 days). An alternative experimental design for future studies would be to dissect the wasps upon their death instead of at fixed intervals.

Nevertheless, incorporation of sweet alyssum around canola field edges could be used as a conservation biological control practice to enhance longevity of S. myles.

Habitat management is a common practice to promote natural enemies to support crop production (Patt et al. 1997, Géneau et al. 2012). For instance, biological

control of pests by natural enemies can be increased with the addition of flowering plants to increase longevity and parasitism rates (Patt et al. 1997, Géneau et al. 2012).

Caution should be taken to ensure plant selection favours natural enemies but not their host, the targeted pest (Baggen et al. 1999, Landis et al. 2000, Géneau et al. 2012).

Next steps towards a biological control strategy to use S. myles against C. nasturtii would be to determine whether sweet alyssum, a brassicaceous plant, provides any benefit to C. nasturtii, and then to determine the impact of artificially-supplemented floral resources on both C. nasturtii and S. myles in field trials (Patt et al. 1997). Other flowering plants commonly used in conservation biological programs could also be evaluated for their effects on the fecundity of S. myles, including those from the family

Asteraceae (common yarrow, Achillea millefolium L.; cornflower, Centaurea cyanus L.; common sunflower, Helianthus annuus L.; camphorweed, Heterotheca subaxillaris

Lamarck; and Cutler’s alpine goldenrod, Solidago cutleri de Candolle), Apiaceae

(fennel, Foeniculum vulgare Miller; dill, Anethum graveolens L.; Queen Anne's lace,

Daucus carota L.; coriander Coriandrum sativum L.; and purplestem angelica, Angelica atropurpurea L.), or Paeoniaceae (common garden peony, Paeonia lactiflora Pallas)

(Patt et al. 1997, Picket and Bugg 1998, Bugg et al. 2008, Landis et al. 2000, Legrand

2009a, Legrand 2009b, Legrand 2010). The potential for other conservation biological control methods, including establishing and maintaining non-crop habitat (forest, flowered field margins, meadows) (Tscharntke et al. 2008); avoiding the use of pesticides (Heimpel et al. 2017); and intercropping, cover-cropping, or use of

polycultures (Heimpel et al. 2017), to sustain and promote the reproduction and survival of S. myles should also be investigated (McCravy 2008, Begg et al. 2017).

In conclusion, this study found weak but significant associations between parasitism rates and managed surroundings, presence of flowers adjacent to field edge, soil silt content, and C. nasturtii population size. Parasitism rates increased as the field surroundings were more managed, and as there were more flowers in the field edges.

Parasitism rates decreased with soil silt content and C. nasturtii population size.

Additionally, sweet alyssum showed potential to improve the longevity of S. myles.

Agronomic and pest management practices should be also be investigated to see how they impact S. myles. Canola growers provided their field history and seasonal management information at the end of each season (Appendix A). Preliminary observations suggested that the relationship between pesticide applications

(insecticides, herbicides and fungicides) and parasitism rates should be examined more closely. Field trials are required to investigate the potential of augmentative or conservation biological control of C. nasturtii by S. myles, and the economic feasibility of these approaches in commercial production within the context of an integrated pest management program. Crop yield data should also be examined to determine whether a scenario exists in which biological control, biodiversity, and crop yield can be optimized with habitat management initiatives (Karp et al. 2018).

4 GENERAL DISCUSSION

This study found that Synopeas myles was widespread throughout canola growing regions in Ontario. While S. myles did parasitize up to 1/3 of sampled C. nasturtii populations occasionally, the mean parasitism rate (approximately 6%) was similar to the one found in Europe (approximately 3%) (Abram et al. 2012a). The abundance of S. myles was highly variable across sites and years. Field edges were typically sampled, and when the interior of a field was sampled, S. myles was only detected within the first 10 m. Contarinia nasturtii adults emerged approximately 2 weeks from the date of collection of infested plants, while S. myles took approximately 3 weeks to emerge. Analyses of environmental factors found weak but significant associations between parasitism rates and managed surroundings, presence of flowers adjacent to field edge, soil percent silt, and C. nasturtii population size. Synopeas myles appeared to benefit from the presence of managed surroundings around canola fields and from the presence of flowers. Parasitism rates decreased with percent silt in the field soil and number of C. nasturtii captured in pheromone traps. Sweet alyssum was found to increase the longevity of S. myles; however, it did not have significant effects on its potential and realized fecundity. The potential fecundity experiment (Figure 3.5) cast doubts on whether or not S. myles is a strictly proovigenic species, but no significant differences in total egg load were observed among treatments. Nevertheless, these findings suggest potential for habitat management to promote S. myles through conservation biological control. The major research subjects that should be investigated

next, with respect to biological control of C. nasturtii by S. myles, are (1) field studies on habitat management, (2) the phenology of S. myles, and (3) the negative effects of an alien parasitoid with a non-specific host range on native biodiversity.

Field studies on habitat management

The dispersion of S. myles through canola fields has not yet been thoroughly investigated. Observations from this study suggest that the field edges were a preferred habitat for the parasitoids. This phenomenon may be due to the presence of flowering plants within the vicinity of the edges that provide them with a supply of nectar and pollen, and also to the typically higher levels of infestation of C. nasturtii in field edges.

Spatial patterns of S. myles parasitism rates on C. nasturtii more than 10 m inside the crop remain unknown. Because this study was performed in low S. myles density fields, it would be useful to validate the patterns of C. nasturtii infestation depending on the distance from the edge of the canola field where S. myles is more common, as well as to determine the ability of S. myles to disperse and match the dispersal ability of its host. Many studies report the merits of intercropping with flower strips inside the crop

(Haenke et al. 2009, Brion 2015, Tschumi et al. 2015, Westphal et al. 2015); the benefits of interspersing transects of flowering plants within and/or between canola fields as appropriate to promote the abundance of natural enemies could be investigated, with respect to their dispersion beyond 10 m inside the field. The parasitism rate and yield of quadrats from a field with flowering strips could then be compared to the yield of quadrats from a field without floral supplement. Moreover, it

has not yet been investigated whether canola nectar and pollen (or other field crops) can act as sources of carbohydrates for S. myles. This information would be relevant in the logistics of planning floral supplements to ensure the parasitoid has access to some source of food during its period of activity. To facilitate the bioassays, rearing methods for S. myles should be developed. Mass rearing would allow cage studies to be performed to evaluate the functional response of S. myles to different densities of C. nasturtii, with or without floral supplements, on yield and parasitism rate. Functional response studies (Holling 1959), investigating the rise in prey numbers consumed per predator, as prey density increases, should be pursued. Moreover, parasitism dynamics of C. nasturtii by S. myles at sites with different landscape characteristics should continue to be investigated over additional growing seasons, supplementary to the three years of data collected thus far. Longer-term studies more accurately reveal the influence of environmental factors on parasitoids than do short-term studies (Menalled et al. 2003). Furthermore, insecticide effects and compatibility in S. myles also need to be examined in the context of a conservation biological control program.

Phenology of S. myles

Patterns of S. myles diapause are currently unknown in North America, specifically the date at which the first generation of S. myles emerges. By collecting soil samples in the spring from fields that were sown to canola in the previous growing season (as well as from areas outside of the field to determine overwintering location), it should be possible to determine the date of emergence and parasitism rate of S. myles

from C. nasturtii that overwintered in diapause, similar to the methods of Chavalle et al.

