Title: Murgantia histrionica (Hahn): new trapping tactics and insights on overwintering survival Name: Anthony Stephen DiMeglio

Thesis submitted to the faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of Master of Science in Life Sciences In Entomology

Thomas P. Kuhar, Chair Donald C. Weber, Co-Chair Dorothea Tholl

September 12, 2018

Keywords: (Harlequin bug, Stink bug, Brassica, Trap, Aggregation pheromone, Murgantiol, Low temperature biology, pest management)

Title Murgantia histrionica (Hahn): new trapping tactics and insights on overwintering survival ABSTRACT The harlequin bug, Murgantia histrionica (Hahn), is a serious pest of brassicaceous vegetables in southern North America, with limited establishment north of the 40°N latitude presumably due to low overwintering survival. Integrated Pest Management (IPM) requires knowledge of pest populations and tools to monitor them. For harlequin bug, knowledge of the number of successfully overwintered bugs, and development of an effective trap to monitor populations, are essential to its management. To gain insight into overwintering survival, I determined the supercooling points (SCPs) for Maryland and Virginia adult populations and found no significant difference between these populations. SCPs were similar for adults (푋̅ = -10.4oC;

nd rd th th o 휎푋 = 2.5) and early (2 – 3 ) and late (4 – 5 ) instar nymphs (푋̅ = -11.0 C; 휎푋 = 4.9) and between

st o adult males and females. However, SCPs for 1 instars (푋̅ = -21.6 C; 휎푋 = 1.5) and eggs (푋̅ =

o -23.2 C; 휎푋 = 1.0) were significantly lower. Field survival of overwintering harlequin bug adults was significantly impacted (with 80-96% mortality) during widespread air temperatures lower than

-15oC and sub-freezing soil temperatures in the mid-Atlantic region. Our results provide new information on M. histrionica overwintering biology, and thermal limitations to its distribution, which leads to improved predictive capabilities to forecast pest severity. To monitor harlequin bug activity for IPM recommendations an effective trap is necessary; at this time no such trap exists.

The research presented herein contributes new knowledge of harlequin bug visual ecology, which will aid in the development of an effective trap. In both lab and field color choice experiments, harlequin bug adults and large nymphs responded positively to green and black colors, and statistically less frequently to yellow, white, purple or red with the exception of adult females, which were most attracted to red and green in the lab, but green and black in the field. To develop an effective trapping device for this pest, experiments were conducted in Virginia to assess factors

to increase harlequin bug attraction to and arrestment at fixed artificial stimuli (“traps”) positioned within an agricultural landscape. In a laboratory experiment, harlequin bugs were effectively killed or severely impaired after a few minutes exposure to a synthetic pyrethroid-incorporated netting

(D-Terrence®, Vestergaard-frandsen), and, thereafter, the netting was used as a toxicant on the trapping devices in the field. In one experiment, square corrugated plastic panels were wrapped with the insecticide netting and baited with harlequin bug aggregation pheromone, murgantiol.

Bugs were effectively drawn to the panels, with green panels having significantly more dead harlequin bugs and fewer dead lady beetles (Coleoptera: Coccinellidae) at their base than yellow panels. Thus, green was chosen as the ideal trap color to use for another field experiment that evaluated three trap types – a corrugated plastic square panel, a pyramidal trap, and a ramp trap – each with three lure treatments, murgantiol alone or murgantiol plus a low or high rate of benzyl isothiocynate. More bugs were killed with the pyramidal trap than with the panel trap or the ramp trap, and more bugs were killed at traps containing murgantiol combined with benzyl isothiocyanate than at those with murgantiol alone. This research demonstrated that, with use of proper visual and semiochemical stimuli, harlequin bugs can be drawn to trapping devices and effectively killed after contact with deltamethrin-incorporated netting.

GENERAL AUDIENCE ABSTRACT Harlequin bugs are orange and black aggregation pheromone emitting stink bug pests, specifically of cole crops such as kale, broccoli and collards. This nearly loyal crop preference makes an interesting challenge for trapping them and helping farmers predict pest severity. Harlequin bugs can be found in much of North America, and are a serious problem in the southeastern United

States. Presumably their persistence into northern regions is limited by extreme winters. In 2014 and 2015 the arctic polar vortex extended into mid-latitudes bringing a blanket of sustained sub- freezing temperatures to much of the United States. We used these events to determine effects of extreme winter weather on harlequin bug survival. In both years we observed nearly identical low temperatures of -15oC and this linked to high (80-96%) harlequin bug mortality. In the lab we measured exact lethal freezing temperatures in harlequin bugs (i.e. supercooling points) to see if a physiological metric could be used to predict overwinter survival. Harlequin bug adults froze and died at -10.4oC, and similarly, their larger juvenile stages freeze at -11.0oC. Freshly hatched harlequin bugs and unhatched eggs froze at considerably lower temperatures with eggs forming ice crystals at -23.2oC and recent hatches at -21.6oC. Now with an understanding of how harlequin bugs likely survive winter extreme, we can then work on developing a trap to tally their populations in the spring and predict summer and fall pest severity. In the lab and field, harlequin bug adults and large nymphs were more likely found on green and black colors, and statistically less frequently on yellow, white, purple or red colors with the exception of adult females, which were most attracted to red and green in the lab, but green and black in the field. To increase harlequin bug attraction to and termination at traps square corrugated plastic panels were wrapped with an insecticide netting and baited with harlequin bug aggregation pheromone, murgantiol. Bugs were effectively drawn to the panels, with green panels having significantly more dead harlequin bugs and fewer dead beneficial lady beetles (Coleoptera: Coccinellidae) at their base than yellow

panels. Thus, green was chosen as the ideal trap color to use for another field experiment that evaluated three trap types – a corrugated plastic square panel, pyramidal trap, and ramp trap – each with three lure treatments, murgantiol alone or murgantiol plus a low or high rate of mustard oil.

More bugs were killed with the pyramidal trap than with the panel trap or the ramp trap, and more bugs were killed at traps containing murgantiol combined with benzyl isothiocyanate than at those with murgantiol alone. This research demonstrated that with the proper visual elements and odors, harlequin bugs can be drawn to traps and effectively killed after contact with insecticide- incorporated netting.

Acknowledgements

It is worth a listen: RadioLab host, Jab Abumrad, uses a theoretical biologist’s idea of the “adjacent possible” as a back-drop to describe how thoughtful stories culminate. Like an organism’s evolutionary history, this thesis is a fragment of a larger story of those most adjacent that provided the necessary environment to succeed. The ideas, experiments and narrative in this thesis were stimulated by many patient and understanding individuals. Dr. Donald C. Weber and many others at the USDA-ARS provided the mentorship and invaluable time in developing my critical eye on biological sciences. Dr. Thomas P. Kuhar, and his outstanding leadership, was vital in honing in many of my obscure ideas into formable research questions that emphasized on “do- good” outcomes. Dr. Dorothea Tholl shared a world novel to many, molecular biochemistry, and inspired me to push my limits of thought, inquiry and measurements. Dr. Tim Kring was an understanding force who kept the process going. A loud “thank you” is more than well deserved to my fellow Vegetable Insect Pest Research lab mates, and Department of Entomology and Biological Science colleagues for their support.

As someone with bipolar, I understand how science can create both stimulating and degrading environments—and this thesis carries a legacy of extreme ups and downs. The people adjacent to me were affected greatly while I was ignorant to my condition. Learned from this thesis are lessons of openness and vulnerability. Through the patience of others, and surviving the struggles first hand, I was able to build the tools necessary to cope and grow past the condition. For those that are learning about their struggle with bipolar/depression as a daily survivor I encourage you to push forward and open up to those adjacent. The feelings you feel are OK and people will understand, and share their warmth.

These financial donors made the successful defense of this thesis possible: Carolyn Elmore, Barbra Hiscock, Deborah Troehler, Meredith McQuoid-Greason, Gail Enright, Corinne Noirot, and Liane Ripley. Without their financial gifts, this thesis would be years delayed. Many grants and agencies supported this project and my coursework.

Words cannot articulate well enough how the beautiful Lauris McQuoid-Greason was instrumentally vital in bringing this all together. She was there in the field, on the phone, and as a reassuring push to make it through to the end. Lauris, I deeply appreciate your forgiving and endless patience, undivided love, and commitment to building a genuine fire.

vi

Chapter 1: Literature Review Harlequin bug pest status and control tactics …………………………………... 1 Understanding climate as a population control mechanism: cold temperatures and climate change in mid latitude regions ….……………………….….……… 5 Glucosinolate-Myrosinase associations in harlequin bug ……………………… 6 Discovery, identification and synthesis of murgantiol ……………………...….. 7 Pentatomid trap development ……………………………………....……………10 Developing a trap for harlequin bug: a preface …………………………….…. 10 References …………………………………………....…………….…………….. 11 Chapter 2: Supercooling Points of Murgantia histrionica (Hemiptera: Pentatomidae) and Field Mortality in the Mid-Atlantic United States following Lethal Low Temperatures Abstract …………………………………………………………………………. 17 Introduction ….……………………………………………….………..……….. 18 Materials and Methods ……………………………………………..………….. 21 Results and Discussion ………………………………………….….………...… 27 References …………………………………………...……………....………….. 33 Chapter 3: Color Preference of Harlequin Bug, Murgantia histrionica (Heteroptera: Pentatomidae) Abstract ………………………………………………………………………….. 40 Introduction ….……………………………………………….………..………... 41 Materials and Methods ……………………………………………..…………... 42 Results ………………..………………………………………….….………...…. 44 Discussion ………………………………………………………………………... 45 References …………………………………………...……………....…………... 48 Chapter 4: Optimizing Attract-and-Kill Technology for Managing Harlequin Bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae) Abstract …………………………………………………………………………. 52 Introduction ….……………………………………………….………..……….. 53 Materials and Methods ……………………………………………..………….. 56 Results and Discussion …………………………………………………………. 61 References …………………………………………...………………...... ……... 70 Chapter 5: Conclusions and Future Considerations ………………………………...... 84

vii

List of Figures Chapter 2: Supercooling Points of Murgantia histrionica (Hemiptera: Pentatomidae) and Field Mortality in the Mid-Atlantic United States following Lethal Low Temperatures 2.1 – Soil and Air Temperatures ………………………………………..………. 36 2.2 – Supercooling Point Comparison….………………………..…………..….. 37 2.3 – Mean Egg Mortality ………………………………………………………. 38 2.4 – Cumulative Distribution of Supercooling Points ……………………...… 39

Chapter 3: Color Preference of Harlequin Bug, Murgantia histrionica (Heteroptera: Pentatomidae) 3.1 – Mean Bugs Collected ………………………..…………………….………. 51

Chapter 4: Optimizing Attract-and-Kill Technology for Managing Harlequin Bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae) 4.1 – Field Layout for Nine Treatment Test …….……………………..………. 74 4.2 – Field Layout for Fence Post Test …….……………………..………….…. 75 4.3 – Visitation and Tenure Exp. Layout ………….…………..…………..…... 76 4.4 – Green versus Yellow Trap Captures ...…………………...…………….… 77 4.5 – Semiochemical and Shape Field Test ………………………...……...…… 78 4.6 – D-Terrence versus Tanglefoot Test ………………………………………. 79 Supplemental 4.1 -- D-Terrence® Mortality Assay ………………………..….. 82 Supplemental 4.2 – Images from Green and Yellow Trap Study …..……….... 83

viii

List of Tables Chapter 3: Color Preference of Harlequin Bug, Murgantia histrionica (Heteroptera: Pentatomidae) 3.1 – Mean Harlequin Bug Response in Lab Color Choice Test...………...... 50

Chapter 4: Optimizing Attract-and-Kill Technology for Managing Harlequin Bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae) 4.1 -- Results from a two-way ANOVA (p<0.05) performed on trap collection during a field experiment investigating efficacy of semiochemical baits and trap shape on dispersing harlequin bugs ………………..…………………………….... 80 4.2 -- Results from a two-way ANOVA (p<0.05) performed on trap collection during a field experiment investigating efficacy of semiochemical baits and trap shape on dispersing harlequin bugs …………………...... 81

ix

Chapter One

Literature Review

Harlequin bug pest status and control tactics

Three of the twenty-five documented genera (e.g. Eurydema, Bagrada, and Murgantia ) within Pentatomidae are Brassicaceae specialists (Panizzi and Lucini 2017). In North America the harlequin bug, Murgantia histrionica Hahn, is a major pest of Brassica spp., with two to three generations per year causing severe economic loss in crops such as collards, broccoli and kale

(Wallingford et al. 2011). Native to Central America (Paddock 1918), its expansion and persistent pest status across North America after it was first collected in Texas in 1864 (Paddock 1915) was well documented in the late 19th and early 20th centuries. Although typically only a pest south of latitude 40° N (Hodson and Cook 1960), depending on extreme winter temperatures, harlequin bugs occur throughout much of the continental US from New Hampshire and New York south to

Florida and west to North and South Dakota, Nebraska, and California (Froeschner 1988, Rider

2012).

Due to a limited range of control tactics prior to the advent of synthetic insecticides, the harlequin bug was a serious agricultural pest over the course of seven decades. During this time multiple authors (Paddock 1915, Thomas 1915, Paddock 1918, Walker and Anderson 1933, 1939,

White and Brannon 1939) voiced similar suggestions on remedial, artificial and natural controls which are still applicable tactics today when synthetic insecticide options are restricted, banned or not allowed within certain agricultural practices (e.g. organic approaches).

Thomas (1915) defined remedial measures to include (1) destroying overwintering habitat and host plant refuge in late fall or early winter, and (2) trap crops where alternate host plants

(e.g. an early planting of radish, kale, mustard as a trap crop for less preferred cabbage) reduce

- 1 - injury to food crops; trap crops would then be managed with kerosene or burning. Paddock (1915,

1918) expanded cultural control recommendations with emphasis on clean culture and hand- picking bugs off host plants. A spray formulation comprised of a mixture of nicotine sulfate, fish oil soap, and water was also described to kill 65–75% of nymphs and 45–50% of adults by

Paddock (1918). Paddock (1918) also showed that the inorganic arsensical preparations Paris

Green and London Purple were ineffective at controlling harlequin bugs.

Natural controls of harlequin bug populations include incidental mortality from rainfall, low overwintering survival (Paddock 1918), and predators such as the wheel bug—Arilus cristatus (L) (Hemiptera: Reduviidae) (White and Brannon 1939), and hymenoptera egg parasitoids Trissolcus euschisti (Ashmead) (Hymenoptera: Scelionidae) (Paddock 1918),

Trissolcus brochymenae (Ashmead) (Hymenoptera: Scelionidae) (Paddock 1918) and Ooencyrtus johnsoni (Ashmead) (Hymenoptera: Encyrtidae) (White and Brannon 1939). Walker and

Anderson (1933) observed 35–55% of the eggs parasitized by O. johnsoni collected during

August and September. White and Brannon (1939) reported similar egg parasitism rates within those months.

As chemical control tactics became more effective at reducing harlequin bug populations the literature placed less emphasis on cultural and natural controls, and directed its focus to chemical efficacy. When nicotine, pyrethrum and oil emulsion sprays were found relatively ineffective (Walker and Anderson 1933), research turned to more hazardous insecticide treatments of derris (Derris elliptica Fabaceae) or cubé (Deguelia utilis Fabaceae) powder

(rotenone content of 5 or 6%) for greater lethality (Walker and Anderson 1939). Shortly after this work, synthetic chemistries began to dominate harlequin bug population control recommendations

- 2 - with the advent of DDT and other chlorinated hydrocarbons (Brooks and Anderson 1947, Gaines and Deane 1948).

Once chlorinated hydrocarbon insecticides became notorious for their ecotoxicology and persistence, regulatory restrictions came into effect, and research attention redirected to alternate chemistries. In response, carbamates and organophosphates entered the literature in the 1960s and

1970s for broad spectrum control of harlequin bugs. These broad spectrum cholinesterase inhibitors such as parathion, carbaryl, acephate, and diazinon, were effective at controlling harlequin bugs (Hofmaster 1959, Rogers and Howell 1972). Organophosphates and carbamates raised human health concerns, and in 1996 the United States Environmental Protection Agency’s

Food Quality Protection Act further restricted these chemistries. Synthetic pyrethroids became the predominate chemistry for insect control throughout the 1970s and into the 1990s. Pyrethroid insecticides including permethrin, bifenthrin, lambda-cyhalothrin, zeta-cypermethrin, cyfluthrin, and several others that were shown to be highly efficacious at low concentrations for control of harlequin bugs (Edelson 2004a, 2004b, Edelson and Mackey 2006a). Pyrethroids, however, are also quite lethal to most beneficial organisms (Michaud and Grant 2003), and their excessive wide-scale use has led to resistance development in many key insect pests (Li et al. 2007).

These continuing concerns surrounding insecticide development and use led entomologists to develop the integrated control concept (Stern et al. 1959), and later, integrated pest management, during the late 1960’s and throughout 1970s (Kogan 1998). The integrated control concept, as introduced by Stern (1959), was not to campaign against the use of chemical control in agricultural systems. Instead it focused on limiting chemical control to when it was really needed, in concert with cultural and biological control of pest arthropods within agricultural landscapes.

- 3 -

This interest towards combing biological and chemical control with intent to decrease ecological risks encouraged the development and commercialization of neonicotinoids, which target the nicotinic acetylcholine receptors. Neonicotinoids such as acetamiprid, clothianidin, dinotefuran, imidacloprid, thiacloprid, and thiamethoxam are extremely effective against multiple pests (Edelson 2004a, Edelson and Mackey 2005a,b,c, 2006b; Walgenbach and Schoof 2005).

Once researchers discovered neonicotinoids were water soluble and can be absorbed by xylem tissue (Abbink 1991, Elbert et al. 1998, Sur and Stork 2003, Tomizawa and Casida 2005) they became a primary tool against piercing sucking insect pests. Wallingford et al. (2012) showed that soil drench treatments of imidacloprid, thiamethoxam, clothianidin, or dinotefuran were effective at controlling harlequin bugs weeks after application.

Neonicotinoids are neurotoxins, and use of such toxicants in food systems poses risks, some of them subtle and indirect. In 2013 Europe applied a temporary ban on the use of common neonicotinoids due to much debated (Cressey 2013) research on pollinator health, specifically with its effect on weakening immunity in honey bees (Di Prisco et al 2013). Recently, Dewar

(2017) showed that the ban on neonicotinoids imposes adverse pressure on agricultural productivity with increases in pest activity on farms. Chemical controls in pest management tactics play an important role in agricultural productivity. Therefore, in order to provide a greater diversity of pest management tools future research is needed to support the next tier of pest control tactics for harlequin bugs that creatively recombines natural, cultural and chemical control tactics.

- 4 -

Understanding climate as a population control mechanism: cold temperatures and climate change in mid-latitude regions

In temperate climates, where harlequin bugs regularly persist as a pest, severity of winter extreme temperatures varies widely. Extreme winter events in the Northern Hemisphere are governed by the amplification of the circumpolar vortex. When the amplification of the polar vortex is prolonged, widespread extremely low temperatures are likely in central and mid-Atlantic

States (Frauenfeld and Davis 2003). In January 2014 and 2015, much of the United States experienced extreme lows from such events back to back, which was colloquially referred to as the “polar vortex”.