(2018). Additionally, infested plant tissues, after having been through the insect emergence and monitoring step, could be kept over winter to compare parasitism rates of non-diapausing larvae to diapausing larvae, similar to Abram (2012a).

Negative effects of an alien parasitoid with a non-specific host range on native biodiversity

Synopeas myles would probably not have been intentionally introduced in North

America because of concerns over host specificity (Abram et al. 2012a). However, now that it has been found in Ontario, it was unclear whether its presence should be promoted through conservation biocontrol. Host specificity is important to consider when screening potential biological control agents to avoid impacts on non-target species

(Bigler et al. 2006, Abram et al. 2012a). There may be negative consequences to promoting the presence and abundance of a non-native parasitoid with a non-specific host range. For instance, the generalist carabid beetle Pterostichus melanarius Illiger not only feeds on pea aphids (Acyrthosiphon pisum Harris), but also on Aphidius ervi

Haliday (Snyder and Ives 2001, Snyder and Ives 2003), a commonly used biological control agent of aphids in greenhouses. Certain gall midges are biological control agents of insects (e.g., Aphidoletes aphidimyza Rondani against aphids (Marshall 2012,

Watanabe et al. 2016), and Feltialla acarisuga Vallot against several species of spider mites (Mo and Liu 2006)) or weeds (Julien and Griffiths 1998). Considering that the host range of S. myles is not yet well understood, it is currently unknown whether promoting

the presence and abundance of S. myles could disrupt native and/or beneficial species.

It remains unknown whether or not S. myles poses a threat to beneficial Cecidomyiidae, such as Aphidoletes aphidimyza, a specialist aphidophagous midge commonly used in greenhouse biological control programs (Markulla and Tiittanen 1985) and commonly present in soybean fields (Hallett et al. 2013), which S. myles most likely encounters.

Since much remains to be investigated within the Cecidomyiidae (Marshall

2012), evaluating negative effects of an alien parasitoid with a non-specific host range is a challenge, but one could begin with host specificity studies with beneficial midges or species closely related to the 9 known cecidomyiid hosts among which C. nasturtii is found (Abram et al. 2012a, Chavalle et al. 2018). While it may make sense to promote use of S. myles from a strictly economic and agricultural perspective, an ecological perspective would argue to proceed with caution. Adventively introduced natural enemies have resulted in unforeseen environmental and economic benefits (e.g., reduction in insecticide use) to managing invasive pests, but they may also pose potential threats to native biodiversity (Mason et al. 2017). Investigating the host range of such biological control agents and monitoring them when they have been detected in the invaded region is necessary to improve management of invasive alien species, and understanding of their risks and benefits (Mason et al. 2017).

For the aphid parasitoid Binodoxys communis Gahan (Hymenoptera:

Braconidae), phylogenetic proximity of potential hosts to known hosts was a good indicator of host suitability (Desneux et al. 2012, Zepeda-Paulo et al. 2013). Synopeas myles parasitizes at least 9 cecidomyiid species, all pests: Contarinia tritici Kirby, C. pisi

Winnertz, C. medicaginis Kieffer, Dasineura ignorata Wachtl, D. marginemtorquens

Bremi, D. crataegi Winnertz, D. mali Kieffer, and D. viciae Kieffer (Abram et al. 2012a,

Chavalle et al. 2018). If the phylogenetic proximity hypothesis holds true for S. myles, then other members of the genera Contarinia and Dasineura could be evaluated for host suitability. This would be an enormous labour-intensive and time-consuming task; the issue with using the phylogenetic proximity approach in this case is that both

Contarinia and Dasineura are considered to be cosmopolitan polyphyletic categories

(Gagné 1994), which means that taxonomic revision might result in further division of these genera into multiple groups. Further taxonomic work in these genera might help us to narrow down the potential host range of S. myles. In any case, the current known host range only encompasses herbivorous species deemed as pests. Given that S. myles has established in Ontario, and that it is currently the only known parasitoid of C. nasturtii, perhaps the best course of action would be to direct efforts towards improving the state of knowledge on its potential as a biological control agent. This strategy would avoid allocating resources to promoting a parasitoid whose biological control potential is still uncertain. Whether S. myles can control C. nasturtii to a threshold which makes an economic difference in terms of crop yield or insecticide costs must be investigated through field studies.

REFERENCES CITED

Abbes, K., A. Biondi, A. Kurtulus, M. Ricupero, A. Russo, G. Siscaro, B. Chermiti, and L. Zappalà. 2015. “Combined Non-Target Effects of Insecticide and High Temperature on the Parasitoid Bracon nigricans”. PLoS ONE 10(9): e0138411.

Abram, P. K., T. Haye, P. G. Mason, N. Cappuccino, G. Boivin, and U. Kuhlmann. 2012a. “Identity, Distribution, and Seasonal Phenology of Parasitoids of the Swede Midge, Contarinia nasturtii (Kieffer) (Diptera: Cecidomyiidae) in Europe.” Biological Control 62(3): 197–205.

Abram, P. K., T. Haye, P. G. Mason, N. Cappuccino, G. Boivin, and U. Kuhlmann. 2012b. “Biology of Synopeas myles, a Parasitoid of the Swede Midge, Contarinia nasturtii, in Europe.” Biological Control 57: 789–800.

Abram, P. K., G. Boivin, T. Haye, and P. G. Mason. 2013. “Ch. 18. Contarinia nasturtii Kieffer, Swede Midge (Diptera: Cecidomyiidae).” In Biological Control Programmes in Canada 2001 – 2012, edited by P.G. Mason and D.R. Gillespie, 134-138. Croydon (UK): CAB International.

Abram, P. K., J. Brodeur, A. Urbaneja, and A. Tena. 2019. “Nonreproductive Effects of Insect Parasitoids on Their Hosts.” Annual Review of Entomology 64(1): 259–276.

Alvarez, R. P., B. A. Nault, and K. Poveda. 2019. “Effectiveness of Augmentative Biological Control Depends on Landscape Context.” Nature Scientific Reports 9: 1–15.

Aparicio, Y., R. Gabarra, and J. Arnó. 2018. “Attraction of Aphidius ervi (Hymenoptera: Braconidae) and Aphidoletes aphidimyza (Diptera: Cecidomyiidae) to Sweet Alyssum and Assessment of Plant Resources Effects on their Fitness.” Journal of Economic Entomology 111(2): 533–541.

Atlas of living Australia. 2019. “Lobularia maritima.” Accessed November 11, 2019. https://bie.ala.org.au/species/http://id.biodiversity.org.au/node/apni/2911683

Austin, A. D., N. F. Johnson, and M. Dowton. 2005. “Systematics, evolution, and biology of scelionid and platygastrid wasps.” Annual Review of Entomology 50: 553–582.

Baggen, L. R., and G. M. Gurr. 1998. “The Influence of Food on Copidosoma koehleri (Hymenoptera: Encyrtidae), and the Use of Flowering Plants as a Habitat Management Tool to Enhance Biological Control of Potato Moth, Phthorimaea operculella (Lepidoptera: Gelechiidae).” Biological Control 11(1): 9–17.

Baggen, L. R., G. M. Gurr, and A. Meats. 1999. “Flowers in Tri-Trophic Systems: Mechanisms Allowing Selective Exploitation by Insect Natural Enemies for Conservation Biological Control.” Entomologia Experimentalis et Applicata 91: 155–161.

Bagyaraj, D. J., C. J. Nethravathi, and K. S. Nitin. 2016. “Soil Biodiversity and Arthropods: Role in Soil Fertility”. In Economic and Ecological Significance of Arthropods in Diversified Ecosystems, edited by A. Chakravarthy and S. Sridhara, 17– 51. Singapore: Springer.

Barnes, H.F. 1946. Gall Midges of Economic Importance, Vol. I: Gall Midges of Root and Vegetable Crops. London (UK): Crosby, Lockwood & Son.

Barnes, H. F. 1950. “The Identity of the Swede Midge, with Notes on its Biology.” Annals of Applied Biology 37(2): 241–248.

Barnes, H. F. 1952. “Studies of Fluctuations in Insect Populations XII. Further Evidence of Prolonged Larval Life in the Wheat‐Blssom Midges.” Annals of Applied Biology 39(3): 370–373.

Bates, D., M. Maechler, B. Bolker, and S. Walker. 2019. “Lme4 Package”. Accessed November 10, 2019. https://cran.r-project.org/web/packages/lme4/lme4.pdf

Beddington, J. R., C. A. Free, and J. H. Lawton. 1978. “Characteristics of Successful Natural Enemies in Models of Biological Control of Insect Pests.” Nature 273: 513–519.

Begg, G. H., S. M. Cook, R. Dye, M. Ferrante, P. Franck, C. Lavigne, G.L. Lövei, A. Mansion-Vaquie, J. K. Pell, S. Petit, N. Quesada, B. Ricci, S. D. Wratten, and A. N. E. Birch. 2017. A Functional Overview of Conservation Biological Control. Crop Protection 97:145–158.

Benelli, G., G. Giunti, A. Tena, N. Desneux, A. Caselli, and A. Canale. 2017. “The Impact of Adult Diet on Parasitoid Reproductive Performance”. Journal of Pest Science 90(3): 807-823.

Bianchi, F. J. J. A., and F. L. Wäckers. 2008. “Effects of Flower Attractiveness and Nectar Availability in Field Margins on Biological Control by Parasitoids”. Biological Control 46(3): 400-408.

Bigler, F., D. Babendreier, and U. Kuhlmann. 2006. “Environmental Impact of Invertebrates for Biological Control of Arthropods: Methods and Risk Assessment”. Wallingford: CAB International.

Brennan, E. B. 2013. “Agronomic Aspects of Strip Intercropping Lettuce with Alyssum for Biological Control of Aphids.” Biological Control 65(3): 302–311.

Brennan, E. B. 2016. “Agronomy of Strip Intercropping Broccoli with Alyssum for Biological Control of Aphids.” Biological Control 97: 109–119.

Brion, G. L. 2015. “Swede Midge, Contarinia nasturtii (Diptera: Cecidomyiidae), Response to Brassica oleracea in Simulated Intercropping Systems.” MSc Thesis, University of Vermont, Burlington, Vermont.

Bugg, R., R. Colfer, W. Chaney, H. Smith, and J. Cannon. 2008. “Flower Flies (Syrphidae) and Other Biological Control Agents for Aphids in Vegetable Crops.” University of California, Publication 8285. Accessed December 10, 2019. https://escholarship.org/uc/item/37k3d5g8#author

Buhl, P. N., and D. Notton. 2009. “A Revised Catalogue of the Platygastridae of the British Isles (Hymenoptera: Platygastroidea).” Journal of Natural History 43(25/28): 1651–1703.

Buhl, P. N., and P. Jałoszyński. 2016. “Two New European Species of Platygastrinae and Discovery of the Male of Synopeas burgeri Buhl, 2012 (Hymenoptera: Platygastridae).” Annales zoologici 66(3): 411–416.

Buhl, P. N. 2016. “Keys to Species of Ceratacis and Synopeas from New Zealand, with the Description of New Species (Hymenoptera: Platygastridae).” International Journal of Environmental Studies 73(6): 1016–1036.

CABI. 2017. “Biocontrol: What is it?” Accessed November 20, 2017. https://www.cabi.org/projects/controlling-invasive-species/biocontrol/

Canola Council of Canada 2017a. “Canola Now Worth $26.7 Billion to the Canadian Economy.” Accessed November 23, 2017. https://www.canolacouncil.org/news/canola- now-worth-$267-billion-to-canadian-economy/

Canola Council of Canada. 2017b. “Swede Midge (Contarinia nasturtii).” Accessed November 23, 2017. https://www.canolacouncil.org/canola-encyclopedia/insects/swede- midge/

Chavalle, S., P. N. Buhl, G. S. Martin y Gomez, and M. De Proft. 2018. “Parasitism Rates and Parasitoid Complexes of the Wheat Midges, Sitodiplosis mosellana, Contarinia tritici and Haplodiplosis marginata.” BioControl 63: 641–653.

Chen, M., J. Z. Zhao, and A. M. Shelton. 2007. “Control of Contarinia nasturtii Kieffer (Diptera: Cecidomyiidea) by Foliar Sprays of Acetamiprid on Cauliflower Transplants.” Crop Protection 26(10): 1574–1578.

Chen, M., A. M. Shelton, P. Wang, C. A. Hoepting, W. C. Kain, and D. C. Brainard. 2009. “Occurrence of the New Invasive Insect Contarinia nasturtii (Diptera: Cecidomyiidae) on Cruciferous Weeds.” Journal of Economic Entomology 102(1): 115– 120.

Chen, M., and A. M. Shelton. 2010. “Effect of Insect Density, Plant Age, and Residue Duration on Acetamiprid Efficacy Against Swede Midge.” Journal of Economic Entomology 103(6): 2107–2111.

Chen, M., A. M. Shelton, R. H. Hallett, C. A. Hoepting, J. R. Kikkert, and P. Wang. 2011. “Swede Midge (Diptera: Cecidomyiidae), Ten Years of Invasion of Crucifer Crops in North America.” Journal of Economic Entomology 104(3): 709–716.

Colautti, R. I., A. Ricciardi, I. A. Grigorovich, and H. J. MacIsaac. 2004. “Is Invasion Success Explained by the Enemy Release Hypothesis?” Ecology Letters 7(8): 721–733.

Corlay, F., G. Boivin, and G. Blair. 2007. “Efficiency of Natural Enemies Against the Swede Midge Contarinia nasturtii (Diptera: Cecidomyiidae), a New Invasive Species in North America.” Biological Control 43(2): 195–201.

Dara, S. K. 2017. “Biology and Management of Sucking Pests of Canola.” In Integrated Management of Insect Pests on Canola and Other Brassica Oilseed Crops, edited by G. V. P. Reddy, 68–76. Croydon (UK): CAB International.

Des Marteaux, L. 2012. “A Method for Induction and Quantification of Diapause Entry in the Swede Midge (Diptera: Cecidomyiidae)”. The Canadian Entomologist 144(6): 792- 800.

Desneux, N., R. Blahnik, C. J. Delebecque, and G. E. Heimpel. 2012. “Host Phylogeny and Specialisation in Parasitoids.” Ecology Letters 15(5): 453–460.

Ellers, J., and M. A. Jervis. 2004. “Why Are So Few Parasitoid Wasp Species Pro- Ovigenic?” Evolutionary Ecology Research 6: 993–1002.

Elliot, N. C., R. W. Kieckhefer, J. H. Lee, and B. W. French. 1999. “Influence of Within- Field and Landscape Factors on Aphid Predator Populations in Wheat.” Landscape Ecology 14: 239–252.

Elton, C. S. 1958. The Ecology of Invasions by Animals and Plants. London (UK): Methuen.