Over 110 years earlier, in mid-February 1899, a record-breaking polar vortex event covered much of the US (Kocin et al. 1988). Sanderson (1908) reported noteworthy reductions in harlequin bug populations following this bitterly cold event, when minimum temperatures reached -28.9oC in Ohio and -26.1oC in Washington D.C. (Sanderson 1908), and -15oC in the

Carolinas (Kocin et al. 1988). According to Sinclair et al. (2003), most temperate insects in the

Northern Hemisphere are freeze intolerant, with internal ice formation lethal to this group of insects. Given that harlequin bug populations decreased following extreme winters, we can hypothesize that harlequin bugs are freeze intolerant. Events of prolonged sub-freezing temperatures may occur more frequently in the southeastern United States than in the past decades (Bryson 1975, Franuenfeld and Davis 2003, Limpasuvan et al. 2004, Francis and Vavurs

2012), so it is important understand how insect populations in temperate climates are affected by potentially lethal lower temperatures. Hemolymph will instantaneously crystallize within a range of temperatures; the temperature at which this occurs is referred to as the supercooling point

(Baust and Rojas 1985). Elsey (1993), reported a supercooling point of -14.7oC for harlequin bug,

- 5 - but details of how they determined it and the population source and sample size are not clear. We need to know more about harlequin bug low temperature biology for all life stages to relate laboratory measurements with field survival studies.

Glucosinolate-Myrosinase associations in harlequin bug

Harlequin bugs are attracted to glucosinolate-containing plants (Brassicaceae and

Capparaceae) and sequester plant based glucosinolates in prothoracic region; upon disturbance they secrete a frothy mixture of aglucones of glucosinolates and alkylrnethoxypyrazines as a defense against vertebrate predators (Aldrich et al. 1996, Aliabadi et al. 2002). In the plant, glucosinolate breakdown products (e.g. nitriles and thiocyanates) defend the plant against fungal infection, and aboveground and belowground herbivory (Wittstock et al. 2003); however, numerous insects including members of the Lepidoptera, Hemiptera, Diptera, and Coleoptera, evolved to exploit the glucosinolate-myrosinase system for their own chemical ecology and predator defense systems (Wittstock et al. 2003). The glucosinolate-myrosinase system produces a “mustard-bomb” effect when the enzyme, myrosinase—normally isolated in the plant—comes in contact with glucosinolate sugars; this happens upon plant cell wall disturbance.

Ludwig and Kok (1998) reported a general preference of harlequin bug for mustards

(Brassica juncea L.) over broccoli, and Wallingford et al. (2013) showed a preference for mustards over collards, which they suggest is due to a relationship between harlequin bug host attraction and glucosinolate levels. Species of Brassicaceae are highly variable with regard to their glucosinolate constituents (Fahey et al. 2002) with more than 120 glucosinaolate glucosides. Plant age is a major determinant of glucosinolate quality and quantity (Bones and Rossiter 2006), but age is not as influential as genotype (Farnham et al. 2004). While glucosinolates make up less than 1% biomass dry weight (Fahey et al. 2002), the major and minor glucosinolate components

- 6 - likely serve as significant signals that drive harlequin bug host preference, attraction, retention and dispersal. Thrift et al. (2018) recently found that harlequin bug attraction to host and non-host plants is enhanced with the presence of the glucosinolate breakdown products allyl isothiocyanate and benzyl isothiocyanate, with benzyl outperforming allyl isothiocyanate as an attractant.

Since isothiocyanates are predominant volatile products from myrosinase glucosinolate hydrolysis (e.g. “mustard-bomb”) (Fahey et al. 2002), and Thrift et al. (2018) found harlequin attraction to be significantly stronger to benzyl than allyl isothiocyanate, I focused on benzyl isothiocyanate, as a candidate host attractant for murgantiol baited traps.

Discovery, identification and synthesis of murgantiol

Pheromones are volatile cues (i.e. semiochemicals) used for communication within a species. In the literature a pheromone is characterized by its function. For example, a pheromone that attracts both sexes is referred to an aggregation pheromone (Weber et al. 2018).

With gas chromatography-mass spectrometry Zahn et al. (2008, 2012) discovered that the sexually mature male harlequin bug emits a 10,11-epoxy-1-bisabolen-3-ol as its aggregation pheromone, dubbed murgantiol. In the bug, murgantiol is synthesized de novo via a terpene synthase (Lancaster et al. 2018).

Working in describing pheromone synthesis for both H. halys and harlequin bugs,

Khrimian et al. (2014) found two specific diastereomers of 10,11-epoxy-1-bisabolene-3-ol that constitute the harlequin bug aggregation pheromone. In finer detail, the dual isomers naturally emitted at a 1.4:1 ratio ([3S,6S,7R,10S]: [3S,6S,7R,10R]) were preferred by female harlequin bugs over more commercially viable 1:1 ratios (Weber et al. 2014b). All life stages are attracted to murgantiol isomers regardless of natural or crude blends. Khrimian et al. (2014) can be further credited for aiding the increased commercially viability of pentatomid lure synthesis with their

- 7 - work on H. halys and harlequin bugs. Their most notable contributions led to (1) a direct synthesis of 1,10-bisaboladien-3-ol and 10,11-epoxy-1-bisabolen-3-ol isomers using diastereomeric mixtures of cyclohex-2-enones and ketones synthesized from (R)- and (S)- citronellals, with a rhodium-catalyzed asymmetric addition of trimethylaluminum, and (2) - additions of minor and non-attractive stink bug aggregation pheromone isomers were not repellant to harlequin bugs and H. halys, making the issues of commercial lures being species specific less of a concern. In summary, murgantiol isomers can be included with isomers for H. halys aggregation pheromone, which is high favorable in producing commercial lures because a single mixed isomer lure is attractive to both.

Pentatomid trap development

Currently a harlequin bug specific trap does not exist. Fundamental research, however, has determined harlequin bug specific “attract-and-kill” factors utilizing murgantiol in combination with various attractants and trapping materials. With mark-recaptures studies, Cabrera Walsh et al. (2016) discovered that murgantiol increased harlequin bug densities when paired with a collard plant, but did not enhance retention times. This important finding reveals the limitations of murgantiol in increasing harlequin bug retention on traps and trap crops, and further adding to our understanding of harlequin bug dispersal behavior. Prior to murgantiol discovery and synthesis,

English-Loeb and Collier (1987) observed migratory behavior on Isomeris arborea (Capparaceae) in coastal California, with harlequin bug tenure positively correlated with plant flowering activity rather than harlequin bug sex ratios on release plants; they also found that females were more likely to disperse greater distances from release sites than males. With this information from

Cabrera Walsh et al. (2016) and English-Loeb and Collier (1987), we can broadly infer that harlequin bug tenure may be brief and they are unlikely loyal to a single host plant, and even

- 8 - briefer when landing on an artificial stationary stimulus. With recent work conducted by

Wallingford et al. (2018) we have more evidence that suggests that once harlequin bugs are recruited to a stationary stimulus (i.e. baited plants) they will likely remain within the vicinity/patch of the targeted stimuli (as expected and discussed per Cardé, 2014), but retention at fixed stimuli is brief. This perpetuates a problem for harlequin bug spill-over onto neighboring crops if harlequin bugs are not quickly terminated at the targeted stimuli, as shown by

Wallingford et al. (2018), further indicating effective prompt and chemical controls are necessary to reduce crop injury from an attract-and-kill IPM strategy. Semiochemical baits recruit harlequin bugs from a distance, but the role of attractants in “halo” and “spillover” (i.e. vicinity effects) is still poorly understood (Wallingford et al. 2018).

Pentatomid trap research has progressed from yellow pyramidal traps--i.e. the Tedder Trap

(Tedder and Wood 1994)—intended to capture Curculio caryae (Horn) (Coleoptera:

Curculionidae) due to the traps’ “tree trunk” visual profile (Mizell and Tedders 1995, Mizell et al.

1996. Leskey and Hogmire (2005) expanded the uses of the Tedder Traps for other stink bug species and orchard systems, which advanced the development of species specific trap design for stink bugs. As H. halys became a widespread polyphagous global pest with economic priority,

Leskey et al. (2012) questioned the yellow Tedder traps, specifically its color in trapping H. halys, finding that instead of the yellow traps black was more effective at trapping H. halys. Shimat et al. (2014) go further as to specifically look into trap-top containers to increase H. halys retention at the trap.

The Tedder Trap is well matched to the innate ascending behavior of stink bugs (Morrison et al. 2016). With improved trap-top entry refined by Shimat et al. (2014) and trap design further described by Morrison et al. (2015) effective H. halys trap captures increased, but there were still

- 9 - reports that H. halys would exit the trap. Kuhar et al. (2017) went on further to ensure effective kills once in the trap container by introducing an insecticide-impregnated polyethylene screen mesh. The Tedder Trap, unfortunately, with all its improvements had its flaws; H. halys would still be found alighting at the trap and not entering the collection trap-top device (Leskey et al.

2012b, Morrison et al. 2015).

Developing a trap for harlequin bug: a preface

Given the known pest status of harlequin bug and limited knowledge on its overwintering survival and specifics on visual and chemical ecology, my work focused on bridging these knowledge gaps. My research began with overwintering survival field studies, and basic low temperature biology (i.e. supercooling point measurements). This was intended to develop a

“grower-oriented” pest prediction index to estimate pest severity, and is detailed in Chapter 2 and published as DiMeglio et al. (2016). Monitoring populations for IPM decisions and validating said pest prediction index would require a well-developed trap—and no such trap currently exists.

I began developing a trap design concept with field and lab studies evaluating relative color preference, which is presented in Chapter 3 as published in DiMeglio et al. (2017). Altogether, I sought to develop an attract-and-kill strategy that offered negligible impact on natural enemy populations while optimizing harlequin bug suppression. Harlequin bug attraction to a stationary artificial stimulus (i.e. traps) was investigated by researching relative field attraction to trap shapes with and without various semiochemical baits; this work is described in Chapter 4.

- 10 -

References

Abbink, J. 1991. The biochemistry of imidacloprid. Pflanzenschutz-Nachrichten Bayer (Germany, FR). Aldrich, J. R., J. W. Avery, C. J. Lee, J. C. Graf, D. J. Harrison, and F. Bin. 1996. Semiochemistry of cabbage bugs (Heteroptera: Pentatomidae: Eurydema and Murgantia). J. Entomol. Sci. 31: 172–182. Aliabadi, A., J. A. Renwick, and D. W. Whitman. 2002. Sequestration of glucosinolates by harlequin bug Murgantia histrionica. J. Chem. Ecol. 28: 1749-1762. Baust, J. G., and R. R. Rojas. 1985. Review—Insect cold hardiness: Facts and fancy. J. Insect Physiol. 31: 755–759. Bones, A. M., and J. T. Rossiter. 2006. The enzymic and chemically induced decomposition of glucosinolates. Phytochemistry. 67: 1053–1067. Brooks, W., and L. D. Anderson. 1947. Toxicity tests of some new insecticides. J.Econ. Entomol. 40: 220–228. Bryson, R. A. 1975. The Lessons of Climatic History. Environ. Conser. 2(3), 163–170. Cabrera Walsh, G., A. S. DiMeglio, A. Khrimian, and D. C. Weber. 2016. Marking and retention of harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae), on pheromone baited and unbaited plants. J. Pest Sci., DOI 10.1007/s10340-015-0663-1. Cardé, R.T. 2014. Defining attraction and aggregation pheromones: teleological versus functional perspectives. J. Chem. Ecol. 40: 519-520. Cressey, D. 2013. Europe debates risk to bees. Nature, 496(7446), 408 Dewar, A. M. 2017. The adverse impact of the neonicotinoid seed treatment ban on crop protection in oilseed rape in the United Kingdom. Pest Manag.Sci., 73(7), 1305-1309.

DiMeglio, A. S., A. K. Wallingford, D. C. Weber, T. P. Kuhar, and D. Mullins. 2016. Supercooling points of Murgantia histrionica (Hemiptera: Pentatomidae) and field mortality in the mid-Atlantic United States following lethal low temperatures. Environ. Entomol. 45: 1294-1299.

DiMeglio, A. S., T. P. Kuhar, and D. C. Weber. 2017. Color preference of harlequin bug (Heteroptera: Pentatomidae). J. Econ. Entomol. 110: 2275–2277. Di Prisco, G., V. Cavaliere, D. Annoscia, P. Varricchio, E. Caprio, F. Nazzi, G. Gargiulo, and F. Pennacchio. 2013. Neonicotinoid clothianidin adversely affects insect immunity and promotes replication of a viral pathogen in honey bees. Proc. Natl. Aca.d Sci. U. S. A., 110(46), 18466-18471.

- 11 -

Edelson, J. V. 2004a. Comparison of nicotinoid insecticides for controlling harlequin bug, 2003. Arthropod Management Tests 29: E22. Edelson, J. V. 2004b. Harlequin bug control on collard, 2003. Arthropod Management Tests 29: E24 Edelson, J. V., and C. Mackey. 2005a. Comparison of nicotinoid insecticides for controlling harlequin bug, 2004. Arthropod Management Tests 30: E23. Edelson, J. V., and C. Mackey. 2005b. Comparison of nicotinoid insecticides for controlling harlequin bug and thrips 2004. Arthropod Management Tests 30: E24. Edelson, J. V., and C. Mackey. 2005c. Comparison of nicotinoid insecticides for controlling harlequin bug and thrips, 2004. Arthropod Management Tests 30: E94. Edelson, J. V., and C. Mackey. 2006a. Comparison of nicotinoid insecticides for controlling harlequin bug, 2005. Arthropod Management Tests 31: E17. Edelson, J. V., and C. Mackey. 2006b. Comparison of pyrethroid insecticides for controlling harlequin bug, 2005. Arthropod Management Tests 31: E18. Elbert, A., Nauen, R., & Leicht, W. 1998. Imidacloprid, a novel chloronicotinyl insecticide: biological activity and agricultural importance. In Insecticides with novel modes of action (pp. 50-73). Springer, Berlin, Heidelberg Elsey, K. D., 1993. Cold tolerance of the southern green stink bug (Heteroptera: Pentatomidae). Environ. Entomol. 22: 567-570. English-Loeb, G. M., and B. D. Collier. 1987. Nonmigratory movement of adult harlequin bugs Murgantia histrionica (Hemiptera: Pentatomidae) as affected by sex, age and host plant quality. Am. Midl. Nat. 118: 189–197. Fahey, J. W., A. T. Zalcmann, P. Talalay, and J. W. Fahey. 2002. The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry. 59. Farnham, M. W., P. E. Wilson, K. K. Stephenson, and J. W. Fahey. 2004. Genetic and environmental effects on glucosinolate content and chemoprotective potency of broccoli. Plant Breeding 123(1): 60-65. Francis, J. A. and S. J. Vavrus. 2012. Evidence linking Arctic amplification to extreme weather in mid‐latitudes. Geophys. Res. Lett. 39: L06801. Frauenfeld, O. W. and R. E. Davis. 2003. Northern Hemisphere circumpolar vortex trends and climate change implications. J. Geophysical Res.: Atmospheres 108(D14): 4423. Froeschner, R. C. 1988. Family Pentatomidae Leach, 1815. The stink bugs, pp. 544–597. In T. J. Henry and R. C. Froeschner (Eds.), Catalog of the Heteroptera, or true bugs, of Canada and the continental United States. E. J. Brill, New York, New York. 958 pp. Gaines, J. C., and H. A. Dean. 1948. Comparison of insecticides for control of harlequin bugs. J. Econ. Entomol. 41: 808–809. Hodson, A. C., and E. F. Cook. 1960. Long-range aerial transport of the harlequin bug and the greenbug into Minnesota. J. Econ. Entomol. 53: 604–608.

- 12 -

Hofmaster, R. N. 1959. Effectiveness of new insecticides against the harlequin cabbage bug on collards. J. Econ. Entomol. 52: 777–778.

Joseph, S. V., J. C. Bergh, S. E. Wright, and T. C. Leskey. 2013. Factors affecting captures of brown marmorated stink bug, Halyomorpha halys (Hemiptera: Pentatomidae), in baited pyramid traps. J. Entomol. Sci. 48(1): 43-51. Khrimian, A., A. Zhang, D. C. Weber, H.-Y. Ho, J. R. Aldrich, K. E. Vermillion, M. A. Siegler, S. Shirali, F. Guzman, and T. C. Leskey. 2014a. Discovery of the aggregation pheromone of the brown marmorated stink bug (Halyomorpha halys) through the creation of stereoisomeric libraries of 1-bisabolen-3-ols. J. Nat. Prod. 77: 1708–1717. Khrimian, A., S. Shirali, K. E. Vermillion, M. A. Siegler, F. Guzman, K. Chauhan, J. R. Aldrich, and D. C. Weber. 2014b. Determination of the stereochemistry of the aggregation pheromone of harlequin bug, Murgantia histrionica. J. Chem. Ecol. 40:1260– 1268.

Kocin, P. J., A. D. Weiss, and J. J. Wagner. 1988. The great arctic outbreak and east coast blizzard of February 1899. Weather Forecast 3: 305–318.

Kogan, Marcos. 1998. Integrated pest management: historical perspectives and contemporary developments. Ann. Rev. Entomol. 43.1 243-270. Kuhar, T. P., Short, B. D., Krawczyk, G., & Leskey, T. C. 2017. Deltamethrin-incorporated nets as an integrated pest management tool for the invasive Halyomorpha halys (Hemiptera: Pentatomidae). J. Econ. Entomol.110.2: 543-545. Lancaster, J., A. Khrimian, S. Young, B. Lehner, K. Luc, A. Wallingford, S. K. B. Ghosh, P. Zerbe, A. Muchlinski, P. E. Marek, M. E. Sparks, J. G. Tokuhisa, C. Tittiger, T. G. Köllner, D.C. Weber, D. E. Gundersen-Rindal, T.P. Kuahr, and D. Tholl. 2018. De novo formation of an aggregation pheromone precursor by an isoprenyl diphosphate synthase-related terpene synthase in the harlequin bug. Proc. Natl. Acad. Sci. 115 (37) E8634-E8641; https://doi.org/10.1073/pnas.1800008115 Leskey, T. C., and H. W. Hogmire. 2005. Monitoring stink bugs (Hemiptera: Pentatomidae) in Mid-Atlantic apple and peach orchards. J. Econ. Entomol. 98: 143-153. Leskey, T. C., S. E. Wright., B. D. Short and A. Khrimian. 2012b. Development of behaviorally based monitoring tools for the brown marmorated stink bug, Halyomorpha halys (Stål) (Heteroptera: Pentatomidae) in commercial tree fruit orchards. J. Entomol. Sci. 47: 76–85.

Li, X., M. A. Schuler, and M. R. Berenbaum. 2007. Molecular mechanisms of metabolic resistance to synthetic and natural xenobiotics. Annu. Rev. Entomol. 52: 231-253. Limpasuvan, V., D. W. Thompson, and D. L. Hartmann. 2004. The life cycle of the Northern Hemisphere sudden stratospheric warmings. J. Climate 17: 2584-2596.