Environment Canada. 2019. “Résultats de station - Données historiques”. Accessed February 20, 2019.

http://climat.meteo.gc.ca/historical_data/search_historic_data_stations_f.html?searchTy pe=stnName&timeframe=1&txtStationName=Mono&searchMethod=contains&optLimit= yearRange&StartYear=2016&EndYear=2018&Year=2019&Month=6&Day=17&selRowP erPage=25

Evans, B. G., K. S. Jordan, M. Brownbridge, and R. H. Hallett. 2015. “Effect of Temperature and Host Life Stage on Efficacy of Soil Entomopathogens Against the Swede Midge (Diptera: Cecidomyiidae).” Biological and Microbial Control 8(2): 473– 483.

Fiedler, A. K., D. A. Landis, and S. D. Wratten. 2008. “Maximizing Ecosystem Services from Conservation Biological Control: The Role of Habitat Management.” Biological Control 45(2): 254–271.

Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. “DNA Primers for Amplification of Mitochondrial Cytochrome C Oxidase Subunit I From Diverse Metazoan Invertebrates.” Molecular Marine Biology and Biotechnology 3(5): 294–299.

Foti, M. C., M. Rostas, E. Peri, K. C. Park, T. Slimani, S. D. Wratten, and S. Colazza. 2017. “Chemical Ecology Meets Conservation Biological Control: Identifying Plant Volatiles as Predictors of Floral Resource Suitability for an Egg Parasitoid of Stink Bugs.” Journal of Pest Science 90(1): 299–310.

Fox, J., and S. Weisberg. 2019. “Car package”. Accessed November 12, 2019. https://cran.r-project.org/web/packages/car/car.pdf

Gagné, R. J. 1989. The Plant-Feeding Gall Midges of North America. Ithaca: Cornell University Press.

Gagné, R. J. 1994. The Gall Midges of the Neotropical Region. Ithaca: Cornell University Press.

Gagné, R. J., and M. Jaschhof. 2014. “A Catalog of the Cecidomyiidae (Diptera) of the World, 3rd Edition.” Digital version 2. Accessed December 22, 2019. https://www.ars.usda.gov/ARSUserFiles/80420580/Gagne_2014_World_Cecidomyiidae _Catalog_3rd_Edition.pdf

Gavloski, J. 2017. “Integrated Pest Management in Canola and Other Brassica Oilseed Crops: How Far We Have Come and What Is Still Needed.” In Integrated Management of Insect Pests on Canola and Other Brassica Oilseed Crops, edited by G. V. P. Reddy. 157–175. Croydon (UK): CAB International.

Géneau, C. E., F. L. Wäckers, H. Luka, C. Daniel, and O. Balmer. 2012. “Selective Flowers to Enhance Biological Control of Cabbage Pests by Parasitoids.” Basic and Applied Ecology 13(1): 85–93.

Gillespie, M. A. K., G. M. Gurr, and S. D. Wratten. 2016. “Beyond Nectar Provision: The Other Resource Requirements of Parasitoid Biological Control Agents.” Entomologia Experimentalis et Applicata 159: 207–221.

Goodfellow, S. 2005. Population Dynamics and Predictive Modeling of the Swede Midge, Contarinia nasturtii Kieffer. M.Sc. Thesis, University of Guelph, Guelph, Ontario.

Gordh, G., and J. W. Beardsley. 1999. “Taxonomy and Biological Control”. In Handbook of Biological Control: Principles and Applications of Biological Control, edited by T. S. Bellows and T.W. Fisher. 45–55. Cambridge (USA): Academic Press.

Goulet, H., and J. Huber. 1993. Hymenoptera of the World: An Identification Guide to Families. Ottawa: Agriculture Canada.

Haenke, S., B. Scheid, M. Schaefer, T. Tscharntke, and C. Thies. 2009. “Increasing Syrphid Fly Diversity and Density in Sown Flower Strips Within Simple vs. Complex Landscapes.” Journal of Applied Ecology 46(5): 1106–1114.

Hajek, A. E., and J. Eilenberg. 2018. Natural Enemies: An Introduction to Biological Control. Cambridge (UK): Cambridge University Press.

Hallett, R. H., and J. D. Heal. 2001. “First Nearctic Record of the Swede Midge (Diptera: Cecidomyiidae), a Pest of Cruciferous Crops from Europe.” The Canadian Entomologist 133(5): 713–715.

Hallett, R. H. 2007. “Host Plant Susceptibility to the Swede Midge (Diptera: Cecidomyiidae).” Journal of Economic Entomology 100(4): 1335–1343.

Hallett, R. H., M. Chen, M. K. Sears, and A. M. Shelton. 2009a. “Insecticide Management Strategies for Control of Swede Midge (Diptera: Cecidomyiidae) on Cole Crops.” Journal of Economic Entomology 102(6): 2241–2254.

Hallett, R. H., S. A. Goodfellow, R. M. Weiss, and O. Olfert. 2009b. “MidgEmerge, a New Predictive Tool, Indicates the Presence of Multiple Emergence Phenotypes of the Overwintered Generation of Swede Midge.” Entomolia Experimentalis et Applicata 130(1): 81–97.

Hallett, R. H. 2017. “The Challenge of Swede Midge Management in Canola.” In Integrated Management of Insect Pests on Canola and Other Brassica Oilseed Crops, edited by G. V. P. Reddy, 44–67. Croydon (UK): CAB International.

Hassell, M. P., R. M. May, S. W. Pacala, and P. L. Chesson. 1991. “The Persistence of Host-Parasitoid Associations in Patchy Environments: A General Criterion.” The American Naturalist 138(3): 568–583.

Harper, F. R., and B. Berkenkamp. 1975. “Revised Growth-Stage Key for Brassica campestris and B. napus.” Canadian Journal of Plant Science 55(2): 657–658.

Harris, K. M. 1966. “Gall Midge Genera of Economic Importance (Diptera: Cecidomyiidae) Part 1: Introduction and Subfamily Cecidomyiinae; Supertribe Cecidomyiidi.” Ecological Entomology 118(10): 313–358.

Hawkins, B. A. 1993. “Parasitoid Species Richness, Host Mortality, and Biological Control.” The American Naturalist 141(4): 634–641.

Hawkins, B. A. 1994. Pattern & Process in Host-Parasitoid Interactions. Cambridge (UK): Cambridge University Press.

Hawkins, B. A., and H. V. Cornell. 1994. “Maximum Parasitism Rates and Successful Biological Control.” Science 266(5192): 1886.

Hay-Roe, M. M., R. L. Meagher, R. N. Nagoshi, and Y. Newman. 2016. “Distributional Patterns of Fall Armyworm Parasitoids in a Corn Field and a Pasture Field in Florida.” Biological Control 96: 48–56.

Heger, T., and J. M. Jeschke. 2018. Invasion Biology: Hypotheses and Evidence. Preston: CAB International.

Heimpel, G. E., N. J. Mills, and J. Brodeur. 2017. Biological Control: Ecology and Applications. Cambridge (UK): Cambridge University Press.

Holland, J. M., and S. R. Thomas. 2000. “Do Polyphagous Predators Help Control Orange Wheat Blossom Midge, Sitodiplosis mosellana Géhin (Dipt., Cecidomyiidae) in Winter Wheat?” Journal of Applied Entomology 124(7-8): 325–329.

Rolling, C. S. 1959. “The Components of Predation as Revealed by a Study of Small Mammal Predation of the European Pine Sawfly.” The Canadian Entomologist 91(5): 293–332.