- 13 -

Ludwig, S. W. and L. T. Kok. 1998. Evaluation of trap crops to manage harlequin bugs, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae) on broccoli. Crop Protection 17: 123–128. Michaud, J. P., and A. K. Grant. 2003. IPM-compatibility of foliar insecticides for citrus: indices derived from toxicity to beneficial insects from four orders. J. Insect Sci. 3.1. McBrien, H. L., J. G. Millar, L. Gottlieb, X. Chen, and R. E. Rice. 2001. Male-produced sex attractant pheromone of the green stink bug, Acrosternum hilare (Say). J. Chem. Ecol. 27: 1821–1839. Mizell, R. F., and W. L. Tedders. 1995. A new monitoring method for detection of the stinkbug complex in pecan orchards. Proc. Southeast. Pecan Growers Assoc. Vol. 88. Mizell, R. F., H. C. Ellis, and W. L. Tedders. 1996. Traps to monitor stink bugs and pecan weevils. Pecan Grower 7 (1996): 17-20. Morrison III, W. R., J. P. Cullum, and T. C. Leskey. 2015. Evaluation of trap designs and deployment strategies for capturing Halyomorpha halys (Hemiptera: Pentatomidae). J. Econ. Entomol. 108(4): 1683-1692. Morrison III, W. R., D. Lee, B. D. Short, and A. Khrimian. 2016. Establishing the behavioral basis for an attract-and-kill strategy to manage the invasive Halyomorpha halys in apple orchards. J. Pest. Sci. 89: 81–96. Morrison III, W. R., Allen, M., & Leskey, T. C. 2018. Behavioural response of the invasive Halyomorpha halys (Hemiptera: Pentatomidae) to host plant stimuli augmented with semiochemicals in the field. Agricultural and Forest Entomology, 20(1), 62-72. Paddock, F. B. 1915. The harlequin cabbage-bug. Texas Agricultural Experiment Station Bulletin 179: 1-9. Paddock, F. B. 1918. Studies on the harlequin bug. Texas Agricultural Experiment Station Bulletin 227: 1-65. Panizzi, A.R., and T. Lucini. 2017. Host Plant-Stink Bug (Pentatomidae) Relationships, pp. 32-58. In Čokl, A., and Borges, M. (eds.), Stink Bugs: Biorational Control Based on Communication Processes. Taylor and Francis Group, Boca Raton, FL. Rider, D. R. 2012. The Heteroptera (Hemiptera) of North Dakota 1: Pentatomomorpha: Pentatomoidea. The Great Lakes Entomologist 45: 312–380. Rogers, C. E., and G. R. Howell. 1972. Toxicity of acephate and diazinon to harlequin bugs on cabbage. J. Econ. Entomol. 66: 827–828. Sanderson, D. E. 1908. The influence of minimum temperatures in limiting the northern distribution of insects. J. Econ. Entomol. 1: 245-262. Sinclair, B. J., A. Addo-Bediako, and S. L. Chown. 2003. Climatic variability and the evolution of insect freeze tolerance. Biol. Rev. 78: 181–195.

- 14 -

Stern, V. M., R. Smith, R. Van den Bosch, and K. Hagen. 1959. The integration of chemical and biological control of the spotted alfalfa aphid: the integrated control concept. Hilgardia, 29(2), 81-101. Sur, R., and A. Stork. 2003. Uptake, translocation and metabolism of imidacloprid in plants. Bulletin of Insectology 56: 35–40. Tedders, W. L., and B. W. Wood. 1994. "A new technique for monitoring pecan weevil emergence (Coleoptera: Curculionidae)." J. Entomol. Sci. 29.1 (1994): 18-30. Thomas, W. A. 1915. The cabbage harlequin bug or calico bug. (Murgantia histrionica, Hahn). South Carolina Agricultural Experiment Station Bulletin 28: 1–3. Thrift, Emma M., M.V. Herlihy, A. K. Wallingford, and D. C. Weber. 2018. Fooling the harlequin bug (Hemiptera: Pentatomidae) using synthetic volatiles to alter host plant choice. Environ. Entomol. 47(2): 432-439. doi: 10.1093/ee/nvy013. Tomizawa, M., and J. E. Casida. 2005. Neonicotinoid insecticide toxicology: mechanisms of selective action. Annu. Rev. Pharmacol.45: 247–268. Walgenbach, J. F., and S. C. Schoof. 2005. Insect control on cabbage, 2004. Arthropod Management Tests 30: E14.Walker, H. G., and L. D. Anderson. 1933. Report on the control of the harlequin bug, Murgantia histrionica Hahn, with notes on the severity of an outbreak of this insect in 1932. J. Econ. Entomol.26: 129–135. Walker, H. G., and L. D. Anderson. 1939. Further notes on the control of the harlequin bug. J. Econ. Entomol. 32: 225–228. Wallingford, A. K., T. P. Kuhar, P. B. Schultz, and J. H. Freeman. 2011. Harlequin bug biology and pest management in brassicaceous crops. J. Integr. Pest Manag. 2(1): 1–4. Wallingford, A. K., T. P. Kuhar, and P. B. Schultz. 2012. Toxicity and field efficacy of four neonicotinoids on harlequin bug (Hemiptera: Pentatomidae). Fla. Entomol. 95: 1123– 1126. Wallingford, A. K., T. P. Kuhar, D. G. Pfeiffer, D. B. Tholl, J. H. Freeman, H. B. Doughty, and P. B. Schultz. 2013. Host plant preference of harlequin bug (Hemiptera: Pentatomidae), and evaluation of a trap cropping strategy for its control in collard. J. Econ. Entomol. 106: 283–288. Wallingford, A. K., T. P. Kuhar, D. C. Weber. 2018. Avoiding unwanted vicinity effects with attract-and-kill tactics for harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae). J. Econ. Entomol. toy109 Weber, D. C., T. C. Leskey, G. Cabrera Walsh, and A. Khrimian. 2014a. Synergy of aggregation pheromone with methyl (E,E,Z)-2,4,6-decatrienoate in attraction of Halyomorpha halys (Hemiptera: Pentatomidae). J. Econ. Entomol. 107: 1061–1068. Weber, D. C., G. Cabrera Walsh, A. S. DiMeglio, M. M. Athanas, T. C. Leskey, and A. Khrimian. 2014b. Attractiveness of harlequin bug, Murgantia histrionica (Hemiptera: Pentatomidae), aggregation pheromone: field response to isomers, ratios and dose. J. Chem. Ecol. 40: 1251–1259

- 15 -

Weber, D.C., W.R. Morrison, A. Khrimian, K.B. Rice, T.C. Leskey, C. Rodriguez-Saona, A.L. Nielsen, and B.R. Blaauw. 2017. Chemical ecology of Halyomorpha halys: discoveries and applications. J. Pest Sci. 90.4: 989-1008. Weber, D. C., A. Khrimian, M.C. Blassioli-Moraes & J.G., Millar 2018. Semiochemistry of Pentatomoidea. Pp. 677-725 in: Invasive Stink Bugs and Related Species (Pentatomoidea): Biology, Higher Systematics, Semiochemistry, and Management, ed. by J. E. McPherson. CRC Press, Boca Raton, Florida.

White, W. H., and L. W. Brannon. 1939. The harlequin bug and its control. United States Department of Agriculture Farmers’ Bulletin 1712: 1–10. Wittstock, U., D. J. Kliebenstein, V. Lambrix, M. Reichelt, and J. Gershenzon. 2003. Chapter five: Glucosinolate hydrolysis and its impact on generalist and specialist insect herbivores. Recent Adv. Phytochem. 37: 101–125. Zahn, D. K., J. A. Moreira, and J. G. Millar. 2008. Identification, synthesis, and bioassay of a male-specific aggregation pheromone from the harlequin bug, Murgantia histrionica. J. Chem. Ecol. 34: 238–251.

Zahn, D. K., J. A. Moreira, and J. G. Millar. 2012. Erratum to: Identification, synthesis, and bioassay of a male-specific aggregation pheromone from the harlequin bug, Murgantia histrionica. J. Chem. Ecol. 38: 126.

- 16 -

Chapter Two

Supercooling Points of Murgantia histrionica (Hemiptera: Pentatomidae) and Field

Mortality in the Mid-Atlantic United States following Lethal Low Temperatures

(As Published: Environ Entomol. 2016 Oct;45(5):1294-1299. DOI: 10.1093/ee/nvw091)

Abstract

The harlequin bug, Murgantia histrionica (Hahn), is a serious pest of brassicaceous vegetables in southern North America. While this insect is limited in its northern range of North

America, presumably by severe cold winter temperatures, specific information on its cold hardiness remains unknown. We determined the supercooling points (SCPs) for Maryland and

Virginia adult populations and found no significant difference among these populations. SCPs

o o were similar for adults (푋̅ = -10.35 C; 휎푋 = 2.54) and early and late instar (푋̅ = -11.00 C; 휎푋 =

st o 4.92) and between adult males and females. However, SCPs for 1 instars (푋̅ = -21.56 C; 휎푋 =

o 1.47) and eggs (푋̅ = -23.24 C; 휎푋 = 1.00) were significantly lower. We also evaluated field survival of overwintering harlequin bug adults during extreme cold episodes of January 2014 and

January 2015, which produced widespread air temperatures lower than -15oC and sub-freezing soil temperatures in the Mid-Atlantic Region. After the 48h episode in 2014, bug mortality in exposed field sites averaged 88%, compared to <5% mortality of bugs sheltered in an unheated greenhouse (recorded minimum temperature 4.4oC). After the 2015 episode, approximately 80% of adults that were established in the field the previous November and then sheltered in an unheated garage during the episode, died, in contrast to 96% mortality in exposed field sites. Our results provide new information on M. histrionica overwintering biology, and thermal limitations to its distribution, which leads to improved predictive capabilities to forecast pest severity.

- 17 -

Introduction

Murgantia histrionica (Hahn), native to Central America (Paddock 1918), has expanded widely into the United States (US) since it was first collected in Texas in 1864 (Paddock 1915).

With two to three generations per year (Wallingford et al. 2011), this key pest of Brassica spp. can quickly infest farms, damaging vegetables such as cabbage, broccoli, collards, radishes, and kale. Occasional extreme winter events have historically suppressed its distribution in northern latitudes just south of the Mason-Dixon Line (Walker and Anderson 1933). A combination of physiological and behavioral adaptations allow for overwintering insects, like M. histrionica, to survive a range of winter extremes common in temperate regions (Duman et al. 1991). As

Vennette et al. (2010) explained, understanding implications of insect cold-hardiness can help us better understand pest severity and outbreaks and forecast geographic distribution and potential range expansions.

Extreme winter weather events, such as sudden shifts in Arctic air currents, can have significant effects on temperate insect pest abundance and distributions (Bale and Hayward 2010).

A record-breaking cold air mass blanketed much of the US with a high pressure system propagating from the Arctic in mid-February 1899 (Kocin et al. 1988). Sanderson (1908) reported significant reductions in M. histrionica populations following this bitterly cold Arctic air mass, when minimum temperatures reached -28.9oC in Ohio and -26.1oC in Washington D.C.

(Sanderson 1908), and -15oC in the Carolinas (Kocin et al. 1988). Although field mortality of M. histrionica was not formally measured before and after this winter event, the 1899 North

American winter serves as anecdotal evidence of the negative impact of extreme low temperatures on M. histrionica abundance.

- 18 -

Events of prolonged sub-freezing temperatures may occur more frequently in the southeastern US than in the past decades due to altered atmospheric thermal gradients between northern and mid-latitudes (Franuenfeld and Davis 2003, Limpasuvan et al. 2004, Francis and

Vavurs 2012). Therefore, it is important to understand how freeze-avoiding insects are affected by potentially lethal lower temperatures (Elsey 1993), especially for temperate insects in a changing climate (Bradshaw 2010).

Saulich and Musolin (2012) reviewed literature and experimental data on diapause in 43 pentatomid species overwintering in temperate climates, and concluded that the majority of species undergo facultative imaginal diapause—that is, extrinsic cues such as temperature and photoperiod trigger arrested development in adults. Unfortunately, this study did not include M. histrionica, Murgantia histrionica undergo homodynamic development, and therefore activity during winter months is primarily dictated by temperature; they are active in mild winters

(Thomas 1915, Paddock 1915, 1918) with feeding, copulation, and oviposition possible (Sullivan and Brett 1974).

Diapause and categories of cold-hardiness occur conjointly. Salt (1961) and Lee (1991) divided cold-hardiness of insects into three groups: chilling intolerant, freezing tolerant, and freeze avoidant. According to Sinclair et al. (2003), most temperate insects in the Northern

Hemisphere are freeze intolerant. Internal ice formation is lethal to this group of insects; thus, they are often classified as freeze avoidant (Lee 1991). Hemolymph will instantaneously crystallize below certain temperature thresholds; the temperature at which this occurs is referred to as the supercooling point (Baust and Rojas 1985).

Supercooling points vary in insects according to body size, water content, dissolved solutes such as sugars and ions, and composition of amino acid and proteins in the hemolymph

- 19 -

(Lee 2010). The supercooling point is conventionally measured thermoelectrically to detect the exotherm resulting from the latent heat of fusion. Measuring supercooling points for freeze avoiding insects yields a concrete value for beginning to understand cold hardiness (Carrillo et al.

2004).

Elsey (1993), without providing the recorded range, approximated a supercooling point of

Nezara viridula (L.) (Hemiptera: Pentatomidae) at -11oC with thermocouples in direct contact with their abdomen, insulated by a polystyrene box and held in a -20oC reach-in freezer. There was no difference of supercooling points between males and females, nor between diapausing and non-diapausing adults. However, the author did note that N. viridula had a much higher SCP than two other pentatomids, M. histrionica at -14.7oC and Euschistus servus (Say) at -15.0oC.

Unfortunately, since E. servus and M. histrionica SCPs were sourced from unpublished results, and the important details (e.g. sample size, source population, diet, and environmental preconditioning) are unknown.

Our research investigated supercooling points of M. histrionica, by life stage and sex, as well as field mortality during extreme cold weather episodes in the Mid-Atlantic U.S. If M. histrionica are truly freeze intolerant, the distribution of supercooling points within a sample population will correspond to observed field mortality during extreme cold weather episodes. We assayed field mortality in wild populations of M. histrionica in Beltsville, MD and Blacksburg,

VA during extreme winter weather events of January 2014 and January 2015, and related the observed temperatures and population impacts to SCPs determined in the lab. Egg SCPs were determined thermoelectrically, which we then related to hatch rates after exposure to subfreezing temperatures in an effort to determine a range of lethal lower temperatures for the species.

- 20 -

Materials and Methods

Polar Vortex Climate Observations

The extreme cold “polar vortex” episodes in Beltsville, MD and Blacksburg, VA were documented using climate monitoring instruments (Decagon Devices, Pullman, WA) positioned within overwintering habitats. For both 2014 and 2015, microclimate sensors were installed to record temperature at 1cm soil depth, and air within the plant canopy (0.05m) and above (1m) the

M. histrionica overwintering habitat. After “polar vortex” episodes, both control and exposed bugs were monitored for 3d at ambient temperature (20 – 25oC), to visually characterize external ice formation on insect cuticle, and to record activity and mortality.

Insects

SCP determination

M. histrionica adults were collected in early September 2013 in Beltsville, MD (39°

1'37.37"N, 76°56'0.96"W; Elevation = 38m)) and were used to start a colony, which was maintained on potted collards (Brassica oleracea group acephala cv. ‘Champion’ and ‘VATES’) at 25oC 16:8 L:D. Only bugs from generations F1-F5 were used in these experiments. Bugs were carefully isolated from colonies in ventilated plastic containers with an excised collard leaf, and held at ambient temperature (20 -25oC) up to 5 hours prior to SCP determination. Eggs laid in these colonies within the previous 24h were shipped overnight to Geneva, NY (42o52’38.40”N,

77o00’26.14”W; Elevation = 188m) where egg clusters and first instars that hatched after arrival were assayed for SCPs. An additional field-collected population in early April 2014 from Painter,

VA (37o35’04.61”N, 75o49’15.64”W; Elevation = 9m) was tested with individuals acclimated

(25oC and 16:8 L:D) for at least 14d following the same procedures.

- 21 -

2014 field survival

In late September 2013, wild M. histrionica adults collected on summer cultivated collards were isolated in aluminum mesh cages used by Cabrera Walsh et al. (2016) over patches of volunteer cultivated Brassica species. The habitat also included wild grasses and weeds, which provided overwintering refuges. Bug densities per cage did not exceed 500 individuals per square meter. Three days prior to the extreme cold episode, caged bugs were recovered and held at approximately 10°C in an unheated greenhouse for 24h to segregate active versus inactive, presumably dead, adults. Active adults were then returned to the field 20 hours prior to the extreme cold episode to allow bugs to move unhindered in overwintering habitat. A sample of 75 field collected adults were exposed to extreme winter weather confined within 0.15 m2, and, as a control, 50 adults from the same population were sheltered in an unheated glasshouse where minimum temperatures reached 4.4oC and never exceeded 15.5oC.

2015 field survival

Randomly-selected M. histrionica adults from summer cultivated collards were isolated in

November 2014 from Kentland Farm, Blacksburg, VA (37o12’06.73”N 80o33’52.27”W;

Elevation = 517m) in groups of 10 individuals on large collard plants (leaves radiating out 20 cm from the stem). Twelve isolated groups were randomly arranged in the collard plot. Bugs were confined on plants with 18.9 liter paint strainer bags (Trimaco, Morrisville, NC) and restricted with a plastic zip tie at the base of the plant, to allow bugs to naturally move unhindered to all portions of the plant.

In the afternoon, prior to the January 2015 “polar vortex” episode, 6 of the 12 groups of confined adults were randomly selected and collected by severing the base of the collard plant

- 22 - below the mesh bag, and transported to an unheated garage to shelter them from extreme subfreezing temperatures (indoor temperatures ranged from 1.4oC to 7.0oC). The other six groups remained on plants in the field to serve as our treatment group.

Egg hatch rate following cold treatments

Insect cultures were reared from field collected adults (sourced early Sept. 2015,

Blacksburg, VA) on potted collards (B. oleracea var. VATES or Champion) in summer conditions in a climate controlled greenhouse (25oC - 30oC, providing supplemental light at 16:8

L:D). Eggs from the parental generation were collected 24 – 48 hours after laying, and then held at ambient temperatures (20oC – 25oC) before conducting experiments described below.

Lethal low temperature determination methods

SCPs of adults and early and late instars

Supercooling points of adults and nymphs were measured with braided coiled copper- constantan thermocouples (Hanson and Venette 2013) on the ventral abdominal segments, with the syringe plunger gently immobilizing the bug. Thermocouples were attached to a multichannel data logger (USB-TC, Measurement Computing, Norton, MA) recording at 0.5Hz and logged by using Tracer-DAQ software (Measurement Computing, Norton, MA). Individual supercooling points were determined as the minimum temperature reached before the dramatic increase in temperature, which is associated with the exotherm from latent heat of fusion.

In 2013, we modified a protocol developed by Carrillo et al. (2004), where individual bugs and braided thermocouple were confined in a 35ml syringe that was placed at the center of a 0.19

× 0.19 × 0.19 m expanded polystyrene cube (Foam-Control TM Type IX, Cellofoam North

America Inc., Winchester, VA) and then into a -80°C freezer for a projected cooling rate of

- 23 -

-0.5oC min-1, with a realized cooling rate of -0.75 +/- 0.35oC min-1. Each assay was run until internal temperature of the cube reached -30oC.

One hour prior to SCP measurements for adults and 2nd through 5th instars, bugs were weighed to the nearest μg (Ohaus, Explorer Pro, Parsippany, NJ). After SCP determination, each bug was stored at -25oC until subsequent oven drying. Bugs were dried at 65oC for 48hr and weighed to determine water content.