Hothorn, T., F. Bretz, P. Westfall, R. M. Heiberger, A. Schuetzenmeister, and S. Scheibe. 2019. “Multcomp package”. Accessed November 10, 2019. https://cran.r- project.org/web/packages/multcomp/multcomp.pdf

Hoy, M. A. 2008. “Augmentative Biological Control.” In Encyclopedia of Entomology. Springer. Accessed September 21, 2019. https://link-springer- com.subzero.lib.uoguelph.ca/content/pdf/10.1007/978-1-4020-6359-6_10394.pdf

Huang, R., D. Liu, M. Zhao, Z. Li, M. Li, and S. Sui. 2015. “Artificially Induced Polyploidization in Lobularia maritima (L.) Desv. and Its Effect on Morphological Traits.” HortScience 50(5): 636–639.

Jervis, M. A., G. E. Heimpel, P. N. Ferns, J. A. Harvey, and N. A. C. Kidd. 2001. “Life- History Strategies in Parasitoid Wasps: A Comparative Analysis of ‘Ovigeny’.” Journal of Animal Ecology 70(3): 442–458.

Julien, M. H., and M. W. Griffiths. 1998. Biological Control of Weeds. A World Catalogue of Agents and their Target Weeds, 4th Edition. Wallingford: CAB International.

Karp, D. S., R. Chaplin-Kramer, T. D. Meehan, E. A. Martin, F. DeClerck, H. Grab, C. Gratton et al. 2018. Crop Pests and Predators Exhibit Inconsistent Responses to Surrounding Landscape Composition. Proceedings of the National Academy of Sciences of the United States of America 115(33): E7863–E7870. DOI: 10.1073/pnas.1800042115

Keane, R. M., and M. J. Crawley. 2002. “Exotic Plant Invasions and the Enemy Release Hypothesis.” Trends in Ecology & Evolution 17(4): 164–170.

Keller, S., and N. Wilding. 1985. “Entomophthora brevinucleata sp. nov. [Zygomycetes, Entomophthoraceae], a Pathogen of Gall Midges [Dip.: Cecidomyiidae].” Biological Control 30(1): 55-63.

Kenis, M., B. P. Hurley, A. E. Hajek, and J. W. Cock. 2017. “Classical Biological Control of Insect Pests of Trees: Facts and Figures.” Biological Invasions 19(11): 3401–3417.

Kieffer, J. J. 1888. “Beitraege zur Kenntniss der Gallmuecken”. Entomologische Nachrichten 14: 200–314.

Kikkert, J. R., C. A. Hoepting, Q.-J. Wu, P. Wang, R. Baur, and A. M. Shelton. 2006. “Detection of Contarinia nasturtii (Diptera: Cecidomyiidae) in New York, a New Pest of Cruciferous Plants in the United States.” Journal of Economic Entomology 99(4): 1310– 1315.

Landis, D. A., S. D. Wratten, and G. M. Gurr. 2000. “Habitat Management to Conserve Natural Enemies of Arthropod Pests in Agriculture.” Annual Review of Entomology 45:175–201.

Lahiri, S., D. Orr, C. Sorenson, and Y. J. Cardoza. 2017. “Behavior of Telenomus podisi (Hymenoptera: Platygastridae) Adults Under Overwintering Conditions.” Journal of Entomological Science 52(1): 15–28.

Lee, J. C., and G. E. Heimpel. 2008. “Floral Resources Impact Longevity and Oviposition Rate of a Parasitoid in the Field.” Journal of Animal Ecology 77(3): 565–572.

Legrand, A. 2009a. “Evaluation of Ornamental Plants as Nectar Sources for Summer Tiphia.” In 2009 Annual Turfgrass Research Report, College of Agriculture and Natural Resources, edited by K. Guillard, 77–79. Storrs: University of Connecticut.

Legrand, A. 2009b. “Evaluation of Landscape Ornamental Plants as Nectar Plants for Tiphia vernalisand as Host plants for Pest Scarab Beetles.” In 2009 Annual Turfgrass Research Report, College of Agriculture and Natural Resources, edited by K. Guillard, 73–76. Storrs: University of Connecticut.

Legrand, A. 2010. “Evaluation of Landscape Ornamental and Herb Plants as Nectar Sources for Tiphia Parasitoid Wasps.” In 2010 Annual Turfgrass Research Report, College of Agriculture and Natural Resources, edited by K. Guillard, 73–76. Storrs: University of Connecticut.

Li, F. R., J. L. Liu, C. A. Liu, Q. J. Liu, and R. X. Niu. 2013. “Shrubs and Species Identity Effects on the Distribution and Diversity of Ground-Dwelling Arthropods in a Gobi Desert.” Journal of Insect Conservation 17(2): 319–331.

Liu, J., B. Mori, O. Olfert, and R. H. Hallett. 2018. “Determining Temperature-Dependent Development and Mortality Parameters of the Swede Midge (Diptera: Cecidomyiidae).” Journal of Economic Entomology 112(4): 1665–1675.

Louda, S. M., R. W. Pemberton, M. T. Johnson, and P. A. Follett. 2003. “Nontarget Effects—the Achilles' Heel of Biological Control? Retrospective Analyses to Reduce Risk Associated with Biocontrol Introductions.” Annual Review of Entomology, 48: 365– 396.

Luckmann, W. H., and R. L. Metcalf. 1994. Introduction to Insect Pest Management. Hoboken, New Jersey: Wiley.

Macdonald, P. J., and C. R. Ellis. 1990. “Survival Time of Unfed First-Instar Western Corn Rootworm (Coleoptera: Chrysomelidae) and the Effects of Soil Type, Moisture, and Compaction on their Mobility in Soil.” Environmental Entomology 19(3): 666–671.

Mahmood, T., E. Hasan, M. Ali, and M. Hussain. 2012. “Faisal Canola: A New High Yielding Canola Variety for General Cultivation in Punjab.” Journal of Agricultural Research 50(3): 321–328.

Markulla, M., and K. Tiittanen. 1985. “Biology of the Midge Aphidoletes and Its Potential for Biological Control”. In Biological Pest Control: The Glasshouse Experience, edited by N. W. Hussey and N. Scopes, 74–81. Ithaca: Cornell University Press.

Marshall, S.A. 2012. Flies: The Natural History and Diversity of Diptera. Richmond Hill: Firefly Books.

Mason, P. G., O. O. Olfert, T. Haye, T. D. Gariepy, P. K. Abram, and D. R. Gillespie. 2017. “Risks and Benefits of Accidental Introductions of Biological Control Agents in Canada.” In Proceedings of the 5th International Symposium on Biological Control of Arthropods, edited by P. G. Mason, D. R. Gillespie and C. Vincent, 6–7. Croydon (UK): CABI International.

McCravy K. W. 2008. “Conservation Biological Control”. In Encyclopedia of Entomology, edited by J. L. Capinera, 56. Dordrecht: Springer.

Menalled, F. D., P. C. Marino, S. Gage, and D. A. Landis. 1999. “Does Agricultural Landscape Structure Affect Parasitism and Parasitoid Diversity?” Ecological Applications 9(2): 634–641.

Menalled, F. D., A. C. Costamagna, P. C. Marino, and D. A. Landis. 2003. Temporal Variation in the Response of Parasitoids to Agricultural Landscape Structure. Agriculture, Ecosystems and Environment 96: 29–35.

Mendonca, M. D. S., and H. P. Romanowski. 2002. “Natural Enemies of the Gall-Maker Eugeniamyia dispar (Diptera: Cecidomyiidae): Predatory Ants and Parasitoids.” Brazilian Journal of Biology 62(2): 269-275.