SCP of eggs and first instars

Supercooling points for egg masses and groups of recently hatched M. histrionica were estimated using a method modified from Mills et al. (2006) to detect the exotherms from spontaneous freezing. Unhatched egg masses (36±12hr since oviposition to represent our egg census) or recently hatched masses (approximately 24hr to represent our first instars census) were moved to thermal electric modules (TEM), trays of 10 wells holding Peltier plates (4 × 4 cm). The

TEMs were then moved to a programmable freezer (Thermal Product Solutions, Williamsport,

PA) and held 0°C for 1hrh, to bring plate temperature down from room temperature. The temperature was then reduced at a rate of 0.5°C min-1 to -40°C and then held at -40°C for 1hr to ensure all plates reached this lower temperature. Peltier plates detect temperature gradients on either side of the plate and convert the thermal signals to voltage outputs recorded by a Keithley

Multimeter Data Acquisition System (model 2700-DAQ-40; Keithley Instruments, Cleveland,

OH) at the same time that temperature is recorded at the plate surface by thermistors (model

44212; YSI, Dayton, OH) in one of the wells of each TEM. A visible peak in voltage (mV) graphed over time indicated an exotherm and the corresponding temperature recorded at the plate surface indicates the temperature at which ice formation occurred. The output from at least one empty well in each TEM served as a control to compare voltage amplitude.

- 24 -

Egg hatches after cold treatments

We selected three subfreezing temperatures of -5, -16 and -23oC to test for egg mortality under simulated field conditions (i.e. rates of cooling and warming mirrored trends observed during winter weather anomalies of 2014 and 2015). Temperatures were selected based on observed M. histrionica egg SCP of -23oC, thermal minimum during extreme cold episodes (-

16oC), and typical winter low temperatures (-5oC) in Mid-Atlantic US. Egg masses up to 48hr old were obtained from adult M. histrionica that were collected in early fall from collards near

Blacksburg, VA and held in a greenhouse under conditions mentioned above.

Eggs were subjected to sub-freezing temperatures in 5mL centrifuge tubes

(MacroTubes™, Orange County, CA) sealed to exclude liquid with Parafilm (Bemis NA, Neenah,

WI). Sealed tubes with egg masses were submerged in 50/50 water and ethanol mixture in a refrigerated water bath (Fisher Scientific Isotemp, Waltham, MA) where the temperature declined at a rate of -0.16oC min-1, holding at treatment temperature for 30 min before returning at a rate of

2.00oC min-1 temperature to 25oC. Control eggs (n = 171) were held in Parafilm sealed centrifuge tubes at ambient temperature (20oC – 25oC). After each cold treatment, all treatment and control egg masses were isolated in Petri dishes with filter paper in a growth chamber held at 28oC at 50-

80% RH, 16:8 L:D. Successful egg hatches were tallied by counting the number of nymphs after

20 days; Abbott’s formula was used to correct for control mortality (Abbott 1925).

- 25 -

Statistics

Data were analyzed with one way ANOVAs, Student’s t-test, Tukey-Kramer HSD, and

Fisher's exact test, using R version 3.2.2 (R Core Team 2015) in RStudio version 0.99.484

(RStudio 2015), and analyzed with an α value set at 0.05.

SCP comparisons

The SCPs for adult bugs collected from Painter, VA and Beltsville, MD were inspected for normality using histogram analysis to measure skewness and kurtosis, followed by a Levene

Test for homogeneity of variance among locations and sex. We assumed Gaussian processes without violations within a histogram analysis (skewness = 0.422 and 1.21, and kurtosis = 1.93 and 4.01 for Painter and Beltsville, respectively) and variance between groups was homogenous

(F = 0.2055; df = 3, 33; p = 0.8919).

Predicted supercooling point distributions

The R package fitdistrplus (Delignette-Muller et al. 2015) was used to model adult supercooling point distributions. Within this package, a Cullen-and-Frey Graph was used to approximate theoretical distribution using observed supercooling points. Based on Akaike information criterion, a gamma distribution model was used to predict cumulative supercooling points.

- 26 -

Results and Discussion

Field Survival

Maryland 2014:

Soil surface temperatures where the exposed adults were found below plant litter reached a minimum of -3.9°C during the episode, whereas air temperature reached a minimum of -15.8°C

(Fig. 1A). Because of this thermal gradient, exposed bugs likely experienced a range of sub- freezing temperatures, and cooling rates. Nonetheless, these conditions were lethal to M. histrionica. Compared to sheltered group held in a small greenhouse (recorded minimum temperature 4.4oC), the mortality in the exposed group was significant (Fisher's Exact Test, p<0.0001) with 66 of 75 (88%) compared to 2 of 50 (4%) mortality. However, presence of ice crystals on 8 January 2014 was significantly correlated with mortality (Fisher's Exact Test, p = 0.00089); only 2% of bugs (1 of 48) survived if surrounded by ice, but almost 30% (8 of 27) of those not showing external ice were still alive several days after the cold-temperature episode.

Virginia 2015:

In the field, soil surface temperatures reached a minimum of -6.2°C and air temperature rapidly fell to a minimum of -16.0°C (Fig. 1B). Control bugs sheltered in an unheated garage were exposed to a minimum ambient temperature of 1.4°C during the event. Field exposed bugs experienced high mortality at 26 of 27 (~96%), but according to Fisher's exact test, p = 0.0620, field mortality was not significantly higher than the control group mortality at 23 of 29 (~79%) dead after the cold weather event. However, since these bugs were isolated in the field in early

November 2014, cumulative mortality is unlikely related to the 2015 weather event we assayed,

- 27 - but more likely related to repeated cold weather events because the they were exposed to winter conditions longer than the 2014 outdoor group.

Supercooling Point Comparison and Egg Hatch Rates

Supercooling points in a small number of our observations were at least two standard deviations from than their means, most likely due to fecal or other material with high water content; accordingly we censored six outlying observations ranging from -2.15oC to -3.74oC just as Cira et al. (2016) eliminated similar outliers in their study of Halyomorpha halys (Stål).

A Student’s t-test revealed no significant effect of location (t = 1.0876; df = 35; p =

0.2842) and, therefore, SCP of adults from Beltsville, MD, and Painter, VA, were pooled as one group labeled “adults” in all further analyses. SCP values did not correlate with total water content (t = 0.55849; df = 126; p = 0.5775, Pearson’s R = 0.04969314) among adults and 2nd through 5th instars, and, therefore, were not factored in further analysis.

One-way ANOVAs determined significant differences in SCP among life stages (F=155.6; df = 6; p <0.0001), but not sexes (F = 0.743; df = 1; p = 0.395). These results complement our field survival observations, suggesting that lab-measured SCPs may approximate field outcomes.

A post-hoc Tukey-Kramer HSD showed significant differences among life stages (Fig. 2). The

o o o SCPs of adults ranged from -6.85 C to -18.34 C (푋̅ = -10.35 C; 휎푋 = 2.54) and was similar to

o o o values recorded from later instar nymphs (-6.09 C to -21.14 C (푋̅ = -11.00 C; 휎푋 = 4.92)).

Supercooling points occurred at lower temperatures for eggs and 24h-old 1st instars than for adults and late instars, and lower for eggs than 1st instars (Fig. 2). Murgantia histrionica 1st instars are non-feeding (Zahn et al. 2008, Wallingford et al. 2011), and therefore, may not possess enriched levels of ice nucleating agents as later life stages may since some nucleating agents may

- 28 - be present from food sources (Sømme 1982). First instar N. viridula absorb atmospheric water, aggregating on their egg mass to resist desiccation, before molting (Hirose 2006); M. histrionica may be similarly adapted. Water absorption may not have an impact on nymph supercooling points, since total water content and SCPs did not correlate in later nymphs and adults.

When M. histrionica eggs masses were subjected to sub-freezing conditions, egg mortality

(Fig. 3) reflected egg SCP observations, indicating that eggs are likely more cold-hardy than other life stages. Egg cold-hardiness is not uncommon in insects. For instance, Rosales et al. (1994) found similar dramatic differences among life stages of Musca autumnalis De Geer (Diptera:

Muscidae), eggs being especially cold hardy. Aphid (Homoptera: Aphididae) eggs are widely known to be cold hardy (Strathdee et al. 1995). In addition, Olson et al. (2013) reported SCPs in

Cimex lectularius L. (Hemiptera: Cimicidae) with eggs having the lowest value.

Most reports (Thomas 1915, Paddock 1915, 1918) indicate that adults are the most common stage to overwinter. However, late fall reproductive activity is not uncommon in M. histrionica (Sullivan and Brett 1974), therefore sudden extreme cold weather episodes in the late fall is likely to negatively impact the population. M. histrionica egg age and maternal life history may cause seasonal differences in egg SCP and survivorship. Regardless, the significantly lower

SCPs of M. histrionica eggs and first instar nymphs suggests that during mild winters they are more likely to survive and develop into adults, which can exacerbate pest populations during the following growing seasons.

Cira et al. (2016) found a range of supercooling points in H. halys, and documented chill- intolerance as the result of significant differences in mortality at temperatures above their supercooling points. This pentatomid has one to two generations per year in the Mid-Atlantic

States, and undergoes reproductive diapause before overwintering (Rice et al. 2014). As a

- 29 - survival strategy, the bugs seek anthropogenic structures, fallen and standing dead trees and cliff outcroppings to remain buffered against extreme temperature fluctuations during winter months

(Lee et al. 2014).

Murgantia histrionica do not seek structures that protect them against subfreezing temperatures as H. halys do. Instead we regularly observe M. histrionica overwintering nearby host plant communities at the soil surface blanketed with leaf litter and crop residue, or within winter cover crops. Because of their overwintering habitats, supercooling points in M. histrionica were related to field mortality during extreme cold weather events. Results from our 2014 field winter mortality observations (88% mortality at -15.8oC) are consistent with the modeled M. histrionica cumulative SCP distribution (Fig. 4) indicating that SCPs have predictive value for mortality in adults, and presumably early and late stage nymphs since their mean SCP are statistically similar to adults.

Our investigations were limited to Beltsville, MD, and Painter, VA, populations; geographically distinct phenotypes in other climates should be investigated and compared to understand phenotypic plasticity that may accelerate or restrict M. histrionica geographic range and abundance (Sgrò et al. 2015). A recent invading pentatomid in the Southwestern US, Bagrada hilaris (Burmeister), is a closely-related (Pentatomidae: Pentatominae: Strachiini) brassica specialist native to temperate regions of Africa, India, and the Middle East. This species has similar overwintering biology (Singh and Malik 1993) to M. histrionica and could be another noteworthy pentatomid to study in comparison with M. histrionica; B. hilaris range in the

Southwestern US is expanding both eastward and northward (Reed et al. 2013).

Future work on M. histrionica low temperature biology should continue to combine lab and field studies, aimed at investigating the plastic responses that drive genetic variations within

- 30 - fluctuating climates (Sgrò et al. 2015). Since M. histrionica quiescence is dependent on extrinsic cues, fluctuations in temperature during fall and winter months may have an effect on survival even when temperatures are above physiological lower limits described in this paper. Studies investigating specific seasonal acclimation periods, factors related to population origins, and heat- stress effects on winter survival, are needed to better knowledge of microclimates in overwintering habitats of harlequin bug of M. histrionica thermal tolerance. Low-molecular weight sugars and other cryoprotectants such as glycerols and inositol were not measured in this study. Seasonal changes in water content and cryoprotectants accumulations in winter versus summer populations may lead to a better understanding of M. histrionica cold hardiness and yield better understanding on the specific physiological factors that improve overwintering survival, such as documented in Harmonia axyridis Pallas (Watanabe 2002).

Global climate change is expected to accelerate in the coming decades (Smith et al. 2015), leading to dramatic fluctuation in regional climates. As Limpasuvan et al. (2004) discuss, stratospheric anomalies impose the greatest amount of energy on the circumpolar vortex, which are factors for the prolonged “polar-vortex” episodes we observed in 2014 and 2015 (Frauenfeld and Davis 2003, Francis and Vavrus 2012). Our study related a climate-driven weather anomaly to field survival and low temperature biology laboratory studies in M. histrionica. Future research is needed to understand how to build well suited pest-prediction models to adjust for altered climate patterns (Bale and Hayward 2010).

- 31 -

Acknowledgements

The authors are extremely grateful for the numerous dedicated individuals that made this project possible, especially colleagues at USDA-ARS Invasive Insect Biocontrol and Behavior

Laboratory who shared valuable -80oC freezer space for our supercooling point measurements.

We appreciate Robert Venette (University of Minnesota) for his advice in adapting a supercooling point determination protocol for M. histrionica, as well as Brennan Bathauer for enduring extreme cold temperatures during our field studies, and Madeleine Kahle, Sandra Galbert, and Hallie

Harriman for providing technical support for our lab tests. Michael Richard Greason (NASA) was key in simulating and computing theoretical cooling rates for our supercooling point determination set up. We thank Lauris McQuoid-Greason for reviewing this paper before submission.

- 32 -

References

Abbott, W. S. 1925. A method of computing the effectiveness of an insecticide. J. Econ. Entomol. 18: 265 – 267. Bale, J. S., and S.A.L. Hayward. 2010. Insect overwintering in a changing climate. J. Exp. Biol. 213: 980–94. Baust, J. G., and R. R. Rojas. 1985. Review—Insect cold hardiness: Facts and fancy. J. Insect Physiol. 31: 755–759. Bradshaw, W. E. 2010. Insects at not so low temperatures: Climate change in the temperate zone and its biotic consequences, pp. 260–262. In Denlinger, D. L., Lee, R. E. (eds.), Low Temperature Biolology of Insects. Cambridge Univeristy Press, New York. Cabrera Walsh, G., A.S. Dimeglio, A. Khrimian, and D.C. Weber. 2016. Marking and retention of harlequin bug, Murgantia histrionica (Hahn)(Hemiptera: Pentatomidae), on pheromone baited and unbaited plants. J. Pest Sci. 89: 21-29. Carrillo, M. A., N. Kaliyan, C. A. Cannon, R. V. Morey, and W. F. Wilcke. 2004. A simple method to adjust cooling rates for supercooling point determination. CryoLetters. 25: 155– 160. Cira, T. M., R. C. Venette, J. Aigner, T. Kuhar, D. E. Mullins, S. E. Gabbert, and W. D. Hutchison. 2016. Cold tolerance of Halyomorpha halys (Hemiptera: Pentatomidae) across geographic and temporal scales. Environ. Entomol. 45: 484–491. Delignette-Muller, M. L., C. Dutang, R. Pouillot, and J. B. Denis. 2015. fitdistrplus: Help to Fit of a Parametric Distribution to Non-Censored or Censored Data. Version 1.0-6. https://cran.r-project.org/web/packages/fitdistrplus/index.html Duman, J. G., D. W. Wu, L. Xu, D. Tursman, and T. M. Olsen. 1991. Adaptations of insects to subzero temperatures. Q. Rev. Biol. 66: 387-410. Elsey, K. D., 1993. Cold tolerance of the southern green stink bug (Heteroptera: Pentatomidae). Environ. Entomol. 22: 567-570. Francis, J. A. and S. J. Vavrus. 2012. Evidence linking Arctic amplification to extreme weather in mid‐latitudes. Geophys. Res. Lett. 39: L06801. Frauenfeld, O. W. and R. E. Davis. 2003. Northern Hemisphere circumpolar vortex trends and climate change implications. J. Geophysical Res.: Atmospheres 108(D14): 4423. Hanson, A. A., and R. C. Venette. 2013. Thermocouple design for measuring temperatures of small insects. CryoLetters. 34: 261–266. Hirose, E., A. R. Panizzi, and A. J. Cattelan. 2006. Effect of relative humidity on emergence and on dispersal and regrouping of first instar Nezara viridula (L.)(Hemiptera: Pentatomidae). Neotropical Entomol. 35: 757-761.

- 33 -

Kocin, P. J., A. D. Weiss, and J. J. Wagner. 1988. The great arctic outbreak and east coast blizzard of February 1899. Weather Forecast 3: 305–318. Lee, R. E. 1991. Principles of insect low temperature tolerance, pp. 17–46. In Lee, R.E., Denlinger, D. L. (eds.), Insects at Low Temperature. Chapman and Hall, New York. Lee, R. E. 2010. A primer on insect cold-tolerance, pp. 4–6. In Denlinger, D.L., Lee, R.E. (eds.), Low Temp. Biol. Insects. Cambridge Univeristy Press, New York. Lee, D.-H., J. P. Cullum, J. L. Anderson, J. L. Daugherty, L. M. Beckett, and T. C. Leskey. 2014. Characterization of overwintering sites of the invasive brown marmorated stink bug in natural landscapes using human surveyors and detector canines. PLoS One. 4(9): e91575. Limpasuvan, V., D. W. Thompson, and D. L. Hartmann. 2004. The life cycle of the Northern Hemisphere sudden stratospheric warmings. J. Climate 17: 2584-2596. Mills, L. J., J. C. Ferguson, M. Keller. 2006. Cold-hardiness evaluation of grapevine buds and cane tissues. Am. J. Enol. Vitic. 57: 194-200. Olson, J. F., M. Eaton, S. A. Kells, V. Morin, and C. Wang. 2013. Cold tolerance of bed bugs and practical recommendations for control. J. Econ. Entomol. 106: 2433-2441. Paddock, F. B. 1915. The harlequin cabbage-bug. Texas Agric. Exp. Station Bull. 179: 1–9. Paddock, F. B. 1918. Studies on the harlequin bug. Texas Agric. Exp. Station Bull. 227: 1–65. R CoreTeam 2014. R: A language and environment for statistical computing. Vienna, Austria. (http://www.r-project.org/). Reed, D.A., J.C. Palumbo, T.M. Perring, and C. May. 2013. Bagrada hilaris (Hemiptera: Pentatomidae), an invasive stink bug attacking cole crops in the southwestern United States. J. Integr. Pest Manage. 4: 1-7. Rice, K.B., C.J. Bergh, E.J. Bergmann, D.J. Biddinger, C. Dieckhoff, G. Dively, H. Fraser, T. Gariepy, G. Hamilton, T. Haye, A. Herbert, K. Hoelmer, C.R. Hooks, A. Jones, G. Krawczyk, T. Kuhar, H. Martinson, W. Mitchell, A.L. Nielsen, D.G. Pfeiffer, M.J. Raupp, C. Rodriguez-Saona, P. Shearer, P. Shrewsbury, P.D. Venugopal, J. Whalen, N.G. Wiman, T.C. Leskey, and J. F. Tooker. 2014. Biology, ecology, and management of brown marmorated stink bug (Hemiptera: Pentatomidae). J. Integr. Pest Mgmt. 5: A1- A13. Rosales, A.L., E. S. Krafsur, and Y. Kim. 1994. Cryobiology of the face fly and house fly (Diptera: Muscidae). J. Med. Entomol. 31: 671-680. RStudio. 2014. Integrated development environment for R, Version 0.98.1102. Boston, MA. (http://www.rstudio.org/). Salt, R. W. 1961. Principles of insect cold-hardiness. Annu. Rev. Entomol. 6: 55–74.