Meyer, J. R. 2003. “Chapter Nineteen: Pest Control Tactics (Biological Control).” Accessed June 5, 2019. https://projects.ncsu.edu/cals/course/ent425/text19/biocontrol.html

Miall, J. H., P. K. Abram, N. Cappuccino, and P. G. Mason. 2019. “Effects of Floral Resources on the Efficacy of a Primary Parasitoid and a Facultative Hyperparasitoid.” Journal of Applied Entomology 143(7): 776–786.

Mika, A. M., R. M. Weiss, O. Olfert, R. H. Hallett, and J.A. Newman. 2008. “Will Climate Change be Beneficial or Detrimental to the Invasive Swede Midge in North America? Contrasting Predictions Using Climate Projections from Different General Circulation Models.” Global Change Biology 14(8): 1721–1733.

Mo, T. L., and T. X. Liu. 2006. “Biology, Life Table and Predation of Feltiella Acarisuga (Diptera: Cecidomyiidae) Feeding on Tetranychus cinnabarinus eggs (Acari: Tetranychidae).” Biological Control 39(3): 418–426.

Mori, B. A., L. Andreassen, J. D. Heal, J. R. Dupuis, J. J. Soroka, and B. J. Sinclair. 2019. “A New Species of Contarinia Rondani (Diptera: Cecidomyiidae) that Induces Flower Galls on Canola (Brassicaceae) in the Canadian Prairies.” The Canadian Entomologist 151(2): 1–18.

Murchie, A. K. 1996. Parasitoids of Cabbage Seed Weevil and Brassica Pod Midge in Oilseed Rape. PhD Thesis, University of Keele, Newcastle, UK.

Nacro, S., and J. P. Nénon. 2008. “Female Reproductive Biology of Platygaster diplosisae (Hymenoptera: Platygastridae) and Aprostocetus procerae (Hymenoptera: Eulophidae), Two Parasitoids Associated With the African Rice Gall Midge, Orseolia oryzivora (Diptera: Cecidomyiidae).” Entomological Science 11(2): 231–237.

Nakagawa, S., P. C. D. Johnson, and H. Schielzeth. 2017. “The Coefficient of Determination R2 and Intra-Class Correlation Coefficient from Generalized Linear Mixed-Effects Models Revisited and Expanded.” Journal of the Royal Society Interface 14(134): 1–11.

O’Brien, R. D. 2014. Insecticides: Action and Metabolism. Cambridge (USA): Academic Press.

Ogah, E., J. Odebiyi, A. Omoloye, and F. Nwilene. 2010. “Parasitism and Development of Platygaster diplosisae (Hymenoptera: Platygastridae) on the African Rice Gall Midge Orseolia oryzivora (Diptera: Cecidomyiidae).” International Journal of Tropical Insect Science 30(2): 93–100.

Olfert, O., R. H. Hallett, R. M. Weiss, J. J. Soroka, and S. Goodfellow. 2006. “Potential Distribution and Relative Abundance of Swede Midge, Contarinia nasturtii, an Invasive Pest in Canada.” Entomologia Experimentalis et Applicata 120(3): 221–228.

Pacchioli, M. A., and A. A. Hower 2004. “Soil and Moisture Effects on the Dynamics of Early Instar Clover Root Curculio (Coleoptera: Curculionidae) and Biomass of Alfalfa Root Nodules.” Environmental Entomology 33(2): 120–127.

Pan, C., H. Zhao, Q. Feng, J. Liu, L. Liu, Y. Cai, C. Liu, and J. Li. 2015. “Temporal Variations of Ground-Dwelling Arthropods in Relation to Grassland Salinization.” European Journal of Soil Biology 68: 25–32.

Pan, C., Q. Feng, J. Liu, Y. Li, and X. Yu. 2018. “Community Structure of Grassland Ground-Dwelling Arthropods Along Increasing Soil Salinities.” Environmental Science and Pollution Research 25(8): 7479–7486.

Patt, J. M., G. C. Hamilton, and J. H. Lashomb. 1997. “Foraging Success of Parasitoid Wasps on Flowers: Interplay of Insect Morphology, Floral Architecture and Searching Behavior.” Entomologia Experimentalis et Applicata 83(1): 21–30.

Philips, C., A. Ambourn, and L. Christianson. 2017. “First Detections of Swede Midge (Diptera: Cecidomyiidae) in Minnesota.” Journal of Entomological Science 52(3): 297– 300.

Picket, C., and R. Bugg. 1998. Enhancing Biological Control: Habitat Management to Promote Natural Enemies of Agricultural Pests. Berkeley: University of California Press.

Rand, T. A., W. L. Morrill, J. B. Runyon, K. A. Hoelmer, T. G. Shanower, J. L. Littlefield, and D. K. Weaver. 2016. “Assessing Phenological Synchrony Between the Chinese Sawfly, Cephus fumipennis (Hymenoptera: Cephidae), its Egg-Larval Parasitoid, Collyria catoptron (Hymenoptera: Ichneumonidae), and the North American Sawfly, Cephus cinctus: Implications for Biological Control.” The Canadian Entomologist 148(4): 482–492.

Readshaw, J. L. 1961. The Biology and Ecology of the Swede Midge, Contarinia nasturtii (Kieffer), (Diptera; Cecidomyidae). PhD Thesis, University of Durham, Durham, UK.

Readshaw, J. L. 1966. “The Ecology of the Swede Midge, Contarinia nasturtii (Kieffer) (Diptera, Cecidomyiidae). I. Life-History and Influence of Temperature and Moisture on Development.” Bulletin of Entomological Research 56(4): 685–700.

Rohrig, E., J. Sivinski., P. Teal, C. Stuhl, and M. Aluja. “A Floral-Derived Compound Attractive to the Tephritid Fruit Fly Parasitoid Diachasmimorpha longicaudata (Hymenoptera: Braconidae).” Journal of Chemical Ecology 34: 549–557.

Rosenheim, J. A. 1996. “An evolutionary argument for egg limitation”. Evolution 50: 2089–2094.

Roy, H. E., L. J. Lawson Handley, K. Schönrogge, R. L. Poland, and B. V. Purse. “Can the Enemy Release Hypothesis Explain the Success of Invasive Alien Predators and Parasitoids?” BioControl 56: 451–468.

Sandanayaka, W. R. M., and J. G. Charles. 2006. “Influence of Temperature and Flowers on Longevity of Adult Platygaster demades.” New Zealand Plant Protection 59: 184–189.

Sampson, B. J., S. J. Stringerand, and J. M. Spiers. 2002. “Integrated Pest Management for Dasineura oxycoccana (Diptera: Cecidomyiidae) in Blueberry.” Environmental Entomology 31(2): 339–347.

Schmale, I., F. L. Wäckers, C. Cardona, and S. Dorn. 2001. “Control Potential of Three Hymenopteran Parasitoid Species against the Bean Weevil in Stored Beans: The Effect of Adult Parasitoid Nutrition on Longevity and Progeny Production.” Biological Control 21(1): 134–139.

Sivinski, J., D. Wahl, T. Holler, S. A. Dobai, and R. Sivinski. 2011. “Conserving Natural Enemies with Flowering Plants: Estimating Floral Attractiveness to Parasitic Hymenoptera and Attraction’s Relationship to Flower and Plant Morphology”. Biological Control 58 (3): 208–214.

Sivinski, J. 2013. “Augmentative Biological Control: Research and Methods to Help Make it Work.” CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources 8(26): 1–11.