- 34 -

Sanderson, D. E. 1908. The influence of minimum temperatures in limiting the northern distribution of insects. J. Econ. Entomol. 1: 245-262. Saulich, A. K. and D. L. Musolin. 2012. Diapause in the seasonal cycle of stink bugs (Heteroptera, Pentatomidae) from the temperate zone. Entomol. Rev. 92: 1-26. Sgrò, C.M., J. S. Terblanche, and A. A. Hoffmann. 2015. What can plasticity contribute to insect responses to climate change?. Annu. Rev. Entomol. 61: 433-451 Sinclair, B. J., A. Addo-Bediako, and S. L. Chown. 2003. Climatic variability and the evolution of insect freeze tolerance. Biol. Rev. 78: 181–195. Singh, H., and V. S. Malik. 1993. Biology of painted bug (Bagrada cruciferarum). Indian J. Agric. Sci. 63: 672-674. Smith, S. J., J. Edmonds, C. A. Hartin, A. Mundra, and K. Calvin. 2015. Near-term acceleration in the rate of temperature change. Nature Clim. Change 5: 333-336. Sømme, L. 1982. Supercooling and winter survival in terrestrial arthropods. Comp. Biochem. and Physiol. Part A: Physiol. 73: 519-543. Strathdee, A. T., G. G. Howling, and J. S. Bale. 1995. Cold hardiness of overwintering aphid eggs. J. Insect Physiol. 41: 653-657. Sullivan, M. J., and C. H. Brett. 1974. Resistance of commercial crucifers to the harlequin bug in the coastal plain of North Carolina. J. Econ. Entomol. 67: 262–264. Thomas, W. A. 1915. The cabbage harlequin or calico bug. Bull. S. C. Agric. Exp. Stn. 28. Venette, R. C., D. J. Kriticos, R. D. Magarey, F. H. Koch, R. H. A. Baker, S. P. Worner, N. N. Gómez Raboteaux, D. W. McKenney, E. J. Dobesberger, D. Yemshanov, P. J. De Barro, W. D. Hutchison, G. Fowler, T. M. Kalaris, and J. Pedlar. 2010. Pest risk maps for invasive alien species: A roadmap for improvement. Bioscience. 60: 349–362. Wallingford, A. K., T. P. Kuhar, P. B. Schultz, and J. H. Freeman. 2011. Harlequin bug biology and pest management in Brassicaceous crops. J. Integr. Pest Mgmt. 2: H1-H4. Walker, H. G. and L. D. Anderson. 1933. Report on the control of the harlequin bug, Murgantia histrionica Hahn, with notes on the severity of an outbreak of this insect in 1932. J. Econ. Entomol. 26: 129-135. Watanabe, M.A. 2002. Cold tolerance and myo-inositol accumulation in overwintering adults of a lady beetle, Harmonia axyridis (Coleoptera: Coccinellidae). Eur. J. Entomol. 99:5-10. Zahn, D.K., R. D. Girling, J. S. McElfresh, R. T. Cardé, and J. G. Millar. 2008. Biology and reproductive behavior of Murgantia histrionica (Heteroptera: Pentatomidae). Ann. Entomol. Soc. Am. 101: 215-228.

- 35 -

Fig. 2.1.

Soil and air temperature readings from 2014 and 2015 “Polar Vortex” episodes occurring in early

January. In both panels dark black lines represent air temperature measurements, whereas light gray indicate soil surface temperatures. All labeled dates (date and month) are delimited at 12am

(0:00hours).

Top panel (A): In 2014 soil surface temperatures where adults were found reached a minimum of

-3.9°C, and air temperature -15.8°C.

Bottom panel (B): In 2015 soil surface temperatures reached a minimum of -6.2°C and air temperature -16.0°C.

- 36 -

Fig. 2.2.

Comparison of supercooling points observed for M. histrionica among life stages. One-way

ANOVA indicates significant differences among life stages (p < 0.001). Means for life stages with same letter do not differ according to Tukey-Kramer HSD. Box plots show mean ±25% and

±40% of sample with outliers as dots.

- 37 -

Fig. 2.3.

Mean egg mortality in M. histrionica after simulated exposures to sub-freezing temperatures. Egg masses laid within 48h introduced by a cooling rate of approximately 0.16oC min-1 to -23°, -16°, and -5oC and held for 30mins, then re-acclimated to 20oC at approximately 2.00oC min-1. Total hatches tallied over a 20d period post cold treatment; mortality was defined by the proportion that did not hatch. Means corrected for control mortality (where 149 of the 171 control eggs successfully hatched) with Abbott’s formula, with standard deviations annotated respectively around the means.

- 38 -

Fig. 2.4

Cumulative distribution of supercooling points observed for M. histrionica adults (black dots), fitted with a predictive curve. Predicted cumulative supercooling points calculated with a gamma distribution (continuous line) based on maximum likelihood parameter estimates with a shape of

19.293612 (SE = 4.4474076), and following a rate of 1.863324 (SE = 0.4351432) with an AIC of

171.1608.

- 39 -

Chapter Three

Color Preference of Harlequin Bug, Murgantia histrionica (Heteroptera: Pentatomidae)

(As published: J Econ Entomol. 2017 Oct 1;110(5):2275-2277. doi: 10.1093/jee/tox179)

Abstract:

Harlequin bug, Murgantia histrionica (Hahn), is an important pest of Brassica crops in the southern United States. Regional populations are highly variable and unpredictable from farm-to- farm, and therefore accurate monitoring of activity would greatly improve IPM decision-making and the timing of control tactics. To our knowledge, there is no monitoring device or proven trapping system for this pest. We contribute new knowledge of harlequin bug visual ecology, which will aid in the development of an effective trap. In both lab and field color choice experiments, harlequin bug adults and large nymphs responded positively to green and black colors, and statistically less frequently to yellow, white, purple or red with the exception of adult females, which were most attracted to red and green in the lab, but green and black in the field.

We conclude that future trapping devices for harlequin bug should be green or black in color.

Keywords: Trap, color preference, visual stimuli, harlequin bug, Brassica, cole crops

- 40 -

Introduction

Harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae), is a gregarious and damaging pest of brassica crops, particularly in the southern U.S. (Wallingford et al. 2011b).

Feeding by adults and nymphs results in conspicuous white blotches on the leaves, which can affect crop quality particularly for leafy brassicas, and high pest densities can seriously damage or kill the host plant (Ludwig and Kok 2001, Wallingford et al. 2011). Local populations of this species can be heavily impacted by extreme winter weather events (DiMeglio et al. 2016) and thus, pest infestation levels on a given farm can be quite variable or unpredictable from year to year. To our knowledge, there is no monitoring device or proven trapping system for this pest; thus, an effective monitoring tool would be useful to assess the presence and abundance of this stink bug to determine the need for and timing of control tactics.

Zahn et al. (2008) reported that harlequin bug males produce a sesquiterpenoid aggregation pheromone termed “murgantiol.” Later Khrimian et al. (2014) synthesized and characterized murgantiol as a two-component pheromone consisting of (3S,6S,7R,10S)- and

(3S,6S,7R,10R)-10,11-epoxy-1-bisabolen-3-ol.. The 8-isomer preparation of (7S)-murgantiol is highly attractive to harlequin bug female and male adults and nymphs, and shows great promise as an attractant for targeted recruitment of dispersing harlequin bugs in the field, especially in combination with a host plant (Weber et al. 2014, Cabrera Walsh et al. 2016).

Compared with olfactory stimuli, considerably less is known about the visual stimuli for harlequin bug. Previous research on stink bug trapping devices predominantly focused on multispecies pentatomid complexes and showed that “industrial safety yellow” was generally most attractive for several stink bug species (Mizell and Tedders 1995, Leskey and Hogmire

2005). More recently, Leskey et al. (2012) evaluated the color of pyramid traps and reported that

- 41 - black was most attractive to the brown marmorated stink bug, Halyomorpha halys Stål

(Hemiptera: Pentatomidae). Joseph (2014) also reported that black was most attractive to Bagrada hilaris (Burmeister) (Hemiptera: Pentatomidae) over purple, red, yellow, and white colors. This research highlights the importance of color for trap devices targeting pestiferous stink bug species.

To our knowledge, this has not been investigated for harlequin bug. In this study, we contribute new knowledge of trap color preference in harlequin bugs to understand response to color in the absence of semiochemical attractants. Our study investigates harlequin bug response to lab arena color choice and natural field response to trap color in a vegetable cropping system.

Materials and Methods

Insect cultures

Field-collected harlequin bugs adults from Painter, VA (37o 35’04.61” N, 75o 49’15.64”

W; elevation = 9 m) in mid-April 2015 initiated a colony for this study. Bugs were reared on potted collards (Brassica oleracea: VATES or Champion) in a greenhouse under summer-like conditions (16:8 L:D and 20 to 30oC). Adults and large nymph (4th through 5th instars) were sourced in our laboratory color choice experiments.

Laboratory Color Choice Assay

Corrugated polypropylene (Coroplast, Inteplast Group, Vanceburg, KY) panels

(dimensions at 0.6m × 0.6m × 4mm thickness) were assembled with clear packing tape into a hexagonal choice-test arena with an interior diameter = 1.42m. Reflectance spectra data was collected using a DH-2000-BAL light source (Ocean Optics Inc., Dunedin, FL), Ocean Optics

SpectraSuite software (Ocean Optics Inc., Dunedin, FL) and a USB4000 Spectrometer (Ocean

Optics Inc., Dunedin, FL) calibrated with a Labsphere Reflectance Standard AS-01158-060

- 42 -

USRS-99-010 BZ37A (Labsphere, Inc, North Sutton, NH). The arena was designed for a 6-way color choice test including these specific colors: green (510nm), red (624nm), black (no spectral peaks), white (420 - 880nm), yellow (568nm), and purple (440nm, 677nm). The hexagonal arena was illuminated under “cool-white” (4100K) fluorescent lighting (Philips Lighting Corporation,

Somerset, NJ) with ambient temperatures ranging from 20 – 23oC, and rotated in a six times in order to eliminate panel location bias.

Randomly-selected harlequin bugs were grouped by sex and life stage and starved 12 to

16 h before conducting experiments. Grouped bugs were organized into units of five individuals in a 200ml centrifuge tube and immobilized on ice for three minutes before gently introducing them into the center of the arena. Five minutes were allotted for all bugs to disperse and select a color panel once one of the five bugs regained motion. Color selection “choice” was determined if an individual bug made contact anteriorly or oriented within 2 cm of the colored panel.

With three replicates of the experiment, a total of 89 females, 86 males and 91 late-instar nymphs were assayed. Bugs tested had no prior experience to the arena and were not reused in subsequent assays.

Field experiment

A six-way color choice experiment using corrugated polypropylene , with the same dimension and colors as our lab color choice assay, was conducted at Kentland Farm in

Whitethorne, VA (37o 12’06.73”N, 80o 33’52.27”W; elevation = 517 m) in early autumn 2015.

Color panels were deployed as traps in a bare tilled field without host plants to intercept local populations of harlequin bugs on September 25th. Trap surfaces were treated with a pyrethroid insecticide bifenthrin (Bifenture 2E, United Phosphorous Corporation, King of Prussia, PA), at

0.2 g per liter applied with a manual pressurized backpack sprayer (Solo Sprayers, Newport

- 43 -

News, VA) to knock down intercepted bugs. Naturally dispersing populations were assumed to be highest near a 2015 spring-and-summer mustard and collard field, and therefore we situated the trap transect 6 m distant from the edge of, and parallel to, the east-west axis of the field with the faces of the square traps exposed north and south.

Six weeks before the field color choice experiment, the collard and mustard field was mowed (leaving ca. 0.5 m tall residue) and seeded with rye (Secale cereale L.) cover crop to encourage a slow dispersal of harlequin bugs out of the field. The color choice test transect was a randomized block design of five replicate series with traps spaced 3.66 m on center. One week before deploying the traps, the infested field was mowed low (leaving ca. 0.15 m host residue, at this point fairly desiccated) to encourage dispersal of harlequin bugs. Dead harlequin bugs were collected within a 0.2 m radius of each panel trap centroid 48 h after setting out traps, and every

72 h for 9 days.

Statistical Analysis

All data were analyzed in JMP 11.0.0 (SAS Institute, Cary, NC). Residuals conformed to normality, one-way and two-way ANOVA were used to test for statistical difference in color response from lab and field experiments, and Fisher’s exact test for determining group differences in lab assay. Data reported were non-transformed with α = 0.05. Mean numbers of bug responding were separated with a post-hoc Tukey’s HSD test.

Results

Laboratory Color Choice Assay

Approximately 92.5% of the released harlequin bugs (246 of the 266 bugs tested) dispersed in the lab color choice arena and made a color choice according to the choice criterion.

However, adults were significantly more likely (Fisher’s exact test, P = 0.0035) to respond than

- 44 - nymphs with 7 of the 175 (4%) adults and 13 of 91(14.3%) nymphs unresponsive (i.e. did not disperse) in this assay. Color discrimination among red, yellow, green, purple, black and white in the lab was statistically significant for females, males and nymphs (two-way ANOVA: F = 21.93; df = 5, 306; p < 0.0001). A two-way ANOVA indicated a significant interaction (F = 2.04; df =

10, 306; p = 0.0296) between color panel preference and responsive life stages (i.e. males, females, and nymphs)., A post-hoc Tukey’s HSD test revealed that red and green were equally attractive colors for females in the lab experiment. Green alone was the most attractive color for females, nymphs and males; females and males were least likely to respond to white and yellow, and nymphs were similarly least responsive (see Table 1 for complete post-hoc analysis).

Field experiment

There was a significant effect of color treatment on numbers of harlequin bugs arriving to insecticide-treated panels in the field (F = 4.28 df = 5, 24; p = 0.0063). Nymphs were overall more numerous relative to adults in this field study, with 38 adults and 145 nymphs collected at traps. Total bugs intercepted by traps in the field mirrored lab choice responses with green panels attracting significantly more harlequin bugs than the other five colors tested; black panels ranked second. (Fig. 1). Red was among the least preferred colors along with yellow, purple and white.

Discussion

Our experiments were conducted in the absence of added olfactory stimuli. Stenberg and

Ericson (2007) suggested that there may be greater sensitivity to visual over olfactory cues when specialist herbivores are starved than when satiated. Bugs were starved prior to our lab color choice test experiments and may have been more sensitive to visual cues and lab conditions than under natural field conditions. Regardless, our field and lab experiments yield consistent results

- 45 - with darker colors, specifically green and black, showing the most promise for trap construction materials for managing and monitoring harlequin bug populations.

A positive attraction to green in harlequin bug may be associated with their attraction to

Brassica host plants of high nutritional quality (Prokopy and Owens 1983, Loader and Damman

1991). The addition of red color components to a green or black trap could perhaps enhance attraction to harlequin bug females, but this idea still needs testing to confirm. Species-specific visual stimuli increases trap captures when added to a monochromatic in other insect pest systems. For example, Domingue et al. (2015) showed that 3-D printed emerald ash borer, Agrilus planipennis, (Coleoptera: Buprestidae) decoy baited branch traps with a (Z)-3-hexen-1-ol lure captured significantly more A. planipennis than without the visual-decoy. They also observed sex bias for the visual decoys with males becoming ensnared closer to decoy than females (Domingue et al. 2015).

The interaction of color with other factors such as trap shape, background contrasts, and olfactory stimuli, needs to be evaluated before determining the best application of our results to an optimal trap design for monitoring and managing harlequin bug populations. Mixed-isomer syntheses of 10,11-epoxy-1-bisabolen-3-ol (i.e. murgantiol) (Khrimian et al. 2014), are powerful attractants for harlequin bugs in the field, especially in combination with a host plant. Weber et al.

(2014) reported a subtle but significant female preference for natural murgantiol isomer ratios over more economically viable crude isomer mixtures with 1:1 (SSRS: SSRR) isomer ratios.

Selection of optimal color stimuli would be expected to enhance female attraction to a trap baited with natural or crude mixtures of murgantiol isomers. Future studies should evaluate the combination of both olfactory and visual stimuli of a trap for harlequin bugs.

- 46 -

Acknowledgements

This research was funded by Southern SARE (Sustainable Agriculture Research and

Education) graduate student grant GS15-144. We thank Paul Marek and Jackson Means (Virginia

Tech) for allowing us to use their equipment to measure spectral reflectance of the colored panels.

Paul Marek has spectral reading of other taxa organized on Virginia Tech Structural Color

Database (http://iridescent.life/). This work could not have been completed without the numerous talented individuals that support field research at Virginia Tech’s Kentland Farm.

- 47 -

References Cited

Cabrera Walsh, G., A. S. Dimeglio, A. Khrimian, and D. C. Weber. 2016. Marking and retention of harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae), on pheromone-baited and unbaited plants. J. Pest Sci. 89: 21–29. DiMeglio, A. S., A. K. Wallingford, D. C. Weber, T. P. Kuhar, and D. Mullins. 2016. Supercooling points of Murgantia histrionica (Hemiptera: Pentatomidae) and field mortality in the mid-Atlantic United States following lethal low temperatures. Environ. Entomol. 45: 1294-1299. Domingue, M. J., D. P. Pulsifer, A. Lakhtakia, J. Berkebile, K. C. Steiner, J. P. Lelito, L. P. Hall, and T. C. Baker. 2015. Detecting emerald ash borers (Agrilus planipennis) using branch traps baited with 3D-printed beetle decoys. J. Pest Sci. (2004). 88: 267–279. Khrimian, A., S. Shirali, K. E. Vermillion, M. A. Siegler, F. Guzman, K. Chauhan, J. R. Aldrich, and D. C. Weber. 2014. Determination of the stereochemistry of the aggregation pheromone of harlequin bug, Murgantia histrionica. J. Chem. Ecol. 40: 1260– 1268. Leskey, T. C., and H. W. Hogmire. 2005. Monitoring stink bugs (Hemiptera: Pentatomidae) in Mid-Atlantic apple and peach orchards. J. Econ. Entomol. 98: 143-53. Leskey, T. C., S. E. Wright, B. D. Short, and A. Khrimian. 2012. Development of behaviorally-based monitoring tools for the brown marmorated stink bug (Heteroptera: Pentatomidae) in commercial tree fruit orchards. J. Entomol. Sci. 47: 76–85. Loader, C., and H. Damman. 1991. Nitrogen content of food plants and vulnerability of Pieris rapae to natural enemies. Ecol. 72: 1586–1590. Mizell III, R. F. and W. L. Tedders. 1995. A new monitoring method for detection of the stinkbug complex in pecan orchards. Proc. Southeastern Pecan Growers Assoc. 88:36–40. Prokopy, R. J., and E. D. Owens. 1983. Visual detection of plants by herbivorous insects. Annu. Rev. Entomol. 28: 337–364. Stenberg, J. A., and L. Ericson. 2007. Visual cues override olfactory cues in the host-finding process of the monophagous leaf beetle Altica engstroemi. Entomol. Exp. Appl. 125: 81– 88. Wallingford, A. K., T. P. Kuhar, P. B. Schultz, and J. H. Freeman. 2011. Harlequin bug biology and pest management in brassicaceous crops. J. Integr. Pest Manag. 2: 1–4. Weber, D. C., G. Cabrera Walsh, A. S. DiMeglio, M. M. Athanas, T. C. Leskey, and A. Khrimian. 2014. Attractiveness of harlequin bug, Murgantia histrionica, aggregation pheromone: field response to isomers, ratios, and dose. J. Chem. Ecol. 40: 1251–1259.

- 48 -

Zahn, D. K., J. A. Moreira, and J. G. Millar. 2008. Identification, synthesis, and bioassay of a male-specific aggregation pheromone from the harlequin bug, Murgantia histrionica. J. Chem. Ecol. 34: 238–251.

- 49 -

Table 3.1

Mean Harlequin Bug Response in Lab Color Choice Test

Females Males Nymphs x s x s x s ̅ ̅ ̅ Red 1.56a 1.20 0.89abcd 0.96 0.67abcd 0.69 Yellow 0.17d 0.38 0.17d 0.38 0.22cd 0.43 Green* 1.44a 1.15 1.39a 1.24 1.50a 1.04 Purple 0.72abcd 0.83 0.78abcd 0.88 0.39bcd 0.50 Black 0.78abcd 0.55 1.33ab 1.14 1.17abc 0.92 White 0.17d 0.38 0.00d 0.00 0.39bcd 0.50

Mean harlequin bugs (x̅ ) and corresponding standard deviation (s). Numbers with a letter in common are not significantly different according to a Tukey’s HSD, P < 0.05 * indicates most preferred color shared by all groups tested

- 50 -

Fig. 3.1

Mean harlequin bugs collected at insecticide treated traps during a field experiment occurring over consecutive 72hr collection periods over the course of 11days in early-October 2015 located in Whitethorn, VA . Six colors in a randomized complete block design with a five replicate series intercepted naturally dispersing bugs. Bars with the same letter are not significantly different (Tukey’s HSD test p < 0.05).

- 51 -

Chapter Four Optimizing Attract-and-Kill Technology for Managing Harlequin Bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae)

Abstract:

Harlequin bug, Murgantia histrionica (Hahn), is an important pest of brassica crops in the United

States. To develop an effective trapping device for this pest, experiments were conducted in

Virginia to assess factors to increase harlequin bug attraction to and arrestment at stationary artificial stimuli (“traps”) positioned within an agricultural landscape. In a laboratory experiment, harlequin bugs were effectively killed or severely impaired after a few minutes exposure to deltamethrin-incorporated netting (D-Terrence®, Vestergaard-Frandsen), and, thereafter, the netting was used as a toxicant on the trapping devices in the field. In one experiment, square corrugated plastic panels were wrapped with the insecticide netting and baited with harlequin bug aggregation pheromone, murgantiol. Bugs were effectively drawn to the panels, with green panels wrapped in black D-Terrence® having significantly more dead harlequin bugs and fewer dead lady beetles (Coleoptera: Coccinellidae) at their base than did yellow panels wrapped in yellow D-Terrence®. Thus, green was chosen as the ideal trap color to use for another field experiment that evaluated three trap types coated with a residual spray treatment of bifentrin – a corrugated plastic square panel, pyramidal trap, and ramp trap – each with three lure treatments, murgantiol alone or murgantiol plus a low or high rate of benzyl isothiocynate. More bugs were killed with the pyramidal trap than with the panel trap or the ramp trap, and, on average 230% more bugs were killed at traps containing murgantiol combined with benzyl isothiocyanate than at those with murgantiol alone. This research demonstrated that, with use of proper visual and semiochemical stimuli, harlequin bugs can be drawn to trapping devices and effectively killed after contact with deltamethrin-incorporated netting.

- 52 -

Keywords: attract-and-kill, murgantiol, isothiocyanate, trap

Introduction

Harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae), is notorious for damaging brassicaceous vegetables, such as broccoli, collards, and kale, particularly in the southeastern United States (Wallingford et al. 2011, McPherson et al. 2018). The piercing- sucking feeding by both nymphs and adults, scar the leaves and can lead to leaf wilting and even plant death (Wallingford et al. 2011). The pest is a major impediment to commercial and fresh market brassica production because, unlike the major lepidopteran pests and aphids that also attack these crops, there are relatively few natural enemies of harlequin bug (Wallingford et al.

2011) and, in general, the only insecticides that are effective for harlequin bug control also negatively impact beneficial organisms, and thus, are not compatible with integrated pest management programs (Wallingford et al. 2011, 2012, McPherson et al. 2018).

Previous studies have shown that harlequin bugs respond to chemical cues (Aldrich et al.

1996) and will aggregate on certain preferred host plants such mustard (Brassica juncea

‘Southern Giant Curled’), rapeseed (B. napus ‘Athena’), rapini (B. rapa), and arugula (Eruca sativa) over others even within the Brassicaceae family (Sullivan and Brett 1974, Wallingford et al. 2013). Utilizing this strong host preference by harlequin bug, researchers have developed various trap cropping strategies that have demonstrated success for controlling harlequin bug

(Ludwig and Kok 1998, Bender et al. 1999, Wallingford et al. 2013). However, trap cropping has not been widely adopted by growers as a pest management strategy. It is not always a viable option for growers as it requires land space, which can be at a premium on small farms, and the tactic is not always effective as a control measure because of spillover of bugs from the trap plants (Wallingford et al. 2013). A commercial trap would likely be more widely adopted if it

- 53 - were shown to be effective. Currently, a harlequin bug specific trap does not exist (DiMeglio et al. 2017).

The recent identification and synthesis of the harlequin bug aggregation pheromone (two stereoisomers ([3S,6S,7R,10S]: [3S,6S,7R,10R]) of 10,11-epoxy-1-bisabolen-3-ol), has led to the commercial availability of a potent attractant for this important pest (Zahn et al. 2008, 2012,

Khrimian et al. 2014, Weber et al. 2014). The aggregation pheromone is emitted by males upon feeding on a suitable host plant (Zahn et al. 2008, 2012); therefore, it is important to understand the combined attraction of murgantiol and host plants to further develop attract-and-kill technology with synthetic stimuli in mind. In a murgantiol dose-response experiment comparing trap captures with and without whole collard plant associations, Weber et al. (2014) found the effect of a whole collard plant paired with 10mg loadings of murgantiol to provide nearly 30 times the trapping potential as did simply increasing murgantiol doses alone. Cabrera Walsh et al.

(2016) provide evidence that harlequin bugs will reach a peak density on whole collard plants baited with murgantiol and make a clear distinction between the role murgantiol and mustard derivatives in retention and attraction. In their mark-recapture study harlequin bug retention was brief on baited collard plants.

Commercial vegetable crops such as mustard, broccoli, kale and collards belong to the family (Brassicaceae), which are uniquely defended against herbivory by the glucosinolate- myrosinase system (Fahey et al. 2002, Wittstock et al. 2003, Velasco et al. 2008). With plant tissue disturbance, the segregated enzyme, myrosinase, in the presence of ascorbic acid, hydrolyzes structurally diverse glucosinolates, producing many volatile breakdown products such as thiocyanates, nitriles, and isothiocyanates (Bones and Rossiter 2006). Thrift et al. (2018) found that benzyl and allyl isothiocyanates increase harlequin bug attraction to a variety of plants baited

- 54 - with murgantiol, with benzyl isothiocyanate outperforming allyl isothiocyanate. Therefore, I chose benzyl isothiocyanate as an attractant to pair with murgantiol in this trap development research.

In designing an effective mass attract-and-kill strategy for harlequin bug, I evaluated a long-lasting insecticide net (LLIN) material, D-Terrence (Vestergaard-Frandsen, Washington,

D.C.), to kill bugs attracted to a stimulus. This specific technology was originally developed to control malaria vectors (Graham et al. 2005) and has recently shown promise as a horticultural

IPM tool for managing the turnip aphid (Aphididae: Lipaphis pseudobrassicae Davis) (Licciardi et al. 2008), Colorado potato beetle (Chrysomelidae: Leptinotarsa decemlineata (Say)), Plum curculio (Curculionidae: Conotrachelus nenuphar (Herbst)) (Gökçe et al. 2018), and brown marmorated stink bug Halyomorpha halys (Stål) (Kuhar et al. 2017). I evaluated lethal contact times of harlequin bug nymphs, and utilized this LLIN for trapping applications. Just as H. halys is reported to alight at pheromone baited traps (Leskey et al. 2012, Morrison et al. 2015), I hypothesized adult harlequin bug tenure and interaction with LLIN baited traps is relatively brief in respect to lethal exposure times. With video recorded a mark and recapture study I report relative tenure and visitation times for harlequin bug adults on LLIN square panel traps.

Harlequin bug attraction to visual and semiochemical stimuli can be incorporated into a viable pest management strategy with LLIN technology. With this research, I add new knowledge for understanding factors to increase harlequin bug visitations at stationary artificial stimuli (i.e., traps) positioned within a commercial vegetable grower landscape.

- 55 -

Materials and Methods

Field performance of deltamethrin-netting on murgantiol-baited color panels

In a preliminary lab assay >90% nymphs were either killed or moribund, as defined by

Kuhar et al. 2017, when exposed to deltamethrin netting for 1 minute (Supplemental 4.1).

Wallingford et al. (2018) also showed that the netting was toxic to both nymphs and adults.

Therefore, corrugated plastic panels, as used by DiMeglio et al. (2017), were wrapped in a single layer of deltamethrin netting and affixed with metal staples. Black mesh was used for green panels since commercially produced green insecticide treated mesh was unavailable at the time; and yellow netting from Vestergaard was used for the yellow panels.

Green and yellow panels were grouped in a randomized two-way choice test blocked in four replicates at borders of an overwintered collard (Brassica oleracea group acephala cv.

‘Champion’) field (~0.1 ha, (37°12'0.23"N 80°33'53.55"W,; Elevation = 518m)). Treatments were spaced 10m apart and replicates spaced 33m apart. The overgrown, weedy, collard plot was inspected for harlequin bug activity prior to setting up the experiment on June 7, 2015. Panels were positioned 8m from field edge and centered flush with 1.0m2 of bare soil to allow for unobstructed contact with the trap. In the early-morning on June 8, 2015 panels were cleared of all insects and murgantiol lures (single septa loaded at 10mg murgantiol isomers) were affixed

60cm aboveground to each panel treatment; following immediately with mowing the harlequin bug infested overwintered collard crop to 0.5m (Supplemental 4.2) At 48hrs, numbers of dead harlequin bugs and coccinellid species were counted within 30cm of the trap.

- 56 -

Trap shape and semiochemical bait test

Three insect trap types were selected based on insect behavior: a pyramid trap as used by

Leskey et al. (2015) for Halyomorpha halys trapping, ramp trap used in trapping the banana root borer Cosmopolites sordidus (Germar) (Reddy et al. 2008), and a square panel used in harlequin bug color preference research (DiMeglio et al. 2017). Not all traps were available in green; all three traps were coated evenly with dark green spray paint (Rust-Oleum, Gloss Hunter Green,

UltraCover Paint+Primer, Vermon Hills, IL) and allowed 24h to dry before treating all surfaces with bifenthrin (Bifenture 2EC, United Phosphorous Corporation), at 12.1ml product/L + surfactant mixed in a backpack sprayer) as a contact insecticide sprayed until runoff and given fours to air dry.

Each trap type was baited with a single murgantiol septum loaded at 10mg total for murgantiol isomers provided by and as described by Khrimian et al. 2014 and field tested by

Weber et al. 2014. Separate benzyl isothiocyanate lure dispensers were modeled after Soroka et al. (2005) with and without either low or high emission patterns of benzyl isothiocynate (technical grade, 94%; Sigma Aldrich, St. Louis, MO). Low and high volatile emissions rates were dispensed, respectively, with one and two 75mm capillary tubes with 2.59 to 2.81mm interior diameter (Fischer Scientific Co., Pittsburgh, Pa.; cat. no. 02-668-68). All together this design allowed for nine treatments:

 ramp + murgantiol only (RZ)

 ramp + murgantiol with single isothiocyanate loading (RL)

 ramp + murgantiol with double isothiocyanate loading (RH)

 pyramid + murgantiol only (PZ)

- 57 -

 pyramid + murgantiol with single isothiocyanate loading (PL)

 pyramid + murgantiol with double isothiocyanate loading (PH)

 square + murgantiol only (SZ), square + murgantiol with single isothiocyanate

loading (SL)

 square + murgantiol with double isothiocyanate loading (SH)

Treatments were spaced at 3m to be randomized in four blocks along the east-west transect of a field margin south of two fields (37°12'12.81"N, 80°33'48.97"W; Elevation = 517m) with summer cole crops (Fig 4.1); the most adjacent were collards (Brassica oleracea group acephala cv. ‘Champion’) neighbored by a mustard trap crop experiment (mixed cultivars of

Brassica juncea and B. oleracea cv. ‘Champion’).

At field deployment of 200ul of benzyl isothiocyanate was injected into a 5ml self-sealing autosampler vial (Agilent Technologies, Santa Clara, CA) with capillary tube(s) pierced through

Teflon seal, similar to Soroka et al. (2005). Individual vials and murgantiol septum were affixed

20cm from the soil surface to traps with a small green zip-tie, and stapled to traps.

One week prior to establishing treatments (Aug 31, 2016), host residue in neighboring cole crop plots was mowed using a brush mower to 0.5m to initiate crop decline and encourage harlequin bug dispersal. Dead harlequin bugs were collected within 20cm radius of trap at two

72h intervals (Sept 8 and Sept 11) after setting out the traps. Since the bugs were so numerous

(several thousand individuals), bugs were collected from the ground and counts were tallied later in the laboratory by sex and life stage (adult versus nymph).

- 58 -

Exploring a fence post “trap-line” concept

We utilized a preexisting property line fence bordering Whitethorne Road (37°12'05.0"N,

80°33'44.9"W; Elevation = 517m) east-west leading to Kentland Farm, Blacksburg, VA. The fence line was located 90 to 200 meters south of spring and summer cultivated brassicaceous crops, including the harlequin bug infested host material used as a population sink in our semiochemical and shape test. A regularly managed hay field divided brassica crops and fence line making it a viable trap-line to intercept dispersing harlequin bugs at property margins (Fig

4.2). Twelve fence posts measuring 1.5m tall, spaced 10-12m apart, were wrapped entirely with black D-Terrence net. The base of the fence posts were “skirted” 8cm (i.e. excess mesh fanned- out) to extend over the bare ground at the circumference of the fence post base. Relative attraction was compared in randomized block design with six replicates of murgantiol lure alone versus single capillary tubes dispensing low rate (as described in semiochemical and shape field test) of benzyl isothiocynate. On September 10, 2015, semiochemical lures were affixed to north face of fence posts 60cm above the ground. Dead harlequin bugs accumulated at the base were counted and removed on September 14, 2015. Semiochemical baits were removed to allow air space to clear for 24h before rerandomizing the lure locations on September 15, 2015; dead bugs were counted again on September 20, 2015.

Comparison of D-Terrence® versus Tanglefoot® for harlequin bug capture

The experiment was conducted at Kentland Farm in Whitethorne, VA (37°12'0.23"N,

80°33'53.55"W, elevation = 518m) in late September 2016. A cabbage (Brassica oleracea group capitata, cv. = ‘Bravo’) field 50m in length planted at eight rows on 0.91m centers every 0.31m was assumed the most likely source for naturally dispersing harlequin bugs in late-September

2016. Murgantiol and benzyl isothiocyanate baited square panel traps (as deployed in our trap

- 59 - shape and bait study) with either Tanglefoot® (The Tanglefoot Company, Grand Rapids, MI) or

D-Terrence were deployed at 4.3m spacing along an east-west transect 1.0m south of a ripening pepper, Capsicum annuum (cv. Aristotle) —without harlequin bugs—located 6.0m north of the declining cabbage field. Treatments were arranged in a randomized block design replicated six times with trap surfaces facing north and south, and semiochemcial baits (murgantiol and benzyl isothiocyanate) were affixed 50cm above ground on the southern face.

Nymphal and adult harlequin bugs were collected and tallied within 0.3m radius of the trap base, and enumerated separately per collection position (i.e. panel surface associated with either cabbage or pepper plants) during early evenings daily for 5 days.

Harlequin bug visitation and tenure on D-Terrence

Traps (deployed in Tanglefoot vs D-Terrence experiment) were cleared of all bugs and semiochemical baits two days before conducting this field experiment to relate observed residency of bugs on D-Terrence panels and dead bugs collected at the trap base. One-hundred harlequin bug adults were collected from the infested cabbage plot the evening prior (September

26, 2016) and retained outside in mesh cages without host plants to encourage host seeking dispersal in the experiment. The following morning, bugs were dusted with fluorescent marking powder according to methods described Cabrera Walsh et al. (2016). Marked bugs were grouped as 20 individuals in 150mm diameter Petri dishes prior to releasing bugs 30cm from the south- facing D-Terrence treated trap panels (Fig 4.3).

Square panel traps were video recorded for 5 minutes immediately upon releasing bug to enumerate initial harlequin bug residency on insecticide treated surfaces. Six hours later dead fluorescent marked bugs within 30cm of the trap were collected and tallied; and again at 36 hr.

- 60 -

Statistics

Mortality observations from 36h after exposure to D-Terrence were corrected with respect to control mortality using Abbott’s formula (Abbott 1925). Mortality data were analyzed in R version 3.2.2 (R Core Team 2015) in RStudio version 0.99.484 (RStudio 2015) and for all analyses α=0.05 was used. The R package “drc” (Ritz and Strebig 2015) was used to plot results and predict lethal times; estimated lethal times from back-transformed log-scale-based confidence intervals, and “stats” (R Core Team 2015) to test generalized linear models on the mortality data.

All remaining data were analyzed with one-way and two-way ANOVAs, Student’s t-test,

Tukey-Kramer HSD, and Fisher's exact test, using R Version (R Core Team 2018) in RStudio version 1.1.383 (RStudio 2018). Data were inspected for homogeneity of variance via histogram analysis and, as needed, with a Levene Test; appropriate transformations were performed according to allow for parametric tests.

Results and Discussion

The murgantiol-baited two treatment color field test comparing green versus yellow demonstrates that color, even in the presence of a highly potent aggregation pheromone, is extremely important in harlequin bug field attraction during a mass dispersing event (i.e. crop/host decline) (Figure 4.3). Murgantiol-baited green traps were significantly more attractive to harlequin bugs than were murgantiol-baited yellow traps (Student’s t-test; t = -3.761, df =

26.96, p = 0.0008). In agreement with others observing natural enemy attraction to yellow

(Dowell and Cherry 1981, Colunga-Garcia and Gage 1998, Adedipe and Park 2010), beneficial coccinellid species survival was significantly impacted (Figure 4.4) by trap color selection; with

- 61 - yellow D-Terrence ® treated traps more attractive (Student’s t-test; t =2.135, df = 29.50, p =

0.0412) to coccinellid species (which included Harmonia axyridis, Colemegilla maculata De

Geer, & Coccinella septempunctata L.) than green traps. Attraction and then successful retention are critical limiting factors in targeted attract-and-kill strategies within commercial agroecosystems (Witzgall et al. 2010), which is most often determined by olfactory, visual stimuli and contact stimuli (Prokopy and Owens 1983).

We found no significant interaction between trap shape and semiochemical additions in our field test comparing relatively high, low and zero emissions of benzyl isothiocyanate paired with murgantiol in either Sept. 8 or Sept. 11 (Table 4.1). Therefore, captures were analyzed separately as a function of trap type and semiochemical bait per life stage and sex group. Trap shape also is a significant factor in harlequin bug capture, as determined by a two-way ANOVA

(Table 4.2). Out of the three trap shapes tested in our harlequin bug trap design field test on Sept

8, pyramid traps captured the most harlequin bugs, over square panels, and ramps (Figure 4.5A).

Field observations are limited in determining whether trap shape played a role in field attraction or whether the trap shape prolonged harlequin retention on the insecticide treated surface, and thus, resulted in a significant difference in trap captures. The innate behavior of many pentatomids is to ascend upwards (Leskey and Hogmire 2005, Tillman and Cottrell 2015) and therefore a pyramid trap likely exploits this behavior more effectively than traps with a lower vertical profiles (e.g. ramp and square panel tested in our experiments).