Shililu, J. I., W. B. Grueber, C. M. Mbogo, J. I. Githure, L. M. Riddiford, and J. C. Beier. 2004. “Development and Survival of Anopheles gambiae Eggs in Drying Soil: Influence of the Rate of Drying, Egg Age, and Soil Type.” Journal of the American Mosquito Control Association 20(3): 243–247.

Smith, S. M. 1996. “Biological Control with Trichogramma: Advances, Successes, and Potential of their Use.” Annual Review of Entomology 41(1): 375–406.

Snyder, W. E., and A. R. Ives. 2001. “Generalist Predators Disrupt Biological Control by a Specialist Parasitoid.” Ecology 82(3): 705–716.

Snyder, W. E., and A. R. Ives. 2003. “Interactions Between Specialist and Generalist Natural Enemies: Parasitoids, Predators, and Pea Aphid Biocontrol.” Ecology 84(1): 91– 107.

Solbreck, C., and O. Windenfalk. 2012. “Very Long Diapause and Extreme Resistance to Population Disturbance in a Galling Insect.” Ecological Entomology 37(1): 51–55.

Statistics Canada. 2017. “Principal Field Crop Areas, June 2017”. Accessed November 20, 2017. http://www.statcan.gc.ca/daily-quotidien/170629/dq170629c-eng.htm

Stireman, J. O., and M. S. Singer. 2002. “Spatial and Temporal Variation in the Parasitoid Assemblage of an Exophytic Polyphagous Caterpillar.” Ecological Entomology 27(5): 588–600.

100

Stokes, B. M. 1953a. “The Host Plant Range of the Swede Midge (Contarinia nasturtii Kieffer) with Special Reference to Types of Plant Damage.” European Journal of Plant Pathology 59: 82–90.

Stokes, B. M. 1953b. Biological Investigations into the Validity of Contarinia species Living on the Cruciferae, with Special Reference to the Swede Midge, Contarinia nasturtii (Kieffer). Annals of Applied Biology 40(4): 726–741.

Tena, A., A. Pekas, D. Cano, F. L. Wäckers, and A. Urbaneja. 2015. “Sugar Provisioning Maximizes the Biocontrol Service of Parasitoids.” Journal of Applied Ecology 52(3): 795–804.

Therneau, T. M. 2019. “Package ‘survival’”. Accessed November 10, 2019. https://cran.r-project.org/web/packages/survival/survival.pdf

Thies, C., and T. Tscharntke. 1999. “Landscape Structure and Biological Control in Agroecosystems”. Science New Series 285 (5429): 893–895.

Tillman, P. G. 2016. “Diversity of Stink Bug (Hemiptera: Pentatomidae) Egg Parasitoids in Woodland and Crop Habitats in Southwest Georgia, USA.” The Florida Entomologist 99(2): 286–291.

Tscharntke, T. 1992. “Cascade Effects Among Four Trophic Levels: Bird Predation on Galls Affects Density-Dependent Parasitism.” Ecology 73(5): 1689–1698.

Tscharntke, T., R. Bommarco, Y. Clough, T. O. Crist, D. Kleijn, T. A. Rand, J. M. Tylianakis, S. van Nouhuys, and S. Vidal. 2008. “Conservation Biological Control and Enemy Diversity on a Landscape Scale.” Biological Control 45(2): 238–253.

Tschumi, M., M. Albrecht, M. H. Entling, and K. Jacot. 2015. “High Effectiveness of Tailored Flower Strips in Reducing Pests and Crop Plant Damage.” Proceedings of the Royal Society B 282(1814): 1–8.

United States National Plant Germplasm System. 2019. “Lobularia maritima (L.) Desv.” Accessed November 11, 2019. https://npgsweb.ars- grin.gov/gringlobal/taxonomydetail.aspx?id=22478 van den Bosch, R. 1971. “Biological Control of Insects.” Annual Review of Ecology and Systematics 2: 45–66.

Van Driesche, R. G. 1983. “Meaning of ‘Percent Parasitism’ in Studies of Insect Parasitoids.” Environmental Entomology 12(6): 1611–1622. van Lenteren, J. C. 2012. “The State of Commercial Augmentative Biological Control:

101

Plenty of Natural Enemies, but a Frustrating Lack of Uptake.” BioControl 57: 1–20.

Various contributors. 2017. “Hymenoptera Online”. Accessed November 23, 2017. http://hol.osu.edu

Walker, F. 1836. “On the Species of Platygaster.” Entomological Magazine 3: 217–274.

Walkley, A. J., and I. A. Black. 1934. “Estimation of Soil Organic Carbon by the Chromic Acid Titration Method.” Soil Science 37(1): 29–38.

Watanabe, H., E. Yano, K. Higashida, S. Hasegawa, J. Takabayashi, and R. Ozawa. 2016. “An Attractant of the Aphidophagous Gall Midge Aphidoletes aphidimyza from Honeydew of Aphis gossypii.” Journal of Chemical Ecology 42: 149–155.

Weber, D. C., A. E. Hajek, and K. A. Hoelmer. 2017. “Accidental Introductions of Natural Enemies: Causes and Implications. Session 1: Accidental Introductions of Natural Enemies.” In Proceedings of the 5th International Symposium on Biological Control of Arthropods, edited by P. G. Mason, D. R. Gillespie and C. Vincent, 2–5. Boston: CAB International.

Westphal, C., S. Vidal, F. G. Horgan, G. M. Gurr, M. Escalada, H. V. Chien, T. Tscharntke, K. L. Heong, and J. Settele. 2015. “Promoting Multiple Ecosystem Services with Flower Strips and Participatory Approaches in Rice Production Landscapes.” Basic and Applied Ecology 16(8): 681–689.

Vincent, C. 2018. “Charles-Étienne Ferland ou le petit gars qui traversait la rue pour venir me poser des questions sur les insectes.” Antennae: Bulletin de la Société d’entomologie du Québec 25(3): 16–17.

Weaver, D. K., S. E. Sing, J. B. Runyon, and W. L. Morrill. 2004. Potential Impact of Cultural Practices on Wheat Stem Sawfly (Hymenoptera: Cephidae) and Associated Parasitoids. Journal of Agricultural and Urban Entomology 21(4): 271–287.

Wilde, G. E., R. A. Shufran, S. D. Kindler, H. L. Brooks, and P. E. Sloderbeck. 2001. “Distribution and Abundance of Insecticide Resistant Greenbugs (Homoptera: Aphididae) and Validation of a Bioassay to Assess Resistance.” Journal of Economic Entomology 94(2): 547–551.

Williams III, L., and D. L. Hendrix. 2008. “Comparing Different Floral Resources on the Longevity of a Parasitic Wasp.” Agricultural and Forest Entomology 10(1): 23–28.

Winkler, K., F. Wäckers, G. Bukovinszkine-Kiss, and J. van Lenteren. 2006. “Sugar Resources Are Vital for Diadegma semiclausum Fecundity Under Field Conditions.” Basic and Applied Ecology 7(2): 133–140.

102

Yuan, L. C., Y. B. Luo, L. B. Thien, J. H. Fan, H. L. Xu, and Z. D. Chen. 2007. “Pollination of Schisandra henryi (Schisandraceae) by Female, Pollen-Eating Megommata Species (Cecidomyiidae, Diptera) in South-Central China.” Annals of Botany 99(3): 451–460.