Ramp traps outperformed other trap shapes for nymphs when relatively lower harlequin bug populations were intercepted on our Sept. 11. Interestingly, on the second collection date ramp and pyramid traps perform nearly similar for females and males (Figure 4.5B); whereas, pyramid and square panels were more productive on the Sept 8th collection date. Given the

- 62 - difference in trap captures from Sept. 8 to Sept 11 we can infer that harlequin bugs were likely dispersing in high numbers immediately after the “high-mow”. Our square and pyramid traps likely intercepted more individuals in flight than walking toward the traps on Sept 8th and individual intercepted on Sept. 11 were likely walking individuals, which if this is the case, may account for the difference in trap type.

Depending on collection date, sex and life stage of the bug, trap captures were significantly affected by semiochemical bait levels (Table 4.2), with a single capillary tube, rather than two capillary tube dispensers of benzyl isothiocyanate resulting in more harlequin bug captures on average (Figure 4.4B). At the second field collection date (Sept. 11), semiochemical lures did not have a significant effect on either nymphs or adults. We suspect this is related to benzyl isothiocynate depletion, or reduced lure quality from desiccation and sun exposure. These results indicate that regulated emission of volatiles is important in developing isothiocyanate lures when combining with murgantiol for attraction of harlequin bugs. Murgantiol alone (i.e., without volatile plant compound additions) was least attractive (Figure 4.5 A&B). Weber et al. (2014) found similar results with whole collard plant additions to murgantiol baited pyramid traps playing a greater role in harlequin bug recruitment than increases of murgantiol dose alone.

In the fence post study, there were no significant differences between murgantiol (푋̅ = 2.0 bugs; 휎푋 = 1.7 bugs) and murgantiol + isothiocyanate (푋̅ = 2.8 bugs; 휎푋 = 1.8 bugs) baited fence posts (F=4.965; df = 1; p = 0.469); however, significant differences were observed in the Sept 20 experiment with significantly more harlequin bugs at murgantiol + isothiocyanate (푋̅ = 1.7 bugs;

휎푋 = 0.8 bugs) than murgantiol baited (푋̅ = 0.0 bugs; 휎푋 = 0.0 bugs) posts (F=4.965; df = 1; p

<0.0001). It is loosely understood whether harlequin bugs can be trapped at a distance, and trapping at property margins 150 – 300m from probable source populations provides significant

- 63 - insights on attract-and-kill strategies for harlequin bugs. Compared with trends observed in our close proximity semiochemical bait and shape experiment, harlequin bug captures at baited traps seems to vary is based on unknown factors; however, the additions of benzyl isothiocyanate does not decrease harlequin bug attraction to murgantiol baited traps.

When I’ve interacted with fresh market vegetable growers in the mid-Atlantic U.S., there is skepticism for aggregation pheromone-based pest management. Most of these growers cultivate on 2 -10 acres, and agronomic scale may influence their perceptions. For example, Sargent et al.

(2014) found a correlation between aggregation pheromone baited traps and H. halys feeding injury on tomatoes in urban gardens—which were relatively small cropping systems. Improper semiochemical baited trap deployment can lead to reduced intended plant protection as reported in multiple insect pest complexes (Laidlaw et al. 2003, Leskey et al. 2008, Switzer et al. 2009,

Witzgall et al. 2010, Sargent et al. 2014), specifically in the case of “spill-over” documented in japonilure baited traps for Popillia japonica Newman (Switzer et al. 2009). Future development that fine-tuning productive deployment of attract-and-kill strategies for harlequin bug may benefit from understanding trapping at distant field margin. Pheromone-mediated integrated pest management strategies have been widely successful in managing lepidopteran pests, and Witzgall et al. (2010) cite several practical attract-and-kill successes. It is, however, important to factor in geographical scale of crop production and acceptable crop loss relative to economic injury levels in designing trapping strategies.

My fence-post study provides evidence that low population sampling is informative for developing an attract-and-kill tactic at a distance from host crops, but would require multiple replicates and repeated deployments to use as a method in surveying candidate attractants. This same methodology was successful within a nationwide study comparing attractive isomer blends

- 64 - of an aggregation pheromone for H. halys Stål (Leskey et al. 2015). The results from a single host plant volatile are limited for immediate grower use because the lack of experimental data on relative efficacy. With at least 120 known glucosinolates (Fahey et al. 2002), more research is needed to comprehensively understand harlequin bug attraction to host volatiles, in combinations and at proper emissions, when paired with murgantiol.

When comparing harlequin bug captures at murgantiol and benzyl isothiocyanate baited

D-Terrence® versus Tanglefoot® treated square panel traps (Figure 4.6), significantly more adults were captured on Tanglefoot® than D-Terrence treated traps (Student’s t-test; t = -2.2; df

= 5; p = 0.0396); and nearly the inverse with nymphs (Student’s t-test; t = 4.034; df = 5; p =

0.00499). Thus, the trap surfaces are critical to intercepting the full swath of responding bugs to semiochemical-baited traps, and therefore further showing a need for an improved trap design that accounts for dispersal behaviors. More harlequin bug (nymphs and adults) were found clustered on the adjacent row of peppers, a non-host plant, than on our traps. Bugs on peppers were observed feeding on ripening fruit.

These differences among nymphs and adults, and observed harlequin bug “spill-over” to a non-host crop, may be linked to life stage specific mobility factors. For instance, adults collected on Tanglefoot treatments were consistently intercepted at height of ~0.46m above ground; whereas nymphs were prominently at the base of the traps. Since Tanglefoot is a known insect trapping material for pentatomids (Jang and Park 2010, Blaauw et al. 2016, 2017), we can infer that these adults made contact with the trap via flight. Thus far, all elements of our traps are attractive to harlequin bugs, but “spill-over” to cash crops is still a potential problem.

Within the first five minutes of video recordings, 25% of harlequin bug adults walked

30cm from their release point to the base of the trap and quickly ascended upon contact. Adults

- 65 - were recorded navigating trap surfaces freely before either crawling down (11/25bugs) or taking flight and departing from the trap (14/25bugs). However, there was no significant difference in residency time between these two groups (Student’s t-test; t = 2.0856; df = 20; p = 0.560), and no significant difference (Fisher’s Exact test, p = 0.807) between departure fate within the observed sample. The fluorescent mark-and-release experiment confirmed that adult harlequin bug residency on D-Terrence traps baited with murgantiol and benzyl isothiocynates is brief (푋̅ =

61.7s; 휎푋 = 67.2s). Residency times ranged from 5 to 236 s with three mentionable outliers of

236, 219, and 217 s; whereas, these bugs resided <1cm from semiochemical baits for an extended period of time. We observed zero immediate knockdown cases within these first five minutes; so therefore in all factors considered, field captures at D-Terrence-treated traps does not accurately represent absolute captures (i.e., actual number of bugs intercepted by the trap and picking up lethal dose).

When we inspected the traps for fluorescent marked harlequin bug adults at 6h and 36h post release we observed mean captures of 1.8 bugs (휎푋 = 1.47) and 0.8 bugs (휎푋 =0.75), respectively. Based on video recordings of the first 5 minutes after release, there was over a two- fold difference in numbers of harlequin bugs intercepting the trap and then departing than total bugs captured 36h post release. It is likely that absolute trap visit is greater than documented in the video recordings since 20 marked bugs were released at each replicate totaling 100 bugs, and thus our 36h counts represent a small portion of the absolute attract-and-kill efficacy. Whether or not bugs that come into contact with D-Terrence then alight actually survive and cause noticeable crop injury is unknown.

Designing a device intended for targeted attract-and-kill tactics in vegetable crop production should consider species-specific toxicants. D-Terrence formulation is currently

- 66 - deltamethrin, a broadspectrum pyrethroid insecticide. Including additional insecticides into D-

Terrence formulations may abbreviate exposure times needed for effective knockdown when used in stink bug attract-and-kill strategies.

Visual cues—such as color, surface area, and shape—are all important in attracting harlequin bugs. When comparing green and yellow traps baited with murgantiol, yellow traps killed more non-target coccinellids. Just as visual cues have a noticeable impact on non-targets, so too do volatile host compounds. James (2005) documents that coccinellids, hymenopterans, hemipterans and dipterans are attracted to herbivore-induced plant volatiles (e.g. trans-2-hexen-1- a, linalool, cis-3-hexen-1-ol, and methyl salicylate); including these in harlequin bug attract-and- kill technology may compromise sustainable IPM tactics by increasing non-target captures.

Beneficial coccinellid species were not present in captures from our semiochemical and shape test, therefore the current combination of green traps with murgantiol and benzyl isothiocynate is likely to have relatively negligible impacts on beneficial insect communities.

Late-summer harlequin bug populations peak at a time when most available host crops are in decline, as we observed prolonged harlequin bug aggregations and feeding on ripening peppers when murgantiol + benzyl isothiocyanate traps were deployed in immediate proximity.

Peppers are out of the host range of reproductive harlequin bugs and others have observed harlequin bugs on ripening non-host fruits including raspberries (Basnet et al. 2014). Therefore, it is likely that harlequin dispersal behavior in the late-summer months resembles an opportunistic feeder, and less of a specialist.

Seasonal variation in host volatile attraction is not uncommon in other brassica specialists. Gruber et al. (2009) report that Phyllotreta flea beetles were inhibited by allyl isothiocynates in the late-fall, but allyl isothiocynates were significantly attracted in the spring

- 67 - and early-fall. Behavioral tendency shifts due to physiological conditioning are not uncommon in harlequin bugs. In the field, mated females were observed moving from mustards to collards.

Helmey-Hartman and Miller (2014) found that harlequin bug mating and female choice are affected by natal host plants, also that mating success can be dependent on host plant context.

In conclusion, I provided evidence that semiochemical and visual stimuli are both important factors to consider in designing trapping strategies for harlequin bug. Additions of controlled release rates of benzyl isothiocyanate to murgantiol lures increased the likelihood of trapping more bugs. Mass attract-and-kill strategies that leverage toxicant exposure require pinpoint (i.e. a defined geometric area which effectively intercepts all recruited bugs) of attraction and sufficient retention to reduce the inevitable risk of trap “spill-over”. Given what we now know about harlequin bug mobility factors, a pyramid trap with ramp traps features may effectively capture both walking and flying bugs. In future studies, a wider range in candidate semiochemical attractants should be considered to expand knowledge on harlequin bug behavior, as well as to optimize trapping efficacy. It would be interesting to also investigate seasonally driven variability in behavioral responses to baited traps to further understand ethological drivers in harlequin bug attraction and dispersal behavior.

Acknowledgements

Murgantiol lures were produced and provided by Dr. Ashot Khrimian and Mr. Fil Guzman at the USDA-ARS Invasive Insect Biocontrol and Behavior Laboratory. This research was funded by Southern SARE (Sustainable Agriculture Research and Education) graduate student grant

GS15-144, and a USDA-ARS cooperative agreement. This work could not have been completed without the numerous talented individuals that support field research at Virginia Tech’s Kentland

Farm.

- 68 -

References

Abbott, W. S. 1925. A method of computing the effectiveness of an insecticide. J. Econ. Entomol. 18: 265 – 267. Adedipe, F., and Y.-L. Park. 2010. Visual and olfactory preference of Harmonia axyridis (Coleoptera: Coccinellidae) adults to various companion plants. J. Asia. Pac. Entomol. 13: 319–323. Aigner, J. D., J. M. Wilson, L. B. Nottingham, J. A. Morehead, A. DiMeglio, and T. P. Kuhar. 2015. Bioassay evaluation of IKI-3106 (Cyclaniliprole) for control of brown marmorated stink bug and harlequin bug, 2014. Arthropod Manag. Tests. 40: L4. Aldrich, J. R., J. W. Avery, C.-J. Lee, J. C. Graf, and D. J. Harrison. 1996. Semiochemistry of cabbage bugs (Heteroptera: Pentatomidae) Eurydema and Murgantia. J. Entomol. Sci. 31: 172–182. Basnet, S., L. M. Maxey, C. A. Laub, T. P. Kuhar, and D. G. Pfeiffer. 2014. Stink bugs (Hemiptera: Pentatomidae) in primocane-bearing raspberries in southwestern Virginia. J. Entomol. Sci. 49: 304–312. Bender, D. A., W. P. Morrison, and R. E. Frisbe. 1999. Intercropping cabbage and Indian mustard for potential control of lepidopterous and other insects. Hortscience 34: 275–279. Blaauw, B. R., V. P. Jones, and A. L. Nielsen. 2016. Utilizing immunomarking techniques to track Halyomorpha halys (Hemiptera: Pentatomidae) movement and distribution within a peach orchard. PeerJ. 4: e1997. Blaauw, B. R., W. R. Morrison, C. Mathews, T. C. Leskey, and A. L. Nielsen. 2017. Measuring host plant selection and retention of Halyomorpha halys by a trap crop. Entomol. Exp. Appl. 163: 197–208. Bones, A. M., and J. T. Rossiter. 2006. The enzymic and chemically induced decomposition of glucosinolates. Phytochemistry. 67: 1053–1067. Cabrera Walsh, G., A. S. DiMeglio, A. Khrimian, and D. C. Weber. 2016. Marking and retention of harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae), on pheromone baited and unbaited plants. J. Pest Sci., DOI 10.1007/s10340-015-0663-1.

Colunga-Garcia, M., and S. H. Gage. 1998. Arrival, establishment, and habitat use of the multicolored asian lady beetle (Coleoptera: Coccinellidae) in a Michigan landscape. Environ. Entomol. 27: 1574–1580. DiMeglio, A. S., J. D. Aigner, J. M. Wilson, L. B. Nottingham, J. A. Morehead, and T. P. Kuhar. 2015. Field efficacy of IKI 3106 50SL for the control of harlequin bugs in collards in Virginia, 2014: Table 1. Arthropod Manag. Tests. 40: E4. DiMeglio, A. S., T. P. Kuhar, and D. C. Weber. 2017. Color preference of harlequin bug (Heteroptera: Pentatomidae). J. Econ. Entomol. 110: 2275–2277.

- 69 -

Dowell, R. V., and R. H. Cherry. 1981. Survey traps for parasitoids, and coccinellid predators of the citrus blackfly, Aleurocanthus woglumi. Entomol. Exp. Appl. 29: 356–362. Fahey, J. W., A. T. Zalcmann, P. Talalay, and J. W. Fahey. 2002. The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry. 59. Gökçe, A., G. Bingham V., and M. Whalon E. 2018. Impact of long-lasting insecticide- incorporated screens on Colorado potato beetle and plum curculio. Turkish J. Agric. For. 42. Graham, K., M. H. Kayedi, C. Maxwell, H. Kaur, H. Rehman, R. Malima, C. F. Curtis, J. D. Lines, and M. W. Rowland. 2005. Multi-country field trials comparing wash-resistance of PermaNet™ and conventional insecticide-treated nets against anopheline and culicine mosquitoes. Med. Vet. Entomol. 19: 72–83. Gruber, M. Y., N. Xu, L. Grenkow, X. Li, J. Onyilagha, J. J. Soroka, N. D. Westcott, and D. D. Hegedus. 2009. Responses of the crucifer flea beetle to brassica volatiles in an olfactometer. Environ. Entomol. 38: 1467–1479. Helmey-Hartman, W. L., and C. W. Miller. 2014. Context-dependent mating success in Murgantia histrionica (Hemiptera: Pentatomidae). Ann. Entomol. Soc. Am. 107: 264–273. James, D. G. 2005. Further field evaluation of synthetic herbivore-induced plant volatiles as attractants for beneficial insects. J. Chem. Ecol. 31: 481–495. Jang, S. A., and C. G. Park. 2010. Gymnosoma rotundatum (Diptera: Tachinidae) attracted to the aggregation pheromone of Plautia stali (Hemiptera: Pentatomidae). J. Asia. Pac. Entomol. 13: 73–75. Khrimian, A., S. Shirali, K. E. Vermillion, M. A. Siegler, F. Guzman, K. Chauhan, J. R. Aldrich, and D. C. Weber. 2014. Determination of the stereochemistry of the aggregation pheromone of harlequin bug, Murgantia histrionica. J. Chem. Ecol. 40: 1260–1268.

Kuhar, T.P., A. Morehad, T. DiMeglio. 2015 Diagnosing stink bug injury to vegetables. Vir. Coop. Exten. Ento-173NP Kuhar, T. P., B. D. Short, G. Krawczyk, and T. C. Leskey. 2017. Deltamethrin-incorporated nets as an integrated pest management tool for the invasive Halyomorpha halys (Hemiptera: Pentatomidae). J. Econ. Entomol. 110: 543–545. Laidlaw, W. G., B. G. Prenzel, M. L. Reid, S. Fabris, and H. Wieser. 2003. Comparison of the efficacy of pheromone-baited traps, pheromone-baited trees, and felled trees for the control of Dendroctonus pseudotsugae (Coleoptera: Scolytidae). Environ. Entomol. 32: 477–483. Leskey, T. C., A. Agnello, J. C. Bergh, G. P. Dively, G. C. Hamilton, P. Jentsch, A. Khrimian, G. Krawczyk, T. P. Kuhar, D.-H. Lee, W. R. Morrison, D. F. Polk, C. Rodriguez-Saona, P. W. Shearer, B. D. Short, P. M. Shrewsbury, J. F. Walgenbach, D. C. Weber, C. Welty, J. Whalen, N. Wiman, and F. Zaman. 2015. Attraction of the invasive Halyomorpha halys (Hemiptera: Pentatomidae) to traps baited with semiochemical stimuli across the United States. Environ. Entomol. 44: 746–756.

- 70 -

Leskey, T. C., and H. W. Hogmire. 2005. Monitoring stink bugs (Hemiptera: Pentatomidae) in Mid-Atlantic apple and peach orchards. J. Econ. Entomol. 98. Leskey, T. C., J. C. Pinero, and R. J. Prokopy. 2008. Odor-baited trap trees: A novel management tool for plum curculio (Coleoptera: Curculionidae). J. Econ. Entomol. 101: 1302–1309. Leskey, T. C., S. E. Wright., B. D. Short and A. Khrimian. 2012. Development of behaviorally based monitoring tools for the brown marmorated stink bug, Halyomorpha halys (Stål) (Heteroptera: Pentatomidae) in commercial tree fruit orchards. J. Entomol. Sci. 47: 76–85 Licciardi, S., F. Assogba-Komlan, I. Sidick, F. Chandre, J. M. Hougard, and T. Martin. 2008. A temporary tunnel screen as an eco-friendly method for small-scale farmers to protect cabbage crops in Benin. Int. J. Trop. Insect Sci. 27: 152–158. Ludwig, S. W., and L. T. Kok. 1998. Evaluation of trap crops to manage harlequin bugs, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae) on broccoli. Crop Prot. 17: 123– 128. McPherson, J. E., C. S. Bundy, and T. P. Kuhar. 2018. Chapter 6: Murgantia histrionica (Hahn) pp. 331-348 in J. E. McPherson (Ed.) Invasive Stink Bugs and Related Species (Pentatomoidea): Biology, Higher Systematics, Semiochemistry, and Management, CRC Press. Morrison III, W. R., Cullum, J. P., & Leskey, T. C. 2015. Evaluation of trap designs and deployment strategies for capturing Halyomorpha halys (Hemiptera: Pentatomidae). J. Econ. Entomol. 108(4), 1683-1692. Nauen, R. 2006. Insecticide mode of action: return of the ryanodine receptor. Pest Manag. Sci. 62: 690–692. Prokopy, R. J., and E. D. Owens. 1983. Visual detection of plants by herbivorous insects. Annu. Rev. Entomol. 28: 337–364. R CoreTeam 2015. R: A language and environment for statistical computing. Vienna, Austria. (http://www.r-project.org/). RStudio. 2015. Integrated development environment for R, Version 0.99.484. Boston, MA. (http://www.rstudio.org/).

Ritz and Strebig 2016. "drc" : Analysis of Dose-Response Curves. Version 3.0-1. Reddy, G. V. P.; Cruz, Z. T.; Guerrero, A 2009. Development of an efficient pheromone-based trapping method for the banana root borer Cosmopolites sordidus. Journal of Chemical Ecology. 35:111-117. Sargent, C., H. M. Martinson, and M. J. Raupp. 2014. Traps and trap placement may affect location of brown marmorated stink bug (Hemiptera: Pentatomidae) and increase injury to tomato fruits in home gardens. Environ. Entomol. 43: 432–438.

- 71 -

Soroka, J. J., R. J. Bartelt, B. W. Zilkowski, and A. A. Cossé. 2005. Responses of flea beetle Phyllotreta cruciferae to synthetic aggregation pheromone components and host plant volatiles in field trials. J. Chem. Ecol. 31: 1829–1843. Sullivan, M. J., and C. H. Brett. 1974. Resistance of commercial crucifers to the harlequin bug in the coastal plain of North Carolina. J. Econ. Entomol. 67: 262–264. Switzer, P. V, P. C. Enstrom, and C. A. Schoenick. 2009. Behavioral explanations underlying the lack of trap effectiveness for small-scale management of Japanese beetles (Coleoptera: Scarabaeidae). J. Econ. Entomol. 102: 934–940. Thrift, Emma M., M.V. Herlihy, A. K. Wallingford, and D. C. Weber. 2018. Fooling the harlequin bug (Hemiptera: Pentatomidae) using synthetic volatiles to alter host plant choice. Environ. Entomol. 47: 432-439. Tillman, P. G., and T. E. Cottrell. 2015. Spatiotemporal distribution of Chinavia hilaris (Hemiptera: Pentatomidae) in peanut-cotton farmscapes. J. Insect Sci. 15. Velasco, P., P. Soengas, M. Vilar, and M. Elena-Cartea. 2008. Comparison of glucosinolate profiles in leaf and seed tissues of different brassica napus crops. J. Am. Soc. Hortic. Sci. 133: 551–558. Velikova, V., G. Salerno, F. Frati, E. Peri, E. Conti, S. Colazza, and F. Loreto. 2010. Influence of feeding and oviposition by phytophagous pentatomids on photosynthesis of herbaceous plants. J. Chem. Ecol. 36(6), 629-641.edit Wallingford, A. K., T. P. Kuhar, and D. C. Weber. 2018. Avoiding Unwanted Vicinity Effects With Attract-and-Kill Tactics for Harlequin Bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae). J. Econ. Entomol, 111(3), 2018, 1–8; doi: 10.1093/jee/toy10 Wallingford, A. K., T. P. Kuhar, P. B. Schultz, and J. H. Freeman. 2011. Harlequin bug biology and pest management in brassicaceous crops. J. Integr. Pest Manag. 2. Wallingford, A. K., T. P. Kuhar, and P. B. Schultz. 2012. Toxicity and field efficacy of four neonicotinoids on harlequin bug (Hemiptera: Pentatomidae). Florida Entomol. 95: 1123– 1126. Wallingford, A. K., T. P. Kuhar, D. G. Pfeiffer, D. B. Tholl, J. H. Freeman, H. B. Doughty, and P. B. Schultz. 2013. Host plant preference of harlequin bug (Hemiptera: Pentatomidae), and evaluation of a trap cropping strategy for its control in collard. J. Econ. Entomol. 106: 283–288. Wallingford, A.K., T. P. Kuhar, and D. C. Weber. 2018. Avoiding unwanted vicinity effects with attract-and-kill tactics for harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae). J. Econ. Entomol. 111: 1780-1787.Walsh, G. C., A. S. Dimeglio, A. Khrimian, and D. C. Weber. 2016. Marking and retention of harlequin bug, Murgantia histrionica (Hahn) (Hemiptera: Pentatomidae), on pheromone-baited and unbaited plants. J. Pest Sci. (2004). 89: 21–29. Weber, D. C., G. Cabrera Walsh, A. S. DiMeglio, M. M. Athanas, T. C. Leskey, and A. Khrimian. 2014. Attractiveness of harlequin bug, Murgantia histrionica, aggregation pheromone: Field response to isomers, ratios, and dose. J. Chem. Ecol. 40: 1251–1259.

- 72 -

Wittstock, U., D. J. Kliebenstein, V. Lambrix, M. Reichelt, and J. Gershenzon. 2003. Chapter five: Glucosinolate hydrolysis and its impact on generalist and specialist insect herbivores. Recent Adv. Phytochem. 37: 101–125. Witzgall, P., P. Kirsch, and A. Cork. 2010. Sex pheromones and their impact on pest management. J. Chem. Ecol. 26: 80–100. Zahn, D. K., J. A. Moreira, and J. G. Millar. 2008. Identification, synthesis, and bioassay of a male-Specific aggregation pheromone from the harlequin bug, Murgantia histrionica. J. Chem. Ecol. 34: 238–251. Zahn, D. K., J. A. Moreira, and J. G. Millar. 2012. Erratum to: Identification, synthesis, and bioassay of a male-specific aggregation pheromone from the harlequin bug, Murgantia histrionica. J. Chem. Ecol. 38: 126.

- 73 -

Fig 4.1 Nine treatments of traps, as shown in green boxes with pairs of letters, were randomized within four blocks. Trap shapes pyramid (P), ramp (R) and square panel (S) were paired with murgantiol lure treatments (Z) and combined with benzyl isothiocyanate emitted with two capillary tubes (H) or one tube (L). Both mustard and collard crops were mowed 0.5m high to disperse resident harlequin bugs a week prior to arranging this experimental design.

- 74 -

Fig 4.2

Twelve fence posts measuring 1.5m tall, spaced 10-12m apart, were wrapped entirely with black

D-Terrence® net and baited with 10mg loading of murgantiol. A regularly managed hay field divided spring and summer brassica crops (bright green bars) and fence line (tan circles) making it a viable trap-line to intercept dispersing harlequin bugs at property margins. Relative attraction was compared in a randomized block design with six replicates of murgantiol lure alone versus single capillary tubes dispensing low rate (as described in semiochemical and shape field test) of benzyl isothiocynate.

- 75 -

Fig 4.3 Square panel traps (one shown here in side-profile as a single black line) were cleared of all bugs and for this field experiment to relate observed residency and tenure of harlequin bugs on

D-Terrence panels and dead bugs collected at the trap base. One-hundred harlequin bug adults were collected from the infested cabbage plot the evening prior (September 26, 2016) and retained outside in mesh cages without host plants. The following morning, bugs were dusted with fluorescent marking powder and grouped as 20 individuals prior to releasing bugs 30cm from the south-facing D-Terrence treated trap panels. Square panel traps were video recorded for

5 minutes immediately upon releasing bugs to enumerate initial harlequin bug residency on insecticide treated surfaces. Six hours later dead fluorescent marked bugs within 30cm of the trap were collected and tallied; and again at 36 hr.

- 76 -

Fig. 4.4

(Left) Total number of dead harlequin bugs recorded over 48h period from green versus yellow murgantiol-baited square panel traps wrapped with deltamethrin-incorporated netting. (Right)

Total number of dead coccinellids recorded over 48h period from green versus yellow murgantiol-baited square panel traps wrapped with deltamethrin-incorporated netting. Treatments were positioned within a randomized complete block design with a four replicates. June 8, 2015.

Asterisk (*) indicates significantly different trap captures (Fisher’s Exact Test, p<0.05).

- 77 -

Fig. 4.5 Mean number of harlequin bugs (+/-sd) collected at insecticide treated traps during a field experiment occurring over two consecutive 72h collection periods over the course of 6 days in early-September 2016 located in Whitethorn, VA . Three trap types (square, pyramid, and ramp) and three semiochemical bait combinations ([zero = murgantiol], [low = murgantiol + single capillary tube emitting benzyl isothiocyanate], and [high = murgantiol + two capillary tubes emitting benzyl isothiocyanate]) were positioned within a randomized complete block design with a four replicate series intercepting dispersed bugs from mowed harlequin bug infested cole crops.

Bars with the same letter are not significantly different (Tukey’s HSD test p < 0.05).

- 78 -

Fig. 4.6.Mean (+/- sd) harlequin bug captures per treatment respective trap face (i.e. pepper v. cabbage). Dark textured grey (face associated with peppers) and light grey with diagonal patterns

(face associated with cabbage) communicates trap context and likely explains “spill-over” from murgantiol + benzyl isothiocyanate baited trap.

- 79 -

Table 4.1. Results from a two-way ANOVA (p<0.05) performed on trap collection during a field experiment investigating efficacy of semiochemical baits and trap shape on dispersing harlequin bugs.

Sept. 8 Sept. 11

F-Value df p-value F Value df p-value

Females Bait*Trap 0.242 4, 27 0.912 0.21 4, 27 0.931

Males Bait*Trap 0.325 4, 27 0.859 0.165 4, 99 0.956

Nymphs Bait*Trap 0.066 4, 27 0.992 1.275 4, 27 0.304

- 80 -

Table 4.2. Results from a two-way ANOVA (p<0.05) performed on trap collection during a field experiment investigating efficacy of semiochemical baits and trap shape on dispersing harlequin bugs.

Sept. 8 Sept. 11

F-Value df p-value F Value df p-value

Females Bait Levels 2.658 2 0.088 1.182 2 0.322 Trap Shape 5.749 2 0.008 9.27 2 0.001

Males Bait Levels 4.284 2 0.024 0.488 2 0.616 Trap Shape 2.909 2 0.072 4.815 2 0.010

Nymphs Bait Levels 12.944 2 0.000 1.278 2 0.295 Trap Shape 4.278 2 0.024 19.875 2 0.000

- 81 -

Supplemental

Supplemental 4.1

Mean (+/- sd) percentage mortality (black bars) and moribundity (grey bars) of harlequin bug nymphs from timed (minutes) ventral contact with deltamethrin-incorporated netting. Bar graphs describe observations (A) immediately and (B) 36h post exposure. Nymph mortality and moribundity was high (푋̅ = 22.5%; 휎푋 = 16.4% and 푋̅ = 62.5%; 휎푋 = 8.3%, respectively) when in contact with D-Terrence for one minute when evaluated 36 hours post treatment (Figure 4.1B).

Our mortality evaluation at 36h post treatment for all exposure times led to a predicted LT50 of

0.107mins 95% CI [0.0117, 0.9696] (or ~5.34sec) in harlequin bug nymphs.

- 82 -

B

A A

C D

Supplemental 4.2. (A) green panel trap, (B) yellow panel trap, (C) mowing harlequin bug refuge to drive insects out to traps, (D) after mow HB still cling to cut host material.

- 83 -

Chapter 5: Conclusions and Future Considerations

There are always afterthoughts with a thesis; additional experiments, experiments that could have been run differently; another field season of data, or just more time. In this thesis, I advance our knowledge of harlequin bug cold hardiness and determine key factors for the development of a trap for this pest. I conducted my experiments with the intent to translate results to field outcomes, with a particular emphasis on developing a grower friendly application.

Therefore, some basic details of harlequin bug biology may have been overlooked.

The cold hardiness studies provided valuable insights relating internal ice formation and survivability to cold weather events. Both 2014 and 2015 winter weather extreme events were serendipitous scenarios. The weather forecast only allowed a few days to prepare the experiments.

I was thankful to have microclimate sensors installed precisely where harlequin bugs overwintered. In 2014 there were thousands of local harlequin bugs available to sequester in cages and therefore, I was able to recover and test a large number of bugs from a single farm plot. In

2015, however, available bugs were limited and I conservatively handled and sequestered them on overwintering collards. Because of this I did not separate dead versus live harlequin bugs, limiting the ability to tease out the exact lethal effects from the cold weather episode itself. Nonetheless, the overall results relate extreme winter weather to harlequin bug survival.

Extreme winter weather episodes, are simply that—episodic. If I had the forethought and sufficient bug supply, it would have been insightful to explore relative impacts of episode-free winters on harlequin bug survival. Experimentally, this could have been accomplished by simply isolating large numbers (several hundred) of overwintering adults in an unheated greenhouse during those 2014 and 2015 episodes, then returning them to the field with a control population remaining out in the field to experience extreme cold conditions. In addition to survival, I could

- 84 - have assessed their longevity, how they dispersed from their winter refuge, as well as feed and reproduce in the following spring. I hypothesize that bugs sheltered from extreme winter weather and held at warmer winter temperatures would be more fecund, likely to disperse and live longer during the following spring.

I surveyed supercooling points (SCPs) of Virginia and Maryland populations to describe physiological limits of cold hardiness. The insects were summer (16:8, L:D) conditioned within a greenhouse environment. At the time of the studies, I knew very little on the effects of photoperiod and temperature in seasonally-conditioning harlequin bugs. After these studies were conducted, Dr. Anna Wallingford (pers. communication) found that photoperiod is a critical cue for egg laying in harlequin bug. In future supercooling point studies, I would recommend surveying preconditioned bugs under various seasonal conditions.

For one of my field trapping experiments, in early spring (March), I ran a preliminary study looking at potential traps paired with harlequin bug aggregation pheromone, murgantiol, and a collard plant. The study was conducted over the course of 5 days and consisted of a yellow crossvane trap paired with murgantiol and coupled with and without a collard plant. Collard plants with a yellow crossvane + murgantiol caught significantly fewer harlequin bugs than a collard plant + murgantiol alone. This suggested yellow could possibly be a deterrent or repellant to harlequin bugs. I did not pursue this idea further, instead continuing with experiments describing harlequin bug color preference and its importance for a trap.

A bulk of my field trapping studies investigated the role of semiochemical baits in trapping harlequin bugs. In early planning phase of my experiments I considered working with herbivore induced plant volatiles in investigating harlequin bug recruitment at traps. I retracted this idea given the assumed impact on beneficial insects. Benzyl isothiocyanate was selected over

- 85 - allyl isothiocyanate because of earlier (2013) preliminary work with Dr. Guillermo Cabrera

Walsh and studies comparing recruitment performance of benzyl isothiocyanate and allyl isothiocynate; we found that benzyl isothiocyanate outperformed allyl isothiocyanate. The mustard oils in these studies were applied in aliquots on a cotton dental wick. In my studies presented in this thesis I used a capillary tube system to regulate emission of mustard oil compounds. I attempted to calculate an average emission rate with this set up, but found it difficult and found errors in the calculations. I used field exposed vials and calculated emission rate via weight differences at a given period time of exposure. In the field, I observed condensation build up in the sampler vials, which is a probable explanation for the errors in my calculations. I would like to see future research to relate emission rates of isothiocyanates to harlequin attraction. It would also be insightful if research is conducted on multiple candidate attractants.

Future experiments with harlequin bug trap design should glean more information on harlequin bug behavior. I would recommend follow up experiments to this thesis include factors related to harlequin bug preconditioning. For example, in the visitation and tenure experiments, I did not account for bug age, reproductive status of females in particular, or nutritional status.

These are key physiological conditions that could affect interaction time with a trap, and possibly help us further understand vicinity effects such as trap spill-over.

Harlequin bug field experiments are innately problematic, but yet valuable in yielding insights on overall on-farm behaviors. In my trapping factor studies, I relied on crop destruction to trigger large population dispersals. I decided on this method since we were looking at developing trap that would specifically attract dispersing bugs. Mass destruction of a harlequin bug infested crop is also a scenario that I personally witnessed on vegetable farms, and therefore

- 86 - realistic for field studies to give relevant on-farm information. The high-mow method allowed for sufficient data interpretation and statistical test.

I did run other field tests that did not involve crop destruction, that instead represented dispersal patterns throughout the growing season, in the absence of crop destruction. Many of these studies, such as one comparing red versus black in female attraction, resulted in either zero bugs arriving at the traps or limited numbers for statistical analyses. I would like to see someone in the future develop field experiment methodology that can decisively infer the nuances of host plant selection behavior under different conditions and seasons for harlequin bug. This could also help answer the question I posed in Chapter 4 of whether or not harlequin bugs are truly year- round specialists or if they shift to opportunistic feeding in early fall.

Low population sampling is powerful in making inferences with field data—it’s just more realistic. Oftentimes in applied sciences we gravitate to parametric statistics, which yield clear results when the data is just right. Several of my studies were a “flop” meaning I did not generate enough data to make a reasonable inference on treatment effect. If I had a stronger background and preparation in non-Gaussian statistical models I would have designed experiments differently to analyze treatment effects within a low population size. There is value in becoming trained in spatial statistics.

The harlequin bug is a fascinating organism for behavioral experiments. Although not discussed in my earlier chapters, I observed behavioral patterns that make me think a short-range aggregation pheromone (other than murgantiol, which seems to function as a long-range semiochemical) is important to harlequin behavior. In terms of short range behavior studies, most are only focused on courtship and mating behavior. Throughout the course of my studies I observed behavior anomalies in harlequin bug. One instance involved hundreds of adult bugs in a

- 87 - mesh bag all seeming normal in appearance except for one, which was half the size of the other adults and mottled with abnormal hues. The bugs in the bag seem to literally “bully” this abnormal individual pushing it out of the main cluster of bugs in the bag. This is an observation that would be extremely difficult to replicate, but is noteworthy as an afterthought for better understating harlequin bug behavior.

One of the other behaviors that make me think a short-range aggregation pheromone is important to harlequin bug behavior include a seemingly important overwintering strategy where bugs bask immediately after the first frost on brown necrotic lesions on their host plants. In this scenario the bugs (both nymph and adults) are in tight aggregations. I observed this nearly every year; and I would suspect that if this behavior is linked to a short range aggregation pheromone, that could be also an important semiochemical in their overwintering strategy.

Lastly, I observed harlequin bugs exhibiting a rescue behavior to each other. I documented this behavior with video reduction and am currently working up a publication describing the observed behavior. During my studies I also attempted to identify a pheromone to explain the behavior using solid phase microextraction fibers. The volatile sampling process was crude, so my controls contained undecipherable levels of background noise. I did, however, make a clear reading of a four component emission that match the proportional distributions of the common bed bug’s short range (i.e. 30cm or less) aggregation pheromone. When the rescue behavior work is published I will make note of this pheromone idea. If it is true that harlequin bugs produce a short range aggregation pheromone when distressed, and simultaneously engage in a rescue behavior, that would add important information to understanding how these bug remain in tight aggregations and how they disperse.

- 88 -

Overall, harlequin bug is an attractive species for agroecological studies: quite conspicuous, all life stages found on brassica crops, responsive to both visual and chemical cues that impact its behavior, and a significant problem for vegetable growers. Essentially, every organic farmer that I interacted with during my research considered harlequin bug to be a serious pest, which offers a great opportunity for doing impactful and relevant research. In closing, I feel that I have contributed to our overall understanding of the biology of this important pest and how to best attract and kill them in a management strategy.

- 89 -