Zepeda‐Paulo, F. A., S. A. Ortiz‐Martínez, C. C. Figueroa, and B. Lavandero. 2013. “Adaptive Evolution of a Generalist Parasitoid: Implications for the Effectiveness of Biological Control Agents.” Evolutionary Applications 6(6): 983–999.

Zhao He, X., and Q. Wang. 2011. “Phenological Dynamics of Dasineura mali (Diptera: Cecidomyiidae) and its Parasitoid Platygaster demades (Hymenoptera: Platygasteridae) in Apple Orchards.” Journal of Economic Entomology 104(5): 1640– 1646.

Zhao He, X., and Q. Wang. 2015. “Ability of Platygaster demades (Hymenoptera: Platygastridae) to Parasitize Both Eggs and Larvae Makes it an Effective Natural Enemy of Dasineura mali (Diptera: Cecidomyiidae).” Journal of Economic Entomology 108(4): 1884–1889.

103

APPENDIX A

Impact of agricultural practices on parasitism of Contarinia nasturtii

Objectives and Methods

Growers that cultivated the canola fields sampled weekly in 2017 and 2018 provided information about their production practices including seeding date, seed variety, seed treatment, phytosanitary treatments (applications of insecticides, herbicides, and fungicides), crop rotation (history of past crops in the sampled field per year), and presence of adjacent canola in previous year (number of years passed since the sampled field was seeded in canola) (Figure A.1) at the end of each season. This information was received for 11 of 12 fields, and was used to make a preliminary assessment of the the impact of agricultural practices on parasitism of C. nasturtii by

Synopeas myles.

Producer: Field name: Field coordinates: When was this field planted? What variety was planted? Was there a seed treatment applied? Would it be possible to obtain spray records of 2018 for this field? (number of insecticides, herbicides, fungicides) Is it possible to obtain a history of this field (what was planted in the previous years)? If yes, could you list previous crops planted to the best of your knowledge? Is it possible to obtain information regarding adjacent crops in the previous year? If you know, could you please write them down below? If not, do you know who managed adjacent fields in the previous year? Figure A. 1. Canola growing practices questionnaire submitted to producers. 104

Results and Discussion7

Data from 11 fields were received (Table A.1). Growers seeded their fields between April 27 and May 28. There were 6 different seed varieties or mixes used, including LibertyLink varieties InVigor 5440 and L130 (Bayer, Leverkusen, Germany),

Pod Shatter Reduction (PSR) InVigor L140P and L223P (Bayer, Leverkusen,

Germany), and a mix of InVigor L252 (BASF, Ludwigshafen, Germany) - InVigor 255PC

(Bayer, Leverkusen, Germany), and a mix of 5440-L252-L223P. Seeds were treated with Lumiderm (DuPont, Midland, United States) or Prosper (Bayer, Leverkusen,

Germany). The number of insecticide, herbicide, and fungicide applications varied between 0 and 2, 1 and 3, and 0 and 1, respectively. The number of summers without canola in a given sampled field varied between 1 and 5. One field was seeded adjacent to a field that had canola in the previous year, while 9 were not.

Early seeding is recommended for better management of C. nasturtii (Hallett

2017), but it is unknown whether seeding date impacts S. myles populations by shifting host availability for C. nasturtii. The preliminary data indicate that seeding date did not influence parasitism rates (Figure A.2).

The impact of seed variety on C. nasturtii and, in turn, S. myles needed to be considered in a study of environmental factors and parasitism because insects can demonstrate herbivory and oviposition preferences and certain plant varieties exhibit

7 N.B. Throughout this Appendix, differences are numerical, and are not statistically significant. 105

resistance to insects or pathogens (Mahmood et al. 2012, Dara 2017, Gavloski 2017).

Varieties L223P8 and L252 seemed to have the highest parasitism rate, while fields including variety 5440 had the lowest parasitism rate (Figure A.3). Both L and LP varieties have resistance to Liberty® herbicides, and 5440 is an older variety which has been retired because better varieties are now available. The impact of seed variety genetics on insect development could be examined further.

Seed treatments and foliar applications of insecticides, herbicides and fungicides can have negative impacts on beneficial insects (Wilde et al. 2001, O’Brien 2014).

Therefore, the compatibility of pesticides with S. myles should be investigated. The higher parasitism rates were observed in canola fields sown with seed treatments though only one field had no seed treatment (Figure A.4). Parasitism rates were numerically lower in fields with 2 insecticide applications (Figure A.5), 3 herbicide applications (Figure A.6) and 1 fungicide application (Figure A.7), suggesting that it would be worthwhile examining the effect of pesticides on parasitoids further. The highest parasitism rates were observed in fields which had not been sown in canola for

2 – 3 years (Figure A.8), and in fields which were not adjacent to a canola field in the previous year (Figure A.9).

Data on agricultural practices from 11 fields is insufficient to draw strong conclusions; however, the relationship between pesticide use and parasitoid performance should be investigated further, as well as the effect of seed variety. Along

8 “L” stands for the LibertyLink® technology system, “P” for Pod Shatter Reduction. 106

with the data that was provided by the canola growers , future studies could collect information on yield, seed count per pod, seed weight, and racemes per plant, to investigate the economical impact, if any, of S. myles.

Table A. 1. Sample sizes and production practice information provided by canola growers. Question Answer n Seeding date (day-of-year) 135, 117, 136, 131 10 138, 147, 148, 146 132, 138

Seed varieties 5440 1 a mix of 5440-L252-L223P 1 L130 1 L140P 2 L223P 1 L252 4 a mix of L252-255PC 1

Seed treatment Lumiderm 3 Prosper 7 None 1

Number of insecticide applications 0 4 1 5 2 2

Number of herbicide applications 1 5 2 5 3 1

Number of fungicide applications 0 7 1 4

Number of summers without canola in a given sampled field 1 3 2 3 3 2 4 1 5 1

Field seeded next to a field that had canola in the previous year No 9 Yes 1

107

25 y = 0.0022x + 3.3443 20 R² = < 0.01

5 Parasitism rate (%) rateParasitism

0 110 120 130 140 150 Seeding date (day of year)

Figure A. 2. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the date at which the crop was seeded (day of year) (±SE) (n sites = 11).

5 Parasitism rate (%) rateParasitism 0 5440 5440 L130 L140P L223P L252 L252 L252 255PC L233P Seed variety

Figure A. 3. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the seed variety used (±SE) (n sites = 11).

108

5 Parasitism rate (%) rateParasitism

0 Lumiderm None Prosper Seed treatment

Figure A. 4. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the seed treatment applied (±SE) (n sites = 11).

5 Parasitism rate (%) rateParasitism

0 0 1 2 Number of insecticide applications

Figure A. 5. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of insecticide applications (±SE) (n sites = 11).

109

10 Parasitism rate (%) rateParasitism

0 1 2 3 Number of herbicide applications

Figure A. 6. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of herbicide applications (±SE) (n sites = 11).

10 Parasitism rate (%) rateParasitism

0 0 1 Number of fungicide applications

Figure A. 7. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of fungicide applications (±SE) (n sites = 11).

110

4 Parasitism rate (%) rateParasitism

0 1 2 3 4 5 Summers without canola in the field

Figure A. 8. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the number of summers without canola in that field (±SE) (n sites = 11).

35 30 25 20 15 10

Parasitism rate (%) rateParasitism 5 0 No Yes Canola adjacent in previous year

Figure A. 9. Parasitism rate (%) of Contarinia nasturtii by Synopeas myles according to the presence of adjacent canola in previous year (±SE) (n sites = 11).

111